LongBench v2 / 66f3e58c821e116aacb2fabc / Which of the following statements below are false according to the three…

Government policy: meaning, types, manifestations, theories, and policy cycles
Article  in  Insights into Regional Development · June 2023
DOI: 10.9770/IRD.2023.5.2(6)
CITATIONS
6
READS
10,345
4 authors, including:
Adetayo Olaniyi Adeniran
Federal University of Technology, Akure
98 PUBLICATIONS   580 CITATIONS   
SEE PROFILE
Adedayo Adeniran
University of Ibadan
7 PUBLICATIONS   27 CITATIONS   
SEE PROFILE
 
 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195(online) https://jssidoi.org/ird/ 
 
 
 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
         
       
            
 
            Publisher 
    http://jssidoi.org/esc/home 
 
 
 
 
  
 
83 
 
GOVERNMENT POLICY: MEANING, TYPES, MANIFESTATIONS, THEORIES, AND POLICY 
CYCLES 
 
Adetayo Olaniyi Adeniran 1, Joseph Mosunmola Muraina 2, Joseph Olanrewaju Ilugbami 3, 
Adedayo Ayomide Adeniran 4  
 
1Department of Logistics and Transport Technology, Federal University of Technology Akure, Nigeria 
2Department of Geography and Planning Science, Ekiti State University, Ekiti, Nigeria 
3Rufus Giwa Polytechnic-Owo Rector Office, Ondo State, Nigeria 
4Department of Geography and Planning, University of Ibadan, Nigeria 
 
E-mails:adeniranao@futa.edu.ng1; jmosun07@gmail.com2; ilugbamijoseph@gmail.com3; ddone2@gmail.com4 
 
Received 10 March 2023; accepted 10 June 2023; published 30 June 2023 
 
 
Abstract. In any democracy, it is strongly advised that effective policies be created since they are crucial to how democracies operate. 
Government policy definitions and categories were widened. Government policy types were discussed concerning the sectoral groups 
comprising each given government. This is important because a policy’s or its objective elements frequently suggest different meanings for 
different stakeholders. Policymaking is a process impacted by socio-political and other factors and is not a governmental function. Thus, 
there is a need to comprehend the theoretical underpinnings on which government policymaking and its execution may be evaluated and 
characterized. According to the elite/mass hypothesis, there are two groups in society: those who occupy positions of power and those who 
do not. Government policy is more influenced by those with access to knowledge and influence. It is a remarkable characteristic of group 
theory which is ideally in line with the legislative because the legislatures are where the voices of the people are expressed. Governmental 
institutions and government policy are closely related, claims institutional theory. The rational choice theory may need to be more accurate 
since participants in government policy must have access to all information to make informed judgments. The systems theory offers a more 
straightforward method for categorizing and comprehending the contributions and interrelationships made by institutions and policy 
players, including the function played by the external environment in policy formulation. Lastly, since democracy is a system of 
governance built on extensive public engagement, any ideology that supports any type of citizen participation (particularly in a democracy) 
should be endorsed by both politicians and public officeholders. 
 
Keywords: Government policy; Policy manifestations; Policy execution; Policy underpinnings; Policy context and consequences 
 
Reference to this paper should be made as follows: Adeniran, A.O., Muraina, J.M., Ilugbami, J.O., Adeniran, A.A. (2023). Government 
policy: meaning, types, manifestations, theories, and policy cycles. Insights into Regional Development, 5(2), 83-99. 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
JEL Classifications: J58, J68, J78  
 
Additional discipline: Government policy 
 
 
 
 
 
 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
84 
 
1. Introduction 
 
Each democracy needs sound policies. Additionally, in a democracy, the proper application of those policies is 
crucial. According to Delamaza (2015), democracy is a kind of government. Under a democratic political 
administration, among the issues facing governance is erecting a foundation that enhances the practice of 
democracy without undermining the freedom to embark on purpose and functions and to ensure that social 
demands and conflicts arising from various interest groups and civil societies are tackled is one of their tasks 
(Dunne, 2021; Forcher-Mayr and Mahlknecht, 2020). 
 
Given that both government policies and how they are carried out may strengthen a democracy, there is a need to 
define government policy more broadly and the ingredients of government policy execution (GPE) (Adeniran, 
2016; Delamaza and Palma, 2022; Matuku-Mphahlele and Zandamela, 2022). These terminologies are essential 
due to the elements involved in the execution of government policy. Hence, the word government policy 
execution is a subset of the primary term government policy. Government policy may be described as a cycle or 
process with several steps to be taken before achieving a policy’s goals. Typically, there are four or five stages:  
a) Stage for issue/ problem identification; 
b) Stage for setting agenda; 
c) Stage for policy formulation or policymaking; 
d) Stage for policy execution; and  
e) Stage for policy evaluation. 
 
Policy phases will be significantly influenced by the particular technique employed (Zeb-un et al., 2021). 
Government policy is, first and foremost, a persuasive art, as Deygers and Vanbuel (2022) claimed. It is so named 
because it calls for the selection, enactment of legislation, and consultation of all relevant parties (Kofele-Kale, 
2006; Nunes et al., 2019). According to Oyadiran and Akintola (2014), one objective of government policy is to 
guarantee that persons responsible for carrying out significant decisions in society, regardless of their position, are 
well-trained. This opinion was also agreed upon by Myrczik et al. (2022), De-Marchi, Lucertini and Tsoukiàs 
(2014), and Ozturk (2015). 
 
According to Galli (2015), government policy should be viewed as both a declaration of goals and a negotiated 
outcome resulting from the execution process. One of government policy’s most distinguishing features is how 
unstable and changeable it is (Deygers and Vanbuel, 2022). The assertion that proposed or envisioned government 
policies lacks any evident beginning or end is maintained in the study of Ashmore et al. (2020), which noted that 
they should be understood as analogous to seashells or jelly. It flows almost circularly at times. Myrczik et al. 
(2022) assert that when the policy is discussed, it implies addressing pertinent issues germane to human existence. 
 
Falk and Tally (2016) identified the features of government policy, such as the intended direction that the 
legislator would want to guide the public, including the description of how the country’s resources are to be used 
(Díaz-Llamas et al., 2023). It was also revealed by Oyadiran and Akintola (2014) that several variables might 
influence the overall government policy process. These include the legislators in charge, noting what the 
Constitution stands for. The issues that need to be resolved should be known to politicians or bureaucrats Koelble 
and Siddle (2014). 
 
Also, a significant portion of those involved in the government policy process are local (government excluded). 
Consequently, it is crucial to get in touch with these influential individuals who know the community's situation, 
their challenges, and the issues that need to be fixed. The act of fashioning, enacting, monitoring, reviewing, or 
revising government policies is covered by Imenda (2014). Nokele (2022) argues that because it is crucial to the 
efficacy and reach of government policies, its execution should be the primary emphasis of the whole process. 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
85 
 
Macheridis and Paulsson (2019) noted that government policies are centred on presumptions about what 
governments can do and what the effects of those actions would be since they would otherwise be the product of 
political activity and would, as a result, have political ramifications (Mellaard and van Meijl, 2017). Maggetti and 
Gilardi (2016) assert that it is uncommon to get a thorough explanation of the assumptions underlying 
government policy as a theory or model, much alone the context in which those assumptions must be employed or 
understood. But, as with every procedure, an idea or model is always presupposed (McCann and Ward, 2013). 
 
Every story has two sides, and the government policy level confirms this truism. Government policy is two-
dimensional or contains two storylines; given that politics and administration are a component of it, it has a two-
dimensional structure. Creese, Dutton and Esteve-Gonzalez (2021) refer to this reality as the more significant 
number of pertinent legislative and administrative operations. Knill and Tosum’s viewpoints on government 
policy may be contrasted to show how interdependent politics and the administrative side are. Government policy 
and politics should adhere to the same course (Molossi et al., 2023). 
 
The role that legislators play in deciding the resource utilization of a country in the government mentioned above 
policy is regarded as the political side of the government policy process (Gray, 2018). On the other hand, the 
administrative side of the government policy process focuses on the executive and their actions to realize the 
stated objectives established by the government (Mellaard and van Meijl, 2017). The administrative side of the 
public process is responsible for ensuring that the adopted policy will persist throughout time, according to 
Deygers and Vanbuel (2022). Policymakers, administrators, and bureaucrats should encourage all significant 
stakeholders of the necessity of a specific policy and the reasons for that requirement for that execution to take 
place (Purtle et al., 2023; Mellaard and van Meijl, 2017). 
 
2. Literature Review 
 
2.1. Manifestations of Government Policy 
Everyday life is a manifestation of government policy. Also, it starts in casual conversations when regular people 
talk about things like how to improve government policy. As stated in the introduction chapter, creating 
government policies is a complex, multi-layered process (Mellaard and van Meijl, 2017). For the creation and 
execution of government policy, two guiding concepts (or significant areas of study) are essential. Public 
administration and political sciences/studies fall under this category. According to Andrews-Speed (2021), 
government policy encompasses several political science subfields. Implementing government policies, which 
come from the political (or policymaking) facets of government and are backed and endorsed by political 
administrators, is the priority over public administration’s primary goal (Uddin et al., 2023). 
 
Wilson’s dualism (quoted by Guidi et al., 2020) contends that politics and administration cannot be divided into 
distinct roles when determining government policy from both the structural and functional perspectives. There, 
the line thins out to the consistency of a spider’s web thread. According to Simeon (1976), institutions and 
practices that are exposed in and through economic, social, and political dynamics shape government policy. 
Government policy can also result from issue articulation (acknowledging a policy challenge), finding 
alternatives, and the political processes (Crabolu, Font and Eker, 2023).  
 
According to Bertram (2020) the focus of political studies on government policy has been around for a while. 
Mellaard and van Meijl (2017) contend that the academic study of the government policy process is a part of 
political studies/sciences since politics deals with who gets what, when, and how. Political science may be 
necessary to government policy issues while maintaining its dedication to scientific investigation (Mellaard and 
van Meijl, 2017; Deygers and Vanbuel, 2022; Fischer et al., 2015). Politicians, pressure organizations, and 
‘passive beneficiaries of policy’ are only a few stakeholders engaged in the government policy process (Jiang, 
2018). 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
86 
 
 
2.2. Underpinnings of Government Policy 
Government policy is often regarded as being first and primarily a course of action (Makhetha, 2015). This course 
of action must demonstrate logical decision-making, as doing so will lead to responsible behaviour (Daniell, 
2014). Wu (2022) define government policy as the process or series of actions taken by the government to solve a 
particular societal issue that was originally recognized. According to Guidi et al. (2020), the people whose lives 
will eventually be impacted by the outcomes of policy action are represented in the specialized policy subsystems 
where government policy is developed, implemented, and evaluated.  
 
According to Paulsson and Macheridis (2022), who also concurs with Makhetha (2015) and Wu (2022), 
government policies are the result of a combination of systematic forces, political processes, institutional 
influences, rivalry among groups, elite preferences, belief in or advocacy of change through small steps, and 
rational planning (Fischer et al., 2015; Deygers and Vanbuel, 2022). Whatever decisions the government agrees 
on will fall under this (Kharel and Kharel, 2020; Jakonen and Sokka, 2022). Simeon (1976) concluded that 
policies are the climax of a complicated negotiation process and the outcome of several modest judgments made 
by decision-makers. Yet, Simeon (1976) maintains that ideology is at play both in the formulation of policies and 
during the policymaking process, suggesting that government policies indeed reflect ideology (or have a symbolic 
repertoire; Steven, 2021; Molossi et al., 2023). 
 
The socioeconomic circumstances present in a particular geographic area that the government policy must address 
impact how the framework of government policies is developed claims Kharel and Kharel (2020). However, 
several factors, such as institutional frameworks, a country’s party system, or the overall relationship between the 
government and the populace, can affect the process of formulating government policy (Díaz-Llamas et al., 2023). 
Government policy also incorporates a society's dominant ideas, dogmas, and beliefs (Simeon, 1976). Because of 
this, these components offer a framework for the underlying assumptions and arrangements that permit the 
examination of policies (Simeon, 1976). 
 
Recognizing social issues and how societies choose to handle and solve them are essential elements of 
government policy, according to Parsons (2002) and Steinert (2016). Government facilitates reducing or removing 
these issues that society has identified (Parsons, 2002; Crabolu, Font and Eker, 2023). Guidi et al. (2020) assert 
that two features or functions, namely structural terms and/or functional words, can be used to conceptualize 
government policy. The structural component of government policy includes the interactions that may occur 
between the governments' policy players in the setting of the several specialized areas of the subject (Guidi et al., 
2020). The many policy types considerably influence how government policy is framed (Crabolu, Font and Eker, 
2023).  
 
2.3. Types of Government Policy 
Government policies can be created in several styles and/or types to address the need on the policy agenda. Lowi 
refers to this classification of policies as a policy categorization (1972). As a specific policy type would be 
associated with a variety of politics, categorizing policies is essential for studying politics (Oyadiran and 
Akintola, 2014; Aritz et al., 2017). Hence, a politically appropriate policy classification has been developed 
(Oyadiran and Akintola, 2014). The classification of policies must, however, be founded on intellectual and 
theoretical considerations that have an influence on actual political situations (Oyadiran and Akintola, 2014). The 
policy categorization aims to ensure that it supports the study of politics while avoiding omitting the public 
administration component or having a detrimental impact on the political environment as a whole (Oyadiran and 
Akintola, 2014). 
 
The aim of government policy classifications or taxonomies5, according to Bertram, Maleki and  Karsten (2019), 
is to comprehend the basic contrasts between policies and the political settings that influence the various types of 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
87 
 
policies in place. It is simpler to express the typifications that role-players typically utilize to characterize 
government policies when approaches are categorized, according to Aritz et al. (2017). This suggests that using 
policy taxonomies makes it possible to accurately describe government policies (Aritz et al., 2017). Sol (2023) 
claims that employing policy taxonomies may assist in determining the scope and presentation of a policy. 
 
According to Simeon (1976), policy taxonomies offer the chance to consider the amount of coercion and the 
equilibrium between individual and collective activities leisurely. Simeon (1976) thinks Lowi’s (1972) proposed 
policy taxonomies are essential and fundamental for political science students. According to Munzhedzi (2020), 
there are four types of government policies, or policy taxonomies: distributive, redistributive, regulatory, and 
component government policies. Some taxonomies or classifications are considered to be governmental functions. 
According to Nico (2015), these policy categories may be used to pinpoint the specific effects of a policy, which 
might promote political discourse about how decisions are made and how to execute policies. 
 
Also, the sectoral categories or clusters should serve as the foundation for policy classifications (Ahmad et al., 
2021). The terms types and categories of policies were used interchangeably throughout the study. For instance, 
there may be a collection of regulatory or protective policies. 
  
2.3.1. Distributive government policy 
Guidi Guardiancich and Levi-Faur (2020) claim that the primary objective of distributive policies is issue-solving; 
as a result, they typically function in the most supportive political climate. The strong clientele, knowledge, 
leadership, and coherence characterize the context in which distributive policies are carried out (Rakšnys and 
Valickas, 2023). It involves acting to address issues facing the general population (Rakšnys and Valickas, 2023). 
Significantly, distributive policies may also be described as dealing with how additional resources, expenses, and 
advantages from the government are distributed to specific population demography (Díaz-Llamas et al., 2023). 
Bertram, Maleki and Karsten (2019) revealed that distributive strategies address Lasswell’s (1936) maxim of who 
receives what, when, and how. 
 
According to Bertram, Maleki and Karsten (2019), distributive policies use general public funds (instead of user 
fees) to help a particular segment of a social group without considering resource limitations or financial 
constraints (Rakšnys and Valickas, 2023; Díaz-Llamas et al., 2023). As shown in election manifestos, when 
different political parties seek voters to approve of the resources and services they can deliver to them if they are 
elected to power (or held in power), the constituencies of elected politicians also benefit from distributive policies 
(Kraft and Furlong, 2013). 
 
2.3.2. Redistributive government policy 
Allocative government policies, sometimes referred to as redistributive government policies, deal with necessities 
like the funding of the welfare system, health care system, and education system (Ahmad et al., 2021). Guidi 
Guardiancich and Levi-Faur (2020) claim that redistributive policies occur when the government levies taxes on 
one group of people to benefit another. These resources are distributed between the wealthy and the socially 
disadvantageous and destitute groups (Díaz-Llamas et al., 2023). A redistributive strategy can be implemented 
despite ideological cleavages, following Guidi et al. (2020). 
 
Concerning the aforementioned, Guidi Guardiancich and Levi-Faur (2020) assert that direct taxation and the 
transfer of resources from one socioeconomic group to another lead to the emergence of a distinctive 
characteristic that distinguishes distributive and redistributive policies from one another. A dispute is this trait 
(Jutta, 2016). Redistributive policies are exceedingly political, difficult, unfavourable, and polarizing to design 
and implement, which causes this conflict. They cause disputes that polarize the population along party lines 
(Rakšnys and Valickas, 2023). Redistributive programs face this challenge since one group gains at the expense 
of another (Yanow, 2015). The discussion around distributive government policies is heightened because they 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
88 
 
explicitly allude to an ideology or a class war. According to Donnelly (2015), the disadvantage of redistributive 
policy is that the government typically lacks the means to implement such a program. 
 
2.3.3. Regulatory government policy 
Regulatory policies, according to Munzhedzi (2020), typically address the need for policies relating to 
transportation, infrastructure, health, and other regulations and standards, or they prohibit people from acting in 
certain ways, such as selling illegal goods like dangerous drugs, participating in unfair competition in the market 
(Rakšnys and Valickas, 2023). According to Anyebe (2018), regulatory policies are laws carried out by 
government agencies without any interference or money inducement. 
 
A regulation policy can be a form of competitive regulation to regulate individual industries and their activities. It 
can also be a protective regulation meant to protect the general public. Bertram, Maleki and Karsten (2019) 
contend that regulatory approaches are questionable because they let the government meddle in private enterprises 
and people’s daily lives. Another disadvantage of regulatory government policies, according to Creese, Dutton 
and Esteve-Gonzalez (2021), is that they significantly impact how much money is spent and how much assistance 
from other social actors is needed. 
 
2.4. Constitution government policy 
Oyadiran and Akintola (2014) developed the component policy as a subset of constituent policy. Both the 
government and/or the nation as a whole are considered to be two constituents of government policy, according to 
Oyadiran and Akintola (2014) and Guidi Guardiancich and Levi-Faur (2020). According to Meier’s additional 
definition from 2007, constituent policies aim to advance the interests of the nation-state and the broader public. 
Constituency policies are portrayed by Guidi Guardiancich and Levi-Faur (2020) as being exceedingly detailed, 
meticulous, and in charge of significant initiatives. Meier’s (2000) notion of component policies may be used to 
depict the presidential department where policies are executed, monitored, and coordinated. Constituent policies 
also cover governmental operations, including defence and foreign policy (Rakšnys and Valickas, 2023). 
 
The present democratic society can be classified under constituent policies because of their method of operation 
and provision for election laws (Yanow, 2015). According to Creese, Dutton and Esteve-Gonzalez (2021), there is 
a fact that constituent government policies only have an impact on the executive branch of government. As was 
said above, Oyadiran and Akintola (2014) identified the many kinds of government policies and found just four 
policy taxonomies. Not all government policies will fall within Lowi’s (1972) taxonomy of approaches, as 
(Rakšnys and Valickas, 2023) indicates. These policy taxonomies thus have the disadvantage of excluding 
alternative policies that might not fit the policy classification. A few new categories of approaches have been 
included in the classification of procedures since Lowi’s (1972) policy taxonomies were first introduced. 
 
The following section briefly discusses one more policy type that is mainly referred to as substantive government 
policy. 
 
2.5. Substantive government policy 
Government policies are crucial in a wide range of substantive sectors, according to Paulsson and Macheridis 
(2022). These substantive sectors include, but are not limited to, environmental issues, economic development, 
security, public service, international relations, primary education, social development and domestic affairs 
(Fischer et al., 2015). A substantive policy focuses on what the government should do (Simeon, 1976). A 
substantive policy may incorporate specific overarching goals (such as describing the anticipated results of the 
policy while it is being produced, for example) (Marie-Kim and Marie-Hélène, 2020). It might also consist of 
more concrete objectives the policy must achieve. Yudiatmaja et al. (2022) conclude that successful substantive 
solutions can resolve a policy issue. 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
89 
 
2.6. Government policy Execution 
The policy as it is carried out is the genuine policy of a government, according to Peters (2001). One of the main 
reasons why government policy execution is one of the most crucial stages in the whole government 
policymaking process is because it refers to the point at which a policy is implemented (Díaz-Llamas et al., 2023; 
Aguerre and Hernan, 2015). According to Peters (2001), one of the main issues with our current political systems 
is how government policies are carried out. Many behaviours in the administrative and political settings where 
government policy is being implemented are taken into account throughout the execution process, claims Hurel 
and Rocha (2018). According to Galli (2015), politics substantially influences every action or step taken during 
the cycle of policy execution, with both a macro and micro political context (Galli, 2015). 
 
The macro-political backdrop, which includes factors like legislation, economy, and what is happening or moving 
worldwide, is what Galli (2015) refers to as the external environment. On the other hand, according to Galli 
(2015), the micro-political context comprises things like the policy’s mission, the competencies needed, the 
organizational culture, and the external environment. GPE is rather challenging since several factors must be 
considered, some of which the implementers have influence over and others of which they do not. Falk and Tally 
(2016) argue that it is incorrect to assume that implementing government policy only entails putting previously 
developed procedures into action since there is more to it than that. Implementing government policy involves 
using important inherent information. 
 
Creese, Dutton and Esteve-Gonzalez (2021) assert that GPE bridges policymakers and policy addresses. The 
implementers aid this relationship. This GPE phase is essential because it makes it possible to execute the 
proposed or envisioned policy (Steven, 2021). This suggests that the result of the policy is transformed into its 
production. Aguerre and Hernan (2015) contend that policies and practices must be separated to understand the 
whole process of producing policies. As mentioned earlier, the role-players in charge of implementing the policies 
must thus not act entirely independently. As a result, they offer guidelines for applying already created and 
authorized policies. 
 
3. 
Theoretical Review 
 
Government policy theories are essential in the social, environmental, technological and engineering literature. 
These theories offer unique characteristics of political and human development. Among the theories of 
government policy are the political systems theory, group theory, institutional theory, rational choice theory, and 
the policy process model. 
 
Many of the previous and present policies are formulated and implemented because they are influenced by 
systemic variables, political processes, institutional influences, game-playing, incrementalism, interest group, 
rational planning, elite preferences, and interest group interests. These theories will pose further issues regarding 
government policy and the primary channels from which sound decisions are formed. The following section will 
cover these theories. 
 
3.1. The elite/mass theory 
The elite notion holds that a small elite group controls the bulk (Zeb-un et al., 2021). This idea works best in the 
countries of Africa. Because the interests and well-being of the elite are prioritized under this theory, elite 
viewpoints that diverge from those of the general public can affect the development of government policy (Zeb-un 
et al., 2021; Jutta, 2016). The elite thesis is based on the notion that because the general public is allegedly 
uninformed and indifferent, their opinions shouldn’t have any bearing on how government policy is formulated 
(Fischer et al., 2015). 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
90 
 
The elite notion holds that only a caste that is acknowledged throughout society should influence government 
policy (Kraft and Furlong, 2013). This elite caste includes members of the governing class, political parties, 
business executives, wealthy individuals, and educated segments of society (Jutta, 2016). One way that the elite 
ideology is implemented is by whom the most influence over how government policy is decided (Jutta, 2016). Not 
all elites have an outsized impact on shaping government policy. Each elite aims to have a significant effect on a 
specific niche market. For instance, business executives would want to weigh in on decisions regarding tax 
legislation and import and export laws. The governing class, however, would like to have a voice in how the 
general public is governed, how money is allocated, and how resources are utilized.  
 
It’s also conceivable that these two exclusive groups come into contact with one another and interact as they use 
their influence and power. Zeb-un et al. (2021) assert that public administrators’ perceived importance is 
influenced by the idea that they are members of the ruling class rather than citizens’ servants. This idea may be 
explained as a small elite making decisions that cascade down to an uneducated civil society (Fischer et al., 2015). 
Zeb-un et al. (2021) assert that political power influences these decisions and that the bureaucracy is necessary to 
carry them out. The idea holds that only a select group of experts possess the authority. 
 
3.2. Group theory  
Politics is characterized by the interaction of groups, and the group theory incorporates organized interest groups 
in the creation of government policy (Jutta, 2016). These actors are shown as tenacious voice-hearing warriors. In 
light of this, it is possible, to sum up group theory as a battle between the voices of organized interest groups. 
Group theory includes, for example, individuals working in the agriculture sector and companies producing 
music. Organizations should be allowed to make a major contribution and have a say in determining government 
policy. In order to dispute the abuse, poor administration, and fraudulent execution of policies, as well as hold 
those responsible accountable, people should be able to challenge laws that are thought to be illogical, unsuited, 
or ineffective for the intended purpose. 
 
Organizations ought to promote justice, transparency, the participation of the citizenry, and awareness in 
policymaking. The group theory is crucial and pertinent to government policy as a result. This is done so that 
organizations may play a big part in setting policy and assisting with enforcing previously approved or ratified 
legislation like the Constitution. Groups have an impact on government policy, whether it be a policy regarding 
environmental concerns or the welfare of the populous as a whole. This exemplifies how several interest groups 
from diverse socioeconomic domains may all voice their opinions on the policies they believe the government 
should adopt or reject and play a significant role in their creation. According to Zeb-un et al. (2021), group theory 
has some implications for political decisions. For instance, the dynamics of the cabinet are changed. 
 
The disadvantage of the group theory is that it rewards more organized groups, has more members, has access to 
resources, has political allegiances, is well-liked, and has built ties with decision-makers (Galli, 2015). The less 
fortunate members of society lack all of the aforementioned resources and are at the other extreme of the spectrum 
(Jutta, 2016). According to Bertram, Maleki and Karsten (2019), group theory is criticized by academics studying 
government policy for giving organized interest groups too much sway and leaving it up to them to decide policy 
(Oyadiran and Akintola, 2014). Government employees also seem to be left on the side of the road (Tacon and 
Hanson, 2011). 
 
Researchers in government policy believe that the degree of impact that organized interest groups have on 
policymaking tends to worsen the complexity and dynamic character of policymaking that is already present 
(Kraft and Furlong, 2013). It is also important to acknowledge that the elite/mass does have some roots in group 
theory (Carroll and Common, 2013). Although all groups (regardless of socioeconomic level or prominence) may 
be accommodated under the group theory, those with access to more resources are often the ones whose opinions 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
91 
 
are heard when policies are being developed (Chaudhary, 2018). Their voices tend to be aristocratic. The voices 
of those groups that lack access to the same resources as the privileged are so muffled. 
 
The term "extra influence" refers to the elite groups’ intrusion into group theory and the creation of government 
policy. This is characterized by Guidi Guardiancich and Levi-Faur (2020) as having a solid clientele, knowledge, 
and leadership. Also, this increases the pressure on public servants and policymakers, which tips the balances in 
their favour when deciding the course of government policy. 
 
3.3. Institutional-based theory  
This theory is often known as the "classical theory" since it is interpreted classically to study government policy 
(Zeb-un et al., 2021). It is not a coincidence that Minkman, van Buuren and Bekkers (2018) state that the 
institutional approach arose as awareness of the importance of enshrining government policy-making in the 
framework of institutions expanded. This implies that the government’s concerns about welfare should take 
precedence over other issues (Zeb-un et al., 2021). Institutional theory is deeply rooted in the formal and legal 
aspects of the governmental system (Díaz-Llamas et al., 2023). The institutional model’s purpose is to evaluate 
the structures that regulate how the government is structured, its legal power and the norms of behaviour it 
adheres to while making decisions (Dunne et al., 2021). The institutional theory focuses primarily on the public’s 
access to decision-making, government transparency, and, eventually, the separation of powers between the 
various levels of government (Zeb-un et al., 2021).  
 
The institutional theory rationally asserts that the structures and codes of conduct that regulate the government 
and its departments significantly impact the various types of policy processes that take place, as well as how role-
players in those processes will ultimately affect those processes (Kraft and Furlong, 2013). Political, economic, 
and sociological institutionalism are the three frameworks that institutional theory embraces (Minkman, van 
Buuren and Bekkers, 2018). Economic institutionalism stresses applying economic analysis to political 
institutions and government policy, whereas political institutionalism looks beyond the traditional forms of 
institutions to pay more attention to (Díaz-Llamas et al., 2023). The institutional theory is essential in ensuring 
government policies' legitimacy, applicability, and coerciveness (especially true of regulatory laws, which impose 
obligations on the general populace) (Díaz-Llamas et al., 2023). 
 
3.4. Rational choice theory 
This is a contemporary theory used in social sciences. The public choice theory is another name for the theory of 
rational choice (Cagnin, 2017). It has a strong economic foundation (Jutta, 2016). Generally, it uses complex 
mathematical modelling, which has only been moderately helpful in evaluating marginal behaviours in 
competitive circumstances and is typically seen throughout an election period (Ashmore et al., 2020). 
 
This theory is thoroughly developed and rigorous, and it could be used to address many government policy-
related issues (Kraft and Furlong, 2013) and used to conclude. Opponents of this theory claim that the decisions 
made based on rational choice are faulty, unrealistic and unworkable. 
 
Cagnin (2017) identified two distinct features of the rational choice theory. Its main focuses are methodological 
individualism and the assumption that people are reasonable. The sensible perspective contends that the ability to 
make decisions indicates a person’s capability for logical reasoning. On the other hand, Ashmore et al. (2020) 
argue that a broad account of human behaviour supports all rational choice theories. According to Ashmore et al. 
(2020), the basic hypothesis holds that individuals are complicated, flawed mortals who strive for perfection 
despite whatever challenges they may encounter. 
 
 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
92 
 
3.5. Political systems theory 
This theory is the most complete among popular approaches (Kraft and Furlong, 2013). The theory aids 
government initiatives and institutions in transforming public inputs (such as environmental needs) into policy 
outputs (such as public opinion and pressure from interest groups) (Cagnin, 2017). The theory was designed to 
raise public awareness of policy issues and give the populace a platform to express grievances (Cagnin, 2017), 
allowing for the problems to be presented on the government’s policy agenda (Uminska-Woroniecka, 2022). 
Moreover, it represents the wider, shared socioeconomic, cultural, and political framework that serves as the 
foundation for decisions on politics and policy (Jutta, 2016). According to Bertram, Maleki and Karsten (2019), 
the language employed in political and policy studies has expanded as a result of the systems theory. 
 
3.5.1. Government policy Cycle  
In accordance with the four government policy functions, the policy process model (Jutta, 2016) recommends an 
analytical progression of the occasions that impact the formulation of government policies (Guidi et al., 2020). At 
each level of the policy process model, the connections between policy players are shown (Jutta, 2016). 
According to Appiah-Agyekum (2020), the policy model explains how decisions were made, makes 
understanding the timeline of events simpler, and supports the pragmatic nature of government policy (Guidi et 
al., 2020). 
 
Moreover, it explains how these results in the understanding that can be applied to any political system and its 
decision-making procedures (Jutta, 2016). The best method to begin a discussion of policy theories and a strategy 
to organize the study of policymaking, according to Cagnin (2017), is to use the policy cycle. According to 
Bertram, Maleki and Karsten (2019), the traditional model is cyclical since formulating policies is continuous and 
always in “motion" as a rolling wheel. 
 
The policy cycle's main lesson is that just because an issue has been identified and a decision has been taken, it 
doesn’t mean everything has been fixed (Cagnin, 2017). That only denotes the beginning of the policymaking 
process. The model’s stages are linked to each other like links in a chain cycle (Appiah-Agyekum et al., 2022). As 
Bertram, Maleki and Karsten (2019) noted, no policy decision or solution is ever final. The policy process model 
does succeed in capturing the essence of policymaking despite all of its flaws, and as a consequence, it correlates 
to political reality. 
 
According to Lerma, Díaz-Baca and Burkart (2022), the conventional model of the policy process consists of four 
functional processes or phases:  
i. 
Agenda setting; 
ii. 
Policy development;  
iii. 
Policy execution; and  
iv. 
Policy assessment 
Two additional steps that Bertram, Maleki and Karsten (2019) add to the concept of the policy process are:  
i. 
Policy legitimization; and  
ii. 
Policy modification.  
Guidi et al. (2020) postulate the results of policies and the related subsystems that must be implemented. This 
suggests that the stages theory serves as an example of how a government policy develops (or comes into 
existence). Guidi et al. (2020) assert that knowledge and information are the main forces behind policy 
construction. 
 
In a significant sense, this is the reason why everyone involved in the policy process has to be sufficiently 
informed of how government policy is produced, as well as possess the knowledge, skills, and competence 
necessary to see the process through to the end (Rakšnys and Valickas, 2023). Moreover, it serves as a tool for 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
93 
 
guiding and educating decision-makers on the procedures involved in carrying out government policy (Cairney 
2012). 
 
3.5.2. Agenda setting 
Any social issue that the public brings up should be taken seriously (Aguerre and Hernan, 2015). But more 
importantly, agenda-setting in democracies is expected to be characterized by a high level of citizenry 
participation (Blackstock et al., 2020). Various media influences can manage, shape, and define the issues on the 
policy agenda (Fischer et al., 2015). 
The topics included on the policy agenda can be influenced, controlled, shaped, and defined using these platforms 
or the participation of experts from a particular subject (Crabolu, Font and Eker, 2023). These are the three steps 
that makeup agenda setting: 
a) Identification of issues; 
b) verifying which problem is of significant essence; and  
c) Outlining the dynamics of an issue (Cagnin, 2017).  
According to Díaz-Llamas et al. (2023) and Falk and Tally (2016), only one element determines whether 
policymakers should pay attention at this early stage of the policy process. That aspect is the availability of 
information about social issues/issues. According to Steinert (2016), the public media’s assessment and awareness 
of a societal issue has an effect on agenda shaping. This is due to the possibility that the press might impact public 
opinion, given the variety of media outlets available (Chetty, 2015). 
 
3.5.3. Policy formulation 
Before creating a policy, one must create a strategy for responding to the suggestions made during the first phase 
of the policymaking process. According to Cagnin (2017), creating policies entails defining goals, estimating 
costs, and assessing the specific outcomes this policy will produce. As a result, the suggested course of action and 
the policymaker's (s) ' intentions are both stated at this point in the policy cycle (Steven, 2021). Rational, logical 
solutions are chosen. Following the conclusion of this process, crucial policy instruments are selected (Cagnin, 
2017). Falk and Tally (2016) assert that all necessary stakeholders, such as interest organizations, elected 
officials, legislators, and the public, should participate in policy development. 
 
3.5.4. Policy execution 
Only the events in the early stages of the policy process result in government policy. At this stage, it may be 
anticipated that a government policy will undergo changes, such as revision; the government policy may even be 
rejected at this stage (Kustec and Mcardle, 2012). A significant feature of government policy execution is that it 
may take on many shapes and forms depending on the institutional and cultural context (Welsh, 2019). Attention 
was called to an essential facet of carrying out government policy, especially given that it operates or is carried 
out at a time when "government" procedures are seen as having been transformed into "governance". 
 
Moreover, Jaishia et al. (2023) classify government policy execution research as a political science and 
administration subject. This suggests that overly-simplistic hierarchical models are being abandoned and that a 
broad spectrum of stakeholders is starting to participate in policymaking (Iroulo and Boateng, 2023). Also, 
politics ends when administration begins. Politics and administration are related. According to Mügge and 
Alenda-Demoutiez (2019), the institutions of democracy and the rule of law have entrenched a tight hierarchy in 
the relationship between these two disciplines. 
 
3.5.5. Policy Outcomes and Evaluation 
At this stage, a policy is evaluated to determine its success (or failure) (Cagnin, 2017). The effective execution of 
the procedure, the judgments taken about the policy, and whether the policy generated the intended results as 
described in the stage of defining the agenda and formulating the policy are all crucial factors to take into account 
when evaluating policies (Cagnin, 2017; Steven, 2021). Lessons will be drawn from this, recognized, and 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
94 
 
typically serve as the basis for future policy choices (Appiah-Agyekum et al., 2022; Cagnin, 2017). This is done 
by carefully reviewing all of the information gleaned from evaluating the policy’s outcomes (Mügge and Alenda-
Demoutiez, 2019). 
 
Many government policy actors, such as think tanks, government organizations, external consultants, nonprofit 
groups, the media, and the general public, can engage in this activity (Fischer et al., 2015). When this process is 
finished, the policy can be sent back to the legislator, who can then choose whether to change it (possibly 
signalling the start of a new policy cycle) (Steven, 2021). 
 
The policy cycle is advantageous. Welsh (2019) identified the main reasons for this:  
a) Since it is a logical process that may depict the variety of reality, it is plain and easy to grasp.  
b) Each phase disseminates knowledge to a particular section of the setting in which government policy is 
produced (Welsh, 2019). This could aid the policymaker in selecting the many variables and tactics 
available. 
c) The policy cycle shows that policymaking is flexible.  
d) The process of establishing policies follows a chronological order. 
The policy cycle also identifies the point at which the policymaking process should start, which makes it a helpful 
tool for the decision-maker. 
 
Conclusions 
 
Definitions and classifications of government policy were expanded in this study. It is essential to keep in mind 
different types of policies are defined in different ways but must be simple to comprehend. As the objective of a 
policy is more likely to shape society, the distributive or substantive policy may be seen by one group of 
participants as a regulatory policy. Still, another group may not see it as such. Therefore, policy classifications aid 
in outlining the various ways that policy stakeholders frequently describe policies and the development of 
practicability and reality of the policy that will be implemented. 
 
The elite/mass idea holds that society is divided into two groups: those in positions of authority and those who do 
not. Those with access to and influence take a more active role in creating government policy, which is in line 
with the elite/mass theory. The exciting aspect of group theory is that it aligns more with the legislative branch of 
government than the bureaucracy. This could be because the legislature is where the general population's opinions 
are represented. 
 
For the institutional theory, it was revealed that institutions and policy are closely related. The institutional 
theory's foundation is procedural legislation and how it could help or impede political goals in various 
governmental structure sectors. Although the rational choice theory assumes that government policy actors have 
access to all the information necessary to make well-informed decisions, this theory can be misleading and 
unrealistic because it believes that government policy actors will have the knowledge and ability to make rational 
decisions. 
 
To categorize and understand the contributions and linkages made by institutions and policy players and the role 
played by the external environment in producing policy, however, the systems theory offers a more 
understandable method. It was contended that as democracy is an administrative system built on broad public 
involvement, politicians and people in public office should support any concept that fosters citizen engagement in 
any form (especially in democracies). 
 
The participation of the citizenry at all proper steps of the policy cycle is only fair because government policy is 
created with the general public in mind; nonetheless, caution against dismissing any of these models or theories. 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
95 
 
They provide different viewpoints on politics and government policy and information on how these two function 
in the institutional and political domains. They directly give rise to theories of politics and government policy, 
which provide light on how issues are discussed during policymaking. 
 
It is essential to remember that government policy and how it is carried out are about ‘outcomes’ for the policies 
being implemented. It entails gathering all the inputs (needs) from the community and rating each demand 
according to priority. These inputs from the community or other role-players decide the issues listed on the 
policy’s agenda. Second, the bureaucracy must recognize the outside world since external variables, except for the 
community, significantly impact government policy. Laws, the environment on a global scale, technology, 
traditional views, politics, diversity, and complexity are some of these external elements. 
 
 
 
References 
 
Adeniran, A. O. (2016). The rational approach of Obafemi Awolowo’s Principles in enhancing efficient leadership and good governance 
through intellectual enormity. Journal of Law, Policy and Globalization, 49, 101-107. 
 
Aguerre, C., and Hernan, G. H. (2015). Internet policy formation in Latin America: understanding the links between the national, the 
regional, and the global. SSRN Electronic Journal. http://doi.org/10.2139/ssrn.2809883  
 
Ahmad, F., Draz, M. U., Chang, W. Y., Yang, S. C., and Su, L. (2021). More than there source curse: exploring the nexus of natural 
resource 
abundance 
and 
environmental 
quality 
in 
north-western 
China. 
Resour 
Policy, 
70, 
101902. 
https://doi.org/10.1016/j.resourpol.2020.101902  
 
Andrews-Speed, P. (2021). How may National culture shape public policy? The case of energy policy in China. The Energy Journal 43(3). 
http://doi.org/10.5547/01956574.43.3.pand    
 
Anyebe, P. A. (2018). An overview of approaches to the study of public policy. International Journal of Political Science (IJPS), 08-17. 
 
Appiah-Agyekum, N. N. (2020). Primary healthcare implementation in practice: evidence from primary healthcare managers in Ghana. 
African Journal of Primary Health Care and Family Medicine, 12, 1-7. https://doi.org/10.4102/phcfm.v12i1.2183  
 
Appiah-Agyekum, N. N., Kayi, E. A., Appiah-Agyekum, J., and Otoo, D. D. (2022). Capacity issues in primary health care 
implementation: examples from Ghana. Health Education, 122(1). http://doi.org/10.1108/HE-06-2021-0095  
 
Aritz, J., Walker, R., Cardon, P., and Li, Z. (2017). Discourse of leadership: The power of questions in organizational decision making. 
International Journal of Business Communication, 54(2), 161-181. http://doi.org/10.1177/2329488416687054  
 
Ashmore D. P., Thoreau, R., Kwami, C., Christie, N., and Tyler, N. A. (2020). Using thematic analysis to explore symbolism in transport 
choice across national cultures. Transportation, 47, 607-640. https://doi.org/10.1007/s11116-018-9902-71  
 
Bertram, D. (2020). Accounting for culture in policy transfer: A blueprint for research and practice. Political Studies Review, 1-18. 
http://doi.org/10.1177/1478929920965352  
 
Bertram, D., Maleki, S., and Karsten, N. (2019). Factoring in societal culture in policy transfer design: the proliferation of private 
sponsorship of refugees. Journal of International Migration and Integration, 21(5). http://doi.org/10.1007/s12134-019-00738-0  
 
Blackstock, K., Novo, P., Byg, A., and Waylen, K. A. (2020). Policy instruments for environmental public goods: Interdependencies and 
hybridity. Land Use Policy, 107(3), 104709. http://doi.org/10.1016/j.landusepol.2020.104709  
  
Cagnin, I. F. (2017). The study of macro and micro implementation of social policy, de Paul Bierman. Revista Brasileira de Políticas 
Públicas e Internacionais - RPPI, 2(2), 175-180. http://doi.org/10.22478/ufpb.2525-5584.2017v2n2.37386  
 
Carroll, P., and Common, R. (2013). Policy Transfer and Learning in Public Policy and Management, London: Routledge 
https://doi.org/10.4324/9780203749395  
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
96 
 
Chaudhary, D. (2018). Agriculture policy and rural development in Nepal. Research Nepal Journal of Development Studies, 1(2), 34-46. 
https://doi.org/10.3126/rnjds.v1i2.22425  
 
Chetty, R. (2015). Behavioral economics and public policy. American Economic Review: Paper and Proceedings, 105(5), 1-33. 
https://doi.org/10.1257/aer.p20151108  
 
Crabolu, G., Font, X., and Eker, S. (2023). Evaluating policy complexity with Causal Loop Diagrams. Annals of Tourism Research, 
100(11), 103572.    http://doi.org/10.1016/j.annals.2023.103572  
 
Creese, S., Dutton, W. H., and Esteve-Gonzalez, P. (2021). The social and cultural shaping of cybersecurity capacity building: a 
comparative study of nations and regions. Personal and Ubiquitous Computing 25(2). http://doi.org/10.1007/s00779-021-01569-6  
 
Daniell, K. (2014). The role of national culture in shaping public policy: a review of the literature. HC Coombs Policy Forum Crawford 
School of Public Policy. ANU College of Asia & the Pacific 
 
Delamaza, G. (2015). Enhancing democracy: Public policies and citizen participation in Chile. New York: Berghahn. 
 
Delamaza, G., and Palma, J. F. (2022). From above or from below? Chilean NGOs, the State and education reforms. Journal of 
Educational Administration and History. http://doi.org/10.1080/00220620.2022.2137480  
 
Deygers, B., and Vanbuel, M. (2022). Advocating an empirically-founded university admission policy. Language Policy, 21(3), 1-22. 
http://doi.org/10.1007/s10993-022-09615-6  
 
Díaz-Llamas, J. L., Soto-Ceja, E., Bravo-Olivas, M. L., and Chávez-Dagostino, R. M. (2023). Ecological footprint of an artisanal fishing 
cooperativein a lagoon body. Revista Bio Ciencias, 10 e1399. http://doi.org/10.15741/revbio.10.e1399  
 
Dunne, L., Craig, N., Miller, S., Connolly, P., Lackman, J., Aber, L., Yoshikawa, H., Fitzpatrick, S., Pham, P., Vinck, P., and Walmsley, P. 
(2021). Promoting Social Cohesion and Peacebuilding Through Investment in Early Childhood Development Programs. In book: 
Transitioning to Peace. http://doi.org/10.1007/978-3-030-77688-6_17  
 
Falk, M., and Tally, K. (2016). Cultural participation in Europe: Can we identify common determinants? Journal of Cultural Economics, 
40, 127-162. https://doi.org/10.1007/s10824-015-9242-9  
 
Fischer, F., Torgerson, D., Durnová, A., and Orsini, M. (2015). Introduction to critical policy analysis.” In Handbook of Critical Policy 
Studies, edited by F. Fischer, D. Torgerson, A. Durnová, and M. Orsini. Cheltenhamn, UK: Edward Elgar Publishing. 1-26. 
https://doi.org/10.4337/9781783472352.00005  
 
Forcher-Mayr, M., and Mahlknecht, S. (2020). Critical entrepreneurship education in general education and TVET: two concepts of 
practice in a South African township. Discourse and Communication for Sustainable Education, 11(2), 65-84. http://doi.org/10.2478/dcse-
2020-0019  
 
Galli, A. (2015). On the rationale and policy usefulness of ecological footprint accounting: The case of Morocco. Environmental Science & 
Policy, 48, 210-224. http://doi.org/10.1016/j.envsci.2015.01.008  
 
Gray, S. (2018). Achieving compliance with the World Anti-Doping Code: learning from the implementation of another international 
agreement. International Journal of Sport Policy, 11(4), 1-14. http://doi.org/10.1080/19406940.2018.1550798  
 
Guidi, M., Guardiancich, I., and Levi-Faur, D. (2020). Modes of regulatory governance: A political economy perspective. Governance, 
33(1), 5-19. http://doi.org/10.1111/gove.12479  
 
Hurel, L. M., and Rocha, M. S. (2018). Brazil, China and Internet Governance. Mapping Divergence and Convergence, 22, 67. 
http://doi.org/10.5278/ojs.jcir.v0i0.2267  
 
Imenda, S. (2014). Is there a conceptual difference between conceptual and theoretical frameworks? Journal of Social Science, 38(2), 185-
195. https://doi.org/10.1080/09718923.2014.11893249  
 
Iroulo, L. C., and Boateng, O. A. (2023). Bureaucratic acquiescence as an institutional strategy in the African Union. African and Asian 
Studies, 22(1-2), 113-134. http://doi.org/10.1163/15692108-12341583  
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
97 
 
Jaishia, M., Sharmab, G. P., Nepalic, P. B., Gauchana, D., Shresthad, R. K., Timsinae, K. P., and Neupanea, H. (2023). Government 
framework for agriculture service delivery at the local level in Nepal. Nepal Public Policy Review, 3(1). https://doi.org/10.59552/nppr.v3 
i1.28  
 
Jakonen, O., and Sokka, S. (2022). Finnish cultural policy as public funding: Regime view across policy domains. Nordisk Kulturpolitisk 
Tidsskrift 25(3), 293-313. http://doi.org/10.18261/nkt.25.3.9  
 
Jiang, Y. H. (2018). Confucian political theory in contemporary China. Annual Review of Political Science, 21, 155-173. 
https://doi.org/10.1146/annurev-polisci-041916-020230  
 
Jutta, V. (2016). Participatory turn in cultural policy? An analysis of the concept of cultural participation in Finnish cultural policy. Nordic 
Journal of Cultural Policy, 19(1), 59-77. https://doi.org/10.18261/ISSN2000-8325-2016-01-04  
 
Kharel, K. R., and Kharel, S. (2020). Analyze the effectiveness of resource mobilization of local government. Management Dynamics, 
23(1), 169-182. https://doi.org/10.3126/md.v23i1.35577    
 
Koelble, T. A., and Siddle, A. (2014). Institutional complexity and unanticipated consequences: The failure of decentralization in South 
Africa. Democratization, 21(6), 1117-1133. http://doi.org/10.1080/13510347.2013.784270  
 
Kofele-Kale (2006). The international law of responsibility for economic crimes: holding state officials individually liable for acts of 
fraudulent enrichment. Hampshire, England: Ashgate Publishing Ltd. 
 
Kraft, M. E., and Furlong, S. R. (2013). Politics, analysis and alternatives. Los Angeles: SAGE. Landsberg, C. n.d. South Africa’s African 
agenda: challenges of policy and implementation. 
 
Kustec. S., and Mcardle, D. (2012). National law, domestic governance and global policy: A case study of anti-doping policy in Slovenia. 
International Journal of Sport Policy, 6(1), 1-17. http://doi.org/10.1080/19406940.2012.662692  
 
Lerma, L M., Díaz-Baca, M. F., and Burkart, S. (2022). Public policies for the development of a sustainable cattle sector in Colombia, 
Argentina, 
and 
Costa 
Rica: 
a 
comparative 
analysis 
(2010-2020). 
Front. 
Sustain. 
Food 
Syst., 
6, 
722522. 
https://doi.org/10.3389/fsufs.2022.722522  
 
Macheridis, N., and Paulsson, A. (2019). Professionalism between profession and governance: How University teachers’ professionalism 
shapes coordination. Studies in Higher Education, 44(3), 470-485. http://doi.org/10.1080/03075079.2017.1378633  
 
Maggetti, M., and Gilardi, F. (2016). Problems and solutions in the measurement of policy diffusion mechanisms. Journal of Public Policy, 
36 1), 87-107. https://doi.org/10.1017/S0143814X1400035X  
 
Marie-Kim, C., and Marie-Hélène, G. (2020). Changes in practitioners’ attitudes, perceived training needs and self-efficacy over the 
implementation process of an evidence-based parenting program. BMC Health Services Research, 20(1). http://doi.org/10.1186/s12913-
020-05939-3  
 
Matuku Mphahlele, M., and Zandamela, H. L. (2022). Capacity Development-oriented Service Delivery Structures and Programmes: Case 
Study 
of 
a 
Mineral-rich 
District 
Municipality. 
Journal 
of 
Public 
Administration 
and 
Governance, 
12(1), 
78. 
http://doi.org/10.5296/jpag.v12i1.19363  
 
McCann, E., and Ward, K. (2013). A multi-disciplinary approach to policy transfer research: geographies, assemblages, mobilities and 
mutations. Policy Studies, 34 (1), 2-18. https://doi.org/10.1080/01442872.2012.748563  
 
Mellaard, A., and va/n Meijl, T. (2017). Doing policy: enacting a policy assemblage about domestic violence. Critical Policy Studies, 
11(3), 330–348. http://doi.org/0.1080/19460171.2016.1194766  
 
Minkman, E., van Buuren, A., and Bekkers, V. J. J. M. (2018). Policy transfer routes: An evidence-based conceptual model to explain 
policy adoption. Policy Studies, 3(2), 222-250. https://doi.org/10.1080/01442872.2018.1451503  
 
Molossi, L., Hoshide, A. K., de Abreu, D. C., and de Oliveira, R.A. (2023). Agricultural support and public policies improving 
sustainability in Brazil's beef industry. Sustainability, 15(6), 4801. http://doi.org/10.3390/su15064801  
 
Mügge, D., and Alenda-Demoutiez, J. (2019). The lure of ill-fitting unemployment statistics: How South Africa’s discouraged work 
seekers disappeared from the unemployment rate. New Political Economy, 25(104), 1-17. http://doi.org/10.1080/13563467.2019.1613355  


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
98 
 
 
Munzhedzi, P. H. (2020). Evaluating the Efficacy of municipal policy implementation in South Africa: challenges and prospects. African 
Journal of Governanceand Development, 1(9), 89-105. 
 
Myrczik, E., Heikkilä, R., Kristensen, N. N., and Purhonen, S. (2022). Missing out on culture-or not: Danes and Finns’ cultural 
participation, the pandemic, and cultural policy measures. Nordisk Kulturpolitisk Tidsskrift, 25(2), 1-24. http://doi.org/10.18261/nkt.25.2.4  
 
Nico, 
C. 
(2015). 
Differentiating 
between 
access, 
interaction 
and 
participation. 
Conjunctions, 
2(2), 
7-28. 
https://doi.org/10.7146/tjcp.v2i2.23117  
 
Nokele, K. (2022). Context and Realities of Policy Implementation in South Africa: Quo Vadis. American Journal of Multidisciplinary 
Research and Innovation (AJMRI), 1(6), 10.54536/ajmri.v1i6.898 
  
Nunes, S., Barlow, J., Gardner, T., Sales, M., Monteiro, D., and Souza Jr. C. (2019). Uncertainties in assessing the extent and legal 
compliancestatus 
of 
riparian 
forests 
in 
the 
eastern 
Brazilian 
Amazon. 
Land 
Use 
Policy, 
82, 
37-47. 
https://doi.org/10.1016/j.landusepol.2018.11.051  
 
Oyadiran, P. O., and Akintola, O. E. (2014). Public policy making and implementation in Nigeria: the Millennium Development Goals in 
perspectives. International Journal of Diversity, (1), 1-22. 
 
Ozturk (2015). Sustainability in the food-energy-water nexus: evidence from BRICS (Brazil, the Russian Federation, India, China, and 
South Africa) countries. Energy, 93. https://doi.org/10.1016/j.energy.2015.09.104  
 
Paulsson, A., and Macheridis, N. (2022). The policy unconscious: educational labor, the research-and-teaching relationship and the 
unquestioned meaning of higher education. Critical Policy Studies. http://doi.org/10.1080/19460171.2022.2064889  
 
Purtle, J., Stadnick, N. A., Wynecoop, M., and Aarons, G. (2023). A policy implementation study of earmarked taxes for mental health 
services: study protocol. Implementation Science Communications, 4(1). http://doi.org/10.1186/s43058-023-00408-4  
 
Rakšnys, A. V., and Valickas, A. (2023). The application of complexity theory in the context of public governance challenges. Viešoji 
politika ir administravimas, 21(4), 195-205. 10.13165/VPA-22-21-4-14 
 
Sol, S. L. (2023). Participation is important, but-” Professional rationalities of balancing acts in publicly funded cultural institutions. 
Nordisk Kulturpolitisk Tidsskrift, 26(1), 42-56. http://doi.org/10.18261/nkt.26.1.4  
 
Steinert, C. (2016). The impact of societal preferences on national monetary policy outcomes: A revival of economic pluralism. Politikon, 
31, 2. http://doi.org/10.22151/politikon.31.2   
 
Steven, H. (2021). Audience development and cultural policy. New directions in cultural policy research. Cham, Switzerland: Palgrave 
Macmillan 
 
Tacon, R., and Hanson, A. (2011). Sport policy and the structure of sport in the UK. In book: Managing sport business: An introduction: 
Routledge 
 
Uddin, J., Ullah, A., Saqib, N., Kousar, R., and Usman, M. (2023). Heterogeneous role of energy utilization, financial development, and 
economic development in ecological footprint: How far away are developing economies from developed ones. Environmental Science and 
Pollution Research, 30(20). http://doi.org/10.1007/s11356-023-26584-3  
 
Uminska-Woroniecka, A. (2022). We were taken on the ride. Leadership trait analysis in the circumstances of international crisis-the 
decision 
of 
Aleksander 
Kwasniewski 
in 
the 
context 
of 
intervention 
in 
Iraq. 
Teoria 
polityki, 
6, 
215-235. 
http://doi.org/10.4467/25440845TP.22.007.16007  
 
Welsh, J. (2019). Ranking academics: toward a critical politics of academic rankings. Critical Policy Studies, 13(2), 153-173. 
http://doi.org/10.1080/19460171.2017.1398673  
 
Wu, D. (2022). Implementation and public diplomacy. In book: U.S. Public Diplomacy Towards China. http://doi.org/10.1007/978-3-030-
95644-8_3    
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
99 
 
Yanow, D. (2015). Making sense of policy practices: interpretation and meaning. In handbook of critical policy studies, edited by F. 
Fischer, 
D. 
Torgerson, 
A. 
Durnová, 
and 
M. 
Orsini.Cheltenhamn, 
UK: 
Edward 
Elgar 
Publishing. 
401-421. 
https://doi.org/10.4337/9781783472352.00031  
 
Yudiatmaja, W. E., Solina, E., Yudhi, I., and Samnuzulsari, T. (2022). Paid plastic shopping bags policy: What the factors drive the 
implementation? Journal of Governance and Public Policy, 9(1), 13-23. http://doi.org/10.18196/jgpp.v9i1.11152    
 
Zeb-un, N., Mustafa, G., Yaseen, Z., Arslan, M., and Imran, M. (2021). Theoretical approaches to study the public policy: an analysis of 
the cyclic/stages heuristic model. Palarch’s Journal of Archaeology of Egypt/Egyptology, 18(10), 1307-1321. 
 
 
 
 
Funding: This research was supported by the project, which has received funding from the European Union’s Horizon 2020 
research and innovation programme European Research Council (ERC) under the European Union’s Horizon 2020 research 
and innovation programme Marie Sklodowska-Curie Research and Innovation Staff Exchanges ES H2020-MSCA-RISE-
2014 CLUSDEVMED (2015-2019) Grant Agreement Number 645730730 
 
 
Author Contributions: Conceptualization: Adetayo Adeniran, Joseph Muraina, writing-original draft preparation: Adetayo 
Adeniran, Joseph Muraina, Joseph Ilugbami, writing; review and editing: Joseph Ilugbami, Adedayo Adeniran. All authors 
have read and agreed to the published version of the manuscript. 
 
 
 
Adetayo Olaniyi ADENIRAN Department of Logistics and Transport Technology, Federal University of Technology 
Akure, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0000-0002-6870-1212  
 
Joseph Mosunmola MURAINA Department of Geography and Planning Science, Ekiti State University, Ekiti, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0009-0006-5764-3594  
 
Joseph Olanrewaju ILUGBAMI Rufus Giwa Polytechnic-Owo Rector Office, Ondo State, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0009-0005-8114-5264  
 
Adedayo Ayomide ADENIRAN Department of Geopgraphy and Planning, University of Ibadan, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0009-0001-0241-6232  
 
 
 
Make your research more visible, join the Twitter account of INSIGHTS INTO REGIONAL DEVELOPMENT:  
@IntoInsights 
 
This is peer-reviewed scientific journal https://jssidoi.org/ird/page/peer-review-policy 
________________________________________________________________________________________________________________ 
________________________________________________________________________________________________________________ 
Copyright © 2023 by author(s) and VsI Entrepreneurship and Sustainability Center 
This work is licensed under the Creative Commons Attribution International License (CC BY). 
http://creativecommons.org/licenses/by/4.0/ 
 
 
 
View publication stats




This book systematically analyzes how and why China has expectedly lost and then 
surprisingly gained ground in the quest to solve the complicated environmental 
problem of air pollution over the past two decades.
Yuan Xu shines a light on how China’s sulfur dioxide emissions rose quickly 
in tandem with rapid economic growth but then dropped to a level not seen for 
at least four decades. Despite this favorable mitigation outcome, Xu details how 
this stemmed from a litany of policy stumbles within the Chinese context of no 
democracy and a lack of sound rule of law. Throughout this book, the author 
examines China’s environmental governance and strategy and how they shape 
environmental policy. The chapters weave together a goal-­
centered governance 
model that China has adopted of centralized goal setting, decentralized goal 
attainment, decentralized policy making and implementation. Xu concludes that 
this model provides compelling evidence that China’s worst environmental years 
reside in the past.
This book will be of great interest to students and scholars of Chinese 
environmental policy and governance, air pollution, climate change and sustainable 
development, as well as practitioners and policy makers working in these fields.
Yuan Xu is Associate Professor in the Department of Geography and Resource 
Management, The Chinese University of Hong Kong.
Environmental Policy and 
Air Pollution in China


Strategic Designs for Climate Policy Instrumentation
Governance at the Crossroads
Gjalt Huppes
The Right to Nature
Social Movements, Environmental Justice and Neoliberal Natures
Edited by Elia Apostolopoulou and Jose A. Cortes-­
Vazquez
Guanxi and Local Green Development in China
The Role of Entrepreneurs and Local Leaders, 1st Edition
Chunhong Sheng
Environmental Policy in India
Edited by Natalia Ciecierska-­
Holmes, Kirsten Jörgensen, Lana Ollier 
and D. Raghunandan
Mainstreaming Solar Energy in Small, Tropical Islands
Cultural and Policy Implications
Kiron C. Neale
EU Environmental Governance
Current and Future Challenges
Edited by Amandine Orsini and Elena Kavvatha
The European Union and Global Environmental Protection
Transforming Influence into Action
Edited by Mar Campins Eritja
Environmental Policy and Air Pollution in China
Governance and Strategy
Yuan Xu
For more information about this series, please visit: www.routledge.com/
Routledge-­
Studies-­
in-­
Environmental-­
Policy/book-­
series/RSEP
Routledge Studies in Environmental Policy


Environmental Policy and 
Air Pollution in China
Governance and Strategy
Yuan Xu


First published 2021
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2021 Yuan Xu
The right of Yuan Xu to be identified as author of this work has been 
asserted by him in accordance with sections 77 and 78 of the Copyright, 
Designs and Patents Act 1988.
The Open Access version of this book, available at www.taylorfrancis.
com, has been made available under a Creative Commons Attribution-­
Non 
Commercial-­
No Derivatives 4.0 license.
Trademark notice: Product or corporate names may be trademarks or 
registered trademarks, and are used only for identification and explanation 
without intent to infringe.
British Library Cataloguing-­
in-­
Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-­
in-­
Publication Data
A catalog record for this book has been requested 
ISBN: 978-­
1-­
138-­
32232-­
5 (hbk)
ISBN: 978-­
0-­
429-­
45215-­
4 (ebk)
Typeset in Times New Roman
by Apex CoVantage, LLC


List of figures
vi
List of tables
ix
Preface
x
Acknowledgments
xii
1	
Introduction
1
2	
Political will
17
3	
Environmental governance
25
4	
Mobilizing the government
42
5	
Policy making
77
6	
Policy implementation
105
7	
Environmental technology and industry
149
8	
Goal-­
centered governance
179
Index
193
Contents


	
1.1	
Environmental Performance Index in the baseline year
2
	
1.2	
China’s premature deaths due to air and water pollution 
in the Global Burden of Disease study
2
	
1.3	
Disability-­
adjusted life years (DALYs) in China due to air 
and water pollution in the Global Burden of Disease study
3
	
1.4	
DALYs in days (or disability-­
adjusted life days [DALDs]) 
per person per year in China and India
4
	
1.5	
Polity Democracy Index for China, South Korea, Singapore, 
India and the United States
5
	
1.6	
GDP per capita in PPP (purchasing power parity) in China, 
South Korea, Japan and the United States
7
	
1.7	
Governance indicators of China, India and the United States
8
	
1.8	
SO2 emissions in China
10
	
1.9	
SO2 emissions by sector in China (from two different data 
sources for 1970–2012 and 2010–2017, respectively)
11
	
1.10	 The power sector’s shares of coal consumption and SO2 
emissions in China and the United States
11
	
1.11	 SO2 emissions in the United States and SO2 intensities 
in China and the United States
12
	
2.1	
Sectoral employment changes and GDP growth rates across 
China’s administrations
19
	
2.2	
Employment and population structures in China
19
	
3.1	
Environmental protection personnel at four governmental levels 
in China
29
	
3.2	
Governmental revenue and expenditure to GDP ratios by central 
and local governments in China
33
	
3.3	
Budget balance of central and local governments in China 
as a proportion of GDP
35
	
3.4	
Governmental budget balance by provinces as a proportion 
of governmental expenditures in 2018
36
	
3.5	
The central and local governments’ shares of expenditures 
by budgetary items in 2018
37
	
3.6	
Central, local and overall governmental expenditures 
by budgetary items in 2018
37
Figures


Figures  vii
	
3.7	
Shares in governmental expenditures
38
	
4.1	
Designated SO2 emission intensity in distributing SO2 emissions 
quota to coal-­
fired power plants for 2010 in the 11th Five-­
Year Plan
58
	
4.2	
Daily SO2 concentrations in Shijiazhuang
68
	
4.3	
Daily PM2.5 concentrations in Shijiazhuang
69
	
4.4	
Daily 8-­
hour O3 concentrations (daily maximum concentration 
over 8 hours) in Shijiazhuang
70
	
4.5	
Monthly average AQI in Shijiazhuang
70
	
4.6	
Monthly average AQI in Beijing
71
	
4.7	
Monthly average AQI in Shenzhen
72
	
5.1	
Economic growth in China, Japan and the United States
85
	
5.2	
Primary energy consumption and energy efficiency
86
	
5.3	
Prices of coal (Qinhuangdao spot price), oil and natural gas
87
	
5.4	
The annual growth of primary energy consumption in China 
by fuels
88
	
5.5	
China’s primary energy consumption by fuel and the shares 
of coal and fossil fuels
88
	
5.6	
Primary energy consumption and its electrification rate
89
	
5.7	
Electricity generation by fuels in China
90
	
5.8	
The annual growth of electricity generation in China by fuels 
and coal’s share
91
	
5.9	
The decomposition of China’s SO2 emissions
91
	
5.10	 Distribution of sulfur contents in coal power plants in China
95
	
5.11	 Coal-­
fired power and SO2 scrubber capacities in China
97
	
5.12	 The annual growth of coal-­
fired power and SO2 scrubber 
capacities in China
98
	
5.13	 Annually increased SO2 scrubber capacity and unit sizes
99
	
5.14	 The annual growth of SO2 scrubber capacity by regions
100
	
5.15	 SO2 scrubbing technologies by unit sizes
101
	
6.1	
The operation of SO2 scrubbers in Jiangsu Province, including 
self-­
reported operation rates and later confirmed operation 
rates
106
	
6.2	
A conceptual model of environmental compliance monitoring
121
	
6.3	
Model simulation of compliance rates in the diagnosing and 
screening systems with available compliance monitoring 
resources and initial compliance rates
128
	
6.4	
Model simulation of equilibrium compliance rates in the 
screening and diagnosing systems in relation to available 
inspection staff
128
	
6.5	
Model simulation of equilibrium compliance rates in the 
screening and diagnosing systems in relation to (a) the number 
of polluters; (b) the ratios between pollution abatement costs 
and noncompliance penalty; (c) available inspection staff, where 
the pollution abatement cost-­
to-­
noncompliance penalty ratio has 
a lognormal distribution; and (d) the relative resource intensity 
of screening and diagnosing technologies.
130


viii  Figures
	
6.6	
Model simulation of equilibrium compliance rates in the 
screening and diagnosing systems in relation to the probabilities 
that (a) the screening technology recognizes compliance 
cases as being compliant, (b) the diagnosing technology 
recognizes compliance cases as being compliant, (c) the 
screening technology recognizes noncompliance cases as being 
noncompliant and (d) the diagnosing technology recognizes 
noncompliance cases as being noncompliant.
133
	
7.1	
The progressive paths on the deployment and operation of SO2 
scrubbers in China and the United States
150
	
7.2	
Annual average unit capital costs of SO2 scrubbers in China and 
the United States
151
	
7.3	
Model projection of the SO2 mitigation path in China’s coal-­
fired power plants: (a) deployment and operation of SO2 
scrubbers under goal-­
centered governance (the dots refer to 
actual data); (b) avoided SO2 emissions under goal-­
centered and 
rule-­
based governance
154
	
7.4	
Yearly university graduates in China from four-­
year 
undergraduate programs by subjects
157
	
7.5	
R&D personnel, expenditure and market value (in 2018 RMB) 
in China
161
	
7.6	
Patents on environmental technology by filing office in the world
162
	
7.7	
Wind energy development in China and the United States
168
	
7.8	
Companies in the Chinese and U.S. markets installing 100-­
MW-­
scale or greater SO2 scrubbers
171
	
7.9	
Average prices of wind turbines in China and the United States
172
	
8.1	
An illustration of the goal-­
centered governance model
183


	
4.1(a)	
Correlation coefficients of key factors for 27 provinces
53
	
4.1(b)	
Summary of variables
54
	
4.2	
Regression model results for distributing the national goal 
to provinces
56
	
4.3	
Regression model results for distributing provincial goals 
to municipalities
59
	
4.4	
Provincial goal distribution matrix
61
	
5.1	
Applied fractions of sulfur retained in ash
93
	
5.2	
Effluent SO2 emissions and necessary SO2 removal rates
97
	
6.1	
Data on SO2 scrubbers in China’s seven coal-­
fired power 
plants
114
	
6.2	
Decision scenarios for the managers of coal-­
fired power plants
117
	
6.3	
Key parameters in the model and their empirical values
126
	
7.1	
Up-­
front lump-­
sum fees of SO2 scrubber technology licenses
158
Tables


China is puzzling to read.
After the Cultural Revolution and a short transitional period, China entered 
the era of Reform and Open-­
up in December 1978. The size of China’s economy 
has skyrocketed by more than 30 times. Despite numerous benefits, this rapid 
economic growth also brought immense pressure on the environment. China’s 
environmental crises are multifaceted, stretching across air, water, soil, ecosystem 
and climate change.
Hope was not readily available. As a public good, environmental protection 
requires effective governmental intervention. However, China is not a democracy, 
and sound rule of law has not been established. The country’s governance quality 
has been ranked consistently and significantly lower than that in developed coun­
tries that are liberal democracies, where environmental quality first deteriorated 
with economic growth and then fundamentally improved. Their experiences sug­
gest that China’s environmental crises are expected, while their solutions are hard 
to reach.
Then what happened in China in the past 15 years became surprising as the 
environmental trajectory deviated away from the projections. Sulfur dioxide (SO2) 
is one air pollutant that is crucial for air quality but very difficult to control. Since 
reaching their peak in the mid-­
2000s, SO2 emissions in China have been declin­
ing, and the downward pace accelerated in the past few years to reach a level not 
seen in more than four decades. A large coal-fired power sector appeared to install 
and operate SO2 scrubbers that mitigate emissions from polluting sources. Simi­
lar desirable outcomes are also observed in other environmental and renewable 
energy fields. However, China has not changed seriously from the perspectives of 
democracy and the rule of law, although environmental policy has been improv­
ing and strengthening. The legal system still does not play any major role in envi­
ronmental protection. Policy making lacks transparency and public consultation, 
while policy blunders are not rare. Policy implementation still has considerable 
problems and is often selective. It is not unusual to hear about the abuse of gov­
ernmental authorities.
This book aims to provide a theoretical understanding to explain how China 
achieved deep and sustained pollution mitigation without democracy and sound 
rule of law. Causal relationships are explored between the favorable outcome and 
Preface


Preface  xi
the unfavorable path. The major puzzle is why China frequently witnesses both 
sides at the same time or whether the conventional insights may have missed 
something important in reading China. China’s strategy is theorized into goal-­
centered governance. China is both highly centralized – in goal setting – and 
highly decentralized – in goal attainment, policy making and implementation. 
Unlike the rule-­
based governance in developed countries as indicated in their 
well-­
established rule of law, China places goals in the first place, while deficien­
cies in policy making and implementation are much tolerated as long as goals 
can be attained. The mitigation trajectory was not centrally planned but gradu­
ally evolved through decentralized pathfinding under centralized goals. In other 
words, the Chinese puzzle should primarily be explained from the perspective of 
its governance strategy but not individual policies. A strategic mistake is often a 
lot more devastating and far-­
reaching than any policy stumble, while an effective 
strategy can accommodate many policy mistakes without compromising much 
the final outcome.
The research and thinking for this book stretched over a dozen years. When 
I first started studying China’s SO2 mitigation around 2007, the hypothesis was 
that the environmental crisis was rooted in policy failures and, more fundamen­
tally, the lack of democracy and the rule of law. However, what unfolded later 
forced me to rethink this causal relationship, especially in the 2010s when the 
mitigation pace dashed forward. As a former physicist, I hope to find a theoretical 
explanation to the Chinese puzzle that is simple, like one equation, and rich. The 
goal-­
centered governance in this book reflects such a new attempt.


I owe a tremendous amount of debts to many people. This book is dedicated 
to Robert H. Socolow, the supervisor of my PhD thesis at Princeton Univer­
sity’s Woodrow Wilson School of Public and International Affairs. His inspi­
ration is vital in my research journey. Much of this book is rooted although 
widely extended from my PhD study over a decade ago. I am grateful for Rob­
ert H. Williams, Denise L. Mauzerall, Eric D. Larson, Yiguang Ju, Gregory 
C. Chow, Edward S. Steinfeld, Richard K. Lester and Kin-­
Che Lam, whose 
support and insights were crucial to sustain and enlighten this research. My 
deep appreciation also goes to numerous interviewees who kindly shared their 
knowledge. I thank Matthew Shobbrook of Routledge, whom I worked with 
to finally complete this book.
My wife, Jing Song, and our two children, Anlan Xu and Antao Song, are per­
petual motivation and sources of encouragement for my research. My parents, 
Meilan Yuan and Yicai Xu, and parents-­
in-­
law, Meiyu Song and Changfa Song, 
provide patient and unconditional support. My family made this work possible, 
especially under the ongoing COVID-­
19 pandemic.
Funding support throughout this research in the past dozen years was provided 
by Princeton University, Massachusetts Institute of Technology, The Chinese 
University of Hong Kong, and Hong Kong Research Grants Council (General 
Research Fund, 14654016).
Parts of the book were adapted with permissions from the author’s several pub­
lished journal articles, including Xu, Y. 2011. The use of a goal for SO2 mitigation 
planning and management in China’s 11th five-­
year plan. Journal of Environmen­
tal Planning and Management, 54, 769–783 [in Chapter 4; Copyright (2011) Tay­
lor & Francis]; Xu, Y. 2011. Improvements in the operation of SO2 scrubbers in 
China’s coal power plants. Environmental Science & Technology, 45, 380–385 [in 
Chapter 6; Copyright (2011) American Chemical Society]; Xu, Y. 2011. China’s 
functioning market for sulfur dioxide scrubbing technologies. Environmental Sci­
ence & Technology, 45, 9161–9167 [in Chapter 7; Copyright (2011) American 
Chemical Society]; Xu, Y. 2013. Comparative advantage strategy for rapid pol­
lution mitigation in China. Environmental Science & Technology, 47, 9596–9603 
[in Chapter 7; Copyright (2013) American Chemical Society]. Much has been 
revised and expanded on. 
Acknowledgments


1  
China’s environmental crises
China faces colossal, multifaceted environmental challenges, many at crisis lev­
els. Its environmental degradation has been widely documented and analyzed in 
academic studies as well as in public media. China is now the largest energy 
consumer, supplier and emitter of most major air and water pollutants as well as 
various greenhouse gases. Together with its geographically high population and 
economic densities, especially in the eastern half of the country, China was cat­
egorized at the very bottom of air quality among the 180 countries and regions in 
the Environmental Performance Index (Wendling et al., 2018; Figure 1.1). Few 
readers would be surprised to know that China’s air quality is among the most 
polluted in the world (Figure 1.1).
Air and water pollution in China have certainly taken a serious toll. China has 
made steady progress in the past decades to significantly reduce premature deaths 
due to water-­
related environmental factors and indoor air pollution, but ambi­
ent particulate matter (PM) pollution has been deteriorating. The Global Burden 
of Disease study elaborates in great detail the causes and risk factors of deaths 
across individual countries (Institute for Health Metrics and Evaluation, 2018). In 
1990, China accounted for 22.2% of the global population, and in 2017, the share 
dropped to 18.5% despite an 18.0% increase in absolute population (Figure 1.2). 
In premature deaths that are due to environmental risk factors, China’s share in the 
world in 1990 was 29.2% for household air pollution from solid fuels and 4.6% 
for unsafe water, sanitation and handwashing. In other words, an average Chinese 
was 31.5% more likely and 79.3% less likely to die prematurely due to the two 
risks than an average person in the world. The shares were significantly reduced 
to 16.5% and 0.6% in 2017, respectively, to make an average Chinese 10.6% and 
96.8% less likely to die prematurely. In absolute terms, they were reduced by 
65.7% and 92.5%, respectively. However, ambient PM pollution caused 404,000 
premature deaths in 1990 and 852,000 in 2017, more than double. Its global share 
climbed from 23.0% to 29.0% over the period. In 2000, indoor air pollution was 
overtaken by ambient PM pollution in causing more premature deaths. In com­
parison to China’s share of the global population, in 1990, an average Chinese 
faced only a slightly greater risk, 3.8%, from ambient PM pollution than an aver­
age person in the world, but in 2017, the risk premium was enlarged to 56.9%.
1	

Introduction


2  Introduction
China
US
India
Japan
South Korea
UK
0
10
20
30
40
50
60
70
80
90
100
0
20
40
60
80
100
Air quality
Air pollution
Figure 1.1  
Environmental Performance Index in the baseline year
Source: Wendling et al. (2018).
Note: “Air pollution” at the x-axis refers to sulfur dioxide (SO2) and nitrogen oxide (NOx) emission inten­
sities, and its baseline year is 2006. “Air quality” in the y-axis indicates household solid fuels (baseline 
year: 2005), fine particulate matter (PM2.5) exposure and PM2.5 exceedance (baseline year: 2008).
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
0
150,000
300,000
450,000
600,000
750,000
900,000
1990
1995
2000
2005
2010
2015
China’s share in the world
)
s
n
o
s
r
e
p
(
 
s
h
t
a
e
d
 
e
r
u
t
a
m
e
r
P
Year
Ambient particulate matter pollution
Household air pollution from solid fuels
Unsafe water, sanitation and handwashing
Share of population
Figure 1.2  
China’s premature deaths due to air and water pollution in the Global Burden 
of Disease study
Source: Institute for Health Metrics and Evaluation (2018).
Note: Solid lines indicate absolute numbers in persons with the left y-axis, while dashed lines refer to 
China’s shares in the world with the right y-axis.


Introduction  3
Another measurement of pollution’s health impact is the disability-­
adjusted life 
years (DALYs) that quantifies the loss of “healthy” life years. It combines the lost 
life years due to both premature deaths and illnesses. Various types of environmental 
pollution in different countries may cause premature deaths and illnesses that cor­
respond to different life expectancies, ages and other situations. The ratio between 
DALYs and premature deaths is much higher for water pollution than for air pollu­
tion. For example, in 2017, China lost 19.8 million, 6.46 million and 0.85 million 
DALYs due to ambient PM pollution, household air pollution from solid fuels, and 
unsafe water, sanitation and handwashing, respectively. The corresponding ratios 
between DALYs and premature deaths were 23.3, 23.8 and 89.0, respectively, to 
indicate the more severe health impacts of water pollution for an average case.
Nevertheless, the indicator of DALYs does not change the conclusion that was 
presented with the examination of premature deaths (Figure 1.3). Substantial pro­
gress was also made on indoor air pollution and water, with their DALYs being 
reduced by 77.3% and 92.0%, while the deterioration trend for ambient PM pol­
lution is distinguished with an increase of DALYs by 47.3%. In terms of China’s 
shares in the world, ambient PM pollution is still the only risk factor among the 
three to surpass that of its population, which accounted for 23.8% of the world’s 
total in 2017. For all DALYs due to the three environmental risk factors, ambient 
PM pollution’s share rose from 25.6% in 1990 to 73.0% in 2017. Accordingly, 
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
0
5
10
15
20
25
30
1990
1995
2000
2005
2010
2015
China’s share in the world
)
s
r
a
e
y
 
0
0
0
,
0
0
0
,
1
(
 
s
Y
L
A
D
Year
Ambient particulate matter pollution
Household air pollution from solid fuels
Unsafe water, sanitation and handwashing
Share of population
Figure 1.3  
Disability-adjusted life years (DALYs) in China due to air and water pollution 
in the Global Burden of Disease study
Source: Institute for Health Metrics and Evaluation (2018).


4  Introduction
environmental pollution in China is more and more dominated by ambient air 
pollution and especially PM pollution.
On average, the DALYs due to various environmental risks indicate that an 
average Chinese loses a significant number of healthy life days for every year liv­
ing in these environmental risks. In 1990, household air pollution from solid fuels 
was the most severe environmental risk in China to incur the loss of 8.7 disability-­
adjusted life days (DALDs) per person, while the damages from ambient PM 
pollution and from unsafe water, sanitation and handwashing were similar at 4.1 
and 3.2 DALDs per person, respectively (Figure 1.4). In other words, an average 
Chinese lost 16.0 health life days due to the three air and water pollution risk 
factors for living through 1990. In 2017, ambient PM pollution became the most 
severe risk factor, being responsible for 5.1 DALDs per person or 1.0 DALDs 
more, after the other two experienced dramatic improvement in the past decades. 
The total loss was 7.0 DALDs for living through 2017.
China is not a unique country to witness the diverging progress of different risk 
factors. India had similar paths for distinguishing the rising importance of ambi­
ent PM pollution in environmental protection. Ambient PM pollution in India 
has remained stable throughout the years to account for 5.7 and 5.6 DALDs per 
person in 1990 and 2017, respectively. Although household air pollution from 
solid fuels still claimed greater health damages in 2017, its steady declining trend 
0
5
10
15
20
25
30
1990
1995
2000
2005
2010
2015
)
r
a
e
y
 
r
e
p
 
n
o
s
r
e
p
 
r
e
p
 
s
y
a
d
(
 
s
Y
L
A
D
Year
China: Ambient particulate matter pollution
China: Household air pollution from solid fuels
China: Unsafe water, sanitation and handwashing
India: Ambient particulate matter pollution
India: Household air pollution from solid fuels
India: Unsafe water, sanitation and handwashing
Figure 1.4  
DALYs in days (or disability-adjusted life days [DALDs]) per person per year 
in China and India
Source: Institute for Health Metrics and Evaluation (2018).


Introduction  5
suggests that ambient PM pollution will soon become the most damaging environ­
mental risk among the three in India as well (Figure 1.4).
2  
China’s expected rise of SO2 emissions and unexpected 
success in SO2 mitigation
China has been rapidly industrializing in the past four decades. Environmental cri­
ses can be empirically expected in the contexts of its rapid economic development, 
rising energy consumption and coal dominance. The expectation also comes from 
crucial governance factors that are believed to be favorable for environmental tran­
sition but that China is especially weak at. First, democracy is believed to be good 
for environmental protection by many scholars (e.g., Payne, 1995). Unfortunately, 
China is not a democracy, and thus, society’s demand for cleaner air is often not 
believed to be able to effectively influence policy making as in a democracy. It is 
generally ranked at the bottom of various democracy indexes. According to Polity’s 
ratings that can reflect the common views of democracy evaluation at least in West­
ern liberal democracies, modern-­
day China, under the communist rule, is debatably 
less democratic than the imperial days in the 19th-­
century Qing dynasty, when the 
emperors still held absolute power, with the Polity index being −6 (Marshall et al., 
2019). China’s economic reform era after the Cultural Revolution only slightly 
–10
–8
–6
–4
–2
0
2
4
6
8
10
1980
1985
1990
1995
2000
2005
2010
2015
Polity Index (–10 ~10) 
Year
China
South Korea
South Korea
India
United States
Singapore
Figure 1.5  
Polity Democracy Index for China, South Korea, Singapore, India and the 
United States (−10 being the most autocratic and 10 the most democratic)
Source: Marshall et al. (2019).


6  Introduction
improved its Polity index from −8 to −7 (Figure 1.5). In comparison, South Korea 
was fundamentally transformed from an authoritarian regime to a democratic one 
after the reform in the 1980s. Singapore is steadily ranked toward the authoritarian 
side. India and the United States are standard democracies despite slight fluctuations.
Democratic states are argued to be more responsive to the public’s demands. If the 
public in a democracy gives top priority to environmental matters, strong political 
will is more likely to be generated (Li and Reuveny, 2006; Payne, 1995; Downey and 
Strife, 2010). Furthermore, the public in a democracy could be more pro-­
environment 
than are the elites in an autocracy; this could be because of better access to informa­
tion, a more developed civil society and a longer time horizon of planning (Li and 
Reuveny, 2006; Payne, 1995). Democracy is generally closely associated with the 
rule of law, and therefore, there should be better enforcement of environmental regu­
lations (Li and Reuveny, 2006). Nevertheless, democracy might also be associated 
with weakness in environmental protection. People’s self-­
interest and the interests of 
business are more difficult to overcome in a democracy (Li and Reuveny, 2006). If 
the public gives only a low priority to having a clean environment, then a democracy 
could be less likely to heavily focus on environmental protection.
Empirical statistical studies have found no conclusive relationship between 
democracy and the environment. Congleton (1992) and Neumayer (2002) found 
that democracy contributes positively to international environmental commit­
ments. Midlarsky (1998) discovered that democracy leads to more protected 
areas of land, but that it tends to negatively influence deforestation and carbon 
dioxide (CO2) emissions per capita. Winslow (2005) found only good effects of 
democracy, whereas Pellegrini and Gerlagh (2006) found that it had insignificant 
impacts. The mixed results of the relationship could be at least partly caused by 
the difference in environmental indicators. For example, CO2 is more difficult 
to abate, but it has much less local influence than urban particulate pollution. 
Studies that used panel data also reported mixed results regarding the relation­
ship (Torras and Boyce, 1998; Barrett and Graddy, 2000). Different democracy 
indexes do not differ greatly in their relationship to the environment. A prob­
lem in the literature is that a linear relationship is generally assumed between 
democracy and the environment. However, theoretical arguments might suggest 
that both democracy and autocracy could have a beneficial effect on environ­
mental protection, while regimes in between make the situation worse. Among 
control variables, the most common one is income. Considering the literature on 
the Environmental Kuznets Curve and a plausible relationship between income 
and the environment (Grossman and Krueger, 1995; Stern and Common, 2001), 
income together with its squared and cubed terms are necessary control vari­
ables. One study that did not include income as an independent variable could 
suffer from potential missing-­
variable problems (Winslow, 2005). In addition, 
two studies controlled a governance index, namely, that of corruption (Pellegrini 
and Gerlagh, 2006; Buitenzorgy and Mol, 2011), but most of them disregarded 
governance. Various studies differ greatly from each other in how they control 
other variables, including trade openness (Li and Reuveny, 2006), inequality/
Gini ratio (Torras and Boyce, 1998), energy resource endowment (Congleton, 


Introduction  7
1992), country size in gross domestic product (GDP; Winslow, 2005), population 
size (Neumayer, 2002; Congleton, 1992) and literacy (Torras and Boyce, 1998).
Case studies found no conclusive relationship either. A case study in Kenya 
found that democracy is benign to the environment; this is because the government 
responded mainly to the “environmental and developmental civil society” and 
“Western supporters” rather than to the “marginalized poor” (Njeru, 2010). On the 
other hand, democratization in a number of southern African countries, particu­
larly Malawi, South Africa and Mozambique, has resulted in greater destruction 
of the environment for short-­
term economic and social reasons (Walker, 1999). 
In Mexico City, it has been found that democratic elections do not assist in stop­
ping local deforestation (Hagene, 2010). Through studying China and Southeast 
Asia, it is even proposed that “ ‘good’ authoritarianism” is essential for solving 
our urgent environmental problems (Beeson, 2010). A case study in Guatemala 
found that the relationship between democracy and the environment is complex 
and not straightforward (Sundberg, 2003).
In addition, the empirical relationship between economic development and 
environmental quality did not expect that China would be able, or willing, to 
pull down its pollutant emissions and improve air quality. Environmental Kuznets 
Curve – an empirical bell-­
shaped relationship between income level and environ­
mental quality – predicts that before a country becomes rich enough to reach a 
certain level of income (or GDP per capita), its environmental quality will keep 
0
10,000
20,000
30,000
40,000
50,000
60,000
1980
1985
1990
1995
2000
2005
2010
2015
)
p
a
c
/
$
S
U
 
1
1
0
2
 
,
P
P
P
(
 
a
t
i
p
a
c
 
r
e
p
 
P
D
G
Year
South Korea
China
Japan
United States
Figure 1.6  
GDP per capita in PPP (purchasing power parity) in China, South Korea, Japan 
and the United States
Source: IMF (2019).


8  Introduction
deteriorating (Grossman and Krueger, 1995). China’s GDP per capita in purchas­
ing power parity and constant 2011 dollars in 2018 was US$16,100, and the Inter­
national Monetary Fund projected that it would rise to US22,200in2024,whilethelevelwasUS29,100 in the United States in 1980 (Figure 1.6). In other words, 
China is about five decades behind the United States in terms of economic devel­
opment status. Different studies report different turning points, and the lowest one 
for SO2 emissions is at about US3,000(in1990US and nominal exchange rates) 
(Stern and Common, 2001). China’s GDP per capita only surpassed US$3,000 per 
capita in nominal terms in 2008 (IMF, 2019), which was still much lower than the 
empirical minimum turning point.
Furthermore, environmental governance is critical to provide better environ­
mental quality as a public good. As suggested in the World Bank’s six governance 
indicators, comparatively China is poorly governed (Kaufmann and Kraay, 2019). 
The indicators assigned a score between −2.5 (worst) and 2.5 (best) to indicate 
governance performance. On “voice and accountability,” China scored consist­
ently and significantly lower than democracies, such as the United States and India. 
Their average scores from 1996 to 2018 were −1.58, 1.18 and 0.42, respectively 
–2.50
–2.00
–1.50
–1.00
–0.50
0.00
0.50
1.00
1.50
2.00
1996
2000
2003
2005
2007
2009
2011
2013
2015
2017
Governance indicators (–2.5 ~ 2.5)
Year
Voice_China
Voice_India
Voice_US
Law_China
Law-India
Law-US
Figure 1.7  
Governance indicators of China, India and the United States
Source: Kaufmann and Kraay (2019).
Note: “Voice”: Voice and accountability “reflects perceptions of the extent to which a country’s citi­
zens are able to participate in selecting their government, as well as freedom of expression, freedom of 
association, and a free media.” “Law”: Rule of law measures “perceptions of the extent to which agents 
have confidence in and abide by the rules of society, and in particular the quality of contract enforce­
ment, property rights, the police, and the courts, as well as the likelihood of crime and violence.”


Introduction  9
(Kaufmann and Kraay, 2019; Figure 1.7). It suggests that Chinese citizens are 
less able to directly participate in selecting a government and that their voices are 
less likely to be heard. In terms of “political stability and absence of violence/
terrorism,” China scored −0.44, better than India’s −1.13 but worse than United 
States’ 0.48. “Government effectiveness” measures the provision of public and 
civil services as well as the quality of policy making and implementation. It is the 
governance indicator that China had the best performance. It is also the only one 
that China’s score is positive, being 0.09 on average, and consistently improved 
from −0.35 in 1996 to 0.48 in 2018 (Kaufmann and Kraay, 2019). Nevertheless, 
China is still much behind the United States that scored 1.58 in 2018. For the “rule 
of law” indicator, China performs poorly with an average score of −0.46, much 
lower than the United States’ 1.58 and India’s 0.07 (Figure 1.7). Although slight 
progress was made in China from −0.55 in 1996 to −0.20 in 2018, it was always 
located in the negative territory. Little progress was achieved on “corruption” as 
the score remained consistently low with an average of −0.41, which was poorer 
than the 1.47 in the United States and −0.38 in India (Kaufmann and Kraay, 2019). 
China performed steadily poor in “regulatory quality” that focuses on the private 
sector. The United States scored 1.51 on average for the 1996–2018 period, much 
better than China’s −0.25 or India’s −0.36 (Kaufmann and Kraay, 2019). These 
governance indicators quantitatively measure various aspects of governance in a 
country to enable comparison across countries and years. As a classical example 
of market failure to demand governmental intervention, environmental protec­
tion cannot be effective without effective governance. However, none of the six 
governance indicators suggest that the Chinese government can sustainably, effec­
tively and efficiently enact and implement environmental policies and laws.
With all the unfavorable conditions and rising environmental pressures from 
energy consumption, little hope existed to make China’s environmental cleanup 
promising. SO2 is one of the most important air pollutants, and it was also the 
first air pollutant explicitly included in the national Five-­
Year Plans for serious 
mitigation (National People’s Congress, 2006). Its emissions were more than 
doubled from 1980 to the 2000s to echo such expectations (Figure 1.8). How­
ever, something has obviously worked as indicated in the more recent trajectory 
of SO2 emissions (Figure 1.8). Multiple data sources – from Chinese official sta­
tistics, independent bottom-­
up and top-­
down estimates inside and outside of the 
country to satellite and remote sensing data – all point to the same trend: China’s 
SO2 emissions have been rapidly decreasing in the past decade (Li et al., 2017; 
Zheng et al., 2018; Lu et al., 2011; Crippa et al., 2018; Fioletov et al., 2019; 
National Statistics Bureau and Ministry of Ecology and Environment, 2019). 
Although different emission inventories still show gaps between each other on 
when peak SO2 emissions happened and how high they reached, China should 
have completely wiped out all additional SO2 emissions that accompanied its 
unprecedented economic growth in the past four decades (Figure 1.8). Although 
China’s economy has expanded by more than 30-­
fold since the Open-­
up policy 
was initiated in 1978, the country now emits significantly less SO2 (Figure 1.8). 
It seems to have taken China less than one decade to remove all the additional 


10  Introduction
SO2 emissions that the country increased with its economic development and 
energy consumption.
With the rapid electrification trend of energy consumption and the power sec­
tor’s increasing share of coal consumption, the power sector is becoming more 
and more important in deciding the trajectory of China’s SO2 mitigation. In 
1980, its share of SO2 emissions was only 22.5%, less than the industrial sec­
tor’s 50.0% and the residential sector’s 23.5% (Figure 1.9). The relatively less 
significance was due to the power sector’s low share of coal consumption, 20.2% 
(Figure 1.10). In the following two decades, the power sector’s share climbed 
continuously to peak in 2002 at 45.7% and surpass that of the industrial and resi­
dential sectors (Figure 1.9) together with its 52.2% share of coal consumption 
(Figure 1.10). However, these two trajectories started to diverge from each other 
afterward (Figure 1.10). In 2017, the power sector consumed 57.3% of China’s 
coal but only accounted for 17.4% of SO2 emissions (Figure 1.9). The industrial 
and residential sectors’ shares rebounded to reach 56.8% and 22.6%, respectively. 
Accordingly, the power sector now emits much less SO2 for consuming one unit 
of coal than the industrial and residential sectors do.
Although energy transition away from coal is favorable for SO2 mitigation, coal 
consumption in China still remains at a high level, with only a slight decrease 
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
1980
1985
1990
1995
2000
2005
2010
2015
SO2 emissions (1,000 tons)
Year
Official
EDGAR
Lu et al., 2011
Zheng et al., 2018
Fioletov et al., 2019
Figure 1.8  
SO2 emissions in China
Source: Data from Fioletov et al. (2019) refer to large power plants, while others are for China as a 
whole (Zheng et al., 2018; Lu et al., 2011; Crippa et al., 2018; Fioletov et al., 2019; National Statistics 
Bureau and Ministry of Ecology and Environment, 2019).


Introduction  11
0%
10%
20%
30%
40%
50%
60%
70%
0
5
10
15
20
25
30
35
1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 2010 2010 2014
Shares in SO2 emissions
SO2 emissions (million tons)
Power
Industry
Residential
Others
Power’s share (right)
Industry’s share (right)
Residential’s share (right)
Year
Figure 1.9  
SO2 emissions by sector in China (from two different data sources for 1970–
2012 and 2010–2017, respectively)
Source: Crippa et al. (2018); Zheng et al. (2018).
0
500
1,000
1,500
2,000
2,500
3,000
3,500
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1950
1960
1970
1980
1990
2000
2010
Coal consumption (Mtce)
 
e
r
a
h
s
 
s
r’
o
t
c
e
s
 
r
e
w
o
P
Year
Power’s share of coal consumption: China
Power’s share of coal consumption: U.S.
Power’s share of SO2 emissions (EDGAR)
Power’s share of SO2 emissions (Zheng et al., 2018)
Total coal consumption: China (right)
Figure 1.10  
The power sector’s shares of coal consumption and SO2 emissions in China 
and the United States
Source: EIA (2019); Fridley and Lu (2016); National Bureau of Statistics (2019).


12  Introduction
in recent years (Figure  1.10). Most mitigation of absolute SO2 emissions was 
because much greater SO2 emissions are avoided per unit of coal consumption. 
The power sector’s high efficiency in removing SO2 is also reflected in its SO2 
emission intensity of coal-­
fired electricity. Since the enactment of the Clean Air 
Act Amendments (1990), the United States has substantially reduced its overall 
SO2 emissions from 20.9 million tons in 1990 to 2.48 million tons in 2018 (Fig­
ure 1.11). The power sector has consistently been the largest contributor, and its 
SO2 emissions dropped from 14.4 million tons to 1.19 million tons over the same 
period, while its share declined from 68.9% to 47.8%. The much higher share than 
China’s reflects the power sector’s greater importance in U.S. coal consumption 
(Figure 1.10).
In reference to the successful progress in the United States, China’s SO2 miti­
gation trajectory was even steeper. In 1990, for generating 1 kWh of coal-­
fired 
electricity, 8.4 g of SO2 were emitted in the United States, while the rate was 
58.5% higher, or 13.3 g in China. In 2017, as calculated with independent emis­
sion inventory data, the SO2 intensity decreased to be 0.96 g in the United States 
and 0.41 g in China, 56.9% lower (Figure 1.11).
0%
10%
20%
30%
40%
50%
60%
70%
80%
0
5
10
15
20
25
1985
1990
1995
2000
2005
2010
2015
Power sector’s share of SO2 emissions
SO2
O
S
 
&
 
)
s
e
n
n
o
t
 
n
o
i
l
l
i
m
(
 
.
S
.
U
 
e
h
t
 
n
i
 
s
n
o
i
s
s
i
m
e
2
l
a
o
c
 
f
o
 
y
t
i
s
n
e
t
n
i
-
O
S
 
g
(
 
y
t
i
c
i
r
t
c
e
l
e
 
d
e
r
i
f
2/kWh)
Year
Power
Industry
Others
Intensity: U.S.
Intensity: China (EDGAR)
Intensity: China (Zheng et al., 2018)
Power’s share (right)
Figure 1.11  
SO2 emissions in the United States and SO2 intensities in China and the United 
States
Source: Crippa et al. (2018); Zheng et al. (2018); BP (2019); U.S. EPA (2019).


Introduction  13
3  
The organization of this book
Democracy and rule of law have played prominent and indispensable roles in 
environmental cleanup in developed countries. However, China is not a democ­
racy and political freedom is indeed highly constrained, but why environmen­
tal protection became the country’s priority to witness a dramatic drop in SO2 
emissions? Furthermore, many policies are not implemented well, and the legal 
system plays an essentially negligible role in China’s environmental protection. 
But why the government was able to effectively bend down pollutant emissions at 
such an astonishing pace? This book focuses on how China defied the empirical 
expectations in SO2 mitigation, especially in the coal-­
fired power sector. It aims to 
provide an explanation at the strategic level for understanding how environmental 
governance is organized and implemented in China.
This book also aims to imply China’s governance in general. Observers on 
China’s governance often have polarized views and each side seems to have 
ample supporting evidence. Regardless of what the focused perspective is, 
China is full of puzzles and controversies. The country has made many remark­
able achievements in the past 40 years, with much higher income and living 
standards, much better infrastructures, much wider social safety nets, much less 
control of individuals’ private lives and much less poverty. It leads the world 
in renewable energy development and electric vehicles. However, rules are 
much less respected in China than in developed countries. The parliament – the 
National People’s Congress – is often referred to as a “rubber stamp,” although 
in the Chinese Constitution, it has the utmost authority beyond any governmen­
tal entity. The judicial system is not independent. Political liberty is much con­
strained without genuine elections. The Chinese Communist Party has almost 
unchecked power, and the authoritarian country is ruled from the top, but an 
often-­
heard sentence in China goes that “policies and orders cannot go beyond 
Zhongnanhai” (the compound where the central government is located). How 
should we explain China’s governance and reconcile the polarized observations 
that are both well documented and evidence-­
based? Are the two sides caus­
ally connected? How can China achieve those favorable outcomes with such 
an unfavorable policy pathway? If we repair all recognized deficiencies in the 
governance, are we going to throw away the baby together with the bathwater? 
Most important, does China follow a different governance model from that in 
developed countries, and thus, is the explanatory power of many theories and 
historical experiences reduced?
The rest of the book is organized as follows: Chapters 2, 3 and 4 examine how 
the Chinese government is organized for environmental protection, especially in the 
contexts of neither democracy nor sound rule of law. Chapter 2 explores how the 
political will for environmental protection has been centrally evolving without 
democracy. Chapter 3 discusses China’s environmental governance structure that 
combines high degrees of both centralization and decentralization from different 


14  Introduction
perspectives. Primary focuses are on the evolution of the Ministry of Ecology 
and Environment and the relationships between the central and local govern­
ments. Chapter 4 studies how prioritized environmental protection is transmitted 
from the central government to local governments for their effective mobilization 
against the background of a weak rule of law. The environmental governance is 
organized to center on goals, specifically on SO2 emissions and environmental 
protection in Five-­
Year Plans. This book calls the governance strategy in China 
as the goal-­
centered governance model that features centralized goal setting and 
decentralized goal attainment.
Chapters 5, 6 and 7 analyze the impacts of China’s goal-­
centered governance 
model. Chapter 5 focuses on decentralized policy making for SO2 mitigation that 
is guided by centralized, top-­
down goals. This integration of centralization and 
decentralization has generated not only profound outcomes, with active policy 
making, innovation and competition, but also many policy deficiencies. China’s 
governance is tolerant of mistakes or even abuses in policy making, as long as 
goals can be achieved. Such tolerance then significantly reduces the requirements 
for policy making quality, choices of policy instruments and inter-­
policy coordi­
nation. Chapter 6 explores how this goal-­
centered governance has exerted impacts 
on decentralized policy implementation. From unfavorable backgrounds of inad­
equate capacity, effectiveness and efficiency of environmental policy implemen­
tation, local governments make gradual and steady improvements that aim for 
approaching their assigned goals. Chapter 7 addresses how China overcame sup­
ply constraints and established its domestic SO2 scrubber industry for meeting the 
skyrocketing demand. Decentralized market entities were able to actively seek 
and capture market opportunities under goal-­
centered governance. Goals on envi­
ronmental protection and economic development could thus achieve better syner­
gies than conflicts.
Chapter  8 concludes this book and discusses the goal-­
centered governance 
model. This theoretical framework can integrate the polarized observations on 
China within a systematic and compatible understanding. The rule-­
based govern­
ance model is the primarily applied strategy in countries with sound rule of law 
that emphasizes on making good, often centralized policies as means, but the 
final outcome is less explicit. In contrast, this goal-­
centered governance model 
emphasizes centralized goals as ends but is more relaxed on the means to result 
in many policy deficiencies. In the contexts of China’s backgrounds of no democ­
racy and weak rule of law, this governance strategy has been proved effective 
not only on SO2 mitigation but also very likely on other prioritized governmental 
affairs. China is also applying the same strategy in governing CO2 mitigation. 
Other countries may also find this alternative governance model helpful in con­
tributing solutions to their major public problems.
References
Barrett, S. & Graddy, K. 2000. Freedom, growth, and the environment. Environment and 
Development Economics, 5, 433–456.


Introduction  15
Beeson, M. 2010. The coming of environmental authoritarianism. Environmental Politics, 
19, 276–294.
BP. 2019. Statistical review of world energy [Online]. Available: https://www.bp.com/
content/dam/bp/business-­
sites/en/global/corporate/pdfs/energy-­
economics/statistical-­
review/bp-­
stats-­
review-­
2019-­
full-­
report.pdf.
Buitenzorgy, M. & Mol, A. P. J. 2011. Does democracy lead to a better environment? 
Deforestation and the democratic transition peak. Environmental & Resource Economics, 
48, 59–70.
Congleton, R. D. 1992. Political-­
institutions and pollution-­
control. Review of Economics 
and Statistics, 74, 412–421.
Crippa, M., Guizzardi, D., Muntean, M., Schaaf, E., Dentener, F., Van Aardenne, J. A., 
Monni, S., Doering, U., Olivier, J. G. J., Pagliari, V. & Janssens-­
Maenhout, G. 2018. 
Gridded emissions of air pollutants for the period 1970–2012 within Edgar v4.3.2. Earth 
System Science Data, 10, 1987–2013.
Downey, L.  & Strife, S. 2010. Inequality, democracy, and the environment. Organiza­
tion & Environment, 23, 155–188.
EIA. 2019. International energy statistics. Washington, DC: U.S. Energy Information 
Administration.
Fioletov, V., McLinden, C., Krotkov, N., Li, C., Leonard, P., Joiner, J. & Carn, S. 2019. 
Multi-­
satellite air quality sulfur dioxide (SO2) database long-­
term L4 global V1. God­
dard Earth Science Data and Information Services Center (GES DISC) [Online]. Avail­
able: https://disc.gsfc.nasa.gov/datasets/MSAQSO2L4_1/summary.
Fridley, D. & Lu, H. 2016. China energy databook version 9.0. Berkeley, CA: Lawrence 
Berkeley National Laboratory.
Grossman, G. M. & Krueger, A. B. 1995. Economic-­
growth and the environment. Quar­
terly Journal of Economics, 110, 353–377.
Hagene, T. 2010. Everyday political practices, democracy and the environment in a native 
village in Mexico city. Political Geography, 29, 209–219.
IMF. 2019. World economic outlook database October 2019 [Online]. Available: https://
www.imf.org/en/Publications/SPROLLs/world-­
economic-­
outlook-­
databases#sort=
%40imfdate%20descending.
Institute for Health Metrics and Evaluation. 2018. Global burden of disease (GBD) 2017 
study [Online]. Available: https://www.thelancet.com/journals/lancet/article/PIIS0140-
­
6736(18)32279-­
7/fulltext.
Kaufmann, D.  & Kraay, A. 2019. The worldwide governance indicators 2019 update: 
Aggregate governance indicators 1996–2018 [Online]. Available: https://papers.ssrn.
com/sol3/papers.cfm?abstract_id=1682130.
Li, C., McLinden, C., Fioletov, V., Krotkov, N., Carn, S., Joiner, J., Streets, D., He, H., Ren, 
X. R., Li, Z. Q. & Dickerson, R. R. 2017. India is overtaking China as the world’s largest 
emitter of anthropogenic sulfur dioxide. Scientific Reports, 7.
Li, Q.  & Reuveny, R. 2006. Democracy and environmental degradation. International 
Studies Quarterly, 50, 935–956.
Lu, Z., Zhang, Q. & Streets, D. G. 2011. Sulfur dioxide and primary carbonaceous aero­
sol emissions in China and India, 1996–2010. Atmospheric Chemistry and Physics, 11, 
9839–9864.
Marshall, M. G., Gurr, T. R. & Jaggers, K. 2019. Polity IV project: Political regime char­
acteristics and transitions, 1800–2018. Center for Systemic Peace [Online]. Available: 
http://www.columbia.edu/acis/eds/data_search/1080.html.


16  Introduction
Midlarsky, M. I. 1998. Democracy and the environment: An empirical assessment. Journal 
of Peace Research, 35, 341–361.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National Statistics Bureau & Ministry of Ecology and Environment. 2019. China statisti­
cal yearbook on environment 2018. Beijing, China: China Statistics Press.
Neumayer, E. 2002. Do democracies exhibit stronger international environmental commit­
ment? A cross-­
country analysis. Journal of Peace Research, 39, 139–164.
Njeru, J. 2010. ‘Defying’ democratization and environmental protection in Kenya: The 
case of Karura forest reserve in Nairobi. Political Geography, 29, 333–342.
Payne, R. A. 1995. Freedom and the environment. Journal of Democracy, 6, 41–55.
Pellegrini, L. & Gerlagh, R. 2006. Corruption, democracy, and environmental policy – an 
empirical contribution to the debate. The Journal of Environment & Development, 15, 
332–354.
Stern, D. I. & Common, M. S. 2001. Is there an environmental Kuznets curve for sulfur? 
Journal of Environmental Economics and Management, 41, 162–178.
Sundberg, J. 2003. Conservation and democratization: Constituting citizenship in the Maya 
biosphere reserve, Guatemala. Political Geography, 22, 715–740.
Torras, M. & Boyce, J. K. 1998. Income, inequality, and pollution: A reassessment of the 
environmental Kuznets curve. Ecological Economics, 25, 147–160.
U.S. EPA. 2019. Air pollutant emissions trends data: Criteria pollutants national tier 1 for 
1970–2018. Washington, DC: U.S. EPA.
Walker, P. A. 1999. Democracy and environment: Congruencies and contradictions in 
Southern Africa. Political Geography, 18, 257–284.
Wendling, Z. A., Emerson, J. W., Esty, D. C., Levy, M. A., De Sherbinin, A. et al. 2018. 
2018 environmental performance index. New Haven, CT: Yale Center for Environmen­
tal Law & Policy.
Winslow, M. 2005. Is democracy good for the environment? Journal of Environmental 
Planning and Management, 48, 771–783.
Zheng, B., Tong, D., Li, M., Liu, F., Hong, C. P., Geng, G. N., Li, H. Y., Li, X., Peng, L. Q., 
Qi, J., Yan, L., Zhang, Y. X., Zhao, H. Y., Zheng, Y. X., He, K. B. & Zhang, Q. 2018. 
Trends in China’s anthropogenic emissions since 2010 as the consequence of clean air 
actions. Atmospheric Chemistry and Physics, 18, 14095–14111.


1  
Centralized political will
Which governmental affairs can become national priorities and their relative rank­
ings are highly centralized in the Chinese context without democracy. In contrast 
to the path argued by Payne (1995), in which a democracy develops its political 
will regarding the environment, China has taken a different route. The state is far 
more dominant in China than it is in a democracy. Even nongovernmental organi­
zations (NGOs) in China actively seek alliances with the government (Hsu, 2010). 
The lack of free elections also reduces the need for the government to directly 
respond to the public’s demands.
The Chinese Communist Party holds tremendous authority in deciding, for 
example, how important environmental protection is among all governmental 
affairs. The party is closely intertwined with the Chinese government, but they 
are also very different. The party makes key decisions while the government takes 
almost all implementation tasks. Although the party has about 90 million mem­
bers and is organized into multiple levels, the authority is very much centralized 
upward and eventually into the Central Committee. The 19th cohort was inau­
gurated in October 2017 after the corresponding National Party’s Congress. It 
has 204 members, and their tenure will last for five years, until 2022 when the 
next National Party’s Congress convenes to form another Central Committee. It 
further forms the Political Bureau, currently with 25 members, and then, most 
crucially, the 7-­
member Standing Committee as China’s top leadership. Many of 
these members, but not all, also hold positions in the Chinese government. Two 
are most important. The secretary general, currently Xi Jinping, is at the center 
and generally assumes the position of president in the Chinese government. The 
prime minister, currently Li Keqiang, leads the Chinese administration. This hier­
archy ensures China’s high degree of centralization in making most important 
decisions. The Chinese government and, specifically, environmental administra­
tion are mainly focused on environmental policy making and implementation. On 
those prioritized governmental affairs that decisions have been made by the top 
leadership of the Party, the government is in charge of implementation.
In the past seven decades after the establishment of the People’s Republic of 
China, each top leadership of the Chinese Communist Party has left a phrase in 
2	

Political will


18  Political will
the party’s Constitution, with their ideologies written as the party’s “guiding com­
pass,” which not only guides their own leadership’s rule but also summarizes a 
legacy. The line has become longer over time to include “Mao Zedong Thoughts,” 
“Deng Xiaoping Theory,” “Three Representativeness” (headed by President Jiang 
Zemin), “Scientific View of Development” (headed by President Hu Jintao) and 
“Socialistic Thoughts with Chinese Characteristics in the Xi Jinping Era” (Chi­
nese Communist Party, 2017).
This chapter mainly focuses on how the political will for environmental pro­
tection has evolved since the 15th Central Committee was formed in 1998. The 
period transcended three top leaderships of the party, including President Jiang 
Zemin and Prime Minister Zhu Rongji (1998–2002), President Hu Jintao and 
Prime Minister Wen Jiabao (2003–2012) and President Xi Jinping and Primer 
Minister Li Keqiang (2013–2022). Chapter 3 examines the environmental gov­
ernance of the Chinese government for implementing the political will.
2  
Economy, jobs and the environment (1998–2002)
The period was under the 15th Central Committee and the leadership of President 
Jiang Zemin and Prime Minister Zhu Rongji. Although China’s environmental 
pollution had already reached high levels, more urgent issues were present to 
suppress forceful political will for environmental protection. Difficult economic 
conditions slowed down energy consumption to witness a decline of sulfur diox­
ide (SO2) emissions in the 9th Five-­
Year Plan (1996–2000; Figure 1.8).
China’s economy was still at the early stage of industrialization, while the 
Asian financial crisis of 1997 hit China badly. In comparison with the previous 
years (1992–1997), the average annual gross domestic product (GDP) growth rate 
declined significantly from 11.8% to 8.3% (Figure 2.1). GDP per capita was still 
at low levels, US$3,185 (purchasing power parity [PPP] in 2011 US)in1998andUS4,276 in 2002, or 7.4% and 9.3% of the U.S. levels, respectively (Figure 1.6). 
Job creation was more important than GDP growth. As will be introduced in 
Chapter 3, the following year, 1998, witnessed China’s several far-­
reaching fun­
damental reforms with a key focus on state-­
owned enterprises and a better-­
defined 
boundary between the state and the market. Many of these state-­
owned enter­
prises were substantially overstaffed and loss-­
making and operated more like gov­
ernmental agencies and less like market-­
oriented entities. The Chinese financial 
sector and, specifically, the state-­
owned banks had extremely high levels of bad 
debts. This period witnessed large-­
scale privatization and the bankruptcy of small 
and medium-­
sized state-­
owned enterprises, mainly in the secondary sector. As a 
result, the secondary sector shed 8.7 million jobs from 1998 to 2002 to reflect the 
massive reform’s side effects (Figure 2.1). Overall, 3.3 million jobs were annu­
ally added to the secondary and tertiary sectors. With many more people entering 
than leaving the workforce as indicated in the rapidly enlarging age group of 
15-­
 to 64-­
year-­
olds (Figure 2.2), many of the unemployed should have returned 
to rural regions as the primary sector added 18.0 million jobs over the five years 
(Figure 2.1). China’s job and demographic structures were still dominated by the 


Political will  19
–16
–12
–8
–4
0
4
8
12
16
–90
–60
–30
0
30
60
90
1992–1997
1998–2002
2003–2007
2008–2012
2013–2018
Annual increase/decrease of nonprimary jobs 
(million) & annual GDP growth rate (%)
)
n
o
i
l
l
i
m
(
 
s
b
o
j
 
f
o
 
e
s
a
e
r
c
e
d
/
e
s
a
e
r
c
n
I
 
c
i
d
o
i
r
e
P
Primary
Secondary
Tertiary
Nonprimary jobs per year (right)
GDP growth rate (right)
Figure 2.1  
Sectoral employment changes and GDP growth rates across China’s administrations
Source: National Bureau of Statistics (2019).
600
650
700
750
800
850
900
950
1,000
1,050
1,100
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1980
1985
1990
1995
2000
2005
2010
2015
Population in the 15–64 age group (million)
 
n
o
i
t
a
l
u
p
o
p
 
r
o
 
t
n
e
m
y
o
l
p
m
e
 
f
o
 
s
e
r
a
h
S
Year
Primary
Secondary
Tertiary
Urban population
Rural population
Population (15–64; right)
Figure 2.2  
Employment and population structures in China
Source: National Bureau of Statistics (2019).


20  Political will
primary sector and rural regions. The primary sector’s share of total jobs hovered 
stably between 49.8% and 50.1%, while the share of the rural population declined 
from 68.1% in 1997 to 60.9% in 2002 (Figure 2.2). As a result, most Chinese 
were less exposed to seriously polluted urban air pollution because they were not 
breathing urban air.
Generally speaking, over the period from 1998 to 2002, environmental pro­
tection was ranked high neither in governmental affairs nor by society. In the 
aftermath of the Asian financial crisis, economic downturn and unemployment 
were more imminent and highly politicized problems to occupy the top leader­
ship’s mind. This top leadership’s guiding ideology, “three representativeness,” 
was mainly engaged in expanding the party’s base from the conventional working 
class to other categories of the society. Environmental protection did not occupy 
any important role in this ideology, while slower industrial development also 
reduced the deterioration rate of environmental pollution.
3  
SARS and the prioritization of environmental 
protection (2003–2012)
Over the ten years (two terms with the 16th and 17th Central Committee) between 
2003 and 2012, when President Hu Jintao and Prime Minister Wen Jiabao were in 
power, China added 75.6 million new jobs in the secondary sector and 67.3 million 
in the tertiary sector, while the primary sector had a decrease of 108.7 million jobs 
(Figure 2.1). To keep pace with the growing working-­
age population, the annual 
increase of nonprimary jobs was 14.3 million, much faster than the 3.3 million 
new jobs annually between 1998 and 2002 (Figure 2.1). The primary sector still 
accounted for 50.0% of China’s overall employment in 2002 and remained the 
largest among the three sectors in 2007 at 40.8%. China’s entry into the World 
Trade Organization in 2001 and multiple major economic reforms led to unprec­
edented growth in the economy, energy consumption and pollution. The global 
financial crisis of 2008 did exert great and negative impacts on China’s economy 
to slow it down. Comparing the two Hu-­
Wen administrations (2003–2007 and 
2008–2012), the annual economic growth rate came down from 11.7% to 9.4%, 
and the annual increase of nonprimary jobs was from 15.9 million to 12.7 million.
Environmental protection started to emerge as a nationally prioritized govern­
mental affair. The 11th Five-­
Year Plan (2006–2010) was completely formulated 
and implemented under this top leadership of the party. It not only included the 
10% mitigation goals of SO2 and chemical oxygen demand but actually achieved 
them (National People’s Congress, 2011), defying challenges from the rapid 
growth of economy and energy consumption and reversing the humiliating fail­
ures in the 10th Five-­
Year Plan (Figure 1.8). Deeper mitigation of SO2 emissions 
followed in later years, while the turning point of environmental protection hap­
pened within this period (Figure 1.8).
Society might not have been ready to put the environment as a high priority 
with strong cleanup determination. For example, despite the dire situation of air 
pollution, a survey in 2010 by Gallup, a U.S. research-­
based consulting company, 


Political will  21
found that only 26% of the Chinese were dissatisfied, and 73% were satisfied, 
with the air quality (English, 2010). The potentially insufficient support from soci­
ety for pollution mitigation, if China were a democracy, might not have generated 
strong political will.
The much stronger political will for environmental protection reflected more 
the intention of the top leadership of the party. As is examined in detail in Chap­
ter 4, the direct involvement of the top leadership was crucial in enacting the 
environmental goals in the 11th Five-­
Year Plan after the failures in the 10th Five-­
Year Plan. The 16th Central Committee was formed in November 2002 at the 16th 
National Party’s Congress. The Standing Committee of the Political Bureaus was 
headed by Secretary General Hu Jintao and included Wen Jiabao. In March 2003, 
at the 10th National People’s Congress, they assumed the positions of president 
and prime minister, respectively, in the Chinese central government. In the transi­
tional period between these two key conferences, they had only party leadership 
roles but officially not those later government positions.
SARS (severe acute respiratory syndrome), a new infectious disease, emerged 
almost exactly over this transitional period in November  2002 and became 
increasingly damaging over the winter (WHO, 2003). The timing of the devastat­
ing pandemic coincided well with the top leadership’s search for a new ideology 
to distinguish themselves from their predecessors. This public health crisis taught 
a painful lesson to the Chinese leadership that public goods should be prioritized 
together with economic development. The overemphasis of the latter may actu­
ally backfire to result in slow economic growth as the Chinese economy was sig­
nificantly damaged, especially in the second quarter of 2003, by the impacts of 
the SARS pandemic (Rawski, 2005; Hai et al., 2004; Xu et al., 2009). After the 
pandemic was over and society returned to normal, a new ideology was gradually 
formed, titled “Science View of Development,” to emphasize development from 
multiple aspects to achieve a “harmonious society.” Environmental protection is 
a natural extension from public health and became one pivotal component in this 
new development direction.
The authorities of the top leadership and this new ideology were hardly distin­
guishable. From this perspective, whether China could achieve serious mitiga­
tion of environmental pollution and reverse the deterioration trend became more 
politicized. This significantly increased the political will of the top leadership 
to start taking environmental protection into the inner core of key governmental 
affairs. In other words, the political will resulted from a more top-­
down rather 
than bottom-­
up approach, although the pressure from society grew over the years.
Key international events also played a role in shaping China’s environmen­
tal protection. One of the most important events over the Hu–Wen administra­
tions was the 29th Summer Olympic Games in August 2008. To ensure good air 
quality over Beijing, China shut down many polluting factories across several 
neighboring provinces around Beijing. Environmental information was increas­
ingly available over this period. The Internet played a key role in distributing 
information. The U.S. Embassy in Beijing started monitoring fine particulate 
matter (PM2.5) levels in 2008. Environmental NGOs, notably the IPE (Institute of 


22  Political will
Public & Environmental Affairs) that was established in 2006, started systemati­
cally collecting, publicizing and distributing environmental information to the 
public.
4  
The sustainability of environmental political will 
(2013–present)
President Xi Jinping and Prime Minister Li Keqiang assumed their top leadership 
roles of the Chinese Communist Party in November 2012 at the 18th National 
Party’s Congress and then of the central government in March 2013 at the 12th 
National People’s Congress. As usual, the change of leadership did raise questions 
about whether environmental protection could be further strengthened or weak­
ened in relation to new economic conditions and new leaders’ ideas. The Chinese 
economy entered a “new normal,” or a stabilized but lower level after 2013. The 
annual GDP growth rate from 2013 to 2018 was 7.0%, even lower than the level 
during the aftermath of the Asian financial crisis. Nevertheless, the economy had 
already reached a wealthier status before the new leadership came into power and 
the progress since 2013 has also been decent. In 2002, China’s GDP per capita 
was US4,276(PPPin2011US), and it increased to US11,049in2012andUS16,098 in 2018 (IMF, 2019). The ratios between China and the United States 
were 9.3%, 21.8% and 28.8%, respectively.
With the working-­
age population stabilized at about 1  billion people (Fig­
ure 2.2), job creation was still at a healthy pace with 10.7 million new nonpri­
mary jobs added annually. Over the six years, in total, the tertiary sector added 
82.5 million new jobs, while the secondary and primary sectors had 18.5 million 
and 55.2 million fewer jobs (Figure 2.1). In contrast to the economic downturn 
between 1998 and 2002, Chinese labor did not return to rural regions. The tertiary 
sector accelerated significantly to account for 46.3% of all employment in 2018, 
up from 36.1% in 2012 (Figure 2.2). The primary sector accounted for 31.4% of 
all jobs in 2013 and further declined to only 26.1% in 2018 (Figure 2.2). Further­
more, China has been urbanizing fast to have 53.7% of people in urban regions 
in 2013. In 2018, the urbanization rate further increased to 59.6% (Figure 2.1). In 
other words, China’s employment and demographic structures have been much 
more urbanized, which also brought more people under the impacts of more pol­
luted urban air.
Rapid economic development and escalating living standards have been key 
foundations for the Chinese people to maintain support to the Chinese Communist 
Party’s holding of power. The Chinese middle class has expanded rapidly in the 
past decades to indicate that this demand was to a great extent satisfied. Given 
the higher income and more intimate exposure of an average Chinese to urban air 
pollution, society started to place environmental quality at a significantly higher 
priority than before. The balance between environmental protection and economic 
growth has thus been shifting gradually toward the former’s end. In the leader­
ship transitional period in January 2013, North China suffered from severe smog 
with PM2.5 concentration levels reaching hazardous levels (Wang et al., 2014). 


Political will  23
Although Hebei Province had worse air quality, it was Beijing, as China’s capi­
tal, that attracted most international and domestic attention. Air pollution mitiga­
tion started to be widely recognized as one crucial demand by society. People are 
increasingly willing to sacrifice economic opportunities for a better environment. 
Environmental protection and especially urban air quality have been significantly 
politicized, now by society, and implicitly linked with the legitimacy of the Chi­
nese Communist Party as the ruling political party.
In addition, environmental protection also became a more and more visible 
business to create jobs and economic outputs. The initial efforts in the Hu–Wen 
administrations started to bear fruits. China’s environmental and renewable 
energy industries are competitive not only domestically but also internationally 
(Xu, 2013; Zhu et al., 2019). They have grown into another pollical force to push 
for China’s continuous environmental cleanup. For example, China now has the 
world’s largest solar, wind and electric vehicle industries. They play increas­
ingly counterbalancing roles against those who are concerned about the negative 
impacts of environmental protection on their businesses.
In the formation of this top leadership’s governing ideology, the party was also 
keen to significantly elevate the priority of environmental protection. The 18th 
National Party’s Congress in 2012 emphasized ecological civilization, while the 
19th National Party’s Congress in 2017 listed “harmony of people and nature” as 
one of the 14 basic things to insist on, which primarily features ecological civi­
lization and the “two mountains” theory. Previously, in the relationship between 
economic development and environmental protection, the statement was that we 
want not only “gold and silver mountain” but also “clear water and green moun­
tain.” In other words, these two were placed as trade-­
offs to each other. The new 
statement of “two mountains” became that “clear water and green mountain” are 
“gold and silver mountain.” The pursuit of environmental quality became equiva­
lent to economic development. Environmental protection does offer opportunities 
to satisfy the demands for both economic development and a better environment, 
for example, when new industries emerge for pollution mitigation or resource 
conservation. Environmental policies have also been playing an active role in 
encouraging innovation and economic transformation, as elaborated in greater 
detail in Chapter 7.
Overall, in this period, both the top leadership of the party and society came 
together with a common and prioritized stake in a cleaner environment. Environ­
mental protection is increasingly politicized to form an unprecedented political will 
for pollution mitigation. The top leadership should meet the growing demand of 
the society for not just economic growth but also environmental cleanup. Because 
“ecological civilization” is a key component in the top leadership’s “Socialis­
tic Thoughts with Chinese Characteristics in the Xi Jinping Era,” significant 
improvement of environmental quality also became crucial for the establishment 
of this new governing ideology. New economic opportunities and environmental 
industries have been serving as an increasingly visible force to counterbalance 
the negative economic impacts of environmental protection. The rapid growth of 
income has also transformed society’s preference between economic development 


24  Political will
and environmental quality. They are crucial forces to make the political will sus­
tainable, even when top leadership changes again in the future.
References
Chinese Communist Party. 2017. The party’s constitution (Revised by the 19th national par­
ty’s congress) [Online]. Available: https://www.researchgate.net/publication/322421053_
The_19th_Congress_of_the_Communist_Party_of_China_and_Its_Aftermath.
English, C. 2010. More than 1 billion worldwide critical of air quality. Washington, DC: 
Gallup.
Hai, W., Zhao, Z., Wang, J. & Hou, Z. G. 2004. The short-­
term impact of SARS on the 
Chinese economy. Asian Economic Papers, 3, 57–61.
Hsu, C. 2010. Beyond civil society: An organizational perspective on state – NGO relations 
in the people’s Republic of China. Journal of Civil Society, 6, 259–277.
IMF. 2019. World economic outlook databases October  2019 [Online]. Available: 
https://www.imf.org/en/Publications/SPROLLs/world-­
economic-­
outlook-­
databases#
sort=%40imfdate%20descending.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
Payne, R. A. 1995. Freedom and the environment. Journal of Democracy, 6, 41–55.
Rawski, T. G. 2005. SARS and China’s economy. In: Kleinman, A. & Watson, J. L. (eds.) 
SARS in China: Prelude to pandemic? Stanford: Stanford University Press.
Wang, Y. S., Yao, L., Wang, L. L., Liu, Z. R., Ji, D. S., Tang, G. Q., Zhang, J. K., Sun, Y., 
Hu, B. & Xin, J. Y. 2014. Mechanism for the formation of the January 2013 heavy haze 
pollution episode over central and eastern China. Science China-­
Earth Sciences, 57, 
14–25.
WHO. 2003. Update 95 – SARS: Chronology of a serial killer [Online]. Available: https://
www.who.int/csr/don/2003_07_04/en/.
Xu, Y. 2013. Comparative advantage strategy for rapid pollution mitigation in China. Envi­
ronmental Science & Technology, 47, 9596–9603.
Xu, Y., Williams, R. H. & Socolow, R. H. 2009. China’s rapid deployment of SO2 scrub­
bers. Energy & Environmental Science, 2, 459–465.
Zhu, L., Xu, Y. & Pan, Y. J. 2019. Enabled comparative advantage strategy in China’s solar 
PV development. Energy Policy, 133.


1  
Evolution of environmental administration
Environmental protection in China could be traced back to the United Nations 
Conference on the Human Environment in June 1972 in Stockholm, Sweden. In 
the turmoil of the Cultural Revolution (1966–1976) and after the United Nations 
voted in 1971 that the People’s Republic of China is the sole representative of 
China, China sent an official delegation to this conference. In August 1973, the 
First National Conference on Environmental Protection was held to mark that 
environmental protection had formally been recognized as a governmental affair. 
However, in the early stage of the Cultural Revolution, the leaders and organiza­
tions of the Chinese Communist Party and the Chinese government at various 
levels were generally toppled by Red Guards (hong wei bin) and Rebels (zao fan 
pai). Although the Chinese government was rebuilt at a later stage, the primary 
focus was not on economic or social affairs but on class struggle. As a result, 
China did not demonstrate a significant conflict between economic development 
and environmental protection because neither mattered.
When the Cultural Revolution ended in 1976, after a short transitional period, 
China entered the new era of Reform and Open-­
up in December 1978. Economic 
development quickly gained prominence in governmental affairs, while class 
struggle and other political affairs wound down. Soon afterward, the impacts of 
economic development on environmental quality started to emerge. As a pub­
lic affair that requires governmental intervention, environmental protection was 
announced as one Basic National Policy in the Second National Conference on 
Environmental Protection from 31 December 1983 to 7 January 1984. Since then, 
dedicated governmental entities have been established in the Chinese government 
to regulate and implement environmental protection. The agency in the Chinese 
central government that oversees environmental protection has evolved over the 
years in terms of organization, power and jurisdiction. The authority of environ­
mental protection has been increasingly strengthened in the past four decades. 
In 1984, the State Environmental Protection Agency was established under the 
then Ministry of Construction. In 1988, it was pulled out to be directly led by the 
State Council, thus with an elevated status and authority at the vice-­
ministry level. 
Environmental protection then became not just an issue for one single ministry 
3	

Environmental governance


26  Environmental governance
but also one key state affair that was widely relevant and one level closer to the 
center of the governmental authority.
China’s key reforms in the past four decades have one crucial central theme for 
adjusting the relationship between the state and the market. There were essentially 
no real markets in the Cultural Revolution because markets were deemed as too 
capitalistic. Prices did not reflect any balance between demand and supply but 
were decided directly by the government. Purchases should be accompanied by 
permits, not just money. Despite fluctuations, the overall trend in the past dec­
ades was the reemergence, creation and maturity of various markets, as well as 
the refocusing of the state from everything to strategic and public affairs. With 
the government giving up its original authority, prices have become much better 
indicators of supply and demand balances. The production and consumption are 
increasingly guided by market signals and little by orders from central planners.
In the 1998 reform of the State Council that featured a better-­
clarified demarca­
tion between the state and the market, 14 ministries that mainly took direct charge 
of the economic sectors were abolished, and 4 new ministries were formed. The 
government then became more focused on public affairs and much less on direct 
management of businesses. In this reform, the then State Environmental Protec­
tion Agency was promoted to the ministerial level and renamed the State Environ­
mental Protection Administration (SEPA). Other significant reforms in the same 
period marked the reorganization of large state-­
owned enterprises, the privatiza­
tion of small ones and the widened space for private businesses.
Although environmental protection gained increasingly higher statuses in the 
previously mentioned reforms, it was still kept away from the core of the Chinese 
central government, in which the State Council is in charge of the country’s routine 
administration. According to China’s Constitution, the State Council comprises 
the following members: prime minister and deputies, state councilors, ministers, 
directors of commissions and the auditor general. Although the SEPA had been 
elevated to the ministerial level after the 1998 reform, it was not a ministry, and 
thus, its director was not a constitutional member of the State Council. He or she 
could be present in the meeting only by invitation, with much constrained author­
ity on other ministries’ affairs even if they may be closely relevant to environmen­
tal protection. The 2008 reform became crucial when the SEPA was reorganized 
as the Ministry of Environmental Protection and thus became a formal comprising 
ministry of the State Council. This reform indicated that environmental protection 
was recognized as one of the key governmental affairs. The enhanced authority 
also gave the new ministry and its counterparts in local governments more force­
ful power in enacting and implementing environmental policies.
In 2018, a new round of major reforms further concentrated environmental 
authorities that scattered in several ministries into the newly formed Ministry of 
Ecology and Environment (MEE; State Council, 2018). Climate change was nota­
bly transferred out of the National Development and Reform Commission to fall 
under the MEE’s jurisdiction. The MEE now combines the original functions of 
(1) Ministry of Environmental Protection, (2) climate change and mitigation under 
the National Development and Reform Commissions, (3) groundwater pollution 


Environmental governance  27
under the Ministry of Land and Resources, (4) water environment management 
under the Ministry of Water Resources, (5) agricultural pollution under the Minis­
try of Agriculture, (6) ocean environment under the State Oceanic Administration 
and (7) south–north water diversion project’s environmental protection under its 
office. This reform further strengthened the authority of environmental protection. 
The significantly wider duties are expected to create better synergies among their 
regulations and solutions.
2  
Chain of command for environmental protection
Environmental protection administration in China has four major levels, being 
central, provincial, municipality and county. The latter three levels are generally 
categorized as local governments, although provincial governments are often not 
directly involved in local administration. Local governments take primary respon­
sibilities for implementing environmental policies and achieving environmental 
protection. The sequential reforms at the central government were followed by 
corresponding reforms in local governments that generally resemble the struc­
tures of the central government, despite differences contingent on local contexts.
Although the MEE and its predecessors had a clear chain of command under 
the State Council of the central government, it is not straightforward whether local 
environmental protection bureaus (EPBs) should be led by corresponding local 
governments or environmental protection agencies at a higher governmental level 
for achieving more effective environmental administration. On one hand, environ­
mental protection is far beyond the authority of the EPBs to involve industrial pol­
icy, urban planning and other policies. Environmental enforcement heavily relies 
on other agencies and budget allocation from local governments. Accordingly, it 
is reasonable to have local governments as the major office-­
bearers. On the other 
hand, local governments may create barriers to environmental protection due to 
the possible conflicts between economic growth and environmental protection. If 
local EPBs could be vertically controlled, they may better serve the purpose of 
environmental protection as local economic growth is not the central considera­
tion of upper-­
level EPBs.
China’s administrative reform in the past four decades has one key trend: more 
and more remaining governmental authorities are being decentralized from the 
central government to local governments, especially regarding the regulation of 
economic activities and the provision of social public goods such as health care, 
education, housing and urban/rural infrastructure and community services. In 
the environmental administrative system, the chain of command for local EPBs 
reflected such a decentralization trend to recognize that environmental protection 
is generally a localized governmental affair. In 1999, the Department of Organi­
zation of the Chinese Communist Party reformed the institutional arrangements 
and specified that the leaders of local EPBs should be jointly appointed by pri­
marily local governments and, to a lesser extent, upper-­
level EPBs (Department 
of Organization of the Central Committee of the Communist Party of China, 
1999). The “double administration” arrangement aimed for a balance between the 


28  Environmental governance
vertical – or “tiao” based on the function of environmental administration – and 
horizontal – or “kuai” based on the location of environmental protection. The 
1999 reform was accordingly mainly horizontally oriented with decentralization. 
EPBs were under local governments with their directors and budgets controlled 
by their corresponding local governments. They were also advised by EPBs in the 
immediate upper-­
level governments.
The general decentralization trajectory in the past decades also engaged 
another argument for recentralization. In the era of Reform and Open-­
up, local 
governments often have to face the conflicts between environmental protec­
tion and economic development. In evaluating the performance of local leaders, 
economic indicators tended to occupy much heavier weights than environmen­
tal protection, especially in the early years. Accordingly, for the sake of the 
local economy, the environment has often been sacrificed. Together with the 
rising status of environmental protection in the central government as described 
earlier, environmental protection started to climb higher on the priority list. 
The MEE as well as its predecessors and local counterparts are less bound by 
such evaluation because economic development is not their direct job duty, but 
environmental protection is their primary responsibility. In 2016, another major 
and more centralization-­
oriented reform was initiated with several provinces 
for pilot implementation (The General Office of the CPC Central Committee 
and The General Office of the State Council, 2016). The authority of appoint­
ing local EPB leaders and their budgets were shifted more toward upper-­
level 
EPBs. Environmental monitoring and inspection agencies were more directly 
controlled vertically.
3  
Division of labor for policy making and implementation
Environmental agencies in China’s central and local governments have distinct 
functional focuses. The central government is mainly in charge of policy mak­
ing. It also supervises local governments, primarily provincial governments, 
for implementing environmental protection. Provincial governments heavily 
focus on policy making within their individual provinces. They also adapt poli­
cies from the central government to their own situations and supervise mainly 
municipality governments. The municipality level has a further diminished 
capacity in policy making and a much heavier focus on policy implementa­
tion, while the tasks of county governments fall almost exclusively on the 
implementation of policies from the upper levels within localized contexts. 
Implementation is primarily the responsibility of municipality and county gov­
ernments. They can also make decisions that are applied within their specific 
jurisdictions, mainly on how to implement policies with greater efficiency and 
effectiveness.
The clear division of labor among the four levels of governments is reflected 
in their composition of environmental protection personnel. Their personnel com­
positions are accordingly different among the four categories: administration, 
inspection, monitoring and others. “Administration” mainly refers to the MEE 


Environmental governance  29
in the central government as well as corresponding bureaus at the three levels 
of local governments. “Inspection” personnel are those who work in Inspection 
Bureaus, while “monitoring” personnel are based in Monitoring Stations. “Oth­
ers” are the remaining personnel, such as those in the Academy of Environmental 
Sciences and Academy of Environmental Planning at the four levels. They pro­
vide research and expertise to support environmental policy and decision making. 
Between 2004 and 2015, using available data, the compositions at the four gov­
ernmental levels were largely stable (Figure 3.1). The only significant exception 
is the share of “inspection” at the central level, which experienced a dramatic 
increase in 2009 (Figure 3.1).
The environmental authority in the central government is not organized for 
shouldering implementation tasks but primarily for making policies and super­
vising local governments (SCOPSR, 2018). At the central level, “others” is the 
largest category. It accounted for 64.1% of all 3,023 environmental protection 
personnel in 2015, while the share was over 80% before 2009 (Figure 3.1). 
Their dominant share indicates that environmental policy making in China 
requires and has been receiving significant intellectual support. “Administra­
tion” hosted only 362 personnel in 2015, and its share remained stable at about 
12% over the period between 2004 and 2015 based on available data. After 
the 2018 reform and the reorganization, the new MEE was allowed to have 
478 personnel, the addition for accommodating expanded functions (SCOPSR, 
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2004 2009 2014 2004 2009 2014 2004 2009 2014 2004 2009 2014
Total personnel (number)
l
e
n
n
o
s
r
e
p
l
a
t
o
t
f
o
e
r
a
h
S
Year
Administration
Inspection
Monitoring
Others
Total personnel (right)
Central
Provincial
Municipality
County
Figure 3.1  
Environmental protection personnel at four governmental levels in China (for 
2004–2015 using available data)
Source: Ministry of Environmental Protection (2002–2016).


30  Environmental governance
2018). Partly as a result of the establishment of six Regional Supervision Cent­
ers, “inspection” had a major shift with its personnel jumping from 41 in 2008 
to 294 in 2009 and further to 542 in 2015. The 2018 reform further formalized 
and upgraded them into Regional Supervision Bureaus, with a total person­
nel capacity of 240 officers (SCOPSR, 2018). “Monitoring” had about 6% of 
all personnel throughout the years. The inspection and monitoring personnel 
provide crucial data support for supervising the environmental protection per­
formance of local governments.
China’s provincial environmental authorities are also structured to have a 
heavy focus on policy making and supervision and less a focus on direct pol­
icy implementation. At the provincial level, “others” remains the largest to have 
46.4% of all provincial environmental protection personnel in 2015 (Figure 3.1). 
It is the largest category to reflect the desired functions in policy making. “Moni­
toring” occupied 19.9%, which was a decline from 26.9% in 2004 (Figure 3.1). 
The share of “inspection” increased from 6.2% in 2004 to 9.0% in 2015, but the 
increase was much less significant in comparison with that at the central level 
(Figure 3.1). Between monitoring and inspection, the central government now 
puts more emphasis on inspection while provincial governments have a heavier 
focus on monitoring.
The municipality and county levels are structured with much lower capacities 
for policy making and primarily for policy implementation. At the municipal­
ity level, “monitoring” is the largest category, with 34.5% of all its environmen­
tal protection personnel in 2015 (Figure  3.1). “Inspection,” “administration” 
and “others” each took about one fifth of the personnel. Their primary tasks 
are, accordingly, sharply different from the central and provincial levels, with a 
heavy focus on actually implementing policies, although they also build decent 
knowledge support for initiating policy innovations. The county level is almost 
exclusively for implementation, with “others’ accounting for only 6.4% of envi­
ronmental protection personnel in 2015. “Inspection” became the largest func­
tional group with 37.0% of personnel, while “monitoring” and “administration” 
had 28.0% and 28.6%, respectively (Figure 3.1).
The differentiated functions of environmental authorities at the four levels indi­
cate that China’s environmental protection requires their close cooperation. From 
the MEE in the central government to environmental protection bureaus at the 
county level, policy making is more concentrated at the top while implementation 
is mainly at the lower levels. However, their cooperation should not be taken for 
granted, even though China has a conventional image of top-­
down administra­
tion. As examined in later sections, local governments and their leaders have their 
own self-­
interests. If environmental policy implementation is against such inter­
ests, the implementation will not be expected to be effective. As expected from 
China’s weak rule of law, regardless of how stringent environmental policies are, 
their weak implementation was one of the primary reasons that led to China’s 
environmental crises. Without forceful enforcement efforts of local governments 
and widespread compliance of polluting sources, environmental cleanup cannot 
be realized.


Environmental governance  31
4  
Decentralized policy making
From social, economic, industrial and environmental perspectives, China has 
been evolving at an astonishing speed in the past four decades. Laws, policies 
and regulations should continuously adapt to the rapidly changing situations. As 
indicated in the World Bank’s governance indicators, the rule of law in China has 
not been well established (Kaufmann and Kraay, 2019). Laws and courts have not 
been playing important roles in daily environmental protection. Instead, policies 
and regulations are much more closely relevant.
Laws in China are enacted by the National People’s Congress. They tend to take 
many years to formulate, enact or amend. For example, the Law of Environmental 
Protection is the basic law to regulate China’s environmental protection. It was first 
enacted in 1989, and then it took 25 years to get amended in 2014. However, China’s 
environmental conditions and pollution had dramatically changed during the 25 years, 
which should have indicated that the older version was seriously outdated. In addition, 
a variety of specific laws are enacted to regulate individual categories of the environ­
ment. For example, the Law of Atmospheric Pollution Prevention and Control was 
enacted in 1987, and two amendments have been done since then, in 2000 and 2015 
(two other minor corrections were done in 1995 and 2018; National People’s Congress, 
2018). The Law of Water Pollution Prevention and Control was enacted in 1984. Only 
one amendment has been done in 2008, while two minor corrections were made in 
1996 and 2017 (National People’s Congress, 2017). Accordingly, many environmental 
policies in China do not have clear corresponding items in environmental laws.
The slow motion of laws’ enactment and amendment may make them at a great 
distance from the rapidly evolving pollution conditions. This could also partly 
explain why many of China’s policies were applied before their legal foundations 
were established. For example, the eco-­
compensation policy got its legal backing 
only in 2014 in the newly amended Environmental Protection Law, but by then, it 
had already been experimented with and applied widely (Wang et al., 2016). Fur­
thermore, courts do not play significant roles in environmental enforcement and 
compliance. The laws are often written to mainly state principles without enough 
details for direct implementation. The situations reflect China’s situation of weak 
rule of law. The Chinese central government does not file lawsuits against local 
governments for not implementing laws and its policies.
In addition, China’s environmental laws are often intentionally vague in order to 
allow more flexibility for the administration, while environmental policies contain 
more implementable details. Compared with the U.S. Clean Air Act Amendments 
(CAAA, 1990), China’s goal and initial plan were much less detailed. The CAAA 
clearly developed a cap-­
and-­
trade system with detailed rules and schedules (The 
U.S. Congress, 1990). Such details were absent in China’s plans. China’s laws are 
often drafted by a ministry, not the National People’s Congress. For example, a 
key task of the MEE is to draft laws and regulations on environmental protection 
(SCOPSR, 2018). A vague law can provide a legal foundation but not constrain 
the enactment of policies. For example, China’s Law of Atmospheric Pollution 
Prevention and Control entitles the environmental authority to enact ambient air 
quality standards and effluent emission standards without further clarification on 


32  Environmental governance
when and how (National People’s Congress, 2000). The State Council gets the 
legal power to collect effluent emission charges and the freedom to enact any 
relevant regulation (National People’s Congress, 2000).
Other than the National People’s Congress, the State Council can enact Regula­
tions. Various ministries, as well as their internal departments, frequently churn 
out policies, standards, projects and other incentives/commands that are relevant 
to environmental protection. (For simplicity, they are referred to as environmen­
tal policies in the following discussion.) Local governments and their environ­
mental authorities also hold the right to enact their own environmental policies 
or to adapt those from the central government into their corresponding jurisdic­
tions and contexts. As shown in Figure 3.1, local governments, especially at the 
provincial and (to a lesser extent) municipality levels, do have decent capacities 
for making policies. All these environmental policies could have very different 
scopes, stringency, instruments, targets and intellectual support. In comparison to 
laws, environmental policies are much more flexible. Its enactment takes much 
less time and faces much lower hurdles. The entire process is also much less cen­
tralized with numerous governmental bodies at ministerial and local levels who 
can independently enact environmental policies. China’s weak rule of law indi­
cates that these policies are rarely challenged in courts or through other channels 
by affected interest groups, although their legal foundation might be porous and 
shaky in vague and slowly updated environmental laws. In order to understand 
China’s rules for environmental protection, laws are not the most reliable sources.
Nevertheless, ironically the weak status of rule of law in China further strength­
ened the decentralization of environmental policy making. Although the National 
People’s Congress is distinctly different from that in a democracy, laws are nev­
ertheless more stable and more authoritative than policies by the administration. 
Laws are based on wider participation, and the legislative process is more transpar­
ent. If strong enough incentives are present, the variety of policy-­
making entities 
at different levels will be able to actively innovate new policies, learn the lessons 
and experiences from other policy making entities, adapt top-­
down policies and 
adopt policies from other regional contexts. Not all policy making is necessarily 
backed by sound research or intellectual support. Nevertheless, the decentralized 
policy making makes active bottom-­
up policy innovation and diffusion possible.
5  
Decentralized policy implementation
From the perspectives of human resources and fiscal expenditures, China’s capac­
ity for environmental policy implementation is heavily tilted toward local govern­
ments, rather than the central government.
5.1  
Decentralized human resources
Policy implementation demands substantially more resources and personnel than 
policy making. Corresponding to the designed focuses between policy making 
and implementation, most of China’s environmental protection officials are at 


Environmental governance  33
the municipality and county levels. China had 232,388 government employees 
on environmental protection in 2015, a 62.8% increase from 142,766 in 2001 to 
reflect the elevated priority of environmental protection in all government affairs. 
The distributions across the four levels of governments have been quite consist­
ent over the years, with 1.3%, 6.8%, 21.5% and 63.1% of the total environmen­
tal protection personnel in 2015 in central, provincial, municipality and county 
governments, respectively. Corresponding to the four categories, the municipality 
and county levels accounted for 92.5% personnel for administration, 97.0% for 
inspection, 94.6% for monitoring and 69.5% for others (Figure 3.1). As a result, 
the environmental authorities at the central and even the provincial levels do not 
have an adequate human resource capacity to implement environmental policies 
in millions of polluting sources that are scattered in China’s wide geographic ter­
ritories (Ministry of Environmental Protection et al., 2010).
5.2  
Decentralized fiscal expenditure and centralized fiscal revenue
Fiscal revenue and expenditure are other key perspectives for understanding the 
central–local relationship in China. The governmental expenditure-­
to-­
GDP ratio in 
China is not high in comparison to that in developed countries. In 2018, the ratio was 
24.5%, in which the central government accounted for 3.6% and local governments 
20.9% (Figure 3.2). The ratio dropped significantly from 26.8% to 11.1% from 1980 
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
1980
1985
1990
1995
2000
2005
2010
2015
 
o
i
t
a
r
 
P
D
G
 
o
t
 
e
r
u
t
i
d
n
e
p
x
e
/
e
m
o
c
n
i
 
l
a
t
n
e
m
n
r
e
v
o
G
Year
Income: Central
Expenditure: Central
Income: Total
Income: Local
Expenditure: Total
Expenditure: Local
Figure 3.2  
Governmental revenue and expenditure to GDP ratios by central and local gov­
ernments in China
Source: National Bureau of Statistics (2019).


34  Environmental governance
to the mid-­
1990s but has since gradually recovered (Figure 3.2). The ratio between 
governmental revenue and GDP had a similar trend, initially falling from 25.3% in 
1990 to 10.2% in 1995 and then rising back to 20.4% in 2018 (Figure 3.2). The gaps 
between revenue and expenditure indicate fiscal surplus or deficit.
In the current fiscal arrangement, the central government has far more revenue 
than it spends while the local governments in general have to rely on fiscal trans­
fers from the central government for meeting their expenditures. In 2018, the cen­
tral government received 46.6% of total general fiscal revenue but accounted for 
only 14.8% of total fiscal expenditures. Local governments, in contrast, received 
nearly half of the revenue but had to shoulder 85.2% of the expenditures.
The fiscal relationship between the central and local governments have expe­
rienced dramatic changes in the past four decades. In 1980, local governments 
directly received an overall revenue of 87.5 billion RMB (current price), but their 
spending was 56.2 billion RMB (National Bureau of Statistics, 2019). In contrast, 
the central government had a revenue of 28.4 billion RMB but spent 66.7 billion 
RMB. It was the central government, not local governments, that spent most of 
the government budget, ranging from 52.5% to 55.0% between 1980 and 1984 
(Figure 3.2). Accordingly, the central government ran a huge deficit, and local 
governments, a huge surplus. The fiscal transfer was then from local govern­
ments to the central government. It reflected that China’s governance remained 
very much centralized in the immediate years after the Cultural Revolution. The 
central government was directly engaged in providing a significant proportion 
of government services and subsidies. Correspondingly, fiscal expenditures were 
required to support such a provision.
The situation was dramatically changed in 1985. When the governmental 
expenditure-­
to-­
GDP ratio started to drop significantly together with market-­
oriented economic reforms, the central government saw a much steeper decline 
(Figure 3.2). The budgets for both the central and local governments became indi­
vidually more balanced (Figure 3.3). The expenditures of the central and local 
governments were only 3.3% above and 2.1% lower than their revenues in 1985. 
Local governments since then have consistently accounted for more than 60% of 
total governmental expenditures, dwarfing the share of the central government. 
Although local governments’ fiscal conditions remain generally balanced in the 
following years, the central government again started to see a widening gap. In 
1993, its expenditures exceeded revenue by 37.0% while its shares in total gov­
ernment revenue and expenditures had dropped to 22.0% and 28.3%, respectively. 
The budget deficit of the central government fiscally constrained it from exerting 
authority on rich provinces and tackling widening regional disparities across the 
country.
In China’s central–local fiscal relationship, 1994 was a crucial watershed when 
a fundamental tax reform entered into effect in January (State Council, 1993). The 
central government’s share of total governmental revenue skyrocketed to 55.7% 
in 1994 while its share of expenditures remained at 30.3%. For the first time, the 
central government ran a budget surplus, with revenue exceeding expenditures by 
65.7%. In contrast, local governments’ fiscal revenue could cover only 57.2% of 


Environmental governance  35
their expenditures. Then a large fiscal transfer became necessary from the central 
government to local governments. With further decentralization of governmental 
affairs and service provision, this newly formed central–local fiscal relationship 
has been kept increasingly entrenched in the past two decades. In 2018, local gov­
ernments accounted for 85.2% of expenditures but only 53.4% of revenue. The 
gap has significantly widened.
The current central–local relationship that features significant fiscal transfer 
from the central government to local governments reflects their differentiated 
roles in policy making and implementation as discussed earlier. The central 
government is primarily in charge of policy making while the implementa­
tion is largely in the hands of local governments. The former requires much 
less expenditure than the latter. All provinces have their expenditures exceed­
ing revenues, but poor provinces tend to rely on the central government’s 
fiscal transfer much more than rich ones (Figure  3.4). For example, Tibet’s 
governmental revenue covered only 11.7% of its expenditures in 2018, while 
the revenue–expenditure gap for Shanghai was only 14.9%. Accordingly, the 
central government could use fiscal transfer as an incentive for local govern­
ments to implement policies or achieve goals that are enacted from the top. 
It is one of the key incentives that the central government can rely on for the 
cooperation of local governments.
–10.0%
–8.0%
–6.0%
–4.0%
–2.0%
0.0%
2.0%
4.0%
6.0%
8.0%
1980
1985
1990
1995
2000
2005
2010
2015
Budget balance (% of GDP)
Year
Total
Central
Local
Figure 3.3  
Budget balance of central and local governments in China as a proportion of 
GDP
Source: National Bureau of Statistics (2019).


36  Environmental governance
With increasing decentralization in the economic reform, more and more budg­
etary items were shifted with local governments as primary entities of governmen­
tal expenditures. Reflecting the division of governmental affairs, the central and 
local governments now have distinct responsibilities on a variety of expenditure 
items. Foreign affairs and national defense are two budgetary items that the cen­
tral government takes almost exclusive responsibility to account for 99.5% and 
98.1%, respectively, of total governmental expenditures. Of the central govern­
ment’s expenditure in 2018, 33.8% was devoted to national defense. Grain storage 
is for the country’s food security, and thus, the central government remained more 
important, being responsible for 66.8% of all governmental expenditures in 2018 
(Figure 3.5). Science and technology is another classical category of public good 
that the market underinvests in to require public expenditures, in which the central 
government took a share of 37.5% in 2018 (Figure 3.5). Health care and urban 
and rural communities are almost exclusively the responsibility of local govern­
ments. Environmental protection was responsible for 2.9% of total governmental 
expenditures in 2018 (Figure 3.6), while local governments accounted for 93.2% 
(Figure 3.5). It occupied 3.1% of local governments’ expenditures and 1.3% of the 
central government’s (Figure 3.6).
The expenditure structures between the central and local governments have 
remained generally unchanged for environmental protection in the past decade. 
However, this largely decentralized budgetary item has also witnessed signs of 
Beijing
Tianjin
Hebei
Shanxi
Inner Mongolia
Liaoning
Jilin
Heilongjiang
Shanghai
Jiangsu
Zhejiang
Anhui
Fujian
Jiangxi
Shandong
Henan
Hubei
Hunan
Guangdong
Guangxi
Hainan
Chongqing
Sichuan
Guizhou
Yunnan
Tibet
Shaanxi
Gansu
Qinghai
Ningxia
Xinjiang
–100.0%
–90.0%
–80.0%
–70.0%
–60.0%
–50.0%
–40.0%
–30.0%
–20.0%
–10.0%
0.0%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
)
s
e
r
u
t
i
d
n
e
p
x
e
 
l
a
i
c
n
i
v
o
r
p
 
f
o
 
%
(
 
e
c
n
a
l
a
b
 
t
e
g
d
u
B
GDP per capita (RMB/person)
Figure 3.4  
Governmental budget balance by provinces as a proportion of governmental 
expenditures in 2018
Source: National Bureau of Statistics (2019).


Environmental governance  37
0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%
 Total
 Interregional aid
 Urban & rural communities
 Health care
 Agriculture, forestry & water
 Social security & employment
 Commercial services
 Education
 Environmental protection
 Housing
 Resource exploration & information
 Culture, sports & communication
 General public service
 Transportation
 Public security
 Land, ocean & meteorology
 Others
 Science & technology
 Debt interest
 Financial
 Debt issuance
 Grain storage
 National defense
 Foreign affairs
Share of governmental expenditures
Local’s share
Central’s share
Figure 3.5  
The central and local governments’ shares of expenditures by budgetary items 
in 2018
Source: National Bureau of Statistics (2019).
0%
5%
10%
15%
20%
25%
30%
35%
 Interregional aid
 Urban & rural communities
 Health care
 Agriculture, forestry & water
 Social security & employment
 Commercial services
 Education
 Environmental protection
 Housing
Resource exploration & information
 Culture, sports & communication
 General public service
 Transportation
 Public security
 Land, ocean & meteorology
 Others
 Science & technology
 Debt interest
 Financial
 Debt issuance
 Grain storage
 National defense
 Foreign affairs
Share of governmental expenditures
Local
Central
Total
Figure 3.6  
Central, local and overall governmental expenditures by budgetary items in 2018
Source: National Bureau of Statistics (2019).


38  Environmental governance
slight recentralization. Recent reforms as described earlier reflected and enabled 
the central government to be keener in improving environmental quality and more 
directly involved in supervising local governments. Environmental protection has 
been listed as a separate budgetary item in the data from the China Statistical 
Yearbook since the 2008 edition (for 2007 data). Its share in total governmental 
expenditures has inched up from 2.0% in 2007 to 2.7% in 2010 and then fluctu­
ated narrowly to reach 2.9% in 2018. The share in local governments’ budgets has 
also been quite stable, within a narrow range between 2.5% and 3.2% over the 
period. However, the central government had a significant shift, allocating a much 
greater share of its budget for environmental protection. It ranged between 0.2% 
and 0.5% from 2007 to 2013 but then jumped to 1.5% in 2014 and has remained 
at the level since then (Figure 3.7). Correspondingly, the central government’s 
share in total environmental protection expenditures was lifted from 2.9% in 2013 
to 9.0% in 2014, while the local governments’ share dropped although their envi­
ronmental protection expenditures were increased every year in absolute terms.
The significant uplifting in 2014 indicates that environmental protection has 
been increasingly prioritized in China’s public affairs (Figure 3.7). The additional 
budget mainly corresponded to the strengthened functions of top-­
down supervi­
sion, monitoring and inspection of local governments’ performance. Because the 
shares in governmental expenditure for China as a whole and for local governments 
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
9.0%
10.0%
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
s
e
r
u
t
i
d
n
e
p
x
e
 
l
a
t
n
e
m
n
r
e
v
o
g
 
n
i
 
s
e
r
a
h
S
Year
Central’s share in environmental
protection expenditures
Environmental protection’s share in total
expenditures
Environmental protection’s share in
central’s total expenditures
Environmental protection’s share in
local’s total expenditures
Figure 3.7  
Shares in governmental expenditures
Source: National Bureau of Statistics (2019).
Note: The National Statistical Yearbook listed environmental protection as a separate budgetary item 
for the first time in 2007.


Environmental governance  39
did not change significantly over the period and especially in 2014, environmen­
tal administrative capacities were not expected to be upgraded disproportionally 
against other governmental affairs. The emphasis on environmental protection 
thus targeted the relationship between the central and local governments to more 
effectively mobilize implementation capacities and to assign a heavier weighting 
to environmental protection relative to local economic development.
6  
Centralized and decentralized personnel management
According to the Chinese Constitution, local leaders are elected by corresponding 
local People’s Congress. Then they are supposed to mainly please their local elec­
torate. Because China’s weak rule of law does not ensure the local implementa­
tion of environmental laws and policies from the National People’s Congress and 
the central government, the central government should only be able to exert very 
constrained authority over the selection of local government leaders and what 
governmental affairs they decide to pursue in their local contexts. Even if local 
leaders refused to implement policies from the top, only local People’s Congress 
can remove them. However, this very decentralized arrangement presents a sharp 
contrast with reality. Far-­
reaching reforms in the past four decades have featured 
economic reforms on the relationship between the state and the market and admin­
istrative reforms on the relationship between the central and local governments, 
but the relationship between the Chinese Communist Party and the Chinese gov­
ernment has witnessed fewer changes. In the 1980s, their separation was debated 
and explored in tentative reforms, but the progress has been much slower.
The party plays a crucial role in shaping the central–local leadership relation­
ship in reality. The party and the Chinese government have overlapped organiza­
tions in the governmental bureaucracy, while the party is even more prevalent 
to be present in enterprises and other nongovernmental organizations. Although 
local leaders should be elected by local People’s Congress, the party, and espe­
cially its Department of Organization, controls the nominations. For the four gov­
ernmental levels, each level has the authority to appoint leaders at one lower level. 
For example, the party’s Department of Organization at the central level controls 
the nomination of provincial-­
level leaders (including those in central ministries). 
Each provincial Department of Organization nominates municipality-­
level lead­
ers within the province. The appointment decisions are in the hands of their cor­
responding party committees. As a result, personnel decisions are one crucial 
channel for the central government to influence local governments. Numerous 
studies have confirmed that China does put governance performance in the deci­
sions to promote or remove officials, especially local government leaders (Li and 
Zhou, 2005; Zhou, 2007). Without the party’s role, China’s governance would be 
substantially different from the current institutional arrangement.
Furthermore, a reform did bring a major change in this personnel relationship 
with significant decentralization. In 1984, the Central Committee of the party 
reformed its personnel management system (Gao and Zou, 2007). Before then, the 
Department of Organization at each level managed two levels down. For example, 


40  Environmental governance
the Central Department of Organization was in charge of nominating and manag­
ing leaders at the provincial and municipality levels. After the reform, the leaders 
at the municipality level are left to the sole responsibility of the Provincial Depart­
ment of Organization, while the Central Department of Organization only takes 
care of the provincial-­
level leaders. As a result, the provincial leaders will have 
much stronger control of their staff and other local government leaders below 
them. Such reform substantially reinforces local leaders’ authorities within their 
jurisdictions. The arrangement coincides with the decentralization of governmen­
tal affairs and expenditures but still maintains a powerful channel through the 
party for the central government to control local leaders.
References
Department of Organization of the Central Committee of the Communist Party of China. 
1999. On reforming the institutions of managing environmental officials. Beijing, China: 
Central Committee of the Communist Party of China.
Gao, X. & Zou, Q. 2007. Research on intra-­
party democracy – evaluation from history and 
reality. Shandong, China: Qingdao Press.
The General Office of the CPC Central Committee & The General Office of the State 
Council. 2016. Guiding advice on the pilot vertical reform in sub-­
provincial monitoring, 
inspection and enforcement agencies. Beijing, China: CPC Central Committee, State 
Council.
Kaufmann, D.  & Kraay, A. 2019. The worldwide governance indicators 2019 update: 
Aggregate governance indicators 1996–2018 [Online]. Available: https://info.world­
bank.org/governance/wgi/.
Li, H. B. & Zhou, L. A. 2005. Political turnover and economic performance: The incentive 
role of personnel control in China. Journal of Public Economics, 89, 1743–1762.
Ministry of Environmental Protection. 2002–2016. Annual statistical report on the envi­
ronment in China. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection, National Statistics Bureau  & Ministry of Agri­
culture. 2010. Public report on the first national census of polluting sources. Beijing, 
China: Ministry of Environmental Protection, National Statistics Bureau.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National People’s Congress. 2000. Law of atmospheric pollution prevention and control 
of people’s republic of China. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2017. Law of water pollution prevention and control. Beijing, 
China: The 4th Conference of the 10th National People’s Congress.
National People’s Congress. 2018. Law of atmospheric pollution prevention and control 
of people’s republic of China. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
SCOPSR. 2018. The function, internal organization and personnel of the ministry of ecol­
ogy and environment. Beijing, China: SCOPSR.
State Council. 1993. Decision on implementing the tax sharing mechanism in fiscal man­
agement. Beijing, China: State Council.
State Council. 2018. Reform plan on the state council. Beijing, China: State Council.


Environmental governance  41
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. 
Congress.
Wang, H., Dong, Z., Xu, Y. & Ge, C. 2016. Eco-­
compensation for watershed services in 
China. Water International, 41, 271–289.
Zhou, L. 2007. Governing China’s local officials: An analysis of promotion tournament 
model. Economic Research Journal, 7, 36–50.


1  
Goals in China’s Five-­
Year Plans
China’s top leadership has gradually gained strong enough political will for 
environmental protection over the past decades (Chapter  2). However, the 
decentralization of policy making and, to a greater extent, policy implementa­
tion requires the cooperation between the central and local governments to 
realize the environmental political will with concrete improvement of environ­
mental quality and pollution mitigation (Chapter 3). This chapter is devoted 
to understanding how the entire Chinese government, from central to local 
governments, is mobilized through environmental goals, especially in Five-­
Year Plans.
Goals have been widely used in governance. For example, UNFCCC (United 
Nations Framework Convention on Climate Change) defines its goal as “stabili­
zation of greenhouse gas concentrations in the atmosphere at a level that would 
prevent dangerous anthropogenic interference with the climate system” (United 
Nations, 1992). President Barack Obama set up a goal to withdraw all U.S. troops 
from Iraq by the end of 2011 (DeYoung, February 28, 2009). Many studies are 
about environmental goals, including those on negotiating goals, distributing 
goals (Chakravarty et al., 2009), policies to achieve goals (such as on emission tax 
and cap-­
and-­
trade) and technological achievability of goals (Pacala and Socolow, 
2004).
A theoretical foundation of using goals as a governance tool can be traced to 
studies in social psychology: through experiments on individuals, the impact of 
various goals on task performance is examined. Locke et al. (1981) reviewed 
the literature and concluded that “specific and challenging goals lead to higher 
performance than easy goals, ‘do your best’ goals, or no goals.” Furthermore,
goal setting is most likely to improve task performance when . . . the subjects 
have sufficient ability, . . . feedback is provided to show progress in rela­
tion to the goal, rewards such as money are given for goal attainment, the 
experimenter or manager is supportive, and assigned goals are accepted by 
the individual.
(Locke et al., 1981)
4	

Mobilizing the government1


Mobilizing the government  43
In the experiments, goals are distributed to individuals and individuals try to accom­
plish the goals. The situation is not much different from an environmental goal 
in a big country like China. The Chinese central government plays a similar role 
as experimenters: it decides a goal and distributes it to local governments. Three 
components could be distinguished: (1) goal setting, (2) goal distribution and (3) 
goal attainment. Goal setting refers to what type of goals should be set up and how 
stringent they are. Because a global or national goal often requires the cooperation 
of different political or administrative entities, goal distribution is necessary. For 
example, a global goal of carbon dioxide (CO2) mitigation should be distributed to 
individual countries, and a Chinese national goal should be distributed to provinces. 
Furthermore, these goals need to be accepted before serious efforts are made. The 
third component of a goal process focuses on evaluating goal attainment. Strong-­
enough incentives should be put into place to mobilize goal implementers.
The seven-­
decade history of the People’s Republic of China can be divided 
into two periods: a centrally planned economy in the first three decades and a later 
era of market-­
oriented economic reforms. Since the 1950s, originally adopted 
from the Soviet Union, Five-­
Year Plans have become pivotal to guide China’s 
economic development. Although China’s economy was strictly state-­
controlled 
before the economic reforms began in 1978, only the first of the earliest five Five-­
Year Plans was actually completed (Liu et al., 2006). The other four were not able 
to be performed due to frequent political movements, with the Cultural Revolu­
tion as the most notable one (Liu et al., 2006).
Five-­
Year Plans gained momentum only in the second period when China tried 
to establish a market-­
oriented economy. Starting from the 6th Five-­
Year Plan 
(1981–1985), China has gradually formed a set of rules to design these plans 
(State Council, 2005b). The 11th Five-­
Year Plan (2006–2010) was the first to 
change its name from “jihua” (more forceful plans) to “guihua” (more directional 
plans). Goals are the most important indicators in the Plans. From the 11th Five-­
Year Plan, goals are distinguished into foreseeable ones (such as the growth rates 
of gross domestic product [GDP] and population) and legally binding ones (such 
as pollutant mitigation; National People’s Congress, 2006). In addition, China’s 
Five-­
Year Plans are not just one document but a system composed of many layers. 
For example, for the nation as a whole, there was a National 11th Five-­
Year Plan 
that included a 10% reduction goal of sulfur dioxide (SO2) emissions. Another 
11th Five-­
Year Plan on Environmental Protection provided further details. At one 
more layer lower, the 11th Five-­
Year Plan on Acid Rain and SO2 Pollution Control 
specifically addressed the mitigation of SO2 emissions. There were also 11th Five-­
Year Plans at all governmental levels.
Goals are playing more and more prominent roles in China’s environmental 
protection, especially in Five-­
Year Plans, to mobilize local governments and the 
Chinese bureaucracy. If expressed in percentage terms, the baseline year is the 
final year of the previous Five-­
Year Plan. For example, China’s energy intensity 
goal in the 11th Five-­
Year Plan (2006–2010) was a 20% reduction (National Peo­
ple’s Congress, 2006); it indicates that China planned to reduce energy intensity, 
or energy consumption per unit of GDP, by 20% in 2010 from the 2005 level.


44  Mobilizing the government
In regulating SO2 emissions that mainly come from the burning of coal, China 
relies on absolute emission goals, which were a 3.8% increase, a 10% reduction, a 
10% reduction, an 8% reduction and a 15% reduction, respectively, for the 9th, 10th, 
11th, 12th and 13th Five-­
Year Plans (National People’s Congress, 2001, 2006, 2011; 
NEPA et al., 1996; National People’s Congress, 2016). The actual growth rates of 
SO2 emissions were a 15.8% reduction, a 27.8% increase, a 14.3% reduction and a 
14.9% reduction, respective for the 9th, 10th, 11th and 12th Five-­
Year Plans, indi­
cating goal attainment in all but the 10th Five-­
Year Plan (National Statistics Bureau 
and Ministry of Ecology and Environment, 2019). This chapter specifically analyzes 
the 10% reduction goal of SO2 emissions in the 11th Five-­
Year Plan as it reversed 
the humiliating failure in the 10th Five-­
Year Plan. The national quantitative goal 
was centrally set up to involve the Chinese top leadership and the then State Envi­
ronmental Protection Administration (SEPA, presently the Ministry of Ecology and 
Environment). The mitigation tasks were distributed to provincial and other local 
governments with their individual goals. Mechanisms were put into place to moni­
tor the goal compliance statuses of local governments and take enforcement actions 
for their cooperation. Goals have also been rapidly evolving to reflect the status and 
intended emphasis of SO2 mitigation and air pollution control.
2  
Centralized goal setting
2.1  Setting up the national goal
China’s goal process involves three overlapping cycles: Five-­
Year Plans, National 
Party’s Congresses and National People’s Congresses. The 11th Five-­
Year Plan for­
mally started in 2006 and concluded in 2010. The 16th National Party’s Congress 
lasted from October 2002 to October 2007. The 10th National People’s Congress 
lagged half a year behind, from March 2003 to March 2008. The 11th Five-­
Year Plan 
did not begin until the middle of the two Congresses. Under China’s present political 
reality, the two Congresses have a reasonable sequence. The National Party’s Con­
gress selects party leaders. After a further distribution of power, these leaders assume 
various governmental jobs in the following National People’s Congress. The first 
gatherings of these two Congresses are mainly about determining the leadership of 
the party and the country. Then China’s leaders reshuffle every five years. As a result, 
the three cycles are actually two: the Five-­
Year Plans and the change of leadership.
The cycles have existed in the present form for about four decades, espe­
cially since 1992. The most stable cycle is the Five-­
Year Plan. All Five-­
Year 
Plans are targeted for five years, even in the most irrational period of the Cul­
tural Revolution. Since the 3rd Five-­
Year Plan (1966–1970), the period has been 
consecutive. The National Party’s Congress formed its own five-­
year cycle in 
1977, and the National People’s Congress, in 1978. But the leadership change did 
not match the Congresses’ cycles until 14 years later. Jiang Zemin was formally 
elected as the secretary general of the party in 1992 and the president of China in 
1993. Since then, China’s top leaders also have established their five-­
year cycles, 
formally synchronized with the Congresses.


Mobilizing the government  45
The cycles of Five-­
Year Plans do not match China’s change of leadership. The 
anchor year of a Five-­
Year Plan is the previous year before the plan starts. But 
because the plan has to be formed before all information in the anchor year is 
known and China’s SO2 emissions are very volatile, relative goals are much bet­
ter than absolute goals to address the huge uncertainty. China’s failure to attain 
the 10% reduction goal of SO2 emissions in the 10th Five-­
Year Plan (2001–2005) 
may partly reflect the mismatch among cycles. A new administration took full 
charge in March 2003 when the 10th Five-­
Year Plan had been going on for over 
two years. Almost immediately afterward, China’s SO2 emissions went out of con­
trol. During 2001–2002, SO2 emissions went down by 3.4%, but in the remaining 
three years (2003–2005), the emissions surged by 32.3% (SEPA, 2001–2009). 
On the other hand, the sharp contrast was not obvious from the perspective of 
economic growth. In annual terms, China’s economy expanded at an annual rate 
of 8.7% in the first two years and 10.2% later (National Bureau of Statistics of 
China, 1999). Although the surge could be simply a coincidence with the change 
of leadership, if 2003 through 2005 had been under the same administration as in 
2001–2002, the result might be different due to a better unification of planning 
and implementation.
The Outline of the National 11th Five-­
Year Plan on Economic and Social 
Development (hereafter referred to as the Outline) was the title of an official 
document ratified by the National People’s Congress, the nominally highest 
authority in China, in March  2006 (National People’s Congress, 2006). The 
10% reduction goal of SO2 emissions was clearly included to be legally bind­
ing. The process to reach the Outline can be divided into three periods: (1) mid-­
2003 to December 2004, concluded with the formation of The Basic Thoughts 
of the National 11th Five-­
Year Plan (hereafter referred to as the Basic Thoughts; 
National Development and Reform Commissions, or NDRC, was responsible); 
(2) February 2005 to October 2005, ended with the ratification of The Sugges­
tions on Designing the National 11th Five-­
Year Plan (hereafter referred to as 
the Suggestions; the Central Committee of the Chinese Communist Party was 
in charge); (3) October 2005 to March 2006, indicated by the enactment of the 
Outline (State Council took the hold).
The Basic Thoughts contemplated the strategic direction of the Outline. This 
idea-­
framing period was initiated in mid-­
2003 and completed by the end of 2004 
(Xinhua News Agency, 2006; NDRC, 2003). For environmental protection, the 
job of the 11th Five-­
Year Plan was to “decelerate the trend of ecological and envi­
ronmental deterioration and strengthen the ability of sustainable development” 
(NDRC, 2005). The wording clearly differs from, for example, “improving envi­
ronmental quality.” It may be reflected later in the Basic Thoughts on Environ­
mental Protection with a flat SO2 emission goal proposed (SEPA, 2006d; Chinese 
Academy for Environmental Planning [CAEP], 2004).
The then named SEPA was responsible for writing the 11th Five-­
Year Plan for 
Environmental Protection. The SEPA understood the specific difficulty of control­
ling SO2 emissions. For example, in 2002, Wang Xinfang, a deputy administrator 
of the SEPA, admitted that it was hard to achieve the 10% reduction goal of SO2 


46  Mobilizing the government
emissions in the 10th Five-­
Year Plan (2001–2005; Wang, 2002). The final result in 
2005 confirmed his concern: goals on other pollutants were either met or slightly 
missed, but SO2 emissions were 27.8% higher than the level in 2000 and 42% higher 
than the original goal (Zou et al., 2006). The SEPA distributed The Basic Thoughts 
on Environmental Protection on December 23, 2004, and proposed a flat goal for 
the 11th Five-­
Year Plan (SEPA, 2006d; CAEP, 2004). The midterm assessment on 
the 10th Five-­
Year Plan that was completed in 2004 could have played a guiding 
role in the proposal: the available data showed an 8.2% increase of SO2 emissions 
in 2003 compared with those in 2000 (SEPA, 2001–2009). The midterm assessment 
believed that the 10% reduction goal had fallen out of reach but still expected that 
SO2 emissions in 2005 could remain the same as the level in 2000 (Zou et al., 2004).
With the tentative Basic Thoughts, the top leadership in the Central Committee of 
the Chinese Communist Party got directly involved. The period was formally initi­
ated with the establishment of a high-­
profile drafting team on February 16, 2005, 
headed directly by Premier Wen Jiabao (Xinhua News Agency, 2005). A prominent 
feature is the multiple meetings presided by President Hu Jintao in the Political 
Bureau or its Standing Committee and by Premier Wen Jiabao in the drafting team 
(Xinhua News Agency, 2005). The Suggestions was finally passed and endorsed 
on October 11, 2005, by the Central Committee of the Chinese Communist Party 
(Xinhua News Agency, 2005). Sharply different from the Basic Thoughts, the Sug­
gestions clearly declared to “reduce total emissions of pollutants,” which essentially 
indicated a goal of improving environmental quality (Xinhua News Agency, 2005).
After the Suggestions tightened the goal for environmental protection in Octo­
ber 2005, the third period started with the establishment of a drafting team that 
comprised various ministries in the central government (Xinhua News Agency, 
2006). An expert committee was summoned to comment on the drafts of the Out­
line (Ma, 2005). The public was also consulted for advice (Ma, 2005). Presi­
dent Hu Jintao and Premier Wen Jiabao organized several meetings to discuss the 
drafts (Xinhua News Agency, 2006). In November 2005, the SEPA drafted a plan 
on acid rain and SO2 emission control (SEPA, 2005). Although SO2 emissions in 
2004 had been 13% higher than the 2000 level, the 10% reduction goal for the 
11th Five-­
Year Plan first appeared (SEPA, 2005, 2001–2009). On December 3, 
2005, State Council enacted Decisions on Realizing Scientific View of Develop­
ment and Strengthening Environmental Protection (State Council, 2005a), which 
linked the new ideology of Scientific View of Development with environmental 
protection. It confirmed the importance of environmental protection in the estab­
lishment of the new ideology. When the 4th Conference of the 10th National Peo­
ple’s Congress was in session, the Outline was submitted on March 5, 2006, and 
approved on March 14, 2006 (Xinhua News Agency, 2006).
2.2  
Methods of goal setting
A Five-­
Year Plan anchors at the previous year of its planning period. For exam­
ple, a goal in the 11th Five-­
Year Plan (2006–2010) is to compare 2010 with 
2005. In practice, the anchor year’s data cannot be fully utilized in setting up 


Mobilizing the government  47
the goals. China generally published environmental data for the previous year in 
around June (SEPA, 2001–2009). Although the public may get the information 
later than the Chinese government, several months could elapse for the collection 
and compilation of data. Accordingly, the anchor year’s information cannot be 
fully employed in planning but has to be the foundation for the next Five-­
Year 
Plan. China’s annual change of SO2 emissions varied greatly: the 2004 emissions 
were 4.5% up from the 2003 level, but the figure surprisingly jumped 13.1% 
in 2005 (SEPA, 2001–2009). At the same time, however, the economic growth 
rates were quite stable with 10.1% in 2004 and 11.4% in 2005 (National Bureau 
of Statistics, 2019). Because of the substantial volatility, the absence of data in 
the most relevant and important anchor year could cause significant trouble in 
calibrating goals.
Two components were important in setting up China’s SO2 emission goals in 
Five-­
Year Plans: long-­
term goals and appropriate mitigation paces. China relied 
on a concept called “environmental capacity” to decide long-­
term SO2 emission 
goals (Yang et al., 1998, 1999). “Environmental capacity” refers to the upper-­
limit 
emissions of a pollutant without degrading a kind of environmental quality below 
a minimum level. The environmental capacity for SO2 emissions is a function of 
three variables: (1) the amount and distribution of SO2 emissions, or emission 
inventories; (2) the transport and sinks of SO2; and (3) an acceptable level of some 
environmental quality. The second variable is largely determined by atmospheric 
circulation and chemistry. The third variable was used as an external choice. If 
society would like to live in a better environment, the limit of ambient SO2 con­
centration could be lowered and SO2 emissions have to be further reduced.
To set up an SO2 goal in a Five-­
Year Plan, China first decided on a long-­
term 
goal and then found an appropriate mitigation pace to attain the goal. The long-­
term goals were determined with models of atmospheric transport and chemistry. 
The implicit long-­
term goal for the 10th Five-­
Year Plan (2001–2005) was 12 mil­
lion tons and was scheduled to get attained in 2020 (Wang, 2002). For the 11th 
Five-­
Year Plan (2006–2010), the long-­
term goal became 18 million tons and the 
goal attainment year would also be 2020 (SEPA, 2005). Although both goals were 
supported by scientific research with different constraint conditions, the signifi­
cant upward revision of the long-­
term goal probably arose as a result of the sharp 
increase in coal use that led to an unanticipated rise of SO2 emissions in the 10th 
Five-­
Year Plan.
The long-­
term goals have certain scientific foundations. China’s Law of Envi­
ronmental Protection clearly holds local governments responsible for local envi­
ronmental quality (National People’s Congress, 1989). Because ambient air quality 
standards are also “mandatory standards” in the Law of Standardization (State 
Council, 1990), local government leaders should be mobilized to enforce SO2 miti­
gation policies if the law were well respected. In 1996, the then State Environmen­
tal Protection Agency enacted ambient air quality standards (NEPA and SBTS, 
1996). Most of China’s land area with economic and human activities should 
have ambient SO2 concentration in annual mean below 0.060 mg/m3. One key 
study showed that only to achieve this average concentration within grid boxes of 



48  Mobilizing the government
0.2° × 0.2°, China has to control its SO2 emissions at 12 million tons (Yang et al., 
1999). Another study for the 11th Five-­
Year Plan selected critical acid deposition 
within grid boxes of 1° × 1° (Zou et al., 2006). Although the number was based 
on several heavy assumptions (most important, the geographical distribution of 
SO2 emission sources), it signaled the stringency of the ambient SO2 concentra­
tion standard. For example, China’s goal in the 11th Five-­
Year Plan was to reduce 
SO2 emissions from 25.5 million tons in 2005 by 10% in 2010, still far above the 
12-­
million-­
ton level (National People’s Congress, 2006).
The distribution of SO2 emissions matters greatly for any national SO2 miti­
gation goal that is based on SO2 concentration. For example, with SO2 concen­
tration of 0.060 mg/m3 as the constraint condition, Shanghai could emit up to 
0.63 million tons of SO2 (Yang et al., 1999), but its actual emissions in 2007 were 
0.50 million tons (Ministry of Environmental Protection, 2008). Then if a pollut­
ing source was located in Shanghai, it would have no necessity to mitigate. But if 
the same source were moved to Jiangsu, a neighboring province with its emission 
limit below actual emissions (Ministry of Environmental Protection, 2008; Yang 
et al., 1999), it would be subject to serious abatement. The 1998 study revealed a 
goal based on SO2 ambient concentration: if not counting the excess environmen­
tal capacity in Tibet compared with its emissions (0.50 million tons vs. 1.5 thou­
sand tons), China’s national goal was to reduce SO2 emissions to about 12 million 
tons (Yang et al., 1999). China planned to attain the goal in 2020 (Wang, 2002). 
The goal for the 10th Five-­
Year Plan was then established as a 10% reduction, or 
18 million tons (SEPA, 2001).
After the big failure in the 10th Five-­
Year Plan on SO2 mitigation, China still 
held 2020 as the attainment year of a long-­
term goal. However, the original goal 
would be too difficult. In 2005, China emitted 25.5 million tons of SO2 (SEPA, 
2001–2009). To achieve the goal of 12 million tons in 2020, a 53% reduction in 
15 years would be required. Even if from the 2004 level when a new goal for the 
11th Five-­
Year Plan was formed, the reduction rate should still be 47% (SEPA, 
2001–2009). By replacing the constraints of SO2 concentration with critical acid 
deposition, a new environmental capacity was worked out to be 17.3 million tons 
(Zou et al., 2006). Then 18 million tons were chosen to be the new long-­
term 
goal (SEPA, 2005). These two long-­
term goals assumed a similar pace of about 2 
to 2.5 million tons reduction per five years. Because of the relatively stable pace 
and a common attainment year of the long-­
term goals, China’s long-­
term goals 
seemed to be reversely decided from current emission levels. Interestingly, both 
long-­
term goals were supported by scientific research. The history could indicate 
that the results of the scientific research were selected beforehand by nonscientific 
factors.
In deciding goals for Five-­
Year Plans, the emission trends in previous years 
were also considered (Wang et al., 2004). Because the 9th Five-­
Year Plan achieved 
a 15.8% reduction (NEPA et al., 1996; SEPA, 2001–2009), even a similar trend 
was thought to be too stringent (Wang et al., 2004). Probably the 10% reduction 
goal was established because it stood between the 15.8% reduction and the origi­
nal goal of a 3.8% increase in the 9th Five-­
Year Plan. A middle ground, closer to 


Mobilizing the government  49
the stringent end, was taken. On the other hand, the same historical trend would 
be too relaxed for the 11th Five-­
Year Plan. SO2 emissions went up by 27.8% in 
the 10th Five-­
Year Plan (SEPA, 2001–2009). Certainly this was not an acceptable 
trend, but it might be an important factor that drove the initial flat goal for the 
11th Five-­
Year Plan (CAEP, 2004). The same principle could have been followed: 
0% change was closer to the stringent end between a 27.8% increase and a 10% 
reduction. As a result, the goal attainment in the previous Five-­
Year Plan should 
have played an important role in framing a goal for the next.
The United States’ goal of SO2 emissions in Clean Air Act Amendments 
(CAAA; 1990) was also expressed in relative terms. Relative to the emission 
level in the anchor year of 1980, SO2 emissions were planned for reduction by 
10 million tons (The U.S. Congress, 1990). Although an intensive 10-­
year study 
was performed in the 1980s (National Acid Precipitation Assessment Program), 
it failed to answer relevant questions for policy making and was not closely con­
nected to the goal-­
setting process (Roberts, 1991; Pouyat and McGlinch, 1998). 
For a fixed long-­
term goal, different anchor years only correspond to different rel­
ative reductions or different expressions of the figures. Furthermore, 1980 was not 
a baseline year for allowance allocation. Rather, 1985 was a much more impor­
tant year with real implications in grandfathering emission permits. However, if 
the 10-­
million-­
ton reduction was fixed, the choice of 1980 did have important 
implications. In 1980, the U.S. emitted 23.5 million tons of SO2 and the figures in 
1985 and 1990 were, respectively, 21.1 and 20.9 million tons (U.S. Environmental 
Protection Agency, 2007). Essentially, the choice of 1985 and 1990 would have 
no difference. But anchoring in 1980 could effectively relax the long-­
term goal by 
about 2.4 to 2.6 million tons. The goal was planned for attainment in 2010. The 
anchor year 1980 was ten years ahead of the legislation and 15 years before the 
program formally started in 1995. Although whether a goal was expressed in rela­
tive or absolute terms matters greatly in China, it was generally not quite relevant 
for the United States’ goal setting. The United States had much less volatility in 
annual SO2 emissions. The burden to achieve the goal – the difference between 
business-­
as-­
usual emissions and the goal – was accordingly much less uncertain 
than China’s. The major benefit of relative terms was to reduce the uncertainty of 
surprising emission growth or reduction. However, less uncertainty in the United 
States and the longer goal cycle did not distinguish this benefit. Furthermore, the 
Acid Rain Program’s goal cycle was much longer than China’s Five-­
Year Plans. 
Because of the well-­
established rule of law, the law ensured that the SO2 mitiga­
tion efforts would continue regardless of who was the president or which political 
party he or she belonged to.
3  
Top-­
down goal distribution
Goal implementation refers to a process for goal implementers to receive, accept 
and work for goal attainment. It is quite different from policy implementation. 
Goal implementation deals with the relationship among different governments or 
their agencies, while policy implementation focuses on the relationship between 


50  Mobilizing the government
the government and polluters, including industrial plants and individuals. Goal 
setters and goal implementers are usually separate in the Chinese government. 
Since goal setters are not directly in charge of achieving the goal, they have 
to find a way to get goal implementers to accept the goal and to work hard for 
it. In order for effective goal implementation, subgoals should be created from 
the national goal to demand an appropriate distribution scheme. The UNFCCC 
defines a principle of sharing the duty of reducing greenhouse gas emissions 
among countries according to “common but differentiated responsibilities and 
respective capabilities” (United Nations, 1992). Which applicable principles 
should be followed has attracted negotiation debates and academic studies 
(Chakravarty et al., 2009; Li, 2010).
A national goal and its distribution to local governments often fall into separate 
decision-­
making processes in the Chinese setting. Taking the SO2 goal in the 11th 
Five-­
Year Plan (2006–2010) as an example, the national 10% reduction goal was 
largely decided by the top leadership of the party, but provincial goals came from 
a bargaining process between the central government – mainly the then SEPA – 
and provincial governments.
3.1  
Goal distribution from the central to provincial governments
Chinese local governments are divided into several levels, mainly provinces, 
municipalities and counties. To implement SO2 emission goals, the central gov­
ernment distributed subgoals to provincial and local governments and issued 
incentives to mobilize their leaders. A good national goal is hard to implement 
without a fair distribution of the burden. After a national goal is framed, provinces 
will negotiate with the central government for their shares of the burden. The 
details of the negotiation and their applied principles are not publicly available but 
could be reversely examined from the outcome.
China qualitatively disclosed principles to distribute the national goal to 
31 provinces. Key influential factors included environmental quality, environmen­
tal capacity, current emission level, economic development status, SO2 mitigation 
capability, requirements of various pollution control plans and regional category 
(west, middle, east; State Council, 2006). An explicit formula was less likely to 
exist that connected these factors with a province’s goal. However, published pro­
vincial information could at least lead to an evaluation of potentially quantitative 
relationships. Econometric analysis was applied here with a linear assumption.
The dependent variable was provincial SO2 emission goals in percentage 
terms: SO2 emission target in 2010 / SO2 emissions in 2005 – 100%. The dis­
tributed national goal, 11.9% reduction, was actually a little more stringent than 
a 10% reduction (State Council, 2006). All provinces combined could only emit 
22.47 million tons, not 22.94 million tons for the nation. The difference (0.47 mil­
lion tons) was reserved for experimenting with SO2 emission cap-­
and-­
trade (State 
Council, 2006).
Independent variables included all those factors indicated by the Chinese govern­
ment (State Council, 2006). Because 2005 was the anchor year of the 11th Five-­
Year 


Mobilizing the government  51
Plan, independent variables all referred to this year unless otherwise specified. Envi­
ronmental quality was represented by both the annually average SO2 concentra­
tion in provincial capitals and nonpower sectors’ emission density (expressed in 
tons/km2). The capitals’ SO2 concentration data were published in China Statistical 
Yearbooks (National Bureau of Statistics, 2006). In addition, China divided SO2 
emissions into two big categories: power and nonpower. Associated with shorter 
chimneys, non-­
power-­
sector emissions were believed to be more closely associ­
ated with local air quality (SEPA, 2006a). Their emission densities in provinces 
were employed to represent another perspective of environmental quality (National 
Bureau of Statistics of China, 1999; Zou et al., 2006). Environmental capacity is a 
term indicating allowed maximum emissions to maintain a certain environmental 
quality. The data used in this section came from a study that calculated long-­
term 
SO2 goals for the 11th Five-­
Year Plan (Zou et al., 2006). Critical acid deposition 
was the targeted environmental quality. The corresponding upper-­
limit national 
emissions were 17.3 million tons, and each province had its own figure (Zou et al., 
2006). Current emission levels were represented by provincial SO2 emissions in 
2005. Provincial goals were formally distributed in August 2006 (State Council, 
2006). Because data for 2005 had been published in June 2006 (SEPA, 2001–2009), 
they should be available for negotiating the goal distribution. Provincial GDP per 
capita stood for economic development status (National Bureau of Statistics, 2006).
No definition had been clearly displayed by the Chinese authorities on SO2 
mitigation capability. Two variables were used. First, higher provincial SO2 
removal rates in 2005 could indicate fewer opportunities for the future. From 
another aspect, they also represented previous efforts in SO2 mitigation. Second, 
SO2 scrubbers (or flue-gas desulfurization facilities, FGD) had been designated as 
a key measure to reduce SO2 emissions in the 11th Five-­
Year Plan (State Council, 
2007a). The power sector’s shares of total emissions would then serve as another 
indicator of mitigation capability (National Bureau of Statistics of China, 1999; 
Zou et al., 2006). Higher shares may lead to a more effective reduction of total 
SO2 emissions through SO2 scrubbers.
China’s policies and emission control plans targeting individual emission 
sources could decide provincial goals in a bottom-­
up way. Nevertheless, it may 
not coincide with the top-­
down results. For example, effluent emission standards 
and SO2 scrubber planning, respectively, were expected to lead to national power 
sector’s emissions of 8.9 and 9.7 million tons in 2010, while the finally assigned 
goal was 9.5 million tons in the 11th Five-­
Year Plan (Zou et al., 2006). To evalu­
ate their impact on goal distribution, two independent variables were generated 
for each province: (1) (Emission standard-­
designated levels in the power sector 
in 2010 + Nonpower emission goals in 2010) / Provincial emissions in 2005 – 
100% (State Council, 2006; Zou et al., 2006; National Bureau of Statistics of 
China, 1999) and (2) (Scrubber planning-­
projected emissions in power sector + 
Nonpower emission goals in 2010) / Provincial emissions in 2005 – 100% (State 
Council, 2006; Zou et al., 2006; National Bureau of Statistics of China, 1999).
According to geographical locations and economic advancement, China 
divides its provinces into three regional groups: west, center and east. To alleviate 


52  Mobilizing the government
regional disparity in economic growth and income, China treats the three cat­
egories differently. For example, “Great West Development” aimed to develop 
western provinces, particularly through building infrastructure. Dummy variables 
were generated to indicate a province’s location. In addition, because China’s 
prevalent wind generally transports air pollutants from the west to the east, SO2 
emissions in western provinces could cause more damage than those in eastern 
provinces. The dummy variables then evaluated the overall impacts of these two 
opposite concerns.
Besides these variables, several others that were not mentioned in the official 
distribution plan were also tested, including SO2 emissions per capita, goal attain­
ment in the 10th Five-­
Year Plan and electricity export. One argument for China 
not to accept a legally binding goal on carbon mitigation in the Kyoto Protocol 
was its low carbon emissions per capita. Whether China applied this principle in 
domestic practice was examined through provincial SO2 emissions per capita in 
2005 (National Bureau of Statistics of China, 1999).
China failed substantially to achieve its 10% reduction goal of SO2 emis­
sions in the 10th Five-­
Year Plan (2001–2005): the actual emissions in 2005 
were 42% higher than the original goal (SEPA, 2001–2009, 2001). But some 
provinces did better than others. Whether better performance in the past was 
recognized is tested through a ratio: Provincial emissions in 2005 / Provincial 
emission targets in the 10th Five-­
Year Plan for 2005 (State Council, 2006; 
National Bureau of Statistics of China, 1999). In addition, for the 27 prov­
inces used in models (discussed later), this variable was highly correlated with 
the provincial growth rates of SO2 emissions in the 10th Five-­
Year Plan and 
the correlation coefficient is 0.98. Accordingly, the model results on this goal 
attainment variable could be almost identically applied to a variable on the 
growth rates.
Pollutant emissions and product consumption are not necessarily in the same 
location. Electricity is a clear and important case. SO2 comes out of coal-­
fired 
power plants, but electricity could be lighting bulbs in another province. This 
effect was examined through provincial electricity trade: Provincial electricity 
generation / Provincial electricity consumption – 100% (National Bureau of Sta­
tistics, 1997–2008).
Although mainland China has 31 provinces, only 27 were used for the statisti­
cal models. Four provinces were kept out. Hainan and Tibet had too-­
insignificant 
SO2 emissions in 2005, respectively, 22,000 and 2,000 tons. Qinghai had the least 
emissions among provinces except the two previously mentioned, and its data on 
avoided industrial emissions were not available in China Statistical Yearbooks. 
Shanghai had its nonpower SO2 emission density in 2005 much higher than 
other provinces (32.7 tons/km2; the next highest was 7.9 tons/km2), a far outlier 
(National Bureau of Statistics of China, 1999; Zou et al., 2006).
The correlation coefficients between the variables are given in Table 4.1. Pro­
vincial goals were highly correlated negatively with nonpower emission density, 
total SO2 emissions and GDP per capita – indicating that higher levels of these 
variables were closely associated with more stringent provincial goals  – and 


Mobilizing the government 53
Electricity
export
in 2005
1.00
Goal
attainment
Five−
in the
10th 
Year Plan
1.00
0.31
 
 
2
emission
 
SO
per
capita
in 2005
1.00
0.17
0.35
est
1.00
0.25
0.06
W
−0.17
Middle
1.00
−0.46
−0.03
0.42
0.52
 
Emission
 
d
 
standar
decided
goals
1.00
−0.08
−0.54
−0.44
−0.07
−0.44
Scrubber
planning
 
decided
goals
1.00
0.02
0.16
0.21
−0.03
0.32
0.10
 
s
’
Power
emission
e
0.30
shar
in 2005
1.00
−0.23
0.10
0.21
−0.29
0.18
0.28
Tibet, and Shanghai.
2
emoval
Rate
in 2005
1.00
−0.22
−0.03
0.10
0.05
SO
r
−0.08
−0.40
−0.14
−0.05
Correlation coefficients of key factors for 27 provinces
GDP
per capita
in 2005
1.00
−0.13
0.25
−0.33
0.75
−0.28
−0.46
−0.14
−0.22
−0.54
otal
in 2005
1.00
−0.16
−0.16
0.05
−0.39
−0.41
0.01
−0.17
0.00
0.21
T
emissions
0.24
 
Long−
term
1.00
0.15
0.14
goal
−0.30
−0.08
−0.35
0.31
−0.19
−0.18
−0.01
−0.05
−0.02
Nonpower
 
emission
density
in 2005
1.00
−0.04
0.22
0.59
−0.01
0.00
−0.36
0.29
−0.32
−0.26
0.03
−0.60
−0.47
s
Capital’
 conc.
2
1.00
0.17
0.05
0.16
SO
In 2005
−0.04
−0.12
−0.34
1
0.1
−0.17
−0.23
0.39
0.24
−0.30
−0.05
Reduction
1.00
0.00
goal
−0.74
−0.06
−0.53
−0.48
0.18
−0.10
0.63
−0.06
0.35
0.24
0.01
0.44
0.40
Reduction goal
 
 
able 4.1(a)
s SO2
concentration
s mainland has 31 provincial regions. Four are not included here: Qinghai, Hainan, 
 
 
 
emission
 
Capital’
Nonpower
density
Long−term goal
 
T
GDP/capita
 
Total emissions
Removal rate
 
s
’
emission
 
Power
share
Scrubber
planning
Emission
standard
Middle
est
Emission/capita
Electricity
export
W
Goal attainment
Note: China’


54  Mobilizing the government
0.0
0.12
7.9
3.7
200.2
4.5
0.6
0.7
0.1
0.0
1
1
61.0
1.3
0.6
Max
−20.4
0.02
0.2
−0.3
19.0
0.5
0.1
0.3
−0.3
−0.2
0
0
9.3
0.0
−0.6
Min
.
Dev
Std. 
5.7
0.020
2.09
1.02
48.39
0.91
0.16
1
1
0.1
0.1
0.09
0.48
0.47
13.26
0.34
0.24
0.057
−10.1
2.97
0.81
91.97
1.55
0.28
0.52
−0.09
−0.13
0.33
0.30
22.95
0.51
0.03
Mean
No. of 
observations
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
ear Plan
ear Plan
ear Plan
ear Plan
ear Plan
Period
11th Five-Y
2005
2005
1th Five-Y
1
2005
2005
2005
2005
1th Five-Y
1
11th Five-Y
2005
10th Five-Y
2005
Unit
%
ton/km2
Summary of variables
mg/m3
%
10,000 tons
10,000 RMB/person
%
%
%
%
dummy
dummy
kg/person
%
%
 conc.
 
s emission share
Table 4.1(b)
ariables
V
Reduction goal
2
s SO
Capital’
Nonpower emission density
long-term goal
’
Total emissions
GDP/capita
Removal rate
Power
Scrubber planning
Emission standard
Middle
est
W
Emission/capita
goal attainment
Electricity export


Mobilizing the government  55
positively with SO2 emissions from scrubber planning, goal attainment in the 10th 
Five-­
Year Plan and electricity export.
Although nonlinear terms could show consistent significance in models, such 
as the squared term of nonpower SO2 emissions density, its actual application in 
the negotiation for provincial goals was difficult. China very likely did not use a 
written formula to decide provincial goals. The nonlinear relationship was thus 
too complicated for arguments, especially those with a turning point. In addition, 
once included, several far points could greatly change the overall relationship 
in models. For example, if Shanghai appeared in the models, its big nonpower 
SO2 emission density would make the corresponding coefficient much different. 
These provinces might experience special negotiation. As a result, only linear 
terms were used in the models to examine China’s principles in goal distribution.
Another decision about the regression models was whether a constant vari­
able should be included. If all provinces had to presume a basic reduction goal 
and adjust it according to specific situations, the constant variable would show 
significance and the explained variance, R2, should be higher compared with a 
no-­
constant model. Model runs indicated otherwise (Table 4.2). In response, no 
constant variable appeared in the remaining models.
The model results showed that two variables were the most important in dis­
tributing the national goal to provinces. First, richer provinces tended to receive 
more stringent reduction goals. Provincial GDP per capita in 2005 and provincial 
goals in the 11th Five-­
Year Plan had a correlation coefficient of −0.48 (Table 4.1). 
But statistical models did not consistently show the significance of GDP per capita 
(Table 4.2). However, if either nonpower emissions density in 2005 or provincial 
goals from scrubber planning were excluded, GDP per capita would become sig­
nificant. For every 10,000 RMB/person increase, the province should reduce its 
SO2 emissions by further 1.3% (from the 2005 level). In explaining the model 
results, a problem was that two factors had a high correlation, and both showed 
significance on some occasions. However, it should not have mattered much in the 
negotiation. As long as no clear formula decided goals, a province could always 
argue with one factor to generate a more favorable goal. For example, Shanghai 
had a much higher nonpower SO2 emission density in 2005 than other provinces, 
but its GDP per capita in 2005 was ahead, with a significantly narrower margin 
(52,000 RMB/person compared with the next highest 45,000; National Bureau of 
Statistics of China, 1999; Zou et al., 2006). Comparatively, Shanghai could ask 
for a less stringent goal from GDP-­
per-­
capita point of view. Second, provinces 
with large emissions had tougher goals. China’s big provinces experienced greater 
pressure to reduce their emissions more for achieving the national goal. Coeffi­
cients of provincial emissions in 2005 were consistently significant (Table 4.2). 
Every 100,000 tons more SO2 emissions corresponded to about 0.47% further 
reduction. Third, nonpower emissions density displayed consistent significance. 
For emitting one more ton per square kilometer, a province should further reduce 
total SO2 emission by about 1.3%.
Notably, several other variables did not show much influence. First, provinces 
with worse environmental quality might not have received more stringent goals. 


56  Mobilizing the government
***
Model 9
***
−0.058
**
**
 
 
 
−2.79
 
 
10.09
12.71
 
 
 
3.62
6.33
 
0.94
Model 8
***
**
 
−0.033
**
−1.29
 
−0.97
 
 
15.39
3.33
 
 
 
 
 
 
0.94
***
Model 7
***
 
−1.34
*
 
−0.047
−1.27
 
 
 
 
 
 
 
 
 
 
0.93
***
Model 6
*
 
−0.83
−0.066
*
−1.73
 
 
 
 
 
 
 
 
1.98
4.48
 
0.93
Model 5
**
*
 
−0.042
*
 
−1.00
−1.47
 
 
12.98
12.55
 
 
0.05
0.73
4.43
 
0.94
Regression model results for distributing the national goal to provinces
***
Model 4
−0.80
−0.43
−0.064
*
−1.87
 
 
 
 
 
 
 
0.01
2.16
3.96
 
0.93
Model 3
***
*
**
47.28
−1.22
−0.99
−0.038
−2.25
2.40
1.43
15.35
9.83
0.36
−0.99
 
 
 
 
0.94
Model 2
*
−1.14*
40.03
−0.86
−0.041
−1.95
*
4.48
−2.42
14.10
15.38
−0.13
−0.90
0.07
0.58
3.69
 
0.95
Model 1
*
*
39.51
−1.14
−0.88
−0.041
−1.98
4.36
−2.63
14.08
15.47
−0.18
−0.95
0.07
0.57
3.74
0.31
0.76
density
2
 
 per capita
T
Independent variables
 concentration
2
s SO
s emission share
R
able 4.2
Capital’
Nonpower emission 
Long-term goal
Removal rate
Power
Emission standard
Middle
est
Total emissions
’
GDP
Scrubber planning
W
Emission/capita
Goal attainment
Electricity export
Constant
Adjusted 
* Significant at 10%. ** Significant at 5%. *** Significant at 1%.


Mobilizing the government  57
Provincial capital cities’ SO2 concentration did not significantly affect provincial 
goals (Table 4.2). But in most provinces, capital cities only occupy a fraction 
of the total land area and thus could not represent the general picture. Another 
problem with this variable was its coefficient’s sign. Intuitively, the sign should 
be negative – dirtier air needs more reduction of pollutant emissions. The actual 
coefficient, although not significant, was consistently positive (Models 1–3 in 
Table 4.2). To avoid its impact, the variable was excluded from other models. 
Second, provinces with higher emissions per capita did not face deeper reduc­
tions. Emissions per capita did not have any significant relationship with pro­
vincial reduction goals. Third, earlier efforts on SO2 emission control were not 
awarded later with relaxed goals. Neither of the two relevant variables – SO2 
removal rates in 2005 and goal attainment in the 10th Five-­
Year Plan – showed 
any consistent significance. Earlier efforts did not make the future easier in SO2 
emission control, while no failure in the past would get punished through adding 
future burden. Because of the very high correlation between the goal attainment 
variable and provincial growth rates of SO2 emissions in the 10th Five-­
Year Plan, 
the model results also indicated that faster emission growth did not have a sig­
nificant impact on provincial goals. For China’s political reality, this result was 
reasonable. Provincial and other local leaders often rotate every five years. If one 
administration was irresponsible, its failure did not get the next administration 
punished. Similarly, a performing administration should not reduce pressure on 
future leaders. Fourth, more electricity net export consistently led to less strin­
gent goals, but the relationship was not statistically significant. It seemed that 
China did not take serious consideration of the disintegration between emissions 
and consumption in distributing environmental goals. Fifth, no influence was 
found solely due to the location of a province. Regional characteristics should 
have been absorbed into other variables. For example, long-­
term goals already 
considered prevalent wind and more damage from western SO2 emissions. West­
ern and central provinces were poorer than eastern ones, which was reflected in 
GDP per capita.
Three principles were distinguished for distributing the national SO2 emission 
goal in the 11th Five-­
Year Plan: those provinces with heavier pollution, bigger 
total emissions and richer GDP per capita should reduce more. The second prin­
ciple was the most consistently applied. An explicit formula of deciding a provin­
cial goal could be written as
Provincial Goal (−0 to −100) = −1.34 × Nonpower emission density 
(tons/km2) – 0.047 × Total emissions (10,000 tons) − 1.27 × GDP 
per capita (10,000 RMB/person).
The 27 provinces had an arithmetic average goal in the 11th Five-­
Year 
Plan of −10.1%. The formula would lead to −10.2%: GDP per capita, −2.0%; 
nonpower SO2 emissions density, −4.0%; and total emissions, −4.3%. The 
explanatory power was high, with adjusted R2 generally over 0.93 (Model 7 in 
Table 4.2).


58  Mobilizing the government
3.2  
Goal distribution from provincial to municipality governments
The SEPA issued guidance for distributing SO2 emission goals from one govern­
ment level to its subordinate level (SEPA, 2006a). The total emissions are dis­
tinguished into the power sector (capacity no less than 6 MW) and nonpower 
sectors (SEPA, 2006a). The SO2 emission quota was generally assigned to each 
fossil-­
fuel power plant according to provincially homogeneous emission inten­
sity (grams SO2/kWh, varying with plant ages; SEPA, 2006a). As shown in Fig­
ure 4.1, the designated emission intensity was more stringent in new coal power 
plants and those in eastern or richer provinces. From provinces to municipalities, 
polluting sources in nonpower sectors received their upper limits on the basis 
of achieving local air quality – particularly SO2 emissions concentration with a 
threshold of 0.060 mg/m3 (SEPA, 2006a). The guidance did not clarify everything 
for assigning goals. It left decisions to provincial governments, especially in non­
power sectors. More important, the excess emission quota of a region was allowed 
to transfer or trade across regions (SEPA, 2006a).
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
East-1
East-2
Central
Southwest
Northwest
O
S
 
d
e
t
a
n
g
i
s
e
D
2
O
S
 
s
m
a
r
g
(
 
y
t
i
s
n
e
t
n
i
 
n
o
i
s
s
i
m
e
2/kWh)
Period I
Period II
Period III
Figure 4.1  
Designated SO2 emission intensity in distributing SO2 emissions quota to coal-
fired power plants for 2010 in the 11th Five-Year Plan
Source: SEPA (2006a).
Note: Coal-fired power plants falling in Period I refer to those that went online or passed the Environ­
mental Impact Assessment reports before December 31, 1996. Period II spans from January 1, 1997, 
to December 31, 2003. And Period III is from January 1, 2004, to the present. “East-1” includes the 
provinces of Liaoning, Hebei, Shandong, Zhejiang, Fujian and Hainan. “East-2” covers Beijing, Tian­
jin, Shanghai and Jiangsu Provinces. “Central” refers to Heilongjiang, Jilin, Shanxi, Henan, Hubei, 
Hunan, Anhui and Jiangxi Provinces. “Southwest” provinces are Chongqing, Sichuan, Guizhou, Yunnan, 
Guangxi and Tibet. “Northwest” has Inner Mongolia, Shaanxi, Gansu, Ningxia, Qinghai and Xinjiang.


Mobilizing the government  59
In distributing provincial goals, official documents often did not even qualita­
tively declare what factors took effect. Furthermore, municipality data were not 
as publicly available as provincial data. Two aspects receive special attention in 
examining the provincial scheme of goal distribution: (1) whether the same prin­
ciples in the national goal distribution held and (2) whether the SEPA’s guidance 
was followed. This section looks at four provinces: Hebei, Guangdong, Jiangsu 
and Shanxi. Statistical model results are given in Table 4.3. Those models without 
significance are not shown. According to how the four provinces obey the national 
principle – rich and big provinces reduce more – a matrix is generated in Table 4.4.
Provinces have high autonomy in further allocating their goals among munici­
palities. The four provinces are distinguished with different patterns to dem­
onstrate such decentralized authority. In Hebei Province, all municipalities got 
roughly same reduction goals. The provincial goal of Hebei Province was a 15% 
reduction from 1.50 million tons in 2005 (State Council, 2006). It had 11 munici­
palities, and their SO2 emissions in 2005 ranged from 45,000 to 311,000 tons 
(Hebei Provincial Government, 2007). GDP per capita also varied, from 9,900 to 
27,900 RMB/person (Hebei Provincial Statistics Bureau, 2006). The 2005 data on 
the power sector’s share in SO2 emissions are not publicly available. Information 
from the goals for 2010 is applied instead: the share of the power sector would 
range from 13.4% to 53.4% in the plan (Hebei Provincial Government, 2007). 
However, municipality goals only varied from a 14.1% to a 15.8% reduction, 
centering on the provincial goal (Hebei Provincial Government, 2007). Statistical 
models indicated a consistently significant constant (Table 4.3). But neither GDP 
per capita nor SO2 emissions showed any impact on municipality goals. If the 
goal distribution guidance from the central government worked, a municipality 
with more SO2 emissions from the power sector should receive somewhat more 
stringent goals. But no negotiation seemed to have shaped the municipality goals. 
The provincial government, very likely its top leaders, decided that the provincial 
goal was applied to all with minor adjustments.
In Guangdong Province, higher income and more SO2 emissions led to more 
stringent goals. In the 11th Five-­
Year Plan, Guangdong Province received a 
Table 4.3 Regression model results for distributing provincial goals to municipalities
Hebei
Guangdong
Jiangsu
Shanxi
Observations
11
15
13
11
Nonpower emission density in 2005
−1.5**
GDP/capita in 2005
−7.2***
6.2***
SO2 emissions in 2005
−3.3***
−1.4**
−0.98***
Power sector’s emission share in 2005
−58.1***
_Constant
−15.1***
33.9***
21.4**
Adjusted R2
0.90
0.76
0.86
0.87
Note: Six municipalities in Guangdong province with 2005 SO2 emissions no more than 11,000 tons 
are not included in the model.
* Significant at 10%. ** Significant at 5%. *** Significant at 1%.


60  Mobilizing the government
provincial goal of a 15% reduction from 1.29 million tons in 2005 (State Coun­
cil, 2006). Among its 21 municipalities, 5 emitted less than 10,000 tons in 2005, 
and another 13, no more than 60,000 tons (Guangdong Environmental Protection 
Bureau, 2006). The biggest three emitted 475,000 tons, taking 51% of the pro­
vincial emissions that belonged to municipalities (929,000 tons; the remaining 
365,000 tons were directly claimed to the provincial level; Guangdong Environ­
mental Protection Bureau, 2006). Collectively, these three should reduce their 
SO2 emissions by 46% (Guangdong Environmental Protection Bureau, 2006). 
The other 18 municipalities were even allowed to increase their emissions by 23% 
(Guangdong Environmental Protection Bureau, 2006). Regression models could 
better distinguish influential factors. To avoid the heavy impacts of outlying data 
points, six municipalities were excluded, with total SO2 emissions in 2005 being 
no more than 11,000 tons and their goals in 2010 allowed for over 170% growth. 
Models for the rest of the 15 municipalities show significance of GDP per capita 
and SO2 emissions (Table 4.3). Nonpower SO2 emissions density was not tested 
because of data unavailability. Different from the situation for provincial goals, 
the constant variable here is significant. For every 10,000 RMB/person increase 
of GDP per capita and 10,000 tons more of SO2 emissions in 2005, a municipal­
ity goal would be, respectively, 7.2% and 3.3% more stringent (from the 2005 
level). Although the coefficients were different from those for provincial goals, 
the qualitative principles remained the same: rich and big municipalities should 
reduce more.
In Jiangsu Province, municipalities with higher emissions should reduce more, 
but richer ones were allowed to reduce less. Jiangsu Province’s goal was an 18% 
reduction from 1.23 million tons in 2005 (State Council, 2006). The 13 munici­
palities emitted from 28,000 to 243,000 tons of SO2, a much narrower but still 
larger range than in Guangdong Province (Jiangsu Provincial Government, 2008). 
SO2 emission goals varied between 2.6% to 53.6% reduction (Jiangsu Provincial 
Government, 2008). Regression models included four independent variables, all 
for 2005 at municipality level: nonpower SO2 emissions density, GDP per capita, 
SO2 emissions and the power sector’s share in total emissions (Table 4.3). The 
constant variable showed significance, and its appearance in models made the 
adjusted R2 bigger. Corresponding to the increase of, respectively, 10,000 tons of 
SO2 emissions, 1% of the power sector’s share and 1 ton/km2 of nonpower SO2 
emissions density, a municipality goal would become 1.4%, 0.6% and 1.5% more 
stringent (from the 2005 level). The signs of these coefficients were all reasonable 
and consistent with the situation of distributing the national goal to provinces, 
but GDP per capita displayed the opposite effect: for a municipality with 10,000 
RMB/person richer, its SO2 emissions were allowed to grow by 6.2%. For a prov­
ince, this strategy of “rich municipalities reduce less” might maximize its GDP 
as well as tax income through entitling more opportunities to more promising 
municipalities.
In Shanxi Province, municipalities with higher emissions should reduce more, 
but income level did not have significant impacts. Shanxi Province’s goal was a 
14% reduction in the 11th Five-­
Year Plan from 1.52 million tons in 2005 (State 


Mobilizing the government  61
Council, 2006). Its 11 municipalities emitted from 92,000 to 185,000 tons of SO2 
in 2005, a much narrower range compared with the earlier three provinces or the 
national situation (Shanxi Provincial Government, 2006). Their GDP per capita 
in 2005 was between 5,500 and 26,000 RMB/person (Shanxi Bureau of Statistics, 
2006). The municipality goals were scattered from a 8.5% to a 17.8% reduction 
(Shanxi Provincial Government, 2006). Only total SO2 emissions showed a sig­
nificant influence in the statistical models (Table 4.3). For emitting every 10,000 
tons more of SO2, a municipality goal would be about 1% more stringent (from 
the 2005 level). The GDP per capita’s coefficient was negative, although not sta­
tistically significant. The principle – big provinces should reduce more – held 
here. That rich provinces should reduce more was not well applied but could have 
been considered.
In conclusion, provinces differed from each other in adopting principles from 
distributing the national goal (Table 4.4), which closely reflected that China’s 
governance and, especially, environmental governance had been greatly decen­
tralized (see Chapter 3). The most consistent principle across provinces was that 
bigger emitters should reduce more. The guidance from the SEPA did not have to 
be exactly followed.
4  
Decentralized goal attainment
To keep the goal process running, goal attainment assessment is an inalienable 
step. It examines the effectiveness of the goal process and provides feedback. The 
key questions are what can be called goal attainment and how to evaluate it. For 
one Five-­
Year Plan, goal attainment evaluation does not wait until its conclusion. 
In the 11th Five-­
Year Plan, China publicized provincial SO2 emissions every half 
a year (State Council, 2007c). At the end of 2008, a halfway assessment was 
scheduled (State Council, 2007c).
4.1  
Criteria for goal attainment
In the 11th Five-­
Year Plan, China established “three systems” to facilitate SO2 
mitigation, which covered statistics, monitoring and evaluation (State Coun­
cil, 2007b). This capacity building was planned and carried out by the SEPA 
and endorsed by the State Council (State Council, 2007b). Because of China’s 
Table 4.4 Provincial goal distribution matrix
Rich guys reduce
Less
Neutral
More
Big guys reduce
Less
Neutral
Hebei
More
Jiangsu
Shanxi
Guangdong


62  Mobilizing the government
decentralization, that the central government mainly governs through provin­
cial governments, the national system only targeted the provincial level (State 
Council, 2007b). The subordinate governments within provinces were evaluated 
with the rules passed along by provincial governments. For example, Zhejiang 
Province later enacted a more detailed, although not systematically different, 
regulation targeting municipality and county governments (Zhejiang Provincial 
Government, 2008).
Provincial goal attainment was evaluated with three criteria in the 11th Five-­
Year Plan (State Council, 2007b). The first criterion was on the quantitative goal 
itself and environmental quality. It was often a binary judgment: if they were 
attained, that was a mission accomplished. Excessive reduction would not be fur­
ther awarded after the goal had been attained, while more emissions would not be 
punished if the goal was already broken. The second criterion was on the establish­
ment and operation of three institutions: environmental goal setting of major pol­
lutants, monitoring and goal attainment evaluation (State Council, 2007b). They 
were mainly judged by the enactment and distribution of official documents. The 
third one was on mitigation measures, including the completion and operation of 
pollutant removal facilities, the closure of inefficient factories, policy enactment 
and plan implementation (State Council, 2007b). If any of the three criteria failed 
to pass evaluation, the overall goal attainment would be judged a failure (State 
Council, 2007b). Accordingly, one feature was that the attainment of the goal 

itself, despite its central importance, did not ensure overall success. The first crite­
rion focused on the results and the other two on the process. The regulation of the 
Chinese central government on the process provided feedback for local govern­
ments for adjusting policies and monitoring their implementation.
The 9th Five-­
Year Plan barely had any defined scheme to evaluate SO2 goal 
attainment (NEPA et  al., 1996). The official document available in the public 
domain only pointed out that the SO2 control would be annually examined and 
evaluated and the result would be publicized periodically (NEPA et al., 1996). 
The SO2 goal and its attainment process were better defined in the 10th Five-­
Year Plan, but no evaluation scheme was clearly defined either (SEPA, 2001). 
The National 10th Five-­
Year Plan for Environmental Protection only expressed 
several principles, including holding local government leaders responsible and 
linking environmental goal attainment with the leaders’ performance evaluation 
(SEPA, 2001). The evolution path displayed China’s progress in establishing a 
working evaluation scheme of SO2 emission goals. Although still not perfect, the 
much clearer scheme in the 11th Five-­
Year Plan could have significantly contrib­
uted to the SO2 mitigation goal attainment.
Recognizing the importance of credible data collection for achieving SO2 emis­
sion goals in the 11th Five-­
Year Plan, China experienced an intensive capacity-­
building process. In December  2006, the updated Management Methods of 
Environmental Statistics entered into force (SEPA, 2006c). The regulation speci­
fied the organization and personnel for environmental statistics, rules on environ­
mental survey and management and publication of environmental data (SEPA, 
2006c). For goal attainment in the 11th Five-­
Year Plan, China strengthened its 


Mobilizing the government  63
statistical system focusing on data credibility. SO2 emissions were divided into 
three categories: power, nonpower industries and domestic (State Council, 
2007b). The first two categories (industrial sectors) were further distinguished 
into two – key and non-­
key surveyed sources – based on the sizes of emission 
sources, and key surveyed sources covered 65% of total industrial SO2 emissions 
(State Council, 2007b). Three parallel methods were applied under various situ­
ations: direct monitoring, estimation according to sulfur budget and estimation 
according to emission factors (State Council, 2007b). The first method, if appli­
cable, had the highest priority (State Council, 2007b). SO2 emissions from non-­
key surveyed sources were estimated following a similar trend as key surveyed 
sources (State Council, 2007b). Data about coal consumption and sulfur contents 
worked out SO2 emissions from domestic sectors (State Council, 2007b). In addi­
tion, if cheating were caught in an SO2 removal facility more than twice a year, 
no SO2 removal would be recognized in the statistics data from the facility (State 
Council, 2007b). Furthermore, another two significantly more detailed policies 
were enacted for building emission inventories (SEPA, 2007a, 2007d). Not only 
were detailed accounting methods clearly written, but also the data report was 
regulated in specifics (SEPA, 2007a, 2007d).
4.2  
Incentives for goal attainment
The central government has decentralized its power greatly since the economic 
reform started in 1978. As discussed in Chapter 3, three measures could exist 
to incentivize the cooperation of local governments by targeting local leaders, 
administrative constraints and fiscal transfer. Corresponding to the personnel 
relationship that is mainly established across the various levels of the Chinese 
Communist Party, the top national leadership of the party can greatly decide the 
promotion and removal of provincial-­
level leaders. The attainment of key goals, 
including those on environmental protection and SO2 mitigation, had become an 
important aspect in the evaluation of provincial leaders’ job performance. Con­
cerning SO2 mitigation goal implementation since the 11th Five-­
Year Plan, local 
government leaders but not local environmental protection bureau (EPB) leaders 
were targeted. The clear evidence was that provincial deputy governors, not EPB 
directors, were required to sign pollutant emission control contracts with the cen­
tral government (SEPA, 2006b). The failure in the 10th Five-­
Year Plan on surging 
coal consumption demonstrated that pollution control had been far beyond the 
responsibility of EPBs alone.
Officially five characteristics distinguish a leader in the Chinese Communist 
Party for promotion or removal: virtue, ability, diligence, achievements and 
absence of corruption (The Central Committee of the Chinese Communist Party, 
2002). Furthermore, after the formation of “Scientific View of Development,” 
resource consumption, environmental protection and sustainable development 
were clearly pointed out to comprise “achievements” (Department of Organi­
zation of the Chinese Communist Party, 2006). Contracts on pollutant emission 
control and energy conservation clarified even more the responsibilities of local 


64  Mobilizing the government
government leaders (SEPA, 2006b). Two institutions were applied for the attain­
ment of the SO2 emission goal in the 11th Five-­
Year Plan: accountability and 
veto (State Council, 2007a). “Accountability” demanded local government lead­
ers be held accountable for their governance that fell within their jurisdictions. For 
example, the administrator of the SEPA, Xie Zhenghua, was forced to resign in 
2005 for a serious pollution event in the Songhua River. “Veto” meant that local 
government leaders would fail evaluation on their entire job performance if the 
SO2 emission goal were not attained. Promotion became inappropriate for these 
leaders. If goal failure did not degrade the leaders’ ranks, they may still face a risk 
of being removed from original positions to some less significant ones. On the 
other hand, successful goal attainment was an important achievement and could 
help the leaders’ promotion. A recently developed method for targeting local lead­
ers has been gradually promoted by the Ministry of Environmental Protection 
(MEP) and later by the Ministry of Ecology and Environment. Top leaders of 
those provinces and municipalities that show serious environmental problems or 
fail environmental goals are forced to have “interview appointments” with the 
ministry (Ministry of Ecology and Environment, 2020a). Although those local 
leaders may not face immediate consequences of punishment, they will receive 
crucial warnings that darken their future promotion opportunities, especially if no 
quick fix is achieved afterward.
Another mechanism that was applied in the 11th Five-­
Year Plan was to tem­
porarily constrain local administrative authorities as punishment: if a goal was 
not attained, no new construction projects would receive the ratification of their 
environmental impact assessment (EIA) reports for a given period. Over the 11th 
Five-­
Year Plan period, large construction projects still demanded ratification from 
the central government. In terms of environmental protection, every project with 
potential environmental damage should compose an EIA report and submit for 
ratification to various levels of governments (National People’s Congress, 2002). 
The SEPA, and later the MEP, at the central level was responsible for large pro­
jects, such as new coal-­
fired power plants over 200 MW (SEPA, 2002). No project 

without the MEP’s ratification could legally start construction. In early 2007, the 
SEPA temporarily suspended ratifying EIA reports of four municipalities and 
four power corporations (SEPA, 2007c). The suspension took effect for three 
months to force their cooperation (SEPA, 2007b). Afterward, the policy was for­
mally established to target goal failure (SEPA, 2008). A failure to achieve the SO2 
emission goal could result in regional suspension for one month, three months or 
half a year. If no satisfying progress were made, the suspension could even last 
longer until full cooperation. The “suspension” policy may seriously influence 
the regional economy. Since GDP is the most important criterion in evaluating 
local leaders, this mechanism could effectively force cooperation. Capital invest­
ment was a crucial part of China’s GDP. For example, in 2007, China’s over­
all GDP was 24.7 trillion RMB, and capital investment comprised 13.7 trillion 
RMB, about 56% (National Statistics Bureau, 2008). A one-­
month suspension 
could delay construction and significantly affect capital investment and, conse­
quently, the local economy. GDP growth itself occupied the most important status 


Mobilizing the government  65
in evaluating local government leaders. In addition, a booming GDP could pro­
vide growing tax income not only to make officials more powerful but also to 
enable more budgets for poverty alleviation, health care, education and other key 
governmental affairs. Many of these issues are closely connected with the evalu­
ation of leaders.
Fiscal transfer has not been explicitly linked with environmental goal attain­
ment. However, the very significant fiscal transfer from the central to local gov­
ernments (as discussed in Chapter 3), if institutionally associated with pollutant 
emission control, is potentially powerful to mobilize local governments for envi­
ronmental protection.
With China’s further decentralization of governmental authorities, the first 
mechanism to directly target local governments is expected to be even more 
important. In the past four decades, the central government has been continuously 
loosening direct management of local governmental affairs. As a key feature of 
the economic reform, China has greatly reduced the requirements of adminis­
trative ratification and decentralized much remaining authority to local govern­
ments (State Council, 2013b, 2014). Fossil-­
fuel-­
fired power plants were no longer 
required for the MEP’s ratification after 2015 and the authority entirely went to 
provincial governments (MEP, 2015; Ministry of Ecology and Environment, 
2019).
5  
Goal evolution
Corresponding to different strategies for controlling air pollution–induced health 
damages, three major types of goals can be adopted. First, emission mitigation 
goals of key pollutants, prominently SO2, aim to directly target the sources of 
environmental pollution. The second type focuses on controlling air pollutant 
concentrations. Ambient air quality standards are widely adopted across coun­
tries to specify concentration thresholds of key air pollutants individually, such 
as SO2, fine particulate matter (PM2.5) and ozone (O3). As discussed earlier, the 
control of ambient SO2 concentration was a key scientific foundation to decide 
China’s long-­
term SO2 mitigation goal at 12 million tons (Yang et al., 1999). The 
third type targets environmental quality directly through the Air Quality Index 
(AQI) that provides a synthesized measurement of key air pollutant concentra­
tions. The AQI also guides people’s activities corresponding to air quality condi­
tions. Although the three strategies have a similar ultimate goal for protecting 
public health, they have different implications for implementation. Local govern­
ments can only directly mitigate local emissions while local pollutant concentra­
tion is determined by emissions within and outside of their jurisdiction as well as 
weather conditions, land use and other factors. Then their motivation could differ 
significantly under the different types of goals to affect their performance of pol­
lution mitigation.
Over the past two decades, China has been switching back and forth between 
major governance strategies on environmental protection with different types of 
goals. 
As clearly stated in China’s environmental protection law, local governments 


66  Mobilizing the government
are responsible for environmental quality within their jurisdictions (National Peo­
ple’s Congress, 1989). However, environmental protection was not ranked high 
among all governmental tasks in the 1990s. Local leaders generally prioritized 
economic growth for promotion opportunities. The 10th Five-­
Year Plan (2001–
2005) was a transitional period toward the Total Emission Control regime to set 
up environmental goals for reducing major pollutant emissions by 10% (National 
People’s Congress, 2001). However, due to the lack of environmental cleanup 
incentives and the acceleration of economic growth, SO2 emissions went up by 
27.8%, and only 2 out of 31 provinces achieved their allocated goals. Demand 
for serious, effective and efficient compliance monitoring had not been strong. 
The 11th Five-­
Year Plan (2006–2010) was a milestone in China’s environmental 
protection history. The Total Emission Control regime was strengthened, while 
serious and implementable incentives were put into place for local governments 
to achieve their individual mitigation goals (Xu, 2011). A bottom-­
up compliance 
monitoring system on emissions was initiated and established (SEPA, 2007d). 
Although SO2 emissions did decline in the 11th Five-­
Year Plan, data manipulation 
also strained the compliance monitoring system as indicated in the gaps between 
official and independent emission inventories (Lu et al., 2011).
Concerning SO2 emissions, two sets of regulations were most important and 
direct, being effluent emission standards and ambient air quality standards. Pre­
viously, cities were given goals of “blue sky” days. “Blue sky” was defined as 
that air quality reached the Grade 2 standard. One crucial change in the 2012 
version ambient air quality standards was the addition of PM2.5 (MEP, 2012; 
National Environmental Protection Administration and State Bureau of Techni­
cal Supervision, 1996). PM2.5 concentration is more closely related to air quality 
that affects public health, while the emissions of SO2 and other pollutants are 
only indirect measures. In other words, PM2.5 goals are more related to ends of 
air pollution control, while SO2 emissions goals are more about means. PM2.5 
comprises many more pollutants, including sulfate particles that are originated 
from SO2 emissions.
Together with the 2012 update of the ambient air quality standards, China 
enacted the Ambient Air Quality Index (Ministry of Environmental Protection, 
2012). It synthesizes key air pollutant concentrations into one index to indicate 
air quality. The cutoff AQIs between “excellent,” “good” and “polluted” air are 50 
and 100, respectively. Each air pollutant can calculate its individual AQI (IAQI) 
and the composite AQI is the largest IAQI, or the IAQI of the primary air pol­
lutant. An AQI of 50 or lower corresponds to the Grade 1 ambient air quality 
standards, while 100 or lower corresponds to Grade 2. They provide the technical 
foundation for China to adopt regulatory strategies that are based on air qual­
ity rather than pollutant emissions. Both regulations gave nearly four years of 
grace periods and formally entered into force in January 2016. The 12th Five-­
Year 
Plan (2011–2015) initially continued with the Total Emission Control scheme to 
include more pollutants (National People’s Congress, 2011). However, a major air 
pollution episode in January 2013 that badly hit North China, most notably Bei­
jing, pushed the Chinese government to rethink its strategy (State Council, 2013a) 


Mobilizing the government  67
and accelerated the shift toward the air quality approach and the application of the 
two related standards.
Air quality goals and emission mitigation goals of SO2, as well as other major 
air pollutants, both aim for public health benefits. In order to achieve air quality 
goals that focus on ambient air pollution, efforts should still primarily fall on 
the mitigation of pollutant emissions together with their geographic and temporal 
distributions. Due to the atmospheric transport of air pollution, the attainment of 
PM2.5 goals depends not only on a region’s own mitigation efforts but also that of 
neighboring regions. The interregional reliance tends to be greater for geographi­
cally smaller jurisdictions. Accordingly, free riding may be a potential problem 
to compromise the willingness to engage in hard mitigation efforts. Nevertheless, 
data credibility is a key element in enforcing environmental policies as well as 
the top-­
down goals. Emission mitigation data, however, tend to be much more 
conveniently manipulated than air quality data. The number of polluting sources 
in China could easily overwhelm its compliance monitoring resources, especially 
in sparsely populated and less developed regions. In the 11th Five-­
Year Plan, the 
MEP assembled teams to inspect provinces and their polluting firms. However, 
the data had been of unsatisfying quality, and what was reported by local govern­
ments and polluting firms was often seriously discounted. Data on SO2 emissions 
are more prone to manipulation because the bottom-­
up monitoring and reporting 
have to go through many stakeholders who have incentives to underreport emis­
sions and overreport mitigation. Occasional verification from the central govern­
ment often finds big gaps in data and must “squeeze moisture” from the reported 
mitigation amounts. In contrast, ambient air quality data are much more difficult 
to manipulate and any dishonest behavior is much easier to discover. The central 
government also runs its own air quality monitoring network via ground stations 
and remote sensing, such as satellites. Accordingly, China reversed the strategy 
to have air quality improvement targets (State Council, 2013a). Air quality moni­
toring stations are much fewer than polluting sources to substantially reduce the 
resource burden of compliance monitoring. Thus, the probability of compliance, 
together with the better data quality, should be much higher.
The prospective penalty and reward for goal attainment do not differ substan­
tially from the 11th Five-­
Year Plan to the 12th and 13th. However, the 12th and 
13th Five-­
Year Plans achieved much faster SO2 mitigation, even considering the 
slower economic growth rates. It could indicate that the free-­
riding problems 
were less important than data credibility. Furthermore, SO2 emissions are just 
one among many pollutants, while PM2.5 could better serve as a comprehen­
sive air quality indicator. Provincial and local governments could have greater 
flexibility in weighing various technological and policy mitigation alternatives. 
It could also potentially encourage more local policy innovations and probably 
achieve better cost-­
effectiveness through balancing the marginal abatement costs 
of pollutants.
Furthermore, although emission reduction goals have been consistently 
achieved since the 11th Five-­
Year Plan, air quality was not perceived to have 
improved. One possible cause could be the problems in reporting emission data, 


68  Mobilizing the government
0
50
100
150
200
250
300
1/1/2014
1/1/2015
1/1/2016
1/1/2017
1/1/2018
1/1/2019
1/1/2020
(
 
n
o
i
t
a
r
t
n
e
c
n
o
c
 
e
g
a
r
e
v
a
 
y
l
i
a
D
μg/m3)
Date
Figure 4.2  
Daily SO2 concentrations in Shijiazhuang (1 January 2014–29 February 2020)
Source: Ministry of Ecology and Environment (2020b).
Note: The upper and lower dotted horizontal lines indicate the Grade 2 and 1 standard, respectively, in 
China’s ambient air quality standards in 1996 and 2012.
as discussed earlier. Another more important reason for the wide gap between 
the successful attainment of SO2 mitigation goals and the perceived terrible air 
quality was that SO2 has been increasingly less important in ambient air quality. 
For example, Hebei Province often has one of the highest anthropogenic PM2.5 
concentrations in China and the world. In its capital city, Shijiazhuang, air qual­
ity is taken as one example to illustrate the importance of new PM2.5 standards 
and goals. Significant improvements have been made on reducing SO2 emissions 
and concentrations. In the first three months of 2014, SO2 concentrations in Shi­
jiazhuang exceeded the Grade 1 standard (50 μg/m3) in 90% of all days, while a 
strong seasonal cycle indicated that the winter or the heating season as the worst 
season (Figure 4.2). From February 2019 to February 2020, in contrast, the stand­
ard was not exceeded for even a single day (Figure 4.2). It illustrates China’s hard 
and effective efforts in controlling SO2 emissions and bringing down SO2 concen­
trations. Essentially the original long-­
term goal for SO2 mitigation, 0.060 mg/m3 
or 60 μg/m3 (SEPA, 2006a), had been generally achieved. However, from the 
perspective of PM2.5, Shijiazhuang’s performance has been much less impressive. 
Its concentration has regularly exceeded the much more relaxed Grade 2 standard 
(Figure 4.3).


Mobilizing the government  69
0
50
100
150
200
250
300
350
400
450
500
550
600
650
1/1/2014
1/1/2015
1/1/2016
1/1/2017
1/1/2018
1/1/2019
1/1/2020
(
 
n
o
i
t
a
r
t
n
e
c
n
o
c
 
e
g
a
r
e
v
a
 
y
l
i
a
D
μg/m3)
Date
Figure 4.3  
Daily PM2.5 concentrations in Shijiazhuang (1 January 2014–29 February 2020)
Source: Ministry of Ecology and Environment (2020b).
Note: The upper and lower dotted horizontal lines indicate Grade 2 and 1 standard, respectively, in 
China’s 2012 ambient air quality standards.
PM2.5 is not a single pollutant but a set of various pollutants that fall into the 
size range. SO2 is a gaseous pollutant and could be converted into sulfate particles 
in the atmosphere to become one important component of PM2.5. Heating seasons 
in northern China tend to result in more coal consumption and pollutant emis­
sions, while inversion (when warm air is above cold air) is more frequent in the 
winter when the ground is cold, suppressing convection, and thus facilitates the 
accumulation of pollutant concentrations. Although SO2 is one key precursor spe­
cies of PM2.5, other air pollutants are also crucial components in forming PM2.5.
Furthermore, ozone pollution has significantly deteriorated over the period. 
O3 and PM2.5 concentrations tend to have opposite seasonal cycles. Chemical 
reactions to form O3 in the atmosphere involve nitrogen oxides (NOx), volatile 
organic compounds (VOC) and sunlight, while summer months tend to provide 
more favorable conditions. PM2.5 and SO2 concentrations peak in winter months, 
and O3–8h concentration (daily maximum concentration over 8 hours) is the high­
est in summer months (Figure 4.4). As a result, mitigation goals of SO2 emissions 
and SO2 concentrations will be at a greater distance from perceived air quality that 
mainly corresponds to PM2.5 and O3 concentrations.
SO2 has never been the primary pollutant to decide Shijiazhuang’s monthly 
AQI since 2014 (Figure 4.5). PM2.5 dominated the AQI before 2016, while in and 


70  Mobilizing the government
0
50
100
150
200
250
300
350
1/1/2014
1/1/2015
1/1/2016
1/1/2017
1/1/2018
1/1/2019
1/1/2020
Daily 8-
(
 
n
o
i
t
a
r
t
n
e
c
n
o
c
 
e
g
a
r
e
v
a
 
r
h
μg/m3)
Date
Figure 4.4  
Daily 8-hour O3 concentrations (daily maximum concentration over 8 hours) in 
Shijiazhuang (1 January 2014–29 February 2020)
Source: Ministry of Ecology and Environment (2020b).
Note: The upper and lower dotted horizontal lines indicate Grade 2 and 1 standard, respectively, in 
China’s 2012 ambient air quality standards.
0
50
100
150
200
250
300
350
Jan-14
Jan-15
Jan-16
Jan-17
Jan-18
Jan-19
Jan-20
)
I
Q
A
I
(
 
x
e
d
n
I
 
y
t
i
l
a
u
Q
 
r
i
A
 
l
a
u
d
i
v
i
d
n
I
Month -Year
PM2.5
SO2
O3-8h
Figure 4.5  
Monthly average AQI in Shijiazhuang (January 2014–February 2020; calcu­
lated from daily data)
Source: Ministry of Ecology and Environment (2020b).


Mobilizing the government  71
after 2017 with reduced PM2.5 concentration and rising O3–8h concentration, O3 
became the primary pollution in summer months and PM2.5 remained dominant in 
the winter (Figure 4.5). The trend is similar in other Chinese cities. For example, 
Beijing witnessed the rise of O3 in determining summer AQI a few years earlier 
than Shijiazhuang did (Figure 4.6). In southern China, where winter is mild/warm 
with adequate sunshine, the importance of O3 entirely overshadows that of PM2.5 
in the AQI. For example, in Shenzhen, AQI in most months is now decided by 
O3–8h but not PM2.5 (Figure 4.7).
From 2014 to 2020, PM2.5 concentrations and corresponding air quality indexes 
have been reduced throughout major cities in China, but O3–8h generally had a 
rising trend. One reason for their diverging trends in the past years could be traced 
to the presence of PM2.5 goals but not O3 goals. In the 13th Five-­
Year Plan, China 
further enacted air quality goals together with 15% reduction goals on SO2 and 
NOx emissions (National People’s Congress, 2016). The proportion of days that 
the AQI is below 100 in municipalities should reach 80%, while for those cities 
with PM2.5 concentrations not reaching the Grade 2 standard (or 75 μg/m3), they 
should reduce the level by 18% over the five years (National People’s Congress, 
2016).
The AQI is a more comprehensive measure of air pollution to consider both 
PM2.5 and O3. In China’s further goal evolution especially into the 14th Five-­
Year 
0
50
100
150
200
250
Jan-14
Jan-15
Jan-16
Jan-17
Jan-18
Jan-19
Jan-20
)
I
Q
A
I
(
 
x
e
d
n
I
 
y
t
i
l
a
u
Q
 
r
i
A
 
l
a
u
d
i
v
i
d
n
I
Month -Year
PM2.5
SO2
O3-8h
Figure 4.6  
Monthly average AQI in Beijing (January  2014–February  2020; calculated 
from daily data)
Source: Ministry of Ecology and Environment (2020b).


72  Mobilizing the government
Plan (2021–2025), it could play a more prominent role in mobilizing local govern­
ments for air pollution control.
Note
	
1	 This chapter is based on the author’s own material used in Xu, Y. 2011. The use of a goal 
for SO2 mitigation planning and management in China’s 11th five-­
year plan. Journal 
of Environmental Planning and Management, 54, 769–783; much of which has been 
revised and expanded on.
References
The Central Committee of the Chinese Communist Party. 2002. Regulations on selecting 
and appointing leaders of the party and governments. Beijing, China: The Central Com­
mittee of the Chinese Communist Party.
Chakravarty, S., Chikkatur, A., De Coninck, H., Pacala, S., Socolow, R. & Tavoni, M. 
2009. Sharing global CO2 emission reductions among one billion high emitters. Pro­
ceedings of the National Academy of Sciences of the United States of America, 106, 
11884–11888.
Chinese Academy for Environmental Planning (CAEP). 2004. Basic thoughts on national 
11th five-­
year plan on environmental protection. Beijing, China: CAEP.
0
10
20
30
40
50
60
70
80
90
Jan-14
Jan-15
Jan-16
Jan-17
Jan-18
Jan-19
Jan-20
)
I
Q
A
I
(
 
x
e
d
n
I
 
y
t
i
l
a
u
Q
 
r
i
A
 
l
a
u
d
i
v
i
d
n
I
Month -Year
PM2.5
SO2
O3-8h
Figure 4.7  
Monthly average AQI in Shenzhen (January 2014–February 2020; calculated 
from daily data)
Source: Ministry of Ecology and Environment (2020b).


Mobilizing the government  73
Department of Organization of the Chinese Communist Party. 2006. Temporary methods 
to evaluate local leaders for realizing scientific view of development. Beijing, China: 
Chinese Communist Party.
DeYoung, K. 2009. Obama sets timetable for Iraq. Washington Post, February 28.
Guangdong Environmental Protection Bureau. 2006. The 11th five-­
year plan on major pol­
lutants emission goals of municipalities in Guangdong province. Guangzhou, China: 
Guangdong Environmental Protection Bureau.
Hebei Provincial Government. 2007. The 11th five-­
year plan on environmental protection 
of Hebei province. Shijiazhuang, China: Hebei Provincial Government.
Hebei Provincial Statistics Bureau. 2006. Hebei economy statistics yearbook. Beijing, 
China: China Statistics Press.
Jiangsu Provincial Government. 2008. The 11th five-­
year plan on environmental protection 
and ecological construction of Jiangsu province. Nanjing, China: Jiangsu Provincial 
Government.
Li, J. 2010. How to allocate CO2 mitigation responsibilities within China – application of a 
new burden-­
sharing scheme. Ph.D. dissertation, Princeton University Press, Princeton, NJ.
Liu, G., Zhang, Z., Dong, Z. & Wu, L. 2006. Research report on China’s ten five-­
year 
plans. Beijing, China: People’s Press.
Locke, E. A., Saari, L. M., Shaw, K. N. & Latham, G. P. 1981. Goal setting and task-­
performance – 1969–1980. Psychological Bulletin, 90, 125–152.
Lu, Z., Zhang, Q. & Streets, D. G. 2011. Sulfur dioxide and primary carbonaceous aero­
sol emissions in China and India, 1996–2010. Atmospheric Chemistry and Physics, 11, 
9839–9864.
Ma, K. 2005. Strengthen the scientific and democratic planning process and write a good 
11th five-­
year plan. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
MEP. 2012. Ambient air quality standards. GB 3095-­
2012. Beijing, China: MEP.
MEP. 2015. Catalogue of construction projects for environmental impact assessment ratifica­
tion from the ministry of environmental protection (2015 version). Beijing, China: MEP.
Ministry of Ecology and Environment. 2019. Catalogue of construction projects for envi­
ronmental impact assessment ratification from the ministry of environmental protection 
(2019 version). Beijing, China: Ministry of Ecology and Environment.
Ministry of Ecology and Environment. 2020a. Method for interview appointment by the 
ministry of ecology and environment (Draft for comments). Beijing, China: Ministry of 
Ecology and Environment.
Ministry of Ecology and Environment. 2020b. Realtime city air quality data [Online]. 
Available: www.mee.gov.cn/.
Ministry of Environmental Protection. 2008. Assessment reports on provincial emissions of 
major pollutants. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2012. Technical regulation on ambient air quality 
index (on trial). Beijing, China: Ministry of Environmental Protection.
National Bureau of Statistics. 1997–2008. China energy statistical yearbook. Beijing, 
China: China Statistics Press.
National Bureau of Statistics. 2006. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Statistics 
Press.
National Bureau of Statistics of China. 1999. China statistical yearbook. Beijing, China: 
China Statistics Press.


74  Mobilizing the government
National Environmental Protection Administration & State Bureau of Technical Supervi­
sion. 1996. Ambient air quality standard. GB 3095-­
1996. Beijing, China: NEPA.
National People’s Congress. 1989. Law of environmental protection. Beijing, China: The 
4th Conference of the 10th National People’s Congress.
National People’s Congress. 2001. The outline of national 10th five-­
year plan on economic 
and social developments. Beijing, China: The 4th Conference of the 9th National Peo­
ple’s Congress.
National People’s Congress. 2002. Law of environmental impact assessment. Beijing, 
China: The 4th Conference of the 10th National People’s Congress.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. The outline of the 13th five-­
year plan on economic and 
social development. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
National Statistics Bureau. 2008. China’s economic growth in 2007. Beijing, China: China 
Statistics Press.
National Statistics Bureau & Ministry of Ecology and Environment. 2019. China statisti­
cal yearbook on environment 2018. Beijing, China: China Statistics Press.
NDRC. 2003. Establish a new view of development and do well in preparing writing the 
11th five-­
year plan. Beijing, China: NDRC.
NDRC. 2005. On the basic thoughts of the national 11th five-­
year plan. Beijing, China: NDRC.
NEPA, National Planning Commission & NETC. 1996. The national 9th five-­
year plan on 
controlling major pollutants’ emissions. Beijing, China: NEPA, NPC, NETC.
NEPA  & SBTS. 1996. Ambient air quality standard. GB 3095-­
1996. Beijing, China: 
NEPA, SBTS.
Pacala, S. W. & Socolow, R. H. 2004. Stabilization wedges: Solving the climate problem 
for the next 50 years with current technologies. Science, 305, 968–972.
Pouyat, R. V. & McGlinch, M. A. 1998. A legislative solution to acid deposition. Environ­
mental Science & Policy, 1, 249–259.
Roberts, L. 1991. Acid-­
rain program – mixed review. Science, 252, 371.
SEPA. 2001. National 10th five-­
year plan on environmental protection. Beijing, China: 
State Environmental Protection Administration.
SEPA. 2002. Regulation on ratifying environmental impact assessment reports of construc­
tion projects at different government levels. Decree No. 15. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2005. Plan on controlling acid rain and SO2 emissions (draft for comments). Bei­
jing, China: State Environmental Protection Administration.
SEPA. 2006a. Guidelines on calculating SO2 emission quotas. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2006b. A letter on signing up liability contracts of major pollutants. Beijing, China: 
State Environmental Protection Administration.
SEPA. 2006c. Management methods of environmental statistics. Beijing, China: State 
Environmental Protection Administration.
SEPA. 2006d. Monthly information on writing the national 11th five-­
year plan on envi­
ronmental protection. Beijing, China: State Environmental Protection Administration.


Mobilizing the government  75
SEPA. 2007a. Detailed methods to verify major pollutants emission reduction in the 11th 
five-­
year period (on trial). Beijing, China: State Environmental Protection Administration.
SEPA. 2007b. Lvliang and Liupanshui municipalities and Huadian corporation are finally 
removed from suspension. Beijing, China: State Environmental Protection Administration.
SEPA. 2007c. SEPA disclosed illegal construction projects and used ‘regional suspension 
of ratification’ policy. Beijing, China: State Environmental Protection Administration.
SEPA. 2007d. Verification of major pollutants emission reduction in the 11th five-­
year 
period (on trial). Beijing, China: State Environmental Protection Administration.
SEPA. 2008. Management methods of regional suspension of ratifying environmental 
impact assessment reports (on trial) (draft for comments). Beijing, China: State Envi­
ronmental Protection Administration.
Shanxi Bureau of Statistics. 2006. Shanxi statistics yearbook. Taiyuan, China: Shanxi 
Bureau of Statistics.
Shanxi Provincial Government. 2006. Notice on controlling major pollutants emissions in 
the 11th five-­
year plan. Taiyuan, China: Shanxi Provincial Government.
State Council. 1990. Regulations for the implementation of the law of standardization. 
Beijing, China: State Council.
State Council. 2005a. Decisions on realizing scientific view of development and strengthen­
ing environmental protection. Beijing, China: State Council.
State Council. 2005b. Several advices from the state council on strengthening the enact­
ment of economic and social development plans. Beijing, China: State Council.
State Council. 2006. Opinion on pollution emission quota distribution in the 11th five-­
year 
plan. Beijing, China: State Council.
State Council. 2007a. Notice on distributing composite working plan on energy conserva­
tion and pollutant emission reduction. Beijing, China: State Council.
State Council. 2007b. Notice on distributing implementation plans and methods of statis­
tics, monitoring and assessment on energy conservation and pollutant emission reduc­
tion. Beijing, China: State Council.
State Council. 2007c. Notice on distributing the 11th five-­
year plan on environmental pro­
tection. Beijing, China: State Council.
State Council. 2013a. Action plans on air pollution prevention and control. Beijing, China: 
State Council.
State Council. 2013b. Decisions on cancelling and decentralizing administrative ratifica­
tion [2013(19)]. Beijing, China: State Council.
State Council. 2014. Decisions on cancelling and decentralizing administrative ratification 
[2014(5)]. Beijing, China: State Council.
State Environmental Protection Administration (SEPA). 2001–2009. National report on 
environmental statistics. Beijing, China: State Environmental Protection Administration.
United Nations. 1992. United nations framework convention on climate change. New 
York: United Nations.
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. 
Congress.
U.S. Environmental Protection Agency. 2007. National emissions inventory (NEI) air pol­
lutant emissions trends data. Washington, DC: U.S. Environmental Protection Agency.
Wang, J., Wu, X., Cao, D. & Meng, F. 2004. Proposed scenarios for total emission control 
of SO2 during the 10th five-­
year plan period in China. Research of Environmental Sci­
ences, 17, 4.
Wang, X. 2002. Exponent on ‘national 10th five-­
year plan on environmental protection’. 
Beijing, China: Science Press.


76  Mobilizing the government
Xinhua News Agency. 2005. The birth of ‘the suggestions to 11th five-­
year plan’ [Online]. 
Beijing, China [Online]. Available: http://news.xinhuanet.com/politics/2005-­
10/26/con­
tent_3685219.htm.
Xinhua News Agency. 2006. The drafting of 11th five-­
year plan outline. Beijing, China 
[Online]. Available: http://news.xinhuanet.com/politics/2006-­
03/16/content_4308918.
htm.
Xu, Y. 2011. The use of a goal for SO2 mitigation planning and management in China’s 
11th five-­
year plan. Journal of Environmental Planning and Management, 54, 769–783.
Yang, X., Gao, Q., Jiang, Z., Ren, Z., Chen, F., Chai, F. & Xue, Z. 1998. Research on the 
transportation and precipitation regular pattern of sulfur pollutants in China. Research of 
Environmental Sciences, 11, 27–34.
Yang, X., Gao, Q., Qu, J. & Jiang, Z. 1999. The exploration and initial assessment of total 
amount control method for SO2 emission in China. Research of Environmental Sciences, 
12, 17–20.
Zhejiang Provincial Government. 2008. Notice on distributing Zhejiang provincial imple­
mentation plans and methods of statistics, monitoring and assessment on energy con­
servation and pollutant emission reduction. Hangzhou, China: Zhejiang Provincial 
Government.
Zou, S., Wang, J. & Hong, Y. 2006. Research report on national environmental protection 
plan in the 11th five-­
year plan. Beijing, China: China Environmental Science Press.
Zou, S., Zhang, Z., Yan, G. & Tian, R. 2004. Preliminary assessment on medium-­
term 
execution of the national tenth five-­
year plan outline, important eco-­
environmental con­
servation and environmental protection plan. In: Wang, J., Zou, S. & Hong, Y. (eds.) 
Environmental policy research series. Beijing, China: China Environmental Science 
Press.


1  
China’s challenges in policy making
Policies and goals are important in any country’s governance, but their relative 
roles could have two primary patterns under different governance models. Rules 
are set up through policies (and laws), while polluters and other stakeholders 
decide on their own actions according to the rules. In the rule-­
based governance, 
policies are in the first place, while goals are more implicit to take the second 
place. Another strategy explicitly makes goals in the first place, while policies 
are secondary and could be more flexible. With the rule of law not yet well estab­
lished, China would face great challenges in policy supply under the rule-­
based 
governance, especially given its rapidly evolving economy and society.
1.1  
Uncertain linkages between actions and outcomes
China is rapidly industrializing and the economy grows at a fast pace. It encoun­
ters great uncertainties on whether planned actions could achieve intended goals. 
Sulfur dioxide (SO2) emissions as well as other environmental problems tend to 
have a wide scope of influential economic, energy and environmental factors, as 
well as scattered emission sources in numerous important sectors. Many key fac­
tors for SO2 mitigation are beyond the jurisdiction of environmental protection, 
specifically the Ministry of Ecology and Environment (and previously the Minis­
try of Environmental Protection). Implementation is largely under the responsibil­
ity of local governments, while the central government is not designed and well 
equipped for primary policy implementation. In addition, China’s complexities 
can cast substantial uncertainties on whether preplanned actions can achieve their 
goals. China identified enough efforts to achieve the 10% reduction goals of SO2 
emissions in both the 10th and the 11th Five-­
Year Plans, but their outcomes dif­
fered from each other dramatically. In the Outline of the National 10th Five-­
Year 
Plan that was ratified by the National People’s Congress, the 10% reduction goals 
of “major pollutants” were clearly written (National People’s Congress, 2001). 
“Major Pollutants” were later defined to include SO2, dust, COD (chemical oxy­
gen demand), ammonia-­
nitrogen and industrial solid waste (SEPA, 2001). Exter­
nal measures, particularly energy conservation, did not show up in the national 
5	

Policy making


78  Policy making
Outline (National People’s Congress, 2001). However, in the special plan for 
energy development, China did propose a goal to reduce energy intensity by about 
15% to 17% and coal’s share in total energy consumption by 3.88% in the five 
years (NDRC, 2001). China’s annual economic growth rate, another key factor, 
was estimated to be 7% (National People’s Congress, 2001). Between 2001 and 
2005, the reversely calculated sulfur contents in coal from China’s official data 
went down from 1.22% to 1.05% (Xu et al., 2009), and a lot more SO2 scrub­
bers were installed (Figure 5.11). For the 10th Five-­
Year Plan, the SO2 mitigation 
shortfall was mostly due to the unexpected surge in coal consumption as a result 
of accelerated economic growth, 87.6% over the five years that overwhelmed the 
efforts of the State Environmental Protection Agency (SEPA; BP, 2019).
In the 11th Five-­
Year Plan, the planning structure differed only slightly. With 
the same 10% reduction goal, the Outline of the National 11th Five-­
Year Plan nar­
rowed the definition of “major pollutants” to cover only SO2 and COD for primary 
attention (National People’s Congress, 2006). Other pollutants were addressed in 
the special plan for environmental protection (State Council, 2007b). A goal on 
energy conservation, a 20% reduction of energy intensity, got promoted to the 
national Outline. Coal’s share in total energy consumption remained in the special 
plan for energy development, with a 3% drop in the five years (NDRC, 2007). 
China’s economy was estimated, or conservatively planned, to grow 7.5% per 
year (National People’s Congress, 2006). These figures were quite close to those 
in the 10th Five-­
Year Plan. Simply from the planning perspective, these two 10% 
reduction goals of SO2 emissions should both be attained. However, their results 
diverged significantly away from each other, which illustrated the difficulty to 
foresee the effects of policies and actions on goals.
1.2  
Challenges in policy making to induce actions
In the U.S.’s efforts to control SO2 emissions, the Clean Air Act Amendments 
(1990) established the Acid Rain Program that was distinguished as the most 
important law on the issue (The U.S. Congress, 1990). However, no individual 
environmental policy in China could claim an equal share of importance in its SO2 
mitigation cause. In comparison to goal-­
centered governance, policy supply under 
rule-­
based governance features fewer policies (or laws), and some are of crucial 
importance in achieving the intended goals of environmental protection. Each 
policy has a more extended enactment procedure and implementation horizon, 
which makes policy-­
making process lengthy and careful. The failure/success of 
a key policy thus takes on much heavier weight in environmental protection out­
comes. Nevertheless, in developed countries where the rule of law is well estab­
lished, the linkages between policies and polluters’ actions are more predictable, 
while these actions further contribute to intended outcomes. However, because 
China has not established a sound rule of law, confidence is much lower that a 
policy can be implemented well to induce the intended actions.
Many environmental policy instruments have been designed and applied across 
countries. The first major category involves command and control policies, such 


Policy making  79
as mandatorily shutting down polluting sources, setting pollutant emission and 
energy efficiency standards and mandating the application of the best available 
technologies. Another major category is based on economic incentives and mar­
kets. Typical policy instruments include effluent emission discharge fees, taxes, 
tradable permits and subsidies. Information disclosure, such as labeling and cer­
tificates, aims to enable consumers to voluntarily make informed consumption 
choices for minimizing environmental impacts.
Despite some unique features, China’s policy toolbox for SO2 mitigation was 
not fundamentally different from that in developed countries with rule-­
based gov­
ernance. In China, an engineering approach that was based on SO2 scrubbers in 
coal-­
fired power plants involved many command and control policies for their 
deployment and normal operation to meet effluent emission standards (Minis­
try of Environmental Protection and General Administration of Quality Super­
vision Inspection and Quarantine, 2011; the SEPA and General Administration 

of Quality Supervision Inspection and Quarantine, 2003). China has also been 
experimenting with market-­
incentive policies, such as cap-­
and-­
trade, an effluent 
emission fee or an emission tax (Yan et al., 2009; Dong et al., 2011; Ge et al., 2011; 
Zhang et al., 2016). Technological licensing from developed countries, through a 
functioning technology market, was a cornerstone in China’s SO2 mitigation to 
build a domestic industry for rapid deployment and cost reduction (Xu, 2011).
After assessing the effectiveness and efficiency of individual environmen­
tal policy instruments, policies are enacted for tackling a given environmental 
problem (Barron and Ng, 1996; Goulder and Parry, 2008). A few criteria could 
be important in making an optimal policy, including cost-­
effectiveness; capabil­
ity to address uncertainty, synergy or conflict with current policy instruments; 
compliance monitoring and inspection capacity and requirements; and compli­
ance of polluters. The latter two are especially relevant to developing countries 
like China, where the rule of law has not been well established and environmen­
tal noncompliance might be prevalent. In developed countries, there has been 
an increasing trend in the application of market-­
based instruments (Portney and 
Stavins, 2000; Tietenberg, 1990). Cost-­
effectiveness is the most important argu­
ment for their adoption considering particularly the reduced abatement costs 
(Goulder and Parry, 2008). For example, in the U.S. Acid Rain Program in the 
Clean Air Act Amendments (1990), total SO2 emissions from coal power plants 
were capped and emission permits were allowed to trade in a market (The U.S. 
Congress, 1990). The policy substantially reduced the abatement costs compared 
with command-­
and-­
control instruments (Benkovic and Kruger, 2001).
Environmental policy instruments differ from each other in their capability of 
addressing uncertainties. For example, environmental taxes establish certain lev­
els of emission prices but leave the quantities of emissions uncertain. In contrast, 
tradable permits with a fixed cap are more certain about the quantity within the 
defined boundary of emission sources but not about the price. Other instruments 
all have various impacts on uncertainties (Goulder and Parry, 2008). The intro­
duction of a new environmental policy instrument should consider how it interacts 
with existing policies to create synergies or conflicts. If a new emission trading 


80  Policy making
policy is imposed into an area that is already dominated by command and control 
policies, it may not be able to achieve its intended cost-­
effectiveness (Zhang et al., 
2013). China’s experiments of SO2 emission trading schemes encountered major 
problems, including frequent governmental intervention and inter-­
policy conflicts 
together with the quality of policy design (Zhang et al., 2016).
Compared with developed countries, developing countries and specifically 
China have more difficulties in making optimal policies. Research tends to be 
thinner especially in the past to understand how individual policy instruments 
perform in their contexts. Significant constraints on environmental policy imple­
mentation may exist due to the lack of adequate financial resources, personnel 
and necessary expertise (Blackman, 2010). More details on China’s policy imple­
mentation problems are discussed in Chapter 6. The effectiveness of individual 
policies could be very uncertain with unpredictable implementation, which makes 
policy design challenging.
2  
Goal-­
centered policy supply
China’s policy supply follows a very different pattern from that in rule-­
based gov­
ernance. Goals play the central role in environmental governance, while policies 
as means to achieve goals are primarily instrumental and failures of individual 
policies are more accommodated. Important centralized goals as those few in 
National Five-­
Year Plans drive decentralized policies, laws and regulations from 
ministries, local governments, the People’s Congress and other stakeholders. For 
those environmental fields without goals or with goals but at lower priorities, 
policy supply tends to be less adequate and strong.
2.1  
Enabling goal-­
centered policy supply
Goal-­
centered governance in China is enabled by centralized national goals, 
decentralized goal attainment, decentralized policy making and implementation 
and mobilized central and local governments. In the past four decades, China’s 
environmental governance has been heavily decentralized, as discussed in Chap­
ter  3. The four levels of governments  – central, provincial, municipality and 
county – have diverging divisions of governmental authorities and functions. As 
matched by their personnel categories and fiscal expenditures, the central gov­
ernment heavily focuses on policy making, while the county-­
level governments 
are almost entirely on policy implementation. Provincial-­
 and municipality-­
level 
governments have significant authorities and functions on both. Local govern­
ments hold significant decentralized authorities in initiating local policy innova­
tion, learning and adopting policies from other regions and implementing various 
policies. Without the cooperation and mobilization of local governments, the cen­
tral government can hardly achieve serious SO2 mitigation or any environmental 
cleanup.
However, much authority remains substantially centralized, especially setting 
up national goals for environmental protection. Various considerations for or 


Policy making  81
against strong environmental protection are centrally weighed to form strong or 
weak political will by the top leadership of the Chinese Communist Party, as dis­
cussed in Chapter 2. It is then reflected in Five-­
Year Plans. When the top leader­
ship determines to prioritize environmental protection among other governmental 
affairs, pollution mitigation goals started to enter as the key goals into National 
Five-­
Year Plans. These national goals are then decomposed into provincial goals 
for their implementation, as examined in Chapter 4. The goal allocation further 
penetrates into municipality and county levels, one level at a time. The types and 
stringency of goals closely follow the centralized political will for environmental 
protection. In addition, ministries and their internal departments in the central 
government are also directed by those goals to make policies and supervise pro­
vincial and other local governments for goal attainment. If a crucial environmen­
tal goal in the Five-­
Year Plan is missed, the Ministry of Ecology and Environment 
as the primary responsibility bearer will also be held accountable.
Credible mechanisms are established for central and local governments to make 
efforts for their goals. Most important, provincial leaders and ministers in the cen­
tral government have their promotion opportunities controlled centrally through 
the Chinese Communist Party, while the fates of municipality-­
level leaders are 
determined at the provincial level. The clear linkages between their career devel­
opment and goal-­
centered job performance are crucial incentives to motivate their 
genuine efforts but not just lip service.
Under goal-­
centered governance, it is the succession of goals but not individual 
policies that define environmental milestones. The top leadership of the party and 
the central government cares more about whether a certain goal has been achieved 
rather than a certain policy has been effective, efficient or fully implemented. 
In addition, the fact that China has not established a sound rule of law is also 
an important facilitating factor for enabling a goal-­
centered policy supply. Local 
leaders in charge, such as provincial governors, municipality mayors and county 
leaders as well as their corresponding party secretaries, will be less likely to lose 
their jobs or promotion opportunities for having policy failures, but the probabil­
ity will increase significantly if a crucial goal does not get achieved. If one policy 
does not work, new ones will be quickly enacted to inch toward goal attainment.
2.2  
Policy evolution by implementation selection
Policy supply under goal-­
centered governance has two key components: environ­
mental goals to shape policy demand and low policy-­
making barriers and strong 
incentives to enable policy supply. More ambitious environmental goals will cre­
ate stronger demand for pollution mitigation actions and thus a larger number 
of and more stringent policies. Goal-­
centered governance significantly reduces 
barriers for making policies. The much lower policy-­
making barriers result in 
intensive policy-­
making activities, competition among policies and much faster 
policy cycles. With a significant number of policies, each makes a small step 
toward an intended goal, although some are more important than others. The fail­
ure/success of any policy does not determine, but only to a limited extent affects, 


82  Policy making
the final environmental outcome. Besides laws, a large number and wide variety 
of policies can be found in China on environmental protection that are enacted by 
various authorities, including the Central Committee of the Chinese Communist 
Party, the State Council, ministries and their composing departments (www.mee.
gov.cn/zcwj/) and local governments.
Several causes contribute to the low policy-­
making barriers. The significantly 
decentralized policy making effectively reduces the barriers from the perspective 
of policy suppliers as they have a wide variety of sectoral and geographic jurisdic­
tions and authorities. One consequence of this goal-­
centered policy supply is that 
it encourages policy innovation. Local governments have significant flexibility in 
deciding how to achieve top-­
down goals. Decentralized policy makers can weigh 
the significance, costs and benefits of various policies and their suitability to local 
contexts with dramatic regional disparities. Policies are constantly churned out 
from these decentralized policy makers at various levels to try their effectiveness 
in approaching goals. The effective mobilization of local governments not only 
facilitates policy enforcement, but it also creates incentives for even more active 
local environmental policy making if goal attainment so requires.
Furthermore, several key questions should be considered over the making of 
individual policies, while goal-­
centered governance has much lower require­
ments on policy designs to effectively decrease the policy-­
making barriers. 
First, how to ensure the quality of individual policies? Policies may be directly 
adopted from other countries and regions, revised to suit local contexts or inno­
vated from scratch. China’s colossal size and complexity indicate that many 
environmental policies can hardly be applied to fit all situations across the entire 
country. China’s contexts are also sharply different from those in developed 
countries, where many environmental policies were first introduced and imple­
mented. The decentralization of policy makers also indicates that the training 
and knowledge of those who write the policy texts may vary across local gov­
ernments and ministries/departments. The much more greatly decentralized pol­
icy implementation and its unsatisfactory track record add further difficulties in 
understanding how policies could be designed better for more effective imple­
mentation. Accordingly, direct policy adoption is rarely effective, while policy 
localization and innovation are great challenges and require relevant knowledge 
and understanding. In addition, China’s complexity also hinders timely-­
enough 
assessment of the crucial causes of any policy failure and success. Under goal-­
centered governance, the requirements on the quality of making individual poli­
cies are much lower because no policy or law occupies the central stage to solve 
a targeted environmental problem. The lower requirement for policy quality 
enables much swifter design and enactment processes. In other words, read­
ers of China’s environmental policies should not be primarily entangled in the 
enactment and effectiveness of individual policies, because they are of much 
less importance than goals. For example, essentially no SO2 emission trading 
schedules have produced desirable outcomes that dominate SO2 mitigation, but 
the failure had little impact on China’s trajectory of controlling SO2 emissions 
(Zhang et al., 2016).


Policy making  83
Second, how to choose the most effective and efficient policy instrument among 
many alternatives? The choice of policy instruments is a crucial question for pol­
icy making, especially when a single or very few policies dominate the solutions 
to an environmental problem. Under goal-­
centered governance, this question is 
much less significant because policies are much less mutually exclusive. The con­
siderably decentralized policy making also significantly reduces the possibility of 
any policy monopoly or oligopoly. The enactment of one policy instrument does 
not prevent the application of others. Accordingly, China does not need to choose 
a primary policy instrument for dealing with one environmental problem. For 
example, China’s environmental protection tax law formally entered into force 
in January 2018, covering a wide variety of environmental pollutants, including 
SO2 (National People’s Congress, 2016). Many other crucial environmental poli­
cies are simultaneously in effect, such as the effluent emission standards that were 
examined earlier (MEP and AQSIQ, 2011).
Third, how are policies coordinated? While policies are individually made by 
different ministries and their internal departments, as well as various levels of 
governments, they can exert significant impacts on each other to create synergies 
and/or conflicts. Economic and energy policies are far beyond the jurisdiction 
of environmental protection. With local governments rather than their environ­
mental protection bureaus in charge, coordination across these different types of 
policies became more feasible. In an optimized situation, policies should be well 
coordinated to maximize synergies and minimize conflicts. However, such coor­
dination in China is inadequate in the context of decentralized policy making 
and especially policy implementation. Little evidence indicates that China rolls 
out the numerous policies for achieving the SO2 mitigation goals in a system­
atic and coordinated way. Instead, the policy making is messy, with decentralized 
policy makers who have their individual authority in designing or shaping policies 
within their respective jurisdictions. Under goal-­
centered governance, however, 
such prior coordination of policy making is of lesser importance. After policies 
are made and put into implementation, they evolve rapidly. In China’s context of 
weak rule of law and as examined earlier, individual policies have higher prob­
abilities of unsatisfactory implementation. Similar to the natural selection process 
as proposed by Charles Darwin in understanding biological evolution (Darwin, 
1859), policies in China also experience a dynamic evolution process and those fit 
ones are selected through implementation. Policies that have too many conflicts 
with others will be difficult to get effectively implemented. If one policy fails 
to achieve its intended consequences, new policies can be quickly introduced. 
Successful policies in one province can be rapidly adopted by other provinces or 
elevated to the national level.
Although much progress has been made in policy research in the past decade, 
such capacity was especially deficient in the early stages of SO2 mitigation. China 
should still enhance its capability in policy making to improve the quality of indi­
vidual policies, choose more wisely environmental policy instruments especially 
for those of relatively greater importance and scopes and better coordinate across 
policies. Nevertheless, the goal-­
centered policy supply substantially lowered the 


84 Policy making
requirements for achieving desirable environmental protection outcomes such as 
serious mitigation of SO2 emissions. The preceding crucial questions in policy 
making are of much less concern from their perspectives on influencing policy 
outcomes.
3  
Policy scope for achieving SO2 mitigation goals
China faces a wide scope in policy making for SO2 mitigation. SO2 emissions 
are affected by many economic, energy and environmental development factors 
and corresponding policies. Although the coal-fired 
 
power sector is increasingly 
important in coal consumption, still nearly two fifths of coal is consumed in other 
sectors (Figure 1.10). For achieving increasingly stringent SO2 mitigation and 
environmental goals, the decentralized policy makers should evaluate the contri-
butions of individual policies in policy supply.
3.1   
Key factors for SO2 emissions
SO2 emissions can be decomposed with the following formula into various key 
factors:
Energy
Coal
SO2 emissions
SO2 emissions = GDP ×
×
×
GDP
Energy
Coal
Coal
  
Equation 5.1
= GDP ×
×
EI
×
×
h
h
s
s
(
)
1
2
−
×
r
R
×
−
(
)
1
h
Energy
“GDP” (gross domestic product) indicates the scale effect. Rapid economic 
growth in China leads to more SO2 emissions. Energy consumption is a key foun-
Energy 
dation for any modern economy, and thus, energy intensity 


 is another 



GDP 

crucial effect. It measures how much energy is consumed for producing a given 
unit of GDP. Energy conservation and efficiency will reduce energy intensity and 
thus be beneficial for SO2 mitigation. The economic structure also matters greatly. 
A greater proportion of service sectors in an economy could potentially reduce 
the overall energy intensity because in comparison to industrial sectors, they tend 
to consume much less energy for producing the same amount of economic out-
puts (Feng et al., 2009). China had a goal to reduce energy intensity by 20% in 
the 11th Five- 
Year Plan (National People’s Congress, 2006). The Chinese central 
government also declared its intention in the 12th Five- 
Year Plan to “change the 
economic growth pattern,” with a focus on energy conservation and environmen-
tal protection (National People’s Congress, 2011). These two effects are related to 
economic development and energy conservation, on which economic and energy 
policies exert important influences.
Because coal consumption dominates the sources of SO2 emissions, the share 
of coal in the energy mix is thus critical in deciding the sulfur intensity of energy. 


Policy making  85
Coal
Energy is referred to as the energy transition effect. Its reduction is another meas­
ure for bringing down SO2 emissions, which largely falls into the category of 
energy development and the scope of energy policy.
SO emissions
Coal
2  
 refers to the mitigation effect, which is primarily decided by 
environmental policies. In combustion, a certain proportion of sulfur (ηsr ) will 
be retained in ash and thus not emitted. This rate is mainly decided by the coal 
type and combustion technology, but not by policy intervention. Sulfur content in 
coal (ηs) is an important indicator of coal quality. The control of sulfur contents is 
often targeted in early environmental regulations for reducing SO2 emissions. SO2 
scrubbers and other SO2 removal measures can avoid a certain share of SO2 (ηR) 
from being emitted after generation.
China’s economy has been growing at an astonishing pace in the past four dec­
ades. Real GDP in 2018 was 31.7 times of that in 1980 with a growth rate of 9.5% 
annually, while real GDP per capita rose to be 22.4 times or 8.5% annually (Fig­
ure 5.1). As measured in nominal GDP of current U.S. dollars, China overtook 
Japan to become the second-­
largest economy in the world in 2010 and further rose 
to be equivalent to 65.0% of the United States in 2018 (Figure 5.1). China’s much 
larger population indicates that the country’s GDP per capita still trails the global 
average and is a small fraction of that in Japan and the United States. Although 
the GDP growth rate has been significantly slower in the 2010s than in the 2000s, 
0
5,000
10,000
15,000
20,000
25,000
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
1980
1985
1990
1995
2000
2005
2010
2015
Nominal GDP (Billion US dollars)
)
B
M
R
 
5
1
0
2
(
 
a
t
i
p
a
c
 
r
e
p
 
P
D
G
 
&
 
P
D
G
Year
Real GDP (Billion RMB, China; left)
Real GDP per capita (RMB per capita, China; left)
Nominal GDP (Billion US dollars, China; right)
Nominal GDP (Billion US dollars, Japan; right)
Nominal GDP (Billion US dollars, US; right)
Figure 5.1  
Economic growth in China, Japan and the United States
Source: IMF (2019).


86  Policy making
the convergence of average living standards in China toward that of developed 
countries is expected to further intensify economic activities within its geographi­
cal territory and thus to add great environmental pressures.
Energy consumption is not only one key foundation for economic develop­
ment, but it also brings unwanted consequences of environmental pollution. The 
combustion of fossil fuels, especially coal, is the primary source of air pollutant 
emissions that cause ambient particulate matter (PM) pollution. Although China 
has been improving its energy efficiency for producing one unit of GDP espe­
cially in the past decade, its primary energy consumption climbed up quickly. 
When consuming one ton of oil equivalent of primary energy, China in 2018 
produced US4,084ofnominalGDP,whiletheratesforJapanandtheUnitedStateswereUS10,948 and US$8,945, respectively (IMF, 2019; BP, 2019). Due 
to the significantly lower energy efficiency, China overtook the United States 
to become the largest energy consumer in the world in 2009, but its economy 
then was two thirds smaller. A major shift took place in around 2003, and since 
then, China’s energy consumption has been growing at a much faster pace than 
before (Figure 5.2). Not only China’s economic growth accelerated after 2003, 
but also the energy efficiency reversed its earlier improvement trend to decrease 
between 2002 and 2005 (Figure 5.2). In 2018, China consumed 224% more pri­
mary energy than in 2000 to become 42% higher than the United States’ level 
(Figure 5.2).
0.0
5.0
10.0
15.0
20.0
25.0
30.0
0
500
1,000
1,500
2,000
2,500
3,000
3,500
1980
1985
1990
1995
2000
2005
2010
2015
Energy efficiency (1,000 (2015) RMB/toe)
)
e
o
t
M
(
 
n
o
i
t
p
m
u
s
n
o
c
 
y
g
r
e
n
e
 
y
r
a
m
i
r
P
Year
China
India
United States
Energy efficiency in China (right)
Figure 5.2  
Primary energy consumption and energy efficiency
Source: BP (2019).


Policy making  87
China’s low level of GDP per capita might partly explain why its energy mix 
heavily focuses on coal. As shown in Figure 5.3, coal is the cheapest and most 
affordable among the three major fossil fuels. When China’s economy grew very 
fast, especially in the 2000s, to rapidly push up energy consumption, coal became 
the primary choice for meeting the additional energy demand (Figure 5.4), and 
thus, its share in the energy mix even reversed its earlier declining trend to become 
higher in the early 2000s (Figure 5.5).
In the past decade, energy transition has also been playing an increasingly vis­
ible role that led to the mitigation of SO2 emissions. China’s energy mix is heav­
ily tilted toward coal, the most pollution-­
intensive fuel. With more coal burning 
squeezed into China’s territory, the pressure on the environment is mounting. 
Throughout the 1980s and 1990s, the share of coal was continuously above 70% 
(Figure 5.5). The slow declining trend in the 1990s was reversed in early 2000s to 
witness the share climbing up again from 69.5% in 2001 to 73.7% in 2007, further 
intensifying environmental pollution in China. The following decade witnessed 
an unprecedented decrease and coal’s share had dropped to 58.2% in 2018. Never­
theless, China still accounted for 50.5% of global coal consumption in 2018 (BP, 
2019). Although oil and natural gas have increasing shares in China’s primary 
energy consumption, the overall share of fossil fuels experienced an accelerated 
decline from 94.1% in 2007 to 85.3% in 2018. Nonfossil fuels are much more 
0
2
4
6
8
10
12
14
16
18
20
1980
1985
1990
1995
2000
2005
2010
2015
)
J
G
/
$
 
t
n
e
r
r
u
c
(
 
e
c
i
r
p
 
y
g
r
e
n
E
Year
Oil
Gas
Coal
Figure 5.3  
Prices of coal (Qinhuangdao spot price), oil and natural gas
Source: Japan LNG CIF; BP (2019).


88  Policy making
–50
0
50
100
150
200
250
1980
1985
1990
1995
2000
2005
2010
2015
)
e
o
t
M
(
 
n
o
i
t
p
m
u
s
n
o
c
 
y
g
r
e
n
e
 
f
o
 
h
t
w
o
r
g
 
l
a
u
n
n
A
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Figure 5.4  
The annual growth of primary energy consumption in China by fuels
Source: BP (2019).
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
1980
1985
1990
1995
2000
2005
2010
2015
Share in the energy mix
)
e
o
t
M
(
 
n
o
i
t
p
m
u
s
n
o
c
 
y
g
r
e
n
e
 
y
r
a
m
i
r
P
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Coal’s share
Fossil fuels’ share
Figure 5.5  
China’s primary energy consumption by fuel and the shares of coal and fossil fuels
Source: BP (2019).


Policy making  89
important in the energy mix, from 5.9% in 2007 to 14.7% in 2018 (Figure 5.5). 
Nuclear, hydropower and nonhydro renewables witnessed their shares increased 
from 0.7%, 5.1% and 0.2% in 2007 to 2.0%, 8.3% and 4.4% in 2018, respectively. 
Nonhydro renewables were the fastest-­
growing energy type.
As one indicator of energy modernization, China’s primary energy consump­
tion is rapidly electrifying to reshape the major sources and sectors of SO2 emis­
sions. In 1990, only 20.8% of primary energy consumption went through the 
intermediate stage of electricity before final consumption, which was only slightly 
higher than Africa’s 17.8% (Figure 5.6). With rapid energy modernization, this 
ratio increased to 42.5% in 2015, then similar to the United States’ 40.3% (Fig­
ure 5.4). Rapid electrification also happened in other rapidly industrializing coun­
tries such as India, but the progress in Africa has been much slower (Figure 5.4). 
With China’s continuous efforts for electrifying energy consumption – such as the 
push for electric vehicles (IEA, 2019) – this electrification rate is expected to fur­
ther escalate, which will distinguish the importance of the power sector in China’s 
energy consumption and environmental protection.
Energy transition for electricity generation is even more visible. Coal’s share 
has been reduced significantly from the 81.0% peak in 2007 to 66.5% in 2018. 
Other fossil fuels, including oil and natural gas, accounted for only an insignificant 
share at 3.3% in 2018 (Figure 5.7). In contrast, the share of nonhydro renewables, 
mostly wind and solar energy, has achieved the largest growth from 0.5% in 2007 
to 8.9% in 2018 (Figure 5.7). Coal’s share in electricity generation is significantly 
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
0
500
1,000
1,500
2,000
2,500
3,000
1990 2000 2015
1990 2000 2015
1990 2000 2015
1990 2000 2015
Share of power generation
)
e
o
t
M
(
n
o
i
t
p
m
u
s
n
o
c
y
g
r
e
n
E
Power generation
Others
Share of power generation
United States
China
Africa
India
Figure 5.6  
Primary energy consumption and its electrification rate
Source: IEA (2017).


90  Policy making
higher than that in primary energy consumption, being 66.5% and 58.2% in 2018, 
respectively (Figure 5.8). From 2007 to 2018, their drops were 14.4% and 15.4% 
in percentage points, respectively. Nonfossil fuels – such as nuclear, hydro and 
nonhydro renewables – are generally for electricity generation, while oil and natu­
ral gas in China are primarily consumed not in the power sector. Especially in the 
past decade, the advancement of renewables significantly accelerated to account 
for increasingly sizable shares of electricity generation growth (Figure 5.8).
The power sector has been increasing its importance in coal consumption. In 
1980, only 20.2% of China’s coal consumption was in the power sector, while 
this ratio climbed steadily to 52.2% in 2002 before a decade-­
long stabilization 
(Figure 1.10). During 2015~2017, the increasing trend restarted to reach 57.3% in 
2017 from 50.3% in 2014 (Figure 1.10). This ratio is expected to further increase, 
in reference to the situation in the United States, whose power sector accounted 
for 18.6% of coal consumption in 1950 and 92.8% in 2017 (Figure 1.10). The 
trend indicates that the energy mix in nonpower sectors shifts away from direct 
coal consumption faster than that in the power sector, although the former may 
consume more electricity that comes from coal-­
fired power plants.
In China’s trajectory of SO2 mitigation, these economic, energy and envi­
ronmental factors made different contributions in different Five-­
Year Plans 
(Figure  5.9). SO2 emissions went down by 15.8% in the 9th Five-­
Year Plan 
50%
55%
60%
65%
70%
75%
80%
85%
90%
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
1985
1990
1995
2000
2005
2010
2015
Share in the energy mix
)
h
W
T
(
 
n
o
i
t
a
r
e
n
e
g
 
y
t
i
c
i
r
t
c
e
l
E
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Others
Fossil fuels’ share
Coal’s share
Figure 5.7  
Electricity generation by fuels in China
Source: BP (2019).


Policy making  91
50%
55%
60%
65%
70%
75%
80%
85%
–50
50
150
250
350
450
550
650
1985
1990
1995
2000
2005
2010
2015
Coals shares
)
h
W
T
(
 
n
o
i
t
a
r
e
n
e
g
 
y
t
i
c
i
r
t
c
e
l
e
 
f
o
 
h
t
w
o
r
g
 
l
a
u
n
n
A
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Others
Coal’s share in electricity generation
Coal’s share in primary energy
Figure 5.8  
The annual growth of electricity generation in China by fuels and coal’s share
Source: BP (2019).
–60%
–50%
–40%
–30%
–20%
–10%
0%
10%
20%
30%
40%
50%
60%
SO2 emissions
Scale effect
Energy intensity
effect
Energy transition
effect
Mitigation effect
e
v
i
F
 
r
e
v
o
 
e
g
n
a
h
C
-
 
s
n
a
l
P
 
r
a
e
Y
9th FYP (1996–2000)
10th FYP (2001–2005)
11th FYP (2006–2010)
12th FYP (2011–2015)
13th FYP (2016–2017)
Figure 5.9  
The decomposition of China’s SO2 emissions
Source: National Statistics Bureau and Ministry of Ecology and Environment (2019); IMF (2019); 
BP (2019).
Note: Method comes from Ang (2005).


92 Policy making
(1996–2000). Under the influence of the Asian financial crisis of 1997, the scale 
effect would still lead to an increase of 38.0%, while the effects of energy inten-
sity, energy transition and mitigation reduced SO2 emissions by 26.4%, 6.1% and 
21.3%, respectively, over the five years (Figure 5.9). They indicated the varying 
impacts of economic, energy and environmental policies and development. Spe-
cifically, as represented in the mitigation effect, environmental policies made an 
important but certainly not decisive contribution. The 10th Five-Y
 ear Plan had a 
very different picture. With accelerated economic growth, the scale effect would 
boost emissions by 53.0%, while the energy intensity and energy transition effects 
also pushed the emissions upward by 12.6% and 5.7%, respectively. Although 
the mitigation effect of 43.5% reduction was much greater than that in the 9th 
Five- 
Year Plan, the outcome was that SO2 emissions increased by 27.8%. In other 
words, the deterioration was not due to less effective environmental policies, but 
faster economic expansion reversed trends of energy intensity and transition.
Reversing the deterioration trend, the 11th Five- 
Year Plan managed to reduce 
SO2 emissions by 14.3%. The following 12th Five- 
Year Plan registered a similar 
reduction of 14.9%. The four effects also had comparable contributions in these 
two Five- 
Year Plans: the scale effect 49.6% versus 35.0%, the energy intensity 
effect −19.6% versus −17.6%, the energy transition effect −4.2% versus −9.1% 
and the mitigation effect −40.0% versus −23.2% (Figure 5.9). In the first two 
years (2016–2017) of the 13th Five- 
Year Plan with available data, SO2 emissions 
dropped by whopping 52.9%, and the mitigation effect contributed decisively a 
reduction of 52.0% (Figure 5.9).
3.2   
Technological factors for effluent SO2 emissions  
in the power sector
Electrification of energy consumption and the power sector’s increasing share in 
coal consumption distinguish the importance of coal-fired 
 
power plants in control-
ling China’s SO2 emissions. Given the understanding of coal combustion and SO2 

SO  emissions

emissions in electricity generation, the SO  emission intensity 
2

2

 


Coal



can be converted to effluent SO2 concentration, mg/Nm3 (Ministry of Environ-
mental Protection and General Administration of Quality Supervision Inspection 
and Quarantine, 2011). The SO2 concentration is measured under normal condi-
tions (thus “N”), with excess air coefficient being 1.4. The value 1.4 here indi-
cates that 40% more air or, specifically, oxygen will be blown into boilers than 
what is required for complete combustion. Excess air is needed for more complete 
combustion within a short residence time of fuels in boilers, but excess air will 
also take heat away and lower the thermal efficiency. Accordingly, an optimum 
value exists, not necessarily being 1.4 for every power plant. The fixed value is 
for policy purposes and intends to prevent cheating because a convenient option 
of lowering effluent SO2 concentration is to dilute the flue gas with more air. 
21%
Approximately, excess air coefficient a can be calculated as a ≈
, x% 
21%
%
−x
referring to the percentage of O2 in flue gas, when a = 1.4 and x%
%
»
≈6 .
a
−
21
21
%
%
%
x


Policy making  93
As revealed in Equation 5.1, there are three key technological factors to decide 
SO2 emission intensity and effluent SO2 concentration. The first factor is the frac­
tions of sulfur retained in ash (ηsr). When coal is burned in boilers, sulfur is con­
verted into several forms, being gaseous (SO2, SO3, gaseous sulfates) and solid 
(in bottom ash and as particulate sulfate; EPA, 1998). SO2 greatly dominates the 
gaseous forms (EPA, 1998). Higher combustion temperature leads to lower frac­
tions of sulfur retained in ash, which disadvantages pulverized coal (PC) combus­
tion against fluidized bed combustion (FBC; Sheng et al., 2000). Another factor is 
the calcium/sulfur (Ca/S) molar ratio in coal: a higher Ca/S ratio facilitates sulfur 
retention (Cheng et al., 2004; EPA, 1998). (The Ca/S ratio here is different from 
the Ca/S ratio for SO2 scrubbers as discussed later.) For PC combustion, the pres­
ence of calcium is much less important than FBC due to the thermal instability of 
calcium sulfate (CaSO4), the main product responsible for sulfur retention (Sheng 
et al., 2000). The fractions of sulfur retained in the application are summarized 
in Table 5.1. Compared with China’s assumption of 20% (State Council, 2007a), 
fractions of sulfur retained are widely believed to be significantly lower, except 
for the situations of burning lignite and applying FBC technologies. Even another 
Chinese official document believed the rate to be 10% to 15% and recommended 
10% for the purpose of designing SO2 scrubbers (NDRC, 2004). China’s SO2 
emissions could have been underestimated partly because of the choice of this 
parameter (Figure 1.8).
Second, lower sulfur contents in coal are crucial for reducing SO2 generation 
intensity. An official data set is employed to analyze the distribution of sulfur 
contents for SO2 scrubbers. In the 11th Five-­
Year Plan on Acid Rain and SO2 
Pollution Control, China published data for 248 coal power plants with a total 
capacity of 164 GW that covered all SO2 retrofit projects to be completed between 
2006 and 2010 (SEPA and NDRC, 2008). Sulfur contents estimated from this data 
Table 5.1 Applied fractions of sulfur retained in ash
Coal type
Fractions of 
Source
sulfur retained 
in ash
Bituminous, PC*
5%
U.S. Environmental Protection (EPA, 1998)
Sub-bituminous, PC* 12.5%
Agency’s (EPA’s) choice
Lignite, PC*
25%
Coal in general
≤10%
U.S.’s study
(Singer, 1981)
Nonlignite coal
5%
Assumption from the U.S. EPA (Smith et al., 2001)
Lignite
30%
Assumption in the research
Coal
5%~30%
The study’s assumption
(Ohara et al., 2007)
Coal
5%~10%
China’s study
(Zhao et al., 2008)
Coal, PC*
10%~15%
China’s official recommendation (NDRC, 2004)
for scrubber design
Coal
20%
Assumption in compiling 
(State Council, 2007a)
China’s statistical data
Note: PC* refers to pulverized coal power plants.


94  Policy making
set are expected to represent their distribution for all coal power plants with SO2 
scrubbers. Each plant disclosed information on scale (MW), year, annual SO2 
removal capability (tons per year), location and name. Sulfur contents can then 
be estimated under the following assumptions: thermal efficiencies were 370 g 
of coal equivalent per kilowatt-­
hour, or 1,930 kWh per ton coal (the average effi­
ciency in 2005 [China Electricity Council, 2006–2015]); capacity factors were 
5,500 hours per year (SEPA, 2006a); 80% of the sulfur was converted to SO2 in 
combustion and 20% was retained in ash, as recognized in China’s official sta­
tistics (State Council, 2007a); and overall SO2 removal rates were 85% (SEPA, 
2007). The calculation formula is
Sulfur content
 
SO  removal capability
Coal power capacity
2

55
1930
8
85
00
2
0
 

%
%
“2” in the denominator refers to the fact that when sulfur is converted to SO2, the 
mass doubles since the molecular weight of SO2 is twice that of sulfur. A caveat 
is that these numbers are used here to reversely calculate sulfur contents because 
they represent China’s original assumption in compiling the data. One legitimate 
concern is the accuracy of the assumed 80% conversion. Actually, in the com­
bustion of anthracite, bituminous and sub-­
bituminous coal, 90% or more of the 
sulfur is converted to SO2, as discussed earlier. In addition, as discussed in Chap­
ter 6, China’s actual SO2 removal rates should be significantly lower than 85% 
especially before 2007. Actual thermal efficiencies and capacity factors also vary 
across years.
Sulfur contents are closely related to the costs of SO2 mitigation. Generally 
speaking, higher sulfur contents correspond to lower costs for every ton of SO2 
removed but higher costs for every kilowatt-­
hour of electricity generated. The 
distribution of sulfur contents is shown in Figure 5.10 with the national average 
being about 1.0%. Of the coal-­
fired power plants, 68% burned coal with less 
than 1% sulfur and another 26% between 1% and 2%. The remaining 6% of the 
total capacity was associated with higher than 2%-­
sulfur coal. China not only 
installed SO2 scrubbers not only in coal power plants burning high-­
sulfur coal 
but also in those burning low-­
sulfur coal. China’s distribution of sulfur contents 
had a single peak at around 0.75% (Figure 5.10), which reflected the fact that 
most of China’s coal is mined in one region. For example, two thirds of China’s 
coal production in 2007 came from the seven nearby provinces of Shanxi, Inner 
Mongolia, Shaanxi, Shandong, Anhui, Hebei and Henan (National Bureau of 
Statistics, 1997–2008).
The preceding two factors decide how much SO2 is generated when burning 
a unit quantity of coal. SO2 removal rates are the third factor to reduce the SO2 
emission intensity. Before construction begins, a report of environmental impact 
assessment (EIA) had to be submitted to a governmental authority on environ­
mental protection (NPC, 2002). If the plant was believed to bring unacceptable 
environmental damage – for example, seriously worsen ambient air quality – the 


Policy making  95
EIA report would be rejected. Another policy – “three simultaneities” – required 
pollution control facilities to be designed, constructed and completed at the same 
time as the main project (State Council, 1998). For example, if SO2 scrubbers 
were considered necessary in the EIA report, this policy would demand their 
installation.
3.3  
Technical measures for SO2 removal in the power sector
In order to remove SO2 in electricity generation, coal-­
fired power plants in China 
are required to meet effluent emission standards that are made more stringent 
every six or seven years to reflect growing environmental concerns. In the stand­
ards enacted in 1996, new coal-­
fired power plants that passed EIA after Janu­
ary 1997 should achieve 2,100 mg/Nm3 (if burning coal with ≤1% sulfur) or 1,200 
mg/Nm3 (if burning coal with >1% sulfur; SEPA and AQSIQ, 1996). For coal-­
fired power plants burning bituminous coal with 0.5% sulfur, SO2 concentration 
in the non-­
desulfurized flue gas will generally exceed 1,000 mg/Nm3. Essentially 
the 1996 effluent emission standards meant that coal-­
fired power plants burning 
coal with >1% sulfur should have SO2 scrubbers while those with ≤1% sulfur 
did not need to. In the standards enacted in 2003, for the great majority of coal 
power plants, their effluent SO2 concentration should be kept below 400 mg/Nm3 
0%
5%
10%
15%
20%
25%
Share of SO2
y
t
i
c
a
p
a
c
 
r
e
b
b
u
r
c
s
Sulfur Content
Figure 5.10  
Distribution of sulfur contents in coal power plants in China (with retrofitted 
SO2 scrubbers)
Source: SEPA and NDRC (2008).


96  Policy making
on 1 January 2010 (SEPA and General Administration of Quality Supervision 
Inspection and Quarantine, 2003). In addition to China’s shutting down old, small 
power-­
generating units, the effluent emission standard itself would ensure that a 
dominant share of China’s coal power capacity in 2010 would have SO2 scrubbers 
installed and operate normally.
The standards were updated in 2011 for being effective on 1 January 2012 (Min­
istry of Environmental Protection and General Administration of Quality Supervi­
sion Inspection and Quarantine, 2011). New plants should then reduce their effluent 
SO2 emissions below 100  mg/Nm3 while the standard for existing plants was 
200 mg/Nm3. In southwestern provinces, including Guangxi, Chongqing, Sichuan 
and Guizhou, where local coal contains much higher sulfur contents, the standards 
could be relaxed to 200 mg/Nm3 and 400 mg/Nm3, respectively. Natural gas–fired 
power plants tend to be much cleaner, with the standard being 35 mg/Nm3.
In 2014, a new policy, “Upgrading and Retrofitting Action Plan for Energy Con­
servation and Pollution Mitigation in the Coal-­
Fired Power Sector,” was enacted 
jointly by National Development and Reform Commission, Ministry of Environ­
mental Protection and National Energy Administration (National Development 
and Reform Commission et al., 2014). It required newly constructed coal-­
fired 
power plants in eastern provinces to achieve the standard for natural gas–fired 
power plants, that is, 35 mg/Nm3 for SO2 emissions. Central provinces should 
approach this standard, while western provinces were encouraged to reach the 
level. This much more stringent standard is referred to in China as the ultra-­
low 
emissions. In 2015, another policy mandates the ultra-­
low standard to be achieved 
in most new and existing coal-­
fired power plants with only occasional exceptions 
(Ministry of Environmental Protection et al., 2015).
China’s Law of Standardization and its implementation regulations provide 
legal teeth (NPC, 1988; State Council, 1990). Effluent emission standards are 
clearly stated as “mandatory standards” (State Council, 1990), while products 
not meeting “mandatory standards” are forbidden to produce, sell and import 
(NPC, 1988). In this sense, coal power plants should stop generating electricity 
if the effluent SO2 emissions exceeded corresponding standards. The electric grid 
should not accept the electricity if it were not legally generated.
In order to achieve SO2 removal rates as required by the stringent ultra-­
low efflu­
ent emission standard, coal-­
fired power plants should generally achieve very high 
SO2 removal rates, being 98.5% if Huolinhe lignite or Datong bituminous coals are 
burned or 96.9% for Shenfu bituminous coal (Table 5.2). The SO2 emission intensity 
of electricity generation should also be substantially reduced to about 0.10 to 0.11 
g/kWh. The sulfur contents in these three types of coal, from 0.50% to 0.99%, fall 
within the normal range. For high-­
sulfur coal, especially in southwestern provinces, 
the required SO2 removal rates are much higher, generally beyond 99%. The deep 
reduction can only be achieved through SO2 scrubbers if coal remains as the fuel.
Before China started the large-­
scale deployment of SO2 scrubbers in the early 
2000s, the world in total had installed about 200 GW (Taylor et al., 2005). The 
United States accumulated around 100 GW in a 25-­
year period between 1975 and 
2000 (Taylor et al., 2005). Germany and Japan together took 30% of the world’s 


Policy making  97
market, and the remaining 20% were in other countries (Taylor et al., 2005). The 
scrubber capacity numbers presented in Figure 5.11 were calculated from a pub­
licly available plant-­
level data set (Ministry of Environmental Protection, 2014). 
The dataset includes information on the name and location of coal power plants, 
the serial number and power capacity of generators, the dates that generators and 
Table 5.2 Effluent SO2 emissions and necessary SO2 removal rates
Huolinhe 
 
Datong 
Shenfu 
lignite
bituminous
bituminous
LHV (MJ/kg)
13.9
21.0
21.4
Contents in coal (%)
Sulfur
0.61%
0.99%
0.50%
Carbon
34.1%
55.7%
57.0%
Hydrogen
2.7%
3.4%
3.4%
Oxygen
10.5%
8.3%
8.0%
Nitrogen
0.7%
0.9%
1.1%
Effluent SO2 emissions 
Concentration (mg/Nm3)
2,315
2,291
1,133
(without removal)
Emissions (g/kWh)
6.79
7.27
3.60
For achieving the 
Required SO2 removal 
98.5%
98.5%
96.9%
35 mg/Nm3 standard
rate (%)
Emissions (g/kWh)
0.10
0.11
0.11
Note: Assuming the sulfur retention rate in ash, 90%; thermal efficiency of electricity generation 
(42%, or 293 g of coal equivalent/kWh). Coal quality data are from Shi and Yu (2005).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
100
200
300
400
500
600
700
800
2000
2002
2004
2006
2008
2010
2012
Share of coal-fired power capacity 
)
W
G
(
 
y
t
i
c
a
p
a
C
Year
Coal-fired power capacity
SO2 scrubber capacity
Share of coal-fired power capacity with SO2 scrubbers (right)
Figure 5.11  
Coal-fired power and SO2 scrubber capacities in China
Source: Ministry of Environmental Protection (2014); EIA (2019); China Electricity Council (2010, 
2006–2015).


98  Policy making
SO2 scrubbers came online, SO2 scrubber technology type and the name of the 
SO2 scrubber company in charge. The SEPA (Ministry of Environmental Pro­
tection after March 2008) established a standard procedure for registering SO2 
scrubbers (SEPA, 2005, 2006b); for example, an SO2 scrubber had to operate 
continuously for 168 hours to test its performance before registration.
China’s share was negligible with only 5.6 GW of SO2 scrubbers at the end of 
2000, or 2.5% of its coal-­
fired power capacity (Figure 5.11). In the 10th Five-­
Year 
Plan (2001–2005), SO2 scrubber capacity rose to 46.7 GW in 2005. The progress 
was noticeable and the proportion of coal-­
fired power capacity with SO2 scrub­
bers increased to 12.5%. However, because total coal-­
fired power capacity esca­
lated from 218.9 GW in 2000 to 360.6 GW in 2005, essentially China had more 
coal-­
fired power plants without SO2 scrubbers to witness a steady increase of the 
power sector’s SO2 emissions. In the 11th Five-­
Year Plan (2006–2010), coal-­
fired 
power capacity grew at a much faster pace to reach 654.3 GW in 2010, while SO2 
scrubber capacity was lifted even faster to 569.3 GW in 2010. Then only 85.0 
GW of coal-­
fired power plants, or 13.0%, did not have SO2 scrubbers. This rate 
had already been much lower than that in 2000. In the following years, the ratio 
further inched higher to 94.4% in 2013. Essentially, in China, nearly all coal-­
fired 
power plants should have SO2 scrubbers to continue operation.
The 11th Five-­
Year Plan witnessed a large-­
scale campaign to retrofit existing 
coal-­
fired power plants (Figure  5.12), besides shutting down many inefficient 
0%
50%
100%
150%
200%
250%
300%
0
30
60
90
120
150
2000
2002
2004
2006
2008
2010
2012
Ratio
)
W
G
(
 
h
t
w
o
r
g
 
y
t
i
c
a
p
a
c
 
l
a
u
n
n
A
Year
Coal-fired power capacity
SO2 scrubber capacity
Ratio between SO2 scrubber and
coal-fired power capacities (right)
Figure 5.12  
The annual growth of coal-fired power and SO2 scrubber capacities in China
Source: Ministry of Environmental Protection (2014); EIA (2019); China Electricity Council (2006–
2015, 2010).


Policy making  99
small units (Xu et al., 2013). The ratios between the annually increased capacities 
of SO2 scrubbers and coal-­
fired power plants were consistently higher than 100% 
in every year of the 11th Five-­
Year Plan. At the retrofitting peak in 2008, SO2 
scrubber capacity grew by 127.4 GW while coal-­
fired power capacity increased 
only by 43.1 GW. After the 12th Five-­
Year Plan, a great majority of SO2 scrubbers 
were either built together with coal-­
fired power plants or further retrofitted for 
meeting more stringent effluent emission standards.
Most SO2 scrubbers fall into three scale categories: 300 MW, 600 MW and 
1000 MW (Figure  5.13). They correspond to several predominant, standard­
ized unit scales that China’s coal-­
fired power units have. Among the 754.9 GW 
of coal-­
fired power units with SO2 scrubbers in 2013, 62.1 GW, 215.1 GW and 
263.9 GW were within the 1,000~1,050-­
MW, 600~650-­
MW and 300~350-­
MW 
ranges, respectively. Two smaller scales have seen their importance fading after 
China focused more on larger and more efficient units. Respectively, 42.3 GW and 
32.7 GW fell within the 200~220-­
MW and 135~150-­
MW ranges. In total, these 
five standardized unit sizes accounted for 618.6 GW or 81.9% of all SO2 scrub­
bers. These size and technology standardization provided one crucial advantage 
in designing and rapidly deploying SO2 scrubbers.
The geographic distribution of SO2 scrubbers reflects that of coal-­
fired power 
plants. Provinces in East China, North China and South China had 249.2 GW, 
213.0 GW and 104.1 GW (or 33.0%, 28.2% and 13.8%) of SO2 scrubbers, 
0
20
40
60
80
100
120
140
Before
2001
2003
2005
2007
2009
2011
2013
)
W
G
(
 
y
t
i
c
a
p
a
C
Year
>=1000 MW
600 MW ~ 999 MW
300 MW ~ 599 MW
200 MW ~ 299 MW
100 MW ~199 MW
< 100 MW
Figure 5.13  
Annually increased SO2 scrubber capacity and unit sizes
Source: Ministry of Environmental Protection (2014).


100  Policy making
respectively, in 2013 (Figure  5.14). Northeast, Southwest and Northwest had 
188.5 GW in total, or 25.0%. Their vast geographic territories indicate that these 
coal-­
fired power plants are scattered at much greater distances from each other 
to potentially enhance difficulties for environmental compliance monitoring and 
enforcement.
Different SO2 scrubber technologies correspond to a wide range of possible 
SO2 removal rates. China had 1589 units of SO2 scrubbers at or above 200 
MW in 2013. The limestone-­
gypsum wet type is the most applied technol­
ogy especially for large coal-­
fired power units, accounting for 93.6% of units 
>=1000 MW, 96.1% of those between 600 MW and 999 MW, 87.6% of those 
between 300 MW and 599 MW and 81.0% of those between 200 MW and 
299 MW (Figure 5.15). The share dropped significantly for units smaller than 
200 MW, being only 30.0% (Figure 5.15). Due to the same consideration of 
economy of scale, seawater type also heavily tilted toward large units (Fig­
ure 5.15). Only 94.5 GW of SO2 scrubbers in 2013 were individually smaller 
than 200 MW (12.5% of all SO2 scrubbers), but they had 2,878 units (64.4% 
of all; Figure 5.15).
0%
10%
20%
30%
40%
50%
60%
0
20
40
60
80
100
120
140
Before
2001
2003
2005
2007
2009
2011
2013
Proportion as retrofit
)
W
G
(
 
y
t
i
c
a
p
a
C
Year
East
South
Southwest
Northwest
North
Northeast
Porportion as retrofit (right)
Figure 5.14  
The annual growth of SO2 scrubber capacity by regions (as categorized by the 
six Regional Supervision Bureaus of the Ministry of Ecology and Environ­
ment; SO2 scrubbers are called “retrofits” when the online dates of SO2 scrub­
bers and coal power units are over one year)
Source: Ministry of Environmental Protection (2014).


Policy making  101
Reference
Ang, B. W. 2005. The LMDI approach to decomposition analysis: A  practical guide. 
Energy Policy, 33, 867–871.
Barron, W. F. & Ng, G. T. L. 1996. An assessment methodology for environmental policy 
instruments: An illustrative application to solid wastes in Hong Kong. Journal of Envi­
ronmental Management, 48, 283–298.
Benkovic, S. R. & Kruger, J. 2001. US sulfur dioxide emissions trading program: Results 
and further applications. Water Air and Soil Pollution, 130, 241–246.
Blackman, A. 2010. Alternative pollution control policies in developing countries. Review 
of Environmental Economics and Policy, 1–20.
BP. 2019. Statistical review of world energy [Online]. Available: www.bp.com/en/global/
corporate/energy-­
economics/statisticalreview-­
of-­
world-­
energy.html.
Cheng, J., Zhou, J. H., Liu, J. Z., Cao, X. Y., Zhou, Z. J., Huang, Z. Y., Zhao, X. & Cen, 
K. 2004. Physicochemical properties of Chinese pulverized coal ash in relation to sulfur 
retention. Powder Technology, 146, 169–175.
China Electricity Council. 2006–2015. Annual report of national power generation. Bei­
jing, China: China Electricity Council.
China Electricity Council. 2010. The basic information of electricity generation units and 
electric grids. Beijing, China: China Electricity Council.
Darwin, C. 1859. On the origin of species by means of natural selection, or the preserva­
tion of favoured races in the struggle for life. London: J. Murray.
0
500
1,000
1,500
2,000
2,500
3,000
0
50
100
150
200
250
300
>=1000
MW
600 MW ~
999 MW
300 MW ~
599 MW
200 MW ~
299 MW
100 MW
~199 MW
< 100 MW
Number of units
)
W
G
(
 
y
t
i
c
a
p
a
C
Unit Size
Limestone-Gypsum Wet Type
Seawater Type
Others
Number of units
Figure 5.15  
SO2 scrubbing technologies by unit sizes
Source: Ministry of Environmental Protection (2014).


102  Policy making
Dong, Z., Ge, C., Gao, S. & Wang, J. 2011. Assessment on the Chinese pollution charges 
system and proposals to the future reform. In: Shen, M., Ge, C., Dong, Z. & Zhang, B. 
(eds.) Progress on environmental economics. Beijing, China: China Environmental Sci­
ence Press.
EIA. 2019. International energy statistics. Washington, DC: U.S. Energy Information 
Administration.
EPA. 1998. AP 42, chapter 1: External combustion sources, fifth ed. Washington, DC: 
EPA.
Feng, T. W., Sun, L. Y. & Zhang, Y. 2009. The relationship between energy consumption 
structure, economic structure and energy intensity in China. Energy Policy, 37, 5475–5483.
Ge, C., Ren, Y., Gao, S., Sun, G. & Long, F. 2011. Wastewater discharge tax design: From 
pollutant discharge levy to environmental tax. In: Shen, M., Ge, C., Dong, Z. & Zhang, 
B. (eds.) Progress on environmental economics. Beijing, China: China Environmental 
Science Press.
Goulder, L. H. & Parry, I. W. H. 2008. Instrument choice in environmental policy. Review 
of Environmental Economics and Policy, 2, 152–174.
IEA. 2017. World energy outlook. Paris, France: IEA.
IEA. 2019. Global EV outlook 2019. Paris, France: IEA.
IMF. 2019. World economic outlook database October 2019. Washington, DC: IMF.
MEP & AQSIQ. 2011. Emission standard of air pollutants for thermal power plants. Bei­
jing, China: MEP & AQSIQ.
Ministry of Environmental Protection. 2014. The list of China’s SO2 scrubbers in coal-­
fired 
power plants. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection & General Administration of Quality Supervision 
Inspection and Quarantine. 2011. Emission standard of air pollutants for thermal power 
plants. GB 13223-­
2011. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection, National Development and Reform Commission & 
National Energy Administration. 2015. Comprehensive implementation of the upgrading 
and retrofitting action plan for energy conservation and pollution mitigation in the coal-­
fired power sector. Beijing, China: Ministry of Environmental Protection.
National Bureau of Statistics. 1997–2008. China energy statistical yearbook. Beijing, 
China: China Statistics Press.
National Development and Reform Commission, Ministry of Environmental Protection & 
National Energy Administration. 2014. Upgrading and retrofitting action plan for energy 
conservation and pollution mitigation in the coal-­
fired power sector (2014–2020). Bei­
jing, China: Ministry of Environmental Protection.
National People’s Congress. 2001. The outline of national 10th five-­
year plan on economic 
and social developments. Beijing, China: The 4th Conference of the 9th National Peo­
ple’s Congress.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. Environmental protection tax law of the people’s repub­
lic of China. Beijing, China: The 4th Conference of the 10th National People’s Congress.
National Statistics Bureau & Ministry of Ecology and Environment. 2019. China statisti­
cal yearbook on environment 2018. Beijing, China: China Statistics Press.


Policy making  103
NDRC. 2001. Key special 10th five-­
year plan on energy development. Beijing, China: 
NDRC.
NDRC. 2004. Technical code for designing flue gas desulfurization plants of fossil fuel 
power plants. DL/T5196-­
2004. Beijing, China: NDRC.
NDRC. 2007. 11th five-­
year plan on energy development. Beijing, China: NDRC.
NPC. 1988. The standardization law of the people’s republic of China. Beijing, China: 7th 
National People’s Congress of the People’s Republic of China.
NPC. 2002. Environmental impact assessment law of the people’s republic of China. Bei­
jing, China: 9th National People’s Congress of the People’s Republic of China.
Ohara, T., Akimoto, H., Kurokawa, J., Horii, N., Yamaji, K., Yan, X. & Hayasaka, T. 2007. 
An Asian emission inventory of anthropogenic emission sources for the period 1980–
2020. Atmospheric Chemistry and Physics, 7, 4419–4444.
Portney, P. R. & Stavins, R. N. 2000. Public policies for environmental protection. Wash­
ington, DC: Resources for the Future.
SEPA. 2001. National 10th five-­
year plan on environmental protection. Beijing, China: 
State Environmental Protection Administration.
SEPA. 2005. Flue gas limestone/lime – gypsum desulfurization project technical specification 
of thermal power plant. Beijing, China: State Environmental Protection Administration.
SEPA. 2006a. Guidelines on calculating SO2 emission quotas. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2006b. Technical guidelines for environmental protection in power plant capital 
construction project for check and accept of completed project. Beijing, China: State 
Environmental Protection Administration.
SEPA. 2007. Detailed methods to verify major pollutants emission reduction in the 11th five-­
year period (on trial). Beijing, China: State Environmental Protection Administration.
SEPA & AQSIQ. 1996. Emission standard of air pollutants for thermal power plants. GB 
13223-­
1996. Beijing, China: SEPA, AQSIQ.
SEPA & NDRC. 2008. National 11th five-­
year plan on acid rain and SO2 pollution control. 
Beijing, China: State Environmental Protection Administration, NDRC.
Sheng, C. D., Xu, M. H., Zhang, J. & Xu, Y. Q. 2000. Comparison of sulphur retention by 
coal ash in different types of combustors. Fuel Processing Technology, 64, 1–11.
Shi, D. & Yu, C. 2005. Optimization design, technical control and coal quality evaluation 
of contemporary coal pre-­
processing. Beijing, China: China Contemporary Audio and 
Video Press.
Singer, J. G. 1981. Combustion, fossil power systems: A reference book on fuel burning and 
steam generation. Windsor, CT: Combustion Engineering.
Smith, S. J., Pitcher, H. & Wigley, T. M. L. 2001. Global and regional anthropogenic sulfur 
dioxide emissions. Global and Planetary Change, 29, 99–119.
State Council. 1990. Regulations for the implementation of the law of standardization. 
Beijing, China: State Council.
State Council. 1998. Administrative rule on environmental protection of construction pro­
jects. Decree No. 253. Beijing, China: State Council.
State Council. 2007a. Notice on distributing implementation plans and methods of statis­
tics, monitoring and assessment on energy conservation and pollutant emission reduc­
tion. Beijing, China: State Council.
State Council. 2007b. Notice on distributing the 11th five-­
year plan on environmental pro­
tection. Beijing, China: State Council.
State Environmental Protection Administration (SEPA) & General Administration of Qual­
ity Supervision Inspection and Quarantine. 2003. Emission standard of air pollutants for 


104  Policy making
thermal power plants. GB 13223-­
2003. Beijing, China: State Environmental Protection 
Administration.
Taylor, M. R., Rubin, E. S. & Hounshell, D. A. 2005. Control of SO2 emissions from power 
plants: A case of induced technological innovation in the U.S. Technological Forecast­
ing and Social Change, 72, 697–718.
Tietenberg, T. H. 1990. Economic instruments for environmental regulation. Oxford 
Review of Economic Policy, 6, 17–33.
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. 
Congress.
Xu, Y. 2011. China’s functioning market for sulfur dioxide scrubbing technologies. Envi­
ronmental Science and Technology, 45, 9161–9167.
Xu, Y., Williams, R. H. & Socolow, R. H. 2009. China’s rapid deployment of SO2 scrub­
bers. Energy & Environmental Science, 459–465.
Xu, Y., Yang, C. J. & Xuan, X. W. 2013. Engineering and optimization approaches to 
enhance the thermal efficiency of coal electricity generation in China. Energy Policy, 
60, 356–363.
Yan, G., Yang, J., Wang, J., Chen, X. & Xu, Y. 2009. Carry out SO2 emission trading 
program to set up long-­
term mechanism for emission reduction. In: Wang, J., Lu, J., 
Jintian, Y. & Li, Y. (eds.) Environmental policy research series. Beijing, China: China 
Environmental Science Press.
Zhang, B., Fei, H., He, P., Xu, Y., Dong, Z. & Young, O. 2016. The indecisive role of 
the market in China’s SO2 and COD emissions trading. Environmental Politics, 25, 
875–898.
Zhang, B., Zhang, H., Liu, B. B. & Bi, J. 2013. Policy interactions and underperform­
ing emission trading markets in China. Environmental Science  & Technology, 47, 
7077–7084.
Zhao, Y., Wang, S., Duan, L., Lei, Y., Cao, P. & Hao, J. 2008. Primary air pollutant emis­
sions of coal-­
fired power plants in China: Current status and future prediction. Atmos­
pheric Environment, 42, 8442–8452.


1  
Goal-­
centered policy implementation
In a country with effective rule of law, law enactment and policy making are the 
most important step for environmental protection, while implementation is more 
or less expected, although some bumps may still exist. China does not have sound 
rule of law, and thus, its policy implementation could be even more important than 
policy making. In the United States, after the Acid Rain Program in the Clean Air 
Act Amendments (1990) was enacted, law enforcement was largely the respon­
sibility of the administrative branch. The rule of law obliges the administration 
to enforce the law. However, in China, no such tradition has been established to 
ensure that laws and policies will be genuinely implemented. Furthermore, Chi­
na’s policy implementation is heavily decentralized to local governments (Chap­
ter 3), while the U.S. federal government has a relatively much stronger capacity 
for implementing their own policies. A key difference between China and the 
United States is that China should first mobilize its decentralized policy imple­
menters before witnessing significant efforts and sulfur dioxide (SO2) mitigation. 
China relies on the goal system to mobilize ministries at the central government 
and, more important, local governments for policy making and implementation, 
as discussed in Chapter 4.
Environmental compliance in China was indeed weak but has been improving 
steadily. China has made much progress in the past 15 years to reverse the ear­
lier poor implementation of environmental policies (Jin et al., 2016). Coal-­
fired 
power plants in China have nearly universally installed SO2 scrubbers, already 
94.4% as of 2013 (Figure 5.11). Although most SO2 scrubbers in China today 
do operate properly to greatly contribute to the deep reduction of SO2 emissions 
(Figure 1.8), evidence of their misreporting and cheating was widely present to 
indicate serious noncompliance problems. A study showed that many factories in 
China were primarily concerned about minimizing operation costs and only oper­
ated their pollutant removal facilities when an inspection was imminent (OECD, 
2006). Official data reported that SO2 emissions from the power sector in 2007 
were 11.5 million tons (Ministry of Environmental Protection, 2006–2009), but 
an independent study estimated that 16.4 million tons were emitted in that year 
(Lu et al., 2010). In addition, official data announced that in 2007, 73.2% of SO2 
6	

Policy implementation1


106  Policy implementation
was removed from coal-­
fired power plants that had SO2 scrubbers (Ministry of 
Environmental Protection, 2009b). In Jiangsu Province, which had a relatively 
good track record on environmental protection, however, the rate was found to 
be only about one third in the first few months of 2007 (SERC, 2009). Especially 
before June  2007, cheating was widespread (Figure  6.1). Although almost all 
coal-­
fired power plants generally reported that their SO2 scrubbers were operating 
normally, later confirmed data found the operating time to be much shorter (Fig­
ure 6.1). For those in operation, their SO2 removal efficiencies were often much 
lower than required (SERC, 2009). However, after July 2007, a great majority 
of their SO2 scrubbers were operating for more than 90% of the time and were 
achieving SO2 removal efficiencies of over 90% (Jiangsu Department of Environ­
mental Protection, 2007–2009). Data from the Ministry of Environmental Protec­
tion reported that SO2 scrubbers had already been removing 78.7% of SO2 from 
associated coal-­
fired power plants in 2008 (Ministry of Environmental Protection, 
2009b), indicating that they were largely operating as they were supposed to do. 
This chapter evaluates such transition and examines how the compliance deci­
sions were reversed.
After SO2 scrubbers are installed, the managers of coal-­
fired power plants decide 
whether to operate them or not. The willingness to install SO2 scrubbers does not 
0%
20%
40%
60%
80%
100%
1/2006
7/2006
1/2007
7/2007
1/2008
7/2008
e
t
a
R
 
n
o
i
t
a
r
e
p
O
Month-Year
Later confirmed rates
Self-reported rates
Figure 6.1  
The operation of SO2 scrubbers in Jiangsu Province, including self-reported 
operation rates and later confirmed operation rates
Source: The Economic & Trade Commission of Jiangsu Province, 2009; Jiangsu Department of Envi­
ronmental Protection, 2007–2009; State Electricity Regulation Commission (Nanjing office), 2009.


Policy implementation  107
necessarily mean that the incentives are strong enough for their proper operation. 
In addition, China’s SO2 scrubbers vary greatly in sizes, technology types, sul­
fur contents, costs of reagents and local environmental governance effectiveness. 
A significant variance should exist in their operation especially across provinces.
Studies have shown that three conditions are favorable to ensure compliance 
with environmental legislation: low compliance costs, high penalties for noncom­
pliance and a high probability of catching noncompliance (Cohen, 1999; Helland, 
1998; Becker, 1968). The latter two conditions are complementary to each other. 
In 2000, Blackman and Harrington judged that in China, “both the probability 
of getting caught for underreporting and the penalty for doing so are quite low” 
(Blackman and Harrington, 2000). A 2009 IEA (International Energy Agency) 
report claimed that the operation of China’s SO2 scrubbers was problematic due 
to high operation costs and ineffective environmental regulation (IEA, 2009). As 
the crucial factor to decide the probability of catching noncompliance, the impor­
tance of an effective compliance monitoring system has been widely recognized 
for implementing environmental policies and achieving their intended objectives 
(Lu et al., 2006; Raufer and Li, 2009). Monitoring and site inspection are essential 
for catching offenders but are subject to the constraints of high costs and limited 
budgets (Arguedas, 2008). The problem is especially serious in developing coun­
tries (McAllister et al., 2010; Blackman and Harrington, 2000). To enforce the 
SO2 allowance trading scheme in the United States, the strategy was to install 
continuous emissions monitoring systems (CEMSs) with “periodic quality con­
trol tests of monitoring devices to maintain the accuracy of emissions data” (The 
U.S. Congress, 1990; Stranlund and Chavez, 2000). Large polluters can attract 
more attention. A study of the U.S. steel industry found that large polluting plants 
attracted more scrutiny than their smaller counterparts, regardless of how good 
their compliance record was (Gray and Deily, 1996).
Similar to policy making, policy implementation in China is also goal-­
centered, 
under which actions from the central and local governments focus more on whether 
they can contribute to goal attainment and less on whether policies are genuinely 
implemented. Heavily decentralized policy implementation facilitates their selec­
tive and goal-­
centered enforcement efforts. Such selective policy implementa­
tion also indicates that many rules in China are not followed or respected even 
by governments, which results in the weak rule of law. Given the political, eco­
nomic, social and technological feasibility of implementation, as well as probable 
capacity constraints, those policies that can lead to more pollution mitigation have 
higher probabilities of being prioritized in implementation. This will trigger the 
policy evolution through implementation selection, as discussed in Chapter 5. For 
implementing a given policy in China’s context of originally low environmental 
compliance rates, those factors that contribute to their enhancement are strength­
ened selectively and sequentially, depending on how progress can be made more 
effectively and efficiently with corresponding efforts. In a certain period, compli­
ance costs are determined by technological statuses and market conditions, which 
are largely not decided directly by the environmental administration. In the longer 
term, technologies may evolve and costs may go down, as examined in detail in 


108  Policy implementation
Chapter 7. The other two factors, penalties for noncompliance and environmental 
compliance monitoring, are primarily examined in this chapter.
This goal-­
centered policy implementation echoes the comparative advantage the­
ory that was originated by David Ricardo to analyze the development of international 
trade (Ricardo, 1817). In a two-­
country, two-­
product model, even though a country 
may have lower productivity or absolute disadvantage in producing both products, it 
still could specialize in and export one product based on comparative advantage and 
import only the other. Heckscher and Ohlin further developed the model to attribute 
the origin of comparative advantage in a country’s factor endowment (Ohlin, 1967). 
It later became the foundation of a development theory that argued that a country 
should base its development on its comparative advantage (Chenery, 1961). Lin 
et al. employed this theory to explain the rapid economic growth of China and other 
countries (Lin et al., 2003). Before the economic reform in 1978, China adopted a 
leap-­
forward strategy to develop capital-­
intensive heavy industries against its com­
parative advantage, and this resulted in slow and unsustainable economic growth, 
whereas after the reform, the comparative advantage of labor was better utilized to 
achieve rapid economic growth and upgrading (Lin et al., 2003).
From a status of prevalent noncompliance, goal-­
centered policy implementa­
tion suggests making progress according to the contingent comparative advantage 
of alternative paths for achieving goals. When policy implementers decide which 
path could better serve the SO2 mitigation goals, the chosen path should follow 
the direction whereby the effort’s “productivity” is comparatively higher. In other 
words, easier measures are taken first before moving to more difficult measures, 
although many problems exist in the process.
This chapter examines the progress of three key measures. Penalties for non­
compliance were first increased. A 2007 policy provided subsidies to coal-­
fired 
power plants for normally operating their SO2 scrubbers, but nonoperation would 
incur a penalty of five times. Managers of those coal-­
fired power plants, mostly 
state-­
owned, would lose their jobs if cheating were caught. These penalties were 
relatively easier to be made available, while the more difficult environmental 
compliance monitoring was strengthened in following two steps to enhance the 
probability of catching noncompliance. First, more resources were made available 
to support the conventional environmental compliance monitoring system that 
features monitoring, reporting and verification (MRV). CEMSs also played a key 
role to signal potential noncompliance. The strategy worked well for coal-­
fired 
power plants that tend to be large, making frequent inspections little constrained 
by the shortage of inspectors. Collusion was also made more difficult for improv­
ing data quality. Furthermore, new technologies for environmental compliance 
monitoring are rapidly emerging and evolving, including sensors, satellites and 
social media. They tend to be much cheaper in monitoring one polluter but less 
accurate for legally confirming compliance statuses and issuing penalties, while 
conventional technologies are much more expensive but, if working smoothly, 
can meet the legal accuracy requirements. In the 2010s and, especially, since 
2015, the Chinese government has been actively developing and integrating these 
big data technologies into governance. In environmental protection with millions 


Policy implementation 109
of polluting sources scattered across China’s vast geographic landscape, environ-
mental compliance monitoring is one primary field to apply these new technolo-
gies for achieving higher compliance rates without demanding more resources.
2  
Compliance on the operation of SO2 scrubbers
2.1  SO2 scrubber technologies
Compliance costs of SO2 scrubbers are mainly for their operation as well as main-
tenance. A comprehension of SO2 scrubber technologies is accordingly essential 
to understand how coal- 
fired power plants may cheat on compliance and how the 
government could catch such noncompliance.
SO2 scrubbers (or flue gas desulfurization) have various technology types. Wet 
scrubbers are the most applied technology with SO2 removal efficiencies nor-
mally over 90% (Figure 5.15). This section introduces major features associated 
with wet scrubbers and briefly compares these with dry scrubbers. After the flue 
gas comes out of a dust- 
removal facility (generally electrostatic precipitator, or 
ESP, in China’s coal- 
fired power plants), it will be directed to an SO2 scrubber 
system. The first step is often to pass the flue gas through fans or boosters, which 
facilitate the flow and adjust the velocity of the flue gas to be in a desirable range 
for the best performance of the SO2 scrubber. Then the flue gas enters an absorber 
tower, where actual SO2 removal happens. Generally speaking, coal-fired 
 
power 
generation units of 300 MW or over should have their own absorber towers and 
two units of 200 MW or less could share one (NDRC, 2004). The flue gas enters 
the absorber tower at the lower- 
middle part and moves upward. Limestone slurry 
mixed with products of chemical reactions fills the lower part of the absorber 
tower and is lifted by several circulation pumps to the upper part. Special nozzles 
are used to spray the slurry for the maximization of SO2 removal efficiency. The 
falling slurry droplets contact the flue gas physically and remove approximately 
90% to 95% of SO2 through chemical reactions. Simply put, the main and over-
all reaction is CaCO3
2
+
+
SO
H O
2
3
→
+
CaSO
CO2
2
+ H O. Air is blown into the 
slurry pool at the lower part of the absorber tower to force the oxidization of SO2-
3   
1
and make gypsum (CaSO 2
4
2
H O): CaSO3
2
+
+
O
2H O
2
2
4
→
↓
CaSO 2H2O . 
Before the flue gas exits from the top of the absorber tower, it passes through 
equipment that removes mist. Because the processing removes much heat from 
the flue gas and the efficient outflow from a chimney (often over 200 m high for 
coal- 
fired power plants) requires the flue gas to be above a certain temperature, a 
gas–gas heat exchanger may be included to heat the outlet flow gas with the inlet 
flue gas to raise its temperature.
Two important side systems are respectively for the preparation of limestone 
slurry and the production of gypsum. Limestone is crushed and mixed with water 
to make limestone slurry. Fresh limestone slurry enters the absorber tower often 
through circulation pumps. The bottom slurry with gypsum is pumped out and 
filtered to separate gypsum. The wastewater is sent to a treatment system.


110  Policy implementation
SO2 removal efficiency can be controlled by adjusting various factors, includ­
ing the contact time length between the flue gas and the limestone slurry droplets, 
calcium-­
to-­
sulfur ratio (or Ca/S ratio) and liquid-­
to-­
gas ratio (or L/G ratio). The 
contact time depends on the velocity of the flue gas and the path length before its 
leaving the point where limestone slurry is injected. The height of the absorber 
tower is an influential factor determining the path length. Another factor is asso­
ciated with the different injection heights of circulation pumps. Higher injection 
points indicate a longer path for contact and reaction. When the electricity genera­
tion unit is not in full load with less flue gas, not all circulation pumps will have to 
be operated. Then the choice of different circulation pumps could make some dif­
ference in the SO2 removal efficiency. However, higher absorber tower and longer 
contact path correspond to higher electricity consumption to lift limestone slurry.
In the L/G ratio, the liquid refers to the volume of limestone slurry dropping 
from the upper part of the absorber tower, or circulated liquids. The gas is the 
volume of flue gas entering the absorber tower. Higher L/G ratio leads to higher 
SO2 removal efficiency because the chance is higher for an SO2 molecule to be 
absorbed. It is controlled through circulation pumps: if the flue gas volume does 
not change, turning on more pumps indicates a higher L/G ratio. Since the number 
and power of circulation pumps are fixed after an SO2 scrubber comes online, the 
liquid volume has an upper limit, which restrains the maximum contribution of 
enhancing L/G ratio to increase SO2 removal efficiency.
Ca/S ratio is the molar ratio between calcium carbonate (CaCO3) and sulfur 
oxides (SOx, dominantly SO2). In a perfect situation, as predicted in the chemical 
reaction introduced earlier, the ideal Ca/S ratio is 1 to remove all SO2 and use up 
all limestone. But in the actual situation, not all limestone will be consumed and 
not all SO2 will be removed. SO2 wet scrubbers can achieve high efficiencies in 
both aspects. As a result, the actual Ca/S ratio is only a little higher than 1, usually 
around 1.03 for China’s wet scrubbers (Wu and Qian, 2007). Because the inlet 
quantity of SO2 changes with the volume of flue gas and the SO2 concentration, 
even though the Ca/S ratio is maintained stable, the rate of adding limestone to the 
system will still change. On the other hand, the workload of removing SO2 could 
become too heavy when the actual sulfur content exceeds the designed level by 
a significant margin. In this situation, when all circulation pumps are turned on 
and the L/G ratio has reached its maximum, the only major method to maintain 
a required high SO2 removal efficiency is to enhance the Ca/S ratio. However, a 
much higher Ca/S ratio than the designed level will not only add costs but also 
will more likely clog the system and barricade its normal function. The adjustment 
of the Ca/S ratio is through controlling the pH value of the limestone slurry in the 
absorber tower. In daily operation, the pH value should be maintained within a 
range. A pH value above the normal range indicates excessive limestone and that 
the injection rate of fresh limestone slurry should be reduced.
SO2 wet scrubbers consume about 1% of the electricity generated from the cor­
responding power generation units. The rate could be as high as 3.5% when high-­
sulfur coal is burned with a heavy workload of SO2 removal. China’s coal-­
fired 
power plants consumed, on average, 6.79% of the electricity they generated in 


Policy implementation  111
2008 (SERC et al., 2009), in which SO2 scrubbers accounted for a notable share. 
Significant electricity-­
consuming components of SO2 scrubbers include the fans, 
circulation pumps and limestone slurry preparation system.
Two major economies of scale are associated with SO2 scrubbers in construc­
tion. First, the size of an absorber tower is largely determined by the volume of 
flue gas or the scale of the corresponding power generation unit. A larger volume 
of flue gas or a larger scale in megawatts leads to lower average costs for each 
unit of flue gas treated or each megawatt. Because absorber towers are responsi­
ble for a large part of the capital costs, this economy of scale could significantly 
reduce the unit capital costs. Since most nonpower SO2 emission sources consume 
much less coal and do not generate large enough volume of flue gas to provide 
the economy of scale, the unit costs of SO2 scrubbers are often more expensive.
Second, higher SO2 concentration in the inlet flue gas, or higher sulfur input 
rate, raises capital costs for each unit of flue gas treated because they require 
larger systems of limestone preparation and gypsum handling, as well as probably 
higher absorber tower and more circulation pumps. Economy of scale can also 
be realized for these systems to treat each unit of SO2. SO2 concentration in the 
inlet flue gas is mostly determined by two factors: sulfur and thermal contents of 
coal. The link with sulfur contents is quite straightforward: if different types of 
coal only differ in sulfur contents, higher sulfur contents indicate more SO2 in a 
roughly equal amount of flue gas. With the same thermal efficiency, the volume of 
flue gas mainly depends on the thermal input. Then lower thermal contents of coal 
mean that more coal has to be burned for the required thermal input, and accord­
ingly, more SO2 will be generated. Accordingly, sulfur content per unit of energy 
is a better indicator of SO2 concentration in the inlet flue gas of SO2 scrubbers.
The product of SO2 scrubbers is gypsum. Depending partly on the quality, it 
can either be sold in the market or go to landfill. A significant market for gypsum 
is in building materials.
The operation and maintenance (O&M) of SO2 scrubbers are associated with 
costs in materials (including mainly limestone, electricity and water), labor and 
maintenance. The sale of gypsum could earn some revenue but often only at an 
insignificant portion. If the quality of SO2 scrubbers remains about the same, 
maintenance costs are positively related to the capital investment of SO2 scrub­
bers. As a result, larger scales of SO2 scrubbers are linked with lower maintenance 
costs on the bases of megawatt-­
hour or ton SO2 removed. Similarly, economy 
of scale is also relevant to labor costs, but the impact on overall O&M costs is 
constrained by the insignificant share of labor costs; for example, my field trip 
to China’s coal-­
fired power plants found that roughly 15 workers were required 
to run the SO2 scrubber and ESP for a 300-­
MW plant in 2008. Their total annual 
costs could be about 1 million RMB. The average O&M costs of China’s SO2 
scrubbers were about 15 RMB/MWh, indicating that the total O&M costs would 
be approximately 23 million RMB if the capacity factor was about 5,000 hours/
year. Then the share of labor costs was less than 5%.
Materials comprise most of operation costs. In normal operation, the Ca/S ratio 
remains fairly stable, and thus, the limestone consumption is about linearly related 


112  Policy implementation
to sulfur input. Electricity consumption for running SO2 scrubbers can be roughly 
divided into three major parts: in fans that are mainly associated with the flue 
gas volume, in the handling of limestone and gypsum that is affected by sulfur 
input and in circulation pumps connected with both. Most water consumption is 
in the form of evaporation to the flue gas in the absorber tower, and the water is 
released to the atmosphere together with the cleaned flue gas. As a result, water 
consumption is mainly correlated with the volume of flue gas and is also affected 
by economy of scale.
Because of its wide availability and low costs, limestone is the dominant 
absorbing reagent in wet scrubbers. However, other alkaline reagents are some­
times applied, such as seawater and alkaline wastewater. Furthermore, SO2 scrub­
bers can be dry. Lime (CaO) is often used as the absorbing reagent. Because of its 
much lower utilization rate, the Ca/S ratio has to be much higher (e.g., 1.3~1.5). 
Generally, the SO2 removal efficiency is in the range of about 70% to 80%, lower 
than that of wet scrubbers. The capital costs of dry scrubbers are lower, but the 
operation and maintenance costs are higher (EPA, 2003).
2.2  
Noncompliance behaviors
The managers of coal-­
fired power plants have strong incentives to avoid the costs 
of O&M. Data from Jiangsu Province showed that from 2006 to June 2007, the 
self-­
reported operation rates (the percentage of time that an SO2 scrubber is in 
operation alongside the corresponding power generation unit) from coal-­
fired 
power plants were significantly higher than the values that were later confirmed, 
likely through other relevant data such as limestone consumption, gypsum pro­
duction and electricity consumption (respectively, more than 90% and about 60%; 
Figure 6.1). The discrepancy reflects the likely magnitude of misreporting. This 
section discusses several prominent problems that emerged from the author’s 
interviews and the literature. These problems prevented the proper operation of 
SO2 scrubbers and caused very significant uncertainty in estimating SO2 emis­
sions from coal-­
fired power plants.
Typical noncompliance behaviors could include the following: first, SO2 emis­
sions may be underreported and the quality of SO2 scrubbers could be poor. Coal-­
fired power plants underreported SO2 emissions to pay a lower effluent discharge 
fee and to be seen as complying with regulations. In 2007, 98% of coal consumed 
in China’s power plants was raw coal (National Bureau of Statistics, 2008). Coal-­
fired power plants were allowed to pick out coal stones from received raw coal 
to calculate actual coal consumption. In interviews, the author found that coal 
stones were sometimes overreported. This factor could have led to a 1% to 2% 
underestimation of SO2 emissions. Furthermore, China’s coal-­
fired power plants 
usually had to use different coals with sulfur contents that could vary signifi­
cantly. The instability of coal supply was confirmed by Steinfeld et al. (2009). It 
made the underreporting of sulfur contents harder to detect. Interviews in China’s 
SO2 scrubber companies found that many early scrubbers (e.g., before 2005) had 
serious quality problems. In order to reach designed SO2 removal efficiencies, 


Policy implementation  113
besides the replacement of malfunctioning equipment, a few SO2 scrubbers even 
had to have their very expensive absorber towers retrofitted. The main reason 
for the faults in the SO2 scrubbers was that they were designed on the basis of 
underreported sulfur contents. In China’s first public and high-­
profile statement 
to penalize the abnormal operation of SO2 scrubbers, instability and bad quality 
were particularly pointed out, and three power plants were found to use coal with 
much higher sulfur than the designed levels (Ministry of Environmental Protec­
tion, 2008). At the design stage of SO2 scrubbers, if the managers of coal-­
fired 
power plants had been underreporting sulfur contents in the past, they would con­
tinue to do so to conceal their guilt. Some managers did not plan to operate their 
SO2 scrubbers initially and were not concerned about their quality. They installed 
the SO2 scrubbers purely in order to comply with the government’s requirements 
and to qualify for a subsidy for generating desulfurized electricity. The managers 
wanted to minimize capital costs through underreporting sulfur contents. When 
inspections were known in advance, reaching the required SO2 removal efficien­
cies was not a problem because coal-­
fired power plants often kept some low-­
sulfur coal in reserve on-­
site.
Second, illegal bypass ducts may be used to avoid flue gas treatment. Many 
SO2 scrubbers had bypass ducts to allow the flue gas to exit without going through 
the SO2 scrubber systems. The purpose was to enable electricity generation when 
SO2 scrubbers had minor problems and need to be shut down temporarily. In a 
2007 policy, coal-­
fired power plants were not penalized provided that their SO2 
scrubbers were properly operating for at least 90% of the time (NDRC and SEPA, 
2007b). However, bypass ducts also provided opportunities to avoid the operation 
of SO2 scrubbers when they functioned normally. Six coal-­
fired power plants were 
penalized for illegally using bypass ducts and leaving some flue gas untreated in 
2007 and 2008 (Ministry of Environmental Protection, 2008, 2009c).
Third, data from CEMSs may also be inaccurate and manipulated. CEMSs 
could greatly enhance environmental monitoring capacity. China had 60 SO2 
scrubbers at the end of 2004 (Ministry of Environmental Protection, 2010a), while 
a general survey in 2004 found that about 400 CEMSs had been installed in 180 
coal-­
fired power plants (Pan et al., 2005). The author’s site visits and Steinfeld 
et al. also found that CEMSs were being widely used (Steinfeld et al., 2009). As 
far as cost was concerned, there was little reason for coal-­
fired power plants to 
resist the installation of CEMSs. Two CEMSs in Plant 3 in Table 6.1 cost about 
US$132,000, only 0.5% of the capital costs of the plant’s SO2 scrubbers. How­
ever, CEMSs may not report credible and reliable data. One concern was over 
the quality of the equipment used. CEMSs cost much more in the United States: 
according to a cost model from the U.S. Environmental Protection Agency (EPA), 
it generally required more than half a million dollars for one set (The U.S. EPA, 
2007). The 2004 general survey found that only 20% of the CEMSs in China were 
functioning normally (Pan et al., 2005), local environmental protection bureaus 
generally refused to accept data from CEMSs and only one was recognized as a 
credible data source for the purposes of levying the SO2 effluent discharge fee 
(Pan et al., 2005). Later on, CEMSs were officially accepted as data sources after 


114 Policy implementation
 
ge 
US$/MWh
The 
The 
2.5
0.7
6.83 RMB. Upon the request 
 (State Devel-
**
disclosed in Plant 4. Comparing with other plants in the eastern provinces, 1.0% is used here for later analysis. ** 
(State 
standard 
Effluent 
dischar
fee 
2.0
2.8
0.3
0.7
0.4~0.7
2008. 
data 
the 
for 
2
official 
generally 
and 
emium for 
desulfurized 
of US$0.092/kg SO
China’s 
generation; 
Price 
pr
electricity
US$/MWh
2.2
2.2
2.2
2.2
2.2
3.7
estimation 
 
 =
compiling 
gin 
recent 
mar
ofit 
generation
US$/MWh
~ 14.6
> 0
>> 14.6
< 7.3
Pr
of electricity 
most 
their 
reflect 
31, 2008, is used: US$1
 
& 
gin 
 
mar
The fee rate refers to the level 
in 
electricity 
assumed 
as 
efficiencies of 
20%, 
ash, 
thermal 
interviews; 
Operation
maintenance 
(O&M) costs
US$/MWh
~4.1
1.8
2.2
<2.2
(1): >3.7;
(2): <3.7
profit 
and 
 scrubbers. 
in 
3.7
rates 
2
SO
retention 
s 
’
costs 
author
ts
Sulfur content
The 
down 
the 
to 
*
2009. 
shutting 
echnical Supervision, 2007).
s seven coal-fired power plan
sulfur 
are 
%
3.0%
4.0%
3.5%
1.0% 
0.5%
1.0%
0.7~1.1%
T
July 
dollars, the exchange rate on December
and 
assumptions 
according 
95%, 
 
June 
payment if 
 
in 
 
2
t
t
t
t
SO
scrubber 
type
t
t
We
We
We
We
We
We
(1): dry;
(2): wet
additional 
intermediate 
scrubbers, 
interviews 
 scrubbers in China’
wet 
the 
The 
of 
New or 
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
s 
ofit
’
reflect 
2003). 
s 
Retr
New
New
author
calculated to 
al., 
the 
 
et
efficiencie
in 
2
removal 
Region
Southwest
Southwest
Southwest
collected 
is 
 
2
East
East
East
East
Commission 
Data on SO
fee 
were 
original currency units were in Chinese RMB. In the conversion to U.S. 
ge 
data 
Planning 
2007b); SO
 
The 
opment 
Council, 
Table 6.1
Plant 3
Plant 5
Plant 6
Plant 7
clear information on sulfur contents was 
Note: 
of the interviewees, the names of coal-fired power plants are intentionally not shown.
dischar
Plant No.
No 
effluent 
Plant 1
Plant 2
Plant 4
levels at the corresponding unit scales (Zhejiang Bureau of Quality and 
 
*


Policy implementation  115
their online connection with provincial environmental protection bureaus. How­
ever, the author’s interviewees still said that they did not fully trust data from 
CEMSs. The locations of the sensors could affect the readings of CEMSs, and 
data reporting could also be manipulated. In 2008, three coal-­
fired power plants 
were caught illegally setting up ceilings of outlet SO2 concentrations that could be 
reported (Ministry of Environmental Protection, 2009c).
Fourth, coal-­
fired power plants could either cheat or even collude with environ­
mental compliance inspectors. Data from CEMSs were compared quarterly with 
direct measurements to verify accuracy (State Council, 2007b). A survey in China 
found that multiple inspections per annum tended to deter violation, no matter 
which government level inspectors were from (Lu et al., 2006). However, the 
effectiveness of site inspections could be constrained. According to the author’s 
interviews, many plants were able to raise the removal efficiency of their SO2 
scrubbers from zero to the designed level in half an hour and significantly more 
quickly if from an intermediate level. Some plants therefore kept their scrubbers 
either turned off or on low power and put them in full operation only when an 
inspection was imminent, enabling them to keep costs down while also passing 
the inspection. Even if abnormal operation were caught, a solution could be to 
collude with inspectors through bribes.
3  
Reversing noncompliance: penalty
The proper operation of SO2 scrubbers demands strong enough incentives to over­
come the hurdle of the high O&M costs. In the United States, the average O&M 
costs in 2008 were US$1.55/MWh (EPA and DOE, 2010). These figures could 
hardly be extrapolated for China because of the great differences in the capital 
costs of SO2 scrubbers, labor costs and other items. My interviews collected rel­
evant data from six coal-­
fired power plants, as presented in Table 6.1. The O&M 
costs varied from US$1.8 to 4.1/MWh, all above the average in the United States. 
Sulfur contents were the most influential factor: Plants 1 and 3 burned coals with 
approximately 3% to 4% sulfur content, and their O&M costs were roughly twice 
as high as those in Plants 4, 5, 6 and 7, which burned coals with 1% sulfur content 
or less. The O&M costs could be used as the average marginal costs of operating 
SO2 scrubbers. In generating electricity, several coal-­
fired power plants that the 
author visited had gross profit margins from not much above zero to significantly 
over US$14.4/MWh (including the O&M costs of SO2 scrubbers and the price 
premium for desulfurized electricity).
A survey of China’s inspection authorities and polluting firms found that fines 
for noncompliance were often not high enough to deter potential offenders (Lu 
et  al., 2006). The initial SO2 effluent discharge fee was about 0.20 RMB/kg 
(US$0.031/kg) in most provinces and lower than the marginal abatement costs in 
China’s large plants (Cao et al., 1999; Dasgupta et al., 1997). Polluting firms may 
simply pay to pollute. When facing a penalty, the first reaction of polluting firms 
was to negotiate with environmental protection bureaus or ask for the interference 
of local governments (Lu et al., 2006), which compromised the penalty.


116  Policy implementation
In July 2005, China’s SO2 effluent discharge fee was raised from US$0.031/kg 
in 2003 to US0.092/kg(StateDevelopmentPlanningCommissionetal.,2003).However,asconvertedtoUS/MWh in Table 6.1, it was still too low to overcome 
the hurdle of the much higher O&M costs. Another increase was scheduled in 
2007 to reach US$0.18/kg in three years (State Council, 2007a), but the exact 
schedule varied from province to province. In Jiangsu Province, the higher rate 
had been in effect since July 2007 (Jiangsu Department of Environmental Pro­
tection, 2008), but in Henan Province, the lower rate was still being applied in 
the first quarter of 2010 (Henan Department of Environmental Protection, 2010). 
With the higher rate, coal-­
fired power plants burning high-­
sulfur coals would find 
operating SO2 scrubbers cheaper than paying the effluent discharge fee (Plants 1 
and 3 in Table 6.1). However, for others (Plants 4, 5, 6 and 7) when facing only 
this policy, the rational decision was to pay the fee.
Another policy was introduced in 2004. If new coal-­
fired power plants came 
online together with SO2 scrubbers, the desulfurized electricity could enjoy a 
price premium of US$2.2/MWh (NDRC and SEPA, 2007a). In June 2006, the 
policy extended to cover all SO2 scrubbers, including retrofitted ones (NDRC and 
SEPA, 2007a). Some coal-­
fired power plants were awarded higher price premi­
ums, such as Plant 7 in Table 6.1. The price premium and the effluent discharge 
fee together were a little higher than the O&M costs (Table 6.1), but the small 
difference indicated that the proper operation would be a rational decision only 
when most nonoperation cases were caught.
The 11th Five-­
Year Plan witnessed sharp increases in noncompliance penalties. 
In 2007, a harsh penalty measure was associated with the price premium for the 
first time. If the operation rate of an SO2 scrubber were lower than 80%, a penalty 
of US$11.0/MWh would be issued for any additional non-­
desulfurized electric­
ity generation (NDRC and SEPA, 2007b). The required minimum probability of 
catching nonoperation became much lower to induce the proper operation of SO2 
scrubbers. For Plant 3 in Table 6.1 burning high-­
sulfur coal, corresponding to the 
effluent discharge fee of US$0.092/kg SO2, a risk-­
neutral manager would decide 
to operate SO2 scrubbers properly if the probability of catching nonoperation 
exceeded 26% (see Table 6.2 for the calculation formula). For Plants 4 and 5 burn­
ing low-­
 to medium-­
sulfur coals, the minimum probability was about one seventh. 
Furthermore, additional penalties were introduced on the managers of coal-­
fired 
power plants. In China, almost all coal-­
fired power plants were owned by the 
state. The nonoperation of SO2 scrubbers could increase profit and benefit the 
managers’ career and salary. However, according to formal regulations (NDRC 
and SEPA, 2007b) and the author’s interviews, cheating and nonoperation could 
lead to the removal of the managers. They had to calculate the risk for themselves.
The penalties in 2007 also aimed for minimizing potential moral hazard when 
SO2 scrubbers occasionally had to stop operating due to accidents, malfunctions 
or other reasons. While they were out of action, SO2 emissions could be controlled 
either by minimizing the sulfur content of coal or by shutting down electricity 
generation. China would issue no penalty as long as the operation rate were above 
90%, a mild penalty of US$2.2/MWh if the rate were between 80% and 90% and 


Policy implementation 117
Table 6.2 Decision scenarios for the managers of coal-fired power plants
Scenario SO2 scrubbers SO2 scrubbers Electricity Net revenue of a coal-fired power 
functioning
operating
generation plant
(1)
Yes
Yes
Yes
Profit margin
(2)
Yes
No
Yes
Profit margin + O&M costs – C% × 
(Price premium + Discharge fee + 
Penalty)
(3)
No
No
Yes
Profit margin + O&M costs – C% × 
(Price premium + Discharge fee + 
Penalty)
(4)
No
No
No
0
Note: C% is the actual probability of catching the nonoperation of SO2 scrubbers. The 
proper operation of SO2 scrubbers, when they function, requires that the net revenue in sce-
nario (1) is greater than that in scenario (2). The corresponding condition can be calculated as 
O&M costs
C% >
. When SO2 scrubbers do not function, the discon-
Discharge fee
P
+
+
rice premium
Penalty
tinuation of electricity generation becomes a rational decision when the net revenue in scenario 
Profit margin + O&M costs
(4) is greater than that in scenario (3), or C% >
. Because 
Discharge fee
P
+
+
rice premium
Penalt
l y
profit margins are generally positive, it is accordingly easier to push for the proper operation of 
SO2 scrubbers when they function than to ask coal-fired power plants to discontinue electricity gen-
eration when they do not. In order to encourage coal-fired power plants to fix malfunctioning SO2 
scrubbers as soon as possible, the rational decision when SO2 scrubbers function should generate 
greater net revenue than that with malfunctioning SO2 scrubbers. The condition is fulfilled when 
O&M costs
C% >
.
Discharge fee
P
+
+
rice premium
Penalty
a harsh penalty of US$11.0/MWh if the rate were under 80% (NDRC and SEPA, 
2007b). Because it was expensive to restart electricity generation, the O&M costs 
of SO2 scrubbers may not be critical in the decision making when SO2 scrubbers 
could get fixed soon.
When problems have to take much time to fix – for example, several weeks – and 
the penalty of US$11.0/MWh is applied, the economic incentives should make it 
a rational decision to discontinue electricity generation for many coal-fired 
 
power 
plant managers. Electricity generation without operating SO2 scrubbers earned 
a profit margin and avoided the O&M costs of SO2 scrubbers, but if the non-
operation of SO2 scrubbers were caught, coal-fired 
 
power plants would need to 
return the price premium and pay the effluent discharge fee as well as the penalty. 
Many coal- 
fired power plants might continue generating electricity as long as the 
probability of catching the nonoperation of SO2 scrubbers was low enough (see 
Table 6.2 for the specific calculation). For coal-fired 
 
power plants with large profit 
margins (such as Plant 5 in Table 6.1), electricity generation should continue even 
when nonoperation could not be hidden at all. However, when the author visited 
Plant 5, electricity generation in one system had been discontinued for several 
weeks due to its malfunctioning SO2 scrubber. Personal penalties on the manag-
ers could have played a role. Furthermore, even if the decision was to continue 
electricity generation, a high-enough probability 
 
of detection was still necessary 


118  Policy implementation
to encourage coal-­
fired power plants to fix malfunctioning SO2 scrubbers as soon 
as possible (see Table 6.2 for the specific calculation). If the actual probability was 
not expected to reach this level, there would be little concern about the quality of 
SO2 scrubbers, as in the early years.
Furthermore, coal-­
fired power plants should also comply with regulations on 
SO2 removal efficiency and effluent emission standards (NDRC and SEPA, 2007b; 
SEPA and General Administration of Quality Supervision Inspection and Quar­
antine, 2003; MEP and AQSIQ, 2011). Technically in practice, a coal-­
fired power 
plant could choose a designated SO2 removal efficiency. For example, higher 
ratios of Ca/S (the molar ratio between CaCO3 and SOx) or L/G (the liquid-­
to-­
gas 
ratio in volume) would remove more SO2 from the flue gas. Reasonably, if not 
regulated, a coal-­
fired power plant could lower SO2 removal efficiencies to reduce 
costs. On the other hand, because of changing sulfur contents and workload, SO2 
concentration and flue gas volume were not stable. Scrubbers’ capability to track 
the changes – with the same methods of adjusting SO2 removal efficiencies – was 
necessary for their reliable operation.
The Chinese central government mandated minimum SO2 removal efficiencies 
being established (NDRC and SEPA, 2007b) and provincial governments were in 
charge of the details. For example, when SO2 removal efficiencies were lower than 
predetermined levels (generally 90% for wet scrubbers), Henan Province simply 
counted the time as nonoperation (Henan Development and Reform Commission 
and Henan Environmental Protection Bureau, 2007). In normal conditions, the 
incentives were strong enough to make SO2 scrubbers reach the required levels of 
SO2 removal efficiencies. Two actual cases from the author’s field trip could demon­
strate the decisions. In the first case, sulfur contents went up significantly but were 
expected to be a temporary situation. The designed sulfur content for Plant 6’s SO2 
scrubber was 0.84%, but for a period in 2008 when coal supply was constrained, 
the actual sulfur content was higher than 2%. Such a dramatic increase in sulfur 
content became a serious burden. To maintain SO2 removal efficiencies over 90%, 
the solution was to raise the Ca/S ratio from the designed level of 1.03 to 1.3. In the 
second case, when the increased sulfur contents were expected to be long-­
lasting, 
a temporary solution would not be sustainable. One of the eight coal-­
fired power 
plants the author visited had to shut down and modify the original SO2 scrubber to 
handle the much higher sulfur input rate. Particularly, the absorber tower became 
significantly taller by adding another section on the top of the original one. The 
pathway was accordingly longer for the flue gas and limestone slurry to contact and 
react. Additional circulation pumps could also be added to enhance the L/G ratio.
In order to better implement the incentives, responsible government agen­
cies are specified: electric grid corporations were in charge of paying the price 
premium in time; provincial environmental protection bureaus collected effluent 
discharge fees; provincial price agencies were responsible to recover unjustified 
price premium according to actual operation rates (NDRC and SEPA, 2007b). 
Seven coal-­
fired power plants in 2008 and five in 2009 were penalized for cheat­
ing or nonoperation with the US$11.0/MWh penalty applied (Ministry of Envi­
ronmental Protection, 2009c, 2008).


Policy implementation  119
Central and local governments in China are not the only entities that have their 
tasks centered around goals. Because almost all coal-­
fired power plants in China 
were state-­
owned, they were also assigned quota or goals for their total SO2 emis­
sions (SEPA, 2006). Both goals and policies play crucial roles in their compli­
ance decisions on the operation of their SO2 scrubbers. In addition to financial 
penalties, administrative penalties were also applied for noncompliance. In envi­
ronmental enforcement and compliance, decision makers at local governments, 
power corporations and coal-­
fired power plants also kept in mind their SO2 emis­
sion caps or goals. If SO2 removal efficiencies were too low and nonoperation was 
caught, the SO2 emission permits could be used up soon. In addition, seriously 
abnormal operation of SO2 scrubbers was publicly punished (MEP and NDRC, 
2008; Ministry of Environmental Protection, 2009a), which could affect the 
career of the coal-­
fired power plants’ managers. In the words of an interviewee, 
“it is not worthwhile for the managers of a coal-­
fired power plant to risk losing 
the positions to save money for the plant. Anyway, the money is not theirs, but 
the positions are.”
4  
Reversing noncompliance: environmental 
compliance monitoring
The effectiveness of environmental compliance monitoring determines the prob­
ability of catching noncompliance. China’s emission data MRV system is largely 
bottom up, which could potentially suffer from two major challenges. The first 
challenge lies in the system’s high costs. Compliance monitoring resource con­
straints exist in all countries, but the problem is especially daunting in develop­
ing countries, due to the high costs of compliance monitoring, limited resources, 
understaffed environmental agencies, inadequate training and technological sup­
port (Arguedas, 2008; McAllister et al., 2010; Blackman and Harrington, 2000; 
Russell and Vaughan, 2003; Pan et  al., 2005). How to better utilize available 
resources is critical to determine the effectiveness of every domestic policy and 
international environmental treaty. Compliance monitoring is the most resource-­
consuming activity in enforcing environmental policies. For example, an emis­
sion trading scheme should effectively deter cheating and verify actual emission 
levels (Kruger and Egenhofer, 2006), while compliance monitoring was respon­
sible for 69% of transaction costs for German companies in the European Union 
CO2 Emission Trading Scheme (Heindl, 2012). The existence of many small and 
medium-­
sized polluters could seriously attenuate available resources, even in 
developed countries where the rule of law is generally well established. Due to 
the significant economy of scale, large point sources generally have lower com­
pliance monitoring costs on a per-­
ton-­
emission basis and are often prioritized 
(Heindl, 2012; Gray and Deily, 1996). Because of China’s sheer size, the large 
system involves many personnel and occupies substantial resources. In the 12th 
Five-­
Year Plan (2011–2015) alone, the Chinese government planned to invest 
40 billion RMB (~US$5.9 billion) to enhance related environmental regulation 
capacity (MEP, 2013).


120  Policy implementation
The second challenge is intentional data manipulation. Environmental moni­
toring and reporting in China generally must pass through, and be inspected by, 
polluting firms and various levels of local governments and relevant agencies 
before reaching the central government. Most environmental compliance capaci­
ties, such as personnel and governmental expenditure, are in local governments, 
while the central government is mainly in charge of policy making. Emissions 
of CO2, SO2 and NOx are generally calculated via bottom-­
up energy consump­
tion data and emission factors (Liu et al., 2015; Lu et al., 2011; Zhang et al., 
2007). This approach is often subject to the influence of intentional distortions 
for the interest of stakeholders along the path (Tsinghua University, 2010). China 
has been exerting increasingly high pressure on local governments and energy-­
intensive firms to achieve top-­
down energy and emission control goals from the 
central government (Xu, 2011b). In comparison to the technologically challeng­
ing, economically expensive and politically difficult tasks of actual mitigation, it 
would be much more convenient to twist the reported numbers (Jin et al., 2016).
The objective resource constraint and the intentional data manipulation could 
seriously compromise data quality and thus the effectiveness of environmen­
tal compliance monitoring. Facing immense pressure of environmental crises, 
the Chinese government has been actively searching for potential solutions for 
enhancing environmental data quality.
4.1  
Model construction
In order to understand China’s environmental compliance monitoring in greater 
depth, a conceptual, computable model is constructed to simulate the evolution 
of compliance rates under different compliance monitoring strategies and how 
influential factors in three categories – pollution abatement costs, noncompliance 
penalty and, most important, compliance monitoring effectiveness – affect com­
pliance decisions of polluters and thus the compliance rate. Mathematical details 
of the model are provided in the Appendix to this chapter.
This model stands on the shoulders of two pieces of research literature for cre­
ating a theoretical framework. The first well-­
developed economics literature of 
crime and punishment understands crimes as rational choices. Whether a pol­
luter chooses compliance or noncompliance is based on comparing related costs 
and benefits (Polinsky and Shavell, 2000; Becker, 1968; Glaeser, 1999; Xu, 
2011a; Shimshack, 2014; Levitt, 2004). If a polluter pondered not complying 
with an environmental regulation, pollution abatement costs could be saved as its 
expected benefits. However, such behavior would incur expected costs, which is a 
product of (1) penalty on noncompliance and (2) the probability of being caught. 
Risk-­
neutral rational polluters would choose environmental noncompliance if the 
expected benefits were greater than the expected costs. The compliance or non­
compliance decision is assumed to be deliberate but not at random. The second 
mature literature, or a series of related literature, such as on policing, pollution 
control and tax evasion, examines how to enhance the probability of catching non­
compliance. Compliance monitoring could apply various strategies for enhancing 


Policy implementation  121
the probability with a given amount of resources, although the effectiveness is 
mixed. Levitt (2004) found that policing strategies are of only minor signifi­
cance, while the number of police may explain a large proportion of the crime rate 
change. For tax compliance, endogenous audit selection rules screen taxpayers 
for potential auditing, but the impacts on compliance are mixed (Konrad et al., 
2017; Vossler and Gilpatric, 2018). In epidemiology, strategies are developed to 
promote public health and enhance the rate of finding sick patients at early stages 
among a population (Bonita et al., 2006). A population would be first screened, 
and those with positive results would have to go through another round of more 
careful diagnosing for confirming whether they were true or false positive.
These two pieces of literature are integrated together in this study to simulate 
environmental compliance decisions. Two environmental compliance monitoring 
systems are proposed and simulated, as illustrated in Figure 6.2. The conventional 
system that is based on monitoring, reporting and verification is simplified to 
require governmental compliance monitoring resources primarily for site inspec­
tion. Adopting the terminology in epidemiology, the model refers to these activi­
ties as diagnosing. If a polluter were caught as being noncompliant, a penalty 
would be issued. The new compliance monitoring system inserts an additional 
step before diagnosing to actively screen polluters into high-­
risk and low-­
risk 
groups, with higher and lower probabilities of being noncompliant, respectively. 
Diagnosing with higher costs follows with site inspections or other more accurate 
means to confirm noncompliance only in the high-­
risk group. For the convenience 
Polluters in compliance and noncompliance
Screening: cheap but more errors 
High-risk group
Low-risk group
Diagnosing: expensive but accurate
Penalty
No penalty
Polluters making compliance decisions
Compliance 
monitoring 
resources
Screening: c
High-risk grou
Diagnosing: expensive but accurate
Penalty
Compliance 
monitoring
resources
Figure 6.2  
A conceptual model of environmental compliance monitoring
Note: The dash-line arrows indicate the screening system’s flow, while the diagonal-pattern arrows 
refer to the diagnosing system’s flow. Their major difference is the existence/absence of the screen­
ing step with screening technologies. The gray boxes show compliance monitoring resources that not 
only are allocated between screening and diagnosing technologies in the screening system but only to 
diagnosing technologies in the diagnosing system.


122  Policy implementation
of discussion, the conventional system is referred to in this chapter as the diagnos­
ing system, while the new system contains both screening and diagnosing, and it 
will be called the screening system. Numerous studies have applied the economic 
model of crime and punishment for understanding environmental noncompliance 
(Xu, 2011a; Shimshack, 2014; Guo et  al., 2014). The compliance monitoring 
strategy with screening has also been widely applied in multiple fields (Konrad 
et al., 2017; Vossler and Gilpatric, 2018; Bonita et al., 2006).
4.2  
Strengthening the conventional diagnosing system
China has made several prominent improvements in monitoring and site inspec­
tion to address the previously mentioned two challenges for enhancing the prob­
ability of catching the nonoperation of SO2 scrubbers. First, more resources were 
made available for environmental compliance monitoring. The numbers of gov­
ernment employees at all levels increased from 46,984 in 2005 to 52,944 in 2009 
and 61,668 in 2015 for environmental monitoring and from 50,040 in 2005 to 
60,896 in 2009 and 66,379 in 2015 for inspection (Figure 3.1). Although still lim­
ited, the personnel resources had already been enough to have an intensive focus 
on SO2 scrubbers in coal-­
fired power plants. Particularly, only 503 coal-­
fired 
power plants housed 461 GW SO2 scrubbers (1,264 systems) at the end of 2009, 
and the largest 300 had a total capacity share of 82% (Ministry of Environmental 
Protection, 2010a). In 2013, 282 coal-­
fired power plants that were at or greater 
than 1 GW each had 470 GW SO2 scrubbers in total, or 62.3% of all (Ministry of 
Environmental Protection, 2014). Government personnel are sufficient to follow 
these large plants closely and conduct inspections frequently.
The number of polluting sources (N) that require compliance monitoring varies 
dramatically, depending on focused polluter sizes, pollutants and other features. 
China conducted the first census of polluting sources with the census date being 
December 31, 2007, and pollution information for 2007, covering 5,925,576 pol­
luting sources, including 1,575,504 industrial, 2,899,638 agricultural, 1,445,644 
domestic and 4,790 centralized pollution control facilities (Ministry of Environ­
mental Protection et al., 2010). In comparison, China’s annual environmental sta­
tistics report focused on about one tenth of the polluting sources, being 161,598 
industrial sources, 131,837 farms and 7,578 districts for animal husbandry and 
6,910 water treatment plants, 2,315 municipal waste treatment facilities and 866 
hazardous waste treatment facilities in 2015 (Ministry of Environmental Protec­
tion, 2002–2016). Among these sources, 68,121 polluting sources were under spe­
cial supervisory monitoring (Ministry of Environmental Protection, 2002–2016).
Furthermore, in order to make the conventional diagnosing system more effi­
cient, CEMSs have become critical to monitor the operation of SO2 scrubbers 
especially since 2007 (NDRC and SEPA, 2007b). Six plants (all in Table  6.1 
except Plant 2) allowed me to read the computer screens of their CEMSs. The val­
ues of SO2 concentrations changed continuously, and different data were generally 
consistent. Many CEMSs and SO2 scrubbers had been inspected once or twice 
a month. Because CEMSs transmitted data online and in real time, inspections 


Policy implementation  123
often followed abnormal data reporting. Coal-­
fired power plants were informed 
in advance of some inspections, but in many other cases, inspections were unan­
nounced. Inspectors had the right to enter coal-­
fired power plants without being 
delayed. In the plants that I visited, inspection vehicles generally needed just a 
few minutes to drive from the gates to the sites where the SO2 scrubbers were 
installed. China was actively building up its site inspection capacity. The num­
ber of government inspectors at all levels increased steadily (Figure 3.1). China 
focused on monitoring and inspection in its efforts to build capacity. During the 
period between 2006 and 2008, the two functions accounted for 85% of govern­
ment personnel growth for environmental protection (Ministry of Environmental 
Protection, 2006–2009).
Because of the concern about their data accuracy and reliability, as discovered 
in my interviews, CEMSs were not the only data source to track the operation of 
SO2 scrubbers. Other relevant data were collected, including operation and main­
tenance records, load factors of electricity generation, sulfur contents of coal, the 
consumption of limestone and other reagents, electricity consumption, the han­
dling of products from SO2 scrubbers, the opening and closure of bypass dampers 
and records of accidents and responses (NDRC and SEPA, 2007b; SEPA, 2007). 
SO2 concentration in the inlet flue gas corresponds to the sulfur contents within a 
fairly predictable range. The load factors of electricity generation decide the flow 
rate of the flue gas and can check direct measurement with CEMSs. The factors 
together determine the sulfur load to an SO2 scrubber system. For wet scrubbers 
using limestone as the reagent, the molar ratio between CaCO3 and SO2 is nor­
mally quite stable at approximately 1.02 to 1.05 (Ministry of Environmental Pro­
tection, 2010b). Then the sulfur load would decide the consumption of limestone 
and the production of gypsum. The managers of coal-­
fired power plants were 
asked to keep the receipts of limestone purchases, and cheating on receipts was 
considered financial fraud, with harsh penalties on those responsible. Electricity 
is another important input to operate SO2 scrubbers. Because all data should be 
consistent with each other, it became more difficult to cheat.
The problem of collusion appeared under control. Data from CEMSs were sent 
to more than one agency, including environmental protection bureaus and elec­
tric grid corporations. Authorities at China’s four government levels – central, 
provincial, prefectural and county – all inspected SO2 scrubbers. The multiplicity 
of inspection authorities effectively diminished the opportunities of collusion. In 
addition, the pressure to achieve the 10% reduction goal of SO2 emissions in the 
11th Five-­
Year Plan reduced incentives to collude.
4.3  
Building the screening system with big data
The preceding measures to strengthen the diagnosing system indeed worked, 
but for achieving an even deeper reduction of SO2 emissions, China faces much 
more daunting problems in dealing with smaller polluting sources that are a few 
orders of magnitude greater in numbers. New opportunities are emerging with 
newly emerged environmental compliance monitoring technologies (Kitchin, 


124  Policy implementation
2014), which are evolving rapidly in terms of effectiveness in catching noncom­
pliance and efficiency in utilizing compliance monitoring resources. For example, 
CEMSs played a central role in the U.S. Acid Rain Program as well as the Euro­
pean Union Emission Trading Scheme (The U.S. Congress, 1990; Stranlund and 
Chavez, 2000; European Commission, 2012). Remote-­
sensing technologies using 
satellites could provide large-­
scale spatial coverage of multiple pollutants (Streets 
et al., 2013). The measurement extends to areas beyond the current monitoring 
network, although the spatial resolution is coarse (Streets et al., 2013). Social 
media and the prevalent use of smartphones have greatly facilitated and strength­
ened the power of the civil society in monitoring environmental pollution and 
compliance (Stevens and Ochab, 2010; Kay et al., 2015). Various types of sensors, 
in addition to novel carriers such as unmanned aerial vehicles, have been more 
and more widely adopted to measure pollution levels (Snyder et al., 2013; Wang 
and Brauer, 2014).
China has been actively seeking opportunities in big data that can be applied for 
environmental protection. In 2015, State Council formally issued the Action Out­
line for Promoting Big Data Development to encourage the wide integration of 
big data in governance (State Council, 2015). In 2016, the then Ministry of Envi­
ronmental Protection enacted the Comprehensive Plan on Ecological and Envi­
ronmental Big Data Construction (Ministry of Environmental Protection, 2016a). 
It listed a comprehensive plan on how big data could be collected, integrated, 
developed and applied for environmental compliance monitoring, enforcement 
and management.
These new compliance monitoring technologies shed light on new solutions to 
the old challenges. First, in addressing the compliance monitoring resource con­
straint, these technologies could potentially provide a relatively low-­
cost means 
to monitor polluting sources. For example, although one satellite observing the 
Earth’s CO2 and air quality could cost a few hundred million U.S. dollars, such as 
the OCO-­
2 satellite for CO2 monitoring by the National Aeronautical and Space 
Administration with a price tag of US$465 million, its wide spatial and regular 
coverage would substantially reduce the average and, especially, marginal costs 
for one observation (Wall, July 2, 2014; Osterman et al., 2018). Second, many of 
these technologies could circumvent various levels of local governments and pol­
luting sources to provide top-­
down, external and objective data without subjective 
distortions. They are originated from entirely different external sources, not inter­
nal reporting. Satellite or remote-­
sensing data could be gathered in a centralized 
manner without the direct involvement of local governments or polluting sources 
themselves.
Nevertheless, these new technologies also have a critical weakness. Most of 
them generally have not reached the minimum accuracy requirements to legally 
or administratively punish polluters, while conventional technologies currently 
in application (although not all) could fulfill the requirements if intentional data 
manipulation is effectively deterred. Remote-­
sensing data have been successfully 
applied in China to examine the impacts of environmental policies on pollutant 
emissions from coal-­
fired power plants, but the accuracy has not been adequate 


Policy implementation  125
to justify their direct application in legally determining the compliance status of 
individual polluting firms (Zhang et al., 2009; Li et al., 2010).
The trade-­
offs between conventional and new technologies indicate that the 
latter cannot completely replace the former at their current stage, but their clear 
advantages in costs (and objectiveness) are crucial considerations for China’s 
ongoing reform on the conventional diagnosing system to deeply integrate big 
data and other technologies. Section 4.4 mainly focuses on how this reform may 
achieve better efficiency and effectiveness in environmental compliance monitor­
ing. Different technologies are recognized to have different features mainly from 
cost and accuracy perspectives. Their weaknesses and strengths could comple­
ment each other for building a better system than any individual category of tech­
nologies can do alone.
4.4  
Comparing diagnosing and screening systems
Environmental compliance rates are simulated with empirically defined param­
eters as discussed in the Appendix to this chapter. This subsection discusses the 
model simulation and sensitivity analysis results. If any input parameter is not 
targeted in a simulation, it will adopt the empirical value as specified in the current 
scenario as summarized in Table 6.3.
Compliance rates (1−M t) in the screening system depend on their initial levels 
(1
0
−M ; Figure 6.3). For example, if initially with 40,000 inspection staff, the 
screening system results in two equilibrium compliance rates (1−M *) after sev­
eral time steps, about 27% (a very low compliance rate) and 100% (full compli­
ance; Figure 6.3). An equilibrium state is defined as, given the empirical values 
of parameters, the compliance rate remains stable over time and swings back if 
a small disturbance happens (Table 6.3). When the initial compliance rates are 
above a certain level, the final equilibrium compliance rates tend to converge 
to a high level close to full compliance. However, when the initial compliance 
rates are below that level, the available resources would not be adequate to catch 
enough noncompliance cases. Noncompliance will become the dominant choice 
of rational polluters, or the compliance monitoring system falls into a noncompli­
ance trap due to its equilibrium status. The following simulations of the screen­
ing system will primarily report equilibrium compliance rates. In contrast, the 
diagnosing system demonstrates no memory. Its compliance rates at each time-­
step (1−M t) have no relationship with the initial or proceeding levels (1
0
−M  and 
1
1
−
−
M t ). They are decided only by immediately available compliance monitoring 
resources (Rt; Figure 6.3).
The relative effectiveness of the diagnosing and screening systems in enhanc­
ing compliance rates depends heavily on resource availability (Rt; Figure 6.4). 
When resources were too scarce (e.g., less than 30,000 inspection staff or half of 
China’s available personnel in 2015), neither system would be able to result in 
high-­
compliance statuses, although the diagnosing system could achieve slightly 
better outcomes. When resources were abundant (more than 130,000 inspection 
staff or doubling the available personnel in 2015), either system would lead to 


126 Policy implementation
Simulation
6.4
 6.3
 
 6.5(a)
6.5(b)
Figure
Figure
Figure
 
Figure
Empirical ranges in the model 
under special supervisory monitoring 
 
in 2015 (Ministry of Environmental 
Protection, 2002–2016) to 5,925,576 
C
) 
s from 0% (full compliance
s first census of polluting 
simulation
to 100% (complete noncompliance), 
which covers the full range of possible 
compliance rates
number of inspection staff at the 
to 185,108 (the total number of 
environmental officials at all levels 
; 
 varie
0
central and provincial levels in 2015) 
for administration, inspection and 
monitoring in 2015; Ministry of 
Protection, 2002–2016)
Environmental 
in China’
sources with the census date being 
al., 2010)
 scrubbers, 
1a; 
1.6 (very 
31, 2007 (Ministry of 
 
; in the 2007 
P
2
 [Xu, 201
P
C
A, 2007b]) to 
 60% higher than 
for operating SO
s shale-gas development, the 
Environmental Protection et
NDRC and SEP
C
 
 with 
in China’
 was significantly lower than 
 al., 2014])
December
regulation 
 was five times of 
lenient P
P
P
then 
[Guo et
M
From 1,959 inspection staff (the total 
From 68,121 polluting sources that were 
From 0.1 (very harsh 
2018)
s 
, is 
Empirical values in the current 
0
M
scenario
point of the full possible range 
between 0% and 100%.
66,379 inspection staff (in 2015; 
Ministry of Environmental 
Protection, 2002–2016), within 
which 46,800, or 70.5%, were 
environmental inspectors (in 2017) 
as in the “double randomness, one 
publicization” databases (Ministry 
of Ecology and Environment, 
809,500 polluters under compliance 
monitoring as in the “double 
randomness, one publicization” 
databases (Ministry of Ecology and 
Key parameters in the model and their empirical values
Initial noncompliance rate, 
assumed to be 50% as the middle 
Environment, 2018)
 is assumed to be 1.5 times 
P
2/3 (
of the pollution abatement costs, 
being a middle ground in China’
empirical cases as introduced in the 
cell to the right)
 
; 
 
Noncompliance rate (%) at time-step t
the corresponding compliance rate is
 
. Equilibrium noncompliance
, is defined as the level when
When it 
otal available resources for 
compliance monitoring at the time-
, which could be allocated 
between screening and diagnosing, 
 can be changed 
exogenously at a time step. 
.
=
t
R
R
.
0 0
. %
t
R
. 
: the penalty on 
1
P
−
t
d
*
R
t
M
M
rate, 
t
t
−
=
M
M
t
 and 
1 − 
step 
t
Rs
remains a constant: 
 
The number of polluting sources 
under compliance monitoring
 pollution abatement costs (US/sources;TC:ton),whichvaryacrosspollutingnoncompliance(US/ton), which 
is assumed to be fixed for every 
punished polluting source.
R
t
Table 6.3
Parameters
t
 or 
M
R
N
C
P


Policy implementation 127
(b)
(d)
& 
6.5(d)
 
 6.5(c)
 
6.6(a)
6.6(c) & 
Figure
Figure
 
 
Figure
Figure
The impacts of a lognormal distribution 
are also simulated, due to the lack of 
actual information.
Due to inadequate information, the 
ined with a full 
, is exam
s
d
r
r
ratio, 
possible range from 1% to 100%. By 
definition, screening technologies 
must be cheaper than diagnosing 
technologies. Otherwise, the latter will 
be better from both cost and accuracy 
perspectives to make the former 
obsolete.
Due to inadequate information, a full 
range, 0%~100%, is examined.
Due to inadequate information, a full 
range, 0%~100%, is examined.
 normal distribution is assumed with 
a standard deviation of 0.33.
-year per 
 is assumed to be 
 : 0.074 inspector
inspection (see text for empirical 
s
d
r
r
 is equivalently 0.0074 
estimation); 
.
.
s
-year per inspection.
, the corresponding 
10%, or r
, the corresponding 
inspector
1d
 and 
probabilities for screening and 
respectively
2d
K
K
 and 
1s
diagnosing technologies, are 
assumed to be 90% and 99%, 
2s
probabilities for screening and 
diagnosing technologies, are 
assumed to be 70% and 90%, 
respectively
A
d
r
K
K
C
P
 
Cumulative distribution function of 
s
r
vely), which are 
1
Required resources to screen and 
 polluting source (
-
1
K
ype 
 : (T
1
K2
-
one
, respecti
error) the probability that 
diagnose 
r
assumed to remain unchanged over 
d
and 
time
The probability (%) that one 
technology recognizes compliant 
cases as being compliant; 
ype I 
: (T
compliant cases are recognized as 
being noncompliant.
The probability (%) that one 
technology recognizes 
noncompliant cases as being 
noncompliant; 
II error) the probability that 
noncompliant cases are recognized 
as being compliant
•
( )
1
2
Φ
r
K
K


128  Policy implementation
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
5
10
15
20
25
30
35
40
45
50
Inspection staff
e
t
a
r
 
e
c
n
a
i
l
p
m
o
C
Time step
Diagnosing system
Screening system
Campaign
Available inspection
staff (right)
Shock
Figure 6.3  
Model simulation of compliance rates (1−M t) in the diagnosing and screen­
ing systems with available compliance monitoring resources (i.e., exogenously 
determined number of inspection staff in the dashed curve, Rt) and initial com­
pliance rates (1
0
−M , from 0% to 100%)
Note: After compliance rates reach equilibrium levels (1
1
1
0
−
−
−
M
M
M t
*
), a hypothetical envi­
ronmental campaign (temporarily with more inspection staff) is exogenously triggered to run for three 
time-steps and a hypothetical shock (temporarily with fewer inspection staff) for two time-steps. Their 
periods are indicated alongside the dashed curve. All other model parameters adopt the empirical 
values in the current scenario in Table 6.3.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Inspection staff
Screening system
Diagnosing system
Figure 6.4  
Model simulation of equilibrium compliance rates (1−M *) in the screening and 
diagnosing systems in relation to available inspection staff (R)


Policy implementation 129
nearly full compliance and strategies would not matter much. Most situations in 
the real world, including China in 2015 with 66,379 inspection staff, should fall 
in between: resources are constrained but neither unlimited nor depleted. In these 
situations, the two systems would diverge away from each other and the screening 
system could use available resources much more efficiently to achieve signifi-
cantly higher compliance rates (Figure 6.4).
The number of polluters (N) matters greatly for the relative performance of 
the two compliance monitoring systems. With 66,379 environmental inspection 
staff in the current scenario, the screening system shows significantly higher 
compliance rates than the diagnosing system when the number of polluters is 
between 0.5 million to about 1.2 million (Figure 6.5(a)). Both systems could 
effectively handle fewer than 0.5 million polluters for their nearly full compli-
ance, while neither system could be up for the job with more than 1.2 million 
polluters. As discussed in the Appendix at the end of this chapter, the polluting 
sources under the central government’s special supervisory monitoring, gener-
ally large or hazardous polluters, were 68,121 in 2015. The currently available 
inspection staff would be of little resource constraint to achieve their general 
environmental compliance, as in China’s current situation. The “double ran-
domness, one publicization” scheme covered 809,500 polluters, for which the 
screening system with nearly full compliance tends to have a great advantage 
over the diagnosing system with only about half of polluters under compliance. 
If compliance monitoring does not differentiate the 5,925,576 polluting sources 
in the 2007 census, the overall compliance rate would be very low, being less 
than 5%. As in the Chinese practice, compliance monitoring should strategically 
allocate resources to those bigger and more severe polluters. Otherwise, the 
system would be overwhelmed. From another perspective, the model simulation 
also indicates that small polluters have significantly low environmental compli-
ance rates.
As enlightened in the economic theory of crime and punishment, a penalty 
could enhance compliance rates in a similar way as compliance monitoring. When 
C
the penalty level is ten times of the pollution abatement costs (i.e., 
 being 0.1), 
P
both the screening and the diagnosing systems could yield nearly full compli-
ance (Figure 6.5(b)). For example, in ensuring the normal operation of SO2 scrub-
bers, the penalty for noncompliance was five times the pollution abatement costs 
C
(i.e., 
 being 0.2; Xu, 2011a). China’s compliance monitoring system was closer 
P
to the diagnosing system, but it still effectively brought coal-fired 
 
power plants 
under prevalent compliance as projected by the model (Figure 6.5(b); Xu, 2011a). 
C
When the penalty level barely catches up with the abatement costs (i.e., 
  > 1), 
P
neither system would work although the screening system performs even worse 
(Figure 6.5(b)). This was the case in China’s early days in dealing with water pol-
lution in shale-gas development (Guo et
 
 al., 2014).
A deviation of the statistical distribution of the cost/penalty ratio (Φ( )
• ) 
does not seem to cause much difference for the earlier simulation results 


130 Policy implementation
Figure 6.5  
Model simulation of equilibrium compliance rates (1-M *) in the screening 
and diagnosing systems in relation to (a) the number of polluters (N); (b) the 
 
C 
ratios between pollution abatement costs and noncompliance penalty 


; 


P 


(c) available inspection staff (R), where the pollution abatement cost-to-non-
C 
compliance penalty ratio 


 has a lognormal distribution (Φ( )
• ); and (d) the 


P 


r 
relative resource intensity of screening and diagnosing technologies 
s 




r
d 


0%
20%
40%
60%
80%
100%
0
500,000
1,000,000
1,500,000
2,000,000
2,500,000
s
e
t
a
r
e
c
n
a
i
l
p
m
o
c
m
u
i
r
b
i
l
i
u
q
E
Number of polluters
0%
20%
40%
60%
80%
100%
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Pollution abatement costs to noncompliance penalty ratio
(a)
(b)


Policy implementation 131
0%
20%
40%
60%
80%
100%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Inspection staff
0%
20%
40%
60%
80%
100%
0%
10%
20%
30%
40%
50%
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Diagnosing system
Screening system
Unit cost ratio: screening vs diagnosing technologies
(c)
(d)
Figure 6.5 (Continued)
(Figure 6.5(c)). When the pollution abatement cost-to-
 
 
noncompliance penalty 

ratio C 
 is assumed to have a lognormal distribution, the relationship between 



P 



available inspection staff and equilibrium compliance rates is similar to the situ-
ation earlier (Figure 6.5(c)).


132  Policy implementation
Screening technologies should be carefully selected. Otherwise, the screening 
system would not yield higher compliance rates than the diagnosing system. The 
low costs of a screening technology to monitor one polluting source ( rs ) is crucial 
for its better performance (Figure 6.5(d)). It should be no less than 65% cheaper 
than a diagnosing technology (r
d  ; Figure 6.5(d)). Furthermore, in terms of accu­
racy, compliance monitoring technologies for screening and diagnosing have dis­
tinctly different requirements on their Type I and II errors. Screening technologies 
should make fewer Type I errors in wrongly recognizing compliant cases into the 
high-­
risk group (K1s should be generally above 60%; Figure 6.6(a)), while diag­
nosing technologies should make fewer Type II errors in wrongly recognizing 
noncompliant cases as being compliant for them to evade penalties (K2d must be 
generally above 60%; Figure 6.6(d)). The requirements on the other two accuracy 
indicators are much more relaxed. Screening technologies should not put more 
than 80% of noncompliant cases into the low-­
risk group (K2s must be generally 
above 20%; Figure 6.6(c)). The probability of a diagnosing technology to recog­
nize compliant cases as being compliant seems to matter little (K1d; Figure 6.6(b)). 
Although this model assumes only two compliance statuses of a polluter, being 
compliant or noncompliant, polluters do differ in terms of the noncompliance 
severity. In the terminology of this model, compliance monitoring technologies 
should inherently have thresholds on whether to recognize a polluter as being 
compliant or not. The preceding accuracy indicators, especially K2s and K2d, also 
reflect such thresholds. Accordingly, screening technologies only need to catch 
those more severe noncompliant cases or with strong noncompliant signals (due 
to the relaxed requirement of K2s) while diagnosing technologies must convict 
most of these severe noncompliant polluters (K2d). These results could serve as 
the guideline for assessing and selecting screening and diagnosing technologies.
Overall, the screening system in general does show significantly better perfor­
mance than the diagnosing system to achieve higher compliance rates. Depend­
ing on initial compliance rates and available resources, compliance rates in the 
screening system may evolve into two equilibrium levels, being at nearly full 
compliance and prevalent noncompliance. At the 2015 level of inspection staff in 
China, the screening system would be able to yield nearly full compliance for the 
809,500 polluting sources as covered under the “double randomness, one publi­
cization” scheme. However, the diagnosing system that is closer to reality would 
only bring about half of those polluters under compliance.
4.5  
Resilience of screening and diagnosing systems
Campaigns or movements (yundong) are widely used in China’s governance. In 
order to achieve a highly prioritized goal within a short time, the government may 
intensively reallocate unusual amounts of human, financial or political resources 
for certain tasks. These resources are usually “borrowed” from other agencies or 
functions and thus must be “returned” after campaigns conclude. Examples include 
anticrime campaigns, especially “strike hard” (Trevaskes, 2010); anticorruption 
campaigns (Wedeman, 2005); and environmental campaigns (Jahiel, 1998; van 


Policy implementation  133
Figure 6.6  
Model simulation of equilibrium compliance rates (1−M *) in the screening 
and diagnosing systems in relation to the probabilities that (a) the screening 
technology recognizes compliance cases as being compliant (K1s), (b) the diag­
nosing technology recognizes compliance cases as being compliant (K1d), (c) the 
screening technology recognizes noncompliance cases as being noncompliant 
(K2s) and (d) the diagnosing technology recognizes noncompliance cases as 
being noncompliant (K2d)
0%
20%
40%
60%
80%
100%
40%
50%
60%
70%
80%
90%
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Probability (K1s) 
0%
20%
40%
60%
80%
100%
0%
20%
40%
60%
80%
100%
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Probability (K1d)
(a)
(b)


134  Policy implementation
0%
20%
40%
60%
80%
100%
10%
20%
30%
40%
50%
60%
70%
80%
90%
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Probability (K 2s)
0%
20%
40%
60%
80%
100%
50%
60%
70%
80%
90%
100%
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Diagnosing system
Screening system
Probability (K2d)
Figure 6.6  
(Continued)
(d)
(c)


Policy implementation  135
Rooij, 2006). Opposite to environmental campaigns, compliance monitoring might 
also experience shocks as observed in the author’s fieldwork in China, for exam­
ple, when inspection staff in one region or for one environmental task are tempo­
rarily “borrowed” for launching campaigns in another region or for other tasks.
Campaigns can achieve rapid progress on the targeted tasks. However, when 
the temporarily available resources are retreated, such campaign-­
style compliance 
monitoring and enforcement often fail to reach sustained compliance. One nota­
ble example of a largely short-­
lived environmental enforcement campaign is the 
“midnight action” for solving the unacceptable water pollution in the Huai River in 
1997 that shut down about 5,000 small polluting factories (Bai and Shi, 2006; Liu, 
1998). Improvements were achieved in the short term with significantly reduced 
water pollutant emissions and cleaner water quality (Liu, 1998). However, pollu­
tion rebounded quickly after the campaign was over (Bai and Shi, 2006).
The compliance monitoring model as constructed in this study provides an under­
standing of the short-­
lived impacts of environmental compliance monitoring cam­
paigns and shocks in the diagnosing system. The diagnosing system has no memory, 
and its compliance rate at a given time directly corresponds to the immediately avail­
able enforcement resources (Figure 6.3). In contrast, the screening system has a mem­
ory and this feature suggests that short-­
term environmental campaigns might be more 
strategically utilized to establish the screening system for compliance monitoring and 
achieve high compliance rates. As illustrated in Figure 6.3, although compliance moni­
toring resources are kept at the same level before and after environmental campaigns, 
the equilibrium compliance rate will be fundamentally lifted from a low to a high sta­
tus. The compliance rate evolution could be explained with the compliance monitoring 
model. Before a campaign starts, the prevalent noncompliance indicates that a great 
majority of compliance monitoring resources should be spent in the diagnosing step to 
convict polluters. A small proportion of resources will be enough to screen noncompli­
ant cases for the relatively expensive diagnosing. When the environmental campaign is 
launched, with more and more noncompliant polluters being caught in noncompliance, 
their rational decisions will result in higher compliance rates. Then fewer polluting 
sources will be screened into the high-­
risk group in the following time-­
step, which 
requires less resource for diagnosing. In addition, the simulation also suggests that if 
transformed into the screening system, China might reduce the number of environ­
mental inspection staff from the current level but still maintain high compliance rates.
Environmental compliance monitoring shocks have opposite impacts as cam­
paigns. A temporary shortage of inspection staff could destabilize high equilibrium 
compliance rates back to low levels (Figure 6.3). As explained in the model construc­
tion, the compliance rate in the screening system at a time-­
step is only affected by 
that in the previous time-­
step. This short memory leads to the screening system’s 
limited resilience when facing environmental compliance monitoring shocks.
If compliance rates with a longer past contribute to compliance decisions at a 
current time-­
step, the screening system of compliance monitoring will become 
more resilient. A longer memory shows that environmental campaigns should run 
longer for elevating the compliance rate to a higher equilibrium, while temporary 


136  Policy implementation
environmental shocks would be less damaging, with the dipped compliance rate 
quickly rebounding afterward.
Environmental campaigns with temporary increases in inspection staff or 
shocks with their temporary reduction could destabilize the equilibrium rates in 
the screening system with longer-­
term impacts, while the impacts in the diag­
nosing system would be short-­
lived as seen in empirical cases. Environmental 
campaigns might be especially utilized to pull the system out of a possible non­
compliance trap. If polluters have longer memories and their current compliance 
rate is directly determined by those in the past multiple periods, the screening 
system will demonstrate more resilience against the short-­
term campaigns and 
shocks.
Note
	
1	 Adapted with permission from XU, Y. 2011. Improvements in the operation of SO2 scrub­
bers in China’s coal power plants. Environmental Science & Technology, 45, 380–385. 
Copyright (2011) American Chemical Society. Much has been revised and expanded on.
References
Arguedas, C. 2008. To comply or not to comply? Pollution standard setting under costly 
monitoring and sanctioning. Environmental & Resource Economics, 41, 155–168.
Bai, X. & Shi, P. 2006. Pollution control: In China’s Huai river basin: What lessons for sus­
tainability? Environment: Science and Policy for Sustainable Development, 48, 22–38.
Becker, G. S. 1968. Crime and punishment – economic approach. Journal of Political 
Economy, 76, 169–217.
Blackman, A. & Harrington, W. 2000. The use of economic incentives in developing coun­
tries: Lessons From international experience with industrial air pollution. The Journal of 
Environment Development, 9, 5–44.
Bonita, R., Beaglehole, R. & Kjellström, T. 2006. Basic epidemiology. Washington, DC: 
WHO.
Cao, D., Yang, J. & Ge, C. 1999. SO2 charge and tax policies in China: Experiment and 
reform. In: Environmental taxes: Recent developments in China and OECD countries. 
Paris, France: OECD.
Chenery, H. B. 1961. Comparative advantage and development policy. American Eco­
nomic Review, 51, 18–51.
Cohen, M. A. 1999. Monitoring and enforcement of environmental policy. In: Folmer, 
H. & Tietenberg, T. (eds.) The international yearbook of environmental and resource 
economics 1999/2000. Cheltenham: Edward Elgar Publishing Limited.
Dasgupta, S., Wang, H. & Wheeler, D. 1997. Surviving success: Policy reform and the 
future of industrial pollution in China. World Bank Policy Research Working Paper 
1856. Washington, DC: World Bank Group.
The Economic & Trade Commission of Jiangsu Province. 2009. Quarterly thermal effi­
ciencies of electricity generation units. Nanjing, China.
EPA. 2003. Air pollution control technology fact sheet – flue gas desulfurization. EPA-­
452/F-­
03-­
034. Washington, DC: EPA.
EPA & DOE. 2010. Electric power annual 2008. Washington, DC: EPA.


Policy implementation  137
European Commission. 2012. The monitoring and reporting regulation – general guid­
ance for installations [Online]. Available: https://ec.europa.eu/clima/sites/clima/files/
ets/monitoring/docs/gd1_guidance_installations_en.pdf.
Glaeser, E. L. 1999. An overview of crime and punishment. Cambridge, MA: Harvard 
University and NBER.
Gray, W. B. & Deily, M. E. 1996. Compliance and enforcement: Air pollution regulation 
in the US steel industry. Journal of Environmental Economics and Management, 31, 
96–111.
Guo, M. Y., Xu, Y. & Chen, Y. Q. D. 2014. Fracking and pollution: Can China rescue its 
environment in time? Environmental Science & Technology, 48, 891–892.
Heindl, P.  2012. Transaction costs and tradable permits: Empirical evidence from the 
EU emissions trading scheme. ZEW Discussion Paper No. 12-­
021 [Online]. Available: 
https://www.econstor.eu/handle/10419/56029.
Helland, E. 1998. The enforcement of pollution control laws: Inspections, violations, and 
self-­
reporting. Review of Economics and Statistics, 80, 141–153.
Henan Department of Environmental Protection. 2010. Information on the collection of 
SO2 effluent discharge fee from province-­
regulated coal-­
fired power plants in the first 
quarter of 2010. Zhenzhou, China: Henan Department of Environmental Protection.
Henan Development and Reform Commission & Henan Environmental Protection Bureau. 
2007. A notice to transfer NDRC and SEPA’s policy: Management measures on desulfur­
ized electricity price and the operation of desulfurization facilities of coal-­
fired power 
generators (on trial). Zhenzhou, China: NDRC and SEPA.
IEA. 2009. Cleaner coal in China. Paris, France: IEA.
Jahiel, A. R. 1998. The organization of environmental protection in China. The China 
Quarterly, 156, 757–787.
Jiangsu Department of Environmental Protection. 2007–2009. Monthly report on the oper­
ation of SO2 scrubbers at coal-­
fired power plants. Nanjing, China: Jiangsu Department 
of Environmental Protection.
Jiangsu Department of Environmental Protection. 2008. On strengthening the collection of 
effluent discharge fee from coal-­
fired power plants. Nanjing, China: Jiangsu Department 
of Environmental Protection.
Jin, Y. N., Andersson, H. & Zhang, S. Q. 2016. Air pollution control policies in China: 
A retrospective and prospects. International Journal of Environmental Research and 
Public Health, 13.
Kay, S., Zhao, B. & Sui, D. 2015. Can social media clear the air? A case study of the air 
pollution problem in Chinese cities. Professional Geographer, 67, 351–363.
Kitchin, R. 2014. The data revolution. Thousand Oaks, CA: Sage Publications.
Konrad, K. A., Lohse, T. & Qari, S. 2017. Compliance with endogenous audit probabilities. 
Scandinavian Journal of Economics, 119, 821–850.
Kruger, J. & Egenhofer, C. 2006. Confidence through compliance in emissions trading 
markets. Sustainable Development Law & Policy, 6, 2–13, 63–64.
Levitt, S. D. 2004. Understanding why crime fell in the 1990s: Four factors that explain the 
decline and six that do not. Journal of Economic Perspectives, 18, 163–190.
Li, C., Zhang, Q., Krotkov, N. A., Streets, D. G., He, K. B., Tsay, S. C. & Gleason, J. F. 
2010. Recent large reduction in sulfur dioxide emissions from Chinese power plants 
observed by the ozone monitoring instrument. Geophysical Research Letters, 37.
Lin, J. Y., Cai, F. & Li, Z. 2003. The China miracle: Development strategy and economic 
reform. Sha Tin: Published for the Hong Kong Centre for Economic Research and the 
International Center for Economic Growth by the Chinese University Press.


138  Policy implementation
Liu, H. 1998. An overview of water pollution prevention and control in the Huai River val­
ley. Enviornmental Management in China, 5–8.
Liu, Z., Guan, D. B., Wei, W., Davis, S. J., Ciais, P., Bai, J., Peng, S. S., Zhang, Q., Hubacek, 
K., Marland, G., Andres, R. J., Crawford-­
Brown, D., Lin, J. T., Zhao, H. Y., Hong, C. 
P., Boden, T. A., Feng, K. S., Peters, G. P., Xi, F. M., Liu, J. G., Li, Y., Zhao, Y., Zeng, 
N. & He, K. B. 2015. Reduced carbon emission estimates from fossil fuel combustion 
and cement production in China. Nature, 524, 335.
Lu, X., Dudek, D. J., Qin, H., Zhang, J., Lin, H., Yang, Z. & Wang, Y. 2006. Survey on the 
capacity of environmental administrative enforcement in China. Research of Environ­
mental Sciences, 19, 1–11.
Lu, Z., Streets, D. G., Zhang, Q., Wang, S., Carmichael, G. R., Cheng, Y. F., Wei, C., Chin, 
M., Diehl, T. & Tan, Q. 2010. Sulfur dioxide emissions in China and sulfur trends in East 
Asia since 2000. Atmospheric Chemistry and Physics, 10, 6311–6331.
Lu, Z., Zhang, Q. & Streets, D. G. 2011. Sulfur dioxide and primary carbonaceous aero­
sol emissions in China and India, 1996–2010. Atmospheric Chemistry and Physics, 11, 
9839–9864.
McAllister, L. K., Van Rooij, B. & Kagan, R. A. 2010. Reorienting regulation: Pollution 
enforcement in industrializing countries. Law & Policy, 32, 1–13.
MEP. 2013. The 12th five-­
year plan on capacity building of environmental regulations. 
Beijing, China: MEP.
MEP & AQSIQ. 2011. Emission standard of air pollutants for thermal power plants. Bei­
jing, China: MEP, AQSIQ.
MEP & NDRC. 2008. Announcement on punishing coal-­
fired power plants with abnormal 
operation of SO2 scrubbers in 2007. Beijing, China: MEP, NDRC.
Ministry of Ecology and Environment. 2018. The comprehensive implementation of ‘dou­
ble randomnesses, one publicization’ for environmental inspection. Beijing, China: Min­
istry of Ecology and Environment.
Ministry of Environmental Protection. 2002–2016. Annual statistical report on the envi­
ronment in China. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2006–2009. Annual statistical report on the envi­
ronment in China. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2008. Statement to penalize coal-­
fired power plants 
for the abnormal operation of their SO2 scrubbers in 2007. Beijing, China: Ministry of 
Environmental Protection.
Ministry of Environmental Protection. 2009a. 2008 assessment report on provincial major 
pollutant emissions. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2009b. Information on the mitigation of major pol­
lutants in 2008. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2009c. Statement to penalize five coal-­
fired power 
plants for the abnormal operation of their SO2 scrubbers in 2008. Beijing, China: Min­
istry of Environmental Protection.
Ministry of Environmental Protection. 2010a. China’s capacities of water treatment plants 
and SO2 scrubbers at coal-­
fired power plants. Beijing, China: Ministry of Environmen­
tal Protection.
Ministry of Environmental Protection. 2010b. Guideline on best available technologies of 
pollution prevention and control for coal-­
fired power plant industry (on trial). Beijing, 
China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2016a. Comprehensive plan on ecological and envi­
ronmental big data construction. Beijing, China: Ministry of Environmental Protection.


Policy implementation  139
Ministry of Environmental Protection. 2016b. final accounts of ministry of environmental 
protection. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection, National Statistics Bureau  & Ministry of Agri­
culture. 2010. Public report on the first national census of polluting sources. Beijing, 
China: Ministry of Environmental Protection.
National Bureau of Statistics. 2008. China energy statistical yearbook. Beijing, China: 
China Statistics Press.
NDRC. 2004. Technical code for designing flue gas desulfurization plants of fossil fuel 
power plants. DL/T5196-­
2004. Beijing, China: NDRC.
NDRC & SEPA. 2007a. The 11th five-­
year plan on SO2 control in existing coal-­
fired power 
plants. Beijing, China: State Environmental Protection Administration, NDRC.
NDRC & SEPA. 2007b. Management measures on desulfurized electricity price permium 
and the operation of desulfurization facilities in coal-­
fired power generators (on trial). 
Beijing, China: State Environmental Protection Administration, NDRC.
OECD. 2006. Environmental compliance and enforcement in China – an assessment of 
current practices and ways forward. Paris, France: OECD.
Ohlin, B. 1967. Interregional and international trade. Cambridge, MA: Harvard Univer­
sity Press.
Osterman, G., Eldering, A., Avis, C., Chafin, B., O’Dell, C., Frankenberg, C., Fisher, B., 
Mandrake, L., Wunch, D., Granat, R. & Crisp, D. 2018. Orbiting carbon observatory – 2 
(OCO-­
2) data product user’s guide, operational L1 and L2 data versions 8 and lite file 
version 9. Pasadena, CA: Jet Propulsion Laboratory, California Institute of Technology.
Pan, L., Wang, Z. & Wang, Z. 2005. Present status and countermeasure suggestion for 
thermal power plants CEMS in China. Research of Environmental Sciences, 18, 42–45.
Polinsky, A. M. & Shavell, S. 2000. The economic theory of public enforcement of law. 
Journal of Economic Literature, 38, 45–76.
Raufer, R. & Li, S. Y. 2009. Emissions trading in China: A conceptual ‘leapfrog’ approach? 
Energy, 34, 904–912.
Ricardo, D. 1817. On the principles of political economy and taxation. London: J. Murray.
Russell, C. S. & Vaughan, W. J. 2003. The choice of pollution control policy instruments 
in developing countries: Arguments, evidence and suggestions. In: Folmer, H. & Tieten­
berg, T. (eds.) International yearbook of environmental and resource economics. Chel­
tenham: Edward Elgar.
SEPA. 2006. Guidelines on calculating SO2 emission quotas. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2007. Verification of major pollutants emission reduction in the 11th five-­
year 
period (on trial). Beijing, China: State Environmental Protection Administration.
SERC. 2009. New credit to CEMSs at SO2 scrubbers in Jiangsu province. Nanjing, China: 
SERC.
SERC, NDRC, National Energy Bureau & MEP. 2009. Report on electric industry’s energy 
conservation and pollutant mitigation in 2008. Beijing, China: SERC, NDRC, MEP.
Shimshack, J. P.  2014. The economics of environmental monitoring and enforcement. 
Annual Review of Resource Economics, 6, 339–360.
Snyder, E. G., Watkins, T. H., Solomon, P. A., Thoma, E. D., Williams, R. W., Hagler, G. S. 
W., Shelow, D., Hindin, D. A., Kilaru, V. J. & Preuss, P. W. 2013. The changing paradigm 
of air pollution monitoring. Environmental Science & Technology, 47, 11369–11377.
State Commission Office for Public Sector Reform. 2018. Regulations on the authorities, 
organization and personnel of the ministry of ecology and environment. Beijing, China: 
State Commission Office.


140  Policy implementation
State Council. 2007a. Notice on distributing composite working plan on energy conserva­
tion and pollutant emission reduction. Beijing, China: State Council.
State Council. 2007b. Notice on distributing implementation plans and methods of statis­
tics, monitoring and assessment on energy conservation and pollutant emission reduc­
tion. Beijing, China: State Council.
State Council. 2015. Action outline for promoting big data development. Beijing, China: 
State Council.
State Council. 2019. Advice on the comprehensive implementation of joint ‘double ran­
domnesses, one publicization’. In: Market regulation. Beijing, China: State Council.
State Development Planning Commission, Ministry of Finance, SEPA & State Economic 
and Trade Commission. 2003. Measures for the administration of the charging rates for 
pollutant discharge fees. Decree No. 31 [Online]. Available: http://www.lawinfochina.
com/display.aspx?id=2705&lib=law&EncodingName=big5.
State Electricity Regulation Commission (Nanjing Office). 2009. Information on the oper­
ation of SO2 scrubbers in coal-­
fired power plants in Jiangsu Province. Nanjing, China: 
SERC.
State Environmental Protection Administration  & General Administration of Quality 
Supervision Inspection and Quarantine. 2003. Emission standard of air pollutants for 
thermal power plants. GB 13223-­
2003. Beijing, China: State Environmental Protection 
Administration.
Steinfeld, E. S., Lester, R. K. & Cunningham, E. A. 2009. Greener plants, grayer skies? 
A report from the front lines of China’s energy sector. Energy Policy, 37, 1809–1824.
Stevens, M.  & Ochab, B. 2010. Participatory noise pollution monitoring using mobile 
phones. Information Polity: The International Journal of Government & Democracy in 
the Information Age, 15, 51–71.
Stranlund, J. K. & Chavez, C. A. 2000. Effective enforcement of a transferable emissions 
permit system with a self-­
reporting requirement. Journal of Regulatory Economics, 18, 
113–131.
Streets, D. G., Canty, T., Carmichael, G. R., De Foy, B., Dickerson, R. R., Duncan, B. N., 
Edwards, D. P., Haynes, J. A., Henze, D. K., Houyoux, M. R., Jacobi, D. J., Krotkov, 
N. A., Lamsal, L. N., Liu, Y., Lu, Z. F., Martini, R. V., Pfister, G. G., Pinder, R. W., 
Salawitch, R. J. & Wechti, K. J. 2013. Emissions estimation from satellite retrievals: 
A review of current capability. Atmospheric Environment, 77, 1011–1042.
Trevaskes, S. 2010. Policing serious crime in China: From ‘strike hard’ to ‘kill fewer’. 
London and New York: Routledge.
Tsinghua University. 2010. A study on the management system of environmental pollution 
data collection in China. Beijing, China: Tsinghua University Press.
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. Congress.
The U.S. EPA. 2007. CEMS cost model. Washington, DC [Online]. Available: www.epa.
gov/ttn/emc/cem/cems.xls.
Van Rooij, B. 2006. Implementation of Chinese environmental law: Regular enforcement 
and political campaigns. Development and Change, 37, 57–74.
Vossler, C. A. & Gilpatric, S. M. 2018. Endogenous audits, uncertainty, and taxpayer assis­
tance services: Theory and experiments. Journal of Public Economics, 165, 217–229.
Wall, M. 2014. NASA launches satellite to monitor carbon dioxide, July 2 [Online]. Avail­
able: www.space.com/26403-­
nasa-­
oco2-­
carbon-­
dioxide-­
satellite-­
launch.html.
Wang, A. & Brauer, M. 2014. Review of next generation air monitors for air pollution [Online]. 
Available: https://www.semanticscholar.org/paper/Review-­
of-­
Next-­
Generation-­
Air-
­
Monitors-­
for-­
Air-­
Wang-­
Brauer/ca660dbe9fa087b2bea4080064c251914155c177.


Policy implementation  141
Wedeman, A. 2005. Anticorruption campaigns and the intensification of corruption in 
China. Journal of Contemporary China, 14, 93–116.
Wu, J. & Qian, Y. 2007. The construction of FGD at Beilun power plant on its 5 × 600 MWe 
units. Electrical Equipment, 8, 105–107.
Xu, Y. 2011a. Improvements in the operation of SO2 scrubbers in China’s coal-­
fired power 
plants. Environmental Science & Technology, 45, 380–385.
Xu, Y. 2011b. The use of a goal for SO2 mitigation planning and management in China’s 
11th five-­
year plan. Journal of Environmental Planning and Management, 54, 769–783.
Zhang, Q., Streets, D. G., Carmichael, G. R., He, K. B., Huo, H., Kannari, A., Klimont, 
Z., Park, I. S., Reddy, S., Fu, J. S., Chen, D., Duan, L., Lei, Y., Wang, L. T. & Yao, Z. L. 
2009. Asian emissions in 2006 for the NASA INTEX-­
B mission. Atmospheric Chemis­
try and Physics, 9, 5131–5153.
Zhang, Q., Streets, D. G., He, K., Wang, Y., Richter, A., Burrows, J. P., Uno, I., Jang, 
C. J., Chen, D., Yao, Z. & Lei, Y. 2007. NOx emission trends for China, 1995–2004: 
The view from the ground and the view from space. Journal of Geophysical Research-­
Atmospheres, 112.
Zhejiang Bureau of Quality and Technical Supervision. 2007. The quota & calculation 
method of coal consumption for generating station. Hangzhou, China: Zhejiang Bureau 
of Quality and Technical Supervision.


­
Appendix
Modeling environmental compliance 
monitoring systems
Key parameters
This model’s primary output is the compliance rate of polluters under compli-
ance monitoring: M t is the noncompliance rate at the end of the time-step 
 
t, 
while 1-M
 
t is the corresponding compliance rate. For simulating the evolution 
of compliance rates over time, the model is designed to follow time-steps. In
 
 
each time-step, 
 
enforcement activities are first conducted to comprise compli-
ance monitoring and penalty on noncompliance. The compliance rate at the end 
of the previous time-step 
 
could affect the subsequent performance of environ-
mental compliance monitoring, that is, the probability of catching noncompli-
ance. This probability is assumed to be commonly available information for all 
polluters. Based on the expected penalty and compliance costs, polluters make 
compliance decisions to yield an overall compliance rate at the end of the cur-
rent time-step.
As listed in Table 6.3, the model has a series of input parameters, whose values 
are given exogenously. They fall into several major categories: (1) environmental 
compliance monitoring system, including initial noncompliance rate (M 0), the 
total available resources for compliance monitoring (Rt) and the number of pol-
luting sources (N); (2) the ratio between pollution abatement costs and penalty on 

noncompliance C 
 as well as its distribution (

Φ

( )
• ); (3) compliance monitoring 




technologies, including 
P 
required resources for monitoring one polluting source 
(r), the probability that one technology recognizes compliant cases as being com-
pliant (K1) and the probability that one technology recognizes noncompliant cases 
as being noncompliant (K2). Screening and diagnosing technologies are further 
distinguished with subscripts s and d, respectively. The four parameters (K1s, K1d, 
K2s and K2d) are assumed to be specific for a given compliance monitoring technol-
ogy and do not change over time and cases. Compliance-monitor
 
ing technologies 
and systems could make two types of errors in identifying noncompliance (Polin-
sky and Shavell, 2000; Bonita et al., 2006). We assume H0: a polluter is under 
environmental compliance. A Type I error indicates that a polluting firm is under 
compliance, but the environmental compliance monitoring wrongly identifies the 
case as noncompliance to mistakenly punish it. A Type II error refers to the situ-
ation that although a polluter is not complying, the system wrongly recognizes it 
as being compliant. Accordingly, the illegal polluter walks away without penalty. 


Policy implementation 143
Both errors consequently lower the deterrence effect, which might lead to lower 
compliance rates.
The diagnosing system
In this diagnosing-only system, the probability 
 
of noncompliant polluting sources 
Rt
that are rightfully punished is: 
d
1
´
´K
r
N
2d, while the probability of compliant 
polluting sources that are mistakenly punished is 
d
Rt
d
1
×
×(
)
1−K
r
N
1d . A polluter 
Rt
d
Rt
will choose compliance when C
P
+
×
d
1
×
×(
)
1−
<
K
P ×
×
d
1
d
×K
r
N
1
2
r
N
d, or 
C
Rt
d
d
<
×
d
1 ×
+
(
)
K
K
2
1
d
d −1 . Because all resources are devoted to diagnosing, 
P
r
N
d
R
R
t =
t
d. One polluting source could be diagnosed more than once to potentially 
incur a penalty every time that it is caught noncompliance. Corresponding to 
available enforcement resources, the noncompliance rate will be
Rt
M t
1
=
−
1
Φ(
(
×
× K
K
r
N
2
1
d
d
+
−1)) 
Equation 6.1
d
Because M t is not related to M t-1, the noncompliance rate under the diagnosing 
system will not show dynamic evolution over time when other factors remain 
unchanged.
The screening system
When resources are inadequate, some polluting sources may be neither screened 
nor diagnosed, while the optimal allocation of resources will make sure that 
all screened-out polluting 
 
sources in the high- 
risk group are diagnosed and no 
available resource is wasted. Due to the existence of Type I and II errors, each 
group contains compliant and noncompliant sources. The high- 
risk group in the 
Rt
time-step t will comprise 
s ×
×
M
K
t−1
2s noncompliant polluting sources and 
Rt
r
s
s ×
−
(
)
1
1
M
K
t−1 ×
−
(
)
1s  compliant polluting sources. Accordingly, the noncom-
r
pliance rate in the high-
s
M
K
t−1
risk group is 
 
M t
×
h =
2s
t−
−
1
t
1
.
M
K
×
+(
)
1
1
−
×
M
K
(
)
−
The low- 
risk group will contain all remaining polluting 
2s
sources, including 
1s
those 
Rt
Rt
screened out and those not screened, 
N −
×
s
M
K
t−
−
1 ×
−
s ×
−M
K
t
1
2s
(
)
1
1
×
−
(
)
1s , 
r
s
r
Rt
Rt
Rt
s
or (
)
N −
+
s
s ×
×
M
K
t−
−
1
(
)
1
1
−
+
s ×
−
(
)
M
K
t
1
2s
×
1s. The number of noncom-
r
s
r
s
r
s Rt
pliant polluting sources is N
M
×
−
t−
−
1
1
s ×
×
M
K
t
2s. Then the noncompliance 
r
s
Rt
N
M
×
−
t−
−
1
1
s ×
×
M
K
t
r
2s
rate in the low-risk group is 
 
M t =
s
l
Rt
N
s
M
K
t
1
Rt
.
−
×
−
−
×
−
s
2s
×
−
(
)
1
M t
1 ×(1−K
r
s
r
1s)
s
­


144 Policy implementation
After diagnosing, the number of noncompliant polluting sources that are 
Rt
rightfully punished is 
d ´
´
M
K
t
r
h
2d. The probability of noncompliant pollut-
t
d
Rd ×
×
M
K
t
r
h
2d
d
Rt
M t
ing sources that are rightfully punished is 
d
1
=
×
×
×
h
K
N
M
×
t−
−
1
1
r
N
M t
2d. 
The number of compliant polluting sources that are mistakenly pun
d
-
ished is 
Rt
d
M
K
r ´ -
(
)
1
1
t
h ´ -
(
)
1d C : \ wspath\ WS5551\ Math_Preference\ Equat
tion\ pref\ Euclid.eqp
d
, and the corresponding probability is 
Rt
d ×
−
(
)
1
1
M
K
t
r
h ×
−
(
)
1d
Rt
1
M t
d
=
×
d
1−
×
h ×
−
(1
K ). We assume that the two 
N
M
×
−
(
)
1
t−
−
1
1
r
N
1−M t
1d
d
probabilities are known to all polluting sources for their following compliance 
decisions.
Rt
−
Thus, the expected compliance cost is C
P
d
1
1
M t
+
×
×
×
h ×
−
(
)
1
K
r
N
1−M t−1
1d , 
d
Rt
1
M t
while the expected penalty on noncompliance is P ×
×
d
×
×
h
K . 
r
N
d
d
M t−1
2  
For a decision of compliance, the former should be lower than the latter: 
Rt
1
1−M t
Rt
1
M t
t
d
h
1
C
P
+
×
×
×
×
−
(
)
1
K
P
d
h
C
R
1d <
×
×
×
×K
<
×
1
M t−
−
1
r
N
t
1
2d , or 
d
×
r
N
d
−
d
M
P
r
N
d
K
K
2
2
s
d
×
−(
)
1
1
−
×
K
K
1
1
s
d
(
)
−
r
N
M
K
t−
−
1 ×
+
s
1
1
M
K
t
1
.
2
(
)
−
×( −
1s)
Rt
screened 
min  is further defined as a threshold when all polluting sources have just been 
(R
N
t
s =
×r
s ), all polluting sources in the high-risk group are 
 
diagnosed 
Rt
t
(Rt =
×
(
(
s
M
K
t−
−
1
Rs
d
×
+
s
1
1
M
K
t
2
×
−
1)
(
×
−
1s
d
))×r ) and all resources are uti-
r
s
r
lized 
s
R
R
t =
+
t
s
Rt . Then R
N
t
t−
−
1
t
1
d
min =
×(
(
r
M
s +
×K
M
2s +
−
(
)
1
1
×
−
(
)
K
r
1s
d
)
)
×
.
When R
R
t
t
Rt
£
Rt =
min , 
d
. The com-
1
(
+1)
r
d ×
×
(
(
M
K
t−
−
1
+
−
1
2
×
−
r
s
1
1
M
K
t
)
(
1s))
s
C
Rt
K
K
2
2
s
d
×
−(
)
1
1
−
×
K
K
(
)
pliance condition is 
1
1
−
<
×
s
d
r
M t
.
P
N
(
(
s +
×
−
−
1
K
M
s
1
1
t
1
2 +
−
(
)×
−
(
K1s
d
))×r )
Additional compliance monitoring resources beyond Rt
min will be devoted 
to diagnosing those polluting sources in the high- 
risk group. These sources 
could be diagnosed and punished once or multiple times. In this situation, 
R
R
t
t
d =
−R
R
t
t
s =
−N
r
× s.
Then corresponding to available enforcement resources, the noncompliance 
rate at the end of time-step 
 
t will be

t

If R
R
t
t

−
£
t
×
−
, 
R
K
K
(
)
1
1
−
×
K
K
(
)

min M =
−
1
Φ
×
2
2
s
d
1
1


s
d


N
(
(
r
M
+
×
t−
−
1
K
M
+
−
(
)
1
1
t
1
; Equation 6.2

s
2s
×( −
×
K
r 

1s
d
))
)
K
K
M
K
s
d
t
s
×
−
×
+
−
2
2
1
2
(
(
C
P
Rd
t
d
<
×
×
1


Policy implementation 145

If 
t

t
t , 
t

R
N
−
×r
K
K
×
−(
)
1
1
−
×
K
K
(
)
−

R
R
>
min M =
−
1
Φ

s ×
2
2
s
d
1
1
s
d

.  
Equation 6.3




N
r
×
t
d
M
K
−
−
1 ×
+
s
(
)
M t
1
2
1−
×
−
(
)
1
K1 

s
Rt  will be greater if the noncompliance rate at the end of time-step 
 
t-1, M t-1
is higher or screening and diagnosing are more resource-
min
, 
 
intensive with greater 
r
s and r
individual 
d. Given a certain amount of total emissions under regulation, smaller 
polluting sources will result in a greater number of polluting sources, 
N, for compliance monitoring and thus higher demand for resources. Because the 
noncompliance rate, M t, changes over time, Rt  will change accordingly.
More accurate compliance monitoring technologies (
min
K
K
1
1
s
d
,
,K
K
2
2
s
d
,
® 1) with 
lower costs for an average polluting source (r r
,
® 0) tend to induce higher com-
pliance rates. Various factors could affect the availability 
s
d
of enforcement resources 
Rt
per polluting source (
). The economy of scale in compliance monitoring could 
N
have two folds. On one hand, larger polluting sources could lead to an internal 
economy of scale because the required enforcement resources are more related to 
the number of sources. More enforcement resources, larger polluting sources and 
a smaller amount of total emissions will increase the resource availability indica-
tor. Even if with the screening step or effective compliance monitoring strategy, 
the probability of catching enough noncompliance cannot be enhanced to a high 
enough level without sufficient enforcement resources. On the other hand, the 
geographical proximity of polluting sources could provide an external economy 
of scale. The sources could then be equivalently bundled and reduce the compli-
ance monitoring costs for one polluting source.
Corresponding to their required features, screening technologies are less accu-
rate but also less expensive than diagnosing technologies. They must have such 
trade- 
offs to fit in the expected complementary roles. If one technology were both 
cheaper and more accurate than the other, the latter technology would be entirely 
replaced by the former.
Input parameters in China’s empirical case
In order to empirically illustrate and analyze the model, the input parameters will 
adopt empirical values from the Chinese context. A current scenario and the range 
of parameters are defined with the best available empirical data in China’s current 
situation. They are briefly summarized in Table 6.3, and this subsection provides 
a more detailed explanation.
Available resources for compliance monitoring (Rt) are a key input parameter 
that this model focuses on. For simplicity, compliance-monitoring 
 
resources (Rt) 
and costs of screening and diagnosing technologies (r
s and r
d) are counted as the 
number of environmental inspection staff. China has been gradually increasing 
governmental employees for environmental inspection. The resource availabil-
ity still faces constraints, but it does not fall into the situation of extreme scar-
city. From 2001 to 2015, staff for environmental inspection grew from 37,934 to 
66,379 (Ministry of Environmental Protection, 2002–2016). More important, with 
­


146 Policy implementation
the full establishment of regional supervisory centers/bureaus in 2008 by the then 
Ministry of Environmental Protection, the central government has significantly 
strengthened its capacity of environmental inspection, accounting for 0.48% (294 
employees) of inspection staff at all four levels in 2009 and 0.82% (542 employ-
ees) in 2015, up from 0.07% in 2008 (41 employees; Ministry of Environmental 
Protection, 2002–2016). Six regional Supervision Bureaus were allowed to have, 
in total, 240 formal employees for taking charge of supervision tasks within their 
jurisdictions (State Commission Office for Public Sector Reform, 2018). Not all 
staff employed in the inspection section are environmental inspectors, for exam-
ple, to play supporting roles such as office work. In 2017, China had 46,800 envi-
ronmental inspectors in the databases for “double randomness, one publicization” 
(Ministry of Ecology and Environment, 2018). The closest year with available 
data on inspection staff was 2015. Accordingly, about 70.5% of inspection staff
 
were environmental inspectors. The empirical model simulation adopts this ratio 
to examine the impacts of resource availability on environmental compliance 
rates. The current scenario thus has 66,379 inspection staff, or 46,800 environ-
mental inspectors. If not specified, they will remain unchanged over time.
The number of polluting sources (N) was been briefly described in Section 4.2. 
The current scenario takes the intermediate number, 809,500 polluting sources as 
targeted in 2017 under the “double randomness, one publicization” scheme.
Pollution abatement costs and the associated penalty for noncompliance range 
across sectors, technologies and severity of noncompliance. The ratio between 

compliance costs and penalty C 
 is a key variable in this compliance monitoring 



P 



model. In 2007, in order to tackle the long-term 
 
problem of weak environmental 
policy enforcement, China not only subsidized those coal-fired 
 
power plants to 
normally operate their SO2 scrubbers but, more important, also issued a penalty, 
being five times of the subsidy/costs on a per- 
kilowatt-hour 
 
basis (Xu, 2011a; 
NDRC and SEPA, 2007b). In dealing with potential noncompliance on water pol-
lution and withdrawal, however, China’s penalty was barely able to catch up with 
the pollution abatement costs (Guo et al., 2014). In the current scenario, the cost/
penalty ratio is assumed to be 2/3. Furthermore, pollution abatement costs are not 
identical across polluting firms due to, for example, economy of scale, the sulfur 
content of coal and whether the pollution removal facility is a retrofit or built 
together with the main equipment. In compiling China’s SO2 emission inventory, 
Lu et al. (2011) assumed that the sulfur content had a normal distribution. The 
current scenario follows, due to the key influence of sulfur contents on SO2 abate-

ment costs, to assume that the cost/penalty ratio C 
 has a normal distribution 


) among the polluting sources.

P 


(Φ( )
•

The costs of screening and diagnosing technologies are accounted as the 
required number of inspectors in a year per environmental observation, either 
screening or diagnosing inspection (inspector-year 
 
per observation, being noted 
as r
randomness, one publicization” 
s and r
d, respectively). China has comprehensively established the “double 
method for governmental, including environ-
mental and other, inspections on firms (State Council, 2019). For environmental 


Policy implementation  147
inspections, the method had been well established in 2017 (Ministry of Ecology 
and Environment, 2018). Under this method, polluting firms and environmental 
inspectors will both be randomly selected from databases, while the information 
will be publicized to the public. In 2017, 809,500 polluting firms and 46,800 envi­
ronmental inspectors were included in the databases, while 632,600 environmen­
tal inspections were conducted (Ministry of Ecology and Environment, 2018). 
Accordingly, 27 inspections were conducted by an average inspector in 2017. 
According to the author’s earlier fieldwork in China (Guo et al., 2014; Xu, 2011a), 
one inspection generally involves two inspectors. Thus, the cost of environmental 
inspection or diagnosing technology (r
d) was 0.074 inspector-­
year per inspection. 
It is adopted in the current scenario.
Different screening and diagnosing technologies have different cost structures. 
For example, a sophisticated satellite-­
based technology has very high initial 
capital costs, but its marginal costs of monitoring one more pixel are negligi­
ble. For example, OCO-­
2 cost US465milliontosetup,butwithmorethan100,000measurementsofcolumnCO2concentrationseachday(Ostermanetal.,2018),eachmeasurementsinceitslaunchinJuly2014costmerelyaboutUS2 
to US$3 per measurement, considering neither operation and maintenance costs 
that will raise the unit cost nor expected longer lifetime that will reduce the 
unit cost. According to the author’s fieldwork in China’s coal-­
fired power plants, 
continuous emissions monitoring system (CEMS) costs about 500,000 RMB/set 
around 2010. China has been publishing hourly data from CEMSs in key pollut­
ing sources. With an expected lifetime of approximately 5 to 10 years, the unit 
cost would also be about US$1 to US$2 per published data point. Screening often 
requires multiple observations. OCO-­
2 has a 16-­
day ground-­
track repeat cycle 
to result in about 23 repeated observations per year for one pixel, or at a cost of 
roughly US$50 per year. CEMSs in China could provide more than 8,000 hourly 
observations per year and have an annual cost of about US8,000toUS16,000. 
Accordingly, the costs of an average screening technology are assumed to be 
in the range of several hundred U.S. dollars per year for one polluting source. 
In contrast, compliance monitoring by environmental inspectors is cheaper to 
set up but more expensive to operate. For example, China in 2015 at the cen­
tral level had 542 employees for environmental inspections (Figure 3.1), with 
a total cost of 63.5 million RMB (~US10.2millionin2015exchangerate,orUS18,800/person-­
year; Ministry of Environmental Protection, 2016b). Accord­
ingly, the average cost of one inspection was about 0.074 inspector-­
year/inspec­
tion / 70.5% × US$18,800/person-­
year, or US$2,000/inspection. In the current 
scenario, the unit cost of a screening technology (rs) is then assumed to be one 
order of magnitude cheaper than that of screening technology, or equivalently 
0.0074 inspector-­
year per screening round.
The compliance monitoring accuracy of one technology is hard to exactly 
measure, because only data on observed compliance and noncompliance are avail­
able but not those on absolute truth. Furthermore, the dichotomy of compliance 
and noncompliance does not measure the severity of noncompliance, while more 
severe cases, due to their stronger signal-­
to-­
noise ratios, tend to be easier to catch. 


148  Policy implementation
In theory, the screening strategy would only work when the noncompliance rate 
in the high-­
risk group is higher than that in the low-­
risk group. The more different 
their noncompliance rates between these groups gap are, the better the screening 
strategy will be. In the current scenario, K1s, K1d, K2s and K2d are assumed to be 
90%, 99%, 70% and 90%, respectively.


7	

Environmental technology 
and industry1
1  
Goal-­
centered SO2 mitigation path
Besides other critical measures, pollution mitigation often involves facilities 
such as those installed in coal-­
fired power plants to remove sulfur oxide (SO2), 
nitrogen oxide (NOx), particles, mercury and carbon dioxide (CO2), together with 
renewable-­
energy facilities for reducing coal consumption such as wind turbines 
and solar panels and hybrid and electric vehicles. Two major factors determine 
how rapidly a country could utilize these facilities for pollution mitigation. First, 
there must be a strong demand for their rapid deployment and normal operation, as 
examined in detail in Chapters 5 and 6. Second, if the demand is put in place, 
enough supply capacity should be established to meet the demand. A develop­
ing country could take the latecomer’s advantage to utilize the supply capacity 
in developed countries. However, because of China’s sheer size, the rest of the 
world might not be able to accommodate its huge demand. With constrained sup­
ply capacity but significantly greater demand, the international price of pollution 
control facilities could rise sharply, and this would discourage their utilization and 
slow the pollution mitigation process. Rapid pollution mitigation in China relies 
greatly on the rapid establishment of a domestic industry.
SO2 mitigation achieved rapid progress over the past two decades from low 
starting positions. On the supply side, in the late 1990s, China had few domestic 
firms and barely any commercialized technologies. The Chinese markets were 
dominated by foreign firms and foreign technologies. After a decade, a large num­
ber of firms entered the market to meet the newly emerged huge demand for SO2 
scrubbers to even drive down prices substantially.
As an illustration of the differences between goal-­
centered and rule-­
based gov­
ernance, the progressive paths in China and the United States have been dra­
matically different in reaching the wide deployment of SO2 scrubbers in coal-­
fired 
power plants and their normal operation of high SO2 removal rates (Figure 7.1). 
From the very beginning, the normal operation of SO2 scrubbers in the United 
States with rule-­
based governance has been achieved while the progress went 
mainly through the deployment dimension. In contrast, China deployed SO2 
scrubbers with poor operation in the early stage and then proceeded simultane­
ously in the dimensions of deployment and operation until the technical limits of 


150  Environmental technology and industry
SO2 removal rates were roughly reached. Accordingly, the requirements on the 
quality of SO2 scrubbers were initially low in the Chinese market and became 
increasingly higher only later, while in the U.S. market, quality was important 
from the beginning.
In China, under goal-­
centered governance, at the early stage of deployment with 
few SO2 scrubbers and incapable policy implementation, more SO2 mitigation 
would be achieved if the focus were on further deployment rather than on opera­
tional improvement. With more and more SO2 scrubbers in place, any improve­
ment in the operation of the growing stock would lead to a greater reduction in 
SO2 emissions. For achieving their SO2 mitigation goals, the rational choice of 
the Chinese central and local governments led to a path in which initial progress 
was made mainly in deploying more SO2 scrubbers, and it was only afterward that 
their level of operation caught up.
Implementing policies on the deployment and normal operation of SO2 scrub­
bers require different amounts of resources for compliance monitoring. On one 
hand, the compliance monitoring on the physical existence of SO2 scrubbers is 
straightforward and the huge sizes – for example, an absorbing tower is generally 
several meters in diameter and tens of meters high – make them easily visible. The 
one-­
by-­
one inspection indicates that the corresponding compliance monitoring 
0
100
200
300
400
500
600
0%
20%
40%
60%
80%
100%
)
W
M
0
0
0
,
1
(
t
n
e
m
y
o
l
p
e
d
e
v
i
t
a
l
u
m
u
C
SO2 removal rate
China 
(2006–2008, 2010)
United States
(1973–2010)
Figure 7.1  
The progressive paths on the deployment and operation of SO2 scrubbers in 
China and the United States
Source: Lefohn et al. (1999); Xu (2011b); Ministry of Environmental Protection (2011b, 2008–2012); 
EIA (1986–2006, 2007–2011); Xu (2013).


Environmental technology and industry  151
costs for each SO2 scrubber do not greatly differ, regardless of how many have 
been deployed. On the other hand, the compliance monitoring on the installation 
is just a onetime event, but when in operation they demand significantly more 
resources on a day-­
by-­
day basis. A well-­
functioning environmental compliance 
monitoring system has significant initial costs of establishment. A  significant 
proportion of additional costs for monitoring one more SO2 scrubber are largely 
borne by the polluting firms because they are responsible for installing their own 
monitoring equipment. For policy enforcers, the compliance monitoring costs 
have a great economy of scale and they increase relatively modestly with wider 
deployment of SO2 scrubbers.
The political resistance against the deployment and against the normal opera­
tion of SO2 scrubbers also differs. The normal operation and maintenance (O&M) 
costs are significantly higher than the annualized capital costs, especially for SO2 
scrubbers with compromised quality (Xu, 2011b). Data on the capital costs of SO2 
scrubbers were retrieved from two sources to report a dramatic reduction together 
with an expanding domestic market of SO2 scrubbers (Figure 7.2). From February 
to August 2006, China’s Association of Environmental Protection Industries sur­
veyed SO2 scrubber projects in operation or under construction at the end of 2005 
(Xu et al., 2006). One hundred thirteen projects (223 coal-­
fired power units) with 
0
20
40
60
80
100
120
140
160
180
200
0
20
40
60
80
100
120
140
)
W
k
/
$
S
U
(
s
t
s
o
c
l
a
t
i
p
a
c
t
i
n
U
Annual installation of SO2 scrubbers (1,000 MW)
United States, 2000–2009
China, 2000–2008
Figure 7.2  
Annual average unit capital costs of SO2 scrubbers in China and the United States
Source: Xu et al. (2006); EIA (2012–2013); Ministry of Environmental Protection (2008–2012); Xu 
(2013).
Note: China’s average unit capital costs refer to limestone-gypsum wet scrubbers. Annual average 
exchange rates were used for currency conversion. Data from 2000 to the peak year of deployment 
are shown.


152  Environmental technology and industry
a total capacity of 83,850 MW applied limestone-­
gypsum wet scrubber technol­
ogy and had cost information available. Data on projects using other technologies 
are much less continuous to provide longitudinal insights. They had already been 
or were expected to be in operation over the period from 2002 to 2008. Their 
expected time in operation could partly reflect when the contracts were signed and 
accordingly the then market situation. Furthermore, the author’s interviews pro­
vided an independent source to cross-­
check the survey data and to shed light on 
their more recent changes. Data for the United States came from the U.S. Energy 
Information Administration (EIA; 2012–2013).
Quality has a great impact on the capital costs of SO2 scrubbers. For example, 
SO2 scrubbers in Hong Kong’s two coal-­
fired power plants were contracted with 
firms from mainland China, and the unit capital costs were three to four times 
those of similar projects in mainland China, although still at about half of the 
comparable costs in the United States. Hong Kong’s SO2 scrubbers require high-­
quality equipment, engineering and construction and enough redundancy, and 
they take about twice the amount of time from contract to completion. Beyond 
higher labor costs, the higher price in Hong Kong above “The China Price” could 
be mainly explained as a quality premium.
Considering the reduction of capital costs in the Chinese market (Figure 7.2), the 
investment for one more SO2 scrubber would decrease to indicate that the politi­
cal resistance dwindles when many SO2 scrubbers had been deployed. The O&M 
costs for each SO2 scrubber varied less along the deployment dimension because 
of the necessary consumption of electricity, limestone, and water (Table  6.1). 
More SO2 scrubbers led to greater overall O&M costs, and this increased the over­
all political resistance. However, installing SO2 scrubbers without normal opera­
tion wasted financial resources, and it conflicted with environmental policies. The 
associated political pressure for each SO2 scrubber from the civil society, despite 
its underdeveloped status in China, and from within the government increased 
when more SO2 scrubbers were deployed to make the problem more visible. The 
overall net political resistance against the normal operation of existing SO2 scrub­
bers could increase at the very early stage of deployment and then shrink when 
more SO2 scrubbers are in place.
Given China’s then poor record of implementing environmental policies, the 
evolving quality requirements contributed to goal attainment with a rapid path 
that could be theoretically understood. The Chinese government can make a cer­
tain amount of effort to work for pollution mitigation with two choices, either to 
deploy more pollution control facilities or to enhance the operational performance 
of the existing stock. The goal is to maximize the impacts of efforts on pollution 
mitigation at every step. After a certain amount of pollution control facilities have 
been deployed, the net political resistance against the deployment of one more 
facility and against the enhancement of operational performance by 1% could be 
roughly taken as unchanged with the level of deployment. Accordingly, a given 
amount of effort could either raise the deployment rate by α (in the two cases of 
SO2 scrubbers, the unit is megawatts, MW) or the performance of existing facili­
ties by β% (in the SO2 scrubber case, the unit is percentage points of SO2 removal 


Environmental technology and industry 153
rates). The initially deployed facilities have a total capacity of A, and the initial 
performance is B%. Then the initial pollution mitigation effect of the facilities is 
roughly proportional to A × B%. The performance has a technical upper limit, 
B*%.
The option of devoting the efforts to the deployment could raise the pollution 
mitigation effect to (A + α) × B%, and the other option to work on the opera-
tion would have an effect of A × (B% + β%). If there is no constraint, a rational 
decision maker to maximize the impact of his or her efforts will choose the first 
A + α
β
B%
%
+
option when (A + α) × B% > A × (B% + β%), or when 
>
 or 
A
B%
α
β%
α
β%
>
. The second option will be taken when 
<
, and the two options 
A
B%
A
B%
α
β%
are no different when 
=
. With the progress on the deployment and opera-
A
B%
tion, the choice could change. This is what goal-centered 
 
governance would indi-
cate. If adding one constraint that the choice should prioritize policy enforcement, 
the progress should be first made to improve the operation. Only when B% has 
reached B*%, more facilities are allowed to be deployed. This could illustrate 
rule-based governance.
One more constraint could be added to describe the situation on the supply side. 
As examined below with more details, the goal-centered 
 
governance strategy low-
ers technological barriers of market entry to facilitate the rapid establishment of 
a large-enough supply capacity
 
, while the rule-based governance 
 
strategy would 
correspond to higher market- 
entry barriers and discounted supply capacity in the 
Chinese context. To simplify the model, the supply capacity under rule-based 
 
governance is η% less than that in goal-centered
 
 governance, and thus, the same 
amount of efforts could only raise the deployment rate by α
η
×
−
(
%
1
).
The SO2 scrubber case is simulated here to exemplify the usefulness of this 
very simple model. Here are the assumptions of the earlier parameters: (1) A0: the 
initial capacity of SO2 scrubbers, 7,000 MW, equivalent to the level in 2000 (Min-
istry of Environmental Protection, 2008–2012); (2) B0%: the initial SO2 removal 
rate in coal- 
fired power plants with SO2 scrubbers, 31.3%, equivalent to the level 
in Jiangsu Province in 2006 (Xu, 2011b); (3) B*%: 79%, the highest SO2 removal 
α
4,500 MW
rate China achieved in 2010 (Figure 6.1); (4) 
: 
, or the required 
β%
1%
effort from decision-makers 
 
was the same to deploy 4,500 MW of SO2 scrubbers 
and to increase the SO2 removal rate of the existing stock by 1%. The number is 
assumed to fit China’s actual data; (5) η%: 50%, assumed to indicate the impacts of 
higher market-entry barriers in the rule-
 
of-
 
 
law strategy. As illustrated in Figure 7.3, 
the projection with the goal-centered 
 
governance strategy fits well into China’s SO2 
mitigation path for coal-fired 
 
power plants with SO2 scrubbers. If considering no 
constraint from the supply side, rule-based 
 
governance mainly would differ from 
goal- 
centered governance at the early stage of progress. However, if considering 
the potential supply constraints due to higher market-entry barriers, pollution miti
 
-
gation under rule-based governance would proceed at a much slower pace.
 
­


154  Environmental technology and industry
Figure 7.3  
Model projection of the SO2 mitigation path in China’s coal-fired power plants: 
(a) deployment and operation of SO2 scrubbers under goal-centered govern­
ance (the dots refer to actual data); (b) avoided SO2 emissions under goal-
centered and rule-based governance
Source: Xu (2013).
0
100
200
300
400
500
600
0%
20%
40%
60%
80%
100%
SO2
)
W
M
 
0
0
0
,
1
(
 
y
t
i
c
a
p
a
c
 
r
e
b
b
u
r
c
s
SO2 removal rate
2006
2007
2008
2010
(a)
Avoided SO2 emissions
Cumulative efforts
Actual path
Goal-centered governance
Rule-based governance without supply constraint
Rule-based governance with supply constraint
2006
2007
2008
2010
(b)


Environmental technology and industry  155
2  
Technology licensing under goal-­
centered SO2 
mitigation path
The international technology market provided opportunities for China’s domestic 
firms to license foreign technologies and to quickly ramp up their technological 
capabilities, although at a cost. Functioning markets for transferring technologies 
to developing countries not only are important for their economic development 
and upgrading along the value chain but also have critical implications for the 
environment. Due to China’s huge and steadily growing emissions, how fast and 
effective environmentally friendly technologies were adopted was a key determi­
nant for its environmental cleanup. Technology transfer from developed to devel­
oping countries has long been recognized as a key measure in addressing CO2 
mitigation (United Nations, 1992). One important method of technology transfer 
is through technology licensing. With available markets for technologies, a tech­
nology owner could choose between licensing its product or directly investing 
in the client country, and a firm that needs technology could either license in or 
innovate indigenously (Arora et al., 2001a; Teece, 1988; Arora et al., 2001b). In 
international negotiation on transferring low-­
carbon technologies from developed 
to developing countries, developed countries generally argue for market-­
based 
solutions and adequate protection of intellectual property rights (IPR), while 
developing countries often demand nonmarket solutions at lower than market 
rates (Ockwell et al., 2010). The differing positions become an obstacle to the 
agreement of new and effective climate treaties (Ockwell et al., 2010).
Despite unfavorable conditions, the global market for technology has been sig­
nificant, amounting to about US$35 to US$50 billion in the mid-­
1990s (Arora 
et al., 2001b) and roughly US$100 billion in 2002 (Arora and Gambardella, 2010). 
However, only a small portion – less than one third for the United States – of 
technological transactions were between unaffiliated organizations and thus true 
market transactions (Arora and Gambardella, 2010; Saggi, 2002). Most cross-­
border technology licensing happens among developed countries and that from 
developed to developing countries is much rarer (Arora and Gambardella, 2010). 
Product markets in most developing countries are not large enough to attract many 
potential technology licensors. Developing countries generally lag behind devel­
oped countries in human and technological capacities that enable them to effec­
tively absorb licensed foreign technologies and exploit their full value (Metz et al., 
2000). Additionally, effective IPR protection could help address the problems of 
unauthorized use of intellectual property (Gans and Stern, 2010), but developing 
countries often do not have well-­
developed systems of IPR protection and thus are 
placed in relatively disadvantageous positions in creating an attractive market for 
technology (Strokova, 2010). However, large developing countries like China are 
able to access foreign low-­
carbon technologies, although not those at the cutting 
edge (Ockwell et al., 2010; Lewis, 2007). China’s rapid development of many 
industries had roots partly in the importation of foreign technologies, including, 
for example, wind turbines (Lewis, 2007), large hydroelectric turbines (Liang, 
2001) and high-­
speed railways (Chan and Aldhaban, 2009).


156  Environmental technology and industry
An especially prominent case was that of SO2 scrubbers. SO2 scrubber tech­
nologies have been commercially deployed since the mid-­
1970s, mainly in devel­
oped countries. Up until 1998 (expressed in terms of generating capacity of power 
stations thus equipped), the pace of deployment was about 10 GW per year in 
the world and 4 GW per year in the United States (Srivastava et al., 2001). Many 
international firms had established their technological and engineering reputations 
in this field. China began to significantly deploy SO2 scrubbers about three dec­
ades later than developed countries, with a deployment rate of over 100 GW per 
year in the 11th Five-­
Year Plan (Chapter 5). Because of their high SO2 removal 
efficiencies – generally over 90% with wet-­
type technologies – SO2 scrubbers 
became the most vital technology in achieving China’s goal of a 10% reduction 
in SO2 emissions in the 11th Five-­
Year Plan (2006–2010; Xu, 2011b, 2011c). 
Among the more than 500 GW of SO2 scrubbers in China at the end of 2010, more 
than 90% were installed by Chinese firms using licensed foreign technologies 
(Ministry of Environmental Protection, 2011a). Major Chinese firms universally 
licensed foreign technologies and relied heavily on them. Conversely, fewer than 
5% were installed by foreign firms or under joint ventures (Ministry of Environ­
mental Protection, 2011a). Domestic firms dominated the market, in spite of their 
initial lack of proven technologies and experience.
The goal-­
centered SO2 mitigation path created three characteristics of Chi­
na’s SO2 scrubber demand in the early stage. The difficult SO2 mitigation goals 
together with China’s colossal size required more than 100 GW SO2 scrubbers 
annually, which was multiple times as big as the world together had experienced 
before (Figure 5.12). Their initial poor operation significantly relaxed actual qual­
ity requirements (Figure 7.1). The initial one-­
sided emphasis on the deployment 
of SO2 scrubbers indicated that the huge demand for SO2 scrubbers would be cre­
ated swiftly from a low level in the 10th Five-­
Year Plan, which led to stringent 
time constraints for SO2 scrubber firms (Figure 5.12). They played key roles in 
shaping the strategies of domestic technology licensees and foreign technology 
licensors for tapping into the market.
2.1  
The strategy of domestic technology licensees
China’s domestic firms as technology licensees could fall into the three follow­
ing categories: state-­
owned, university-­
established and nonstate. “State-­
owned” 
firms refer to those controlled by state-­
owned power corporations, which could 
have faced less fierce competition to win SO2 scrubber projects because of their 
special “internal” relationship. Indigenous SO2 scrubber technologies had been 
developed by a few research institutes and universities to directly transfer their 
human and technological capabilities to state-­
owned and university-­
established 
firms. Nonstate firms could behave differently due to their relative lack of such 
initial capabilities. In addition, although most of China’s major firms relied heav­
ily on licensed technologies, some concentrated on applying their own. China 
had five large state-­
owned power corporations at the national level in the late 
2010s, four having major SO2 scrubber firms, and two were selected for interview. 


Environmental technology and industry  157
In the available SO2 scrubbers at the end of 2011 with unit scales not smaller 
than 100 MW, the two firms had market shares of 11.9% and 3.3%, respectively. 
Another smaller firm owned by one of the five power corporations was also vis­
ited, and its market share was 0.4%. The special relationship with their parent 
corporations put them in relatively advantageous positions in market competition. 
Eight firms that had no association with power corporations were interviewed. 
Their market shares ranged from 0.7% to 6.0%, being 22.8% in total. In addition, 
two foreign firms and their Chinese representative offices as technology licensors 
were also interviewed to provide an external perspective.
Domestic firms’ decisions to license in SO2 scrubber technologies were heavily 
influenced by the three demand characteristics under the goal-­
centered SO2 miti­
gation path. First, the sheer size of China’s demand for SO2 scrubbers challenged 
the supply capacity. One concern was whether China had enough engineers. This 
condition was met partly through rapidly training many more university students 
(Figure 7.4). In 2000, 496,000 undergraduate students graduated from full-­
time 
four-­
year undergraduate programs, including 213,000 in engineering. In 2010, the 
numbers had grown to 2,591,000 and 813,000, respectively. In 2018, the numbers 
further climbed to 3,868,000 and 1,269,000, respectively. In 2018, about the same 
number of undergraduate students (3,665,000) graduated from other full-­
time pro­
grams with shorter study periods of two or three years. The age group, 20 to 24, 
comprised 5.95% of China’s population in 2018, or 16.6 million for each yearly 
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
)
e
l
p
o
e
p
 
0
0
0
,
1
(
 
s
e
t
a
u
d
a
r
G
Year
Science
Engineering
Agriculture
Medicine
Others
Figure 7.4  
Yearly university graduates in China from four-year undergraduate programs 
by subjects
Source: Ministry of Education (1999–2019).


158  Environmental technology and industry
age (National Bureau of Statistics, 1996–2019). Accordingly, in 2018, about half 
of China’s newly available labor force had a received formal university educa­
tion. Other part-­
time or Internet-­
based undergraduate programs trained another 
4.1 million graduates in that year. These enhanced human resources provide a 
crucial foundation for China’s rapid deployment of pollution-­
removal industrial 
facilities.
The huge market also helps diminish one concern that licensors might not trans­
fer technologies completely after receiving payments (Arora et al., 2001b). In the 
case of SO2 scrubber technology, royalties dominated the revenue stream in tech­
nology licensing and effectively deterred such a moral hazard. By way of exam­
ple, an American firm charged one licensee US$652,118 as the up-­
front lump-­
sum 
fee (Table 7.1): interviews discovered that a license’s approximate royalty rate 
should be 2% of SO2 scrubber contract values. Between 2004 and 2010, the firm’s 
income from royalties was nearly 40 times as much as the up-­
front lump-­
sum fee 
(the licensee completed 34,900-­
MW wet SO2 scrubbers in that period; Ministry 
of Environmental Protection, 2011a), and the national average contract value was 
about US$35/kW (Xu et al., 2006)). From another perspective, as demonstrated 
in the case of a Japanese licensor, a licensee’s loss was limited to approximately 
the up-­
front lump-­
sum fee when the technology transfer was not satisfactory. In 
addition, if a licensor gained a bad reputation, this could limit its future business 
opportunities in the huge and rapidly growing Chinese market.
Second, the quality requirements for SO2 scrubbers were initially low. The 
deployment of SO2 scrubbers took off around 2002, but the normal operation was 
improved significantly only in about 2007 (Xu, 2011b; Xu et al., 2009). In the five 
gap years, many managers of installed SO2 scrubbers did not plan to operate them 
normally and cared very little about the quality, while quality was closely associated 
with the technological advancement of a supply firm. In addition, China’s reform 
in the power sector in 2002 created multiple independent power corporations to 
Table 7.1 Up-front 
 
lump-sum fees of SO2 scrubber technology licenses (the Chinese 
licensees here are all listed on stock markets and the data are from their annual 
reports)
Chinese licensee
Country origin Lump-sum fee*
Year
Technology type
of the foreign 
licensor
Wuhan Kaidi
Germany
US277,3041998DrytypeFujianLongjingGermanyUS3,989,234
2001
Wet type
Circulation fluidized bed
Wuhan Kaidi
United States
US652,1182002WettypeZhejiangFeidaUnitedStatesUS1,250,000
2002
Wet type
Jiulong Electric
Japan
US1,126,5632002WettypeJiulongElectricAustriaUS1,423,765
2004
Wet type
Insigma Technology
France
US$1,200,000
2004
Wet type
* Exchange rates on December 31, 2010 were used: 1 US$ = 6.62 RMB = 0.75 euro.


Environmental technology and industry  159
encourage competition – this was even though all of these were state-­
owned. The 
rapid construction of new power plants strained their available financial resources 
to create strong incentives to minimize capital investment for each new project, 
while the poor quality of SO2 scrubbers could substantially reduce capital costs. 
Furthermore, the low requirement for quality was strengthened by the largely sepa­
rate decisions of capital investment and daily operation and by the different incen­
tives of respective decision-­
makers. Managers of coal-­
fired power plants should 
have an incentive to install high-­
quality SO2 scrubbers while capital investment 
was within the authority of the upper levels of management in power corporations. 
The low requirement for quality and technological advancement substantially low­
ered the technological market-­
entry barrier not just for the SO2 scrubber firms but 
also along the entire supply chain. In contrast, the quality requirement and techno­
logical market-­
entry barrier in the U.S. market were much higher.
China’s regulators also paid attention to the quality requirements, especially with 
the knowledge of domestic firms’ initially unsatisfactory technological statuses. 
Technologies could come from international transfer or in-­
house innovation. Vari­
ous factors could affect the choice of a country or a firm between these two technol­
ogy strategies. China used to focus almost entirely on in-­
house innovation under 
the rule of Chairman Mao when China segregated itself from the world. The “Not 
Invented Here” syndrome – that internally developed technologies are preferred – 
was found to be a barrier to technology licensing (Arora and Gambardella, 2010), but 
it does not seem to be deeply rooted in China in the economic reform era. Secondary 
innovation based on imported technologies, coupled with original and integrated 
innovation, had been established as three cornerstones of China’s indigenous inno­
vation strategy (State Council, 2006). With regard to the installation of SO2 scrub­
bers, China stipulated in tendering documents that established technologies were 
required. As late as 2005, bidders were clearly asked to specify a foreign technology 
provider that had installed SO2 scrubbers of the same or greater scale (Guizhou 
Qiandong Power Station, 2005). Interviews also confirmed the general requirement 
for foreign, commercialized technologies in the early years when almost no Chinese 
firms had any proven experience. This requirement was relaxed only in later years 
after many firms in the market had completed enough projects.
Third, time was a serious constraint. In the late 1990s and early 2000s, few 
domestic firms were capable of designing SO2 scrubbers. The sudden appearance 
of a huge market led to the creation of many new firms and the reorientation of 
existing ones from other industries. Because few firms had any prior experience 
and the market was large enough to accommodate many, most – except those 
owned by coal-­
fired power corporations – were placed on a more or less equal 
footing. Firms would achieve distinction if they could establish engineering and 
management teams and develop their technological capability faster than others. 
Another time constraint was the short period from the issue of tendering docu­
ments to completion of the bidding process; this typically lasted only one to four 
weeks. Additionally, the design process could not take more than a few months 
if the construction was to begin on schedule. Successful firms had to respond 
quickly and provide acceptable quality.


160  Environmental technology and industry
These time constraints helped push domestic firms toward technology licens­
ing, due to their weak technological foundations. When demand for SO2 scrub­
bers started to surge, domestic technologies were generally not able to satisfy 
the time constraints because of their immaturity. Domestic research and devel­
opment generated “naked” technologies, to quote the word of one interviewee. 
Demonstration projects on a commercial scale should be followed by multiple 
projects to make the technology mature and ready for wide commercial deploy­
ment. The commercialization of these “naked” technologies would require at least 
a few years plus significant financial resources and the willingness of coal-­
fired 
power plants to take risks by trying them. The expected short-­
term peak in Chi­
na’s scrubber market diminished the potential return on investment in indigenous 
technology. The easy prospect of licensing foreign technologies also reduced the 
incentive to take risks with indigenous innovations. All the major Chinese firms 
in the market licensed foreign technologies in order to acquire and substantiate 
their technological capabilities. No clear difference could be found among state-­
owned, university-­
established and nonstate firms. Even the nonstate firm that 
mainly applied its own technology had to initially license from abroad.
As tacit knowledge cannot be so easily transferred as codified knowledge, kno­
whow played a positive role in establishing a sound market for technology. The 
contractual acquisition of know-­
how presents more problems than licensing pat­
ents (Arora et al., 2001b). However, in a developing country like China with poor 
IPR protection, the licensing of patents might be unnecessary in the absence of 
know-­
how as the knowledge contained in the patents have already entered the 
public domain. Chinese firms had generally chosen to legally license, rather than 
to illegally acquire, SO2 scrubber technologies. Legal licensing secured a com­
plete package including systematic training, technical documentation and trade 
secrets in a relatively short timescale, without exposing the licensees to legal 
disputes. One alternative option was to recruit experts from foreign firms, but 
the legal risks were not insignificant and the received technologies may not be 
complete because it would be difficult to recruit an entire team. It would also take 
much longer for the acquiring firms to comprehend a technology by this means 
than they would through technology licensing. The associated costs would not be 
low either, because foreign experts generally had to be paid considerably more 
than standard Chinese salaries. Furthermore, illegal acquisition did not provide a 
technological guarantee from a trusted provider, while this guarantee was stipu­
lated by coal-­
fired power plants in their tendering documents.
China’s domestic firms could quickly absorb licensed technologies to meet the 
time constraints. From as early as the 1970s, China had, through its own research 
and development on SO2 scrubbers, built up vital capabilities to establish domes­
tic firms and assimilate imported technology (Shu, 2003). From the mid-­
1970s 
to the mid-­
1980s, China appraised several technologies, although on scales that 
were at least one or two orders of magnitude smaller than any commercial pro­
ject. For example, a 300-­
MW unit corresponds to a flue gas flow rate of about 
1,000,000 Nm3/hour (cubic meter at standard temperature and pressure per hour), 
while the largest Chinese experiment at the time had a flow rate of 70,000 Nm3/hour 


Environmental technology and industry  161
0.00%
0.25%
0.50%
0.75%
1.00%
1.25%
1.50%
1.75%
2.00%
2.25%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
1995
2000
2005
2010
2015
Ratio
l
l
u
f
 
0
0
0
1
(
 
l
e
n
n
o
s
r
e
p
 
D
&
R
-
 
)
e
l
p
o
e
p
 
t
n
e
l
a
v
i
u
q
e
 
e
m
i
t
)
B
M
R
 
8
1
0
2
 
n
o
i
l
l
i
b
(
 
e
r
u
t
i
d
n
e
p
x
e
 
d
n
a
Year
R&D Personnel (Full-time Equivalent)
R&D expenditure
Technology market transaction value
R&D expenditure vs. GDP (%)
Technology market transaction value vs. GDP (%)
Figure 7.5  
R&D personnel, expenditure and market value (in 2018 RMB) in China
Source: National Bureau of Statistics (1996–2019).
(Shu, 2003). From the mid-­
1980s to 2000, foreign technologies were demonstrated 
on a commercial scale (Gu, 2004; Shu, 2003). In 2000, having resulted in a consid­
erable fund of domestic human and technological capability, foreign technologies 
were officially recognized as the basis for further development of SO2 scrubber 
technologies in China (National Economic and Trade Commission, 2000). China’s 
absorptive capacities were effectively distributed to all major firms including non­
state ones through a free labor market of engineers and managers.
Recognizing the constraints of technology licensing such as on expansion 
beyond China, in the past two decades, China has put a much heavier empha­
sis on research and development (R&D). In 2000, China had 922,000 full-­
time 
equivalent personnel on R&D and this number rapidly grew by 375% to 4.4 mil­
lion in 2018. R&D expenditures were raised from 0.60% of gross domestic 
product (GDP) in 1995 to 2.19% in 2018 (Figure 7.5). A much more vibrant mar­
ket for technology emerged and the transaction value increased from 0.46% of 
GDP in 1995 to 1.97% in 2018 (Figure 7.5). Together with the rapid growth of 
China’s GDP, the R&D expenditures and technology market transaction values 
had become 1084% and 1365% greater in 2018 from the levels in 2000 in real 
terms (Figure 7.5). This R&D boom strengthened China’s capacity to absorb for­
eign technologies and innovate domestic intellectual property. In the category of 
environmental technology, China’s residents and nonresidents were granted 103 
and 69 patents, respectively, in 2000 in China’s patent filing office, which were 
about 10% of those in the United States. They grew to 7,459 and 881 patents, 


162  Environmental technology and industry
respectively, in 2018, while the figures in the United States were correspondingly 
1,258 and 1,369 patents (Figure 7.6).
2.2  
The strategy of foreign technology licensors
The strategy of potential foreign technology licensors was also shaped by the pre­
viously mentioned three characteristics of China’s SO2 scrubber demand under a 
goal-­
centered SO2 mitigation path. First, the huge demand for SO2 scrubbers created 
profitable business opportunities. Their decision of technology licensing involves 
the revenue effect (i.e., payments received from licensing) and rent-­
dissipation effect 
(i.e., revenue loss due to a new or strengthened competitor in the product market; 
Arora and Fosfuri, 2003). A stronger revenue effect promotes the decision to license, 
while a stronger rent-­
dissipation effect discourages licensing. For major foreign 
firms that held intellectual property of SO2 scrubber technologies, the option to do 
nothing was rarely attractive because of the temptation of the huge emergent Chi­
nese market. The revenue effect was indeed significant. Technology licensing only 
required a small office in China to monitor licensees and to “service” the partnership. 
For example, each of the two interviewed American firms had an office in Beijing 
with about five staff members, whereas their licensees were in charge of contracts 
worth several hundred million dollars annually. The initial cost in transferring tech­
nologies was covered by up-­
front lump-­
sum fees paid by licensees (Table 7.1). The 
0
3,000
6,000
9,000
12,000
15,000
18,000
21,000
1980
1985
1990
1995
2000
2005
2010
2015
)
y
g
o
l
o
n
h
c
e
t
 
l
a
t
n
e
m
n
o
r
i
v
n
e
(
 
e
c
i
f
f
o
 
g
n
i
l
i
f
 
y
b
 
s
t
n
a
r
g
 
t
n
e
t
a
P
Year
China: Resident
China: Nonresident
U.S.: Resident
U.S.: Nonresident
Others: Resident
Others: Nonresident
Figure 7.6  
Patents on environmental technology by filing office in the world
Source: WIPO (2019).


Environmental technology and industry  163
commercial success of licensees would result in considerable royalties to the licensor 
if the contracts were honored. After the know-­
how and trade secrets were transferred, 
the intellectual property rights were at risk of misuse or infringement, possibly with 
the royalties not being fully paid. Despite this, most foreign firms decided to take this 
risk in order to avoid the much greater risk inherent in direct investment.
After technologies are transferred, one primary concern of technology licensors 
arose on whether licensees paid royalties honestly. Both licensors and licensees 
reported in interviews that major Chinese firms were paying royalties regularly. Also, 
several expiring licenses had been renewed, indicating a good record of royalty pay­
ments. As a preventative measure, design software was encrypted and only specially 
prepared computers could install it with annual reregistration. Several interviewees 
in the Chinese firms said that, after a few years, they had figured out what was inside 
the black box but still chose to pay royalties. It was not very difficult to keep track of 
licensees. The huge size of SO2 scrubbers often made local news and the Ministry of 
Environmental Protection annually published details of every SO2 scrubber and its 
contractor (Ministry of Environmental Protection, 2011a). Besides, a good partner­
ship with licensors suited the long-­
term interests of licensees. Technological sophis­
tication had increased step by step in the Chinese SO2 scrubber market as reflected 
in the unit scales: the 300-­
MW scale was dominant before 2005, but after 2006, the 
600-­
MW scale became crucial and then the 1,000-­
MW scale or greater (Ministry 
of Environmental Protection, 2014). Every significant increase in scale indicated a 
new technical advance. Accordingly, the licensing of scrubber technologies was a 
continuous operation and not a one-­
off process. Good partnerships, strengthened 
by honest royalty payments, could also help licensees expand into new markets 
through future technology licensing. In addition, a partnership may generate busi­
ness opportunities for both sides. For example, when a large coal-­
fired power plant 
in Hong Kong decided to install SO2 scrubbers, it first approached several interna­
tional firms, including one from the United States. But the American firm was fully 
committed in the domestic market and was not willing to take the financial risk of 
an Engineering, Procurement, and Construction (EPC) project in Hong Kong. Its 
Chinese licensee was introduced and finally won the contract.
Royalty rates may decrease over time to reduce the costs of honoring licensing 
contracts. For example, one license divided the ten-­
year contract period into three 
phases with declining royalty rates. In several other cases, the royalty rate was rene­
gotiated when competition in the market became much too fierce to significantly 
shrink the profit margin. Excessively high royalty rates could damage licensees’ 
competitiveness. The final result might be a reduced income from royalties and 
an increased risk of no payment being made at all. The renegotiation strengthened 
the partnerships between licensors and licensees and thus worked for the inter­
ests of both sides. In one licensing contract signed in 1998, the level of royalties 
was originally associated with the volume of flue gases. Because China’s capital 
costs of installing SO2 scrubbers had dropped substantially since then (Figure 7.2), 
the royalty rate would increase significantly as a percentage of the contract value. 
Renegotiation took place to lower the royalty rate. The partnership remained strong 
with both the licensor and the licensee maintaining market success.


164  Environmental technology and industry
Lawsuits, particularly those resolved outside China, were also a deterrent to 
potential infringement, which maintained the strong revenue effect. For example, 
Insigma Technology is a Chinese firm listed on the Shanghai Stock Exchange, and 
it releases information regularly. It signed a technology licensing contract with 
a French firm in December 2004 (Table 7.1). However, in April 2006, Insigma 
declared that it would cancel the contract and thereafter stop using the licensed 
technology. Royalties were paid for six projects in 2005 and 2006 with a total 
capacity of 7,450 MW (Sina Finance, 2010). The firm later signed a new contract 
with an Italian firm in September 2006, which was for one year and was to be 
automatically renewed if no objections were received from either side. The fee 
for royalties was a fixed sum of €20,000 (US$26,600) for every project regardless 
of the contract value (Sina Finance, 2010). The French firm later sued Insigma in 
Singapore (where disputes should be resolved according to the licensing contract). 
The court made a decision in February 2010 and Insigma was ordered to pay com­
pensation of US$2,085,737 for the loss of royalties in 2005 and US$24,566,684 
for the loss afterward (Sina Finance, 2010). The lawsuit may have helped deter 
other significant licensees from not honoring their licensing contracts.
Second, low-­
quality requirements and correspondingly low technological 
market-­
entry barriers led to active market entry of new firms to contain the rent-­
dissipation effect for technology licensors. If the downstream operations of a firm 
are small or the downstream market is in fierce competition, the rent-­
dissipation 
effect will be limited and technology licensing becomes more likely (Arora and 
Gambardella, 2010). Indeed, the Chinese downstream SO2 scrubber market was 
newly created and in fierce competition (Ministry of Environmental Protection, 
2011a). In addition, market evolution also demonstrated that the rent-­
dissipation 
effect should be minimal. Foreign firms tended to lag behind domestic ones in 
understanding the market’s real demand, especially in the early period. Among all 
the foreign firms, the examined Japanese firm ought to be the best prepared for 
the Chinese market. It owned more Chinese patents on flue gas desulfurization 
than any other firm (State Intellectual Property Office, 2010) and, between the late 
1980s to 1990s, had won contracts to install China’s first-­
ever commercial wet 
SO2 scrubbers (four units of 360-­
MW capacity; Gu, 2004). However, up to the 
end of 2010, its technology was only applied to a further 3,300 MW, with the final 
project in 2006 (Mitsubishi Heavy Industries, 2011). Interviews in China revealed 
that many foreign firms generally licensed design software together with other 
know-­
how in order to enable their Chinese licensees to compete independently, 
but this Japanese firm was reluctant to hand over design software and wanted to 
participate more actively. Thus, the technology transfer of know-­
how was not 
complete. The decision could have been influenced by the expectedly significant 
rent-­
dissipation effect due to potentially high rents as a result of its favorable 
position in granted patents. However, partly because the relationship made them 
slower in responding to the market and hampered their competitiveness, its Chi­
nese licensees decided instead to do business with other technology licensors. For 
example, according to the annual reports from a firm listed on the Shanghai Stock 
Exchange – Jiulong Electric, the holding firm of Yuanda Environmental Protection 


Environmental technology and industry  165
Engineering – although US$1.1 million was paid to the Japanese firm as the up-­
front lump-­
sum fee, just two years later it decided to sign another licensing con­
tract with a European firm and gave up the Japanese technology (Table 7.1). Even 
with the tight control of technology licensing, the Japanese firm earned little profit 
or rent from the Chinese market, an indication of a small rent-­
dissipation effect. 
The existence of many technology licensors diminished the rent-­
dissipation effect 
because no single licensor had significant market power.
Third, time constraints discouraged direct participation of foreign technology 
licensors in the Chinese market. Two interviewed American firms each had a 
small representative office in Beijing, but their licensing strategies were notably 
different. They reported that the Chinese government put no restrictions on allow­
ing foreign firms to bid for SO2 scrubber projects, but many foreign firms did not 
expect that they would earn significant profits by establishing subsidiaries or joint 
ventures in China. One major American firm expected the Chinese market to peak 
for only a few years before it began shrinking; this expectation proved prescient 
(Figure 5.12). The initial investment of capital and human resources to establish a 
subsidiary in China would therefore only be of temporary benefit. The firm’s past 
experience in other countries suggested that direct investment could not be freely 
withdrawn, and accordingly, it was not justified in this particular Chinese market. 
In addition, the lack of adequate human resources also constrained some foreign 
firms from choosing direct investment, particularly due to the revived U.S. market 
for SO2 scrubbers (U.S. Energy Information Administration, 2011).
2.3  
Why technology market can emerge in China?
Even in developed countries – as Gans and Stern argue – an effective market for 
technology is difficult to establish because it often fails to satisfy the three criteria 
of effective market design as specified by Roth that successful marketplaces must 
be “thick, uncongested and safe” (Gans and Stern, 2010; Roth, 2008). The Roth 
criteria were proposed to fix broken markets or build new ones if they are missing, 
which could be especially useful for environmental protection as market failure 
is often the cause. First, an efficient market requires many potential buyers and 
sellers, or market thickness, to enhance the chances of effective matching. How­
ever, many ideas are not independent but reliant on other complementary ideas and 
assets to achieve their full value, with notable examples in low-­
carbon technolo­
gies (Harvey, 2008). This problem makes the licensing of a single idea less desir­
able. If the ideas belong to different entities, ineffective coordination could limit 
the willingness of potential buyers and sellers to participate in the market. Second, 
the market should overcome Roth’s “congestion” criterion, whereby buyers and 
sellers should be able to negotiate with a number of possible trading partners and 
have sufficient time to make effective selections. In a congested market, competi­
tion is not sufficient and the price does not reach market equilibrium. Because nec­
essary information disclosure for buyers to assess a technology’s value might lead 
to unwanted diffusion, the information is often kept secret between buyers and sell­
ers to constrain open market competition, thus failing the “congestion” criterion. 


166  Environmental technology and industry
Third, market transactions should be “safe”; that is, conducted in good faith and 
with safeguards that allow the expression of real intention and information and 
result in mutual satisfaction. A drawback on this point is that, after licensors have 
disclosed information, licensees might be able to exploit it independently, without 
signing licensing contracts, creating issues over misuse of intellectual property.
The Chinese market for SO2 scrubber technologies satisfied all three Roth crite­
ria. Key contributing factors could include China’s large market size, the maturity 
of available technologies and goal-­
centered governance. First, because the size of 
the Chinese market for SO2 scrubbers as a downstream market for the technolo­
gies is far greater than any other country, major foreign SO2 scrubber firms, as 
potential licensors, could hardly overlook the potential business opportunities. 
The large market and low technological barriers facilitated by technology licens­
ing have created many domestic firms as potential licensees. Multiple sellers from 
the United States, Europe and Japan actively licensed out their technologies (Xu 
et al., 2009, 2006). In addition, in the Chinese market up to 2010, 16 firms – all 
Chinese – had completed at least 10 GW of SO2 scrubbers, all using licensed-
­
in foreign technologies (Xu et al., 2006; Ministry of Environmental Protection, 
2011a). The three types of Chinese firms – state-­
owned, university-­
established 
and nonstate – did not show significantly different behavior in the market for 
technology. Fierce competition drove down costs and diminished expected profit 
from direct investment, but revenue from technology licensing was significant. 
The rent-­
dissipation effect was overwhelmed by the revenue effect of technology 
licensing, which accordingly became a dominant choice of foreign firms. As a 
large country, China has a strong capacity to absorb new technology due to its 
previous R&D, and this capacity was effectively distributed to all three types of 
firms through a free labor market. Licensors and licensees held multiple bilateral 
negotiations simultaneously to help solve the market congestion problem. Fur­
thermore, the safety of technology licensing also benefited from China’s large 
market size. As a result of the large market, there were significant revenues from 
royalties that encouraged licensors to transfer complete packages of technolo­
gies. The market for SO2 scrubbers at every unit scale was substantial and the 
unit scales escalated over time to require continuous technological support from 
licensors. Such dynamism favored long-­
term partnerships between licensors and 
licensees for their mutual benefit and fostered honest royalty payments.
Second, the maturity of SO2 scrubber technologies played a crucial facilitating 
role. After several decades of commercial deployment in developed countries, 
many firms had acquired complete technology packages. Personal and corporate 
expertise, or know-­
how as tacit knowledge, was a vital part of the technology 
package. Acquiring knowhow raised costs and contracting problems, but given 
the inadequate standard of IPR protection in China, technology licensing became 
necessary in order to acquire complete packages of technologies. Many foreign 
firms had become independent technology holders, and a potential licensee only 
needed to negotiate with one licensor for a complete technology package. When 
deciding whether to license out technologies or set up direct subsidiaries in devel­
oping countries or even just do nothing, firms from developed countries needed 


Environmental technology and industry  167
to compare the expected profits of each market option. The dominant business 
reality in the market was technology licensing. For a potential licensee, the tech­
nology could either be developed internally or acquired externally. Favorable con­
ditions created the demand for foreign technologies in the Chinese market.
The Chinese market also met the second Roth criterion on the lack of con­
gestion. The maturity and wide deployment of SO2 scrubbing technologies also 
enabled a fairly accurate estimation of the technology’s value to facilitate mar­
ket transactions. Interviews revealed that, although the negotiation of technology 
licensing was generally bilateral, without disclosing information to third parties, 
licensors and licensees often negotiated with several entities on the other side at 
the same time for most suitable licensing contracts. IPR protection is recognized 
as a key means to ensure market safety and satisfy the third Roth criterion (Gans 
and Stern, 2010). As examined earlier, know-­
how and credible threat of lawsuits 
ensured the general satisfaction of this criterion. The disclosure of the necessary 
information for value assessment in negotiations caused fewer problems because 
knowhow could not be easily acquired.
The existence of many potential licensees enabled licensors to design their 
strategies to maximize profit. At least three clear strategies emerged among three 
licensors. A major American firm licensed to only two Chinese firms and built up 
long-­
term partnerships through full technical support. One license was restricted 
to the licensee’s home province for a certain period and the other covered the 
whole of mainland China. The licensees had a near monopoly to use the specific 
technology in their assigned market territories. Another significant American firm 
had about eight licensees in China; the strategy was to increase the market share 
of its technology as well as its royalties, but the licensees were still selected so as 
to prevent unqualified ones from ruining the technology’s reputation. In addition, 
as mentioned earlier, a Japanese firm licensed its technology to a few Chinese 
firms but, unlike the two American firms, refused to transfer design software. The 
two American firms had their technologies widely applied but the Japanese tech­
nology was abandoned without much deployment. From the perspective of the 
level of royalties, the two American strategies were clear winners.
An effective market for cutting-­
edge technologies is understandably more 
difficult to establish. It is probable that not many organizations have acquired 
intellectual property as potential licensors. The value of a particular cutting-­
edge 
technology is harder to assess and the accumulation of know-­
how may still be in 
progress with a consequently high price of the final product which will limit its 
deployment. These unfavorable conditions discourage the emergence of potential 
licensees. Information disclosure to facilitate licensing will also raise more con­
cerns on the part of technology owners. As a result, the Roth criteria of effective 
market design will be harder to meet for cutting-­
edge than for mature technologies.
Third, goal-­
centered governance resulted in a path of SO2 mitigation to signifi­
cantly lower market-­
entry barriers for domestic firms. The previous two factors 
are mainly given, while governance strategy could be more deliberately taken. 
For developing countries that have not established a sound rule of law and strong 
domestic industries for pollution removal, goal-­
centered governance may induce 


168  Environmental technology and industry
a feasible path for improvement. In order to meet time constraints and technologi­
cal requirements, major Chinese firms universally licensed in foreign technolo­
gies to quickly build technological strength. In the early period, China had not 
established a system to well implement environmental policies and thus many 
SO2 scrubbers were not operating normally. For meeting governmental regula­
tions, coal-­
fired power plants chose to install the cheapest SO2 scrubbers but did 
not expect to run them. For domestic firms that had no technological advantages, 
this initially low but escalating requirements on the quality of SO2 scrubbers pro­
vided helpful stepping-­
stones to enter the market.
The utilization of wind energy followed a comparable path under goal-­
centered 
governance, which also helped to lower market-­
entry barriers for the establish­
ment of a domestic wind turbine industry. Similar to the SO2 mitigation case, the 
initial stage of wind energy development also focused more on the deployment 
to follow the goal-­
driven demand. In China’s 11th Five-­
Year Plan for Renew­
able Energy Development, the major goal for wind electricity referred to gen­
eration capacity whereas actual electricity generation served as a supplementary 
goal (NDRC, 2008). One average kilowatt-­
hour of wind capacity consistently 
generated much less electricity in a year in China than in the United States, and 
this partly indicated poorer operating conditions in China (Figure 7.7). When the 
deployment of wind turbines became sufficiently wide, the Chinese government 
started to pay more attention to their operation. Problems in the quality and opera­
tion of wind turbines emerged with their deployment to threaten not just wind 
0
20
40
60
80
100
120
140
160
180
200
0
50
100
150
200
250
300
350
400
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
Wind capacity (1,000 MW)
)
h
W
T
(
 
y
t
i
c
i
r
t
c
e
l
e
 
d
n
i
W
Year
Wind electricity: China (left)
Wind electricity: United States (left)
Wind capacity: China (right)
Wind capacity: United States (right)
Figure 7.7  
Wind energy development in China and the United States
Source: BP (2019).


Environmental technology and industry  169
electricity generation but, more important, also the safety of the electric grid and 
to push for greater focus and higher requirements (SERC, 2011). In 2010, the 
National Energy Administration published a plan to enact 247 technical stand­
ards for wind energy development, including several which were already in force 
(National Energy Administration, 2010). Lower technological market-­
entry bar­
riers played a positive role to encourage new firms. In 2006, the Chinese market 
had 12 firms that supplied wind turbines, and the number rose to 29 in 2012 (Shi, 
2007; China Wind Energy Association, 2012). Many component suppliers along 
the supply chain also actively entered the market (Chinese Wind Energy Equip­
ment Association, 2011). Compared to wind turbine manufacturers, market-­
entry 
barriers were even lower and the technologies were less complex for component 
suppliers, and this resulted in fiercer competition and thinner profit margins.
Furthermore, unlike SO2 scrubber firms, the Chinese firms in the wind industry 
licensed their technologies from a very different category of foreign firms. Foreign 
licensors of SO2 scrubber technologies were generally major firms that were closely 
involved in the downstream business of installing SO2 scrubbers (Xu, 2011a). In 
contrast, major foreign wind turbine manufacturers were largely reluctant to license 
technologies to Chinese firms, and most foreign licensors were design firms or small 
manufacturers that focused more on upstream technological development. This 
phenomenon is explained in the theory of markets for technology as the rational 
choice based on the respective industrial structure (Arora and Gambardella, 2010). 
The good-­
enough quality, lower price and no geographic constraints of technology 
licenses made the Chinese domestic wind industry potentially competitive.
The market for technology might also work for other large developing coun­
tries, such as India. They may also have potentially large markets through which 
to spawn many domestic operators and fierce competition. Many other low-­
carbon 
and pollution-­
control technologies have been commercialized with much know-­
how. A caveat is that these large developing countries may not necessarily always 
have large domestic markets for pollution mitigation. These are partly determined 
by government policies and not just by the overall sizes of their economies. Their 
abilities to take on board foreign technologies might not be consistently strong. 
However, there is great potential for large developing countries to make use of mar­
kets for technology to build their industrial prowess with mature technologies. Goal-­
centered governance may provide more feasible pathways for domestic industries in 
these developing countries to take roots and further grow from weak starting points.
3  
Environmental industry under goal-­
centered SO2 
mitigation path
3.1  Market entry and competition
Considering both firms that pollute the environment and others that provide pol­
lution removal facilities, the impacts of the goal-­
centered SO2 mitigation path in 
China may not be straightforward. On one hand, although empirical studies gen­
erated mixed results on the “pollution haven hypothesis” in the Chinese context 


170  Environmental technology and industry
(Levinson and Taylor, 2008; He, 2006; Shen, 2008), its key root cause – poor 
environmental regulation, including weak policies and poor enforcement  – is 
argued to potentially benefit polluting firms for not acting on, delaying or comply 
only partially with pollution control (Harney, 2008). In China, policies on envi­
ronmental protection and business standards were recognized by polluting firms 
as less important barriers to market entry (Niu et al., 2012). Relative to the Euro4 
fuel quality standards, the poorer Euro2 standards in China could reduce costs by 
1.1 and 1.9 U.S. cents per gallon for gasoline and diesel, respectively (Liu et al., 
2008). The cost burden also acts as a political and regulatory hurdle to bring pol­
luting firms under full compliance. On the other hand, from the perspective of 
supplying pollutant removal facilities, weak regulation could lower market-­
entry 
barriers to encourage competition, innovation and the establishment of industrial 
capacities for pollution control (Stigler, 1971; Dean and Brown, 1995).
Two important barriers on the supply side could slow down the deployment 
of SO2 scrubbers in China. No existing supply capacity could meet the unprec­
edented peak demand of over 100 GW a year (Figure 5.12). The capital costs of 
about US$65 to 90/kW (Figure 7.2) were initially too high, being over 10% of 
the costs of building new coal-­
fired power plants (SERC, 2006). If the large labor 
force and industrial base in China could be effectively mobilized for the deploy­
ment of SO2 scrubbers, the supply capacity would not have a major problem in 
meeting the rapidly growing demand. The lack of significant restrictions on for­
eign direct investment indicates that both foreign and domestic firms could tap 
into the labor force.
The huge Chinese market can easily accommodate many SO2 scrubber firms 
without losing economies of scale. Whether the supply potential could be released 
depends on whether existing firms could expand their capacity and (more impor­
tantly) whether new firms could emerge. Although the U.S. market had only 
about ten firms, and with new firms rarely entering, the Chinese market had over 
60 firms – almost all of which were newly established, most being domestic but 
some being foreign – thereby indicating much lower market-­
entry barriers (Fig­
ure 7.8). In the past decade, the annually added capacity of SO2 scrubbers increased 
significantly both in China and the United States, but the evolution of unit capital 
costs showed a rapid cost reduction in China and a cost spike in the United States 
(Figure 7.2). In China, the rapidly rising demand triggered intensive market entry 
to create fierce competition followed by a cost reduction whereas competition in 
the United States was rather limited, and this constrained the expansion of the sup­
ply capacity. When the demand for SO2 scrubbers grew, the price was pushed up.
As discussed earlier, domestic firms did not have technological advantages, 
especially in the early period. Nevertheless, because of the existence of many 
potential licensors in the technology market, no foreign firm was able to prevent 
others from licensing technologies to China. Technologies therefore could not be 
used as a barrier to exclude Chinese firms from competing. The crowded market 
enabled fierce competition not just for providing SO2 scrubbers. Competition also 
took place between foreign firms for licensing to especially promising Chinese 
firms that were expected to win many projects and return significant revenues 


Environmental technology and industry  171
0
10
20
30
40
50
60
70
0
2
4
6
8
10
12
14
2000 2002 2004 2006 2008 2010
2000 2002 2004 2006 2008 2010
Number of FGD companies in the Chinese market
t
e
k
r
a
m
 
.
S
.
U
 
e
h
t
 
n
i
 
s
e
i
n
a
p
m
o
c
 
D
G
F
 
f
o
 
r
e
b
m
u
N
Year
New entry
Existing
United States
China
Figure 7.8  
Firms in the Chinese and U.S. markets installing 100-MW-scale or greater SO2 
scrubbers
Source: Ministry of Environmental Protection (2008–2012); EIA (2007–2011); Xu (2013).
Note: “Existing”: firms have been in the market in the past. “New entry”: firms entering the market for 
the first time. The U.S. numbers use the left axis, and the Chinese numbers use the right axis.
from royalties. Those potential licensees were mainly established by coal-­
fired 
power producers. Interviews showed that financial payments were the most criti­
cal aspect of negotiating licenses, although other aspects were also important, 
such as the suitability of technologies and the scope of licenses. The willingness 
to accept lower up-­
front lump-­
sum fees and lower royalty rates made a licensor 
more competitive. After the significant variance of early contracts, the up-­
front 
lump-­
sum fee stabilized to be about US$1.2 million for wet scrubbers (Table 7.1).
3.2  
International competitiveness of China’s SO2 scrubber industry
Due to specific features in various environmental fields, goal-­
centered govern­
ance may present very different impacts on different environmental industries. 
One significant difference is on the international competitiveness of China’s SO2 
scrubber and wind turbine industries, as could clearly be seen from the reaction 
of the United States to China’s rising industrial prowess. Over the same period 
as China’s rapid growth was taking place, the United States also witnessed sig­
nificantly wider deployment. From 2004 to 2010, its SO2 scrubber capacity grew 
from 100 GW to 181 MW and its wind capacity from 6.8 GW to 40.3 GW (EIA, 


172  Environmental technology and industry
2012–2013). “The China Price” was a critical reason for trade disputes between 
China and the United States. In 2010, the price tag of SO2 scrubbers in China was 
about US$20/kW as revealed in the author’s fieldwork, whereas in the United 
States, it was US$206/kW (EIA, 2012–2013). For wind turbines, the average 
price in 2010 was US700/kWinChinaandUS1,460/kW in the United States 
(Figure 7.9). However, China’s SO2 scrubbers barely made any news in trade dis­
putes between the two countries while those of wind turbines were highly visible 
(Cooper, September 28, 2012). From another perspective, the Chinese SO2 scrub­
ber industry did not contribute to international SO2 mitigation whereas its wind 
industry strengthened the global CO2 mitigation capability.
Despite the success in building up the supply capacity and achieving cost 
reduction, China’s large SO2 scrubber industry did not become competitive in 
the international market as indicated by the nearly tenfold price difference in the 
segregated Chinese and U.S. markets (Figure 7.2). Many SO2 scrubbers were of 
low quality, and this increased the operation and maintenance costs and shortened 
their lifetimes. Although the delayed improvement of the operation of SO2 scrub­
bers was critical for lowering the initial quality requirement and technological 
barriers to market entry, after 2007 when the normal operation of SO2 scrubbers 
was largely expected, the prices stayed low. The gap between 2002 and 2007 
was too long and China was trapped in a low-­
quality bottom. The huge quality 
0
200
400
600
800
1,000
1,200
1,400
1,600
1,800
0
5
10
15
20
)
W
k
/
$
(
 
e
c
i
r
p
 
e
n
i
b
r
u
t
 
d
n
i
W
Annual installation of wind turbines (1,000 MW)
United States (2004–2010)
China (2004–2010)
Figure 7.9  
Average prices of wind turbines in China and the United States
Source: IEA and ERI (2011); Wiser and Bolinger (2012); BP (2019); Xu (2013).


Environmental technology and industry  173
premium presented serious financial challenges to power corporations. In addi­
tion, the quality of SO2 scrubbers was quite opaque to investors, and only the SO2 
scrubber firms had the best knowledge of the product. In the five gap years, a race 
to the bottom had pushed the quality and price of SO2 scrubbers to reach a mini­
mum and stable level. Because no SO2 scrubber firm had established a reputation 
for quality, any significant price increase would put the firm in a disadvantageous 
position in competition. Even when China started to allow BOT (Build, Operate, 
Transfer) contracts for SO2 scrubbers to better integrate the decisions of capi­
tal investment and daily operation (NDRC and SEPA, 2007), the trap remained 
a difficult one to escape from. Another important reason for the segregation of 
the Chinese and U.S. SO2 scrubber markets lay in the restriction of technology 
licensors. Almost every major Chinese SO2 scrubber firm licensed and relied on 
foreign technologies that felt themselves constrained in the Chinese market (Xu, 
2011a). Even projects in Hong Kong required special permission from technology 
licensors.
However, the lower market-­
entry barrier at the early stage of wind energy 
development was still much higher compared to that of SO2 scrubbers. Although 
costs were much lower in China than in the United States, a race to the bottom on 
quality and price did not happen and the price of China’s wind turbines remained 
stable (Figure 7.9). The operational requirement never dropped to a bottom as in 
the SO2 scrubber case. One critical reason lied in their different regulatory foun­
dations. Although the enforcement capacity for the deployment and operation of 
SO2 scrubbers could be built on the existing regulatory system, the weak environ­
mental policy enforcement indicated that such a system had not been satisfactorily 
established in China. In comparison, the compliance monitoring system for wind 
electricity delivery had been largely established despite wind energy being a new 
energy type for electricity supply. Furthermore, because electricity generation has 
direct and significant economic benefits to local governments, the political will 
for greater demand and better management was much stronger than in the case 
of SO2 scrubbers. Because the poor operation or quality of wind turbines would 
affect wind electricity generation and thus the revenue, investors in wind farms 
value quality substantially more than those investing in SO2 scrubbers.
Despite the highly visible trade disputes between China and the United States, 
the actual trade in wind turbines was minimal. In 2011, the total capacity of 
exported wind turbines was equivalent to only 1.3% of that installed domesti­
cally (China Wind Energy Association, 2012). Although four Chinese wind tur­
bine manufacturers had been ranked among the largest ten in the world, unlike the 
other six as regional or global suppliers, they remained largely domestic (Li et al., 
2011). Besides other influential factors, one important reason could be the quality 
gap that made the Chinese wind turbines fail to reach the technological market-­
entry barriers in developed countries. However, the Chinese wind industry could 
have a promising future. If the price difference between China and the United 
States were taken as the upper limit of the quality premium or the depth of the 
quality trap, the wind industry would be much more likely to escape the trap than 
the SO2 scrubber industry.


174  Environmental technology and industry
As demonstrated in the two comparative case studies, the depth of the low-­
quality trap could be determined by how long the operational improvement of 
pollution control facilities is delayed. The delay should be long enough for the 
domestic supply capacity to become established but short enough to prevent a 
race to the bottom on quality and price. Another influential factor on the depth 
of the trap is how strong the initial enforcement capacity is. Because electricity 
generation corresponds to much stronger enforcement capacity than the mitiga­
tion of conventional pollutants, China could have a better chance to build inter­
nationally competitive industries for renewable energy that generally has to be 
converted into electricity. Low market-­
entry barriers for quality and technological 
advancement are a key factor to make the Chinese market and industrial develop­
ment vibrant. In the later upgrading, China could focus more on raising the corre­
sponding requirements but on keeping other barriers low to minimize the negative 
impacts of such enhancement.
4  
Inter-­
goal coordination under goal-­
centered governance
China’s Five-­
Year Plans feature multiple goals in several fields, including econ­
omy, social development, environmental protection and resource conservation. 
Goals on economic growth rates are always the first one in the goal table in each 
Five-­
Year Plan, while they have been listed as “expecting” since the 11th Five-­
Year Plan when goals were first differentiated between “expecting” and “binding” 
(National People’s Congress, 2001, 2006, 2011, 2016, 1996). Although goals on 
environmental protection have been gaining importance and become “binding,” 
the relationship between economic development and environmental protection is 
still crucial to profoundly affect the sustainability of the environmental political 
will and the achievement of environmental goals. One pivotal concern is how to 
coordinate various goals for maximizing their potential synergies and minimiz­
ing conflicts. SO2 mitigation and economic development have two-­
way impacts. 
First, SO2 mitigation is one constraint for economic development. Energy con­
sumption and economic growth are fundamental drivers of SO2 emissions, whose 
mitigation thus reversely becomes a limiting factor. Second, SO2 mitigation also 
relies on the emergence and development of a pollution removal industry to fea­
sibly provide the technological means of SO2 mitigation, which could create new 
jobs and economic opportunities.
Over the past four decades, central economic planning has also gradually 
shifted toward decentralized market evolution. Various local governments are also 
actively competing with each other in establishing local industries that can serve 
the huge national market. One key feature of the four-­
decade economic reform 
has been the gradual peeling of constraints on the market. The state-­
owned sec­
tor has been generally retreating and those remaining ones are more profit-­
driven 
than like governmental agencies. China’s economic reform has created many mar­
kets from a negligible basis after the Cultural Revolution and greatly enhanced 
the importance of the markets. The boundary between the state and the market has 
also become clearer.


Environmental technology and industry  175
China’s SO2 mitigation path as examined earlier surely has contributed to its 
SO2 mitigation goals. At the same time, new economic opportunities emerged and 
were generally seized, which should also have facilitated the advancement of eco­
nomic goals. In comparison with rule-­
based governance, goal-­
centered govern­
ance has resulted in much lower requirements on inter-­
goal coordination. Local 
governments in China are the primary, decentralized entities to bear the respon­
sibilities and incentives for achieving both environmental and economic goals. 
They can have greater flexibility in adapting their policies and actions to take the 
best advantage of changing situations.
These goals are also crucial indicators of how the Chinese central government 
balances between environmental protection and economic development. When 
economic goals were emphasized while environmental goals were not, local gov­
ernments primarily focused on achieving economic goals. These goals are not 
fully coordinated but generally are independently implemented in a bottom-­
up 
manner. They do not demand centrally planned coordination either, as shown pre­
viously in China’s surprising emergence of the SO2 scrubber industry. They will 
seek appropriate ways for balancing how they achieve both goals. Decentralized 
policies and market evolution may utilize unexpected opportunities and circum­
vent unexpected difficulties in a much better way than any intelligent central plan­
ner can foresee in advance. Goal-­
centered governance thus can better maximize 
synergies and minimize conflicts among various goals and government tasks.
Note
	
1	 Adapted with permission from Xu, Y. 2011. China’s functioning market for sulfur diox­
ide scrubbing technologies. Environmental Science & Technology, 45, 9161–9167. Cop­
yright (2011) American Chemical Society; and Xu, Y. 2013. Comparative advantage 
strategy for rapid pollution mitigation in China. Environmental Science & Technology, 
47, 9596–9603. Copyright (2013) American Chemical Society. Much has been revised 
and expanded on.
References
Arora, A. & Fosfuri, A. 2003. Licensing the market for technology. Journal of Economic 
Behavior & Organization, 52, 277–295.
Arora, A., Fosfuri, A. & Gambardella, A. 2001a. Markets for technology and their implica­
tions for corporate strategy. Industrial and Corporate Change, 10, 419–451.
Arora, A., Fosfuri, A. & Gambardella, A. 2001b. Markets for technology: The economics 
of innovation and corporate strategy. Cambridge, MA: MIT Press.
Arora, A. & Gambardella, A. 2010. Ideas for rent: An overview of markets for technology. 
Industrial and Corporate Change, 19, 775–803.
BP. 2019. Statistical review of world energy [Online]. Available: www.bp.com/en/global/
corporate/energy-­
economics/statisticalreview-­
of-­
world-­
energy.html.
Chan, L. & Aldhaban, F. 2019. Technology transfer to China: With case studies in the high-­
speed rail industry. PICMET 2009 Proceedings, Portland, OR, August 2–6, 2858–2867.
China Wind Energy Association. 2012. Statistics on China’s wind turbine installation in 
2011. Beijing, China: China Wind Energy Association.


176  Environmental technology and industry
Chinese Wind Energy Equipment Association. 2011. Chinese wind turbine generator sys­
tem selection manual (edition 2011). Beijing, China: China Wind Energy Association.
Cooper, H. 2012. Obama orders Chinese company to end investment at sites near Drone 
Base. New York Times, September 28.
Dean, T. J. & Brown, R. L. 1995. Pollution regulation as a barrier to new firm entry – initial 
evidence and implications for future-­
research. Academy of Management Journal, 38, 
288–303.
EIA. 1986–2006. Annual steam-­
electric plant operation and design data (Form EIA-­
767). 
Washington, DC: U.S. Department of Energy.
EIA. 2007–2011. Electric generator report data (form EIA-­
860). Washington, DC: U.S. 
Department of Energy.
EIA. 2012–2013. Electric power annual 2010–2011. Washington, DC: U.S. Department 
of Energy.
Gans, J. S. & Stern, S. 2010. Is there a market for ideas? Industrial and Corporate Change, 
19, 805–837.
Gu, X. 2004. A summary of installing SO2 scrubbers in Luohuang power plant and the 
impact on the society and environment. Annual conference of the Chinese Association 
of Science, Hainan, China.
Guizhou Qiandong Power Station. 2005. Tendering document for the flue gas desulfuri­
zation island. Zhenyuan, Guizhou [Online]. Available: www.in-­
en.com/power/html/
power-­
2006200604145584.html [Accessed January 27, 2011].
Harney, A. 2008. The China price: The true cost of Chinese competitive advantage. New 
York: Penguin Press.
Harvey, I. 2008. Intellectual property rights: The catalyst to deliver low carbon technolo­
gies. Breaking the Climate Deadlock, Briefing Paper. London: The Climate Group.
He, J. 2006. Pollution haven hypothesis and environmental impacts of foreign direct invest­
ment: The case of industrial emission of sulfur dioxide (SO(2)) in Chinese provinces. 
Ecological Economics, 60, 228–245.
IEA & ERI. 2011. China wind energy development roadmap 2050. Paris, France: IEA, ERI.
Lefohn, A. S., Husar, J. D.  & Husar, R. B. 1999. Estimating historical anthropogenic 
global sulfur emission patterns for the period 1850–1990. Atmospheric Environment, 
33, 3435–3444.
Levinson, A. & Taylor, M. S. 2008. Unmasking the pollution haven effect. International 
Economic Review, 49, 223–254.
Lewis, J. I. 2007. Technology acquisition and innovation in the developing world: Wind 
turbine development in China and India. Studies in Comparative International Develop­
ment, 42, 208–232.
Li, J., Cai, F., Tang, W., Xie, H., Gao, H., Ma, L., Chang, Y. & Dong, L. 2011. China wind 
power outlook 2011. Beijing, China: China Environmental Science Press.
Liang, W. 2001. Power equipment of the gigantic three Gorges project. Proceedings of the 
Fifth International Conference on Electrical Machines and Systems, 1, 676–678.
Liu, H. A., He, K. B., He, D. Q., Fu, L. X., Zhou, Y., Walsh, M. P. & Blumberg, K. O. 
2008. Analysis of the impacts of fuel sulfur on vehicle emissions in China. Fuel, 87, 
3147–3154.
Metz, B., Turkson, J. K. & Intergovernmental Panel on Climate Change, Working Group 
III. 2000. Methodological and technological issues in technology transfer. Cambridge 
and New York: Cambridge University Press.
Ministry of Education. 1999–2019. Education statistics yearbook of China. Beijing, China: 
Ministry of Education.


Environmental technology and industry  177
Ministry of Environmental Protection. 2008–2012. The list of China’s SO2 scrubbers in 
coal-­
fired power plants. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2011a. China’s capacities of water treatment plants, 
SO2 scrubbers and NOx removal systems at coal power plants. Beijing, China: Ministry 
of Environmental Protection.
Ministry of Environmental Protection. 2011b. Statistical data on the environment. Beijing, 
China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2014. The list of China’s SO2 scrubbers in coal-­
fired 
power plants. Beijing, China: Ministry of Environmental Protection.
Mitsubishi Heavy Industries. 2011. Delivery record [Online]. Available: www.mhi.co.jp/
en/products/pdf/delivery_record.pdf [Accessed January 27, 2011].
National Bureau of Statistics. 1996–2019. China statistical yearbook. Beijing, China: 
China Statistics Press.
National Economic and Trade Commission. 2000. Key planning points on flue gas des­
ulfurization technologies and their localization (2000–2010). Beijing, China: National 
Economic and Trade Commission.
National Energy Administration. 2010. A set of standards for wind energy in the energy 
sector. Beijing, China: National Energy Administration.
National People’s Congress. 1996. Outlines of the 9th five-­
year plan and long-­
term goals 
in 2010 for economic and social development of the people’s republic of China. Beijing, 
China: The 4th Conference of the 10th National People’s Congress.
National People’s Congress. 2001. The outline of national 10th five-­
year plan on economic 
and social developments. Beijing, China: The 4th Conference of the 9th National Peo­
ple’s Congress.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. The outline of the 13th five-­
year plan on economic and 
social development. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
NDRC. 2008. The 11th five-­
year plan on renewable energy. Beijing, China: NDRC.
NDRC & SEPA. 2007. Working plan on experimenting BOT management of flue gas desul­
furization in coal power plants. Beijing, China: State Environmental Protection Admin­
istration, NDRC.
Niu, Y., Dong, L. C. & Chen, R. 2012. Market entry barriers in China. Journal of Business 
Research, 65, 68–76.
Ockwell, D. G., Haum, R., Mallett, A. & Watson, J. 2010. Intellectual property rights and 
low carbon technology transfer: Conflicting discourses of diffusion and development. 
Global Environmental Change-­
Human and Policy Dimensions, 20, 729–738.
Roth, A. E. 2008. What have we learned from market design? Economic Journal, 118, 
285–310.
Saggi, K. 2002. Trade, foreign direct investment, and international technology transfer: 
A survey. World Bank Research Observer, 17, 191–235.
SERC. 2006. Capital costs of electric projects completed in the 10th five-­
year plan. Bei­
jing, China: SERC.
SERC. 2011. Supervision report on wind electricity safety. Beijing, China: SERC.


178  Environmental technology and industry
Shen, J. 2008. Trade liberalization and environmental degradation in China. Applied Eco­
nomics, 40, 997–1004.
Shi, P. 2007. Statistics on China’s wind turbine installation in 2006. Beijing, China: China 
Wind Energy Association.
Shu, H. 2003. SO2 emission control for coal fired power plant. Electrical Equipment, 4, 
4–8.
Sina Finance. 2010. Information on Insigma Technology Co., Ltd. [Online]. Available: 
http://finance.sina.com.cn [Accessed February 9, 2011].
Srivastava, R. K., Jozewicz, W. & Singer, C. 2001. SO2 scrubbing technologies: A review. 
Environmental Progress, 20, 219–228.
State Council. 2006. The outline of national science and technology development plan in 
the middle and long term. Beijing, China: State Council.
State Intellectual Property Office. 2010. Database of patents granted in China. Beijing, 
China [Online]. Available: www.sipo.gov.cn/sipo2008/zljs/ [Accessed July 2, 2010].
Stigler, G. J. 1971. The theory of economic regulation. The Bell Journal of Economics and 
Management Science, 2, 3–21.
Strokova, V. 2010. International property rights index – 2010 report. Washington, DC: 
Americans for Tax Reform Foundation; Property Rights Alliance.
Teece, D. J. 1988. Capturing value from technological innovation – integration, strategic 
partnering, and licensing decisions. Interfaces, 18, 46–61.
United Nations. 1992. United nations framework convention on climate change. New 
York: United Nations.
U.S. Energy Information Administration. 2011. Official energy statistics from the U.S. gov­
ernment. Washington, DC [Online]. Available: www.eia.doe.gov/ [Accessed May 18, 
2011].
WIPO. 2019. WIPO statistics database [Online]. Available: https://www.wipo.int/edocs/
pubdocs/en/wipo_pub_943_2019.pdf.
Wiser, R. & Bolinger, M. 2012. 2011 wind technologies market report. Washington, DC: 
U.S. Department of Energy.
Xu, F., Yi, B., Zhuang, D., Yang, M., Yan, J. & Yan, Z. 2006. Survey report on the construc­
tion and operation of SO2 scrubbers at coal power plants in the 10th five-­
year plan. 
Beijing, China: The 4th Conference on Flue Gas Desulfurization Technologies.
Xu, Y. 2011a. China’s functioning market for sulfur dioxide scrubbing technologies. Envi­
ronmental Science and Technology, 45, 9161–9167.
Xu, Y. 2011b. Improvements in the operation of SO2 scrubbers in China’s coal power 
plants. Environmental Science & Technology, 45, 380–385.
Xu, Y. 2011c. The use of a goal for SO2 mitigation planning and management in China’s 
11th five-­
year plan. Journal of Environmental Planning and Management, 54, 769–783.
Xu, Y. 2013. Comparative advantage strategy for rapid pollution mitigation in China. Envi­
ronmental Science & Technology, 47, 9596–9603.
Xu, Y., Williams, R. H. & Socolow, R. H. 2009. China’s rapid deployment of SO2 scrub­
bers. Energy & Environmental Science, 459–465.


8	

Goal-­
centered governance
1  
Alternative governance models
China is experiencing very serious environmental damage. Nevertheless, the 
country in the past decade has achieved probably the fastest sulfur dioxide (SO2) 
mitigation pace for a large country. Significant progress has been made to clean up 
air and water. Its energy system has been gaining momentum to transition away 
from coal and toward renewables. With the economy more than 30 times bigger, 
SO2 emissions within one decade dropped to a level that was seen only before the 
economic reform era began in the late 1970s. Strong political will was formed to 
increasingly prioritize environmental protection among governmental affairs. The 
entire Chinese government across the central, provincial, municipality and county 
levels has been much better mobilized and committed. Policies are constantly 
enacted by various central and local authorities. The conventional poor policy 
implementation has been more effectively addressed and rapidly evolving to gain 
greater efficiency. In the coal-­
fired power sector, China managed to achieve essen­
tially universal coverage of SO2 scrubbers. More important, the original nonoper­
ation of SO2 scrubbers was also reversed to reach high SO2 removal rates. On the 
other hand, China established the largest SO2 scrubber industry, which provided 
employment and economic outputs. However, two decades ago at the early stage 
of China’s SO2 mitigation, few domestic firms existed with barely any domestic 
commercialized technologies. Although China has not been widely recognized 
by developed countries as a market economy, new firms were actively formed 
and swarmed into the new market to seek profitable opportunities. China’s inad­
equate protection of intellectual property rights did not seem to have prevented 
widespread market-­
based technology licensing from firms in developed countries. 
Despite numerous problems, China can claim great success in SO2 mitigation in 
the past two decades. These different components of environmental governance 
must work together to witness a favorable outcome. This book assesses the out­
come and, most important, aims to explain the trajectory.
Conventional wisdom can easily explain China’s environmental crises but has 
serious difficulties in understanding the cleanup process. Democracy and the 
rule of law are believed to be crucial contributors to forming strong political will 
and enabling the means to achieve pollution mitigation. However, China is not a 


180  Goal-­
centered governance
democracy and often ranked much behind developed countries in the rule-­
of-­
law 
index. Accordingly, we expect that China’s rapid economic growth will result in 
environmental crises and unacceptably high SO2 emissions, but the later, even 
faster SO2 mitigation is surprising because it defies the original expectations. 
China has not been fundamentally changed from the perspective of democracy 
and rule of law. The Chinese Communist Party is still the ruling political party in 
China. Governmental officials at various levels are still appointed but not demo­
cratically elected. Although certain progress has been made, Chinese society is 
still far from reaching the similar rule-­
based status as developed countries.
In one common conventional impression, the Chinese government is authori­
tarian and highly centralized with forceful central planning. Accordingly, in this 
theory, China’s environmental cleanup in the past two decades would be explained 
from the perspective of central planning. The central government might have 
designed the trajectory and its unchallenged authority could then implement such 
a design. This logic goes that the Chinese government does not have the checks 
and balances as in those democratic, developed countries, which enables China’s 
central planners to design an optimized path with good coordination among vari­
ous policy makers and implementers. When few domestic firms existed, the Chi­
nese government did not require the good operation of SO2 scrubbers to enable 
low technological market-­
entry barriers, provide and localize necessary supply 
capacities and reduce costs of SO2 mitigation. When many firms have been well 
established in the market, effluent emission standards and other regulations were 
made more stringent with better implementation for more effective SO2 mitiga­
tion. These newly emerged environmental industries provide economic opportu­
nities and cushion the negative impacts of stringent environmental protection on 
economic growth.
However, this explanation must assume that China’s central planners were 
extremely intelligent and well informed, but little evidence shows that such high-­
quality central planning has ever existed. As a developing country, China’s data 
collection system is less advanced than that in developed countries, especially two 
decades ago, to provide adequate data support for central planning. China’s com­
plexity and scale also make such high-­
level central planning intelligence impossi­
ble to achieve. The extreme centralization under Chairman Mao resulted in social, 
political and economic chaos with disastrous consequences. It is hardly convinc­
ing that central planning can lead to either rapid SO2 mitigation amid momentous 
economic growth or the establishment of a large SO2 scrubber industry.
Furthermore, the rule-­
based environmental governance that accounts for the 
trajectories in developed countries can also experience difficulties if applied to 
provide a primary explanation. As indicated in the World Bank’s governance 
indicators as well as in general impression, China’s performance has not been 
remarkable. China is still unable to make rules as important as developed coun­
tries prevalently do for environmental governance. In addition, under rule-­
based 
governance, although individual entities make their own decisions based on the 
rules, the rules are often centrally enacted by legislatures and/or courts as laws 
and the executive branch as regulations. Even if the rule of law is well established 


Goal-­
centered governance  181
in a society, whether rule-­
based governance can produce good outcomes depends 
on the quality of rulemaking. Rule-­
based governance alone is not a guarantee of a 
good outcome. Poorly designed rules and effective implementation may turn out 
to be undesirable, while policy making in China has not gained a decent reputa­
tion on its soundness, and consultation has also been much less thorough than 
that in developed countries. For example, before 1997, market speculation was a 
serious crime in China that was written into the Criminal Law. The intention was 
to maintain the order of a planned economy.
This book provides a different account of China’s environmental cleanup. 
China today has abandoned the Soviet-­
style central planning that was featured in 
the first three decades of the People’s Republic under the leadership of Chairman 
Mao. However, rule-­
based governance has not been well established. New laws 
and policies take a considerable amount of time to form and settle. For example, 
the Civil Code had just been enacted in May 2020 after many decades of grad­
ual formation. Instead, a new governance strategy has been tried and gradually 
become mature, with various goals taking the central stage. This goal-­
centered 
governance model is a mixture of centralization and decentralization to explain 
China’s SO2 mitigation trajectory much better than the central planning approach 
or rule-­
based governance can.
2  
Goal-­
centered governance
Readings of China are polarized, especially when China becomes bigger and 
more influential. One side profoundly denounces China and accuses the country 
of being messy, of not being a democracy, of having a rubber-­
stamp legislature 
and of being authoritarian without adequate respect to the rule of law. The Chinese 
government has been heavily criticized for breaking many rules that are highly 
valued in liberal democracies, such as those related to political liberty. Freedom 
of speech and civil society are constrained. Rising income inequality and privi­
leges of the wealthy and the powerful add social tensions. However, another side 
supports the Chinese government as they see many positive outcomes in China’s 
development. Together with rapid and sustained economic growth, the social wel­
fare system has been expanded dramatically to widen health care coverage even 
in rural communities, increase retirement pension and alleviate poverty. The Chi­
nese people can now enjoy living standards that were unimaginable one genera­
tion ago. They can largely choose where to live, work or travel as well as what to 
buy and sell. A great majority of the population has received significant returns 
of the economic development, although the distribution is uneven. Both views on 
China seem to have strong evidence to validate their claims. Then how can we 
understand China with these two sharply polarized readings? Are they connected? 
How China may further reform to embrace a better future?
For evidence-­
based researchers, the negative views on China could be mainly 
about rules and their implementation, while the positive views could be primarily 
shaped by outcomes. Although not all arguments on either side are sound, both 
views can find enough evidence to back them up. SO2 mitigation, or environmental 


182  Goal-­
centered governance
protection in general, is one government affair that exemplified such situations. 
The rapid mitigation was surprising but has been verified from multiple independ­
ent data sources, including external satellite data. Although active policy making 
and effective implementation were pivotal for achieving SO2 mitigation goals, 
many policies failed or were not implemented well. Initially, a large fleet of SO2 
scrubbers were built but not normally operating. In any understanding of China’s 
governance, a theoretical explanation should be able to accommodate both sides 
but not ignore the evidence of the other side. Furthermore, how are the two sides 
connected? In China’s case, does the favorable outcome have to be accompa­
nied by numerous policy blunders? If the rules were required to be well designed 
and implementable before putting into practice, would that affect the favorable 
outcomes?
This book explains China’s puzzles into a goal-­
centered governance model. As 
this book has examined in individual chapters on China’s SO2 mitigation, goal-­
centered governance has two foci, including goals and policies. Goals direct poli­
cies and policies achieve goals. Rule-­
based governance also has such two foci, but 
goals become secondary. The decisions in governance are mainly about enacting 
rules that are expected to be genuinely implemented. Fewer policies (or regula­
tions and laws) are enacted and the policy making might be more centralized, but 
they tend to be more carefully drafted. The outcome is an implicit product of such 
rules but not in the form of explicit, binding goals.
The goal-­
centered governance model can be understood from its organization 
mechanisms, features and applicability.
2.1  
Organization mechanisms
China has two hands in environmental governance, one visible and the other 
invisible. SO2 mitigation and environmental cleanup were achieved when the 
two hands cooperated. As a visible hand, the top leadership sets up prioritized 
goals with neither full-­
fledged deliberation nor stringent requirements on the path 
selection. The path results from bottom-­
up efforts of decentralized stakeholders 
as directed by an invisible hand of governance. The invisible hand of the market 
has been widely recognized and utilized. Rational market participants maximize 
their self-­
interests or profits, while this decentralized process also leads to the 
maximization of a society’s overall economic interest. Goal-­
centered governance 
could resemble and enable such an invisible hand to guide the central and local 
governments toward goal attainment. When their self-­
interests are served with 
various incentives for goal attainment, the overall goal will be achieved to sat­
isfy society’s overall interest. If more stringent goals are enacted, the incentives 
should also be strengthened. In order to finally achieve environmental cleanup, 
environmental goals must be prioritized with increasing stringency over a long 
period. If goals are changed, the invisible hand will direct the system away from 
the original goals and toward new ones.
As illustrated in Figure 8.1, goal-­
centered governance comprises three pillars: 
centralized goal setting, decentralized goal attainment and decentralized policy 


Goal-­
centered governance  183
making and implementation. First, the process for setting up goals of nationwide 
priority is highly centralized. The top leadership, with the Political Bureau of the 
Chinese Communist Party and its Standing Committee at the core, is in charge 
of supplying the country with goals as they deem crucial, especially in Five-­
Year 
Plans. The relationship among different goals could be balanced at this stage. 
Some goals could be prioritized that correspond to higher ratings in the perfor­
mance assessment of local leaders. In the case of SO2 mitigation, the Chinese top 
leadership did generally respond to what society wants, although the process was 
not democratic. The goals on SO2 mitigation and environmental protection were 
revised more stringent when such demand escalated.
Second, for decentralized goal attainment, national goals are distributed to 
provincial governments and then lower-­
level local governments, as in the case 
of SO2 mitigation and environmental protection goals. These individualized, 
quantitative goals guide the efforts of local governments and related ministries. 
Strong enough incentives are put into place to reward goal attainment and pun­
ish failures. Because China’s local leaders are appointed but not elected, their 
jobs are explicitly linked to the performance of achieving various goals with 
different priorities. The Chinese Communist Party’s organization plays a crucial 
role in establishing such a crucial personnel relationship between the central and 
provincial governments and their further subsidiaries. In addition, the central 
government receives much greater revenues than it spends, while the situation 
for local governments is generally the opposite: to demand a significant fiscal 
transfer from the central government. If local governments failed their individual 
goals, their leaders would face grim opportunities of promotion and could even 
be removed. Those who outperform others are distinguished for better promotion 
opportunities.
Third, policy making and implementation are heavily decentralized. With the 
responsibility of achieving goals, local governments have sufficient flexibility, 
authority and capacity for policy making and especially implementation, while 
the central government is especially weak in policy implementation. Require­
ments are significantly lowered on the quality of policy making, the optimal 
choice of policy instrument and coordination among policies. As a developing 
Policy Implementation
Policy Making
Local 
Governments
Pollution Control Firms
Polluting Firms
Localized Goals 
& Incentives
Society & Economy
Central Government
National Goals
Top Leadership
Centralized goal setting
Decentralized goal 
attainment
Decentralized policy 
making & implementation
Figure 8.1  
An illustration of the goal-centered governance model


184  Goal-­
centered governance
country, China has not acquired enough strengths from these perspectives despite 
continuous improvement. The weak rule of law indicates that the system neither 
requires nor ensures their genuine implementation. Policies compete with each 
other and evolve with implementation selection.
2.2  
Features
Under goal-­
centered governance, several key features could emerge.
First, not all goals are important and prioritized goals are few. The mobilization 
of the entire Chinese government, from central to local levels, depends on the cred­
ible incentives for their goal attainment performance. Any additional goal could 
dilute the effectiveness of existing ones. Accordingly, the number of nationally 
prioritized goals should be constrained, while provincial governments and central 
ministries may have their second-­
tier goals with lower priorities. Governmental 
efforts are highly concentrated on those goals of high priority, while in areas with 
lesser or no goals, the performance could be significantly compromised.
Second, policy making is active and each makes an incremental contribution 
to goal attainment. Local governments and central ministries are mandated to 
achieve their individualized goals. The incentives are mainly associated with the 
goals’ attainment, while any mistakes in policy making and implementation are 
much more leniently accommodated. Furthermore, they also have great authority 
and flexibility in policy making, adoption, innovation and learning in the decen­
tralized arrangement. These favorable conditions encourage active policy making, 
as witnessed in the case of SO2 mitigation. Because it is local governments but 
not their environmental protection bureaus that bear the responsibility of achiev­
ing goals, they often involve multiple bureaus in making their specialized policies 
that may contribute to SO2 mitigation. Ministry of Ecology and Environment, its 
predecessors and its composing departments, as well as other central ministries, 
have also been actively trying new policy tools. Unlike the situation in the United 
States that the Acid Rain Program in the Clean Air Act Amendments (1990) and 
its previous versions may claim a lion’s share of credits, China does not feature 
any pivotal policy of similar importance for SO2 mitigation, while SO2 mitigation 
goals were achieved through numerous policies and each contributed a small and 
accumulative share.
Third, more policy failures exist and policy implementation is selective. These 
may be seen as the necessary costs of the goal-­
centered governance model, espe­
cially when China is still in the process of strengthening its policy-­
making quality 
and policy implementation effectiveness. Policies in China may fail from multi­
ple perspectives. The design itself may be less mature and flawed. Decentralized 
policy making indicates that not all policy makers, especially those in local gov­
ernments, have adequate intellectual support. Policy implementation may have 
unexpectedly high obstacles from various interest groups or weak enforcement 
capacity. To ensure the faithful implementation of individual policies is only a 
secondary priority for local governments. When good implementation of a certain 
policy contributes significantly to goals, more efforts will be directed to this issue. 


Goal-­
centered governance  185
For SO2 mitigation, those policies on installing SO2 scrubbers were first targeted 
in implementation, while their operation was only made a priority later when the 
significant and growing fleet of SO2 scrubbers increased the impacts of such pol­
icy on reducing SO2 emissions. Environmental policy implementation capacity 
was strengthened, and new environmental compliance monitoring technologies 
were actively adopted with SO2 mitigation goals in primary focus.
Fourth, requirements on goal coordination are lower. With impacts on SO2 mitiga­
tion, industrial, energy and environmental policies are enacted generally indepen­
dently from each other for achieving their specific goals. Various policies for one or 
multiple goals could have synergies and/or conflicts. In goal-­
centered governance 
for SO2 mitigation, policy coordination largely is not centrally organized. Conflicting 
policies may not be implemented well to positively contribute to goal attainment, 
and thus, they would dwindle. Those compatible policies that have synergies will be 
expanded from local to national levels or adopted from one region to another. In other 
words, such policy coordination is not achieved primarily through intentional intel­
ligent design but via bottom-­
up evolution through implementation selection.
Fifth, requirements on information availability and measurability are lower 
with moral hazards better contained. Policy making and implementation are much 
more data-­
intensive than the assessment of goal attainment. Significant uncertain­
ties exist and many potential factors could affect the final outcome, such as in 
the case of SO2 mitigation. Because the efforts of local governments are difficult 
to accurately measure and sometimes hardly observable, local leaders in China 
may simply pay frequent lip service, emphasize constraints and external factors 
other than their own efforts but behave differently in reality. Comparison across 
regions then faces high hurdles to disable effective competition among local gov­
ernments. However, under goal-­
centered governance, goals are primarily on those 
measurable outcome indicators, such as SO2 emissions and air quality, which sig­
nificantly reduce the required information. Lip service is much less helpful than 
actual efforts for achieving goals.
Corresponding to the questions that are raised in the book, the coexistence of 
favorable outcomes and unfavorable policy pathways is only puzzling because 
they cannot be properly explained by the rule-­
based or central planning govern­
ance models, while a decent theoretical understanding can be reached with the 
goal-­
centered governance model. If the system has a very low tolerance for prob­
lems in policy making and implementation, especially for China as a developing 
country, the favorable outcomes might indeed be seriously compromised. Nev­
ertheless, the costs of policy deficiencies can be reduced when China gradually 
acquires the capability and capacity for high-­
quality policy making and effective 
policy implementation.
2.3  
Applicability
Since the Qin dynasty (221–207 BCE) first established centralized rule in China, 
local governments have always been crucial in Chinese governance to distinguish 
the importance of the central–local relationship. The vast territory and population, 


186  Goal-­
centered governance
as well as huge regional differences, weaken direct ruling by the emperors or 
prime ministers who reside in the distant capital. Although China has long been 
enacting laws and policies in texts, such as those by Shang Yang in a major reform 
in the 4th century BCE that led to the rise of the Qin Kingdom, the modern sense 
of the rule of law has never been well established to occupy the central stage of 
governance.
Corresponding to the organization mechanisms of goal-­
centered governance, 
the system may fail under three situations. First, the achievement of governmental 
goals does not lead to outcomes that the society wants. The supply of goals by 
the top leadership may have a lag or lead from the demand, but the gap should 
not be too wide to let the system fail. This concern is closely related to arguments 
in China’s context without democracy. When China was much poorer and the 
public prioritized economic growth and jobs over environmental protection, envi­
ronmental goals were ranked much lower than economic goals. When the public 
started to pay more attention to life quality and clean environment, environmental 
goals should then be ranked high among governmental affairs. It is not neces­
sary that the goals are exactly identical as what the society desires. For example, 
the maximization of long-­
term tax revenues may be compatible with improving 
the living standard of the public. After the Mongol empire under Genghis Khan 
occupied North China in early 13th century, one high-­
ranking official suggested 
eliminating all Han Chinese and using the land for grazing because Han Chinese’s 
primary economic activities were not raising animals. His goal was for the land to 
generate more tax revenues. Another key advisor to Genghis Khan, Yelv Chucai, 
proposed that if the Han Chinese could be left alive to still engage in agriculture 
and business, they would contribute much more tax. His advice was taken, and the 
outcome was favorable to both the Mongol court and the people.
Second, the decentralized goal attainment fails. The central government may 
not be able to impose their prioritized goals onto local governments. A frequent 
complaint in the Chinese government was that “policies and orders cannot go 
beyond Zhongnanhai.” Zhongnanhai, or “Central and Southern Seas,” is a com­
pound in Beijing where the central government of the People’s Republic of China 
is located. This sentence generally means that the central government cannot 
smoothly impose their policies and orders onto local governments. Even Chair­
man Mao complained before the Cultural Revolution that the Beijing municipal 
government was “penetrable by neither water nor needles.” Local governments 
and central ministries may malfunction or no effective incentives are available to 
incentivise or force them to work for their assigned goals. A long-­
lasting ques­
tion in the Chinese history is the collapse of the Ming dynasty (1368–1644) in 
early 17th century. Historians pointed out one crucial reason in Emperor Wanli 
(r. 1572–1620) when he left many key positions vacant and the government could 
not function (Huang, 1981). In the later decades of the Tang dynasty (618–907), 
local leaders had exclusive power over military, civil affairs and fiscal revenue. 
They could also pass their titles to heirs who were chosen by themselves. Essen­
tially, local governments were semi-­
independent kingdoms, which eventually led 
to the collapse of the Tang dynasty.


Goal-­
centered governance  187
Third, policy making and implementation are overcentralized, and local gov­
ernments have very limited flexibility or capability in choosing their own paths 
for achieving goals. One-­
size-­
fits-­
all rules from Beijing may be at a great distance 
from diverging regional realities to undermine their effectiveness and efficiencies. 
Active policy making, innovation and learning could be suppressed with overcen­
tralization or when mistakes were much less accommodated. When policy-­
making 
authorities, fiscal revenues/expenditures and capable officials are concentrated 
into the central government, local governments may be too weak to perform their 
jobs well. Local governments in wealthy regions may experience little difficulty 
in attracting capable employees or building enough capacity in policy making and 
implementation for achieving their goals. However, China has significant regional 
disparity in economic development. If left alone, poor regions would not be able 
to utilize the policy and technological tools effectively and efficiently.
Goal-­
centered governance is mainly for new and evolving governmental affairs 
without well-­
established policies. In comparison to two decades ago, China has 
designed, enacted and implemented many policies for SO2 mitigation and other 
environmental goals. Many will last to make SO2 mitigation a routine governmen­
tal affair, such as the effluent emission standards of thermal power plants. These 
tested policies and correspondingly strengthened implementation systems will 
form an escalating base for the continuous advancement of environmental protec­
tion until reaching fundamental solutions. Then goal-­
centered governance could 
gradually give way to rule-­
based governance and other governmental affairs may 
receive more attention with prioritized goals. In the past two decades, key envi­
ronmental goals in China’s Five-­
Year Plans have been extended from SO2 and 
chemical oxygen demand (COD) in the 11th Five-­
Year Plan (2006–2010), plus 
ammonia-­
nitrogen (NH3–N) and nitrogen oxide (NOx) in the 12th Five-­
Year Plan 
(2011–2015), plus water quality grade, the Air Quality Index and fine particulate 
matter (PM2.5) concentrations in the 13th Five-­
Year Plan (2016–2020; National 
People’s Congress, 2011, 2006, 2016). With the continuous progress, it will not be 
surprising to see that SO2 mitigation goal removed and an ozone (O3) goal added 
in the future, if not in the upcoming 14th Five-­
Year Plan (2021–2025).
This goal-­
centered governance has been tested as an effective strategy for China 
to make rapid advancement from unfavorable situations and to significantly lower 
key requirements on policy making as in a rule-­
based governance system. From 
one perspective, it is an effective and efficient path-­
finding strategy for China 
to reach a more sustainable, rule-­
based future. With new problems continuously 
emerging, it should and will be the crucial strategy in China’s future governance 
even when China reaches the stage of a developed country.
The goal-­
centered governance model may be best utilized in countries with 
the following characteristics: (1) newly prioritized governmental affairs or others 
with rapid evolution to require continuous focus; (2) developing countries where 
policies have not been maturely established and policy making has not achieved 
adequate quality and acquired enough data and intellectual support; (3) being 
large in scale with genuine necessity of multiple governmental levels and where 
the central government can impose adequate incentives on local governments to 


188  Goal-­
centered governance
encourage policy innovation, while goal evaluation is largely fair with good data 
support and rewards are issued based mainly on meritocracy; (4) where the system 
is more tolerant to mistakes in policy making and implementation and pays pri­
mary attention to outcomes and only secondarily on paths; and (5) local govern­
ments are capable of policy innovation and resourceful for policy implementation.
Countries in federal systems may not find this governance model applicable 
because incentives very likely are neither adequately available nor strong enough 
for the federal government to incentivize state governments. Small countries may 
not need this governance strategy as the central government is much closer to the 
society and local governments are not as important as those in large countries. For 
countries that have established sound rule of law, goal-­
centered governance may 
not occupy center stage either because the system is less tolerant of mistakes in 
policy making and implementation, while active policy innovation may indeed 
encounter more mistakes and failures. Highly centralized countries in policy mak­
ing and implementation may constrain such bottom-­
up efforts as well. This goal-­
centered governance model is not necessarily inapplicable in democracies, but 
the application nevertheless may be much constrained if competition across local 
governments may not have enough impetus and incentives.
Despite the constraints of its applicability, governments at various levels across 
countries with different institutional and developmental contexts may still be able 
to draw helpful insights from the goal-­
centered governance model and explicitly 
apply goals in organizing their governance. Decentralized policy innovation and 
competition can be encouraged in countries with sound rule of law, despite vari­
ous constraints of existing rules.
2.4  
Comparison with other theories
This study’s development of the goal-­
centered governance model not only ben­
efits immensely from earlier theoretical explorations but also demonstrates sig­
nificant differences.
Goal-­
setting theory in social psychology is one key intellectual source (Latham 
et al., 2008; Latham and Yukl, 1975; Locke and Latham, 1990, 2002; Locke et al., 
1981). The goal-­
setting theory mainly emphasizes on how goals could enhance 
task performance of individuals, while goal-­
centered governance pays primary 
attention to the performance of local governments, central ministries and other 
governmental agencies. In addition, the latter has a heavy focus on the flexibility 
of those decentralized stakeholders in utilizing policies for achieving those goals.
The goal-­
centered governance model can be regarded as a specific application 
of pragmatism with clear directions (Alford and Hughes, 2008), while it places 
goals at the center and makes policies instrumental. The criteria of assessing pol­
icies are based on whether they contribute, undermine or have no impacts on 
goal attainment in actual contexts but not on prior selection. Policy innovation, 
competition, revision, learning and expansion are common, and specific policies 
will rarely be unequivocally relied on. This governance model is a theoretical 
extension of Deng Xiaoping’s cat theory. Deng Xiaoping was officially accredited 


Goal-­
centered governance  189
as the “chief architect of China’s reform and open-­
up” by the Chinese Commu­
nist Party. However, he did not have a clear long-­
term blueprint on how China’s 
economic reform should proceed when China just got out of the devastation of 
the Cultural Revolution, but many doctrines remained strong. As summarized in 
his famous quote, “regardless of whether the cat is a white cat or a black cat, as 
long as it can catch mice, it is a good cat.” He was less interested in the debate 
about whether China’s economic reform may contain too much capitalism but 
mainly focused on whether the country can prosper at a faster pace. This strategy 
was sharply different from Chairman Mao’s, under whose leadership China had a 
stringent restriction on the choice of paths or “cats.” Another famous quote could 
summarize his main idea: “we would rather have socialistic grass than capitalistic 
grain.” This goal-­
centered governance has clear directions as specified in goals, 
but the pathfinding is much less constrained.
It also echoes adaptive and polycentric governance to address complexity and 
uncertainty that emphasize localized solutions (Dietz et al., 2003; Chaffin et al., 
2014; Ostrom, 2010). This goal-­
centered governance emphasizes more on how 
these solutions could evolve in decentralized and bottom-­
up manners with moti­
vated local governments under the centralized direction of goals. In comparison 
to the comparative advantage strategy that advocates good, incremental improve­
ments but not perfect, once-­
and-­
for-­
all solutions to environmental problems (Xu, 
2013), this goal-­
centered governance model is more incorporative to explain in 
what conditions the comparative advantage strategy will be taken, why it can 
work and what impacts it may exert on governance. The competition among local 
governments and other goal bearers borrows the idea from the Tiebout model 
(Tiebout, 1956), but they are also quite different. The incentives for the competi­
tion are not bottom up from local citizens but are top down from imposed goals. 
For explaining the development of China’s environmental industries, the ecologi­
cal modernization theory may provide an alternative understanding that connects 
environmental protection with economic modernization (Hajer, 1995; Zhang 
et al., 2007). The goal-­
centered governance model, in comparison, explains that 
the impacts on environmental industries were not intentionally planned, and envi­
ronmental and economic policies were not deliberately coordinated for new firms 
in a developing country like China to actively enter the market and grow up.
Incrementalism is another crucial intellectual source to build the goal-­
centered 
governance model (Lindblom, 1959; Lindblom, 1979). Neither emphasizes on 
key, deliberately designed policies with maximized impacts on achieving objec­
tives, but each policy should make incremental but accumulative contributions. 
However, goal-­
centered governance does have explicit goals at the center as ends, 
while policy making is not centralized for finding optimized means. Instead, the 
incremental improvement was made by decentralized local governments, not by 
centralized policy makers. Goal-­
centered governance is compatible with Joseph 
Stigler’s economic theory of regulation (Stigler, 1971). It understands the demand 
for regulations with an additional key source from goals, while the supply of 
regulations is decentralized to witness active policy making, innovation and 
competition.


190  Goal-­
centered governance
3  
Implications
In the past two centuries, China has tried, voluntarily or involuntarily, many dif­
ferent governance models. When one model was proved ineffective, reforms were 
attempted, and frequently, revolutions were started. Even under the rule of the 
Chinese Communist Party since 1949, China has tried sharply different models. 
Under Chairman Mao, the Chinese government was much more centralized. His 
goals significantly deviated away from what the society wanted, but no effective 
checks could counterbalance and prevent his goals from becoming the nation’s. 
The results were disastrous.
Through trial and error and with tremendous costs, China should have found 
an effective model to govern the vast, complex, developing country with a deep 
institutional history. The goal-­
centered governance model has demonstrated its 
effectiveness and efficiency in fundamentally reversing the rising trend of SO2 
emissions as well as China’s multifaceted environmental crises. Nevertheless, the 
governance model has two potentially highly damaging risks. First, goals may 
not be formed to satisfy society’s demands, like what happened under Chairman 
Mao. The current focus on environmental protection could have a chance to be 
disrupted, and thus, the entire governance system would be directed in another 
direction. Second, overcentralization and low tolerance to policy mistakes may 
undermine the system’s effectiveness and efficiency. Local governments and 
other governmental agencies may be weakened on the incentives, authorities and 
capacities of policy making and implementation. One indicator would be whether 
policy innovation and learning are still active.
A famous quote from Voltaire, a French writer, is that “the perfect is the enemy 
of the good.” The goal-­
centered governance model is far from being perfect. Even 
when it achieves great success, the process is full of stumbles, policy deficien­
cies, unsatisfactory policy implementation and even frequent abuse of govern­
mental authorities. However, as China has tried, alternative governance models 
may hardly provide better outcomes due to difficulties from uncertainties, com­
plexities and data inadequacy in China’s contexts, although they may work well 
in another country’s contexts. Rule-­
based governance demands high requirements 
on policy making quality, optimal choice of policy instruments and inter-­
policy 
coordination, but these were not China’s strengths especially in the early stages of 
dealing with major issues such as SO2 mitigation and environmental cleanup. This 
goal-­
centered governance is a “good” model but certainly not a “perfect” one due 
to its numerous weaknesses. Especially for developing countries with many diffi­
culties in policy making and implementation, this proven “good” model provides 
a promising way to organize governance for achieving what the society deems 
significantly desirable, while a “perfect” governance model may be unreachable. 
The pursuit of being perfect should not stop a country from becoming better.
Environmental crises that have accumulated over a few decades cannot be 
solved within a few years. Efforts should be sustained even when the govern­
ment changes after elections or leadership reshuffle. In developed countries, the 
rule-­
based governance model has been effective to achieve economic prosperity 


Goal-­
centered governance  191
and later sustained reduction of pollution with laws at the center. The gradually 
formed and tested goal-­
centered governance model offers a feasible method for 
China to fundamentally solve environmental degradation problems. The SO2 mit­
igation has transcended multiple Five-­
Year Plans since the 9th Five-­
Year Plan 
(1996–2000) under three top leaderships. The demand for environmental quality 
has grown stronger among the public, and China’s top leadership has also been 
largely supplying national goals to match the demand. It is expected that environ­
mental goals will remain highly prioritized among governmental affairs in China.
Climate change is a much greater environmental problem than any conven­
tional air or water pollution. This goal-­
centered governance model has also been 
used in tackling the mitigation of China’s greenhouse gas emissions since the 12th 
Five-­
Year Plan (2011–2015) when a goal to reduce carbon dioxide (CO2) intensity 
by 17% over the five years was first written into the national plan (National Peo­
ple’s Congress, 2011). Goal attainment, policy making and implementation have 
also been heavily decentralized. The market has been actively taking advantage 
of economic opportunities from CO2 mitigation to develop, deploy and innovate 
technologies, such as renewable energy, electric vehicles and energy efficiency. 
Similar to SO2 mitigation, CO2 mitigation has centralized goals, but its actual 
attainment is largely decentralized. It is expected that this goal-­
centered govern­
ance model will also lead to China’s eventual transition of climate mitigation.
References
Alford, J. & Hughes, O. 2008. Public value pragmatism as the next phase of public man­
agement. American Review of Public Administration, 38, 130–148.
Chaffin, B. C., Gosnell, H. & Cosens, B. A. 2014. A decade of adaptive governance schol­
arship: Synthesis and future directions. Ecology and Society, 19.
Dietz, T., Ostrom, E. & Stern, P. C. 2003. The struggle to govern the commons. Science, 
302, 1907–1912.
Hajer, M. A. 1995. The politics of environmental discourse: Ecological modernization and 
the policy process. Oxford and New York: Clarendon Press, Oxford University Press.
Huang, R. 1981. 1587, a year of no significance: The Ming dynasty in decline. New Haven: 
Yale University Press.
Latham, G. P., Borgogni, L. & Petitta, L. 2008. Goal setting and performance management 
in the public sector. International Public Management Journal, 11, 385–403, 113.
Latham, G. P. & Yukl, G. A. 1975. Review of research on application of goal setting in 
organizations. Academy of Management Journal, 18, 824–845.
Lindblom, C. E. 1959. The science of muddling through. Public Administration Review, 
19, 79–88.
Lindblom, C. E. 1979. Still muddling, not yet through. Public Administration Review, 39, 
517–526.
Locke, E. A. & Latham, G. P. 1990. A theory of goal setting & task performance. Engle­
wood Cliffs, NJ: Prentice Hall.
Locke, E. A. & Latham, G. P. 2002. Building a practically useful theory of goal setting and 
task motivation – A 35-­
year odyssey. American Psychologist, 57, 705–717.
Locke, E. A., Saari, L. M., Shaw, K. N. & Latham, G. P. 1981. Goal setting and task-­
performance – 1969–1980. Psychological Bulletin, 90, 125–152.


192  Goal-­
centered governance
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. The outline of the 13th five-­
year plan on economic and 
social development. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
Ostrom, E. 2010. Beyond markets and states: Polycentric governance of complex eco­
nomic systems. American Economic Review, 100, 641–672.
Stigler, G. J. 1971. The theory of economic regulation. The Bell Journal of Economics and 
Management Science, 2, 3–21.
Tiebout, C. M. 1956. A pure theory of local expenditures. Journal of Political Economy, 
64, 416–424.
Xu, Y. 2013. Comparative advantage strategy for rapid pollution mitigation in China. Envi­
ronmental Science & Technology, 47, 9596–9603.
Zhang, L., Mol, A. P. J. & Sonnenfeld, D. A. 2007. The interpretation of ecological mod­
ernisation in China. Environmental Politics, 16, 659–668.


Page numbers in italic indicate a figure and page numbers in bold indicate a table on the 
corresponding page.
Index
3rd Five-Year Plan (1966–1970) 44
6th Five-Year Plan (1981–1985) 43
9th Five-Year Plan (1996–2000) 18, 62
10th Five-Year Plan (2001–2005) 45 – 48, 
62, 66, 92, 98
11th Five-Year Plan (2006–2010) 20, 
43 – 49, 57, 62, 64, 66, 67, 78, 92, 
98 – 99, 116, 168, 187
12th Five-Year Plan (2011–2015) 119, 
187, 191
13th Five-Year Plan (2016–2020) 187
Academy of Environmental Planning 29
Academy of Environmental Sciences 29
accountability 64
acid rain and SO2 pollution control, 11th 
Five-Year Plan on 43, 46, 93
Acid Rain Program, U.S. 49, 78, 79, 105, 
124, 184
Action Outline for Promoting Big Data 
Development 124
administration 28 – 29, 29, 30, 33
Africa, energy consumption and 
electrification rate in 89
agricultural pollution, under Ministry of 
Agriculture 27
air and water pollution: in China 1; 
controlling 65; DALYs in China due 3, 
3; mitigation 23; premature deaths due 
to 1, 2, 3
Air Quality Index (AQI) 65, 66, 71
ambient particulate matter (PM) pollution: 
cause of 86; in China 1 – 4, 2, 3; in India 
4 – 5, 4; PM2.5 goals 66
Asian financial crisis of 1997 18
Association of Environmental Protection 
Industries, China’s 151
autocracy, democracy and 6
Basic Thoughts of the National 11th Five-
Year Plan, The 45, 46
Beijing, AQI in 71, 71
Blackman, A. 107
“blue sky” 66
BOT (Build, Operate, Transfer) contracts, 
for SO2 scrubbers 173
budget balance: of central and local 
governments 34, 35; by provinces 35, 36
calcium/sulfur (Ca/S) molar ratio in coal 
93, 110
campaigns/movements (yundong) 
135 – 136
carbon dioxide (CO2); emissions 6; 
mitigation, goal of 43, 191
Central Department of Organization 40
central government: budget balance 34, 
35, 36; in charge of policy making 28, 
35; environmental authorities at 29, 33; 
governmental revenue and expenditure 
to GDP ratios by 33 – 35, 33; reforms 
at 27; shares of expenditures (2018) 
36 – 38, 37 – 38
centralized: and decentralized personnel 
management 39 – 40; goal setting 44 – 49, 
183; political will 17 – 18
China 25, 43; administrative reform 27; 
air quality 1, 2; average prices of wind 
turbines in 172 – 173, 172; central–local 
fiscal relationship (1994) 34 – 35; central 
planning 180; challenges in policy 
making 77 – 80; coal consumption in 
10, 11, 84, 87, 90; coal-fired power and 
SO2 scrubber capacities in 96 – 99, 97, 
98; DALYs in 3, 3, 4; deployment and 
operation of SO2 scrubbers in 149 – 150, 
150; economic growth in 85 – 87, 85; 


194  Index
economy 9, 18, 22; electricity 
generation by fuels in 89 – 90, 90; 
employees on environmental protection 
(2015) 33; energy consumption 43, 78, 
86 – 90, 86, 88, 92, 174; energy intensity 
goal 43, 78, 84, 92; environmental 
agencies in 28; environmental 
compliance in 105; environmental 
crises 1 – 5, 180; environmental 
policies and laws in 29, 31 – 32, 39; 
environmental/renewable energy 
industries 23; financial sector 18; firms 
in 170, 171; GDP in 7 – 8, 7, 18, 19, 22; 
goal-centered governance in 80 – 84; 
goal-centered policy implementation 
105 – 109; goals in Five-Year Plans 
42 – 44; governance indicators of 
8 – 9, 8; governmental income/
expenditure-to-GDP ratio in 33 – 35, 
33; international competitiveness of 
SO2 scrubber industry 171 – 174; job 
and demographic structures 18 – 20, 19, 
22; Law of Environmental Protection 
47; Law of Standardization 47, 96; 
laws in 31; leadership change 44 – 45; 
market-incentive policies 79; mobilizing 
government 42 – 72; NGOs in 17; 
patents on environmental technology 
161 – 162, 162; policies in 83; polity 
democracy index for 5 – 6, 5; pollution 
mitigation in 149; power sector shares 
10, 11; premature deaths 1, 2; provincial 
environmental authorities 30; R&D 
expenditures in 161, 161; shares in 
governmental expenditure for 36 – 39; 
strategies on environmental protection 
65; sulfur contents distribution in coal 
power plants 93 – 94, 95; unit capital 
costs of SO2 scrubbers in 151 – 152, 151; 
weak rule of law 32, 39, 107; 
wind energy development in 168 – 169, 
168; yearly university graduates in 
157 – 158, 157
“China Price, The” 172
Chinese Communist Party 13, 17, 25, 39, 
180, 183
Civil Code 181
Clean Air Act Amendments (CAAA, U.S.) 
12, 31, 49, 78, 79, 105, 184
climate change 26, 191
coal: consumption 10, 11, 84, 87, 90, 149; 
lower sulfur contents in 93; prices of 87, 
87; share in electricity generation 89 – 90
coal-fired power: annual growth of 98 – 99, 
98; decision scenarios for managers of 
plants 117; plants 95 – 100, 105 – 106, 
112 – 119, 123; and SO2 scrubber 
capacities 96 – 97, 97
competition, market entry and 169 – 171
compliance: costs 107, 109; monitoring 
119, 120 – 121; see also environmental 
compliance monitoring
compliance on SO2 scrubbers operation: 
noncompliance behaviors 112 – 115; SO2 
scrubber technologies 109 – 112
Comprehensive Plan on Ecological 
and Environmental Big Data 
Construction 124
Congleton, R. D. 6
continuous emissions monitoring systems 
(CEMSs) 107, 108, 113, 115, 122
corruption 6, 9
Cultural Revolution (1966–1976) 25
Darwin, C. 83
decentralization 27 – 28; in economic 
reform 36; fiscal revenue and 
expenditure 33 – 39; goal attainment 
61 – 65, 183; of governmental affairs 35, 
40; human resources 32 – 33; personnel 
management 39 – 40; of policy making 
31 – 32, 42, 82 – 83, 183 – 184
Decisions on Realizing Scientific View 
of Development and Strengthening 
Environmental Protection (2005) 46
deforestation 6, 7
democracy: and environment 5 – 7; and 
political will 17; and rule of law 13, 
179 – 180
Deng Xiaoping 188 – 189
Department of Organization of the Chinese 
Communist Party 27, 39
disability-adjusted life years (DALYs) 3; 
in China, due to air and water pollution 
3, 3, 4; in India 4, 4; premature deaths 
and 3
division of labor, for policy making and 
implementation 28 – 30
domestic technology licensees, strategy of 
156 – 162, 158
“double randomness, one publicization” 
scheme 129, 132
eco-compensation policy 31
ecological civilization 23
economic development: and energy 
conservation 84; and environmental 
protection 25, 27, 28, 174; and 
environmental quality 7, 51; Five-Year 
Plans 43; and SO2 mitigation 174


Index  195
effluent emissions: and SO2 removal rates 
97, 118; standards 31 – 32, 51, 66, 79, 
83, 95 – 96, 180, 187
electricity generation: annual growth of 90, 
91; energy consumption/transition for 
10, 89, 89, 92; by fuels in China 89 – 90, 
90; provincial 52; SO2 scrubbers 95, 
112, 117; wind 173
electrostatic precipitator (ESP) 109
employment and population structures, in 
China 18 – 20, 19, 22
energy consumption: annual growth of 87, 
88; economic conditions and 18, 20, 78, 
84, 174; electrification of 10, 89, 89, 92; 
and energy efficiency 86, 86; by fuel 87, 
88; reduction of 43
energy intensity goal 43, 78, 84, 92
energy transition effect 85, 92
Engineering, Procurement, and 
Construction (EPC) project, in Hong 
Kong 163
environment: and democracy 5 – 7; income 
and 6
environmental campaigns 135 – 136
environmental capacity 47, 51
environmental compliance monitoring 
105, 108 – 109, 119 – 136; building 
screening system with big data 123 – 125; 
comparing diagnosing and screening 
systems 125 – 135; conceptual model 
of 121; model construction 120 – 122; 
resilience of screening and diagnosing 
systems 135 – 136; strengthening 
conventional diagnosing system 122 – 123
environmental crises, in China 1 – 5, 180
environmental enforcement 27, 119
environmental governance 8; centralized/
decentralized personnel management 
39 – 40; evolution of environmental 
administration 25 – 27; for implementing 
political will 25 – 40; policy making 
and implementation 28 – 39; see also 
environmental protection
environmental impact assessment (EIA) 
reports 64, 94 – 95
environmental industry under goal-
centered SO2 mitigation path: 
international competitiveness of China’s 
SO2 scrubber industry 171 – 174; market 
entry and competition 169 – 171
Environmental Kuznets Curve 6, 7
Environmental Performance Index 1, 2
environmental policies 23
environmental protection: 11th Five-Year 
Plan for 43, 45; administration 27; 
authority of 25; as Basic National Policy 
25; as budgetary item 36, 37, 38; chain 
of command for 27 – 28; in China 25; 
economic development and 25, 27, 28, 
174; economic growth and 22, 23; goals 
on 43, 174; implementing 28; importance 
in new ideology establishment 46; 
personnel at governmental levels 28 – 29, 
29; political will for 17 – 24; prioritized 
42 – 72; provincial personnel 30; 
recognized as governmental affair 25, 
26; regulations on 31; SARS and 20 – 22; 
share in governmental expenditures 
36 – 38, 38; south–north water diversion 
project 27; strategies on 65; tax law 83; 
urban air quality and 23
Environmental Protection Agency (EPA), 
U.S. 113
environmental protection bureaus (EPBs) 
27 – 28
Environmental Protection Law 31
environmental quality 7, 47, 51, 65, 191
European Union Emission Trading Scheme 
119, 124
expenditures/revenue, of central and local 
governments 33 – 35, 33
financial sector 18
First National Conference on 
Environmental Protection (1973) 25
fiscal revenue and expenditure 33 – 39, 33
Five-Year Plans, goals in 42 – 44; see also 
individual plans
flue gas desulfurization see SO2 scrubbers
fluidized bed combustion (FBC) 93
foreign affairs and national defense 36
foreign technology licensors, strategy of 
162 – 165
fossil-fuel-fired power plants 65
fossil fuels 87, 89
fractions of sulfur retained in ash 93, 93
GDP (gross domestic product): capital 
investment and 64; in China, South 
Korea, Japan and US 7 – 8, 7; 
governmental revenue and 33 – 35, 33; 
growth rates of 18, 19, 22, 43, 85; R&D 
expenditures and 161
Genghis Khan 186
Gerlagh, R. 6
Global Burden of Disease study: China’s 
premature deaths due to air and water 
pollution in 1, 2; DALYs in China due 
to air and water pollution 3, 3
global financial crisis of 2008 20


196  Index
goal(s): in China’s Five-Year Plans 
42 – 44; in environmental protection 
43; evolution 65 – 72; implementation 
49 – 50; types of 65
goal attainment 43, 183; criteria for 
61 – 63; incentives for 63 – 65, 184
goal-centered governance: alternative 
governance models 179 – 181; 
applicability 185 – 188; characteristics 
187 – 188; in China 80 – 84; comparison 
with other theories 188 – 189; features 
184 – 185; illustration of model 
183; implications 190 – 191; inter-
goal coordination under 174 – 175; 
organization mechanisms 182 – 184
goal-centered policy implementation and 
supply 105 – 109; enabling 80 – 81; 
policy evolution by implementation 
selection 81 – 84
goal-centered SO2 mitigation path 
149 – 154; environmental industry under 
169 – 174; technology licensing under 
155 – 169, 158
goal distribution 43; from central to 
provincial governments 50 – 57; 
correlation coefficients of key factors 
for provinces 52, 53 – 54; provincial goal 
57, 61; from provincial to municipality 
governments 58 – 61; regression results 
to provinces/municipalities 56, 59
goal setting 43, 183; methods of 46 – 49; 
setting up national goal 44 – 46; in social 
psychology 188
governance indicators 8 – 9, 8
governmental revenue and expenditure to 
GDP ratios 33 – 35, 33
government effectiveness 9
grain storage 36
Great West Development 52
greenhouse gas concentrations, 
stabilization of 42, 50, 191
groundwater pollution, under Ministry of 
Land and Resources 26 – 27
Guangdong Province, distributing goals to 
municipality 59 – 60, 59
Guatemala 7
gypsum 111
Hainan Province 52
Harrington, W. 107
health care 36
Hebei Province 23, 59, 59, 68
Henan Province 116, 118
household air pollution from solid fuels 
1 – 4, 2, 3
Hu Jintao 18, 20, 21, 46
Human Environment, UN Conference on 
(1972) 25
human resources and fiscal expenditures 
32 – 39
IEA (International Energy Agency) report 
107
income and environment 6
incrementalism 189
India: ambient PM pollution in 4; energy 
consumption and electrification rate 89; 
governance indicators of 8 – 9, 8; market 
for technology 169; polity democracy 
index for 5, 6
indoor air pollution 1, 3
industrial and residential sectors 10
Insigma Technology 164
inspection 29 – 30, 29, 33
inter-goal coordination under goal-
centered governance 174 – 175
international competitiveness of China’s 
SO2 scrubber industry 171 – 174
International Monetary Fund 8
IPE (Institute of Public & Environmental 
Affairs) 21 – 22
Japan: economic growth in 85, 85; GDP in 7
Jiangsu Province 48, 59, 60, 106, 112, 116, 
153
Jiang Zemin 18, 44
Jiulong Electric 164 – 165
job creation 18 – 20, 22
Kenya 7
Kyoto Protocol 52
Law of Atmospheric Pollution Prevention 
and Control 31
Law of Environmental Protection 31, 47
Law of Standardization 47, 96
Law of Water Pollution Prevention and 
Control 31
Levitt, S. D. 121
Li Keqiang 17, 18, 22
limestone 109 – 112
liquid-to-gas ratio (L/G ratio) 110
local governments: achieving top-down 
goals 82; budget balance 34, 35, 36; 
environmental agencies in 28, 32; in 
era of Reform and Open-up 28; goal 
distribution 50 – 61; governmental 
revenue and expenditure to GDP 
ratios by 33 – 35, 33; implementing 
environmental policies 27; mobilization 


Index  197
of 43, 82; at provincial/municipality 
levels 32; responsibility for 
environmental quality 47, 65 – 66; shares 
of expenditures (2018) 36 – 38, 37 – 38
Locke, E. A. 42
major pollutants 77 – 78
Management Methods of Environmental 
Statistics 62
market: entry and competition 169 – 171; 
-oriented economic reforms 43; 
speculation 181; state and 26, 39, 174
Midlarsky, M. I. 6
Ming dynasty (1368–1644) 186
Ministry of Agriculture 27
Ministry of Ecology and Environment 
(MEE) 26 – 27, 28, 31, 44, 64, 77, 81, 184
Ministry of Environmental Protection 
(MEP) 26, 64, 67, 96, 106, 124, 163
Ministry of Land and Resources 26 – 27
Ministry of Water Resources 27
mitigation effect 85, 92
monitoring 29 – 30, 29, 33
National Acid Precipitation Assessment 
Program 49
National Aeronautical and Space 
Administration 124
National Development and Reform 
Commission (NDRC) 26, 45, 96
National Energy Administration 96, 169
National Environmental Protection 
Administration 47
National Party’s Congress 44
National People’s Congress 13, 31, 32, 44, 
45, 77
Neumayer, E. 6
noncompliance, reversing: environmental 
compliance monitoring 119 – 136; 
penalty 115 – 119
noncompliance behaviors, on SO2 
scrubbers operation 112 – 115
nonfossil fuels 90
nongovernmental organizations (NGOs), 
in China 17
nonhydro renewables 89
non-power-sector emissions 51
nonstate firms 156
“Not Invented Here” syndrome 159
Obama, B. 42
ocean environment, under State Oceanic 
Administration 27
oil and natural gas 87, 89
Open-up policy 9, 25
organization mechanisms, of goal-centered 
governance 182 – 184
Outline of the National 11th Five-
Year Plan on Economic and Social 
Development, The 45, 46, 77 – 78
ozone pollution 69, 70, 71
patents on environmental technology 
161 – 162, 162
Payne, R. A. 17
Pellegrini, L. 6
penalties for noncompliance 108, 115 – 119
People’s Republic of China; see China
personnel management, centralized/
decentralized 39 – 40
policing strategies 121
policy making and implementation: 
challenges in 77 – 80; compliance on 
operation of SO2 scrubbers 109 – 115; 
decentralized 31 – 39, 183; division 
of labor for 28 – 30; environmental 
compliance monitoring 119 – 136; goal-
centered 105 – 109; lower barriers 81 – 82; 
overcentralized 187; penalty 115 – 119; 
reversing noncompliance 115 – 136
political stability and absence of violence/
terrorism 9
political will 6; centralized 17 – 18; 
economy/jobs/environment (1998–2002) 
18 – 20, 19; for environmental protection 
17 – 24; SARS and prioritization of 
environmental protection (2003–2012) 
20 – 22; sustainability of (2013–present) 
22 – 24
pollution: abatement costs 120; health 
impact, measurement of 3; mitigation 
149; ozone 69, 70, 71; see also air and 
water pollution; ambient particulate 
matter (PM) pollution
power sector: shares of coal consumption 
and SO2 emissions 10, 11 – 12, 12, 59, 
60, 90; technological factors for effluent 
SO2 emissions in 92 – 95, 97
premature deaths: DALYs and 3; reduction 
and causes of 1, 2
provincial governments, on policy making 28
public: in democracy 6; health, goal for 
protecting 65
pulverized coal (PC) combustion 93
Qin dynasty (221–207 BCE) 185
Qinghai Province 52
Rebels (zao fan pai) 25
Red Guards (hong wei bin) 25


198  Index
Regional Supervision Bureaus 30
regulation, economic theory of 189
rent-dissipation effect 162
research and development (R&D) 
expenditures 161, 161
revenue: and expenditures, of central and 
local governments 33 – 35, 33; effect 162
Ricardo, D. 108
rule-based environmental governance 
180 – 181
SARS and environmental protection 
(2003–2012) 20 – 22
science and technology 36
Scientific View of Development 21, 63
sectoral employment changes and GDP 
growth rates, across China 18, 19
Shanghai: revenue–expenditure gap for 35; 
SO2 emissions 48, 52, 55
Shang Yang 186
Shanxi Province, distributing goals to 
municipality 59, 60 – 61
Shenzhen, AQI in 71, 72
Shijiazhuang: daily O3 concentrations in 
69, 70, 71; daily PM2.5 concentrations in 
68 – 69, 69; daily SO2 concentrations in 
68, 68; monthly average AQI in 
69, 70
Singapore, polity democracy index for 5, 6
SO2 (sulfur dioxide) emissions: in 9th 
Five-Year Plan 18, 62, 90, 92; in China 
9 – 10, 10 – 12, 12; controlling 48; daily 
SO2 concentrations, in Shijiazhuang 
68, 68; decomposition of 90, 91; 
designated intensity, in coal power 
plants 58, 58; economic growth and 84; 
emission mitigation goals of 67 – 69, 
83; environmental capacity for 47; goal 
implementation 50, 58, 63; intensity of 
electricity generation 96; key factors 
for 84 – 92; mitigation of 10, 12, 20, 
43, 51, 65, 67 – 68, 72, 77 – 78, 84, 
181 – 182, 185, 191; policy scope for 
achieving mitigation goals 84 – 100; 
in power and nonpower categories 51, 
63; reduction goal of 43 – 44, 45 – 46, 
48 – 49, 59 – 60; regulations 44, 66, 118; 
removal efficiencies/rates 51, 57, 94, 
118; by sector 10, 11; setting up goals 
47; technical measures for 95 – 100; 
technological factors for 92 – 95; 
underestimation of 112; in United States 
12, 12
SO2 mitigation path: and economic 
development 174; environmental 
industry under goal-centered 169 – 174; 
goal-centered 149 – 154; model 
projection of 154; technology licensing 
under goal-centered 155 – 169, 158
SO2 scrubbers: BOT contracts for 173; 
capacities 95, 97 – 99, 97 – 100, 171; 
capital costs of 112, 115, 151 – 152, 151; 
categories 99; coal-fired power and 79, 
97, 97, 98, 99, 106; compliance costs of 
109; compliance on operation 109 – 115; 
data, in China’s coal-fired power plants 
114; deployment and operation of 
149 – 151, 150; designing 93; economies 
of scale and 111 – 112; effluent discharge 
fee 115 – 116; electricity-consuming 
components of 111; firms 156; 
geographic distribution of 99; goals 
and policies in compliance decisions 
on operation 119; installation 94 – 96, 
95, 106 – 107, 151, 159, 171, 185; 
international competitiveness of industry 
171 – 174; noncompliance behaviors on 
operation 112 – 115; nonoperation of 
116 – 117; O&M costs of 111 – 112, 115, 
151, 152; operation in Jiangsu Province 
106, 106; planning 51, 55; product of 
111; reduction of emissions through 
51, 85, 96, 105; technologies 100, 
101, 109 – 112, 156; see also reversing 
noncompliance
“Socialistic Thoughts with Chinese 
Characteristics in the Xi Jinping Era” 23
social psychology 42
social welfare system 181
South Korea: GDP in 7; polity democracy 
index for 5, 6
south–north water diversion project’s 
environmental protection 27
Standing Committee of the Political 
Bureaus 21
state and market 26, 39, 174
State Council, 1998 reform of 26
State Environmental Protection 
Administration (SEPA) 26, 44, 45 – 46, 
50, 58, 64, 78, 98
State Environmental Protection Agency 
(1984) 25, 26
State Oceanic Administration 27
state-owned enterprises/firms 18, 156
Steinfeld, E. S. 112, 113
Stigler, J. 189


Index  199
Suggestions on Designing the National 
11th Five-Year Plan, The 45, 46
sulfur contents 94, 111; coal consumption 
and 63, 78; control of 85; distribution in 
coal power plants 93 – 94, 95; see also 
SO2 entries
suspension policy 64
sustainability of environmental political 
will (2013–present) 22 – 24
Tang dynasty (618–907) 186
tax compliance 121
technology licensing under goal-centered 
SO2 mitigation path 155 – 169; criteria 
of effective market design 165 – 167; 
strategy of domestic technology 
licensees 156 – 162, 158; strategy of 
foreign technology licensors 162 – 165; 
technology market emerging in China, 
reasons for 165 – 169
thermal contents of coal 111
“three representativeness” 20
Tibet: governmental revenue/expenditures 
35; SO2 emissions 48, 52
top-down goal distribution 49 – 61
Total Emission Control regime 66
unemployment 20
UNFCCC (United Nations Framework 
Convention on Climate Change) 42, 50
United States: average prices of wind 
turbines in 172 – 173, 172; CEMSs cost 
in 113; Clean Air Act Amendments 
(1990) 12, 31, 49, 78, 79, 105, 184; 
coal consumption 12; deployment and 
operation of SO2 scrubbers in 149 – 150, 
150; economic growth in 85, 85; energy 
consumption and electrification rate 89; 
Energy Information Administration 152; 
Environmental Protection Agency (EPA) 
113; firms in 170, 171; GDP in 7; 
governance indicators of 8 – 9, 8; patents 
on environmental technology 161 – 162, 
162; polity democracy index for 5, 6; 
power sector shares of coal consumption 
and SO2 emissions 10, 11; SO2 emissions/
intensities in 12, 12, 49; unit capital costs 
of SO2 scrubbers in 151 – 152, 151; wind 
energy development in 168 – 169, 168
university-established firms 156
unsafe water/sanitation/handwashing 1, 
2, 3, 4
“Upgrading and Retrofitting Action Plan 
for Energy Conservation and Pollution 
Mitigation in the Coal-Fired Power 
Sector” policy 96
veto 64
Wang Xinfang 45 – 46
Wanli (Emperor) 186
water: consumption 112; environment 
management 27; pollution, health 
impacts of 3; see also air and water 
pollution
Wen Jiabao 18, 20, 21, 46
wind: energy development 168 – 169, 168, 
173; and solar energy 89; turbines, 
average prices of 172 – 173, 172
Winslow, M. 6
World Bank 8, 31, 180
World Trade Organization in 2001 20
Xie Zhenghua 64
Xi Jinping 17, 18, 22
Yelv Chucai 186
Zhejiang Province 62
Zhongnanhai (Central and Southern 
Seas) 186
Zhu Rongji 18

Government policy: meaning, types, manifestations, theories, and policy cycles
Article  in  Insights into Regional Development · June 2023
DOI: 10.9770/IRD.2023.5.2(6)
CITATIONS
6
READS
10,345
4 authors, including:
Adetayo Olaniyi Adeniran
Federal University of Technology, Akure
98 PUBLICATIONS   580 CITATIONS   
SEE PROFILE
Adedayo Adeniran
University of Ibadan
7 PUBLICATIONS   27 CITATIONS   
SEE PROFILE
 
 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195(online) https://jssidoi.org/ird/ 
 
 
 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
         
       
            
 
            Publisher 
    http://jssidoi.org/esc/home 
 
 
 
 
  
 
83 
 
GOVERNMENT POLICY: MEANING, TYPES, MANIFESTATIONS, THEORIES, AND POLICY 
CYCLES 
 
Adetayo Olaniyi Adeniran 1, Joseph Mosunmola Muraina 2, Joseph Olanrewaju Ilugbami 3, 
Adedayo Ayomide Adeniran 4  
 
1Department of Logistics and Transport Technology, Federal University of Technology Akure, Nigeria 
2Department of Geography and Planning Science, Ekiti State University, Ekiti, Nigeria 
3Rufus Giwa Polytechnic-Owo Rector Office, Ondo State, Nigeria 
4Department of Geography and Planning, University of Ibadan, Nigeria 
 
E-mails:adeniranao@futa.edu.ng1; jmosun07@gmail.com2; ilugbamijoseph@gmail.com3; ddone2@gmail.com4 
 
Received 10 March 2023; accepted 10 June 2023; published 30 June 2023 
 
 
Abstract. In any democracy, it is strongly advised that effective policies be created since they are crucial to how democracies operate. 
Government policy definitions and categories were widened. Government policy types were discussed concerning the sectoral groups 
comprising each given government. This is important because a policy’s or its objective elements frequently suggest different meanings for 
different stakeholders. Policymaking is a process impacted by socio-political and other factors and is not a governmental function. Thus, 
there is a need to comprehend the theoretical underpinnings on which government policymaking and its execution may be evaluated and 
characterized. According to the elite/mass hypothesis, there are two groups in society: those who occupy positions of power and those who 
do not. Government policy is more influenced by those with access to knowledge and influence. It is a remarkable characteristic of group 
theory which is ideally in line with the legislative because the legislatures are where the voices of the people are expressed. Governmental 
institutions and government policy are closely related, claims institutional theory. The rational choice theory may need to be more accurate 
since participants in government policy must have access to all information to make informed judgments. The systems theory offers a more 
straightforward method for categorizing and comprehending the contributions and interrelationships made by institutions and policy 
players, including the function played by the external environment in policy formulation. Lastly, since democracy is a system of 
governance built on extensive public engagement, any ideology that supports any type of citizen participation (particularly in a democracy) 
should be endorsed by both politicians and public officeholders. 
 
Keywords: Government policy; Policy manifestations; Policy execution; Policy underpinnings; Policy context and consequences 
 
Reference to this paper should be made as follows: Adeniran, A.O., Muraina, J.M., Ilugbami, J.O., Adeniran, A.A. (2023). Government 
policy: meaning, types, manifestations, theories, and policy cycles. Insights into Regional Development, 5(2), 83-99. 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
JEL Classifications: J58, J68, J78  
 
Additional discipline: Government policy 
 
 
 
 
 
 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
84 
 
1. Introduction 
 
Each democracy needs sound policies. Additionally, in a democracy, the proper application of those policies is 
crucial. According to Delamaza (2015), democracy is a kind of government. Under a democratic political 
administration, among the issues facing governance is erecting a foundation that enhances the practice of 
democracy without undermining the freedom to embark on purpose and functions and to ensure that social 
demands and conflicts arising from various interest groups and civil societies are tackled is one of their tasks 
(Dunne, 2021; Forcher-Mayr and Mahlknecht, 2020). 
 
Given that both government policies and how they are carried out may strengthen a democracy, there is a need to 
define government policy more broadly and the ingredients of government policy execution (GPE) (Adeniran, 
2016; Delamaza and Palma, 2022; Matuku-Mphahlele and Zandamela, 2022). These terminologies are essential 
due to the elements involved in the execution of government policy. Hence, the word government policy 
execution is a subset of the primary term government policy. Government policy may be described as a cycle or 
process with several steps to be taken before achieving a policy’s goals. Typically, there are four or five stages:  
a) Stage for issue/ problem identification; 
b) Stage for setting agenda; 
c) Stage for policy formulation or policymaking; 
d) Stage for policy execution; and  
e) Stage for policy evaluation. 
 
Policy phases will be significantly influenced by the particular technique employed (Zeb-un et al., 2021). 
Government policy is, first and foremost, a persuasive art, as Deygers and Vanbuel (2022) claimed. It is so named 
because it calls for the selection, enactment of legislation, and consultation of all relevant parties (Kofele-Kale, 
2006; Nunes et al., 2019). According to Oyadiran and Akintola (2014), one objective of government policy is to 
guarantee that persons responsible for carrying out significant decisions in society, regardless of their position, are 
well-trained. This opinion was also agreed upon by Myrczik et al. (2022), De-Marchi, Lucertini and Tsoukiàs 
(2014), and Ozturk (2015). 
 
According to Galli (2015), government policy should be viewed as both a declaration of goals and a negotiated 
outcome resulting from the execution process. One of government policy’s most distinguishing features is how 
unstable and changeable it is (Deygers and Vanbuel, 2022). The assertion that proposed or envisioned government 
policies lacks any evident beginning or end is maintained in the study of Ashmore et al. (2020), which noted that 
they should be understood as analogous to seashells or jelly. It flows almost circularly at times. Myrczik et al. 
(2022) assert that when the policy is discussed, it implies addressing pertinent issues germane to human existence. 
 
Falk and Tally (2016) identified the features of government policy, such as the intended direction that the 
legislator would want to guide the public, including the description of how the country’s resources are to be used 
(Díaz-Llamas et al., 2023). It was also revealed by Oyadiran and Akintola (2014) that several variables might 
influence the overall government policy process. These include the legislators in charge, noting what the 
Constitution stands for. The issues that need to be resolved should be known to politicians or bureaucrats Koelble 
and Siddle (2014). 
 
Also, a significant portion of those involved in the government policy process are local (government excluded). 
Consequently, it is crucial to get in touch with these influential individuals who know the community's situation, 
their challenges, and the issues that need to be fixed. The act of fashioning, enacting, monitoring, reviewing, or 
revising government policies is covered by Imenda (2014). Nokele (2022) argues that because it is crucial to the 
efficacy and reach of government policies, its execution should be the primary emphasis of the whole process. 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
85 
 
Macheridis and Paulsson (2019) noted that government policies are centred on presumptions about what 
governments can do and what the effects of those actions would be since they would otherwise be the product of 
political activity and would, as a result, have political ramifications (Mellaard and van Meijl, 2017). Maggetti and 
Gilardi (2016) assert that it is uncommon to get a thorough explanation of the assumptions underlying 
government policy as a theory or model, much alone the context in which those assumptions must be employed or 
understood. But, as with every procedure, an idea or model is always presupposed (McCann and Ward, 2013). 
 
Every story has two sides, and the government policy level confirms this truism. Government policy is two-
dimensional or contains two storylines; given that politics and administration are a component of it, it has a two-
dimensional structure. Creese, Dutton and Esteve-Gonzalez (2021) refer to this reality as the more significant 
number of pertinent legislative and administrative operations. Knill and Tosum’s viewpoints on government 
policy may be contrasted to show how interdependent politics and the administrative side are. Government policy 
and politics should adhere to the same course (Molossi et al., 2023). 
 
The role that legislators play in deciding the resource utilization of a country in the government mentioned above 
policy is regarded as the political side of the government policy process (Gray, 2018). On the other hand, the 
administrative side of the government policy process focuses on the executive and their actions to realize the 
stated objectives established by the government (Mellaard and van Meijl, 2017). The administrative side of the 
public process is responsible for ensuring that the adopted policy will persist throughout time, according to 
Deygers and Vanbuel (2022). Policymakers, administrators, and bureaucrats should encourage all significant 
stakeholders of the necessity of a specific policy and the reasons for that requirement for that execution to take 
place (Purtle et al., 2023; Mellaard and van Meijl, 2017). 
 
2. Literature Review 
 
2.1. Manifestations of Government Policy 
Everyday life is a manifestation of government policy. Also, it starts in casual conversations when regular people 
talk about things like how to improve government policy. As stated in the introduction chapter, creating 
government policies is a complex, multi-layered process (Mellaard and van Meijl, 2017). For the creation and 
execution of government policy, two guiding concepts (or significant areas of study) are essential. Public 
administration and political sciences/studies fall under this category. According to Andrews-Speed (2021), 
government policy encompasses several political science subfields. Implementing government policies, which 
come from the political (or policymaking) facets of government and are backed and endorsed by political 
administrators, is the priority over public administration’s primary goal (Uddin et al., 2023). 
 
Wilson’s dualism (quoted by Guidi et al., 2020) contends that politics and administration cannot be divided into 
distinct roles when determining government policy from both the structural and functional perspectives. There, 
the line thins out to the consistency of a spider’s web thread. According to Simeon (1976), institutions and 
practices that are exposed in and through economic, social, and political dynamics shape government policy. 
Government policy can also result from issue articulation (acknowledging a policy challenge), finding 
alternatives, and the political processes (Crabolu, Font and Eker, 2023).  
 
According to Bertram (2020) the focus of political studies on government policy has been around for a while. 
Mellaard and van Meijl (2017) contend that the academic study of the government policy process is a part of 
political studies/sciences since politics deals with who gets what, when, and how. Political science may be 
necessary to government policy issues while maintaining its dedication to scientific investigation (Mellaard and 
van Meijl, 2017; Deygers and Vanbuel, 2022; Fischer et al., 2015). Politicians, pressure organizations, and 
‘passive beneficiaries of policy’ are only a few stakeholders engaged in the government policy process (Jiang, 
2018). 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
86 
 
 
2.2. Underpinnings of Government Policy 
Government policy is often regarded as being first and primarily a course of action (Makhetha, 2015). This course 
of action must demonstrate logical decision-making, as doing so will lead to responsible behaviour (Daniell, 
2014). Wu (2022) define government policy as the process or series of actions taken by the government to solve a 
particular societal issue that was originally recognized. According to Guidi et al. (2020), the people whose lives 
will eventually be impacted by the outcomes of policy action are represented in the specialized policy subsystems 
where government policy is developed, implemented, and evaluated.  
 
According to Paulsson and Macheridis (2022), who also concurs with Makhetha (2015) and Wu (2022), 
government policies are the result of a combination of systematic forces, political processes, institutional 
influences, rivalry among groups, elite preferences, belief in or advocacy of change through small steps, and 
rational planning (Fischer et al., 2015; Deygers and Vanbuel, 2022). Whatever decisions the government agrees 
on will fall under this (Kharel and Kharel, 2020; Jakonen and Sokka, 2022). Simeon (1976) concluded that 
policies are the climax of a complicated negotiation process and the outcome of several modest judgments made 
by decision-makers. Yet, Simeon (1976) maintains that ideology is at play both in the formulation of policies and 
during the policymaking process, suggesting that government policies indeed reflect ideology (or have a symbolic 
repertoire; Steven, 2021; Molossi et al., 2023). 
 
The socioeconomic circumstances present in a particular geographic area that the government policy must address 
impact how the framework of government policies is developed claims Kharel and Kharel (2020). However, 
several factors, such as institutional frameworks, a country’s party system, or the overall relationship between the 
government and the populace, can affect the process of formulating government policy (Díaz-Llamas et al., 2023). 
Government policy also incorporates a society's dominant ideas, dogmas, and beliefs (Simeon, 1976). Because of 
this, these components offer a framework for the underlying assumptions and arrangements that permit the 
examination of policies (Simeon, 1976). 
 
Recognizing social issues and how societies choose to handle and solve them are essential elements of 
government policy, according to Parsons (2002) and Steinert (2016). Government facilitates reducing or removing 
these issues that society has identified (Parsons, 2002; Crabolu, Font and Eker, 2023). Guidi et al. (2020) assert 
that two features or functions, namely structural terms and/or functional words, can be used to conceptualize 
government policy. The structural component of government policy includes the interactions that may occur 
between the governments' policy players in the setting of the several specialized areas of the subject (Guidi et al., 
2020). The many policy types considerably influence how government policy is framed (Crabolu, Font and Eker, 
2023).  
 
2.3. Types of Government Policy 
Government policies can be created in several styles and/or types to address the need on the policy agenda. Lowi 
refers to this classification of policies as a policy categorization (1972). As a specific policy type would be 
associated with a variety of politics, categorizing policies is essential for studying politics (Oyadiran and 
Akintola, 2014; Aritz et al., 2017). Hence, a politically appropriate policy classification has been developed 
(Oyadiran and Akintola, 2014). The classification of policies must, however, be founded on intellectual and 
theoretical considerations that have an influence on actual political situations (Oyadiran and Akintola, 2014). The 
policy categorization aims to ensure that it supports the study of politics while avoiding omitting the public 
administration component or having a detrimental impact on the political environment as a whole (Oyadiran and 
Akintola, 2014). 
 
The aim of government policy classifications or taxonomies5, according to Bertram, Maleki and  Karsten (2019), 
is to comprehend the basic contrasts between policies and the political settings that influence the various types of 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
87 
 
policies in place. It is simpler to express the typifications that role-players typically utilize to characterize 
government policies when approaches are categorized, according to Aritz et al. (2017). This suggests that using 
policy taxonomies makes it possible to accurately describe government policies (Aritz et al., 2017). Sol (2023) 
claims that employing policy taxonomies may assist in determining the scope and presentation of a policy. 
 
According to Simeon (1976), policy taxonomies offer the chance to consider the amount of coercion and the 
equilibrium between individual and collective activities leisurely. Simeon (1976) thinks Lowi’s (1972) proposed 
policy taxonomies are essential and fundamental for political science students. According to Munzhedzi (2020), 
there are four types of government policies, or policy taxonomies: distributive, redistributive, regulatory, and 
component government policies. Some taxonomies or classifications are considered to be governmental functions. 
According to Nico (2015), these policy categories may be used to pinpoint the specific effects of a policy, which 
might promote political discourse about how decisions are made and how to execute policies. 
 
Also, the sectoral categories or clusters should serve as the foundation for policy classifications (Ahmad et al., 
2021). The terms types and categories of policies were used interchangeably throughout the study. For instance, 
there may be a collection of regulatory or protective policies. 
  
2.3.1. Distributive government policy 
Guidi Guardiancich and Levi-Faur (2020) claim that the primary objective of distributive policies is issue-solving; 
as a result, they typically function in the most supportive political climate. The strong clientele, knowledge, 
leadership, and coherence characterize the context in which distributive policies are carried out (Rakšnys and 
Valickas, 2023). It involves acting to address issues facing the general population (Rakšnys and Valickas, 2023). 
Significantly, distributive policies may also be described as dealing with how additional resources, expenses, and 
advantages from the government are distributed to specific population demography (Díaz-Llamas et al., 2023). 
Bertram, Maleki and Karsten (2019) revealed that distributive strategies address Lasswell’s (1936) maxim of who 
receives what, when, and how. 
 
According to Bertram, Maleki and Karsten (2019), distributive policies use general public funds (instead of user 
fees) to help a particular segment of a social group without considering resource limitations or financial 
constraints (Rakšnys and Valickas, 2023; Díaz-Llamas et al., 2023). As shown in election manifestos, when 
different political parties seek voters to approve of the resources and services they can deliver to them if they are 
elected to power (or held in power), the constituencies of elected politicians also benefit from distributive policies 
(Kraft and Furlong, 2013). 
 
2.3.2. Redistributive government policy 
Allocative government policies, sometimes referred to as redistributive government policies, deal with necessities 
like the funding of the welfare system, health care system, and education system (Ahmad et al., 2021). Guidi 
Guardiancich and Levi-Faur (2020) claim that redistributive policies occur when the government levies taxes on 
one group of people to benefit another. These resources are distributed between the wealthy and the socially 
disadvantageous and destitute groups (Díaz-Llamas et al., 2023). A redistributive strategy can be implemented 
despite ideological cleavages, following Guidi et al. (2020). 
 
Concerning the aforementioned, Guidi Guardiancich and Levi-Faur (2020) assert that direct taxation and the 
transfer of resources from one socioeconomic group to another lead to the emergence of a distinctive 
characteristic that distinguishes distributive and redistributive policies from one another. A dispute is this trait 
(Jutta, 2016). Redistributive policies are exceedingly political, difficult, unfavourable, and polarizing to design 
and implement, which causes this conflict. They cause disputes that polarize the population along party lines 
(Rakšnys and Valickas, 2023). Redistributive programs face this challenge since one group gains at the expense 
of another (Yanow, 2015). The discussion around distributive government policies is heightened because they 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
88 
 
explicitly allude to an ideology or a class war. According to Donnelly (2015), the disadvantage of redistributive 
policy is that the government typically lacks the means to implement such a program. 
 
2.3.3. Regulatory government policy 
Regulatory policies, according to Munzhedzi (2020), typically address the need for policies relating to 
transportation, infrastructure, health, and other regulations and standards, or they prohibit people from acting in 
certain ways, such as selling illegal goods like dangerous drugs, participating in unfair competition in the market 
(Rakšnys and Valickas, 2023). According to Anyebe (2018), regulatory policies are laws carried out by 
government agencies without any interference or money inducement. 
 
A regulation policy can be a form of competitive regulation to regulate individual industries and their activities. It 
can also be a protective regulation meant to protect the general public. Bertram, Maleki and Karsten (2019) 
contend that regulatory approaches are questionable because they let the government meddle in private enterprises 
and people’s daily lives. Another disadvantage of regulatory government policies, according to Creese, Dutton 
and Esteve-Gonzalez (2021), is that they significantly impact how much money is spent and how much assistance 
from other social actors is needed. 
 
2.4. Constitution government policy 
Oyadiran and Akintola (2014) developed the component policy as a subset of constituent policy. Both the 
government and/or the nation as a whole are considered to be two constituents of government policy, according to 
Oyadiran and Akintola (2014) and Guidi Guardiancich and Levi-Faur (2020). According to Meier’s additional 
definition from 2007, constituent policies aim to advance the interests of the nation-state and the broader public. 
Constituency policies are portrayed by Guidi Guardiancich and Levi-Faur (2020) as being exceedingly detailed, 
meticulous, and in charge of significant initiatives. Meier’s (2000) notion of component policies may be used to 
depict the presidential department where policies are executed, monitored, and coordinated. Constituent policies 
also cover governmental operations, including defence and foreign policy (Rakšnys and Valickas, 2023). 
 
The present democratic society can be classified under constituent policies because of their method of operation 
and provision for election laws (Yanow, 2015). According to Creese, Dutton and Esteve-Gonzalez (2021), there is 
a fact that constituent government policies only have an impact on the executive branch of government. As was 
said above, Oyadiran and Akintola (2014) identified the many kinds of government policies and found just four 
policy taxonomies. Not all government policies will fall within Lowi’s (1972) taxonomy of approaches, as 
(Rakšnys and Valickas, 2023) indicates. These policy taxonomies thus have the disadvantage of excluding 
alternative policies that might not fit the policy classification. A few new categories of approaches have been 
included in the classification of procedures since Lowi’s (1972) policy taxonomies were first introduced. 
 
The following section briefly discusses one more policy type that is mainly referred to as substantive government 
policy. 
 
2.5. Substantive government policy 
Government policies are crucial in a wide range of substantive sectors, according to Paulsson and Macheridis 
(2022). These substantive sectors include, but are not limited to, environmental issues, economic development, 
security, public service, international relations, primary education, social development and domestic affairs 
(Fischer et al., 2015). A substantive policy focuses on what the government should do (Simeon, 1976). A 
substantive policy may incorporate specific overarching goals (such as describing the anticipated results of the 
policy while it is being produced, for example) (Marie-Kim and Marie-Hélène, 2020). It might also consist of 
more concrete objectives the policy must achieve. Yudiatmaja et al. (2022) conclude that successful substantive 
solutions can resolve a policy issue. 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
89 
 
2.6. Government policy Execution 
The policy as it is carried out is the genuine policy of a government, according to Peters (2001). One of the main 
reasons why government policy execution is one of the most crucial stages in the whole government 
policymaking process is because it refers to the point at which a policy is implemented (Díaz-Llamas et al., 2023; 
Aguerre and Hernan, 2015). According to Peters (2001), one of the main issues with our current political systems 
is how government policies are carried out. Many behaviours in the administrative and political settings where 
government policy is being implemented are taken into account throughout the execution process, claims Hurel 
and Rocha (2018). According to Galli (2015), politics substantially influences every action or step taken during 
the cycle of policy execution, with both a macro and micro political context (Galli, 2015). 
 
The macro-political backdrop, which includes factors like legislation, economy, and what is happening or moving 
worldwide, is what Galli (2015) refers to as the external environment. On the other hand, according to Galli 
(2015), the micro-political context comprises things like the policy’s mission, the competencies needed, the 
organizational culture, and the external environment. GPE is rather challenging since several factors must be 
considered, some of which the implementers have influence over and others of which they do not. Falk and Tally 
(2016) argue that it is incorrect to assume that implementing government policy only entails putting previously 
developed procedures into action since there is more to it than that. Implementing government policy involves 
using important inherent information. 
 
Creese, Dutton and Esteve-Gonzalez (2021) assert that GPE bridges policymakers and policy addresses. The 
implementers aid this relationship. This GPE phase is essential because it makes it possible to execute the 
proposed or envisioned policy (Steven, 2021). This suggests that the result of the policy is transformed into its 
production. Aguerre and Hernan (2015) contend that policies and practices must be separated to understand the 
whole process of producing policies. As mentioned earlier, the role-players in charge of implementing the policies 
must thus not act entirely independently. As a result, they offer guidelines for applying already created and 
authorized policies. 
 
3. 
Theoretical Review 
 
Government policy theories are essential in the social, environmental, technological and engineering literature. 
These theories offer unique characteristics of political and human development. Among the theories of 
government policy are the political systems theory, group theory, institutional theory, rational choice theory, and 
the policy process model. 
 
Many of the previous and present policies are formulated and implemented because they are influenced by 
systemic variables, political processes, institutional influences, game-playing, incrementalism, interest group, 
rational planning, elite preferences, and interest group interests. These theories will pose further issues regarding 
government policy and the primary channels from which sound decisions are formed. The following section will 
cover these theories. 
 
3.1. The elite/mass theory 
The elite notion holds that a small elite group controls the bulk (Zeb-un et al., 2021). This idea works best in the 
countries of Africa. Because the interests and well-being of the elite are prioritized under this theory, elite 
viewpoints that diverge from those of the general public can affect the development of government policy (Zeb-un 
et al., 2021; Jutta, 2016). The elite thesis is based on the notion that because the general public is allegedly 
uninformed and indifferent, their opinions shouldn’t have any bearing on how government policy is formulated 
(Fischer et al., 2015). 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
90 
 
The elite notion holds that only a caste that is acknowledged throughout society should influence government 
policy (Kraft and Furlong, 2013). This elite caste includes members of the governing class, political parties, 
business executives, wealthy individuals, and educated segments of society (Jutta, 2016). One way that the elite 
ideology is implemented is by whom the most influence over how government policy is decided (Jutta, 2016). Not 
all elites have an outsized impact on shaping government policy. Each elite aims to have a significant effect on a 
specific niche market. For instance, business executives would want to weigh in on decisions regarding tax 
legislation and import and export laws. The governing class, however, would like to have a voice in how the 
general public is governed, how money is allocated, and how resources are utilized.  
 
It’s also conceivable that these two exclusive groups come into contact with one another and interact as they use 
their influence and power. Zeb-un et al. (2021) assert that public administrators’ perceived importance is 
influenced by the idea that they are members of the ruling class rather than citizens’ servants. This idea may be 
explained as a small elite making decisions that cascade down to an uneducated civil society (Fischer et al., 2015). 
Zeb-un et al. (2021) assert that political power influences these decisions and that the bureaucracy is necessary to 
carry them out. The idea holds that only a select group of experts possess the authority. 
 
3.2. Group theory  
Politics is characterized by the interaction of groups, and the group theory incorporates organized interest groups 
in the creation of government policy (Jutta, 2016). These actors are shown as tenacious voice-hearing warriors. In 
light of this, it is possible, to sum up group theory as a battle between the voices of organized interest groups. 
Group theory includes, for example, individuals working in the agriculture sector and companies producing 
music. Organizations should be allowed to make a major contribution and have a say in determining government 
policy. In order to dispute the abuse, poor administration, and fraudulent execution of policies, as well as hold 
those responsible accountable, people should be able to challenge laws that are thought to be illogical, unsuited, 
or ineffective for the intended purpose. 
 
Organizations ought to promote justice, transparency, the participation of the citizenry, and awareness in 
policymaking. The group theory is crucial and pertinent to government policy as a result. This is done so that 
organizations may play a big part in setting policy and assisting with enforcing previously approved or ratified 
legislation like the Constitution. Groups have an impact on government policy, whether it be a policy regarding 
environmental concerns or the welfare of the populous as a whole. This exemplifies how several interest groups 
from diverse socioeconomic domains may all voice their opinions on the policies they believe the government 
should adopt or reject and play a significant role in their creation. According to Zeb-un et al. (2021), group theory 
has some implications for political decisions. For instance, the dynamics of the cabinet are changed. 
 
The disadvantage of the group theory is that it rewards more organized groups, has more members, has access to 
resources, has political allegiances, is well-liked, and has built ties with decision-makers (Galli, 2015). The less 
fortunate members of society lack all of the aforementioned resources and are at the other extreme of the spectrum 
(Jutta, 2016). According to Bertram, Maleki and Karsten (2019), group theory is criticized by academics studying 
government policy for giving organized interest groups too much sway and leaving it up to them to decide policy 
(Oyadiran and Akintola, 2014). Government employees also seem to be left on the side of the road (Tacon and 
Hanson, 2011). 
 
Researchers in government policy believe that the degree of impact that organized interest groups have on 
policymaking tends to worsen the complexity and dynamic character of policymaking that is already present 
(Kraft and Furlong, 2013). It is also important to acknowledge that the elite/mass does have some roots in group 
theory (Carroll and Common, 2013). Although all groups (regardless of socioeconomic level or prominence) may 
be accommodated under the group theory, those with access to more resources are often the ones whose opinions 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
91 
 
are heard when policies are being developed (Chaudhary, 2018). Their voices tend to be aristocratic. The voices 
of those groups that lack access to the same resources as the privileged are so muffled. 
 
The term "extra influence" refers to the elite groups’ intrusion into group theory and the creation of government 
policy. This is characterized by Guidi Guardiancich and Levi-Faur (2020) as having a solid clientele, knowledge, 
and leadership. Also, this increases the pressure on public servants and policymakers, which tips the balances in 
their favour when deciding the course of government policy. 
 
3.3. Institutional-based theory  
This theory is often known as the "classical theory" since it is interpreted classically to study government policy 
(Zeb-un et al., 2021). It is not a coincidence that Minkman, van Buuren and Bekkers (2018) state that the 
institutional approach arose as awareness of the importance of enshrining government policy-making in the 
framework of institutions expanded. This implies that the government’s concerns about welfare should take 
precedence over other issues (Zeb-un et al., 2021). Institutional theory is deeply rooted in the formal and legal 
aspects of the governmental system (Díaz-Llamas et al., 2023). The institutional model’s purpose is to evaluate 
the structures that regulate how the government is structured, its legal power and the norms of behaviour it 
adheres to while making decisions (Dunne et al., 2021). The institutional theory focuses primarily on the public’s 
access to decision-making, government transparency, and, eventually, the separation of powers between the 
various levels of government (Zeb-un et al., 2021).  
 
The institutional theory rationally asserts that the structures and codes of conduct that regulate the government 
and its departments significantly impact the various types of policy processes that take place, as well as how role-
players in those processes will ultimately affect those processes (Kraft and Furlong, 2013). Political, economic, 
and sociological institutionalism are the three frameworks that institutional theory embraces (Minkman, van 
Buuren and Bekkers, 2018). Economic institutionalism stresses applying economic analysis to political 
institutions and government policy, whereas political institutionalism looks beyond the traditional forms of 
institutions to pay more attention to (Díaz-Llamas et al., 2023). The institutional theory is essential in ensuring 
government policies' legitimacy, applicability, and coerciveness (especially true of regulatory laws, which impose 
obligations on the general populace) (Díaz-Llamas et al., 2023). 
 
3.4. Rational choice theory 
This is a contemporary theory used in social sciences. The public choice theory is another name for the theory of 
rational choice (Cagnin, 2017). It has a strong economic foundation (Jutta, 2016). Generally, it uses complex 
mathematical modelling, which has only been moderately helpful in evaluating marginal behaviours in 
competitive circumstances and is typically seen throughout an election period (Ashmore et al., 2020). 
 
This theory is thoroughly developed and rigorous, and it could be used to address many government policy-
related issues (Kraft and Furlong, 2013) and used to conclude. Opponents of this theory claim that the decisions 
made based on rational choice are faulty, unrealistic and unworkable. 
 
Cagnin (2017) identified two distinct features of the rational choice theory. Its main focuses are methodological 
individualism and the assumption that people are reasonable. The sensible perspective contends that the ability to 
make decisions indicates a person’s capability for logical reasoning. On the other hand, Ashmore et al. (2020) 
argue that a broad account of human behaviour supports all rational choice theories. According to Ashmore et al. 
(2020), the basic hypothesis holds that individuals are complicated, flawed mortals who strive for perfection 
despite whatever challenges they may encounter. 
 
 
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
92 
 
3.5. Political systems theory 
This theory is the most complete among popular approaches (Kraft and Furlong, 2013). The theory aids 
government initiatives and institutions in transforming public inputs (such as environmental needs) into policy 
outputs (such as public opinion and pressure from interest groups) (Cagnin, 2017). The theory was designed to 
raise public awareness of policy issues and give the populace a platform to express grievances (Cagnin, 2017), 
allowing for the problems to be presented on the government’s policy agenda (Uminska-Woroniecka, 2022). 
Moreover, it represents the wider, shared socioeconomic, cultural, and political framework that serves as the 
foundation for decisions on politics and policy (Jutta, 2016). According to Bertram, Maleki and Karsten (2019), 
the language employed in political and policy studies has expanded as a result of the systems theory. 
 
3.5.1. Government policy Cycle  
In accordance with the four government policy functions, the policy process model (Jutta, 2016) recommends an 
analytical progression of the occasions that impact the formulation of government policies (Guidi et al., 2020). At 
each level of the policy process model, the connections between policy players are shown (Jutta, 2016). 
According to Appiah-Agyekum (2020), the policy model explains how decisions were made, makes 
understanding the timeline of events simpler, and supports the pragmatic nature of government policy (Guidi et 
al., 2020). 
 
Moreover, it explains how these results in the understanding that can be applied to any political system and its 
decision-making procedures (Jutta, 2016). The best method to begin a discussion of policy theories and a strategy 
to organize the study of policymaking, according to Cagnin (2017), is to use the policy cycle. According to 
Bertram, Maleki and Karsten (2019), the traditional model is cyclical since formulating policies is continuous and 
always in “motion" as a rolling wheel. 
 
The policy cycle's main lesson is that just because an issue has been identified and a decision has been taken, it 
doesn’t mean everything has been fixed (Cagnin, 2017). That only denotes the beginning of the policymaking 
process. The model’s stages are linked to each other like links in a chain cycle (Appiah-Agyekum et al., 2022). As 
Bertram, Maleki and Karsten (2019) noted, no policy decision or solution is ever final. The policy process model 
does succeed in capturing the essence of policymaking despite all of its flaws, and as a consequence, it correlates 
to political reality. 
 
According to Lerma, Díaz-Baca and Burkart (2022), the conventional model of the policy process consists of four 
functional processes or phases:  
i. 
Agenda setting; 
ii. 
Policy development;  
iii. 
Policy execution; and  
iv. 
Policy assessment 
Two additional steps that Bertram, Maleki and Karsten (2019) add to the concept of the policy process are:  
i. 
Policy legitimization; and  
ii. 
Policy modification.  
Guidi et al. (2020) postulate the results of policies and the related subsystems that must be implemented. This 
suggests that the stages theory serves as an example of how a government policy develops (or comes into 
existence). Guidi et al. (2020) assert that knowledge and information are the main forces behind policy 
construction. 
 
In a significant sense, this is the reason why everyone involved in the policy process has to be sufficiently 
informed of how government policy is produced, as well as possess the knowledge, skills, and competence 
necessary to see the process through to the end (Rakšnys and Valickas, 2023). Moreover, it serves as a tool for 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
93 
 
guiding and educating decision-makers on the procedures involved in carrying out government policy (Cairney 
2012). 
 
3.5.2. Agenda setting 
Any social issue that the public brings up should be taken seriously (Aguerre and Hernan, 2015). But more 
importantly, agenda-setting in democracies is expected to be characterized by a high level of citizenry 
participation (Blackstock et al., 2020). Various media influences can manage, shape, and define the issues on the 
policy agenda (Fischer et al., 2015). 
The topics included on the policy agenda can be influenced, controlled, shaped, and defined using these platforms 
or the participation of experts from a particular subject (Crabolu, Font and Eker, 2023). These are the three steps 
that makeup agenda setting: 
a) Identification of issues; 
b) verifying which problem is of significant essence; and  
c) Outlining the dynamics of an issue (Cagnin, 2017).  
According to Díaz-Llamas et al. (2023) and Falk and Tally (2016), only one element determines whether 
policymakers should pay attention at this early stage of the policy process. That aspect is the availability of 
information about social issues/issues. According to Steinert (2016), the public media’s assessment and awareness 
of a societal issue has an effect on agenda shaping. This is due to the possibility that the press might impact public 
opinion, given the variety of media outlets available (Chetty, 2015). 
 
3.5.3. Policy formulation 
Before creating a policy, one must create a strategy for responding to the suggestions made during the first phase 
of the policymaking process. According to Cagnin (2017), creating policies entails defining goals, estimating 
costs, and assessing the specific outcomes this policy will produce. As a result, the suggested course of action and 
the policymaker's (s) ' intentions are both stated at this point in the policy cycle (Steven, 2021). Rational, logical 
solutions are chosen. Following the conclusion of this process, crucial policy instruments are selected (Cagnin, 
2017). Falk and Tally (2016) assert that all necessary stakeholders, such as interest organizations, elected 
officials, legislators, and the public, should participate in policy development. 
 
3.5.4. Policy execution 
Only the events in the early stages of the policy process result in government policy. At this stage, it may be 
anticipated that a government policy will undergo changes, such as revision; the government policy may even be 
rejected at this stage (Kustec and Mcardle, 2012). A significant feature of government policy execution is that it 
may take on many shapes and forms depending on the institutional and cultural context (Welsh, 2019). Attention 
was called to an essential facet of carrying out government policy, especially given that it operates or is carried 
out at a time when "government" procedures are seen as having been transformed into "governance". 
 
Moreover, Jaishia et al. (2023) classify government policy execution research as a political science and 
administration subject. This suggests that overly-simplistic hierarchical models are being abandoned and that a 
broad spectrum of stakeholders is starting to participate in policymaking (Iroulo and Boateng, 2023). Also, 
politics ends when administration begins. Politics and administration are related. According to Mügge and 
Alenda-Demoutiez (2019), the institutions of democracy and the rule of law have entrenched a tight hierarchy in 
the relationship between these two disciplines. 
 
3.5.5. Policy Outcomes and Evaluation 
At this stage, a policy is evaluated to determine its success (or failure) (Cagnin, 2017). The effective execution of 
the procedure, the judgments taken about the policy, and whether the policy generated the intended results as 
described in the stage of defining the agenda and formulating the policy are all crucial factors to take into account 
when evaluating policies (Cagnin, 2017; Steven, 2021). Lessons will be drawn from this, recognized, and 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
94 
 
typically serve as the basis for future policy choices (Appiah-Agyekum et al., 2022; Cagnin, 2017). This is done 
by carefully reviewing all of the information gleaned from evaluating the policy’s outcomes (Mügge and Alenda-
Demoutiez, 2019). 
 
Many government policy actors, such as think tanks, government organizations, external consultants, nonprofit 
groups, the media, and the general public, can engage in this activity (Fischer et al., 2015). When this process is 
finished, the policy can be sent back to the legislator, who can then choose whether to change it (possibly 
signalling the start of a new policy cycle) (Steven, 2021). 
 
The policy cycle is advantageous. Welsh (2019) identified the main reasons for this:  
a) Since it is a logical process that may depict the variety of reality, it is plain and easy to grasp.  
b) Each phase disseminates knowledge to a particular section of the setting in which government policy is 
produced (Welsh, 2019). This could aid the policymaker in selecting the many variables and tactics 
available. 
c) The policy cycle shows that policymaking is flexible.  
d) The process of establishing policies follows a chronological order. 
The policy cycle also identifies the point at which the policymaking process should start, which makes it a helpful 
tool for the decision-maker. 
 
Conclusions 
 
Definitions and classifications of government policy were expanded in this study. It is essential to keep in mind 
different types of policies are defined in different ways but must be simple to comprehend. As the objective of a 
policy is more likely to shape society, the distributive or substantive policy may be seen by one group of 
participants as a regulatory policy. Still, another group may not see it as such. Therefore, policy classifications aid 
in outlining the various ways that policy stakeholders frequently describe policies and the development of 
practicability and reality of the policy that will be implemented. 
 
The elite/mass idea holds that society is divided into two groups: those in positions of authority and those who do 
not. Those with access to and influence take a more active role in creating government policy, which is in line 
with the elite/mass theory. The exciting aspect of group theory is that it aligns more with the legislative branch of 
government than the bureaucracy. This could be because the legislature is where the general population's opinions 
are represented. 
 
For the institutional theory, it was revealed that institutions and policy are closely related. The institutional 
theory's foundation is procedural legislation and how it could help or impede political goals in various 
governmental structure sectors. Although the rational choice theory assumes that government policy actors have 
access to all the information necessary to make well-informed decisions, this theory can be misleading and 
unrealistic because it believes that government policy actors will have the knowledge and ability to make rational 
decisions. 
 
To categorize and understand the contributions and linkages made by institutions and policy players and the role 
played by the external environment in producing policy, however, the systems theory offers a more 
understandable method. It was contended that as democracy is an administrative system built on broad public 
involvement, politicians and people in public office should support any concept that fosters citizen engagement in 
any form (especially in democracies). 
 
The participation of the citizenry at all proper steps of the policy cycle is only fair because government policy is 
created with the general public in mind; nonetheless, caution against dismissing any of these models or theories. 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
95 
 
They provide different viewpoints on politics and government policy and information on how these two function 
in the institutional and political domains. They directly give rise to theories of politics and government policy, 
which provide light on how issues are discussed during policymaking. 
 
It is essential to remember that government policy and how it is carried out are about ‘outcomes’ for the policies 
being implemented. It entails gathering all the inputs (needs) from the community and rating each demand 
according to priority. These inputs from the community or other role-players decide the issues listed on the 
policy’s agenda. Second, the bureaucracy must recognize the outside world since external variables, except for the 
community, significantly impact government policy. Laws, the environment on a global scale, technology, 
traditional views, politics, diversity, and complexity are some of these external elements. 
 
 
 
References 
 
Adeniran, A. O. (2016). The rational approach of Obafemi Awolowo’s Principles in enhancing efficient leadership and good governance 
through intellectual enormity. Journal of Law, Policy and Globalization, 49, 101-107. 
 
Aguerre, C., and Hernan, G. H. (2015). Internet policy formation in Latin America: understanding the links between the national, the 
regional, and the global. SSRN Electronic Journal. http://doi.org/10.2139/ssrn.2809883  
 
Ahmad, F., Draz, M. U., Chang, W. Y., Yang, S. C., and Su, L. (2021). More than there source curse: exploring the nexus of natural 
resource 
abundance 
and 
environmental 
quality 
in 
north-western 
China. 
Resour 
Policy, 
70, 
101902. 
https://doi.org/10.1016/j.resourpol.2020.101902  
 
Andrews-Speed, P. (2021). How may National culture shape public policy? The case of energy policy in China. The Energy Journal 43(3). 
http://doi.org/10.5547/01956574.43.3.pand    
 
Anyebe, P. A. (2018). An overview of approaches to the study of public policy. International Journal of Political Science (IJPS), 08-17. 
 
Appiah-Agyekum, N. N. (2020). Primary healthcare implementation in practice: evidence from primary healthcare managers in Ghana. 
African Journal of Primary Health Care and Family Medicine, 12, 1-7. https://doi.org/10.4102/phcfm.v12i1.2183  
 
Appiah-Agyekum, N. N., Kayi, E. A., Appiah-Agyekum, J., and Otoo, D. D. (2022). Capacity issues in primary health care 
implementation: examples from Ghana. Health Education, 122(1). http://doi.org/10.1108/HE-06-2021-0095  
 
Aritz, J., Walker, R., Cardon, P., and Li, Z. (2017). Discourse of leadership: The power of questions in organizational decision making. 
International Journal of Business Communication, 54(2), 161-181. http://doi.org/10.1177/2329488416687054  
 
Ashmore D. P., Thoreau, R., Kwami, C., Christie, N., and Tyler, N. A. (2020). Using thematic analysis to explore symbolism in transport 
choice across national cultures. Transportation, 47, 607-640. https://doi.org/10.1007/s11116-018-9902-71  
 
Bertram, D. (2020). Accounting for culture in policy transfer: A blueprint for research and practice. Political Studies Review, 1-18. 
http://doi.org/10.1177/1478929920965352  
 
Bertram, D., Maleki, S., and Karsten, N. (2019). Factoring in societal culture in policy transfer design: the proliferation of private 
sponsorship of refugees. Journal of International Migration and Integration, 21(5). http://doi.org/10.1007/s12134-019-00738-0  
 
Blackstock, K., Novo, P., Byg, A., and Waylen, K. A. (2020). Policy instruments for environmental public goods: Interdependencies and 
hybridity. Land Use Policy, 107(3), 104709. http://doi.org/10.1016/j.landusepol.2020.104709  
  
Cagnin, I. F. (2017). The study of macro and micro implementation of social policy, de Paul Bierman. Revista Brasileira de Políticas 
Públicas e Internacionais - RPPI, 2(2), 175-180. http://doi.org/10.22478/ufpb.2525-5584.2017v2n2.37386  
 
Carroll, P., and Common, R. (2013). Policy Transfer and Learning in Public Policy and Management, London: Routledge 
https://doi.org/10.4324/9780203749395  
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
96 
 
Chaudhary, D. (2018). Agriculture policy and rural development in Nepal. Research Nepal Journal of Development Studies, 1(2), 34-46. 
https://doi.org/10.3126/rnjds.v1i2.22425  
 
Chetty, R. (2015). Behavioral economics and public policy. American Economic Review: Paper and Proceedings, 105(5), 1-33. 
https://doi.org/10.1257/aer.p20151108  
 
Crabolu, G., Font, X., and Eker, S. (2023). Evaluating policy complexity with Causal Loop Diagrams. Annals of Tourism Research, 
100(11), 103572.    http://doi.org/10.1016/j.annals.2023.103572  
 
Creese, S., Dutton, W. H., and Esteve-Gonzalez, P. (2021). The social and cultural shaping of cybersecurity capacity building: a 
comparative study of nations and regions. Personal and Ubiquitous Computing 25(2). http://doi.org/10.1007/s00779-021-01569-6  
 
Daniell, K. (2014). The role of national culture in shaping public policy: a review of the literature. HC Coombs Policy Forum Crawford 
School of Public Policy. ANU College of Asia & the Pacific 
 
Delamaza, G. (2015). Enhancing democracy: Public policies and citizen participation in Chile. New York: Berghahn. 
 
Delamaza, G., and Palma, J. F. (2022). From above or from below? Chilean NGOs, the State and education reforms. Journal of 
Educational Administration and History. http://doi.org/10.1080/00220620.2022.2137480  
 
Deygers, B., and Vanbuel, M. (2022). Advocating an empirically-founded university admission policy. Language Policy, 21(3), 1-22. 
http://doi.org/10.1007/s10993-022-09615-6  
 
Díaz-Llamas, J. L., Soto-Ceja, E., Bravo-Olivas, M. L., and Chávez-Dagostino, R. M. (2023). Ecological footprint of an artisanal fishing 
cooperativein a lagoon body. Revista Bio Ciencias, 10 e1399. http://doi.org/10.15741/revbio.10.e1399  
 
Dunne, L., Craig, N., Miller, S., Connolly, P., Lackman, J., Aber, L., Yoshikawa, H., Fitzpatrick, S., Pham, P., Vinck, P., and Walmsley, P. 
(2021). Promoting Social Cohesion and Peacebuilding Through Investment in Early Childhood Development Programs. In book: 
Transitioning to Peace. http://doi.org/10.1007/978-3-030-77688-6_17  
 
Falk, M., and Tally, K. (2016). Cultural participation in Europe: Can we identify common determinants? Journal of Cultural Economics, 
40, 127-162. https://doi.org/10.1007/s10824-015-9242-9  
 
Fischer, F., Torgerson, D., Durnová, A., and Orsini, M. (2015). Introduction to critical policy analysis.” In Handbook of Critical Policy 
Studies, edited by F. Fischer, D. Torgerson, A. Durnová, and M. Orsini. Cheltenhamn, UK: Edward Elgar Publishing. 1-26. 
https://doi.org/10.4337/9781783472352.00005  
 
Forcher-Mayr, M., and Mahlknecht, S. (2020). Critical entrepreneurship education in general education and TVET: two concepts of 
practice in a South African township. Discourse and Communication for Sustainable Education, 11(2), 65-84. http://doi.org/10.2478/dcse-
2020-0019  
 
Galli, A. (2015). On the rationale and policy usefulness of ecological footprint accounting: The case of Morocco. Environmental Science & 
Policy, 48, 210-224. http://doi.org/10.1016/j.envsci.2015.01.008  
 
Gray, S. (2018). Achieving compliance with the World Anti-Doping Code: learning from the implementation of another international 
agreement. International Journal of Sport Policy, 11(4), 1-14. http://doi.org/10.1080/19406940.2018.1550798  
 
Guidi, M., Guardiancich, I., and Levi-Faur, D. (2020). Modes of regulatory governance: A political economy perspective. Governance, 
33(1), 5-19. http://doi.org/10.1111/gove.12479  
 
Hurel, L. M., and Rocha, M. S. (2018). Brazil, China and Internet Governance. Mapping Divergence and Convergence, 22, 67. 
http://doi.org/10.5278/ojs.jcir.v0i0.2267  
 
Imenda, S. (2014). Is there a conceptual difference between conceptual and theoretical frameworks? Journal of Social Science, 38(2), 185-
195. https://doi.org/10.1080/09718923.2014.11893249  
 
Iroulo, L. C., and Boateng, O. A. (2023). Bureaucratic acquiescence as an institutional strategy in the African Union. African and Asian 
Studies, 22(1-2), 113-134. http://doi.org/10.1163/15692108-12341583  
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
97 
 
Jaishia, M., Sharmab, G. P., Nepalic, P. B., Gauchana, D., Shresthad, R. K., Timsinae, K. P., and Neupanea, H. (2023). Government 
framework for agriculture service delivery at the local level in Nepal. Nepal Public Policy Review, 3(1). https://doi.org/10.59552/nppr.v3 
i1.28  
 
Jakonen, O., and Sokka, S. (2022). Finnish cultural policy as public funding: Regime view across policy domains. Nordisk Kulturpolitisk 
Tidsskrift 25(3), 293-313. http://doi.org/10.18261/nkt.25.3.9  
 
Jiang, Y. H. (2018). Confucian political theory in contemporary China. Annual Review of Political Science, 21, 155-173. 
https://doi.org/10.1146/annurev-polisci-041916-020230  
 
Jutta, V. (2016). Participatory turn in cultural policy? An analysis of the concept of cultural participation in Finnish cultural policy. Nordic 
Journal of Cultural Policy, 19(1), 59-77. https://doi.org/10.18261/ISSN2000-8325-2016-01-04  
 
Kharel, K. R., and Kharel, S. (2020). Analyze the effectiveness of resource mobilization of local government. Management Dynamics, 
23(1), 169-182. https://doi.org/10.3126/md.v23i1.35577    
 
Koelble, T. A., and Siddle, A. (2014). Institutional complexity and unanticipated consequences: The failure of decentralization in South 
Africa. Democratization, 21(6), 1117-1133. http://doi.org/10.1080/13510347.2013.784270  
 
Kofele-Kale (2006). The international law of responsibility for economic crimes: holding state officials individually liable for acts of 
fraudulent enrichment. Hampshire, England: Ashgate Publishing Ltd. 
 
Kraft, M. E., and Furlong, S. R. (2013). Politics, analysis and alternatives. Los Angeles: SAGE. Landsberg, C. n.d. South Africa’s African 
agenda: challenges of policy and implementation. 
 
Kustec. S., and Mcardle, D. (2012). National law, domestic governance and global policy: A case study of anti-doping policy in Slovenia. 
International Journal of Sport Policy, 6(1), 1-17. http://doi.org/10.1080/19406940.2012.662692  
 
Lerma, L M., Díaz-Baca, M. F., and Burkart, S. (2022). Public policies for the development of a sustainable cattle sector in Colombia, 
Argentina, 
and 
Costa 
Rica: 
a 
comparative 
analysis 
(2010-2020). 
Front. 
Sustain. 
Food 
Syst., 
6, 
722522. 
https://doi.org/10.3389/fsufs.2022.722522  
 
Macheridis, N., and Paulsson, A. (2019). Professionalism between profession and governance: How University teachers’ professionalism 
shapes coordination. Studies in Higher Education, 44(3), 470-485. http://doi.org/10.1080/03075079.2017.1378633  
 
Maggetti, M., and Gilardi, F. (2016). Problems and solutions in the measurement of policy diffusion mechanisms. Journal of Public Policy, 
36 1), 87-107. https://doi.org/10.1017/S0143814X1400035X  
 
Marie-Kim, C., and Marie-Hélène, G. (2020). Changes in practitioners’ attitudes, perceived training needs and self-efficacy over the 
implementation process of an evidence-based parenting program. BMC Health Services Research, 20(1). http://doi.org/10.1186/s12913-
020-05939-3  
 
Matuku Mphahlele, M., and Zandamela, H. L. (2022). Capacity Development-oriented Service Delivery Structures and Programmes: Case 
Study 
of 
a 
Mineral-rich 
District 
Municipality. 
Journal 
of 
Public 
Administration 
and 
Governance, 
12(1), 
78. 
http://doi.org/10.5296/jpag.v12i1.19363  
 
McCann, E., and Ward, K. (2013). A multi-disciplinary approach to policy transfer research: geographies, assemblages, mobilities and 
mutations. Policy Studies, 34 (1), 2-18. https://doi.org/10.1080/01442872.2012.748563  
 
Mellaard, A., and va/n Meijl, T. (2017). Doing policy: enacting a policy assemblage about domestic violence. Critical Policy Studies, 
11(3), 330–348. http://doi.org/0.1080/19460171.2016.1194766  
 
Minkman, E., van Buuren, A., and Bekkers, V. J. J. M. (2018). Policy transfer routes: An evidence-based conceptual model to explain 
policy adoption. Policy Studies, 3(2), 222-250. https://doi.org/10.1080/01442872.2018.1451503  
 
Molossi, L., Hoshide, A. K., de Abreu, D. C., and de Oliveira, R.A. (2023). Agricultural support and public policies improving 
sustainability in Brazil's beef industry. Sustainability, 15(6), 4801. http://doi.org/10.3390/su15064801  
 
Mügge, D., and Alenda-Demoutiez, J. (2019). The lure of ill-fitting unemployment statistics: How South Africa’s discouraged work 
seekers disappeared from the unemployment rate. New Political Economy, 25(104), 1-17. http://doi.org/10.1080/13563467.2019.1613355  


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
98 
 
 
Munzhedzi, P. H. (2020). Evaluating the Efficacy of municipal policy implementation in South Africa: challenges and prospects. African 
Journal of Governanceand Development, 1(9), 89-105. 
 
Myrczik, E., Heikkilä, R., Kristensen, N. N., and Purhonen, S. (2022). Missing out on culture-or not: Danes and Finns’ cultural 
participation, the pandemic, and cultural policy measures. Nordisk Kulturpolitisk Tidsskrift, 25(2), 1-24. http://doi.org/10.18261/nkt.25.2.4  
 
Nico, 
C. 
(2015). 
Differentiating 
between 
access, 
interaction 
and 
participation. 
Conjunctions, 
2(2), 
7-28. 
https://doi.org/10.7146/tjcp.v2i2.23117  
 
Nokele, K. (2022). Context and Realities of Policy Implementation in South Africa: Quo Vadis. American Journal of Multidisciplinary 
Research and Innovation (AJMRI), 1(6), 10.54536/ajmri.v1i6.898 
  
Nunes, S., Barlow, J., Gardner, T., Sales, M., Monteiro, D., and Souza Jr. C. (2019). Uncertainties in assessing the extent and legal 
compliancestatus 
of 
riparian 
forests 
in 
the 
eastern 
Brazilian 
Amazon. 
Land 
Use 
Policy, 
82, 
37-47. 
https://doi.org/10.1016/j.landusepol.2018.11.051  
 
Oyadiran, P. O., and Akintola, O. E. (2014). Public policy making and implementation in Nigeria: the Millennium Development Goals in 
perspectives. International Journal of Diversity, (1), 1-22. 
 
Ozturk (2015). Sustainability in the food-energy-water nexus: evidence from BRICS (Brazil, the Russian Federation, India, China, and 
South Africa) countries. Energy, 93. https://doi.org/10.1016/j.energy.2015.09.104  
 
Paulsson, A., and Macheridis, N. (2022). The policy unconscious: educational labor, the research-and-teaching relationship and the 
unquestioned meaning of higher education. Critical Policy Studies. http://doi.org/10.1080/19460171.2022.2064889  
 
Purtle, J., Stadnick, N. A., Wynecoop, M., and Aarons, G. (2023). A policy implementation study of earmarked taxes for mental health 
services: study protocol. Implementation Science Communications, 4(1). http://doi.org/10.1186/s43058-023-00408-4  
 
Rakšnys, A. V., and Valickas, A. (2023). The application of complexity theory in the context of public governance challenges. Viešoji 
politika ir administravimas, 21(4), 195-205. 10.13165/VPA-22-21-4-14 
 
Sol, S. L. (2023). Participation is important, but-” Professional rationalities of balancing acts in publicly funded cultural institutions. 
Nordisk Kulturpolitisk Tidsskrift, 26(1), 42-56. http://doi.org/10.18261/nkt.26.1.4  
 
Steinert, C. (2016). The impact of societal preferences on national monetary policy outcomes: A revival of economic pluralism. Politikon, 
31, 2. http://doi.org/10.22151/politikon.31.2   
 
Steven, H. (2021). Audience development and cultural policy. New directions in cultural policy research. Cham, Switzerland: Palgrave 
Macmillan 
 
Tacon, R., and Hanson, A. (2011). Sport policy and the structure of sport in the UK. In book: Managing sport business: An introduction: 
Routledge 
 
Uddin, J., Ullah, A., Saqib, N., Kousar, R., and Usman, M. (2023). Heterogeneous role of energy utilization, financial development, and 
economic development in ecological footprint: How far away are developing economies from developed ones. Environmental Science and 
Pollution Research, 30(20). http://doi.org/10.1007/s11356-023-26584-3  
 
Uminska-Woroniecka, A. (2022). We were taken on the ride. Leadership trait analysis in the circumstances of international crisis-the 
decision 
of 
Aleksander 
Kwasniewski 
in 
the 
context 
of 
intervention 
in 
Iraq. 
Teoria 
polityki, 
6, 
215-235. 
http://doi.org/10.4467/25440845TP.22.007.16007  
 
Welsh, J. (2019). Ranking academics: toward a critical politics of academic rankings. Critical Policy Studies, 13(2), 153-173. 
http://doi.org/10.1080/19460171.2017.1398673  
 
Wu, D. (2022). Implementation and public diplomacy. In book: U.S. Public Diplomacy Towards China. http://doi.org/10.1007/978-3-030-
95644-8_3    
 


 
INSIGHTS INTO REGIONAL DEVELOPMENT 
ISSN 2669-0195 (online) https://jssidoi.org/ird/ 
2023 Volume 5 Number 2 (June) 
 
 
http://doi.org/10.9770/IRD.2023.5.2(6) 
 
99 
 
Yanow, D. (2015). Making sense of policy practices: interpretation and meaning. In handbook of critical policy studies, edited by F. 
Fischer, 
D. 
Torgerson, 
A. 
Durnová, 
and 
M. 
Orsini.Cheltenhamn, 
UK: 
Edward 
Elgar 
Publishing. 
401-421. 
https://doi.org/10.4337/9781783472352.00031  
 
Yudiatmaja, W. E., Solina, E., Yudhi, I., and Samnuzulsari, T. (2022). Paid plastic shopping bags policy: What the factors drive the 
implementation? Journal of Governance and Public Policy, 9(1), 13-23. http://doi.org/10.18196/jgpp.v9i1.11152    
 
Zeb-un, N., Mustafa, G., Yaseen, Z., Arslan, M., and Imran, M. (2021). Theoretical approaches to study the public policy: an analysis of 
the cyclic/stages heuristic model. Palarch’s Journal of Archaeology of Egypt/Egyptology, 18(10), 1307-1321. 
 
 
 
 
Funding: This research was supported by the project, which has received funding from the European Union’s Horizon 2020 
research and innovation programme European Research Council (ERC) under the European Union’s Horizon 2020 research 
and innovation programme Marie Sklodowska-Curie Research and Innovation Staff Exchanges ES H2020-MSCA-RISE-
2014 CLUSDEVMED (2015-2019) Grant Agreement Number 645730730 
 
 
Author Contributions: Conceptualization: Adetayo Adeniran, Joseph Muraina, writing-original draft preparation: Adetayo 
Adeniran, Joseph Muraina, Joseph Ilugbami, writing; review and editing: Joseph Ilugbami, Adedayo Adeniran. All authors 
have read and agreed to the published version of the manuscript. 
 
 
 
Adetayo Olaniyi ADENIRAN Department of Logistics and Transport Technology, Federal University of Technology 
Akure, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0000-0002-6870-1212  
 
Joseph Mosunmola MURAINA Department of Geography and Planning Science, Ekiti State University, Ekiti, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0009-0006-5764-3594  
 
Joseph Olanrewaju ILUGBAMI Rufus Giwa Polytechnic-Owo Rector Office, Ondo State, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0009-0005-8114-5264  
 
Adedayo Ayomide ADENIRAN Department of Geopgraphy and Planning, University of Ibadan, Nigeria. 
ORCID ID: https://orcid.org/orcid.org/0009-0001-0241-6232  
 
 
 
Make your research more visible, join the Twitter account of INSIGHTS INTO REGIONAL DEVELOPMENT:  
@IntoInsights 
 
This is peer-reviewed scientific journal https://jssidoi.org/ird/page/peer-review-policy 
________________________________________________________________________________________________________________ 
________________________________________________________________________________________________________________ 
Copyright © 2023 by author(s) and VsI Entrepreneurship and Sustainability Center 
This work is licensed under the Creative Commons Attribution International License (CC BY). 
http://creativecommons.org/licenses/by/4.0/ 
 
 
 
View publication stats




This book systematically analyzes how and why China has expectedly lost and then 
surprisingly gained ground in the quest to solve the complicated environmental 
problem of air pollution over the past two decades.
Yuan Xu shines a light on how China’s sulfur dioxide emissions rose quickly 
in tandem with rapid economic growth but then dropped to a level not seen for 
at least four decades. Despite this favorable mitigation outcome, Xu details how 
this stemmed from a litany of policy stumbles within the Chinese context of no 
democracy and a lack of sound rule of law. Throughout this book, the author 
examines China’s environmental governance and strategy and how they shape 
environmental policy. The chapters weave together a goal-­
centered governance 
model that China has adopted of centralized goal setting, decentralized goal 
attainment, decentralized policy making and implementation. Xu concludes that 
this model provides compelling evidence that China’s worst environmental years 
reside in the past.
This book will be of great interest to students and scholars of Chinese 
environmental policy and governance, air pollution, climate change and sustainable 
development, as well as practitioners and policy makers working in these fields.
Yuan Xu is Associate Professor in the Department of Geography and Resource 
Management, The Chinese University of Hong Kong.
Environmental Policy and 
Air Pollution in China


Strategic Designs for Climate Policy Instrumentation
Governance at the Crossroads
Gjalt Huppes
The Right to Nature
Social Movements, Environmental Justice and Neoliberal Natures
Edited by Elia Apostolopoulou and Jose A. Cortes-­
Vazquez
Guanxi and Local Green Development in China
The Role of Entrepreneurs and Local Leaders, 1st Edition
Chunhong Sheng
Environmental Policy in India
Edited by Natalia Ciecierska-­
Holmes, Kirsten Jörgensen, Lana Ollier 
and D. Raghunandan
Mainstreaming Solar Energy in Small, Tropical Islands
Cultural and Policy Implications
Kiron C. Neale
EU Environmental Governance
Current and Future Challenges
Edited by Amandine Orsini and Elena Kavvatha
The European Union and Global Environmental Protection
Transforming Influence into Action
Edited by Mar Campins Eritja
Environmental Policy and Air Pollution in China
Governance and Strategy
Yuan Xu
For more information about this series, please visit: www.routledge.com/
Routledge-­
Studies-­
in-­
Environmental-­
Policy/book-­
series/RSEP
Routledge Studies in Environmental Policy


Environmental Policy and 
Air Pollution in China
Governance and Strategy
Yuan Xu


First published 2021
by Routledge
2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN
and by Routledge
52 Vanderbilt Avenue, New York, NY 10017
Routledge is an imprint of the Taylor & Francis Group, an informa business
© 2021 Yuan Xu
The right of Yuan Xu to be identified as author of this work has been 
asserted by him in accordance with sections 77 and 78 of the Copyright, 
Designs and Patents Act 1988.
The Open Access version of this book, available at www.taylorfrancis.
com, has been made available under a Creative Commons Attribution-­
Non 
Commercial-­
No Derivatives 4.0 license.
Trademark notice: Product or corporate names may be trademarks or 
registered trademarks, and are used only for identification and explanation 
without intent to infringe.
British Library Cataloguing-­
in-­
Publication Data
A catalogue record for this book is available from the British Library
Library of Congress Cataloging-­
in-­
Publication Data
A catalog record for this book has been requested 
ISBN: 978-­
1-­
138-­
32232-­
5 (hbk)
ISBN: 978-­
0-­
429-­
45215-­
4 (ebk)
Typeset in Times New Roman
by Apex CoVantage, LLC


List of figures
vi
List of tables
ix
Preface
x
Acknowledgments
xii
1	
Introduction
1
2	
Political will
17
3	
Environmental governance
25
4	
Mobilizing the government
42
5	
Policy making
77
6	
Policy implementation
105
7	
Environmental technology and industry
149
8	
Goal-­
centered governance
179
Index
193
Contents


	
1.1	
Environmental Performance Index in the baseline year
2
	
1.2	
China’s premature deaths due to air and water pollution 
in the Global Burden of Disease study
2
	
1.3	
Disability-­
adjusted life years (DALYs) in China due to air 
and water pollution in the Global Burden of Disease study
3
	
1.4	
DALYs in days (or disability-­
adjusted life days [DALDs]) 
per person per year in China and India
4
	
1.5	
Polity Democracy Index for China, South Korea, Singapore, 
India and the United States
5
	
1.6	
GDP per capita in PPP (purchasing power parity) in China, 
South Korea, Japan and the United States
7
	
1.7	
Governance indicators of China, India and the United States
8
	
1.8	
SO2 emissions in China
10
	
1.9	
SO2 emissions by sector in China (from two different data 
sources for 1970–2012 and 2010–2017, respectively)
11
	
1.10	 The power sector’s shares of coal consumption and SO2 
emissions in China and the United States
11
	
1.11	 SO2 emissions in the United States and SO2 intensities 
in China and the United States
12
	
2.1	
Sectoral employment changes and GDP growth rates across 
China’s administrations
19
	
2.2	
Employment and population structures in China
19
	
3.1	
Environmental protection personnel at four governmental levels 
in China
29
	
3.2	
Governmental revenue and expenditure to GDP ratios by central 
and local governments in China
33
	
3.3	
Budget balance of central and local governments in China 
as a proportion of GDP
35
	
3.4	
Governmental budget balance by provinces as a proportion 
of governmental expenditures in 2018
36
	
3.5	
The central and local governments’ shares of expenditures 
by budgetary items in 2018
37
	
3.6	
Central, local and overall governmental expenditures 
by budgetary items in 2018
37
Figures


Figures  vii
	
3.7	
Shares in governmental expenditures
38
	
4.1	
Designated SO2 emission intensity in distributing SO2 emissions 
quota to coal-­
fired power plants for 2010 in the 11th Five-­
Year Plan
58
	
4.2	
Daily SO2 concentrations in Shijiazhuang
68
	
4.3	
Daily PM2.5 concentrations in Shijiazhuang
69
	
4.4	
Daily 8-­
hour O3 concentrations (daily maximum concentration 
over 8 hours) in Shijiazhuang
70
	
4.5	
Monthly average AQI in Shijiazhuang
70
	
4.6	
Monthly average AQI in Beijing
71
	
4.7	
Monthly average AQI in Shenzhen
72
	
5.1	
Economic growth in China, Japan and the United States
85
	
5.2	
Primary energy consumption and energy efficiency
86
	
5.3	
Prices of coal (Qinhuangdao spot price), oil and natural gas
87
	
5.4	
The annual growth of primary energy consumption in China 
by fuels
88
	
5.5	
China’s primary energy consumption by fuel and the shares 
of coal and fossil fuels
88
	
5.6	
Primary energy consumption and its electrification rate
89
	
5.7	
Electricity generation by fuels in China
90
	
5.8	
The annual growth of electricity generation in China by fuels 
and coal’s share
91
	
5.9	
The decomposition of China’s SO2 emissions
91
	
5.10	 Distribution of sulfur contents in coal power plants in China
95
	
5.11	 Coal-­
fired power and SO2 scrubber capacities in China
97
	
5.12	 The annual growth of coal-­
fired power and SO2 scrubber 
capacities in China
98
	
5.13	 Annually increased SO2 scrubber capacity and unit sizes
99
	
5.14	 The annual growth of SO2 scrubber capacity by regions
100
	
5.15	 SO2 scrubbing technologies by unit sizes
101
	
6.1	
The operation of SO2 scrubbers in Jiangsu Province, including 
self-­
reported operation rates and later confirmed operation 
rates
106
	
6.2	
A conceptual model of environmental compliance monitoring
121
	
6.3	
Model simulation of compliance rates in the diagnosing and 
screening systems with available compliance monitoring 
resources and initial compliance rates
128
	
6.4	
Model simulation of equilibrium compliance rates in the 
screening and diagnosing systems in relation to available 
inspection staff
128
	
6.5	
Model simulation of equilibrium compliance rates in the 
screening and diagnosing systems in relation to (a) the number 
of polluters; (b) the ratios between pollution abatement costs 
and noncompliance penalty; (c) available inspection staff, where 
the pollution abatement cost-­
to-­
noncompliance penalty ratio has 
a lognormal distribution; and (d) the relative resource intensity 
of screening and diagnosing technologies.
130


viii  Figures
	
6.6	
Model simulation of equilibrium compliance rates in the 
screening and diagnosing systems in relation to the probabilities 
that (a) the screening technology recognizes compliance 
cases as being compliant, (b) the diagnosing technology 
recognizes compliance cases as being compliant, (c) the 
screening technology recognizes noncompliance cases as being 
noncompliant and (d) the diagnosing technology recognizes 
noncompliance cases as being noncompliant.
133
	
7.1	
The progressive paths on the deployment and operation of SO2 
scrubbers in China and the United States
150
	
7.2	
Annual average unit capital costs of SO2 scrubbers in China and 
the United States
151
	
7.3	
Model projection of the SO2 mitigation path in China’s coal-­
fired power plants: (a) deployment and operation of SO2 
scrubbers under goal-­
centered governance (the dots refer to 
actual data); (b) avoided SO2 emissions under goal-­
centered and 
rule-­
based governance
154
	
7.4	
Yearly university graduates in China from four-­
year 
undergraduate programs by subjects
157
	
7.5	
R&D personnel, expenditure and market value (in 2018 RMB) 
in China
161
	
7.6	
Patents on environmental technology by filing office in the world
162
	
7.7	
Wind energy development in China and the United States
168
	
7.8	
Companies in the Chinese and U.S. markets installing 100-­
MW-­
scale or greater SO2 scrubbers
171
	
7.9	
Average prices of wind turbines in China and the United States
172
	
8.1	
An illustration of the goal-­
centered governance model
183


	
4.1(a)	
Correlation coefficients of key factors for 27 provinces
53
	
4.1(b)	
Summary of variables
54
	
4.2	
Regression model results for distributing the national goal 
to provinces
56
	
4.3	
Regression model results for distributing provincial goals 
to municipalities
59
	
4.4	
Provincial goal distribution matrix
61
	
5.1	
Applied fractions of sulfur retained in ash
93
	
5.2	
Effluent SO2 emissions and necessary SO2 removal rates
97
	
6.1	
Data on SO2 scrubbers in China’s seven coal-­
fired power 
plants
114
	
6.2	
Decision scenarios for the managers of coal-­
fired power plants
117
	
6.3	
Key parameters in the model and their empirical values
126
	
7.1	
Up-­
front lump-­
sum fees of SO2 scrubber technology licenses
158
Tables


China is puzzling to read.
After the Cultural Revolution and a short transitional period, China entered 
the era of Reform and Open-­
up in December 1978. The size of China’s economy 
has skyrocketed by more than 30 times. Despite numerous benefits, this rapid 
economic growth also brought immense pressure on the environment. China’s 
environmental crises are multifaceted, stretching across air, water, soil, ecosystem 
and climate change.
Hope was not readily available. As a public good, environmental protection 
requires effective governmental intervention. However, China is not a democracy, 
and sound rule of law has not been established. The country’s governance quality 
has been ranked consistently and significantly lower than that in developed coun­
tries that are liberal democracies, where environmental quality first deteriorated 
with economic growth and then fundamentally improved. Their experiences sug­
gest that China’s environmental crises are expected, while their solutions are hard 
to reach.
Then what happened in China in the past 15 years became surprising as the 
environmental trajectory deviated away from the projections. Sulfur dioxide (SO2) 
is one air pollutant that is crucial for air quality but very difficult to control. Since 
reaching their peak in the mid-­
2000s, SO2 emissions in China have been declin­
ing, and the downward pace accelerated in the past few years to reach a level not 
seen in more than four decades. A large coal-fired power sector appeared to install 
and operate SO2 scrubbers that mitigate emissions from polluting sources. Simi­
lar desirable outcomes are also observed in other environmental and renewable 
energy fields. However, China has not changed seriously from the perspectives of 
democracy and the rule of law, although environmental policy has been improv­
ing and strengthening. The legal system still does not play any major role in envi­
ronmental protection. Policy making lacks transparency and public consultation, 
while policy blunders are not rare. Policy implementation still has considerable 
problems and is often selective. It is not unusual to hear about the abuse of gov­
ernmental authorities.
This book aims to provide a theoretical understanding to explain how China 
achieved deep and sustained pollution mitigation without democracy and sound 
rule of law. Causal relationships are explored between the favorable outcome and 
Preface


Preface  xi
the unfavorable path. The major puzzle is why China frequently witnesses both 
sides at the same time or whether the conventional insights may have missed 
something important in reading China. China’s strategy is theorized into goal-­
centered governance. China is both highly centralized – in goal setting – and 
highly decentralized – in goal attainment, policy making and implementation. 
Unlike the rule-­
based governance in developed countries as indicated in their 
well-­
established rule of law, China places goals in the first place, while deficien­
cies in policy making and implementation are much tolerated as long as goals 
can be attained. The mitigation trajectory was not centrally planned but gradu­
ally evolved through decentralized pathfinding under centralized goals. In other 
words, the Chinese puzzle should primarily be explained from the perspective of 
its governance strategy but not individual policies. A strategic mistake is often a 
lot more devastating and far-­
reaching than any policy stumble, while an effective 
strategy can accommodate many policy mistakes without compromising much 
the final outcome.
The research and thinking for this book stretched over a dozen years. When 
I first started studying China’s SO2 mitigation around 2007, the hypothesis was 
that the environmental crisis was rooted in policy failures and, more fundamen­
tally, the lack of democracy and the rule of law. However, what unfolded later 
forced me to rethink this causal relationship, especially in the 2010s when the 
mitigation pace dashed forward. As a former physicist, I hope to find a theoretical 
explanation to the Chinese puzzle that is simple, like one equation, and rich. The 
goal-­
centered governance in this book reflects such a new attempt.


I owe a tremendous amount of debts to many people. This book is dedicated 
to Robert H. Socolow, the supervisor of my PhD thesis at Princeton Univer­
sity’s Woodrow Wilson School of Public and International Affairs. His inspi­
ration is vital in my research journey. Much of this book is rooted although 
widely extended from my PhD study over a decade ago. I am grateful for Rob­
ert H. Williams, Denise L. Mauzerall, Eric D. Larson, Yiguang Ju, Gregory 
C. Chow, Edward S. Steinfeld, Richard K. Lester and Kin-­
Che Lam, whose 
support and insights were crucial to sustain and enlighten this research. My 
deep appreciation also goes to numerous interviewees who kindly shared their 
knowledge. I thank Matthew Shobbrook of Routledge, whom I worked with 
to finally complete this book.
My wife, Jing Song, and our two children, Anlan Xu and Antao Song, are per­
petual motivation and sources of encouragement for my research. My parents, 
Meilan Yuan and Yicai Xu, and parents-­
in-­
law, Meiyu Song and Changfa Song, 
provide patient and unconditional support. My family made this work possible, 
especially under the ongoing COVID-­
19 pandemic.
Funding support throughout this research in the past dozen years was provided 
by Princeton University, Massachusetts Institute of Technology, The Chinese 
University of Hong Kong, and Hong Kong Research Grants Council (General 
Research Fund, 14654016).
Parts of the book were adapted with permissions from the author’s several pub­
lished journal articles, including Xu, Y. 2011. The use of a goal for SO2 mitigation 
planning and management in China’s 11th five-­
year plan. Journal of Environmen­
tal Planning and Management, 54, 769–783 [in Chapter 4; Copyright (2011) Tay­
lor & Francis]; Xu, Y. 2011. Improvements in the operation of SO2 scrubbers in 
China’s coal power plants. Environmental Science & Technology, 45, 380–385 [in 
Chapter 6; Copyright (2011) American Chemical Society]; Xu, Y. 2011. China’s 
functioning market for sulfur dioxide scrubbing technologies. Environmental Sci­
ence & Technology, 45, 9161–9167 [in Chapter 7; Copyright (2011) American 
Chemical Society]; Xu, Y. 2013. Comparative advantage strategy for rapid pol­
lution mitigation in China. Environmental Science & Technology, 47, 9596–9603 
[in Chapter 7; Copyright (2013) American Chemical Society]. Much has been 
revised and expanded on. 
Acknowledgments


1  
China’s environmental crises
China faces colossal, multifaceted environmental challenges, many at crisis lev­
els. Its environmental degradation has been widely documented and analyzed in 
academic studies as well as in public media. China is now the largest energy 
consumer, supplier and emitter of most major air and water pollutants as well as 
various greenhouse gases. Together with its geographically high population and 
economic densities, especially in the eastern half of the country, China was cat­
egorized at the very bottom of air quality among the 180 countries and regions in 
the Environmental Performance Index (Wendling et al., 2018; Figure 1.1). Few 
readers would be surprised to know that China’s air quality is among the most 
polluted in the world (Figure 1.1).
Air and water pollution in China have certainly taken a serious toll. China has 
made steady progress in the past decades to significantly reduce premature deaths 
due to water-­
related environmental factors and indoor air pollution, but ambi­
ent particulate matter (PM) pollution has been deteriorating. The Global Burden 
of Disease study elaborates in great detail the causes and risk factors of deaths 
across individual countries (Institute for Health Metrics and Evaluation, 2018). In 
1990, China accounted for 22.2% of the global population, and in 2017, the share 
dropped to 18.5% despite an 18.0% increase in absolute population (Figure 1.2). 
In premature deaths that are due to environmental risk factors, China’s share in the 
world in 1990 was 29.2% for household air pollution from solid fuels and 4.6% 
for unsafe water, sanitation and handwashing. In other words, an average Chinese 
was 31.5% more likely and 79.3% less likely to die prematurely due to the two 
risks than an average person in the world. The shares were significantly reduced 
to 16.5% and 0.6% in 2017, respectively, to make an average Chinese 10.6% and 
96.8% less likely to die prematurely. In absolute terms, they were reduced by 
65.7% and 92.5%, respectively. However, ambient PM pollution caused 404,000 
premature deaths in 1990 and 852,000 in 2017, more than double. Its global share 
climbed from 23.0% to 29.0% over the period. In 2000, indoor air pollution was 
overtaken by ambient PM pollution in causing more premature deaths. In com­
parison to China’s share of the global population, in 1990, an average Chinese 
faced only a slightly greater risk, 3.8%, from ambient PM pollution than an aver­
age person in the world, but in 2017, the risk premium was enlarged to 56.9%.
1	

Introduction


2  Introduction
China
US
India
Japan
South Korea
UK
0
10
20
30
40
50
60
70
80
90
100
0
20
40
60
80
100
Air quality
Air pollution
Figure 1.1  
Environmental Performance Index in the baseline year
Source: Wendling et al. (2018).
Note: “Air pollution” at the x-axis refers to sulfur dioxide (SO2) and nitrogen oxide (NOx) emission inten­
sities, and its baseline year is 2006. “Air quality” in the y-axis indicates household solid fuels (baseline 
year: 2005), fine particulate matter (PM2.5) exposure and PM2.5 exceedance (baseline year: 2008).
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
0
150,000
300,000
450,000
600,000
750,000
900,000
1990
1995
2000
2005
2010
2015
China’s share in the world
)
s
n
o
s
r
e
p
(
 
s
h
t
a
e
d
 
e
r
u
t
a
m
e
r
P
Year
Ambient particulate matter pollution
Household air pollution from solid fuels
Unsafe water, sanitation and handwashing
Share of population
Figure 1.2  
China’s premature deaths due to air and water pollution in the Global Burden 
of Disease study
Source: Institute for Health Metrics and Evaluation (2018).
Note: Solid lines indicate absolute numbers in persons with the left y-axis, while dashed lines refer to 
China’s shares in the world with the right y-axis.


Introduction  3
Another measurement of pollution’s health impact is the disability-­
adjusted life 
years (DALYs) that quantifies the loss of “healthy” life years. It combines the lost 
life years due to both premature deaths and illnesses. Various types of environmental 
pollution in different countries may cause premature deaths and illnesses that cor­
respond to different life expectancies, ages and other situations. The ratio between 
DALYs and premature deaths is much higher for water pollution than for air pollu­
tion. For example, in 2017, China lost 19.8 million, 6.46 million and 0.85 million 
DALYs due to ambient PM pollution, household air pollution from solid fuels, and 
unsafe water, sanitation and handwashing, respectively. The corresponding ratios 
between DALYs and premature deaths were 23.3, 23.8 and 89.0, respectively, to 
indicate the more severe health impacts of water pollution for an average case.
Nevertheless, the indicator of DALYs does not change the conclusion that was 
presented with the examination of premature deaths (Figure 1.3). Substantial pro­
gress was also made on indoor air pollution and water, with their DALYs being 
reduced by 77.3% and 92.0%, while the deterioration trend for ambient PM pol­
lution is distinguished with an increase of DALYs by 47.3%. In terms of China’s 
shares in the world, ambient PM pollution is still the only risk factor among the 
three to surpass that of its population, which accounted for 23.8% of the world’s 
total in 2017. For all DALYs due to the three environmental risk factors, ambient 
PM pollution’s share rose from 25.6% in 1990 to 73.0% in 2017. Accordingly, 
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
0
5
10
15
20
25
30
1990
1995
2000
2005
2010
2015
China’s share in the world
)
s
r
a
e
y
 
0
0
0
,
0
0
0
,
1
(
 
s
Y
L
A
D
Year
Ambient particulate matter pollution
Household air pollution from solid fuels
Unsafe water, sanitation and handwashing
Share of population
Figure 1.3  
Disability-adjusted life years (DALYs) in China due to air and water pollution 
in the Global Burden of Disease study
Source: Institute for Health Metrics and Evaluation (2018).


4  Introduction
environmental pollution in China is more and more dominated by ambient air 
pollution and especially PM pollution.
On average, the DALYs due to various environmental risks indicate that an 
average Chinese loses a significant number of healthy life days for every year liv­
ing in these environmental risks. In 1990, household air pollution from solid fuels 
was the most severe environmental risk in China to incur the loss of 8.7 disability-­
adjusted life days (DALDs) per person, while the damages from ambient PM 
pollution and from unsafe water, sanitation and handwashing were similar at 4.1 
and 3.2 DALDs per person, respectively (Figure 1.4). In other words, an average 
Chinese lost 16.0 health life days due to the three air and water pollution risk 
factors for living through 1990. In 2017, ambient PM pollution became the most 
severe risk factor, being responsible for 5.1 DALDs per person or 1.0 DALDs 
more, after the other two experienced dramatic improvement in the past decades. 
The total loss was 7.0 DALDs for living through 2017.
China is not a unique country to witness the diverging progress of different risk 
factors. India had similar paths for distinguishing the rising importance of ambi­
ent PM pollution in environmental protection. Ambient PM pollution in India 
has remained stable throughout the years to account for 5.7 and 5.6 DALDs per 
person in 1990 and 2017, respectively. Although household air pollution from 
solid fuels still claimed greater health damages in 2017, its steady declining trend 
0
5
10
15
20
25
30
1990
1995
2000
2005
2010
2015
)
r
a
e
y
 
r
e
p
 
n
o
s
r
e
p
 
r
e
p
 
s
y
a
d
(
 
s
Y
L
A
D
Year
China: Ambient particulate matter pollution
China: Household air pollution from solid fuels
China: Unsafe water, sanitation and handwashing
India: Ambient particulate matter pollution
India: Household air pollution from solid fuels
India: Unsafe water, sanitation and handwashing
Figure 1.4  
DALYs in days (or disability-adjusted life days [DALDs]) per person per year 
in China and India
Source: Institute for Health Metrics and Evaluation (2018).


Introduction  5
suggests that ambient PM pollution will soon become the most damaging environ­
mental risk among the three in India as well (Figure 1.4).
2  
China’s expected rise of SO2 emissions and unexpected 
success in SO2 mitigation
China has been rapidly industrializing in the past four decades. Environmental cri­
ses can be empirically expected in the contexts of its rapid economic development, 
rising energy consumption and coal dominance. The expectation also comes from 
crucial governance factors that are believed to be favorable for environmental tran­
sition but that China is especially weak at. First, democracy is believed to be good 
for environmental protection by many scholars (e.g., Payne, 1995). Unfortunately, 
China is not a democracy, and thus, society’s demand for cleaner air is often not 
believed to be able to effectively influence policy making as in a democracy. It is 
generally ranked at the bottom of various democracy indexes. According to Polity’s 
ratings that can reflect the common views of democracy evaluation at least in West­
ern liberal democracies, modern-­
day China, under the communist rule, is debatably 
less democratic than the imperial days in the 19th-­
century Qing dynasty, when the 
emperors still held absolute power, with the Polity index being −6 (Marshall et al., 
2019). China’s economic reform era after the Cultural Revolution only slightly 
–10
–8
–6
–4
–2
0
2
4
6
8
10
1980
1985
1990
1995
2000
2005
2010
2015
Polity Index (–10 ~10) 
Year
China
South Korea
South Korea
India
United States
Singapore
Figure 1.5  
Polity Democracy Index for China, South Korea, Singapore, India and the 
United States (−10 being the most autocratic and 10 the most democratic)
Source: Marshall et al. (2019).


6  Introduction
improved its Polity index from −8 to −7 (Figure 1.5). In comparison, South Korea 
was fundamentally transformed from an authoritarian regime to a democratic one 
after the reform in the 1980s. Singapore is steadily ranked toward the authoritarian 
side. India and the United States are standard democracies despite slight fluctuations.
Democratic states are argued to be more responsive to the public’s demands. If the 
public in a democracy gives top priority to environmental matters, strong political 
will is more likely to be generated (Li and Reuveny, 2006; Payne, 1995; Downey and 
Strife, 2010). Furthermore, the public in a democracy could be more pro-­
environment 
than are the elites in an autocracy; this could be because of better access to informa­
tion, a more developed civil society and a longer time horizon of planning (Li and 
Reuveny, 2006; Payne, 1995). Democracy is generally closely associated with the 
rule of law, and therefore, there should be better enforcement of environmental regu­
lations (Li and Reuveny, 2006). Nevertheless, democracy might also be associated 
with weakness in environmental protection. People’s self-­
interest and the interests of 
business are more difficult to overcome in a democracy (Li and Reuveny, 2006). If 
the public gives only a low priority to having a clean environment, then a democracy 
could be less likely to heavily focus on environmental protection.
Empirical statistical studies have found no conclusive relationship between 
democracy and the environment. Congleton (1992) and Neumayer (2002) found 
that democracy contributes positively to international environmental commit­
ments. Midlarsky (1998) discovered that democracy leads to more protected 
areas of land, but that it tends to negatively influence deforestation and carbon 
dioxide (CO2) emissions per capita. Winslow (2005) found only good effects of 
democracy, whereas Pellegrini and Gerlagh (2006) found that it had insignificant 
impacts. The mixed results of the relationship could be at least partly caused by 
the difference in environmental indicators. For example, CO2 is more difficult 
to abate, but it has much less local influence than urban particulate pollution. 
Studies that used panel data also reported mixed results regarding the relation­
ship (Torras and Boyce, 1998; Barrett and Graddy, 2000). Different democracy 
indexes do not differ greatly in their relationship to the environment. A prob­
lem in the literature is that a linear relationship is generally assumed between 
democracy and the environment. However, theoretical arguments might suggest 
that both democracy and autocracy could have a beneficial effect on environ­
mental protection, while regimes in between make the situation worse. Among 
control variables, the most common one is income. Considering the literature on 
the Environmental Kuznets Curve and a plausible relationship between income 
and the environment (Grossman and Krueger, 1995; Stern and Common, 2001), 
income together with its squared and cubed terms are necessary control vari­
ables. One study that did not include income as an independent variable could 
suffer from potential missing-­
variable problems (Winslow, 2005). In addition, 
two studies controlled a governance index, namely, that of corruption (Pellegrini 
and Gerlagh, 2006; Buitenzorgy and Mol, 2011), but most of them disregarded 
governance. Various studies differ greatly from each other in how they control 
other variables, including trade openness (Li and Reuveny, 2006), inequality/
Gini ratio (Torras and Boyce, 1998), energy resource endowment (Congleton, 


Introduction  7
1992), country size in gross domestic product (GDP; Winslow, 2005), population 
size (Neumayer, 2002; Congleton, 1992) and literacy (Torras and Boyce, 1998).
Case studies found no conclusive relationship either. A case study in Kenya 
found that democracy is benign to the environment; this is because the government 
responded mainly to the “environmental and developmental civil society” and 
“Western supporters” rather than to the “marginalized poor” (Njeru, 2010). On the 
other hand, democratization in a number of southern African countries, particu­
larly Malawi, South Africa and Mozambique, has resulted in greater destruction 
of the environment for short-­
term economic and social reasons (Walker, 1999). 
In Mexico City, it has been found that democratic elections do not assist in stop­
ping local deforestation (Hagene, 2010). Through studying China and Southeast 
Asia, it is even proposed that “ ‘good’ authoritarianism” is essential for solving 
our urgent environmental problems (Beeson, 2010). A case study in Guatemala 
found that the relationship between democracy and the environment is complex 
and not straightforward (Sundberg, 2003).
In addition, the empirical relationship between economic development and 
environmental quality did not expect that China would be able, or willing, to 
pull down its pollutant emissions and improve air quality. Environmental Kuznets 
Curve – an empirical bell-­
shaped relationship between income level and environ­
mental quality – predicts that before a country becomes rich enough to reach a 
certain level of income (or GDP per capita), its environmental quality will keep 
0
10,000
20,000
30,000
40,000
50,000
60,000
1980
1985
1990
1995
2000
2005
2010
2015
)
p
a
c
/
$
S
U
 
1
1
0
2
 
,
P
P
P
(
 
a
t
i
p
a
c
 
r
e
p
 
P
D
G
Year
South Korea
China
Japan
United States
Figure 1.6  
GDP per capita in PPP (purchasing power parity) in China, South Korea, Japan 
and the United States
Source: IMF (2019).


8  Introduction
deteriorating (Grossman and Krueger, 1995). China’s GDP per capita in purchas­
ing power parity and constant 2011 dollars in 2018 was US$16,100, and the Inter­
national Monetary Fund projected that it would rise to US$22,200 in 2024, while 
the level was US$29,100 in the United States in 1980 (Figure 1.6). In other words, 
China is about five decades behind the United States in terms of economic devel­
opment status. Different studies report different turning points, and the lowest one 
for SO2 emissions is at about US$3,000 (in 1990 US$ and nominal exchange rates) 
(Stern and Common, 2001). China’s GDP per capita only surpassed US$3,000 per 
capita in nominal terms in 2008 (IMF, 2019), which was still much lower than the 
empirical minimum turning point.
Furthermore, environmental governance is critical to provide better environ­
mental quality as a public good. As suggested in the World Bank’s six governance 
indicators, comparatively China is poorly governed (Kaufmann and Kraay, 2019). 
The indicators assigned a score between −2.5 (worst) and 2.5 (best) to indicate 
governance performance. On “voice and accountability,” China scored consist­
ently and significantly lower than democracies, such as the United States and India. 
Their average scores from 1996 to 2018 were −1.58, 1.18 and 0.42, respectively 
–2.50
–2.00
–1.50
–1.00
–0.50
0.00
0.50
1.00
1.50
2.00
1996
2000
2003
2005
2007
2009
2011
2013
2015
2017
Governance indicators (–2.5 ~ 2.5)
Year
Voice_China
Voice_India
Voice_US
Law_China
Law-India
Law-US
Figure 1.7  
Governance indicators of China, India and the United States
Source: Kaufmann and Kraay (2019).
Note: “Voice”: Voice and accountability “reflects perceptions of the extent to which a country’s citi­
zens are able to participate in selecting their government, as well as freedom of expression, freedom of 
association, and a free media.” “Law”: Rule of law measures “perceptions of the extent to which agents 
have confidence in and abide by the rules of society, and in particular the quality of contract enforce­
ment, property rights, the police, and the courts, as well as the likelihood of crime and violence.”


Introduction  9
(Kaufmann and Kraay, 2019; Figure 1.7). It suggests that Chinese citizens are 
less able to directly participate in selecting a government and that their voices are 
less likely to be heard. In terms of “political stability and absence of violence/
terrorism,” China scored −0.44, better than India’s −1.13 but worse than United 
States’ 0.48. “Government effectiveness” measures the provision of public and 
civil services as well as the quality of policy making and implementation. It is the 
governance indicator that China had the best performance. It is also the only one 
that China’s score is positive, being 0.09 on average, and consistently improved 
from −0.35 in 1996 to 0.48 in 2018 (Kaufmann and Kraay, 2019). Nevertheless, 
China is still much behind the United States that scored 1.58 in 2018. For the “rule 
of law” indicator, China performs poorly with an average score of −0.46, much 
lower than the United States’ 1.58 and India’s 0.07 (Figure 1.7). Although slight 
progress was made in China from −0.55 in 1996 to −0.20 in 2018, it was always 
located in the negative territory. Little progress was achieved on “corruption” as 
the score remained consistently low with an average of −0.41, which was poorer 
than the 1.47 in the United States and −0.38 in India (Kaufmann and Kraay, 2019). 
China performed steadily poor in “regulatory quality” that focuses on the private 
sector. The United States scored 1.51 on average for the 1996–2018 period, much 
better than China’s −0.25 or India’s −0.36 (Kaufmann and Kraay, 2019). These 
governance indicators quantitatively measure various aspects of governance in a 
country to enable comparison across countries and years. As a classical example 
of market failure to demand governmental intervention, environmental protec­
tion cannot be effective without effective governance. However, none of the six 
governance indicators suggest that the Chinese government can sustainably, effec­
tively and efficiently enact and implement environmental policies and laws.
With all the unfavorable conditions and rising environmental pressures from 
energy consumption, little hope existed to make China’s environmental cleanup 
promising. SO2 is one of the most important air pollutants, and it was also the 
first air pollutant explicitly included in the national Five-­
Year Plans for serious 
mitigation (National People’s Congress, 2006). Its emissions were more than 
doubled from 1980 to the 2000s to echo such expectations (Figure 1.8). How­
ever, something has obviously worked as indicated in the more recent trajectory 
of SO2 emissions (Figure 1.8). Multiple data sources – from Chinese official sta­
tistics, independent bottom-­
up and top-­
down estimates inside and outside of the 
country to satellite and remote sensing data – all point to the same trend: China’s 
SO2 emissions have been rapidly decreasing in the past decade (Li et al., 2017; 
Zheng et al., 2018; Lu et al., 2011; Crippa et al., 2018; Fioletov et al., 2019; 
National Statistics Bureau and Ministry of Ecology and Environment, 2019). 
Although different emission inventories still show gaps between each other on 
when peak SO2 emissions happened and how high they reached, China should 
have completely wiped out all additional SO2 emissions that accompanied its 
unprecedented economic growth in the past four decades (Figure 1.8). Although 
China’s economy has expanded by more than 30-­
fold since the Open-­
up policy 
was initiated in 1978, the country now emits significantly less SO2 (Figure 1.8). 
It seems to have taken China less than one decade to remove all the additional 


10  Introduction
SO2 emissions that the country increased with its economic development and 
energy consumption.
With the rapid electrification trend of energy consumption and the power sec­
tor’s increasing share of coal consumption, the power sector is becoming more 
and more important in deciding the trajectory of China’s SO2 mitigation. In 
1980, its share of SO2 emissions was only 22.5%, less than the industrial sec­
tor’s 50.0% and the residential sector’s 23.5% (Figure 1.9). The relatively less 
significance was due to the power sector’s low share of coal consumption, 20.2% 
(Figure 1.10). In the following two decades, the power sector’s share climbed 
continuously to peak in 2002 at 45.7% and surpass that of the industrial and resi­
dential sectors (Figure 1.9) together with its 52.2% share of coal consumption 
(Figure 1.10). However, these two trajectories started to diverge from each other 
afterward (Figure 1.10). In 2017, the power sector consumed 57.3% of China’s 
coal but only accounted for 17.4% of SO2 emissions (Figure 1.9). The industrial 
and residential sectors’ shares rebounded to reach 56.8% and 22.6%, respectively. 
Accordingly, the power sector now emits much less SO2 for consuming one unit 
of coal than the industrial and residential sectors do.
Although energy transition away from coal is favorable for SO2 mitigation, coal 
consumption in China still remains at a high level, with only a slight decrease 
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
1980
1985
1990
1995
2000
2005
2010
2015
SO2 emissions (1,000 tons)
Year
Official
EDGAR
Lu et al., 2011
Zheng et al., 2018
Fioletov et al., 2019
Figure 1.8  
SO2 emissions in China
Source: Data from Fioletov et al. (2019) refer to large power plants, while others are for China as a 
whole (Zheng et al., 2018; Lu et al., 2011; Crippa et al., 2018; Fioletov et al., 2019; National Statistics 
Bureau and Ministry of Ecology and Environment, 2019).


Introduction  11
0%
10%
20%
30%
40%
50%
60%
70%
0
5
10
15
20
25
30
35
1970 1974 1978 1982 1986 1990 1994 1998 2002 2006 2010 2010 2014
Shares in SO2 emissions
SO2 emissions (million tons)
Power
Industry
Residential
Others
Power’s share (right)
Industry’s share (right)
Residential’s share (right)
Year
Figure 1.9  
SO2 emissions by sector in China (from two different data sources for 1970–
2012 and 2010–2017, respectively)
Source: Crippa et al. (2018); Zheng et al. (2018).
0
500
1,000
1,500
2,000
2,500
3,000
3,500
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1950
1960
1970
1980
1990
2000
2010
Coal consumption (Mtce)
 
e
r
a
h
s
 
s
r’
o
t
c
e
s
 
r
e
w
o
P
Year
Power’s share of coal consumption: China
Power’s share of coal consumption: U.S.
Power’s share of SO2 emissions (EDGAR)
Power’s share of SO2 emissions (Zheng et al., 2018)
Total coal consumption: China (right)
Figure 1.10  
The power sector’s shares of coal consumption and SO2 emissions in China 
and the United States
Source: EIA (2019); Fridley and Lu (2016); National Bureau of Statistics (2019).


12  Introduction
in recent years (Figure  1.10). Most mitigation of absolute SO2 emissions was 
because much greater SO2 emissions are avoided per unit of coal consumption. 
The power sector’s high efficiency in removing SO2 is also reflected in its SO2 
emission intensity of coal-­
fired electricity. Since the enactment of the Clean Air 
Act Amendments (1990), the United States has substantially reduced its overall 
SO2 emissions from 20.9 million tons in 1990 to 2.48 million tons in 2018 (Fig­
ure 1.11). The power sector has consistently been the largest contributor, and its 
SO2 emissions dropped from 14.4 million tons to 1.19 million tons over the same 
period, while its share declined from 68.9% to 47.8%. The much higher share than 
China’s reflects the power sector’s greater importance in U.S. coal consumption 
(Figure 1.10).
In reference to the successful progress in the United States, China’s SO2 miti­
gation trajectory was even steeper. In 1990, for generating 1 kWh of coal-­
fired 
electricity, 8.4 g of SO2 were emitted in the United States, while the rate was 
58.5% higher, or 13.3 g in China. In 2017, as calculated with independent emis­
sion inventory data, the SO2 intensity decreased to be 0.96 g in the United States 
and 0.41 g in China, 56.9% lower (Figure 1.11).
0%
10%
20%
30%
40%
50%
60%
70%
80%
0
5
10
15
20
25
1985
1990
1995
2000
2005
2010
2015
Power sector’s share of SO2 emissions
SO2
O
S
 
&
 
)
s
e
n
n
o
t
 
n
o
i
l
l
i
m
(
 
.
S
.
U
 
e
h
t
 
n
i
 
s
n
o
i
s
s
i
m
e
2
l
a
o
c
 
f
o
 
y
t
i
s
n
e
t
n
i
-
O
S
 
g
(
 
y
t
i
c
i
r
t
c
e
l
e
 
d
e
r
i
f
2/kWh)
Year
Power
Industry
Others
Intensity: U.S.
Intensity: China (EDGAR)
Intensity: China (Zheng et al., 2018)
Power’s share (right)
Figure 1.11  
SO2 emissions in the United States and SO2 intensities in China and the United 
States
Source: Crippa et al. (2018); Zheng et al. (2018); BP (2019); U.S. EPA (2019).


Introduction  13
3  
The organization of this book
Democracy and rule of law have played prominent and indispensable roles in 
environmental cleanup in developed countries. However, China is not a democ­
racy and political freedom is indeed highly constrained, but why environmen­
tal protection became the country’s priority to witness a dramatic drop in SO2 
emissions? Furthermore, many policies are not implemented well, and the legal 
system plays an essentially negligible role in China’s environmental protection. 
But why the government was able to effectively bend down pollutant emissions at 
such an astonishing pace? This book focuses on how China defied the empirical 
expectations in SO2 mitigation, especially in the coal-­
fired power sector. It aims to 
provide an explanation at the strategic level for understanding how environmental 
governance is organized and implemented in China.
This book also aims to imply China’s governance in general. Observers on 
China’s governance often have polarized views and each side seems to have 
ample supporting evidence. Regardless of what the focused perspective is, 
China is full of puzzles and controversies. The country has made many remark­
able achievements in the past 40 years, with much higher income and living 
standards, much better infrastructures, much wider social safety nets, much less 
control of individuals’ private lives and much less poverty. It leads the world 
in renewable energy development and electric vehicles. However, rules are 
much less respected in China than in developed countries. The parliament – the 
National People’s Congress – is often referred to as a “rubber stamp,” although 
in the Chinese Constitution, it has the utmost authority beyond any governmen­
tal entity. The judicial system is not independent. Political liberty is much con­
strained without genuine elections. The Chinese Communist Party has almost 
unchecked power, and the authoritarian country is ruled from the top, but an 
often-­
heard sentence in China goes that “policies and orders cannot go beyond 
Zhongnanhai” (the compound where the central government is located). How 
should we explain China’s governance and reconcile the polarized observations 
that are both well documented and evidence-­
based? Are the two sides caus­
ally connected? How can China achieve those favorable outcomes with such 
an unfavorable policy pathway? If we repair all recognized deficiencies in the 
governance, are we going to throw away the baby together with the bathwater? 
Most important, does China follow a different governance model from that in 
developed countries, and thus, is the explanatory power of many theories and 
historical experiences reduced?
The rest of the book is organized as follows: Chapters 2, 3 and 4 examine how 
the Chinese government is organized for environmental protection, especially in the 
contexts of neither democracy nor sound rule of law. Chapter 2 explores how the 
political will for environmental protection has been centrally evolving without 
democracy. Chapter 3 discusses China’s environmental governance structure that 
combines high degrees of both centralization and decentralization from different 


14  Introduction
perspectives. Primary focuses are on the evolution of the Ministry of Ecology 
and Environment and the relationships between the central and local govern­
ments. Chapter 4 studies how prioritized environmental protection is transmitted 
from the central government to local governments for their effective mobilization 
against the background of a weak rule of law. The environmental governance is 
organized to center on goals, specifically on SO2 emissions and environmental 
protection in Five-­
Year Plans. This book calls the governance strategy in China 
as the goal-­
centered governance model that features centralized goal setting and 
decentralized goal attainment.
Chapters 5, 6 and 7 analyze the impacts of China’s goal-­
centered governance 
model. Chapter 5 focuses on decentralized policy making for SO2 mitigation that 
is guided by centralized, top-­
down goals. This integration of centralization and 
decentralization has generated not only profound outcomes, with active policy 
making, innovation and competition, but also many policy deficiencies. China’s 
governance is tolerant of mistakes or even abuses in policy making, as long as 
goals can be achieved. Such tolerance then significantly reduces the requirements 
for policy making quality, choices of policy instruments and inter-­
policy coordi­
nation. Chapter 6 explores how this goal-­
centered governance has exerted impacts 
on decentralized policy implementation. From unfavorable backgrounds of inad­
equate capacity, effectiveness and efficiency of environmental policy implemen­
tation, local governments make gradual and steady improvements that aim for 
approaching their assigned goals. Chapter 7 addresses how China overcame sup­
ply constraints and established its domestic SO2 scrubber industry for meeting the 
skyrocketing demand. Decentralized market entities were able to actively seek 
and capture market opportunities under goal-­
centered governance. Goals on envi­
ronmental protection and economic development could thus achieve better syner­
gies than conflicts.
Chapter  8 concludes this book and discusses the goal-­
centered governance 
model. This theoretical framework can integrate the polarized observations on 
China within a systematic and compatible understanding. The rule-­
based govern­
ance model is the primarily applied strategy in countries with sound rule of law 
that emphasizes on making good, often centralized policies as means, but the 
final outcome is less explicit. In contrast, this goal-­
centered governance model 
emphasizes centralized goals as ends but is more relaxed on the means to result 
in many policy deficiencies. In the contexts of China’s backgrounds of no democ­
racy and weak rule of law, this governance strategy has been proved effective 
not only on SO2 mitigation but also very likely on other prioritized governmental 
affairs. China is also applying the same strategy in governing CO2 mitigation. 
Other countries may also find this alternative governance model helpful in con­
tributing solutions to their major public problems.
References
Barrett, S. & Graddy, K. 2000. Freedom, growth, and the environment. Environment and 
Development Economics, 5, 433–456.


Introduction  15
Beeson, M. 2010. The coming of environmental authoritarianism. Environmental Politics, 
19, 276–294.
BP. 2019. Statistical review of world energy [Online]. Available: https://www.bp.com/
content/dam/bp/business-­
sites/en/global/corporate/pdfs/energy-­
economics/statistical-­
review/bp-­
stats-­
review-­
2019-­
full-­
report.pdf.
Buitenzorgy, M. & Mol, A. P. J. 2011. Does democracy lead to a better environment? 
Deforestation and the democratic transition peak. Environmental & Resource Economics, 
48, 59–70.
Congleton, R. D. 1992. Political-­
institutions and pollution-­
control. Review of Economics 
and Statistics, 74, 412–421.
Crippa, M., Guizzardi, D., Muntean, M., Schaaf, E., Dentener, F., Van Aardenne, J. A., 
Monni, S., Doering, U., Olivier, J. G. J., Pagliari, V. & Janssens-­
Maenhout, G. 2018. 
Gridded emissions of air pollutants for the period 1970–2012 within Edgar v4.3.2. Earth 
System Science Data, 10, 1987–2013.
Downey, L.  & Strife, S. 2010. Inequality, democracy, and the environment. Organiza­
tion & Environment, 23, 155–188.
EIA. 2019. International energy statistics. Washington, DC: U.S. Energy Information 
Administration.
Fioletov, V., McLinden, C., Krotkov, N., Li, C., Leonard, P., Joiner, J. & Carn, S. 2019. 
Multi-­
satellite air quality sulfur dioxide (SO2) database long-­
term L4 global V1. God­
dard Earth Science Data and Information Services Center (GES DISC) [Online]. Avail­
able: https://disc.gsfc.nasa.gov/datasets/MSAQSO2L4_1/summary.
Fridley, D. & Lu, H. 2016. China energy databook version 9.0. Berkeley, CA: Lawrence 
Berkeley National Laboratory.
Grossman, G. M. & Krueger, A. B. 1995. Economic-­
growth and the environment. Quar­
terly Journal of Economics, 110, 353–377.
Hagene, T. 2010. Everyday political practices, democracy and the environment in a native 
village in Mexico city. Political Geography, 29, 209–219.
IMF. 2019. World economic outlook database October 2019 [Online]. Available: https://
www.imf.org/en/Publications/SPROLLs/world-­
economic-­
outlook-­
databases#sort=
%40imfdate%20descending.
Institute for Health Metrics and Evaluation. 2018. Global burden of disease (GBD) 2017 
study [Online]. Available: https://www.thelancet.com/journals/lancet/article/PIIS0140-
­
6736(18)32279-­
7/fulltext.
Kaufmann, D.  & Kraay, A. 2019. The worldwide governance indicators 2019 update: 
Aggregate governance indicators 1996–2018 [Online]. Available: https://papers.ssrn.
com/sol3/papers.cfm?abstract_id=1682130.
Li, C., McLinden, C., Fioletov, V., Krotkov, N., Carn, S., Joiner, J., Streets, D., He, H., Ren, 
X. R., Li, Z. Q. & Dickerson, R. R. 2017. India is overtaking China as the world’s largest 
emitter of anthropogenic sulfur dioxide. Scientific Reports, 7.
Li, Q.  & Reuveny, R. 2006. Democracy and environmental degradation. International 
Studies Quarterly, 50, 935–956.
Lu, Z., Zhang, Q. & Streets, D. G. 2011. Sulfur dioxide and primary carbonaceous aero­
sol emissions in China and India, 1996–2010. Atmospheric Chemistry and Physics, 11, 
9839–9864.
Marshall, M. G., Gurr, T. R. & Jaggers, K. 2019. Polity IV project: Political regime char­
acteristics and transitions, 1800–2018. Center for Systemic Peace [Online]. Available: 
http://www.columbia.edu/acis/eds/data_search/1080.html.


16  Introduction
Midlarsky, M. I. 1998. Democracy and the environment: An empirical assessment. Journal 
of Peace Research, 35, 341–361.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National Statistics Bureau & Ministry of Ecology and Environment. 2019. China statisti­
cal yearbook on environment 2018. Beijing, China: China Statistics Press.
Neumayer, E. 2002. Do democracies exhibit stronger international environmental commit­
ment? A cross-­
country analysis. Journal of Peace Research, 39, 139–164.
Njeru, J. 2010. ‘Defying’ democratization and environmental protection in Kenya: The 
case of Karura forest reserve in Nairobi. Political Geography, 29, 333–342.
Payne, R. A. 1995. Freedom and the environment. Journal of Democracy, 6, 41–55.
Pellegrini, L. & Gerlagh, R. 2006. Corruption, democracy, and environmental policy – an 
empirical contribution to the debate. The Journal of Environment & Development, 15, 
332–354.
Stern, D. I. & Common, M. S. 2001. Is there an environmental Kuznets curve for sulfur? 
Journal of Environmental Economics and Management, 41, 162–178.
Sundberg, J. 2003. Conservation and democratization: Constituting citizenship in the Maya 
biosphere reserve, Guatemala. Political Geography, 22, 715–740.
Torras, M. & Boyce, J. K. 1998. Income, inequality, and pollution: A reassessment of the 
environmental Kuznets curve. Ecological Economics, 25, 147–160.
U.S. EPA. 2019. Air pollutant emissions trends data: Criteria pollutants national tier 1 for 
1970–2018. Washington, DC: U.S. EPA.
Walker, P. A. 1999. Democracy and environment: Congruencies and contradictions in 
Southern Africa. Political Geography, 18, 257–284.
Wendling, Z. A., Emerson, J. W., Esty, D. C., Levy, M. A., De Sherbinin, A. et al. 2018. 
2018 environmental performance index. New Haven, CT: Yale Center for Environmen­
tal Law & Policy.
Winslow, M. 2005. Is democracy good for the environment? Journal of Environmental 
Planning and Management, 48, 771–783.
Zheng, B., Tong, D., Li, M., Liu, F., Hong, C. P., Geng, G. N., Li, H. Y., Li, X., Peng, L. Q., 
Qi, J., Yan, L., Zhang, Y. X., Zhao, H. Y., Zheng, Y. X., He, K. B. & Zhang, Q. 2018. 
Trends in China’s anthropogenic emissions since 2010 as the consequence of clean air 
actions. Atmospheric Chemistry and Physics, 18, 14095–14111.


1  
Centralized political will
Which governmental affairs can become national priorities and their relative rank­
ings are highly centralized in the Chinese context without democracy. In contrast 
to the path argued by Payne (1995), in which a democracy develops its political 
will regarding the environment, China has taken a different route. The state is far 
more dominant in China than it is in a democracy. Even nongovernmental organi­
zations (NGOs) in China actively seek alliances with the government (Hsu, 2010). 
The lack of free elections also reduces the need for the government to directly 
respond to the public’s demands.
The Chinese Communist Party holds tremendous authority in deciding, for 
example, how important environmental protection is among all governmental 
affairs. The party is closely intertwined with the Chinese government, but they 
are also very different. The party makes key decisions while the government takes 
almost all implementation tasks. Although the party has about 90 million mem­
bers and is organized into multiple levels, the authority is very much centralized 
upward and eventually into the Central Committee. The 19th cohort was inau­
gurated in October 2017 after the corresponding National Party’s Congress. It 
has 204 members, and their tenure will last for five years, until 2022 when the 
next National Party’s Congress convenes to form another Central Committee. It 
further forms the Political Bureau, currently with 25 members, and then, most 
crucially, the 7-­
member Standing Committee as China’s top leadership. Many of 
these members, but not all, also hold positions in the Chinese government. Two 
are most important. The secretary general, currently Xi Jinping, is at the center 
and generally assumes the position of president in the Chinese government. The 
prime minister, currently Li Keqiang, leads the Chinese administration. This hier­
archy ensures China’s high degree of centralization in making most important 
decisions. The Chinese government and, specifically, environmental administra­
tion are mainly focused on environmental policy making and implementation. On 
those prioritized governmental affairs that decisions have been made by the top 
leadership of the Party, the government is in charge of implementation.
In the past seven decades after the establishment of the People’s Republic of 
China, each top leadership of the Chinese Communist Party has left a phrase in 
2	

Political will


18  Political will
the party’s Constitution, with their ideologies written as the party’s “guiding com­
pass,” which not only guides their own leadership’s rule but also summarizes a 
legacy. The line has become longer over time to include “Mao Zedong Thoughts,” 
“Deng Xiaoping Theory,” “Three Representativeness” (headed by President Jiang 
Zemin), “Scientific View of Development” (headed by President Hu Jintao) and 
“Socialistic Thoughts with Chinese Characteristics in the Xi Jinping Era” (Chi­
nese Communist Party, 2017).
This chapter mainly focuses on how the political will for environmental pro­
tection has evolved since the 15th Central Committee was formed in 1998. The 
period transcended three top leaderships of the party, including President Jiang 
Zemin and Prime Minister Zhu Rongji (1998–2002), President Hu Jintao and 
Prime Minister Wen Jiabao (2003–2012) and President Xi Jinping and Primer 
Minister Li Keqiang (2013–2022). Chapter 3 examines the environmental gov­
ernance of the Chinese government for implementing the political will.
2  
Economy, jobs and the environment (1998–2002)
The period was under the 15th Central Committee and the leadership of President 
Jiang Zemin and Prime Minister Zhu Rongji. Although China’s environmental 
pollution had already reached high levels, more urgent issues were present to 
suppress forceful political will for environmental protection. Difficult economic 
conditions slowed down energy consumption to witness a decline of sulfur diox­
ide (SO2) emissions in the 9th Five-­
Year Plan (1996–2000; Figure 1.8).
China’s economy was still at the early stage of industrialization, while the 
Asian financial crisis of 1997 hit China badly. In comparison with the previous 
years (1992–1997), the average annual gross domestic product (GDP) growth rate 
declined significantly from 11.8% to 8.3% (Figure 2.1). GDP per capita was still 
at low levels, US$3,185 (purchasing power parity [PPP] in 2011 US$) in 1998 and 
US$4,276 in 2002, or 7.4% and 9.3% of the U.S. levels, respectively (Figure 1.6). 
Job creation was more important than GDP growth. As will be introduced in 
Chapter 3, the following year, 1998, witnessed China’s several far-­
reaching fun­
damental reforms with a key focus on state-­
owned enterprises and a better-­
defined 
boundary between the state and the market. Many of these state-­
owned enter­
prises were substantially overstaffed and loss-­
making and operated more like gov­
ernmental agencies and less like market-­
oriented entities. The Chinese financial 
sector and, specifically, the state-­
owned banks had extremely high levels of bad 
debts. This period witnessed large-­
scale privatization and the bankruptcy of small 
and medium-­
sized state-­
owned enterprises, mainly in the secondary sector. As a 
result, the secondary sector shed 8.7 million jobs from 1998 to 2002 to reflect the 
massive reform’s side effects (Figure 2.1). Overall, 3.3 million jobs were annu­
ally added to the secondary and tertiary sectors. With many more people entering 
than leaving the workforce as indicated in the rapidly enlarging age group of 
15-­
 to 64-­
year-­
olds (Figure 2.2), many of the unemployed should have returned 
to rural regions as the primary sector added 18.0 million jobs over the five years 
(Figure 2.1). China’s job and demographic structures were still dominated by the 


Political will  19
–16
–12
–8
–4
0
4
8
12
16
–90
–60
–30
0
30
60
90
1992–1997
1998–2002
2003–2007
2008–2012
2013–2018
Annual increase/decrease of nonprimary jobs 
(million) & annual GDP growth rate (%)
)
n
o
i
l
l
i
m
(
 
s
b
o
j
 
f
o
 
e
s
a
e
r
c
e
d
/
e
s
a
e
r
c
n
I
 
c
i
d
o
i
r
e
P
Primary
Secondary
Tertiary
Nonprimary jobs per year (right)
GDP growth rate (right)
Figure 2.1  
Sectoral employment changes and GDP growth rates across China’s administrations
Source: National Bureau of Statistics (2019).
600
650
700
750
800
850
900
950
1,000
1,050
1,100
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
1980
1985
1990
1995
2000
2005
2010
2015
Population in the 15–64 age group (million)
 
n
o
i
t
a
l
u
p
o
p
 
r
o
 
t
n
e
m
y
o
l
p
m
e
 
f
o
 
s
e
r
a
h
S
Year
Primary
Secondary
Tertiary
Urban population
Rural population
Population (15–64; right)
Figure 2.2  
Employment and population structures in China
Source: National Bureau of Statistics (2019).


20  Political will
primary sector and rural regions. The primary sector’s share of total jobs hovered 
stably between 49.8% and 50.1%, while the share of the rural population declined 
from 68.1% in 1997 to 60.9% in 2002 (Figure 2.2). As a result, most Chinese 
were less exposed to seriously polluted urban air pollution because they were not 
breathing urban air.
Generally speaking, over the period from 1998 to 2002, environmental pro­
tection was ranked high neither in governmental affairs nor by society. In the 
aftermath of the Asian financial crisis, economic downturn and unemployment 
were more imminent and highly politicized problems to occupy the top leader­
ship’s mind. This top leadership’s guiding ideology, “three representativeness,” 
was mainly engaged in expanding the party’s base from the conventional working 
class to other categories of the society. Environmental protection did not occupy 
any important role in this ideology, while slower industrial development also 
reduced the deterioration rate of environmental pollution.
3  
SARS and the prioritization of environmental 
protection (2003–2012)
Over the ten years (two terms with the 16th and 17th Central Committee) between 
2003 and 2012, when President Hu Jintao and Prime Minister Wen Jiabao were in 
power, China added 75.6 million new jobs in the secondary sector and 67.3 million 
in the tertiary sector, while the primary sector had a decrease of 108.7 million jobs 
(Figure 2.1). To keep pace with the growing working-­
age population, the annual 
increase of nonprimary jobs was 14.3 million, much faster than the 3.3 million 
new jobs annually between 1998 and 2002 (Figure 2.1). The primary sector still 
accounted for 50.0% of China’s overall employment in 2002 and remained the 
largest among the three sectors in 2007 at 40.8%. China’s entry into the World 
Trade Organization in 2001 and multiple major economic reforms led to unprec­
edented growth in the economy, energy consumption and pollution. The global 
financial crisis of 2008 did exert great and negative impacts on China’s economy 
to slow it down. Comparing the two Hu-­
Wen administrations (2003–2007 and 
2008–2012), the annual economic growth rate came down from 11.7% to 9.4%, 
and the annual increase of nonprimary jobs was from 15.9 million to 12.7 million.
Environmental protection started to emerge as a nationally prioritized govern­
mental affair. The 11th Five-­
Year Plan (2006–2010) was completely formulated 
and implemented under this top leadership of the party. It not only included the 
10% mitigation goals of SO2 and chemical oxygen demand but actually achieved 
them (National People’s Congress, 2011), defying challenges from the rapid 
growth of economy and energy consumption and reversing the humiliating fail­
ures in the 10th Five-­
Year Plan (Figure 1.8). Deeper mitigation of SO2 emissions 
followed in later years, while the turning point of environmental protection hap­
pened within this period (Figure 1.8).
Society might not have been ready to put the environment as a high priority 
with strong cleanup determination. For example, despite the dire situation of air 
pollution, a survey in 2010 by Gallup, a U.S. research-­
based consulting company, 


Political will  21
found that only 26% of the Chinese were dissatisfied, and 73% were satisfied, 
with the air quality (English, 2010). The potentially insufficient support from soci­
ety for pollution mitigation, if China were a democracy, might not have generated 
strong political will.
The much stronger political will for environmental protection reflected more 
the intention of the top leadership of the party. As is examined in detail in Chap­
ter 4, the direct involvement of the top leadership was crucial in enacting the 
environmental goals in the 11th Five-­
Year Plan after the failures in the 10th Five-­
Year Plan. The 16th Central Committee was formed in November 2002 at the 16th 
National Party’s Congress. The Standing Committee of the Political Bureaus was 
headed by Secretary General Hu Jintao and included Wen Jiabao. In March 2003, 
at the 10th National People’s Congress, they assumed the positions of president 
and prime minister, respectively, in the Chinese central government. In the transi­
tional period between these two key conferences, they had only party leadership 
roles but officially not those later government positions.
SARS (severe acute respiratory syndrome), a new infectious disease, emerged 
almost exactly over this transitional period in November  2002 and became 
increasingly damaging over the winter (WHO, 2003). The timing of the devastat­
ing pandemic coincided well with the top leadership’s search for a new ideology 
to distinguish themselves from their predecessors. This public health crisis taught 
a painful lesson to the Chinese leadership that public goods should be prioritized 
together with economic development. The overemphasis of the latter may actu­
ally backfire to result in slow economic growth as the Chinese economy was sig­
nificantly damaged, especially in the second quarter of 2003, by the impacts of 
the SARS pandemic (Rawski, 2005; Hai et al., 2004; Xu et al., 2009). After the 
pandemic was over and society returned to normal, a new ideology was gradually 
formed, titled “Science View of Development,” to emphasize development from 
multiple aspects to achieve a “harmonious society.” Environmental protection is 
a natural extension from public health and became one pivotal component in this 
new development direction.
The authorities of the top leadership and this new ideology were hardly distin­
guishable. From this perspective, whether China could achieve serious mitiga­
tion of environmental pollution and reverse the deterioration trend became more 
politicized. This significantly increased the political will of the top leadership 
to start taking environmental protection into the inner core of key governmental 
affairs. In other words, the political will resulted from a more top-­
down rather 
than bottom-­
up approach, although the pressure from society grew over the years.
Key international events also played a role in shaping China’s environmen­
tal protection. One of the most important events over the Hu–Wen administra­
tions was the 29th Summer Olympic Games in August 2008. To ensure good air 
quality over Beijing, China shut down many polluting factories across several 
neighboring provinces around Beijing. Environmental information was increas­
ingly available over this period. The Internet played a key role in distributing 
information. The U.S. Embassy in Beijing started monitoring fine particulate 
matter (PM2.5) levels in 2008. Environmental NGOs, notably the IPE (Institute of 


22  Political will
Public & Environmental Affairs) that was established in 2006, started systemati­
cally collecting, publicizing and distributing environmental information to the 
public.
4  
The sustainability of environmental political will 
(2013–present)
President Xi Jinping and Prime Minister Li Keqiang assumed their top leadership 
roles of the Chinese Communist Party in November 2012 at the 18th National 
Party’s Congress and then of the central government in March 2013 at the 12th 
National People’s Congress. As usual, the change of leadership did raise questions 
about whether environmental protection could be further strengthened or weak­
ened in relation to new economic conditions and new leaders’ ideas. The Chinese 
economy entered a “new normal,” or a stabilized but lower level after 2013. The 
annual GDP growth rate from 2013 to 2018 was 7.0%, even lower than the level 
during the aftermath of the Asian financial crisis. Nevertheless, the economy had 
already reached a wealthier status before the new leadership came into power and 
the progress since 2013 has also been decent. In 2002, China’s GDP per capita 
was US$4,276 (PPP in 2011 US$), and it increased to US$11,049 in 2012 and 
US$16,098 in 2018 (IMF, 2019). The ratios between China and the United States 
were 9.3%, 21.8% and 28.8%, respectively.
With the working-­
age population stabilized at about 1  billion people (Fig­
ure 2.2), job creation was still at a healthy pace with 10.7 million new nonpri­
mary jobs added annually. Over the six years, in total, the tertiary sector added 
82.5 million new jobs, while the secondary and primary sectors had 18.5 million 
and 55.2 million fewer jobs (Figure 2.1). In contrast to the economic downturn 
between 1998 and 2002, Chinese labor did not return to rural regions. The tertiary 
sector accelerated significantly to account for 46.3% of all employment in 2018, 
up from 36.1% in 2012 (Figure 2.2). The primary sector accounted for 31.4% of 
all jobs in 2013 and further declined to only 26.1% in 2018 (Figure 2.2). Further­
more, China has been urbanizing fast to have 53.7% of people in urban regions 
in 2013. In 2018, the urbanization rate further increased to 59.6% (Figure 2.1). In 
other words, China’s employment and demographic structures have been much 
more urbanized, which also brought more people under the impacts of more pol­
luted urban air.
Rapid economic development and escalating living standards have been key 
foundations for the Chinese people to maintain support to the Chinese Communist 
Party’s holding of power. The Chinese middle class has expanded rapidly in the 
past decades to indicate that this demand was to a great extent satisfied. Given 
the higher income and more intimate exposure of an average Chinese to urban air 
pollution, society started to place environmental quality at a significantly higher 
priority than before. The balance between environmental protection and economic 
growth has thus been shifting gradually toward the former’s end. In the leader­
ship transitional period in January 2013, North China suffered from severe smog 
with PM2.5 concentration levels reaching hazardous levels (Wang et al., 2014). 


Political will  23
Although Hebei Province had worse air quality, it was Beijing, as China’s capi­
tal, that attracted most international and domestic attention. Air pollution mitiga­
tion started to be widely recognized as one crucial demand by society. People are 
increasingly willing to sacrifice economic opportunities for a better environment. 
Environmental protection and especially urban air quality have been significantly 
politicized, now by society, and implicitly linked with the legitimacy of the Chi­
nese Communist Party as the ruling political party.
In addition, environmental protection also became a more and more visible 
business to create jobs and economic outputs. The initial efforts in the Hu–Wen 
administrations started to bear fruits. China’s environmental and renewable 
energy industries are competitive not only domestically but also internationally 
(Xu, 2013; Zhu et al., 2019). They have grown into another pollical force to push 
for China’s continuous environmental cleanup. For example, China now has the 
world’s largest solar, wind and electric vehicle industries. They play increas­
ingly counterbalancing roles against those who are concerned about the negative 
impacts of environmental protection on their businesses.
In the formation of this top leadership’s governing ideology, the party was also 
keen to significantly elevate the priority of environmental protection. The 18th 
National Party’s Congress in 2012 emphasized ecological civilization, while the 
19th National Party’s Congress in 2017 listed “harmony of people and nature” as 
one of the 14 basic things to insist on, which primarily features ecological civi­
lization and the “two mountains” theory. Previously, in the relationship between 
economic development and environmental protection, the statement was that we 
want not only “gold and silver mountain” but also “clear water and green moun­
tain.” In other words, these two were placed as trade-­
offs to each other. The new 
statement of “two mountains” became that “clear water and green mountain” are 
“gold and silver mountain.” The pursuit of environmental quality became equiva­
lent to economic development. Environmental protection does offer opportunities 
to satisfy the demands for both economic development and a better environment, 
for example, when new industries emerge for pollution mitigation or resource 
conservation. Environmental policies have also been playing an active role in 
encouraging innovation and economic transformation, as elaborated in greater 
detail in Chapter 7.
Overall, in this period, both the top leadership of the party and society came 
together with a common and prioritized stake in a cleaner environment. Environ­
mental protection is increasingly politicized to form an unprecedented political will 
for pollution mitigation. The top leadership should meet the growing demand of 
the society for not just economic growth but also environmental cleanup. Because 
“ecological civilization” is a key component in the top leadership’s “Socialis­
tic Thoughts with Chinese Characteristics in the Xi Jinping Era,” significant 
improvement of environmental quality also became crucial for the establishment 
of this new governing ideology. New economic opportunities and environmental 
industries have been serving as an increasingly visible force to counterbalance 
the negative economic impacts of environmental protection. The rapid growth of 
income has also transformed society’s preference between economic development 


24  Political will
and environmental quality. They are crucial forces to make the political will sus­
tainable, even when top leadership changes again in the future.
References
Chinese Communist Party. 2017. The party’s constitution (Revised by the 19th national par­
ty’s congress) [Online]. Available: https://www.researchgate.net/publication/322421053_
The_19th_Congress_of_the_Communist_Party_of_China_and_Its_Aftermath.
English, C. 2010. More than 1 billion worldwide critical of air quality. Washington, DC: 
Gallup.
Hai, W., Zhao, Z., Wang, J. & Hou, Z. G. 2004. The short-­
term impact of SARS on the 
Chinese economy. Asian Economic Papers, 3, 57–61.
Hsu, C. 2010. Beyond civil society: An organizational perspective on state – NGO relations 
in the people’s Republic of China. Journal of Civil Society, 6, 259–277.
IMF. 2019. World economic outlook databases October  2019 [Online]. Available: 
https://www.imf.org/en/Publications/SPROLLs/world-­
economic-­
outlook-­
databases#
sort=%40imfdate%20descending.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
Payne, R. A. 1995. Freedom and the environment. Journal of Democracy, 6, 41–55.
Rawski, T. G. 2005. SARS and China’s economy. In: Kleinman, A. & Watson, J. L. (eds.) 
SARS in China: Prelude to pandemic? Stanford: Stanford University Press.
Wang, Y. S., Yao, L., Wang, L. L., Liu, Z. R., Ji, D. S., Tang, G. Q., Zhang, J. K., Sun, Y., 
Hu, B. & Xin, J. Y. 2014. Mechanism for the formation of the January 2013 heavy haze 
pollution episode over central and eastern China. Science China-­
Earth Sciences, 57, 
14–25.
WHO. 2003. Update 95 – SARS: Chronology of a serial killer [Online]. Available: https://
www.who.int/csr/don/2003_07_04/en/.
Xu, Y. 2013. Comparative advantage strategy for rapid pollution mitigation in China. Envi­
ronmental Science & Technology, 47, 9596–9603.
Xu, Y., Williams, R. H. & Socolow, R. H. 2009. China’s rapid deployment of SO2 scrub­
bers. Energy & Environmental Science, 2, 459–465.
Zhu, L., Xu, Y. & Pan, Y. J. 2019. Enabled comparative advantage strategy in China’s solar 
PV development. Energy Policy, 133.


1  
Evolution of environmental administration
Environmental protection in China could be traced back to the United Nations 
Conference on the Human Environment in June 1972 in Stockholm, Sweden. In 
the turmoil of the Cultural Revolution (1966–1976) and after the United Nations 
voted in 1971 that the People’s Republic of China is the sole representative of 
China, China sent an official delegation to this conference. In August 1973, the 
First National Conference on Environmental Protection was held to mark that 
environmental protection had formally been recognized as a governmental affair. 
However, in the early stage of the Cultural Revolution, the leaders and organiza­
tions of the Chinese Communist Party and the Chinese government at various 
levels were generally toppled by Red Guards (hong wei bin) and Rebels (zao fan 
pai). Although the Chinese government was rebuilt at a later stage, the primary 
focus was not on economic or social affairs but on class struggle. As a result, 
China did not demonstrate a significant conflict between economic development 
and environmental protection because neither mattered.
When the Cultural Revolution ended in 1976, after a short transitional period, 
China entered the new era of Reform and Open-­
up in December 1978. Economic 
development quickly gained prominence in governmental affairs, while class 
struggle and other political affairs wound down. Soon afterward, the impacts of 
economic development on environmental quality started to emerge. As a pub­
lic affair that requires governmental intervention, environmental protection was 
announced as one Basic National Policy in the Second National Conference on 
Environmental Protection from 31 December 1983 to 7 January 1984. Since then, 
dedicated governmental entities have been established in the Chinese government 
to regulate and implement environmental protection. The agency in the Chinese 
central government that oversees environmental protection has evolved over the 
years in terms of organization, power and jurisdiction. The authority of environ­
mental protection has been increasingly strengthened in the past four decades. 
In 1984, the State Environmental Protection Agency was established under the 
then Ministry of Construction. In 1988, it was pulled out to be directly led by the 
State Council, thus with an elevated status and authority at the vice-­
ministry level. 
Environmental protection then became not just an issue for one single ministry 
3	

Environmental governance


26  Environmental governance
but also one key state affair that was widely relevant and one level closer to the 
center of the governmental authority.
China’s key reforms in the past four decades have one crucial central theme for 
adjusting the relationship between the state and the market. There were essentially 
no real markets in the Cultural Revolution because markets were deemed as too 
capitalistic. Prices did not reflect any balance between demand and supply but 
were decided directly by the government. Purchases should be accompanied by 
permits, not just money. Despite fluctuations, the overall trend in the past dec­
ades was the reemergence, creation and maturity of various markets, as well as 
the refocusing of the state from everything to strategic and public affairs. With 
the government giving up its original authority, prices have become much better 
indicators of supply and demand balances. The production and consumption are 
increasingly guided by market signals and little by orders from central planners.
In the 1998 reform of the State Council that featured a better-­
clarified demarca­
tion between the state and the market, 14 ministries that mainly took direct charge 
of the economic sectors were abolished, and 4 new ministries were formed. The 
government then became more focused on public affairs and much less on direct 
management of businesses. In this reform, the then State Environmental Protec­
tion Agency was promoted to the ministerial level and renamed the State Environ­
mental Protection Administration (SEPA). Other significant reforms in the same 
period marked the reorganization of large state-­
owned enterprises, the privatiza­
tion of small ones and the widened space for private businesses.
Although environmental protection gained increasingly higher statuses in the 
previously mentioned reforms, it was still kept away from the core of the Chinese 
central government, in which the State Council is in charge of the country’s routine 
administration. According to China’s Constitution, the State Council comprises 
the following members: prime minister and deputies, state councilors, ministers, 
directors of commissions and the auditor general. Although the SEPA had been 
elevated to the ministerial level after the 1998 reform, it was not a ministry, and 
thus, its director was not a constitutional member of the State Council. He or she 
could be present in the meeting only by invitation, with much constrained author­
ity on other ministries’ affairs even if they may be closely relevant to environmen­
tal protection. The 2008 reform became crucial when the SEPA was reorganized 
as the Ministry of Environmental Protection and thus became a formal comprising 
ministry of the State Council. This reform indicated that environmental protection 
was recognized as one of the key governmental affairs. The enhanced authority 
also gave the new ministry and its counterparts in local governments more force­
ful power in enacting and implementing environmental policies.
In 2018, a new round of major reforms further concentrated environmental 
authorities that scattered in several ministries into the newly formed Ministry of 
Ecology and Environment (MEE; State Council, 2018). Climate change was nota­
bly transferred out of the National Development and Reform Commission to fall 
under the MEE’s jurisdiction. The MEE now combines the original functions of 
(1) Ministry of Environmental Protection, (2) climate change and mitigation under 
the National Development and Reform Commissions, (3) groundwater pollution 


Environmental governance  27
under the Ministry of Land and Resources, (4) water environment management 
under the Ministry of Water Resources, (5) agricultural pollution under the Minis­
try of Agriculture, (6) ocean environment under the State Oceanic Administration 
and (7) south–north water diversion project’s environmental protection under its 
office. This reform further strengthened the authority of environmental protection. 
The significantly wider duties are expected to create better synergies among their 
regulations and solutions.
2  
Chain of command for environmental protection
Environmental protection administration in China has four major levels, being 
central, provincial, municipality and county. The latter three levels are generally 
categorized as local governments, although provincial governments are often not 
directly involved in local administration. Local governments take primary respon­
sibilities for implementing environmental policies and achieving environmental 
protection. The sequential reforms at the central government were followed by 
corresponding reforms in local governments that generally resemble the struc­
tures of the central government, despite differences contingent on local contexts.
Although the MEE and its predecessors had a clear chain of command under 
the State Council of the central government, it is not straightforward whether local 
environmental protection bureaus (EPBs) should be led by corresponding local 
governments or environmental protection agencies at a higher governmental level 
for achieving more effective environmental administration. On one hand, environ­
mental protection is far beyond the authority of the EPBs to involve industrial pol­
icy, urban planning and other policies. Environmental enforcement heavily relies 
on other agencies and budget allocation from local governments. Accordingly, it 
is reasonable to have local governments as the major office-­
bearers. On the other 
hand, local governments may create barriers to environmental protection due to 
the possible conflicts between economic growth and environmental protection. If 
local EPBs could be vertically controlled, they may better serve the purpose of 
environmental protection as local economic growth is not the central considera­
tion of upper-­
level EPBs.
China’s administrative reform in the past four decades has one key trend: more 
and more remaining governmental authorities are being decentralized from the 
central government to local governments, especially regarding the regulation of 
economic activities and the provision of social public goods such as health care, 
education, housing and urban/rural infrastructure and community services. In 
the environmental administrative system, the chain of command for local EPBs 
reflected such a decentralization trend to recognize that environmental protection 
is generally a localized governmental affair. In 1999, the Department of Organi­
zation of the Chinese Communist Party reformed the institutional arrangements 
and specified that the leaders of local EPBs should be jointly appointed by pri­
marily local governments and, to a lesser extent, upper-­
level EPBs (Department 
of Organization of the Central Committee of the Communist Party of China, 
1999). The “double administration” arrangement aimed for a balance between the 


28  Environmental governance
vertical – or “tiao” based on the function of environmental administration – and 
horizontal – or “kuai” based on the location of environmental protection. The 
1999 reform was accordingly mainly horizontally oriented with decentralization. 
EPBs were under local governments with their directors and budgets controlled 
by their corresponding local governments. They were also advised by EPBs in the 
immediate upper-­
level governments.
The general decentralization trajectory in the past decades also engaged 
another argument for recentralization. In the era of Reform and Open-­
up, local 
governments often have to face the conflicts between environmental protec­
tion and economic development. In evaluating the performance of local leaders, 
economic indicators tended to occupy much heavier weights than environmen­
tal protection, especially in the early years. Accordingly, for the sake of the 
local economy, the environment has often been sacrificed. Together with the 
rising status of environmental protection in the central government as described 
earlier, environmental protection started to climb higher on the priority list. 
The MEE as well as its predecessors and local counterparts are less bound by 
such evaluation because economic development is not their direct job duty, but 
environmental protection is their primary responsibility. In 2016, another major 
and more centralization-­
oriented reform was initiated with several provinces 
for pilot implementation (The General Office of the CPC Central Committee 
and The General Office of the State Council, 2016). The authority of appoint­
ing local EPB leaders and their budgets were shifted more toward upper-­
level 
EPBs. Environmental monitoring and inspection agencies were more directly 
controlled vertically.
3  
Division of labor for policy making and implementation
Environmental agencies in China’s central and local governments have distinct 
functional focuses. The central government is mainly in charge of policy mak­
ing. It also supervises local governments, primarily provincial governments, 
for implementing environmental protection. Provincial governments heavily 
focus on policy making within their individual provinces. They also adapt poli­
cies from the central government to their own situations and supervise mainly 
municipality governments. The municipality level has a further diminished 
capacity in policy making and a much heavier focus on policy implementa­
tion, while the tasks of county governments fall almost exclusively on the 
implementation of policies from the upper levels within localized contexts. 
Implementation is primarily the responsibility of municipality and county gov­
ernments. They can also make decisions that are applied within their specific 
jurisdictions, mainly on how to implement policies with greater efficiency and 
effectiveness.
The clear division of labor among the four levels of governments is reflected 
in their composition of environmental protection personnel. Their personnel com­
positions are accordingly different among the four categories: administration, 
inspection, monitoring and others. “Administration” mainly refers to the MEE 


Environmental governance  29
in the central government as well as corresponding bureaus at the three levels 
of local governments. “Inspection” personnel are those who work in Inspection 
Bureaus, while “monitoring” personnel are based in Monitoring Stations. “Oth­
ers” are the remaining personnel, such as those in the Academy of Environmental 
Sciences and Academy of Environmental Planning at the four levels. They pro­
vide research and expertise to support environmental policy and decision making. 
Between 2004 and 2015, using available data, the compositions at the four gov­
ernmental levels were largely stable (Figure 3.1). The only significant exception 
is the share of “inspection” at the central level, which experienced a dramatic 
increase in 2009 (Figure 3.1).
The environmental authority in the central government is not organized for 
shouldering implementation tasks but primarily for making policies and super­
vising local governments (SCOPSR, 2018). At the central level, “others” is the 
largest category. It accounted for 64.1% of all 3,023 environmental protection 
personnel in 2015, while the share was over 80% before 2009 (Figure 3.1). 
Their dominant share indicates that environmental policy making in China 
requires and has been receiving significant intellectual support. “Administra­
tion” hosted only 362 personnel in 2015, and its share remained stable at about 
12% over the period between 2004 and 2015 based on available data. After 
the 2018 reform and the reorganization, the new MEE was allowed to have 
478 personnel, the addition for accommodating expanded functions (SCOPSR, 
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2004 2009 2014 2004 2009 2014 2004 2009 2014 2004 2009 2014
Total personnel (number)
l
e
n
n
o
s
r
e
p
l
a
t
o
t
f
o
e
r
a
h
S
Year
Administration
Inspection
Monitoring
Others
Total personnel (right)
Central
Provincial
Municipality
County
Figure 3.1  
Environmental protection personnel at four governmental levels in China (for 
2004–2015 using available data)
Source: Ministry of Environmental Protection (2002–2016).


30  Environmental governance
2018). Partly as a result of the establishment of six Regional Supervision Cent­
ers, “inspection” had a major shift with its personnel jumping from 41 in 2008 
to 294 in 2009 and further to 542 in 2015. The 2018 reform further formalized 
and upgraded them into Regional Supervision Bureaus, with a total person­
nel capacity of 240 officers (SCOPSR, 2018). “Monitoring” had about 6% of 
all personnel throughout the years. The inspection and monitoring personnel 
provide crucial data support for supervising the environmental protection per­
formance of local governments.
China’s provincial environmental authorities are also structured to have a 
heavy focus on policy making and supervision and less a focus on direct pol­
icy implementation. At the provincial level, “others” remains the largest to have 
46.4% of all provincial environmental protection personnel in 2015 (Figure 3.1). 
It is the largest category to reflect the desired functions in policy making. “Moni­
toring” occupied 19.9%, which was a decline from 26.9% in 2004 (Figure 3.1). 
The share of “inspection” increased from 6.2% in 2004 to 9.0% in 2015, but the 
increase was much less significant in comparison with that at the central level 
(Figure 3.1). Between monitoring and inspection, the central government now 
puts more emphasis on inspection while provincial governments have a heavier 
focus on monitoring.
The municipality and county levels are structured with much lower capacities 
for policy making and primarily for policy implementation. At the municipal­
ity level, “monitoring” is the largest category, with 34.5% of all its environmen­
tal protection personnel in 2015 (Figure  3.1). “Inspection,” “administration” 
and “others” each took about one fifth of the personnel. Their primary tasks 
are, accordingly, sharply different from the central and provincial levels, with a 
heavy focus on actually implementing policies, although they also build decent 
knowledge support for initiating policy innovations. The county level is almost 
exclusively for implementation, with “others’ accounting for only 6.4% of envi­
ronmental protection personnel in 2015. “Inspection” became the largest func­
tional group with 37.0% of personnel, while “monitoring” and “administration” 
had 28.0% and 28.6%, respectively (Figure 3.1).
The differentiated functions of environmental authorities at the four levels indi­
cate that China’s environmental protection requires their close cooperation. From 
the MEE in the central government to environmental protection bureaus at the 
county level, policy making is more concentrated at the top while implementation 
is mainly at the lower levels. However, their cooperation should not be taken for 
granted, even though China has a conventional image of top-­
down administra­
tion. As examined in later sections, local governments and their leaders have their 
own self-­
interests. If environmental policy implementation is against such inter­
ests, the implementation will not be expected to be effective. As expected from 
China’s weak rule of law, regardless of how stringent environmental policies are, 
their weak implementation was one of the primary reasons that led to China’s 
environmental crises. Without forceful enforcement efforts of local governments 
and widespread compliance of polluting sources, environmental cleanup cannot 
be realized.


Environmental governance  31
4  
Decentralized policy making
From social, economic, industrial and environmental perspectives, China has 
been evolving at an astonishing speed in the past four decades. Laws, policies 
and regulations should continuously adapt to the rapidly changing situations. As 
indicated in the World Bank’s governance indicators, the rule of law in China has 
not been well established (Kaufmann and Kraay, 2019). Laws and courts have not 
been playing important roles in daily environmental protection. Instead, policies 
and regulations are much more closely relevant.
Laws in China are enacted by the National People’s Congress. They tend to take 
many years to formulate, enact or amend. For example, the Law of Environmental 
Protection is the basic law to regulate China’s environmental protection. It was first 
enacted in 1989, and then it took 25 years to get amended in 2014. However, China’s 
environmental conditions and pollution had dramatically changed during the 25 years, 
which should have indicated that the older version was seriously outdated. In addition, 
a variety of specific laws are enacted to regulate individual categories of the environ­
ment. For example, the Law of Atmospheric Pollution Prevention and Control was 
enacted in 1987, and two amendments have been done since then, in 2000 and 2015 
(two other minor corrections were done in 1995 and 2018; National People’s Congress, 
2018). The Law of Water Pollution Prevention and Control was enacted in 1984. Only 
one amendment has been done in 2008, while two minor corrections were made in 
1996 and 2017 (National People’s Congress, 2017). Accordingly, many environmental 
policies in China do not have clear corresponding items in environmental laws.
The slow motion of laws’ enactment and amendment may make them at a great 
distance from the rapidly evolving pollution conditions. This could also partly 
explain why many of China’s policies were applied before their legal foundations 
were established. For example, the eco-­
compensation policy got its legal backing 
only in 2014 in the newly amended Environmental Protection Law, but by then, it 
had already been experimented with and applied widely (Wang et al., 2016). Fur­
thermore, courts do not play significant roles in environmental enforcement and 
compliance. The laws are often written to mainly state principles without enough 
details for direct implementation. The situations reflect China’s situation of weak 
rule of law. The Chinese central government does not file lawsuits against local 
governments for not implementing laws and its policies.
In addition, China’s environmental laws are often intentionally vague in order to 
allow more flexibility for the administration, while environmental policies contain 
more implementable details. Compared with the U.S. Clean Air Act Amendments 
(CAAA, 1990), China’s goal and initial plan were much less detailed. The CAAA 
clearly developed a cap-­
and-­
trade system with detailed rules and schedules (The 
U.S. Congress, 1990). Such details were absent in China’s plans. China’s laws are 
often drafted by a ministry, not the National People’s Congress. For example, a 
key task of the MEE is to draft laws and regulations on environmental protection 
(SCOPSR, 2018). A vague law can provide a legal foundation but not constrain 
the enactment of policies. For example, China’s Law of Atmospheric Pollution 
Prevention and Control entitles the environmental authority to enact ambient air 
quality standards and effluent emission standards without further clarification on 


32  Environmental governance
when and how (National People’s Congress, 2000). The State Council gets the 
legal power to collect effluent emission charges and the freedom to enact any 
relevant regulation (National People’s Congress, 2000).
Other than the National People’s Congress, the State Council can enact Regula­
tions. Various ministries, as well as their internal departments, frequently churn 
out policies, standards, projects and other incentives/commands that are relevant 
to environmental protection. (For simplicity, they are referred to as environmen­
tal policies in the following discussion.) Local governments and their environ­
mental authorities also hold the right to enact their own environmental policies 
or to adapt those from the central government into their corresponding jurisdic­
tions and contexts. As shown in Figure 3.1, local governments, especially at the 
provincial and (to a lesser extent) municipality levels, do have decent capacities 
for making policies. All these environmental policies could have very different 
scopes, stringency, instruments, targets and intellectual support. In comparison to 
laws, environmental policies are much more flexible. Its enactment takes much 
less time and faces much lower hurdles. The entire process is also much less cen­
tralized with numerous governmental bodies at ministerial and local levels who 
can independently enact environmental policies. China’s weak rule of law indi­
cates that these policies are rarely challenged in courts or through other channels 
by affected interest groups, although their legal foundation might be porous and 
shaky in vague and slowly updated environmental laws. In order to understand 
China’s rules for environmental protection, laws are not the most reliable sources.
Nevertheless, ironically the weak status of rule of law in China further strength­
ened the decentralization of environmental policy making. Although the National 
People’s Congress is distinctly different from that in a democracy, laws are nev­
ertheless more stable and more authoritative than policies by the administration. 
Laws are based on wider participation, and the legislative process is more transpar­
ent. If strong enough incentives are present, the variety of policy-­
making entities 
at different levels will be able to actively innovate new policies, learn the lessons 
and experiences from other policy making entities, adapt top-­
down policies and 
adopt policies from other regional contexts. Not all policy making is necessarily 
backed by sound research or intellectual support. Nevertheless, the decentralized 
policy making makes active bottom-­
up policy innovation and diffusion possible.
5  
Decentralized policy implementation
From the perspectives of human resources and fiscal expenditures, China’s capac­
ity for environmental policy implementation is heavily tilted toward local govern­
ments, rather than the central government.
5.1  
Decentralized human resources
Policy implementation demands substantially more resources and personnel than 
policy making. Corresponding to the designed focuses between policy making 
and implementation, most of China’s environmental protection officials are at 


Environmental governance  33
the municipality and county levels. China had 232,388 government employees 
on environmental protection in 2015, a 62.8% increase from 142,766 in 2001 to 
reflect the elevated priority of environmental protection in all government affairs. 
The distributions across the four levels of governments have been quite consist­
ent over the years, with 1.3%, 6.8%, 21.5% and 63.1% of the total environmen­
tal protection personnel in 2015 in central, provincial, municipality and county 
governments, respectively. Corresponding to the four categories, the municipality 
and county levels accounted for 92.5% personnel for administration, 97.0% for 
inspection, 94.6% for monitoring and 69.5% for others (Figure 3.1). As a result, 
the environmental authorities at the central and even the provincial levels do not 
have an adequate human resource capacity to implement environmental policies 
in millions of polluting sources that are scattered in China’s wide geographic ter­
ritories (Ministry of Environmental Protection et al., 2010).
5.2  
Decentralized fiscal expenditure and centralized fiscal revenue
Fiscal revenue and expenditure are other key perspectives for understanding the 
central–local relationship in China. The governmental expenditure-­
to-­
GDP ratio in 
China is not high in comparison to that in developed countries. In 2018, the ratio was 
24.5%, in which the central government accounted for 3.6% and local governments 
20.9% (Figure 3.2). The ratio dropped significantly from 26.8% to 11.1% from 1980 
0.0%
5.0%
10.0%
15.0%
20.0%
25.0%
30.0%
1980
1985
1990
1995
2000
2005
2010
2015
 
o
i
t
a
r
 
P
D
G
 
o
t
 
e
r
u
t
i
d
n
e
p
x
e
/
e
m
o
c
n
i
 
l
a
t
n
e
m
n
r
e
v
o
G
Year
Income: Central
Expenditure: Central
Income: Total
Income: Local
Expenditure: Total
Expenditure: Local
Figure 3.2  
Governmental revenue and expenditure to GDP ratios by central and local gov­
ernments in China
Source: National Bureau of Statistics (2019).


34  Environmental governance
to the mid-­
1990s but has since gradually recovered (Figure 3.2). The ratio between 
governmental revenue and GDP had a similar trend, initially falling from 25.3% in 
1990 to 10.2% in 1995 and then rising back to 20.4% in 2018 (Figure 3.2). The gaps 
between revenue and expenditure indicate fiscal surplus or deficit.
In the current fiscal arrangement, the central government has far more revenue 
than it spends while the local governments in general have to rely on fiscal trans­
fers from the central government for meeting their expenditures. In 2018, the cen­
tral government received 46.6% of total general fiscal revenue but accounted for 
only 14.8% of total fiscal expenditures. Local governments, in contrast, received 
nearly half of the revenue but had to shoulder 85.2% of the expenditures.
The fiscal relationship between the central and local governments have expe­
rienced dramatic changes in the past four decades. In 1980, local governments 
directly received an overall revenue of 87.5 billion RMB (current price), but their 
spending was 56.2 billion RMB (National Bureau of Statistics, 2019). In contrast, 
the central government had a revenue of 28.4 billion RMB but spent 66.7 billion 
RMB. It was the central government, not local governments, that spent most of 
the government budget, ranging from 52.5% to 55.0% between 1980 and 1984 
(Figure 3.2). Accordingly, the central government ran a huge deficit, and local 
governments, a huge surplus. The fiscal transfer was then from local govern­
ments to the central government. It reflected that China’s governance remained 
very much centralized in the immediate years after the Cultural Revolution. The 
central government was directly engaged in providing a significant proportion 
of government services and subsidies. Correspondingly, fiscal expenditures were 
required to support such a provision.
The situation was dramatically changed in 1985. When the governmental 
expenditure-­
to-­
GDP ratio started to drop significantly together with market-­
oriented economic reforms, the central government saw a much steeper decline 
(Figure 3.2). The budgets for both the central and local governments became indi­
vidually more balanced (Figure 3.3). The expenditures of the central and local 
governments were only 3.3% above and 2.1% lower than their revenues in 1985. 
Local governments since then have consistently accounted for more than 60% of 
total governmental expenditures, dwarfing the share of the central government. 
Although local governments’ fiscal conditions remain generally balanced in the 
following years, the central government again started to see a widening gap. In 
1993, its expenditures exceeded revenue by 37.0% while its shares in total gov­
ernment revenue and expenditures had dropped to 22.0% and 28.3%, respectively. 
The budget deficit of the central government fiscally constrained it from exerting 
authority on rich provinces and tackling widening regional disparities across the 
country.
In China’s central–local fiscal relationship, 1994 was a crucial watershed when 
a fundamental tax reform entered into effect in January (State Council, 1993). The 
central government’s share of total governmental revenue skyrocketed to 55.7% 
in 1994 while its share of expenditures remained at 30.3%. For the first time, the 
central government ran a budget surplus, with revenue exceeding expenditures by 
65.7%. In contrast, local governments’ fiscal revenue could cover only 57.2% of 


Environmental governance  35
their expenditures. Then a large fiscal transfer became necessary from the central 
government to local governments. With further decentralization of governmental 
affairs and service provision, this newly formed central–local fiscal relationship 
has been kept increasingly entrenched in the past two decades. In 2018, local gov­
ernments accounted for 85.2% of expenditures but only 53.4% of revenue. The 
gap has significantly widened.
The current central–local relationship that features significant fiscal transfer 
from the central government to local governments reflects their differentiated 
roles in policy making and implementation as discussed earlier. The central 
government is primarily in charge of policy making while the implementa­
tion is largely in the hands of local governments. The former requires much 
less expenditure than the latter. All provinces have their expenditures exceed­
ing revenues, but poor provinces tend to rely on the central government’s 
fiscal transfer much more than rich ones (Figure  3.4). For example, Tibet’s 
governmental revenue covered only 11.7% of its expenditures in 2018, while 
the revenue–expenditure gap for Shanghai was only 14.9%. Accordingly, the 
central government could use fiscal transfer as an incentive for local govern­
ments to implement policies or achieve goals that are enacted from the top. 
It is one of the key incentives that the central government can rely on for the 
cooperation of local governments.
–10.0%
–8.0%
–6.0%
–4.0%
–2.0%
0.0%
2.0%
4.0%
6.0%
8.0%
1980
1985
1990
1995
2000
2005
2010
2015
Budget balance (% of GDP)
Year
Total
Central
Local
Figure 3.3  
Budget balance of central and local governments in China as a proportion of 
GDP
Source: National Bureau of Statistics (2019).


36  Environmental governance
With increasing decentralization in the economic reform, more and more budg­
etary items were shifted with local governments as primary entities of governmen­
tal expenditures. Reflecting the division of governmental affairs, the central and 
local governments now have distinct responsibilities on a variety of expenditure 
items. Foreign affairs and national defense are two budgetary items that the cen­
tral government takes almost exclusive responsibility to account for 99.5% and 
98.1%, respectively, of total governmental expenditures. Of the central govern­
ment’s expenditure in 2018, 33.8% was devoted to national defense. Grain storage 
is for the country’s food security, and thus, the central government remained more 
important, being responsible for 66.8% of all governmental expenditures in 2018 
(Figure 3.5). Science and technology is another classical category of public good 
that the market underinvests in to require public expenditures, in which the central 
government took a share of 37.5% in 2018 (Figure 3.5). Health care and urban 
and rural communities are almost exclusively the responsibility of local govern­
ments. Environmental protection was responsible for 2.9% of total governmental 
expenditures in 2018 (Figure 3.6), while local governments accounted for 93.2% 
(Figure 3.5). It occupied 3.1% of local governments’ expenditures and 1.3% of the 
central government’s (Figure 3.6).
The expenditure structures between the central and local governments have 
remained generally unchanged for environmental protection in the past decade. 
However, this largely decentralized budgetary item has also witnessed signs of 
Beijing
Tianjin
Hebei
Shanxi
Inner Mongolia
Liaoning
Jilin
Heilongjiang
Shanghai
Jiangsu
Zhejiang
Anhui
Fujian
Jiangxi
Shandong
Henan
Hubei
Hunan
Guangdong
Guangxi
Hainan
Chongqing
Sichuan
Guizhou
Yunnan
Tibet
Shaanxi
Gansu
Qinghai
Ningxia
Xinjiang
–100.0%
–90.0%
–80.0%
–70.0%
–60.0%
–50.0%
–40.0%
–30.0%
–20.0%
–10.0%
0.0%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
160,000
)
s
e
r
u
t
i
d
n
e
p
x
e
 
l
a
i
c
n
i
v
o
r
p
 
f
o
 
%
(
 
e
c
n
a
l
a
b
 
t
e
g
d
u
B
GDP per capita (RMB/person)
Figure 3.4  
Governmental budget balance by provinces as a proportion of governmental 
expenditures in 2018
Source: National Bureau of Statistics (2019).


Environmental governance  37
0.0% 10.0% 20.0% 30.0% 40.0% 50.0% 60.0% 70.0% 80.0% 90.0% 100.0%
 Total
 Interregional aid
 Urban & rural communities
 Health care
 Agriculture, forestry & water
 Social security & employment
 Commercial services
 Education
 Environmental protection
 Housing
 Resource exploration & information
 Culture, sports & communication
 General public service
 Transportation
 Public security
 Land, ocean & meteorology
 Others
 Science & technology
 Debt interest
 Financial
 Debt issuance
 Grain storage
 National defense
 Foreign affairs
Share of governmental expenditures
Local’s share
Central’s share
Figure 3.5  
The central and local governments’ shares of expenditures by budgetary items 
in 2018
Source: National Bureau of Statistics (2019).
0%
5%
10%
15%
20%
25%
30%
35%
 Interregional aid
 Urban & rural communities
 Health care
 Agriculture, forestry & water
 Social security & employment
 Commercial services
 Education
 Environmental protection
 Housing
Resource exploration & information
 Culture, sports & communication
 General public service
 Transportation
 Public security
 Land, ocean & meteorology
 Others
 Science & technology
 Debt interest
 Financial
 Debt issuance
 Grain storage
 National defense
 Foreign affairs
Share of governmental expenditures
Local
Central
Total
Figure 3.6  
Central, local and overall governmental expenditures by budgetary items in 2018
Source: National Bureau of Statistics (2019).


38  Environmental governance
slight recentralization. Recent reforms as described earlier reflected and enabled 
the central government to be keener in improving environmental quality and more 
directly involved in supervising local governments. Environmental protection has 
been listed as a separate budgetary item in the data from the China Statistical 
Yearbook since the 2008 edition (for 2007 data). Its share in total governmental 
expenditures has inched up from 2.0% in 2007 to 2.7% in 2010 and then fluctu­
ated narrowly to reach 2.9% in 2018. The share in local governments’ budgets has 
also been quite stable, within a narrow range between 2.5% and 3.2% over the 
period. However, the central government had a significant shift, allocating a much 
greater share of its budget for environmental protection. It ranged between 0.2% 
and 0.5% from 2007 to 2013 but then jumped to 1.5% in 2014 and has remained 
at the level since then (Figure 3.7). Correspondingly, the central government’s 
share in total environmental protection expenditures was lifted from 2.9% in 2013 
to 9.0% in 2014, while the local governments’ share dropped although their envi­
ronmental protection expenditures were increased every year in absolute terms.
The significant uplifting in 2014 indicates that environmental protection has 
been increasingly prioritized in China’s public affairs (Figure 3.7). The additional 
budget mainly corresponded to the strengthened functions of top-­
down supervi­
sion, monitoring and inspection of local governments’ performance. Because the 
shares in governmental expenditure for China as a whole and for local governments 
0.0%
1.0%
2.0%
3.0%
4.0%
5.0%
6.0%
7.0%
8.0%
9.0%
10.0%
2007
2008
2009
2010
2011
2012
2013
2014
2015
2016
2017
2018
s
e
r
u
t
i
d
n
e
p
x
e
 
l
a
t
n
e
m
n
r
e
v
o
g
 
n
i
 
s
e
r
a
h
S
Year
Central’s share in environmental
protection expenditures
Environmental protection’s share in total
expenditures
Environmental protection’s share in
central’s total expenditures
Environmental protection’s share in
local’s total expenditures
Figure 3.7  
Shares in governmental expenditures
Source: National Bureau of Statistics (2019).
Note: The National Statistical Yearbook listed environmental protection as a separate budgetary item 
for the first time in 2007.


Environmental governance  39
did not change significantly over the period and especially in 2014, environmen­
tal administrative capacities were not expected to be upgraded disproportionally 
against other governmental affairs. The emphasis on environmental protection 
thus targeted the relationship between the central and local governments to more 
effectively mobilize implementation capacities and to assign a heavier weighting 
to environmental protection relative to local economic development.
6  
Centralized and decentralized personnel management
According to the Chinese Constitution, local leaders are elected by corresponding 
local People’s Congress. Then they are supposed to mainly please their local elec­
torate. Because China’s weak rule of law does not ensure the local implementa­
tion of environmental laws and policies from the National People’s Congress and 
the central government, the central government should only be able to exert very 
constrained authority over the selection of local government leaders and what 
governmental affairs they decide to pursue in their local contexts. Even if local 
leaders refused to implement policies from the top, only local People’s Congress 
can remove them. However, this very decentralized arrangement presents a sharp 
contrast with reality. Far-­
reaching reforms in the past four decades have featured 
economic reforms on the relationship between the state and the market and admin­
istrative reforms on the relationship between the central and local governments, 
but the relationship between the Chinese Communist Party and the Chinese gov­
ernment has witnessed fewer changes. In the 1980s, their separation was debated 
and explored in tentative reforms, but the progress has been much slower.
The party plays a crucial role in shaping the central–local leadership relation­
ship in reality. The party and the Chinese government have overlapped organiza­
tions in the governmental bureaucracy, while the party is even more prevalent 
to be present in enterprises and other nongovernmental organizations. Although 
local leaders should be elected by local People’s Congress, the party, and espe­
cially its Department of Organization, controls the nominations. For the four gov­
ernmental levels, each level has the authority to appoint leaders at one lower level. 
For example, the party’s Department of Organization at the central level controls 
the nomination of provincial-­
level leaders (including those in central ministries). 
Each provincial Department of Organization nominates municipality-­
level lead­
ers within the province. The appointment decisions are in the hands of their cor­
responding party committees. As a result, personnel decisions are one crucial 
channel for the central government to influence local governments. Numerous 
studies have confirmed that China does put governance performance in the deci­
sions to promote or remove officials, especially local government leaders (Li and 
Zhou, 2005; Zhou, 2007). Without the party’s role, China’s governance would be 
substantially different from the current institutional arrangement.
Furthermore, a reform did bring a major change in this personnel relationship 
with significant decentralization. In 1984, the Central Committee of the party 
reformed its personnel management system (Gao and Zou, 2007). Before then, the 
Department of Organization at each level managed two levels down. For example, 


40  Environmental governance
the Central Department of Organization was in charge of nominating and manag­
ing leaders at the provincial and municipality levels. After the reform, the leaders 
at the municipality level are left to the sole responsibility of the Provincial Depart­
ment of Organization, while the Central Department of Organization only takes 
care of the provincial-­
level leaders. As a result, the provincial leaders will have 
much stronger control of their staff and other local government leaders below 
them. Such reform substantially reinforces local leaders’ authorities within their 
jurisdictions. The arrangement coincides with the decentralization of governmen­
tal affairs and expenditures but still maintains a powerful channel through the 
party for the central government to control local leaders.
References
Department of Organization of the Central Committee of the Communist Party of China. 
1999. On reforming the institutions of managing environmental officials. Beijing, China: 
Central Committee of the Communist Party of China.
Gao, X. & Zou, Q. 2007. Research on intra-­
party democracy – evaluation from history and 
reality. Shandong, China: Qingdao Press.
The General Office of the CPC Central Committee & The General Office of the State 
Council. 2016. Guiding advice on the pilot vertical reform in sub-­
provincial monitoring, 
inspection and enforcement agencies. Beijing, China: CPC Central Committee, State 
Council.
Kaufmann, D.  & Kraay, A. 2019. The worldwide governance indicators 2019 update: 
Aggregate governance indicators 1996–2018 [Online]. Available: https://info.world­
bank.org/governance/wgi/.
Li, H. B. & Zhou, L. A. 2005. Political turnover and economic performance: The incentive 
role of personnel control in China. Journal of Public Economics, 89, 1743–1762.
Ministry of Environmental Protection. 2002–2016. Annual statistical report on the envi­
ronment in China. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection, National Statistics Bureau  & Ministry of Agri­
culture. 2010. Public report on the first national census of polluting sources. Beijing, 
China: Ministry of Environmental Protection, National Statistics Bureau.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National People’s Congress. 2000. Law of atmospheric pollution prevention and control 
of people’s republic of China. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2017. Law of water pollution prevention and control. Beijing, 
China: The 4th Conference of the 10th National People’s Congress.
National People’s Congress. 2018. Law of atmospheric pollution prevention and control 
of people’s republic of China. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
SCOPSR. 2018. The function, internal organization and personnel of the ministry of ecol­
ogy and environment. Beijing, China: SCOPSR.
State Council. 1993. Decision on implementing the tax sharing mechanism in fiscal man­
agement. Beijing, China: State Council.
State Council. 2018. Reform plan on the state council. Beijing, China: State Council.


Environmental governance  41
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. 
Congress.
Wang, H., Dong, Z., Xu, Y. & Ge, C. 2016. Eco-­
compensation for watershed services in 
China. Water International, 41, 271–289.
Zhou, L. 2007. Governing China’s local officials: An analysis of promotion tournament 
model. Economic Research Journal, 7, 36–50.


1  
Goals in China’s Five-­
Year Plans
China’s top leadership has gradually gained strong enough political will for 
environmental protection over the past decades (Chapter  2). However, the 
decentralization of policy making and, to a greater extent, policy implementa­
tion requires the cooperation between the central and local governments to 
realize the environmental political will with concrete improvement of environ­
mental quality and pollution mitigation (Chapter 3). This chapter is devoted 
to understanding how the entire Chinese government, from central to local 
governments, is mobilized through environmental goals, especially in Five-­
Year Plans.
Goals have been widely used in governance. For example, UNFCCC (United 
Nations Framework Convention on Climate Change) defines its goal as “stabili­
zation of greenhouse gas concentrations in the atmosphere at a level that would 
prevent dangerous anthropogenic interference with the climate system” (United 
Nations, 1992). President Barack Obama set up a goal to withdraw all U.S. troops 
from Iraq by the end of 2011 (DeYoung, February 28, 2009). Many studies are 
about environmental goals, including those on negotiating goals, distributing 
goals (Chakravarty et al., 2009), policies to achieve goals (such as on emission tax 
and cap-­
and-­
trade) and technological achievability of goals (Pacala and Socolow, 
2004).
A theoretical foundation of using goals as a governance tool can be traced to 
studies in social psychology: through experiments on individuals, the impact of 
various goals on task performance is examined. Locke et al. (1981) reviewed 
the literature and concluded that “specific and challenging goals lead to higher 
performance than easy goals, ‘do your best’ goals, or no goals.” Furthermore,
goal setting is most likely to improve task performance when . . . the subjects 
have sufficient ability, . . . feedback is provided to show progress in rela­
tion to the goal, rewards such as money are given for goal attainment, the 
experimenter or manager is supportive, and assigned goals are accepted by 
the individual.
(Locke et al., 1981)
4	

Mobilizing the government1


Mobilizing the government  43
In the experiments, goals are distributed to individuals and individuals try to accom­
plish the goals. The situation is not much different from an environmental goal 
in a big country like China. The Chinese central government plays a similar role 
as experimenters: it decides a goal and distributes it to local governments. Three 
components could be distinguished: (1) goal setting, (2) goal distribution and (3) 
goal attainment. Goal setting refers to what type of goals should be set up and how 
stringent they are. Because a global or national goal often requires the cooperation 
of different political or administrative entities, goal distribution is necessary. For 
example, a global goal of carbon dioxide (CO2) mitigation should be distributed to 
individual countries, and a Chinese national goal should be distributed to provinces. 
Furthermore, these goals need to be accepted before serious efforts are made. The 
third component of a goal process focuses on evaluating goal attainment. Strong-­
enough incentives should be put into place to mobilize goal implementers.
The seven-­
decade history of the People’s Republic of China can be divided 
into two periods: a centrally planned economy in the first three decades and a later 
era of market-­
oriented economic reforms. Since the 1950s, originally adopted 
from the Soviet Union, Five-­
Year Plans have become pivotal to guide China’s 
economic development. Although China’s economy was strictly state-­
controlled 
before the economic reforms began in 1978, only the first of the earliest five Five-­
Year Plans was actually completed (Liu et al., 2006). The other four were not able 
to be performed due to frequent political movements, with the Cultural Revolu­
tion as the most notable one (Liu et al., 2006).
Five-­
Year Plans gained momentum only in the second period when China tried 
to establish a market-­
oriented economy. Starting from the 6th Five-­
Year Plan 
(1981–1985), China has gradually formed a set of rules to design these plans 
(State Council, 2005b). The 11th Five-­
Year Plan (2006–2010) was the first to 
change its name from “jihua” (more forceful plans) to “guihua” (more directional 
plans). Goals are the most important indicators in the Plans. From the 11th Five-­
Year Plan, goals are distinguished into foreseeable ones (such as the growth rates 
of gross domestic product [GDP] and population) and legally binding ones (such 
as pollutant mitigation; National People’s Congress, 2006). In addition, China’s 
Five-­
Year Plans are not just one document but a system composed of many layers. 
For example, for the nation as a whole, there was a National 11th Five-­
Year Plan 
that included a 10% reduction goal of sulfur dioxide (SO2) emissions. Another 
11th Five-­
Year Plan on Environmental Protection provided further details. At one 
more layer lower, the 11th Five-­
Year Plan on Acid Rain and SO2 Pollution Control 
specifically addressed the mitigation of SO2 emissions. There were also 11th Five-­
Year Plans at all governmental levels.
Goals are playing more and more prominent roles in China’s environmental 
protection, especially in Five-­
Year Plans, to mobilize local governments and the 
Chinese bureaucracy. If expressed in percentage terms, the baseline year is the 
final year of the previous Five-­
Year Plan. For example, China’s energy intensity 
goal in the 11th Five-­
Year Plan (2006–2010) was a 20% reduction (National Peo­
ple’s Congress, 2006); it indicates that China planned to reduce energy intensity, 
or energy consumption per unit of GDP, by 20% in 2010 from the 2005 level.


44  Mobilizing the government
In regulating SO2 emissions that mainly come from the burning of coal, China 
relies on absolute emission goals, which were a 3.8% increase, a 10% reduction, a 
10% reduction, an 8% reduction and a 15% reduction, respectively, for the 9th, 10th, 
11th, 12th and 13th Five-­
Year Plans (National People’s Congress, 2001, 2006, 2011; 
NEPA et al., 1996; National People’s Congress, 2016). The actual growth rates of 
SO2 emissions were a 15.8% reduction, a 27.8% increase, a 14.3% reduction and a 
14.9% reduction, respective for the 9th, 10th, 11th and 12th Five-­
Year Plans, indi­
cating goal attainment in all but the 10th Five-­
Year Plan (National Statistics Bureau 
and Ministry of Ecology and Environment, 2019). This chapter specifically analyzes 
the 10% reduction goal of SO2 emissions in the 11th Five-­
Year Plan as it reversed 
the humiliating failure in the 10th Five-­
Year Plan. The national quantitative goal 
was centrally set up to involve the Chinese top leadership and the then State Envi­
ronmental Protection Administration (SEPA, presently the Ministry of Ecology and 
Environment). The mitigation tasks were distributed to provincial and other local 
governments with their individual goals. Mechanisms were put into place to moni­
tor the goal compliance statuses of local governments and take enforcement actions 
for their cooperation. Goals have also been rapidly evolving to reflect the status and 
intended emphasis of SO2 mitigation and air pollution control.
2  
Centralized goal setting
2.1  Setting up the national goal
China’s goal process involves three overlapping cycles: Five-­
Year Plans, National 
Party’s Congresses and National People’s Congresses. The 11th Five-­
Year Plan for­
mally started in 2006 and concluded in 2010. The 16th National Party’s Congress 
lasted from October 2002 to October 2007. The 10th National People’s Congress 
lagged half a year behind, from March 2003 to March 2008. The 11th Five-­
Year Plan 
did not begin until the middle of the two Congresses. Under China’s present political 
reality, the two Congresses have a reasonable sequence. The National Party’s Con­
gress selects party leaders. After a further distribution of power, these leaders assume 
various governmental jobs in the following National People’s Congress. The first 
gatherings of these two Congresses are mainly about determining the leadership of 
the party and the country. Then China’s leaders reshuffle every five years. As a result, 
the three cycles are actually two: the Five-­
Year Plans and the change of leadership.
The cycles have existed in the present form for about four decades, espe­
cially since 1992. The most stable cycle is the Five-­
Year Plan. All Five-­
Year 
Plans are targeted for five years, even in the most irrational period of the Cul­
tural Revolution. Since the 3rd Five-­
Year Plan (1966–1970), the period has been 
consecutive. The National Party’s Congress formed its own five-­
year cycle in 
1977, and the National People’s Congress, in 1978. But the leadership change did 
not match the Congresses’ cycles until 14 years later. Jiang Zemin was formally 
elected as the secretary general of the party in 1992 and the president of China in 
1993. Since then, China’s top leaders also have established their five-­
year cycles, 
formally synchronized with the Congresses.


Mobilizing the government  45
The cycles of Five-­
Year Plans do not match China’s change of leadership. The 
anchor year of a Five-­
Year Plan is the previous year before the plan starts. But 
because the plan has to be formed before all information in the anchor year is 
known and China’s SO2 emissions are very volatile, relative goals are much bet­
ter than absolute goals to address the huge uncertainty. China’s failure to attain 
the 10% reduction goal of SO2 emissions in the 10th Five-­
Year Plan (2001–2005) 
may partly reflect the mismatch among cycles. A new administration took full 
charge in March 2003 when the 10th Five-­
Year Plan had been going on for over 
two years. Almost immediately afterward, China’s SO2 emissions went out of con­
trol. During 2001–2002, SO2 emissions went down by 3.4%, but in the remaining 
three years (2003–2005), the emissions surged by 32.3% (SEPA, 2001–2009). 
On the other hand, the sharp contrast was not obvious from the perspective of 
economic growth. In annual terms, China’s economy expanded at an annual rate 
of 8.7% in the first two years and 10.2% later (National Bureau of Statistics of 
China, 1999). Although the surge could be simply a coincidence with the change 
of leadership, if 2003 through 2005 had been under the same administration as in 
2001–2002, the result might be different due to a better unification of planning 
and implementation.
The Outline of the National 11th Five-­
Year Plan on Economic and Social 
Development (hereafter referred to as the Outline) was the title of an official 
document ratified by the National People’s Congress, the nominally highest 
authority in China, in March  2006 (National People’s Congress, 2006). The 
10% reduction goal of SO2 emissions was clearly included to be legally bind­
ing. The process to reach the Outline can be divided into three periods: (1) mid-­
2003 to December 2004, concluded with the formation of The Basic Thoughts 
of the National 11th Five-­
Year Plan (hereafter referred to as the Basic Thoughts; 
National Development and Reform Commissions, or NDRC, was responsible); 
(2) February 2005 to October 2005, ended with the ratification of The Sugges­
tions on Designing the National 11th Five-­
Year Plan (hereafter referred to as 
the Suggestions; the Central Committee of the Chinese Communist Party was 
in charge); (3) October 2005 to March 2006, indicated by the enactment of the 
Outline (State Council took the hold).
The Basic Thoughts contemplated the strategic direction of the Outline. This 
idea-­
framing period was initiated in mid-­
2003 and completed by the end of 2004 
(Xinhua News Agency, 2006; NDRC, 2003). For environmental protection, the 
job of the 11th Five-­
Year Plan was to “decelerate the trend of ecological and envi­
ronmental deterioration and strengthen the ability of sustainable development” 
(NDRC, 2005). The wording clearly differs from, for example, “improving envi­
ronmental quality.” It may be reflected later in the Basic Thoughts on Environ­
mental Protection with a flat SO2 emission goal proposed (SEPA, 2006d; Chinese 
Academy for Environmental Planning [CAEP], 2004).
The then named SEPA was responsible for writing the 11th Five-­
Year Plan for 
Environmental Protection. The SEPA understood the specific difficulty of control­
ling SO2 emissions. For example, in 2002, Wang Xinfang, a deputy administrator 
of the SEPA, admitted that it was hard to achieve the 10% reduction goal of SO2 


46  Mobilizing the government
emissions in the 10th Five-­
Year Plan (2001–2005; Wang, 2002). The final result in 
2005 confirmed his concern: goals on other pollutants were either met or slightly 
missed, but SO2 emissions were 27.8% higher than the level in 2000 and 42% higher 
than the original goal (Zou et al., 2006). The SEPA distributed The Basic Thoughts 
on Environmental Protection on December 23, 2004, and proposed a flat goal for 
the 11th Five-­
Year Plan (SEPA, 2006d; CAEP, 2004). The midterm assessment on 
the 10th Five-­
Year Plan that was completed in 2004 could have played a guiding 
role in the proposal: the available data showed an 8.2% increase of SO2 emissions 
in 2003 compared with those in 2000 (SEPA, 2001–2009). The midterm assessment 
believed that the 10% reduction goal had fallen out of reach but still expected that 
SO2 emissions in 2005 could remain the same as the level in 2000 (Zou et al., 2004).
With the tentative Basic Thoughts, the top leadership in the Central Committee of 
the Chinese Communist Party got directly involved. The period was formally initi­
ated with the establishment of a high-­
profile drafting team on February 16, 2005, 
headed directly by Premier Wen Jiabao (Xinhua News Agency, 2005). A prominent 
feature is the multiple meetings presided by President Hu Jintao in the Political 
Bureau or its Standing Committee and by Premier Wen Jiabao in the drafting team 
(Xinhua News Agency, 2005). The Suggestions was finally passed and endorsed 
on October 11, 2005, by the Central Committee of the Chinese Communist Party 
(Xinhua News Agency, 2005). Sharply different from the Basic Thoughts, the Sug­
gestions clearly declared to “reduce total emissions of pollutants,” which essentially 
indicated a goal of improving environmental quality (Xinhua News Agency, 2005).
After the Suggestions tightened the goal for environmental protection in Octo­
ber 2005, the third period started with the establishment of a drafting team that 
comprised various ministries in the central government (Xinhua News Agency, 
2006). An expert committee was summoned to comment on the drafts of the Out­
line (Ma, 2005). The public was also consulted for advice (Ma, 2005). Presi­
dent Hu Jintao and Premier Wen Jiabao organized several meetings to discuss the 
drafts (Xinhua News Agency, 2006). In November 2005, the SEPA drafted a plan 
on acid rain and SO2 emission control (SEPA, 2005). Although SO2 emissions in 
2004 had been 13% higher than the 2000 level, the 10% reduction goal for the 
11th Five-­
Year Plan first appeared (SEPA, 2005, 2001–2009). On December 3, 
2005, State Council enacted Decisions on Realizing Scientific View of Develop­
ment and Strengthening Environmental Protection (State Council, 2005a), which 
linked the new ideology of Scientific View of Development with environmental 
protection. It confirmed the importance of environmental protection in the estab­
lishment of the new ideology. When the 4th Conference of the 10th National Peo­
ple’s Congress was in session, the Outline was submitted on March 5, 2006, and 
approved on March 14, 2006 (Xinhua News Agency, 2006).
2.2  
Methods of goal setting
A Five-­
Year Plan anchors at the previous year of its planning period. For exam­
ple, a goal in the 11th Five-­
Year Plan (2006–2010) is to compare 2010 with 
2005. In practice, the anchor year’s data cannot be fully utilized in setting up 


Mobilizing the government  47
the goals. China generally published environmental data for the previous year in 
around June (SEPA, 2001–2009). Although the public may get the information 
later than the Chinese government, several months could elapse for the collection 
and compilation of data. Accordingly, the anchor year’s information cannot be 
fully employed in planning but has to be the foundation for the next Five-­
Year 
Plan. China’s annual change of SO2 emissions varied greatly: the 2004 emissions 
were 4.5% up from the 2003 level, but the figure surprisingly jumped 13.1% 
in 2005 (SEPA, 2001–2009). At the same time, however, the economic growth 
rates were quite stable with 10.1% in 2004 and 11.4% in 2005 (National Bureau 
of Statistics, 2019). Because of the substantial volatility, the absence of data in 
the most relevant and important anchor year could cause significant trouble in 
calibrating goals.
Two components were important in setting up China’s SO2 emission goals in 
Five-­
Year Plans: long-­
term goals and appropriate mitigation paces. China relied 
on a concept called “environmental capacity” to decide long-­
term SO2 emission 
goals (Yang et al., 1998, 1999). “Environmental capacity” refers to the upper-­
limit 
emissions of a pollutant without degrading a kind of environmental quality below 
a minimum level. The environmental capacity for SO2 emissions is a function of 
three variables: (1) the amount and distribution of SO2 emissions, or emission 
inventories; (2) the transport and sinks of SO2; and (3) an acceptable level of some 
environmental quality. The second variable is largely determined by atmospheric 
circulation and chemistry. The third variable was used as an external choice. If 
society would like to live in a better environment, the limit of ambient SO2 con­
centration could be lowered and SO2 emissions have to be further reduced.
To set up an SO2 goal in a Five-­
Year Plan, China first decided on a long-­
term 
goal and then found an appropriate mitigation pace to attain the goal. The long-­
term goals were determined with models of atmospheric transport and chemistry. 
The implicit long-­
term goal for the 10th Five-­
Year Plan (2001–2005) was 12 mil­
lion tons and was scheduled to get attained in 2020 (Wang, 2002). For the 11th 
Five-­
Year Plan (2006–2010), the long-­
term goal became 18 million tons and the 
goal attainment year would also be 2020 (SEPA, 2005). Although both goals were 
supported by scientific research with different constraint conditions, the signifi­
cant upward revision of the long-­
term goal probably arose as a result of the sharp 
increase in coal use that led to an unanticipated rise of SO2 emissions in the 10th 
Five-­
Year Plan.
The long-­
term goals have certain scientific foundations. China’s Law of Envi­
ronmental Protection clearly holds local governments responsible for local envi­
ronmental quality (National People’s Congress, 1989). Because ambient air quality 
standards are also “mandatory standards” in the Law of Standardization (State 
Council, 1990), local government leaders should be mobilized to enforce SO2 miti­
gation policies if the law were well respected. In 1996, the then State Environmen­
tal Protection Agency enacted ambient air quality standards (NEPA and SBTS, 
1996). Most of China’s land area with economic and human activities should 
have ambient SO2 concentration in annual mean below 0.060 mg/m3. One key 
study showed that only to achieve this average concentration within grid boxes of 



48  Mobilizing the government
0.2° × 0.2°, China has to control its SO2 emissions at 12 million tons (Yang et al., 
1999). Another study for the 11th Five-­
Year Plan selected critical acid deposition 
within grid boxes of 1° × 1° (Zou et al., 2006). Although the number was based 
on several heavy assumptions (most important, the geographical distribution of 
SO2 emission sources), it signaled the stringency of the ambient SO2 concentra­
tion standard. For example, China’s goal in the 11th Five-­
Year Plan was to reduce 
SO2 emissions from 25.5 million tons in 2005 by 10% in 2010, still far above the 
12-­
million-­
ton level (National People’s Congress, 2006).
The distribution of SO2 emissions matters greatly for any national SO2 miti­
gation goal that is based on SO2 concentration. For example, with SO2 concen­
tration of 0.060 mg/m3 as the constraint condition, Shanghai could emit up to 
0.63 million tons of SO2 (Yang et al., 1999), but its actual emissions in 2007 were 
0.50 million tons (Ministry of Environmental Protection, 2008). Then if a pollut­
ing source was located in Shanghai, it would have no necessity to mitigate. But if 
the same source were moved to Jiangsu, a neighboring province with its emission 
limit below actual emissions (Ministry of Environmental Protection, 2008; Yang 
et al., 1999), it would be subject to serious abatement. The 1998 study revealed a 
goal based on SO2 ambient concentration: if not counting the excess environmen­
tal capacity in Tibet compared with its emissions (0.50 million tons vs. 1.5 thou­
sand tons), China’s national goal was to reduce SO2 emissions to about 12 million 
tons (Yang et al., 1999). China planned to attain the goal in 2020 (Wang, 2002). 
The goal for the 10th Five-­
Year Plan was then established as a 10% reduction, or 
18 million tons (SEPA, 2001).
After the big failure in the 10th Five-­
Year Plan on SO2 mitigation, China still 
held 2020 as the attainment year of a long-­
term goal. However, the original goal 
would be too difficult. In 2005, China emitted 25.5 million tons of SO2 (SEPA, 
2001–2009). To achieve the goal of 12 million tons in 2020, a 53% reduction in 
15 years would be required. Even if from the 2004 level when a new goal for the 
11th Five-­
Year Plan was formed, the reduction rate should still be 47% (SEPA, 
2001–2009). By replacing the constraints of SO2 concentration with critical acid 
deposition, a new environmental capacity was worked out to be 17.3 million tons 
(Zou et al., 2006). Then 18 million tons were chosen to be the new long-­
term 
goal (SEPA, 2005). These two long-­
term goals assumed a similar pace of about 2 
to 2.5 million tons reduction per five years. Because of the relatively stable pace 
and a common attainment year of the long-­
term goals, China’s long-­
term goals 
seemed to be reversely decided from current emission levels. Interestingly, both 
long-­
term goals were supported by scientific research. The history could indicate 
that the results of the scientific research were selected beforehand by nonscientific 
factors.
In deciding goals for Five-­
Year Plans, the emission trends in previous years 
were also considered (Wang et al., 2004). Because the 9th Five-­
Year Plan achieved 
a 15.8% reduction (NEPA et al., 1996; SEPA, 2001–2009), even a similar trend 
was thought to be too stringent (Wang et al., 2004). Probably the 10% reduction 
goal was established because it stood between the 15.8% reduction and the origi­
nal goal of a 3.8% increase in the 9th Five-­
Year Plan. A middle ground, closer to 


Mobilizing the government  49
the stringent end, was taken. On the other hand, the same historical trend would 
be too relaxed for the 11th Five-­
Year Plan. SO2 emissions went up by 27.8% in 
the 10th Five-­
Year Plan (SEPA, 2001–2009). Certainly this was not an acceptable 
trend, but it might be an important factor that drove the initial flat goal for the 
11th Five-­
Year Plan (CAEP, 2004). The same principle could have been followed: 
0% change was closer to the stringent end between a 27.8% increase and a 10% 
reduction. As a result, the goal attainment in the previous Five-­
Year Plan should 
have played an important role in framing a goal for the next.
The United States’ goal of SO2 emissions in Clean Air Act Amendments 
(CAAA; 1990) was also expressed in relative terms. Relative to the emission 
level in the anchor year of 1980, SO2 emissions were planned for reduction by 
10 million tons (The U.S. Congress, 1990). Although an intensive 10-­
year study 
was performed in the 1980s (National Acid Precipitation Assessment Program), 
it failed to answer relevant questions for policy making and was not closely con­
nected to the goal-­
setting process (Roberts, 1991; Pouyat and McGlinch, 1998). 
For a fixed long-­
term goal, different anchor years only correspond to different rel­
ative reductions or different expressions of the figures. Furthermore, 1980 was not 
a baseline year for allowance allocation. Rather, 1985 was a much more impor­
tant year with real implications in grandfathering emission permits. However, if 
the 10-­
million-­
ton reduction was fixed, the choice of 1980 did have important 
implications. In 1980, the U.S. emitted 23.5 million tons of SO2 and the figures in 
1985 and 1990 were, respectively, 21.1 and 20.9 million tons (U.S. Environmental 
Protection Agency, 2007). Essentially, the choice of 1985 and 1990 would have 
no difference. But anchoring in 1980 could effectively relax the long-­
term goal by 
about 2.4 to 2.6 million tons. The goal was planned for attainment in 2010. The 
anchor year 1980 was ten years ahead of the legislation and 15 years before the 
program formally started in 1995. Although whether a goal was expressed in rela­
tive or absolute terms matters greatly in China, it was generally not quite relevant 
for the United States’ goal setting. The United States had much less volatility in 
annual SO2 emissions. The burden to achieve the goal – the difference between 
business-­
as-­
usual emissions and the goal – was accordingly much less uncertain 
than China’s. The major benefit of relative terms was to reduce the uncertainty of 
surprising emission growth or reduction. However, less uncertainty in the United 
States and the longer goal cycle did not distinguish this benefit. Furthermore, the 
Acid Rain Program’s goal cycle was much longer than China’s Five-­
Year Plans. 
Because of the well-­
established rule of law, the law ensured that the SO2 mitiga­
tion efforts would continue regardless of who was the president or which political 
party he or she belonged to.
3  
Top-­
down goal distribution
Goal implementation refers to a process for goal implementers to receive, accept 
and work for goal attainment. It is quite different from policy implementation. 
Goal implementation deals with the relationship among different governments or 
their agencies, while policy implementation focuses on the relationship between 


50  Mobilizing the government
the government and polluters, including industrial plants and individuals. Goal 
setters and goal implementers are usually separate in the Chinese government. 
Since goal setters are not directly in charge of achieving the goal, they have 
to find a way to get goal implementers to accept the goal and to work hard for 
it. In order for effective goal implementation, subgoals should be created from 
the national goal to demand an appropriate distribution scheme. The UNFCCC 
defines a principle of sharing the duty of reducing greenhouse gas emissions 
among countries according to “common but differentiated responsibilities and 
respective capabilities” (United Nations, 1992). Which applicable principles 
should be followed has attracted negotiation debates and academic studies 
(Chakravarty et al., 2009; Li, 2010).
A national goal and its distribution to local governments often fall into separate 
decision-­
making processes in the Chinese setting. Taking the SO2 goal in the 11th 
Five-­
Year Plan (2006–2010) as an example, the national 10% reduction goal was 
largely decided by the top leadership of the party, but provincial goals came from 
a bargaining process between the central government – mainly the then SEPA – 
and provincial governments.
3.1  
Goal distribution from the central to provincial governments
Chinese local governments are divided into several levels, mainly provinces, 
municipalities and counties. To implement SO2 emission goals, the central gov­
ernment distributed subgoals to provincial and local governments and issued 
incentives to mobilize their leaders. A good national goal is hard to implement 
without a fair distribution of the burden. After a national goal is framed, provinces 
will negotiate with the central government for their shares of the burden. The 
details of the negotiation and their applied principles are not publicly available but 
could be reversely examined from the outcome.
China qualitatively disclosed principles to distribute the national goal to 
31 provinces. Key influential factors included environmental quality, environmen­
tal capacity, current emission level, economic development status, SO2 mitigation 
capability, requirements of various pollution control plans and regional category 
(west, middle, east; State Council, 2006). An explicit formula was less likely to 
exist that connected these factors with a province’s goal. However, published pro­
vincial information could at least lead to an evaluation of potentially quantitative 
relationships. Econometric analysis was applied here with a linear assumption.
The dependent variable was provincial SO2 emission goals in percentage 
terms: SO2 emission target in 2010 / SO2 emissions in 2005 – 100%. The dis­
tributed national goal, 11.9% reduction, was actually a little more stringent than 
a 10% reduction (State Council, 2006). All provinces combined could only emit 
22.47 million tons, not 22.94 million tons for the nation. The difference (0.47 mil­
lion tons) was reserved for experimenting with SO2 emission cap-­
and-­
trade (State 
Council, 2006).
Independent variables included all those factors indicated by the Chinese govern­
ment (State Council, 2006). Because 2005 was the anchor year of the 11th Five-­
Year 


Mobilizing the government  51
Plan, independent variables all referred to this year unless otherwise specified. Envi­
ronmental quality was represented by both the annually average SO2 concentra­
tion in provincial capitals and nonpower sectors’ emission density (expressed in 
tons/km2). The capitals’ SO2 concentration data were published in China Statistical 
Yearbooks (National Bureau of Statistics, 2006). In addition, China divided SO2 
emissions into two big categories: power and nonpower. Associated with shorter 
chimneys, non-­
power-­
sector emissions were believed to be more closely associ­
ated with local air quality (SEPA, 2006a). Their emission densities in provinces 
were employed to represent another perspective of environmental quality (National 
Bureau of Statistics of China, 1999; Zou et al., 2006). Environmental capacity is a 
term indicating allowed maximum emissions to maintain a certain environmental 
quality. The data used in this section came from a study that calculated long-­
term 
SO2 goals for the 11th Five-­
Year Plan (Zou et al., 2006). Critical acid deposition 
was the targeted environmental quality. The corresponding upper-­
limit national 
emissions were 17.3 million tons, and each province had its own figure (Zou et al., 
2006). Current emission levels were represented by provincial SO2 emissions in 
2005. Provincial goals were formally distributed in August 2006 (State Council, 
2006). Because data for 2005 had been published in June 2006 (SEPA, 2001–2009), 
they should be available for negotiating the goal distribution. Provincial GDP per 
capita stood for economic development status (National Bureau of Statistics, 2006).
No definition had been clearly displayed by the Chinese authorities on SO2 
mitigation capability. Two variables were used. First, higher provincial SO2 
removal rates in 2005 could indicate fewer opportunities for the future. From 
another aspect, they also represented previous efforts in SO2 mitigation. Second, 
SO2 scrubbers (or flue-gas desulfurization facilities, FGD) had been designated as 
a key measure to reduce SO2 emissions in the 11th Five-­
Year Plan (State Council, 
2007a). The power sector’s shares of total emissions would then serve as another 
indicator of mitigation capability (National Bureau of Statistics of China, 1999; 
Zou et al., 2006). Higher shares may lead to a more effective reduction of total 
SO2 emissions through SO2 scrubbers.
China’s policies and emission control plans targeting individual emission 
sources could decide provincial goals in a bottom-­
up way. Nevertheless, it may 
not coincide with the top-­
down results. For example, effluent emission standards 
and SO2 scrubber planning, respectively, were expected to lead to national power 
sector’s emissions of 8.9 and 9.7 million tons in 2010, while the finally assigned 
goal was 9.5 million tons in the 11th Five-­
Year Plan (Zou et al., 2006). To evalu­
ate their impact on goal distribution, two independent variables were generated 
for each province: (1) (Emission standard-­
designated levels in the power sector 
in 2010 + Nonpower emission goals in 2010) / Provincial emissions in 2005 – 
100% (State Council, 2006; Zou et al., 2006; National Bureau of Statistics of 
China, 1999) and (2) (Scrubber planning-­
projected emissions in power sector + 
Nonpower emission goals in 2010) / Provincial emissions in 2005 – 100% (State 
Council, 2006; Zou et al., 2006; National Bureau of Statistics of China, 1999).
According to geographical locations and economic advancement, China 
divides its provinces into three regional groups: west, center and east. To alleviate 


52  Mobilizing the government
regional disparity in economic growth and income, China treats the three cat­
egories differently. For example, “Great West Development” aimed to develop 
western provinces, particularly through building infrastructure. Dummy variables 
were generated to indicate a province’s location. In addition, because China’s 
prevalent wind generally transports air pollutants from the west to the east, SO2 
emissions in western provinces could cause more damage than those in eastern 
provinces. The dummy variables then evaluated the overall impacts of these two 
opposite concerns.
Besides these variables, several others that were not mentioned in the official 
distribution plan were also tested, including SO2 emissions per capita, goal attain­
ment in the 10th Five-­
Year Plan and electricity export. One argument for China 
not to accept a legally binding goal on carbon mitigation in the Kyoto Protocol 
was its low carbon emissions per capita. Whether China applied this principle in 
domestic practice was examined through provincial SO2 emissions per capita in 
2005 (National Bureau of Statistics of China, 1999).
China failed substantially to achieve its 10% reduction goal of SO2 emis­
sions in the 10th Five-­
Year Plan (2001–2005): the actual emissions in 2005 
were 42% higher than the original goal (SEPA, 2001–2009, 2001). But some 
provinces did better than others. Whether better performance in the past was 
recognized is tested through a ratio: Provincial emissions in 2005 / Provincial 
emission targets in the 10th Five-­
Year Plan for 2005 (State Council, 2006; 
National Bureau of Statistics of China, 1999). In addition, for the 27 prov­
inces used in models (discussed later), this variable was highly correlated with 
the provincial growth rates of SO2 emissions in the 10th Five-­
Year Plan and 
the correlation coefficient is 0.98. Accordingly, the model results on this goal 
attainment variable could be almost identically applied to a variable on the 
growth rates.
Pollutant emissions and product consumption are not necessarily in the same 
location. Electricity is a clear and important case. SO2 comes out of coal-­
fired 
power plants, but electricity could be lighting bulbs in another province. This 
effect was examined through provincial electricity trade: Provincial electricity 
generation / Provincial electricity consumption – 100% (National Bureau of Sta­
tistics, 1997–2008).
Although mainland China has 31 provinces, only 27 were used for the statisti­
cal models. Four provinces were kept out. Hainan and Tibet had too-­
insignificant 
SO2 emissions in 2005, respectively, 22,000 and 2,000 tons. Qinghai had the least 
emissions among provinces except the two previously mentioned, and its data on 
avoided industrial emissions were not available in China Statistical Yearbooks. 
Shanghai had its nonpower SO2 emission density in 2005 much higher than 
other provinces (32.7 tons/km2; the next highest was 7.9 tons/km2), a far outlier 
(National Bureau of Statistics of China, 1999; Zou et al., 2006).
The correlation coefficients between the variables are given in Table 4.1. Pro­
vincial goals were highly correlated negatively with nonpower emission density, 
total SO2 emissions and GDP per capita – indicating that higher levels of these 
variables were closely associated with more stringent provincial goals  – and 


Mobilizing the government 53
Electricity
export
in 2005
1.00
Goal
attainment
Five−
in the
10th 
Year Plan
1.00
0.31
 
 
2
emission
 
SO
per
capita
in 2005
1.00
0.17
0.35
est
1.00
0.25
0.06
W
−0.17
Middle
1.00
−0.46
−0.03
0.42
0.52
 
Emission
 
d
 
standar
decided
goals
1.00
−0.08
−0.54
−0.44
−0.07
−0.44
Scrubber
planning
 
decided
goals
1.00
0.02
0.16
0.21
−0.03
0.32
0.10
 
s
’
Power
emission
e
0.30
shar
in 2005
1.00
−0.23
0.10
0.21
−0.29
0.18
0.28
Tibet, and Shanghai.
2
emoval
Rate
in 2005
1.00
−0.22
−0.03
0.10
0.05
SO
r
−0.08
−0.40
−0.14
−0.05
Correlation coefficients of key factors for 27 provinces
GDP
per capita
in 2005
1.00
−0.13
0.25
−0.33
0.75
−0.28
−0.46
−0.14
−0.22
−0.54
otal
in 2005
1.00
−0.16
−0.16
0.05
−0.39
−0.41
0.01
−0.17
0.00
0.21
T
emissions
0.24
 
Long−
term
1.00
0.15
0.14
goal
−0.30
−0.08
−0.35
0.31
−0.19
−0.18
−0.01
−0.05
−0.02
Nonpower
 
emission
density
in 2005
1.00
−0.04
0.22
0.59
−0.01
0.00
−0.36
0.29
−0.32
−0.26
0.03
−0.60
−0.47
s
Capital’
 conc.
2
1.00
0.17
0.05
0.16
SO
In 2005
−0.04
−0.12
−0.34
1
0.1
−0.17
−0.23
0.39
0.24
−0.30
−0.05
Reduction
1.00
0.00
goal
−0.74
−0.06
−0.53
−0.48
0.18
−0.10
0.63
−0.06
0.35
0.24
0.01
0.44
0.40
Reduction goal
 
 
able 4.1(a)
s SO2
concentration
s mainland has 31 provincial regions. Four are not included here: Qinghai, Hainan, 
 
 
 
emission
 
Capital’
Nonpower
density
Long−term goal
 
T
GDP/capita
 
Total emissions
Removal rate
 
s
’
emission
 
Power
share
Scrubber
planning
Emission
standard
Middle
est
Emission/capita
Electricity
export
W
Goal attainment
Note: China’


54  Mobilizing the government
0.0
0.12
7.9
3.7
200.2
4.5
0.6
0.7
0.1
0.0
1
1
61.0
1.3
0.6
Max
−20.4
0.02
0.2
−0.3
19.0
0.5
0.1
0.3
−0.3
−0.2
0
0
9.3
0.0
−0.6
Min
.
Dev
Std. 
5.7
0.020
2.09
1.02
48.39
0.91
0.16
1
1
0.1
0.1
0.09
0.48
0.47
13.26
0.34
0.24
0.057
−10.1
2.97
0.81
91.97
1.55
0.28
0.52
−0.09
−0.13
0.33
0.30
22.95
0.51
0.03
Mean
No. of 
observations
27
27
27
27
27
27
27
27
27
27
27
27
27
27
27
ear Plan
ear Plan
ear Plan
ear Plan
ear Plan
Period
11th Five-Y
2005
2005
1th Five-Y
1
2005
2005
2005
2005
1th Five-Y
1
11th Five-Y
2005
10th Five-Y
2005
Unit
%
ton/km2
Summary of variables
mg/m3
%
10,000 tons
10,000 RMB/person
%
%
%
%
dummy
dummy
kg/person
%
%
 conc.
 
s emission share
Table 4.1(b)
ariables
V
Reduction goal
2
s SO
Capital’
Nonpower emission density
long-term goal
’
Total emissions
GDP/capita
Removal rate
Power
Scrubber planning
Emission standard
Middle
est
W
Emission/capita
goal attainment
Electricity export


Mobilizing the government  55
positively with SO2 emissions from scrubber planning, goal attainment in the 10th 
Five-­
Year Plan and electricity export.
Although nonlinear terms could show consistent significance in models, such 
as the squared term of nonpower SO2 emissions density, its actual application in 
the negotiation for provincial goals was difficult. China very likely did not use a 
written formula to decide provincial goals. The nonlinear relationship was thus 
too complicated for arguments, especially those with a turning point. In addition, 
once included, several far points could greatly change the overall relationship 
in models. For example, if Shanghai appeared in the models, its big nonpower 
SO2 emission density would make the corresponding coefficient much different. 
These provinces might experience special negotiation. As a result, only linear 
terms were used in the models to examine China’s principles in goal distribution.
Another decision about the regression models was whether a constant vari­
able should be included. If all provinces had to presume a basic reduction goal 
and adjust it according to specific situations, the constant variable would show 
significance and the explained variance, R2, should be higher compared with a 
no-­
constant model. Model runs indicated otherwise (Table 4.2). In response, no 
constant variable appeared in the remaining models.
The model results showed that two variables were the most important in dis­
tributing the national goal to provinces. First, richer provinces tended to receive 
more stringent reduction goals. Provincial GDP per capita in 2005 and provincial 
goals in the 11th Five-­
Year Plan had a correlation coefficient of −0.48 (Table 4.1). 
But statistical models did not consistently show the significance of GDP per capita 
(Table 4.2). However, if either nonpower emissions density in 2005 or provincial 
goals from scrubber planning were excluded, GDP per capita would become sig­
nificant. For every 10,000 RMB/person increase, the province should reduce its 
SO2 emissions by further 1.3% (from the 2005 level). In explaining the model 
results, a problem was that two factors had a high correlation, and both showed 
significance on some occasions. However, it should not have mattered much in the 
negotiation. As long as no clear formula decided goals, a province could always 
argue with one factor to generate a more favorable goal. For example, Shanghai 
had a much higher nonpower SO2 emission density in 2005 than other provinces, 
but its GDP per capita in 2005 was ahead, with a significantly narrower margin 
(52,000 RMB/person compared with the next highest 45,000; National Bureau of 
Statistics of China, 1999; Zou et al., 2006). Comparatively, Shanghai could ask 
for a less stringent goal from GDP-­
per-­
capita point of view. Second, provinces 
with large emissions had tougher goals. China’s big provinces experienced greater 
pressure to reduce their emissions more for achieving the national goal. Coeffi­
cients of provincial emissions in 2005 were consistently significant (Table 4.2). 
Every 100,000 tons more SO2 emissions corresponded to about 0.47% further 
reduction. Third, nonpower emissions density displayed consistent significance. 
For emitting one more ton per square kilometer, a province should further reduce 
total SO2 emission by about 1.3%.
Notably, several other variables did not show much influence. First, provinces 
with worse environmental quality might not have received more stringent goals. 


56  Mobilizing the government
***
Model 9
***
−0.058
**
**
 
 
 
−2.79
 
 
10.09
12.71
 
 
 
3.62
6.33
 
0.94
Model 8
***
**
 
−0.033
**
−1.29
 
−0.97
 
 
15.39
3.33
 
 
 
 
 
 
0.94
***
Model 7
***
 
−1.34
*
 
−0.047
−1.27
 
 
 
 
 
 
 
 
 
 
0.93
***
Model 6
*
 
−0.83
−0.066
*
−1.73
 
 
 
 
 
 
 
 
1.98
4.48
 
0.93
Model 5
**
*
 
−0.042
*
 
−1.00
−1.47
 
 
12.98
12.55
 
 
0.05
0.73
4.43
 
0.94
Regression model results for distributing the national goal to provinces
***
Model 4
−0.80
−0.43
−0.064
*
−1.87
 
 
 
 
 
 
 
0.01
2.16
3.96
 
0.93
Model 3
***
*
**
47.28
−1.22
−0.99
−0.038
−2.25
2.40
1.43
15.35
9.83
0.36
−0.99
 
 
 
 
0.94
Model 2
*
−1.14*
40.03
−0.86
−0.041
−1.95
*
4.48
−2.42
14.10
15.38
−0.13
−0.90
0.07
0.58
3.69
 
0.95
Model 1
*
*
39.51
−1.14
−0.88
−0.041
−1.98
4.36
−2.63
14.08
15.47
−0.18
−0.95
0.07
0.57
3.74
0.31
0.76
density
2
 
 per capita
T
Independent variables
 concentration
2
s SO
s emission share
R
able 4.2
Capital’
Nonpower emission 
Long-term goal
Removal rate
Power
Emission standard
Middle
est
Total emissions
’
GDP
Scrubber planning
W
Emission/capita
Goal attainment
Electricity export
Constant
Adjusted 
* Significant at 10%. ** Significant at 5%. *** Significant at 1%.


Mobilizing the government  57
Provincial capital cities’ SO2 concentration did not significantly affect provincial 
goals (Table 4.2). But in most provinces, capital cities only occupy a fraction 
of the total land area and thus could not represent the general picture. Another 
problem with this variable was its coefficient’s sign. Intuitively, the sign should 
be negative – dirtier air needs more reduction of pollutant emissions. The actual 
coefficient, although not significant, was consistently positive (Models 1–3 in 
Table 4.2). To avoid its impact, the variable was excluded from other models. 
Second, provinces with higher emissions per capita did not face deeper reduc­
tions. Emissions per capita did not have any significant relationship with pro­
vincial reduction goals. Third, earlier efforts on SO2 emission control were not 
awarded later with relaxed goals. Neither of the two relevant variables – SO2 
removal rates in 2005 and goal attainment in the 10th Five-­
Year Plan – showed 
any consistent significance. Earlier efforts did not make the future easier in SO2 
emission control, while no failure in the past would get punished through adding 
future burden. Because of the very high correlation between the goal attainment 
variable and provincial growth rates of SO2 emissions in the 10th Five-­
Year Plan, 
the model results also indicated that faster emission growth did not have a sig­
nificant impact on provincial goals. For China’s political reality, this result was 
reasonable. Provincial and other local leaders often rotate every five years. If one 
administration was irresponsible, its failure did not get the next administration 
punished. Similarly, a performing administration should not reduce pressure on 
future leaders. Fourth, more electricity net export consistently led to less strin­
gent goals, but the relationship was not statistically significant. It seemed that 
China did not take serious consideration of the disintegration between emissions 
and consumption in distributing environmental goals. Fifth, no influence was 
found solely due to the location of a province. Regional characteristics should 
have been absorbed into other variables. For example, long-­
term goals already 
considered prevalent wind and more damage from western SO2 emissions. West­
ern and central provinces were poorer than eastern ones, which was reflected in 
GDP per capita.
Three principles were distinguished for distributing the national SO2 emission 
goal in the 11th Five-­
Year Plan: those provinces with heavier pollution, bigger 
total emissions and richer GDP per capita should reduce more. The second prin­
ciple was the most consistently applied. An explicit formula of deciding a provin­
cial goal could be written as
Provincial Goal (−0 to −100) = −1.34 × Nonpower emission density 
(tons/km2) – 0.047 × Total emissions (10,000 tons) − 1.27 × GDP 
per capita (10,000 RMB/person).
The 27 provinces had an arithmetic average goal in the 11th Five-­
Year 
Plan of −10.1%. The formula would lead to −10.2%: GDP per capita, −2.0%; 
nonpower SO2 emissions density, −4.0%; and total emissions, −4.3%. The 
explanatory power was high, with adjusted R2 generally over 0.93 (Model 7 in 
Table 4.2).


58  Mobilizing the government
3.2  
Goal distribution from provincial to municipality governments
The SEPA issued guidance for distributing SO2 emission goals from one govern­
ment level to its subordinate level (SEPA, 2006a). The total emissions are dis­
tinguished into the power sector (capacity no less than 6 MW) and nonpower 
sectors (SEPA, 2006a). The SO2 emission quota was generally assigned to each 
fossil-­
fuel power plant according to provincially homogeneous emission inten­
sity (grams SO2/kWh, varying with plant ages; SEPA, 2006a). As shown in Fig­
ure 4.1, the designated emission intensity was more stringent in new coal power 
plants and those in eastern or richer provinces. From provinces to municipalities, 
polluting sources in nonpower sectors received their upper limits on the basis 
of achieving local air quality – particularly SO2 emissions concentration with a 
threshold of 0.060 mg/m3 (SEPA, 2006a). The guidance did not clarify everything 
for assigning goals. It left decisions to provincial governments, especially in non­
power sectors. More important, the excess emission quota of a region was allowed 
to transfer or trade across regions (SEPA, 2006a).
0.0
1.0
2.0
3.0
4.0
5.0
6.0
7.0
8.0
East-1
East-2
Central
Southwest
Northwest
O
S
 
d
e
t
a
n
g
i
s
e
D
2
O
S
 
s
m
a
r
g
(
 
y
t
i
s
n
e
t
n
i
 
n
o
i
s
s
i
m
e
2/kWh)
Period I
Period II
Period III
Figure 4.1  
Designated SO2 emission intensity in distributing SO2 emissions quota to coal-
fired power plants for 2010 in the 11th Five-Year Plan
Source: SEPA (2006a).
Note: Coal-fired power plants falling in Period I refer to those that went online or passed the Environ­
mental Impact Assessment reports before December 31, 1996. Period II spans from January 1, 1997, 
to December 31, 2003. And Period III is from January 1, 2004, to the present. “East-1” includes the 
provinces of Liaoning, Hebei, Shandong, Zhejiang, Fujian and Hainan. “East-2” covers Beijing, Tian­
jin, Shanghai and Jiangsu Provinces. “Central” refers to Heilongjiang, Jilin, Shanxi, Henan, Hubei, 
Hunan, Anhui and Jiangxi Provinces. “Southwest” provinces are Chongqing, Sichuan, Guizhou, Yunnan, 
Guangxi and Tibet. “Northwest” has Inner Mongolia, Shaanxi, Gansu, Ningxia, Qinghai and Xinjiang.


Mobilizing the government  59
In distributing provincial goals, official documents often did not even qualita­
tively declare what factors took effect. Furthermore, municipality data were not 
as publicly available as provincial data. Two aspects receive special attention in 
examining the provincial scheme of goal distribution: (1) whether the same prin­
ciples in the national goal distribution held and (2) whether the SEPA’s guidance 
was followed. This section looks at four provinces: Hebei, Guangdong, Jiangsu 
and Shanxi. Statistical model results are given in Table 4.3. Those models without 
significance are not shown. According to how the four provinces obey the national 
principle – rich and big provinces reduce more – a matrix is generated in Table 4.4.
Provinces have high autonomy in further allocating their goals among munici­
palities. The four provinces are distinguished with different patterns to dem­
onstrate such decentralized authority. In Hebei Province, all municipalities got 
roughly same reduction goals. The provincial goal of Hebei Province was a 15% 
reduction from 1.50 million tons in 2005 (State Council, 2006). It had 11 munici­
palities, and their SO2 emissions in 2005 ranged from 45,000 to 311,000 tons 
(Hebei Provincial Government, 2007). GDP per capita also varied, from 9,900 to 
27,900 RMB/person (Hebei Provincial Statistics Bureau, 2006). The 2005 data on 
the power sector’s share in SO2 emissions are not publicly available. Information 
from the goals for 2010 is applied instead: the share of the power sector would 
range from 13.4% to 53.4% in the plan (Hebei Provincial Government, 2007). 
However, municipality goals only varied from a 14.1% to a 15.8% reduction, 
centering on the provincial goal (Hebei Provincial Government, 2007). Statistical 
models indicated a consistently significant constant (Table 4.3). But neither GDP 
per capita nor SO2 emissions showed any impact on municipality goals. If the 
goal distribution guidance from the central government worked, a municipality 
with more SO2 emissions from the power sector should receive somewhat more 
stringent goals. But no negotiation seemed to have shaped the municipality goals. 
The provincial government, very likely its top leaders, decided that the provincial 
goal was applied to all with minor adjustments.
In Guangdong Province, higher income and more SO2 emissions led to more 
stringent goals. In the 11th Five-­
Year Plan, Guangdong Province received a 
Table 4.3 Regression model results for distributing provincial goals to municipalities
Hebei
Guangdong
Jiangsu
Shanxi
Observations
11
15
13
11
Nonpower emission density in 2005
−1.5**
GDP/capita in 2005
−7.2***
6.2***
SO2 emissions in 2005
−3.3***
−1.4**
−0.98***
Power sector’s emission share in 2005
−58.1***
_Constant
−15.1***
33.9***
21.4**
Adjusted R2
0.90
0.76
0.86
0.87
Note: Six municipalities in Guangdong province with 2005 SO2 emissions no more than 11,000 tons 
are not included in the model.
* Significant at 10%. ** Significant at 5%. *** Significant at 1%.


60  Mobilizing the government
provincial goal of a 15% reduction from 1.29 million tons in 2005 (State Coun­
cil, 2006). Among its 21 municipalities, 5 emitted less than 10,000 tons in 2005, 
and another 13, no more than 60,000 tons (Guangdong Environmental Protection 
Bureau, 2006). The biggest three emitted 475,000 tons, taking 51% of the pro­
vincial emissions that belonged to municipalities (929,000 tons; the remaining 
365,000 tons were directly claimed to the provincial level; Guangdong Environ­
mental Protection Bureau, 2006). Collectively, these three should reduce their 
SO2 emissions by 46% (Guangdong Environmental Protection Bureau, 2006). 
The other 18 municipalities were even allowed to increase their emissions by 23% 
(Guangdong Environmental Protection Bureau, 2006). Regression models could 
better distinguish influential factors. To avoid the heavy impacts of outlying data 
points, six municipalities were excluded, with total SO2 emissions in 2005 being 
no more than 11,000 tons and their goals in 2010 allowed for over 170% growth. 
Models for the rest of the 15 municipalities show significance of GDP per capita 
and SO2 emissions (Table 4.3). Nonpower SO2 emissions density was not tested 
because of data unavailability. Different from the situation for provincial goals, 
the constant variable here is significant. For every 10,000 RMB/person increase 
of GDP per capita and 10,000 tons more of SO2 emissions in 2005, a municipal­
ity goal would be, respectively, 7.2% and 3.3% more stringent (from the 2005 
level). Although the coefficients were different from those for provincial goals, 
the qualitative principles remained the same: rich and big municipalities should 
reduce more.
In Jiangsu Province, municipalities with higher emissions should reduce more, 
but richer ones were allowed to reduce less. Jiangsu Province’s goal was an 18% 
reduction from 1.23 million tons in 2005 (State Council, 2006). The 13 munici­
palities emitted from 28,000 to 243,000 tons of SO2, a much narrower but still 
larger range than in Guangdong Province (Jiangsu Provincial Government, 2008). 
SO2 emission goals varied between 2.6% to 53.6% reduction (Jiangsu Provincial 
Government, 2008). Regression models included four independent variables, all 
for 2005 at municipality level: nonpower SO2 emissions density, GDP per capita, 
SO2 emissions and the power sector’s share in total emissions (Table 4.3). The 
constant variable showed significance, and its appearance in models made the 
adjusted R2 bigger. Corresponding to the increase of, respectively, 10,000 tons of 
SO2 emissions, 1% of the power sector’s share and 1 ton/km2 of nonpower SO2 
emissions density, a municipality goal would become 1.4%, 0.6% and 1.5% more 
stringent (from the 2005 level). The signs of these coefficients were all reasonable 
and consistent with the situation of distributing the national goal to provinces, 
but GDP per capita displayed the opposite effect: for a municipality with 10,000 
RMB/person richer, its SO2 emissions were allowed to grow by 6.2%. For a prov­
ince, this strategy of “rich municipalities reduce less” might maximize its GDP 
as well as tax income through entitling more opportunities to more promising 
municipalities.
In Shanxi Province, municipalities with higher emissions should reduce more, 
but income level did not have significant impacts. Shanxi Province’s goal was a 
14% reduction in the 11th Five-­
Year Plan from 1.52 million tons in 2005 (State 


Mobilizing the government  61
Council, 2006). Its 11 municipalities emitted from 92,000 to 185,000 tons of SO2 
in 2005, a much narrower range compared with the earlier three provinces or the 
national situation (Shanxi Provincial Government, 2006). Their GDP per capita 
in 2005 was between 5,500 and 26,000 RMB/person (Shanxi Bureau of Statistics, 
2006). The municipality goals were scattered from a 8.5% to a 17.8% reduction 
(Shanxi Provincial Government, 2006). Only total SO2 emissions showed a sig­
nificant influence in the statistical models (Table 4.3). For emitting every 10,000 
tons more of SO2, a municipality goal would be about 1% more stringent (from 
the 2005 level). The GDP per capita’s coefficient was negative, although not sta­
tistically significant. The principle – big provinces should reduce more – held 
here. That rich provinces should reduce more was not well applied but could have 
been considered.
In conclusion, provinces differed from each other in adopting principles from 
distributing the national goal (Table 4.4), which closely reflected that China’s 
governance and, especially, environmental governance had been greatly decen­
tralized (see Chapter 3). The most consistent principle across provinces was that 
bigger emitters should reduce more. The guidance from the SEPA did not have to 
be exactly followed.
4  
Decentralized goal attainment
To keep the goal process running, goal attainment assessment is an inalienable 
step. It examines the effectiveness of the goal process and provides feedback. The 
key questions are what can be called goal attainment and how to evaluate it. For 
one Five-­
Year Plan, goal attainment evaluation does not wait until its conclusion. 
In the 11th Five-­
Year Plan, China publicized provincial SO2 emissions every half 
a year (State Council, 2007c). At the end of 2008, a halfway assessment was 
scheduled (State Council, 2007c).
4.1  
Criteria for goal attainment
In the 11th Five-­
Year Plan, China established “three systems” to facilitate SO2 
mitigation, which covered statistics, monitoring and evaluation (State Coun­
cil, 2007b). This capacity building was planned and carried out by the SEPA 
and endorsed by the State Council (State Council, 2007b). Because of China’s 
Table 4.4 Provincial goal distribution matrix
Rich guys reduce
Less
Neutral
More
Big guys reduce
Less
Neutral
Hebei
More
Jiangsu
Shanxi
Guangdong


62  Mobilizing the government
decentralization, that the central government mainly governs through provin­
cial governments, the national system only targeted the provincial level (State 
Council, 2007b). The subordinate governments within provinces were evaluated 
with the rules passed along by provincial governments. For example, Zhejiang 
Province later enacted a more detailed, although not systematically different, 
regulation targeting municipality and county governments (Zhejiang Provincial 
Government, 2008).
Provincial goal attainment was evaluated with three criteria in the 11th Five-­
Year Plan (State Council, 2007b). The first criterion was on the quantitative goal 
itself and environmental quality. It was often a binary judgment: if they were 
attained, that was a mission accomplished. Excessive reduction would not be fur­
ther awarded after the goal had been attained, while more emissions would not be 
punished if the goal was already broken. The second criterion was on the establish­
ment and operation of three institutions: environmental goal setting of major pol­
lutants, monitoring and goal attainment evaluation (State Council, 2007b). They 
were mainly judged by the enactment and distribution of official documents. The 
third one was on mitigation measures, including the completion and operation of 
pollutant removal facilities, the closure of inefficient factories, policy enactment 
and plan implementation (State Council, 2007b). If any of the three criteria failed 
to pass evaluation, the overall goal attainment would be judged a failure (State 
Council, 2007b). Accordingly, one feature was that the attainment of the goal 

itself, despite its central importance, did not ensure overall success. The first crite­
rion focused on the results and the other two on the process. The regulation of the 
Chinese central government on the process provided feedback for local govern­
ments for adjusting policies and monitoring their implementation.
The 9th Five-­
Year Plan barely had any defined scheme to evaluate SO2 goal 
attainment (NEPA et  al., 1996). The official document available in the public 
domain only pointed out that the SO2 control would be annually examined and 
evaluated and the result would be publicized periodically (NEPA et al., 1996). 
The SO2 goal and its attainment process were better defined in the 10th Five-­
Year Plan, but no evaluation scheme was clearly defined either (SEPA, 2001). 
The National 10th Five-­
Year Plan for Environmental Protection only expressed 
several principles, including holding local government leaders responsible and 
linking environmental goal attainment with the leaders’ performance evaluation 
(SEPA, 2001). The evolution path displayed China’s progress in establishing a 
working evaluation scheme of SO2 emission goals. Although still not perfect, the 
much clearer scheme in the 11th Five-­
Year Plan could have significantly contrib­
uted to the SO2 mitigation goal attainment.
Recognizing the importance of credible data collection for achieving SO2 emis­
sion goals in the 11th Five-­
Year Plan, China experienced an intensive capacity-­
building process. In December  2006, the updated Management Methods of 
Environmental Statistics entered into force (SEPA, 2006c). The regulation speci­
fied the organization and personnel for environmental statistics, rules on environ­
mental survey and management and publication of environmental data (SEPA, 
2006c). For goal attainment in the 11th Five-­
Year Plan, China strengthened its 


Mobilizing the government  63
statistical system focusing on data credibility. SO2 emissions were divided into 
three categories: power, nonpower industries and domestic (State Council, 
2007b). The first two categories (industrial sectors) were further distinguished 
into two – key and non-­
key surveyed sources – based on the sizes of emission 
sources, and key surveyed sources covered 65% of total industrial SO2 emissions 
(State Council, 2007b). Three parallel methods were applied under various situ­
ations: direct monitoring, estimation according to sulfur budget and estimation 
according to emission factors (State Council, 2007b). The first method, if appli­
cable, had the highest priority (State Council, 2007b). SO2 emissions from non-­
key surveyed sources were estimated following a similar trend as key surveyed 
sources (State Council, 2007b). Data about coal consumption and sulfur contents 
worked out SO2 emissions from domestic sectors (State Council, 2007b). In addi­
tion, if cheating were caught in an SO2 removal facility more than twice a year, 
no SO2 removal would be recognized in the statistics data from the facility (State 
Council, 2007b). Furthermore, another two significantly more detailed policies 
were enacted for building emission inventories (SEPA, 2007a, 2007d). Not only 
were detailed accounting methods clearly written, but also the data report was 
regulated in specifics (SEPA, 2007a, 2007d).
4.2  
Incentives for goal attainment
The central government has decentralized its power greatly since the economic 
reform started in 1978. As discussed in Chapter 3, three measures could exist 
to incentivize the cooperation of local governments by targeting local leaders, 
administrative constraints and fiscal transfer. Corresponding to the personnel 
relationship that is mainly established across the various levels of the Chinese 
Communist Party, the top national leadership of the party can greatly decide the 
promotion and removal of provincial-­
level leaders. The attainment of key goals, 
including those on environmental protection and SO2 mitigation, had become an 
important aspect in the evaluation of provincial leaders’ job performance. Con­
cerning SO2 mitigation goal implementation since the 11th Five-­
Year Plan, local 
government leaders but not local environmental protection bureau (EPB) leaders 
were targeted. The clear evidence was that provincial deputy governors, not EPB 
directors, were required to sign pollutant emission control contracts with the cen­
tral government (SEPA, 2006b). The failure in the 10th Five-­
Year Plan on surging 
coal consumption demonstrated that pollution control had been far beyond the 
responsibility of EPBs alone.
Officially five characteristics distinguish a leader in the Chinese Communist 
Party for promotion or removal: virtue, ability, diligence, achievements and 
absence of corruption (The Central Committee of the Chinese Communist Party, 
2002). Furthermore, after the formation of “Scientific View of Development,” 
resource consumption, environmental protection and sustainable development 
were clearly pointed out to comprise “achievements” (Department of Organi­
zation of the Chinese Communist Party, 2006). Contracts on pollutant emission 
control and energy conservation clarified even more the responsibilities of local 


64  Mobilizing the government
government leaders (SEPA, 2006b). Two institutions were applied for the attain­
ment of the SO2 emission goal in the 11th Five-­
Year Plan: accountability and 
veto (State Council, 2007a). “Accountability” demanded local government lead­
ers be held accountable for their governance that fell within their jurisdictions. For 
example, the administrator of the SEPA, Xie Zhenghua, was forced to resign in 
2005 for a serious pollution event in the Songhua River. “Veto” meant that local 
government leaders would fail evaluation on their entire job performance if the 
SO2 emission goal were not attained. Promotion became inappropriate for these 
leaders. If goal failure did not degrade the leaders’ ranks, they may still face a risk 
of being removed from original positions to some less significant ones. On the 
other hand, successful goal attainment was an important achievement and could 
help the leaders’ promotion. A recently developed method for targeting local lead­
ers has been gradually promoted by the Ministry of Environmental Protection 
(MEP) and later by the Ministry of Ecology and Environment. Top leaders of 
those provinces and municipalities that show serious environmental problems or 
fail environmental goals are forced to have “interview appointments” with the 
ministry (Ministry of Ecology and Environment, 2020a). Although those local 
leaders may not face immediate consequences of punishment, they will receive 
crucial warnings that darken their future promotion opportunities, especially if no 
quick fix is achieved afterward.
Another mechanism that was applied in the 11th Five-­
Year Plan was to tem­
porarily constrain local administrative authorities as punishment: if a goal was 
not attained, no new construction projects would receive the ratification of their 
environmental impact assessment (EIA) reports for a given period. Over the 11th 
Five-­
Year Plan period, large construction projects still demanded ratification from 
the central government. In terms of environmental protection, every project with 
potential environmental damage should compose an EIA report and submit for 
ratification to various levels of governments (National People’s Congress, 2002). 
The SEPA, and later the MEP, at the central level was responsible for large pro­
jects, such as new coal-­
fired power plants over 200 MW (SEPA, 2002). No project 

without the MEP’s ratification could legally start construction. In early 2007, the 
SEPA temporarily suspended ratifying EIA reports of four municipalities and 
four power corporations (SEPA, 2007c). The suspension took effect for three 
months to force their cooperation (SEPA, 2007b). Afterward, the policy was for­
mally established to target goal failure (SEPA, 2008). A failure to achieve the SO2 
emission goal could result in regional suspension for one month, three months or 
half a year. If no satisfying progress were made, the suspension could even last 
longer until full cooperation. The “suspension” policy may seriously influence 
the regional economy. Since GDP is the most important criterion in evaluating 
local leaders, this mechanism could effectively force cooperation. Capital invest­
ment was a crucial part of China’s GDP. For example, in 2007, China’s over­
all GDP was 24.7 trillion RMB, and capital investment comprised 13.7 trillion 
RMB, about 56% (National Statistics Bureau, 2008). A one-­
month suspension 
could delay construction and significantly affect capital investment and, conse­
quently, the local economy. GDP growth itself occupied the most important status 


Mobilizing the government  65
in evaluating local government leaders. In addition, a booming GDP could pro­
vide growing tax income not only to make officials more powerful but also to 
enable more budgets for poverty alleviation, health care, education and other key 
governmental affairs. Many of these issues are closely connected with the evalu­
ation of leaders.
Fiscal transfer has not been explicitly linked with environmental goal attain­
ment. However, the very significant fiscal transfer from the central to local gov­
ernments (as discussed in Chapter 3), if institutionally associated with pollutant 
emission control, is potentially powerful to mobilize local governments for envi­
ronmental protection.
With China’s further decentralization of governmental authorities, the first 
mechanism to directly target local governments is expected to be even more 
important. In the past four decades, the central government has been continuously 
loosening direct management of local governmental affairs. As a key feature of 
the economic reform, China has greatly reduced the requirements of adminis­
trative ratification and decentralized much remaining authority to local govern­
ments (State Council, 2013b, 2014). Fossil-­
fuel-­
fired power plants were no longer 
required for the MEP’s ratification after 2015 and the authority entirely went to 
provincial governments (MEP, 2015; Ministry of Ecology and Environment, 
2019).
5  
Goal evolution
Corresponding to different strategies for controlling air pollution–induced health 
damages, three major types of goals can be adopted. First, emission mitigation 
goals of key pollutants, prominently SO2, aim to directly target the sources of 
environmental pollution. The second type focuses on controlling air pollutant 
concentrations. Ambient air quality standards are widely adopted across coun­
tries to specify concentration thresholds of key air pollutants individually, such 
as SO2, fine particulate matter (PM2.5) and ozone (O3). As discussed earlier, the 
control of ambient SO2 concentration was a key scientific foundation to decide 
China’s long-­
term SO2 mitigation goal at 12 million tons (Yang et al., 1999). The 
third type targets environmental quality directly through the Air Quality Index 
(AQI) that provides a synthesized measurement of key air pollutant concentra­
tions. The AQI also guides people’s activities corresponding to air quality condi­
tions. Although the three strategies have a similar ultimate goal for protecting 
public health, they have different implications for implementation. Local govern­
ments can only directly mitigate local emissions while local pollutant concentra­
tion is determined by emissions within and outside of their jurisdiction as well as 
weather conditions, land use and other factors. Then their motivation could differ 
significantly under the different types of goals to affect their performance of pol­
lution mitigation.
Over the past two decades, China has been switching back and forth between 
major governance strategies on environmental protection with different types of 
goals. 
As clearly stated in China’s environmental protection law, local governments 


66  Mobilizing the government
are responsible for environmental quality within their jurisdictions (National Peo­
ple’s Congress, 1989). However, environmental protection was not ranked high 
among all governmental tasks in the 1990s. Local leaders generally prioritized 
economic growth for promotion opportunities. The 10th Five-­
Year Plan (2001–
2005) was a transitional period toward the Total Emission Control regime to set 
up environmental goals for reducing major pollutant emissions by 10% (National 
People’s Congress, 2001). However, due to the lack of environmental cleanup 
incentives and the acceleration of economic growth, SO2 emissions went up by 
27.8%, and only 2 out of 31 provinces achieved their allocated goals. Demand 
for serious, effective and efficient compliance monitoring had not been strong. 
The 11th Five-­
Year Plan (2006–2010) was a milestone in China’s environmental 
protection history. The Total Emission Control regime was strengthened, while 
serious and implementable incentives were put into place for local governments 
to achieve their individual mitigation goals (Xu, 2011). A bottom-­
up compliance 
monitoring system on emissions was initiated and established (SEPA, 2007d). 
Although SO2 emissions did decline in the 11th Five-­
Year Plan, data manipulation 
also strained the compliance monitoring system as indicated in the gaps between 
official and independent emission inventories (Lu et al., 2011).
Concerning SO2 emissions, two sets of regulations were most important and 
direct, being effluent emission standards and ambient air quality standards. Pre­
viously, cities were given goals of “blue sky” days. “Blue sky” was defined as 
that air quality reached the Grade 2 standard. One crucial change in the 2012 
version ambient air quality standards was the addition of PM2.5 (MEP, 2012; 
National Environmental Protection Administration and State Bureau of Techni­
cal Supervision, 1996). PM2.5 concentration is more closely related to air quality 
that affects public health, while the emissions of SO2 and other pollutants are 
only indirect measures. In other words, PM2.5 goals are more related to ends of 
air pollution control, while SO2 emissions goals are more about means. PM2.5 
comprises many more pollutants, including sulfate particles that are originated 
from SO2 emissions.
Together with the 2012 update of the ambient air quality standards, China 
enacted the Ambient Air Quality Index (Ministry of Environmental Protection, 
2012). It synthesizes key air pollutant concentrations into one index to indicate 
air quality. The cutoff AQIs between “excellent,” “good” and “polluted” air are 50 
and 100, respectively. Each air pollutant can calculate its individual AQI (IAQI) 
and the composite AQI is the largest IAQI, or the IAQI of the primary air pol­
lutant. An AQI of 50 or lower corresponds to the Grade 1 ambient air quality 
standards, while 100 or lower corresponds to Grade 2. They provide the technical 
foundation for China to adopt regulatory strategies that are based on air qual­
ity rather than pollutant emissions. Both regulations gave nearly four years of 
grace periods and formally entered into force in January 2016. The 12th Five-­
Year 
Plan (2011–2015) initially continued with the Total Emission Control scheme to 
include more pollutants (National People’s Congress, 2011). However, a major air 
pollution episode in January 2013 that badly hit North China, most notably Bei­
jing, pushed the Chinese government to rethink its strategy (State Council, 2013a) 


Mobilizing the government  67
and accelerated the shift toward the air quality approach and the application of the 
two related standards.
Air quality goals and emission mitigation goals of SO2, as well as other major 
air pollutants, both aim for public health benefits. In order to achieve air quality 
goals that focus on ambient air pollution, efforts should still primarily fall on 
the mitigation of pollutant emissions together with their geographic and temporal 
distributions. Due to the atmospheric transport of air pollution, the attainment of 
PM2.5 goals depends not only on a region’s own mitigation efforts but also that of 
neighboring regions. The interregional reliance tends to be greater for geographi­
cally smaller jurisdictions. Accordingly, free riding may be a potential problem 
to compromise the willingness to engage in hard mitigation efforts. Nevertheless, 
data credibility is a key element in enforcing environmental policies as well as 
the top-­
down goals. Emission mitigation data, however, tend to be much more 
conveniently manipulated than air quality data. The number of polluting sources 
in China could easily overwhelm its compliance monitoring resources, especially 
in sparsely populated and less developed regions. In the 11th Five-­
Year Plan, the 
MEP assembled teams to inspect provinces and their polluting firms. However, 
the data had been of unsatisfying quality, and what was reported by local govern­
ments and polluting firms was often seriously discounted. Data on SO2 emissions 
are more prone to manipulation because the bottom-­
up monitoring and reporting 
have to go through many stakeholders who have incentives to underreport emis­
sions and overreport mitigation. Occasional verification from the central govern­
ment often finds big gaps in data and must “squeeze moisture” from the reported 
mitigation amounts. In contrast, ambient air quality data are much more difficult 
to manipulate and any dishonest behavior is much easier to discover. The central 
government also runs its own air quality monitoring network via ground stations 
and remote sensing, such as satellites. Accordingly, China reversed the strategy 
to have air quality improvement targets (State Council, 2013a). Air quality moni­
toring stations are much fewer than polluting sources to substantially reduce the 
resource burden of compliance monitoring. Thus, the probability of compliance, 
together with the better data quality, should be much higher.
The prospective penalty and reward for goal attainment do not differ substan­
tially from the 11th Five-­
Year Plan to the 12th and 13th. However, the 12th and 
13th Five-­
Year Plans achieved much faster SO2 mitigation, even considering the 
slower economic growth rates. It could indicate that the free-­
riding problems 
were less important than data credibility. Furthermore, SO2 emissions are just 
one among many pollutants, while PM2.5 could better serve as a comprehen­
sive air quality indicator. Provincial and local governments could have greater 
flexibility in weighing various technological and policy mitigation alternatives. 
It could also potentially encourage more local policy innovations and probably 
achieve better cost-­
effectiveness through balancing the marginal abatement costs 
of pollutants.
Furthermore, although emission reduction goals have been consistently 
achieved since the 11th Five-­
Year Plan, air quality was not perceived to have 
improved. One possible cause could be the problems in reporting emission data, 


68  Mobilizing the government
0
50
100
150
200
250
300
1/1/2014
1/1/2015
1/1/2016
1/1/2017
1/1/2018
1/1/2019
1/1/2020
(
 
n
o
i
t
a
r
t
n
e
c
n
o
c
 
e
g
a
r
e
v
a
 
y
l
i
a
D
μg/m3)
Date
Figure 4.2  
Daily SO2 concentrations in Shijiazhuang (1 January 2014–29 February 2020)
Source: Ministry of Ecology and Environment (2020b).
Note: The upper and lower dotted horizontal lines indicate the Grade 2 and 1 standard, respectively, in 
China’s ambient air quality standards in 1996 and 2012.
as discussed earlier. Another more important reason for the wide gap between 
the successful attainment of SO2 mitigation goals and the perceived terrible air 
quality was that SO2 has been increasingly less important in ambient air quality. 
For example, Hebei Province often has one of the highest anthropogenic PM2.5 
concentrations in China and the world. In its capital city, Shijiazhuang, air qual­
ity is taken as one example to illustrate the importance of new PM2.5 standards 
and goals. Significant improvements have been made on reducing SO2 emissions 
and concentrations. In the first three months of 2014, SO2 concentrations in Shi­
jiazhuang exceeded the Grade 1 standard (50 μg/m3) in 90% of all days, while a 
strong seasonal cycle indicated that the winter or the heating season as the worst 
season (Figure 4.2). From February 2019 to February 2020, in contrast, the stand­
ard was not exceeded for even a single day (Figure 4.2). It illustrates China’s hard 
and effective efforts in controlling SO2 emissions and bringing down SO2 concen­
trations. Essentially the original long-­
term goal for SO2 mitigation, 0.060 mg/m3 
or 60 μg/m3 (SEPA, 2006a), had been generally achieved. However, from the 
perspective of PM2.5, Shijiazhuang’s performance has been much less impressive. 
Its concentration has regularly exceeded the much more relaxed Grade 2 standard 
(Figure 4.3).


Mobilizing the government  69
0
50
100
150
200
250
300
350
400
450
500
550
600
650
1/1/2014
1/1/2015
1/1/2016
1/1/2017
1/1/2018
1/1/2019
1/1/2020
(
 
n
o
i
t
a
r
t
n
e
c
n
o
c
 
e
g
a
r
e
v
a
 
y
l
i
a
D
μg/m3)
Date
Figure 4.3  
Daily PM2.5 concentrations in Shijiazhuang (1 January 2014–29 February 2020)
Source: Ministry of Ecology and Environment (2020b).
Note: The upper and lower dotted horizontal lines indicate Grade 2 and 1 standard, respectively, in 
China’s 2012 ambient air quality standards.
PM2.5 is not a single pollutant but a set of various pollutants that fall into the 
size range. SO2 is a gaseous pollutant and could be converted into sulfate particles 
in the atmosphere to become one important component of PM2.5. Heating seasons 
in northern China tend to result in more coal consumption and pollutant emis­
sions, while inversion (when warm air is above cold air) is more frequent in the 
winter when the ground is cold, suppressing convection, and thus facilitates the 
accumulation of pollutant concentrations. Although SO2 is one key precursor spe­
cies of PM2.5, other air pollutants are also crucial components in forming PM2.5.
Furthermore, ozone pollution has significantly deteriorated over the period. 
O3 and PM2.5 concentrations tend to have opposite seasonal cycles. Chemical 
reactions to form O3 in the atmosphere involve nitrogen oxides (NOx), volatile 
organic compounds (VOC) and sunlight, while summer months tend to provide 
more favorable conditions. PM2.5 and SO2 concentrations peak in winter months, 
and O3–8h concentration (daily maximum concentration over 8 hours) is the high­
est in summer months (Figure 4.4). As a result, mitigation goals of SO2 emissions 
and SO2 concentrations will be at a greater distance from perceived air quality that 
mainly corresponds to PM2.5 and O3 concentrations.
SO2 has never been the primary pollutant to decide Shijiazhuang’s monthly 
AQI since 2014 (Figure 4.5). PM2.5 dominated the AQI before 2016, while in and 


70  Mobilizing the government
0
50
100
150
200
250
300
350
1/1/2014
1/1/2015
1/1/2016
1/1/2017
1/1/2018
1/1/2019
1/1/2020
Daily 8-
(
 
n
o
i
t
a
r
t
n
e
c
n
o
c
 
e
g
a
r
e
v
a
 
r
h
μg/m3)
Date
Figure 4.4  
Daily 8-hour O3 concentrations (daily maximum concentration over 8 hours) in 
Shijiazhuang (1 January 2014–29 February 2020)
Source: Ministry of Ecology and Environment (2020b).
Note: The upper and lower dotted horizontal lines indicate Grade 2 and 1 standard, respectively, in 
China’s 2012 ambient air quality standards.
0
50
100
150
200
250
300
350
Jan-14
Jan-15
Jan-16
Jan-17
Jan-18
Jan-19
Jan-20
)
I
Q
A
I
(
 
x
e
d
n
I
 
y
t
i
l
a
u
Q
 
r
i
A
 
l
a
u
d
i
v
i
d
n
I
Month -Year
PM2.5
SO2
O3-8h
Figure 4.5  
Monthly average AQI in Shijiazhuang (January 2014–February 2020; calcu­
lated from daily data)
Source: Ministry of Ecology and Environment (2020b).


Mobilizing the government  71
after 2017 with reduced PM2.5 concentration and rising O3–8h concentration, O3 
became the primary pollution in summer months and PM2.5 remained dominant in 
the winter (Figure 4.5). The trend is similar in other Chinese cities. For example, 
Beijing witnessed the rise of O3 in determining summer AQI a few years earlier 
than Shijiazhuang did (Figure 4.6). In southern China, where winter is mild/warm 
with adequate sunshine, the importance of O3 entirely overshadows that of PM2.5 
in the AQI. For example, in Shenzhen, AQI in most months is now decided by 
O3–8h but not PM2.5 (Figure 4.7).
From 2014 to 2020, PM2.5 concentrations and corresponding air quality indexes 
have been reduced throughout major cities in China, but O3–8h generally had a 
rising trend. One reason for their diverging trends in the past years could be traced 
to the presence of PM2.5 goals but not O3 goals. In the 13th Five-­
Year Plan, China 
further enacted air quality goals together with 15% reduction goals on SO2 and 
NOx emissions (National People’s Congress, 2016). The proportion of days that 
the AQI is below 100 in municipalities should reach 80%, while for those cities 
with PM2.5 concentrations not reaching the Grade 2 standard (or 75 μg/m3), they 
should reduce the level by 18% over the five years (National People’s Congress, 
2016).
The AQI is a more comprehensive measure of air pollution to consider both 
PM2.5 and O3. In China’s further goal evolution especially into the 14th Five-­
Year 
0
50
100
150
200
250
Jan-14
Jan-15
Jan-16
Jan-17
Jan-18
Jan-19
Jan-20
)
I
Q
A
I
(
 
x
e
d
n
I
 
y
t
i
l
a
u
Q
 
r
i
A
 
l
a
u
d
i
v
i
d
n
I
Month -Year
PM2.5
SO2
O3-8h
Figure 4.6  
Monthly average AQI in Beijing (January  2014–February  2020; calculated 
from daily data)
Source: Ministry of Ecology and Environment (2020b).


72  Mobilizing the government
Plan (2021–2025), it could play a more prominent role in mobilizing local govern­
ments for air pollution control.
Note
	
1	 This chapter is based on the author’s own material used in Xu, Y. 2011. The use of a goal 
for SO2 mitigation planning and management in China’s 11th five-­
year plan. Journal 
of Environmental Planning and Management, 54, 769–783; much of which has been 
revised and expanded on.
References
The Central Committee of the Chinese Communist Party. 2002. Regulations on selecting 
and appointing leaders of the party and governments. Beijing, China: The Central Com­
mittee of the Chinese Communist Party.
Chakravarty, S., Chikkatur, A., De Coninck, H., Pacala, S., Socolow, R. & Tavoni, M. 
2009. Sharing global CO2 emission reductions among one billion high emitters. Pro­
ceedings of the National Academy of Sciences of the United States of America, 106, 
11884–11888.
Chinese Academy for Environmental Planning (CAEP). 2004. Basic thoughts on national 
11th five-­
year plan on environmental protection. Beijing, China: CAEP.
0
10
20
30
40
50
60
70
80
90
Jan-14
Jan-15
Jan-16
Jan-17
Jan-18
Jan-19
Jan-20
)
I
Q
A
I
(
 
x
e
d
n
I
 
y
t
i
l
a
u
Q
 
r
i
A
 
l
a
u
d
i
v
i
d
n
I
Month -Year
PM2.5
SO2
O3-8h
Figure 4.7  
Monthly average AQI in Shenzhen (January 2014–February 2020; calculated 
from daily data)
Source: Ministry of Ecology and Environment (2020b).


Mobilizing the government  73
Department of Organization of the Chinese Communist Party. 2006. Temporary methods 
to evaluate local leaders for realizing scientific view of development. Beijing, China: 
Chinese Communist Party.
DeYoung, K. 2009. Obama sets timetable for Iraq. Washington Post, February 28.
Guangdong Environmental Protection Bureau. 2006. The 11th five-­
year plan on major pol­
lutants emission goals of municipalities in Guangdong province. Guangzhou, China: 
Guangdong Environmental Protection Bureau.
Hebei Provincial Government. 2007. The 11th five-­
year plan on environmental protection 
of Hebei province. Shijiazhuang, China: Hebei Provincial Government.
Hebei Provincial Statistics Bureau. 2006. Hebei economy statistics yearbook. Beijing, 
China: China Statistics Press.
Jiangsu Provincial Government. 2008. The 11th five-­
year plan on environmental protection 
and ecological construction of Jiangsu province. Nanjing, China: Jiangsu Provincial 
Government.
Li, J. 2010. How to allocate CO2 mitigation responsibilities within China – application of a 
new burden-­
sharing scheme. Ph.D. dissertation, Princeton University Press, Princeton, NJ.
Liu, G., Zhang, Z., Dong, Z. & Wu, L. 2006. Research report on China’s ten five-­
year 
plans. Beijing, China: People’s Press.
Locke, E. A., Saari, L. M., Shaw, K. N. & Latham, G. P. 1981. Goal setting and task-­
performance – 1969–1980. Psychological Bulletin, 90, 125–152.
Lu, Z., Zhang, Q. & Streets, D. G. 2011. Sulfur dioxide and primary carbonaceous aero­
sol emissions in China and India, 1996–2010. Atmospheric Chemistry and Physics, 11, 
9839–9864.
Ma, K. 2005. Strengthen the scientific and democratic planning process and write a good 
11th five-­
year plan. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
MEP. 2012. Ambient air quality standards. GB 3095-­
2012. Beijing, China: MEP.
MEP. 2015. Catalogue of construction projects for environmental impact assessment ratifica­
tion from the ministry of environmental protection (2015 version). Beijing, China: MEP.
Ministry of Ecology and Environment. 2019. Catalogue of construction projects for envi­
ronmental impact assessment ratification from the ministry of environmental protection 
(2019 version). Beijing, China: Ministry of Ecology and Environment.
Ministry of Ecology and Environment. 2020a. Method for interview appointment by the 
ministry of ecology and environment (Draft for comments). Beijing, China: Ministry of 
Ecology and Environment.
Ministry of Ecology and Environment. 2020b. Realtime city air quality data [Online]. 
Available: www.mee.gov.cn/.
Ministry of Environmental Protection. 2008. Assessment reports on provincial emissions of 
major pollutants. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2012. Technical regulation on ambient air quality 
index (on trial). Beijing, China: Ministry of Environmental Protection.
National Bureau of Statistics. 1997–2008. China energy statistical yearbook. Beijing, 
China: China Statistics Press.
National Bureau of Statistics. 2006. China statistical yearbook. Beijing, China: China Sta­
tistics Press.
National Bureau of Statistics. 2019. China statistical yearbook. Beijing, China: China Statistics 
Press.
National Bureau of Statistics of China. 1999. China statistical yearbook. Beijing, China: 
China Statistics Press.


74  Mobilizing the government
National Environmental Protection Administration & State Bureau of Technical Supervi­
sion. 1996. Ambient air quality standard. GB 3095-­
1996. Beijing, China: NEPA.
National People’s Congress. 1989. Law of environmental protection. Beijing, China: The 
4th Conference of the 10th National People’s Congress.
National People’s Congress. 2001. The outline of national 10th five-­
year plan on economic 
and social developments. Beijing, China: The 4th Conference of the 9th National Peo­
ple’s Congress.
National People’s Congress. 2002. Law of environmental impact assessment. Beijing, 
China: The 4th Conference of the 10th National People’s Congress.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. The outline of the 13th five-­
year plan on economic and 
social development. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
National Statistics Bureau. 2008. China’s economic growth in 2007. Beijing, China: China 
Statistics Press.
National Statistics Bureau & Ministry of Ecology and Environment. 2019. China statisti­
cal yearbook on environment 2018. Beijing, China: China Statistics Press.
NDRC. 2003. Establish a new view of development and do well in preparing writing the 
11th five-­
year plan. Beijing, China: NDRC.
NDRC. 2005. On the basic thoughts of the national 11th five-­
year plan. Beijing, China: NDRC.
NEPA, National Planning Commission & NETC. 1996. The national 9th five-­
year plan on 
controlling major pollutants’ emissions. Beijing, China: NEPA, NPC, NETC.
NEPA  & SBTS. 1996. Ambient air quality standard. GB 3095-­
1996. Beijing, China: 
NEPA, SBTS.
Pacala, S. W. & Socolow, R. H. 2004. Stabilization wedges: Solving the climate problem 
for the next 50 years with current technologies. Science, 305, 968–972.
Pouyat, R. V. & McGlinch, M. A. 1998. A legislative solution to acid deposition. Environ­
mental Science & Policy, 1, 249–259.
Roberts, L. 1991. Acid-­
rain program – mixed review. Science, 252, 371.
SEPA. 2001. National 10th five-­
year plan on environmental protection. Beijing, China: 
State Environmental Protection Administration.
SEPA. 2002. Regulation on ratifying environmental impact assessment reports of construc­
tion projects at different government levels. Decree No. 15. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2005. Plan on controlling acid rain and SO2 emissions (draft for comments). Bei­
jing, China: State Environmental Protection Administration.
SEPA. 2006a. Guidelines on calculating SO2 emission quotas. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2006b. A letter on signing up liability contracts of major pollutants. Beijing, China: 
State Environmental Protection Administration.
SEPA. 2006c. Management methods of environmental statistics. Beijing, China: State 
Environmental Protection Administration.
SEPA. 2006d. Monthly information on writing the national 11th five-­
year plan on envi­
ronmental protection. Beijing, China: State Environmental Protection Administration.


Mobilizing the government  75
SEPA. 2007a. Detailed methods to verify major pollutants emission reduction in the 11th 
five-­
year period (on trial). Beijing, China: State Environmental Protection Administration.
SEPA. 2007b. Lvliang and Liupanshui municipalities and Huadian corporation are finally 
removed from suspension. Beijing, China: State Environmental Protection Administration.
SEPA. 2007c. SEPA disclosed illegal construction projects and used ‘regional suspension 
of ratification’ policy. Beijing, China: State Environmental Protection Administration.
SEPA. 2007d. Verification of major pollutants emission reduction in the 11th five-­
year 
period (on trial). Beijing, China: State Environmental Protection Administration.
SEPA. 2008. Management methods of regional suspension of ratifying environmental 
impact assessment reports (on trial) (draft for comments). Beijing, China: State Envi­
ronmental Protection Administration.
Shanxi Bureau of Statistics. 2006. Shanxi statistics yearbook. Taiyuan, China: Shanxi 
Bureau of Statistics.
Shanxi Provincial Government. 2006. Notice on controlling major pollutants emissions in 
the 11th five-­
year plan. Taiyuan, China: Shanxi Provincial Government.
State Council. 1990. Regulations for the implementation of the law of standardization. 
Beijing, China: State Council.
State Council. 2005a. Decisions on realizing scientific view of development and strengthen­
ing environmental protection. Beijing, China: State Council.
State Council. 2005b. Several advices from the state council on strengthening the enact­
ment of economic and social development plans. Beijing, China: State Council.
State Council. 2006. Opinion on pollution emission quota distribution in the 11th five-­
year 
plan. Beijing, China: State Council.
State Council. 2007a. Notice on distributing composite working plan on energy conserva­
tion and pollutant emission reduction. Beijing, China: State Council.
State Council. 2007b. Notice on distributing implementation plans and methods of statis­
tics, monitoring and assessment on energy conservation and pollutant emission reduc­
tion. Beijing, China: State Council.
State Council. 2007c. Notice on distributing the 11th five-­
year plan on environmental pro­
tection. Beijing, China: State Council.
State Council. 2013a. Action plans on air pollution prevention and control. Beijing, China: 
State Council.
State Council. 2013b. Decisions on cancelling and decentralizing administrative ratifica­
tion [2013(19)]. Beijing, China: State Council.
State Council. 2014. Decisions on cancelling and decentralizing administrative ratification 
[2014(5)]. Beijing, China: State Council.
State Environmental Protection Administration (SEPA). 2001–2009. National report on 
environmental statistics. Beijing, China: State Environmental Protection Administration.
United Nations. 1992. United nations framework convention on climate change. New 
York: United Nations.
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. 
Congress.
U.S. Environmental Protection Agency. 2007. National emissions inventory (NEI) air pol­
lutant emissions trends data. Washington, DC: U.S. Environmental Protection Agency.
Wang, J., Wu, X., Cao, D. & Meng, F. 2004. Proposed scenarios for total emission control 
of SO2 during the 10th five-­
year plan period in China. Research of Environmental Sci­
ences, 17, 4.
Wang, X. 2002. Exponent on ‘national 10th five-­
year plan on environmental protection’. 
Beijing, China: Science Press.


76  Mobilizing the government
Xinhua News Agency. 2005. The birth of ‘the suggestions to 11th five-­
year plan’ [Online]. 
Beijing, China [Online]. Available: http://news.xinhuanet.com/politics/2005-­
10/26/con­
tent_3685219.htm.
Xinhua News Agency. 2006. The drafting of 11th five-­
year plan outline. Beijing, China 
[Online]. Available: http://news.xinhuanet.com/politics/2006-­
03/16/content_4308918.
htm.
Xu, Y. 2011. The use of a goal for SO2 mitigation planning and management in China’s 
11th five-­
year plan. Journal of Environmental Planning and Management, 54, 769–783.
Yang, X., Gao, Q., Jiang, Z., Ren, Z., Chen, F., Chai, F. & Xue, Z. 1998. Research on the 
transportation and precipitation regular pattern of sulfur pollutants in China. Research of 
Environmental Sciences, 11, 27–34.
Yang, X., Gao, Q., Qu, J. & Jiang, Z. 1999. The exploration and initial assessment of total 
amount control method for SO2 emission in China. Research of Environmental Sciences, 
12, 17–20.
Zhejiang Provincial Government. 2008. Notice on distributing Zhejiang provincial imple­
mentation plans and methods of statistics, monitoring and assessment on energy con­
servation and pollutant emission reduction. Hangzhou, China: Zhejiang Provincial 
Government.
Zou, S., Wang, J. & Hong, Y. 2006. Research report on national environmental protection 
plan in the 11th five-­
year plan. Beijing, China: China Environmental Science Press.
Zou, S., Zhang, Z., Yan, G. & Tian, R. 2004. Preliminary assessment on medium-­
term 
execution of the national tenth five-­
year plan outline, important eco-­
environmental con­
servation and environmental protection plan. In: Wang, J., Zou, S. & Hong, Y. (eds.) 
Environmental policy research series. Beijing, China: China Environmental Science 
Press.


1  
China’s challenges in policy making
Policies and goals are important in any country’s governance, but their relative 
roles could have two primary patterns under different governance models. Rules 
are set up through policies (and laws), while polluters and other stakeholders 
decide on their own actions according to the rules. In the rule-­
based governance, 
policies are in the first place, while goals are more implicit to take the second 
place. Another strategy explicitly makes goals in the first place, while policies 
are secondary and could be more flexible. With the rule of law not yet well estab­
lished, China would face great challenges in policy supply under the rule-­
based 
governance, especially given its rapidly evolving economy and society.
1.1  
Uncertain linkages between actions and outcomes
China is rapidly industrializing and the economy grows at a fast pace. It encoun­
ters great uncertainties on whether planned actions could achieve intended goals. 
Sulfur dioxide (SO2) emissions as well as other environmental problems tend to 
have a wide scope of influential economic, energy and environmental factors, as 
well as scattered emission sources in numerous important sectors. Many key fac­
tors for SO2 mitigation are beyond the jurisdiction of environmental protection, 
specifically the Ministry of Ecology and Environment (and previously the Minis­
try of Environmental Protection). Implementation is largely under the responsibil­
ity of local governments, while the central government is not designed and well 
equipped for primary policy implementation. In addition, China’s complexities 
can cast substantial uncertainties on whether preplanned actions can achieve their 
goals. China identified enough efforts to achieve the 10% reduction goals of SO2 
emissions in both the 10th and the 11th Five-­
Year Plans, but their outcomes dif­
fered from each other dramatically. In the Outline of the National 10th Five-­
Year 
Plan that was ratified by the National People’s Congress, the 10% reduction goals 
of “major pollutants” were clearly written (National People’s Congress, 2001). 
“Major Pollutants” were later defined to include SO2, dust, COD (chemical oxy­
gen demand), ammonia-­
nitrogen and industrial solid waste (SEPA, 2001). Exter­
nal measures, particularly energy conservation, did not show up in the national 
5	

Policy making


78  Policy making
Outline (National People’s Congress, 2001). However, in the special plan for 
energy development, China did propose a goal to reduce energy intensity by about 
15% to 17% and coal’s share in total energy consumption by 3.88% in the five 
years (NDRC, 2001). China’s annual economic growth rate, another key factor, 
was estimated to be 7% (National People’s Congress, 2001). Between 2001 and 
2005, the reversely calculated sulfur contents in coal from China’s official data 
went down from 1.22% to 1.05% (Xu et al., 2009), and a lot more SO2 scrub­
bers were installed (Figure 5.11). For the 10th Five-­
Year Plan, the SO2 mitigation 
shortfall was mostly due to the unexpected surge in coal consumption as a result 
of accelerated economic growth, 87.6% over the five years that overwhelmed the 
efforts of the State Environmental Protection Agency (SEPA; BP, 2019).
In the 11th Five-­
Year Plan, the planning structure differed only slightly. With 
the same 10% reduction goal, the Outline of the National 11th Five-­
Year Plan nar­
rowed the definition of “major pollutants” to cover only SO2 and COD for primary 
attention (National People’s Congress, 2006). Other pollutants were addressed in 
the special plan for environmental protection (State Council, 2007b). A goal on 
energy conservation, a 20% reduction of energy intensity, got promoted to the 
national Outline. Coal’s share in total energy consumption remained in the special 
plan for energy development, with a 3% drop in the five years (NDRC, 2007). 
China’s economy was estimated, or conservatively planned, to grow 7.5% per 
year (National People’s Congress, 2006). These figures were quite close to those 
in the 10th Five-­
Year Plan. Simply from the planning perspective, these two 10% 
reduction goals of SO2 emissions should both be attained. However, their results 
diverged significantly away from each other, which illustrated the difficulty to 
foresee the effects of policies and actions on goals.
1.2  
Challenges in policy making to induce actions
In the U.S.’s efforts to control SO2 emissions, the Clean Air Act Amendments 
(1990) established the Acid Rain Program that was distinguished as the most 
important law on the issue (The U.S. Congress, 1990). However, no individual 
environmental policy in China could claim an equal share of importance in its SO2 
mitigation cause. In comparison to goal-­
centered governance, policy supply under 
rule-­
based governance features fewer policies (or laws), and some are of crucial 
importance in achieving the intended goals of environmental protection. Each 
policy has a more extended enactment procedure and implementation horizon, 
which makes policy-­
making process lengthy and careful. The failure/success of 
a key policy thus takes on much heavier weight in environmental protection out­
comes. Nevertheless, in developed countries where the rule of law is well estab­
lished, the linkages between policies and polluters’ actions are more predictable, 
while these actions further contribute to intended outcomes. However, because 
China has not established a sound rule of law, confidence is much lower that a 
policy can be implemented well to induce the intended actions.
Many environmental policy instruments have been designed and applied across 
countries. The first major category involves command and control policies, such 


Policy making  79
as mandatorily shutting down polluting sources, setting pollutant emission and 
energy efficiency standards and mandating the application of the best available 
technologies. Another major category is based on economic incentives and mar­
kets. Typical policy instruments include effluent emission discharge fees, taxes, 
tradable permits and subsidies. Information disclosure, such as labeling and cer­
tificates, aims to enable consumers to voluntarily make informed consumption 
choices for minimizing environmental impacts.
Despite some unique features, China’s policy toolbox for SO2 mitigation was 
not fundamentally different from that in developed countries with rule-­
based gov­
ernance. In China, an engineering approach that was based on SO2 scrubbers in 
coal-­
fired power plants involved many command and control policies for their 
deployment and normal operation to meet effluent emission standards (Minis­
try of Environmental Protection and General Administration of Quality Super­
vision Inspection and Quarantine, 2011; the SEPA and General Administration 

of Quality Supervision Inspection and Quarantine, 2003). China has also been 
experimenting with market-­
incentive policies, such as cap-­
and-­
trade, an effluent 
emission fee or an emission tax (Yan et al., 2009; Dong et al., 2011; Ge et al., 2011; 
Zhang et al., 2016). Technological licensing from developed countries, through a 
functioning technology market, was a cornerstone in China’s SO2 mitigation to 
build a domestic industry for rapid deployment and cost reduction (Xu, 2011).
After assessing the effectiveness and efficiency of individual environmen­
tal policy instruments, policies are enacted for tackling a given environmental 
problem (Barron and Ng, 1996; Goulder and Parry, 2008). A few criteria could 
be important in making an optimal policy, including cost-­
effectiveness; capabil­
ity to address uncertainty, synergy or conflict with current policy instruments; 
compliance monitoring and inspection capacity and requirements; and compli­
ance of polluters. The latter two are especially relevant to developing countries 
like China, where the rule of law has not been well established and environmen­
tal noncompliance might be prevalent. In developed countries, there has been 
an increasing trend in the application of market-­
based instruments (Portney and 
Stavins, 2000; Tietenberg, 1990). Cost-­
effectiveness is the most important argu­
ment for their adoption considering particularly the reduced abatement costs 
(Goulder and Parry, 2008). For example, in the U.S. Acid Rain Program in the 
Clean Air Act Amendments (1990), total SO2 emissions from coal power plants 
were capped and emission permits were allowed to trade in a market (The U.S. 
Congress, 1990). The policy substantially reduced the abatement costs compared 
with command-­
and-­
control instruments (Benkovic and Kruger, 2001).
Environmental policy instruments differ from each other in their capability of 
addressing uncertainties. For example, environmental taxes establish certain lev­
els of emission prices but leave the quantities of emissions uncertain. In contrast, 
tradable permits with a fixed cap are more certain about the quantity within the 
defined boundary of emission sources but not about the price. Other instruments 
all have various impacts on uncertainties (Goulder and Parry, 2008). The intro­
duction of a new environmental policy instrument should consider how it interacts 
with existing policies to create synergies or conflicts. If a new emission trading 


80  Policy making
policy is imposed into an area that is already dominated by command and control 
policies, it may not be able to achieve its intended cost-­
effectiveness (Zhang et al., 
2013). China’s experiments of SO2 emission trading schemes encountered major 
problems, including frequent governmental intervention and inter-­
policy conflicts 
together with the quality of policy design (Zhang et al., 2016).
Compared with developed countries, developing countries and specifically 
China have more difficulties in making optimal policies. Research tends to be 
thinner especially in the past to understand how individual policy instruments 
perform in their contexts. Significant constraints on environmental policy imple­
mentation may exist due to the lack of adequate financial resources, personnel 
and necessary expertise (Blackman, 2010). More details on China’s policy imple­
mentation problems are discussed in Chapter 6. The effectiveness of individual 
policies could be very uncertain with unpredictable implementation, which makes 
policy design challenging.
2  
Goal-­
centered policy supply
China’s policy supply follows a very different pattern from that in rule-­
based gov­
ernance. Goals play the central role in environmental governance, while policies 
as means to achieve goals are primarily instrumental and failures of individual 
policies are more accommodated. Important centralized goals as those few in 
National Five-­
Year Plans drive decentralized policies, laws and regulations from 
ministries, local governments, the People’s Congress and other stakeholders. For 
those environmental fields without goals or with goals but at lower priorities, 
policy supply tends to be less adequate and strong.
2.1  
Enabling goal-­
centered policy supply
Goal-­
centered governance in China is enabled by centralized national goals, 
decentralized goal attainment, decentralized policy making and implementation 
and mobilized central and local governments. In the past four decades, China’s 
environmental governance has been heavily decentralized, as discussed in Chap­
ter  3. The four levels of governments  – central, provincial, municipality and 
county – have diverging divisions of governmental authorities and functions. As 
matched by their personnel categories and fiscal expenditures, the central gov­
ernment heavily focuses on policy making, while the county-­
level governments 
are almost entirely on policy implementation. Provincial-­
 and municipality-­
level 
governments have significant authorities and functions on both. Local govern­
ments hold significant decentralized authorities in initiating local policy innova­
tion, learning and adopting policies from other regions and implementing various 
policies. Without the cooperation and mobilization of local governments, the cen­
tral government can hardly achieve serious SO2 mitigation or any environmental 
cleanup.
However, much authority remains substantially centralized, especially setting 
up national goals for environmental protection. Various considerations for or 


Policy making  81
against strong environmental protection are centrally weighed to form strong or 
weak political will by the top leadership of the Chinese Communist Party, as dis­
cussed in Chapter 2. It is then reflected in Five-­
Year Plans. When the top leader­
ship determines to prioritize environmental protection among other governmental 
affairs, pollution mitigation goals started to enter as the key goals into National 
Five-­
Year Plans. These national goals are then decomposed into provincial goals 
for their implementation, as examined in Chapter 4. The goal allocation further 
penetrates into municipality and county levels, one level at a time. The types and 
stringency of goals closely follow the centralized political will for environmental 
protection. In addition, ministries and their internal departments in the central 
government are also directed by those goals to make policies and supervise pro­
vincial and other local governments for goal attainment. If a crucial environmen­
tal goal in the Five-­
Year Plan is missed, the Ministry of Ecology and Environment 
as the primary responsibility bearer will also be held accountable.
Credible mechanisms are established for central and local governments to make 
efforts for their goals. Most important, provincial leaders and ministers in the cen­
tral government have their promotion opportunities controlled centrally through 
the Chinese Communist Party, while the fates of municipality-­
level leaders are 
determined at the provincial level. The clear linkages between their career devel­
opment and goal-­
centered job performance are crucial incentives to motivate their 
genuine efforts but not just lip service.
Under goal-­
centered governance, it is the succession of goals but not individual 
policies that define environmental milestones. The top leadership of the party and 
the central government cares more about whether a certain goal has been achieved 
rather than a certain policy has been effective, efficient or fully implemented. 
In addition, the fact that China has not established a sound rule of law is also 
an important facilitating factor for enabling a goal-­
centered policy supply. Local 
leaders in charge, such as provincial governors, municipality mayors and county 
leaders as well as their corresponding party secretaries, will be less likely to lose 
their jobs or promotion opportunities for having policy failures, but the probabil­
ity will increase significantly if a crucial goal does not get achieved. If one policy 
does not work, new ones will be quickly enacted to inch toward goal attainment.
2.2  
Policy evolution by implementation selection
Policy supply under goal-­
centered governance has two key components: environ­
mental goals to shape policy demand and low policy-­
making barriers and strong 
incentives to enable policy supply. More ambitious environmental goals will cre­
ate stronger demand for pollution mitigation actions and thus a larger number 
of and more stringent policies. Goal-­
centered governance significantly reduces 
barriers for making policies. The much lower policy-­
making barriers result in 
intensive policy-­
making activities, competition among policies and much faster 
policy cycles. With a significant number of policies, each makes a small step 
toward an intended goal, although some are more important than others. The fail­
ure/success of any policy does not determine, but only to a limited extent affects, 


82  Policy making
the final environmental outcome. Besides laws, a large number and wide variety 
of policies can be found in China on environmental protection that are enacted by 
various authorities, including the Central Committee of the Chinese Communist 
Party, the State Council, ministries and their composing departments (www.mee.
gov.cn/zcwj/) and local governments.
Several causes contribute to the low policy-­
making barriers. The significantly 
decentralized policy making effectively reduces the barriers from the perspective 
of policy suppliers as they have a wide variety of sectoral and geographic jurisdic­
tions and authorities. One consequence of this goal-­
centered policy supply is that 
it encourages policy innovation. Local governments have significant flexibility in 
deciding how to achieve top-­
down goals. Decentralized policy makers can weigh 
the significance, costs and benefits of various policies and their suitability to local 
contexts with dramatic regional disparities. Policies are constantly churned out 
from these decentralized policy makers at various levels to try their effectiveness 
in approaching goals. The effective mobilization of local governments not only 
facilitates policy enforcement, but it also creates incentives for even more active 
local environmental policy making if goal attainment so requires.
Furthermore, several key questions should be considered over the making of 
individual policies, while goal-­
centered governance has much lower require­
ments on policy designs to effectively decrease the policy-­
making barriers. 
First, how to ensure the quality of individual policies? Policies may be directly 
adopted from other countries and regions, revised to suit local contexts or inno­
vated from scratch. China’s colossal size and complexity indicate that many 
environmental policies can hardly be applied to fit all situations across the entire 
country. China’s contexts are also sharply different from those in developed 
countries, where many environmental policies were first introduced and imple­
mented. The decentralization of policy makers also indicates that the training 
and knowledge of those who write the policy texts may vary across local gov­
ernments and ministries/departments. The much more greatly decentralized pol­
icy implementation and its unsatisfactory track record add further difficulties in 
understanding how policies could be designed better for more effective imple­
mentation. Accordingly, direct policy adoption is rarely effective, while policy 
localization and innovation are great challenges and require relevant knowledge 
and understanding. In addition, China’s complexity also hinders timely-­
enough 
assessment of the crucial causes of any policy failure and success. Under goal-­
centered governance, the requirements on the quality of making individual poli­
cies are much lower because no policy or law occupies the central stage to solve 
a targeted environmental problem. The lower requirement for policy quality 
enables much swifter design and enactment processes. In other words, read­
ers of China’s environmental policies should not be primarily entangled in the 
enactment and effectiveness of individual policies, because they are of much 
less importance than goals. For example, essentially no SO2 emission trading 
schedules have produced desirable outcomes that dominate SO2 mitigation, but 
the failure had little impact on China’s trajectory of controlling SO2 emissions 
(Zhang et al., 2016).


Policy making  83
Second, how to choose the most effective and efficient policy instrument among 
many alternatives? The choice of policy instruments is a crucial question for pol­
icy making, especially when a single or very few policies dominate the solutions 
to an environmental problem. Under goal-­
centered governance, this question is 
much less significant because policies are much less mutually exclusive. The con­
siderably decentralized policy making also significantly reduces the possibility of 
any policy monopoly or oligopoly. The enactment of one policy instrument does 
not prevent the application of others. Accordingly, China does not need to choose 
a primary policy instrument for dealing with one environmental problem. For 
example, China’s environmental protection tax law formally entered into force 
in January 2018, covering a wide variety of environmental pollutants, including 
SO2 (National People’s Congress, 2016). Many other crucial environmental poli­
cies are simultaneously in effect, such as the effluent emission standards that were 
examined earlier (MEP and AQSIQ, 2011).
Third, how are policies coordinated? While policies are individually made by 
different ministries and their internal departments, as well as various levels of 
governments, they can exert significant impacts on each other to create synergies 
and/or conflicts. Economic and energy policies are far beyond the jurisdiction 
of environmental protection. With local governments rather than their environ­
mental protection bureaus in charge, coordination across these different types of 
policies became more feasible. In an optimized situation, policies should be well 
coordinated to maximize synergies and minimize conflicts. However, such coor­
dination in China is inadequate in the context of decentralized policy making 
and especially policy implementation. Little evidence indicates that China rolls 
out the numerous policies for achieving the SO2 mitigation goals in a system­
atic and coordinated way. Instead, the policy making is messy, with decentralized 
policy makers who have their individual authority in designing or shaping policies 
within their respective jurisdictions. Under goal-­
centered governance, however, 
such prior coordination of policy making is of lesser importance. After policies 
are made and put into implementation, they evolve rapidly. In China’s context of 
weak rule of law and as examined earlier, individual policies have higher prob­
abilities of unsatisfactory implementation. Similar to the natural selection process 
as proposed by Charles Darwin in understanding biological evolution (Darwin, 
1859), policies in China also experience a dynamic evolution process and those fit 
ones are selected through implementation. Policies that have too many conflicts 
with others will be difficult to get effectively implemented. If one policy fails 
to achieve its intended consequences, new policies can be quickly introduced. 
Successful policies in one province can be rapidly adopted by other provinces or 
elevated to the national level.
Although much progress has been made in policy research in the past decade, 
such capacity was especially deficient in the early stages of SO2 mitigation. China 
should still enhance its capability in policy making to improve the quality of indi­
vidual policies, choose more wisely environmental policy instruments especially 
for those of relatively greater importance and scopes and better coordinate across 
policies. Nevertheless, the goal-­
centered policy supply substantially lowered the 


84 Policy making
requirements for achieving desirable environmental protection outcomes such as 
serious mitigation of SO2 emissions. The preceding crucial questions in policy 
making are of much less concern from their perspectives on influencing policy 
outcomes.
3  
Policy scope for achieving SO2 mitigation goals
China faces a wide scope in policy making for SO2 mitigation. SO2 emissions 
are affected by many economic, energy and environmental development factors 
and corresponding policies. Although the coal-fired 
 
power sector is increasingly 
important in coal consumption, still nearly two fifths of coal is consumed in other 
sectors (Figure 1.10). For achieving increasingly stringent SO2 mitigation and 
environmental goals, the decentralized policy makers should evaluate the contri-
butions of individual policies in policy supply.
3.1   
Key factors for SO2 emissions
SO2 emissions can be decomposed with the following formula into various key 
factors:
Energy
Coal
SO2 emissions
SO2 emissions = GDP ×
×
×
GDP
Energy
Coal
Coal
  
Equation 5.1
= GDP ×
×
EI
×
×
h
h
s
s
(
)
1
2
−
×
r
R
×
−
(
)
1
h
Energy
“GDP” (gross domestic product) indicates the scale effect. Rapid economic 
growth in China leads to more SO2 emissions. Energy consumption is a key foun-
Energy 
dation for any modern economy, and thus, energy intensity 


 is another 



GDP 

crucial effect. It measures how much energy is consumed for producing a given 
unit of GDP. Energy conservation and efficiency will reduce energy intensity and 
thus be beneficial for SO2 mitigation. The economic structure also matters greatly. 
A greater proportion of service sectors in an economy could potentially reduce 
the overall energy intensity because in comparison to industrial sectors, they tend 
to consume much less energy for producing the same amount of economic out-
puts (Feng et al., 2009). China had a goal to reduce energy intensity by 20% in 
the 11th Five- 
Year Plan (National People’s Congress, 2006). The Chinese central 
government also declared its intention in the 12th Five- 
Year Plan to “change the 
economic growth pattern,” with a focus on energy conservation and environmen-
tal protection (National People’s Congress, 2011). These two effects are related to 
economic development and energy conservation, on which economic and energy 
policies exert important influences.
Because coal consumption dominates the sources of SO2 emissions, the share 
of coal in the energy mix is thus critical in deciding the sulfur intensity of energy. 


Policy making  85
Coal
Energy is referred to as the energy transition effect. Its reduction is another meas­
ure for bringing down SO2 emissions, which largely falls into the category of 
energy development and the scope of energy policy.
SO emissions
Coal
2  
 refers to the mitigation effect, which is primarily decided by 
environmental policies. In combustion, a certain proportion of sulfur (ηsr ) will 
be retained in ash and thus not emitted. This rate is mainly decided by the coal 
type and combustion technology, but not by policy intervention. Sulfur content in 
coal (ηs) is an important indicator of coal quality. The control of sulfur contents is 
often targeted in early environmental regulations for reducing SO2 emissions. SO2 
scrubbers and other SO2 removal measures can avoid a certain share of SO2 (ηR) 
from being emitted after generation.
China’s economy has been growing at an astonishing pace in the past four dec­
ades. Real GDP in 2018 was 31.7 times of that in 1980 with a growth rate of 9.5% 
annually, while real GDP per capita rose to be 22.4 times or 8.5% annually (Fig­
ure 5.1). As measured in nominal GDP of current U.S. dollars, China overtook 
Japan to become the second-­
largest economy in the world in 2010 and further rose 
to be equivalent to 65.0% of the United States in 2018 (Figure 5.1). China’s much 
larger population indicates that the country’s GDP per capita still trails the global 
average and is a small fraction of that in Japan and the United States. Although 
the GDP growth rate has been significantly slower in the 2010s than in the 2000s, 
0
5,000
10,000
15,000
20,000
25,000
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
80,000
90,000
1980
1985
1990
1995
2000
2005
2010
2015
Nominal GDP (Billion US dollars)
)
B
M
R
 
5
1
0
2
(
 
a
t
i
p
a
c
 
r
e
p
 
P
D
G
 
&
 
P
D
G
Year
Real GDP (Billion RMB, China; left)
Real GDP per capita (RMB per capita, China; left)
Nominal GDP (Billion US dollars, China; right)
Nominal GDP (Billion US dollars, Japan; right)
Nominal GDP (Billion US dollars, US; right)
Figure 5.1  
Economic growth in China, Japan and the United States
Source: IMF (2019).


86  Policy making
the convergence of average living standards in China toward that of developed 
countries is expected to further intensify economic activities within its geographi­
cal territory and thus to add great environmental pressures.
Energy consumption is not only one key foundation for economic develop­
ment, but it also brings unwanted consequences of environmental pollution. The 
combustion of fossil fuels, especially coal, is the primary source of air pollutant 
emissions that cause ambient particulate matter (PM) pollution. Although China 
has been improving its energy efficiency for producing one unit of GDP espe­
cially in the past decade, its primary energy consumption climbed up quickly. 
When consuming one ton of oil equivalent of primary energy, China in 2018 
produced US$4,084 of nominal GDP, while the rates for Japan and the United 
States were US$10,948 and US$8,945, respectively (IMF, 2019; BP, 2019). Due 
to the significantly lower energy efficiency, China overtook the United States 
to become the largest energy consumer in the world in 2009, but its economy 
then was two thirds smaller. A major shift took place in around 2003, and since 
then, China’s energy consumption has been growing at a much faster pace than 
before (Figure 5.2). Not only China’s economic growth accelerated after 2003, 
but also the energy efficiency reversed its earlier improvement trend to decrease 
between 2002 and 2005 (Figure 5.2). In 2018, China consumed 224% more pri­
mary energy than in 2000 to become 42% higher than the United States’ level 
(Figure 5.2).
0.0
5.0
10.0
15.0
20.0
25.0
30.0
0
500
1,000
1,500
2,000
2,500
3,000
3,500
1980
1985
1990
1995
2000
2005
2010
2015
Energy efficiency (1,000 (2015) RMB/toe)
)
e
o
t
M
(
 
n
o
i
t
p
m
u
s
n
o
c
 
y
g
r
e
n
e
 
y
r
a
m
i
r
P
Year
China
India
United States
Energy efficiency in China (right)
Figure 5.2  
Primary energy consumption and energy efficiency
Source: BP (2019).


Policy making  87
China’s low level of GDP per capita might partly explain why its energy mix 
heavily focuses on coal. As shown in Figure 5.3, coal is the cheapest and most 
affordable among the three major fossil fuels. When China’s economy grew very 
fast, especially in the 2000s, to rapidly push up energy consumption, coal became 
the primary choice for meeting the additional energy demand (Figure 5.4), and 
thus, its share in the energy mix even reversed its earlier declining trend to become 
higher in the early 2000s (Figure 5.5).
In the past decade, energy transition has also been playing an increasingly vis­
ible role that led to the mitigation of SO2 emissions. China’s energy mix is heav­
ily tilted toward coal, the most pollution-­
intensive fuel. With more coal burning 
squeezed into China’s territory, the pressure on the environment is mounting. 
Throughout the 1980s and 1990s, the share of coal was continuously above 70% 
(Figure 5.5). The slow declining trend in the 1990s was reversed in early 2000s to 
witness the share climbing up again from 69.5% in 2001 to 73.7% in 2007, further 
intensifying environmental pollution in China. The following decade witnessed 
an unprecedented decrease and coal’s share had dropped to 58.2% in 2018. Never­
theless, China still accounted for 50.5% of global coal consumption in 2018 (BP, 
2019). Although oil and natural gas have increasing shares in China’s primary 
energy consumption, the overall share of fossil fuels experienced an accelerated 
decline from 94.1% in 2007 to 85.3% in 2018. Nonfossil fuels are much more 
0
2
4
6
8
10
12
14
16
18
20
1980
1985
1990
1995
2000
2005
2010
2015
)
J
G
/
$
 
t
n
e
r
r
u
c
(
 
e
c
i
r
p
 
y
g
r
e
n
E
Year
Oil
Gas
Coal
Figure 5.3  
Prices of coal (Qinhuangdao spot price), oil and natural gas
Source: Japan LNG CIF; BP (2019).


88  Policy making
–50
0
50
100
150
200
250
1980
1985
1990
1995
2000
2005
2010
2015
)
e
o
t
M
(
 
n
o
i
t
p
m
u
s
n
o
c
 
y
g
r
e
n
e
 
f
o
 
h
t
w
o
r
g
 
l
a
u
n
n
A
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Figure 5.4  
The annual growth of primary energy consumption in China by fuels
Source: BP (2019).
55%
60%
65%
70%
75%
80%
85%
90%
95%
100%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
1980
1985
1990
1995
2000
2005
2010
2015
Share in the energy mix
)
e
o
t
M
(
 
n
o
i
t
p
m
u
s
n
o
c
 
y
g
r
e
n
e
 
y
r
a
m
i
r
P
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Coal’s share
Fossil fuels’ share
Figure 5.5  
China’s primary energy consumption by fuel and the shares of coal and fossil fuels
Source: BP (2019).


Policy making  89
important in the energy mix, from 5.9% in 2007 to 14.7% in 2018 (Figure 5.5). 
Nuclear, hydropower and nonhydro renewables witnessed their shares increased 
from 0.7%, 5.1% and 0.2% in 2007 to 2.0%, 8.3% and 4.4% in 2018, respectively. 
Nonhydro renewables were the fastest-­
growing energy type.
As one indicator of energy modernization, China’s primary energy consump­
tion is rapidly electrifying to reshape the major sources and sectors of SO2 emis­
sions. In 1990, only 20.8% of primary energy consumption went through the 
intermediate stage of electricity before final consumption, which was only slightly 
higher than Africa’s 17.8% (Figure 5.6). With rapid energy modernization, this 
ratio increased to 42.5% in 2015, then similar to the United States’ 40.3% (Fig­
ure 5.4). Rapid electrification also happened in other rapidly industrializing coun­
tries such as India, but the progress in Africa has been much slower (Figure 5.4). 
With China’s continuous efforts for electrifying energy consumption – such as the 
push for electric vehicles (IEA, 2019) – this electrification rate is expected to fur­
ther escalate, which will distinguish the importance of the power sector in China’s 
energy consumption and environmental protection.
Energy transition for electricity generation is even more visible. Coal’s share 
has been reduced significantly from the 81.0% peak in 2007 to 66.5% in 2018. 
Other fossil fuels, including oil and natural gas, accounted for only an insignificant 
share at 3.3% in 2018 (Figure 5.7). In contrast, the share of nonhydro renewables, 
mostly wind and solar energy, has achieved the largest growth from 0.5% in 2007 
to 8.9% in 2018 (Figure 5.7). Coal’s share in electricity generation is significantly 
0%
5%
10%
15%
20%
25%
30%
35%
40%
45%
0
500
1,000
1,500
2,000
2,500
3,000
1990 2000 2015
1990 2000 2015
1990 2000 2015
1990 2000 2015
Share of power generation
)
e
o
t
M
(
n
o
i
t
p
m
u
s
n
o
c
y
g
r
e
n
E
Power generation
Others
Share of power generation
United States
China
Africa
India
Figure 5.6  
Primary energy consumption and its electrification rate
Source: IEA (2017).


90  Policy making
higher than that in primary energy consumption, being 66.5% and 58.2% in 2018, 
respectively (Figure 5.8). From 2007 to 2018, their drops were 14.4% and 15.4% 
in percentage points, respectively. Nonfossil fuels – such as nuclear, hydro and 
nonhydro renewables – are generally for electricity generation, while oil and natu­
ral gas in China are primarily consumed not in the power sector. Especially in the 
past decade, the advancement of renewables significantly accelerated to account 
for increasingly sizable shares of electricity generation growth (Figure 5.8).
The power sector has been increasing its importance in coal consumption. In 
1980, only 20.2% of China’s coal consumption was in the power sector, while 
this ratio climbed steadily to 52.2% in 2002 before a decade-­
long stabilization 
(Figure 1.10). During 2015~2017, the increasing trend restarted to reach 57.3% in 
2017 from 50.3% in 2014 (Figure 1.10). This ratio is expected to further increase, 
in reference to the situation in the United States, whose power sector accounted 
for 18.6% of coal consumption in 1950 and 92.8% in 2017 (Figure 1.10). The 
trend indicates that the energy mix in nonpower sectors shifts away from direct 
coal consumption faster than that in the power sector, although the former may 
consume more electricity that comes from coal-­
fired power plants.
In China’s trajectory of SO2 mitigation, these economic, energy and envi­
ronmental factors made different contributions in different Five-­
Year Plans 
(Figure  5.9). SO2 emissions went down by 15.8% in the 9th Five-­
Year Plan 
50%
55%
60%
65%
70%
75%
80%
85%
90%
0
1,000
2,000
3,000
4,000
5,000
6,000
7,000
8,000
1985
1990
1995
2000
2005
2010
2015
Share in the energy mix
)
h
W
T
(
 
n
o
i
t
a
r
e
n
e
g
 
y
t
i
c
i
r
t
c
e
l
E
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Others
Fossil fuels’ share
Coal’s share
Figure 5.7  
Electricity generation by fuels in China
Source: BP (2019).


Policy making  91
50%
55%
60%
65%
70%
75%
80%
85%
–50
50
150
250
350
450
550
650
1985
1990
1995
2000
2005
2010
2015
Coals shares
)
h
W
T
(
 
n
o
i
t
a
r
e
n
e
g
 
y
t
i
c
i
r
t
c
e
l
e
 
f
o
 
h
t
w
o
r
g
 
l
a
u
n
n
A
Year
Coal
Oil
Gas
Nuclear
Hydro
Renewables
Others
Coal’s share in electricity generation
Coal’s share in primary energy
Figure 5.8  
The annual growth of electricity generation in China by fuels and coal’s share
Source: BP (2019).
–60%
–50%
–40%
–30%
–20%
–10%
0%
10%
20%
30%
40%
50%
60%
SO2 emissions
Scale effect
Energy intensity
effect
Energy transition
effect
Mitigation effect
e
v
i
F
 
r
e
v
o
 
e
g
n
a
h
C
-
 
s
n
a
l
P
 
r
a
e
Y
9th FYP (1996–2000)
10th FYP (2001–2005)
11th FYP (2006–2010)
12th FYP (2011–2015)
13th FYP (2016–2017)
Figure 5.9  
The decomposition of China’s SO2 emissions
Source: National Statistics Bureau and Ministry of Ecology and Environment (2019); IMF (2019); 
BP (2019).
Note: Method comes from Ang (2005).


92 Policy making
(1996–2000). Under the influence of the Asian financial crisis of 1997, the scale 
effect would still lead to an increase of 38.0%, while the effects of energy inten-
sity, energy transition and mitigation reduced SO2 emissions by 26.4%, 6.1% and 
21.3%, respectively, over the five years (Figure 5.9). They indicated the varying 
impacts of economic, energy and environmental policies and development. Spe-
cifically, as represented in the mitigation effect, environmental policies made an 
important but certainly not decisive contribution. The 10th Five-Y
 ear Plan had a 
very different picture. With accelerated economic growth, the scale effect would 
boost emissions by 53.0%, while the energy intensity and energy transition effects 
also pushed the emissions upward by 12.6% and 5.7%, respectively. Although 
the mitigation effect of 43.5% reduction was much greater than that in the 9th 
Five- 
Year Plan, the outcome was that SO2 emissions increased by 27.8%. In other 
words, the deterioration was not due to less effective environmental policies, but 
faster economic expansion reversed trends of energy intensity and transition.
Reversing the deterioration trend, the 11th Five- 
Year Plan managed to reduce 
SO2 emissions by 14.3%. The following 12th Five- 
Year Plan registered a similar 
reduction of 14.9%. The four effects also had comparable contributions in these 
two Five- 
Year Plans: the scale effect 49.6% versus 35.0%, the energy intensity 
effect −19.6% versus −17.6%, the energy transition effect −4.2% versus −9.1% 
and the mitigation effect −40.0% versus −23.2% (Figure 5.9). In the first two 
years (2016–2017) of the 13th Five- 
Year Plan with available data, SO2 emissions 
dropped by whopping 52.9%, and the mitigation effect contributed decisively a 
reduction of 52.0% (Figure 5.9).
3.2   
Technological factors for effluent SO2 emissions  
in the power sector
Electrification of energy consumption and the power sector’s increasing share in 
coal consumption distinguish the importance of coal-fired 
 
power plants in control-
ling China’s SO2 emissions. Given the understanding of coal combustion and SO2 

SO  emissions

emissions in electricity generation, the SO  emission intensity 
2

2

 


Coal



can be converted to effluent SO2 concentration, mg/Nm3 (Ministry of Environ-
mental Protection and General Administration of Quality Supervision Inspection 
and Quarantine, 2011). The SO2 concentration is measured under normal condi-
tions (thus “N”), with excess air coefficient being 1.4. The value 1.4 here indi-
cates that 40% more air or, specifically, oxygen will be blown into boilers than 
what is required for complete combustion. Excess air is needed for more complete 
combustion within a short residence time of fuels in boilers, but excess air will 
also take heat away and lower the thermal efficiency. Accordingly, an optimum 
value exists, not necessarily being 1.4 for every power plant. The fixed value is 
for policy purposes and intends to prevent cheating because a convenient option 
of lowering effluent SO2 concentration is to dilute the flue gas with more air. 
21%
Approximately, excess air coefficient a can be calculated as a ≈
, x% 
21%
%
−x
referring to the percentage of O2 in flue gas, when a = 1.4 and x%
%
»
≈6 .
a
−
21
21
%
%
%
x


Policy making  93
As revealed in Equation 5.1, there are three key technological factors to decide 
SO2 emission intensity and effluent SO2 concentration. The first factor is the frac­
tions of sulfur retained in ash (ηsr). When coal is burned in boilers, sulfur is con­
verted into several forms, being gaseous (SO2, SO3, gaseous sulfates) and solid 
(in bottom ash and as particulate sulfate; EPA, 1998). SO2 greatly dominates the 
gaseous forms (EPA, 1998). Higher combustion temperature leads to lower frac­
tions of sulfur retained in ash, which disadvantages pulverized coal (PC) combus­
tion against fluidized bed combustion (FBC; Sheng et al., 2000). Another factor is 
the calcium/sulfur (Ca/S) molar ratio in coal: a higher Ca/S ratio facilitates sulfur 
retention (Cheng et al., 2004; EPA, 1998). (The Ca/S ratio here is different from 
the Ca/S ratio for SO2 scrubbers as discussed later.) For PC combustion, the pres­
ence of calcium is much less important than FBC due to the thermal instability of 
calcium sulfate (CaSO4), the main product responsible for sulfur retention (Sheng 
et al., 2000). The fractions of sulfur retained in the application are summarized 
in Table 5.1. Compared with China’s assumption of 20% (State Council, 2007a), 
fractions of sulfur retained are widely believed to be significantly lower, except 
for the situations of burning lignite and applying FBC technologies. Even another 
Chinese official document believed the rate to be 10% to 15% and recommended 
10% for the purpose of designing SO2 scrubbers (NDRC, 2004). China’s SO2 
emissions could have been underestimated partly because of the choice of this 
parameter (Figure 1.8).
Second, lower sulfur contents in coal are crucial for reducing SO2 generation 
intensity. An official data set is employed to analyze the distribution of sulfur 
contents for SO2 scrubbers. In the 11th Five-­
Year Plan on Acid Rain and SO2 
Pollution Control, China published data for 248 coal power plants with a total 
capacity of 164 GW that covered all SO2 retrofit projects to be completed between 
2006 and 2010 (SEPA and NDRC, 2008). Sulfur contents estimated from this data 
Table 5.1 Applied fractions of sulfur retained in ash
Coal type
Fractions of 
Source
sulfur retained 
in ash
Bituminous, PC*
5%
U.S. Environmental Protection (EPA, 1998)
Sub-bituminous, PC* 12.5%
Agency’s (EPA’s) choice
Lignite, PC*
25%
Coal in general
≤10%
U.S.’s study
(Singer, 1981)
Nonlignite coal
5%
Assumption from the U.S. EPA (Smith et al., 2001)
Lignite
30%
Assumption in the research
Coal
5%~30%
The study’s assumption
(Ohara et al., 2007)
Coal
5%~10%
China’s study
(Zhao et al., 2008)
Coal, PC*
10%~15%
China’s official recommendation (NDRC, 2004)
for scrubber design
Coal
20%
Assumption in compiling 
(State Council, 2007a)
China’s statistical data
Note: PC* refers to pulverized coal power plants.


94  Policy making
set are expected to represent their distribution for all coal power plants with SO2 
scrubbers. Each plant disclosed information on scale (MW), year, annual SO2 
removal capability (tons per year), location and name. Sulfur contents can then 
be estimated under the following assumptions: thermal efficiencies were 370 g 
of coal equivalent per kilowatt-­
hour, or 1,930 kWh per ton coal (the average effi­
ciency in 2005 [China Electricity Council, 2006–2015]); capacity factors were 
5,500 hours per year (SEPA, 2006a); 80% of the sulfur was converted to SO2 in 
combustion and 20% was retained in ash, as recognized in China’s official sta­
tistics (State Council, 2007a); and overall SO2 removal rates were 85% (SEPA, 
2007). The calculation formula is
Sulfur content
 
SO  removal capability
Coal power capacity
2

55
1930
8
85
00
2
0
 

%
%
“2” in the denominator refers to the fact that when sulfur is converted to SO2, the 
mass doubles since the molecular weight of SO2 is twice that of sulfur. A caveat 
is that these numbers are used here to reversely calculate sulfur contents because 
they represent China’s original assumption in compiling the data. One legitimate 
concern is the accuracy of the assumed 80% conversion. Actually, in the com­
bustion of anthracite, bituminous and sub-­
bituminous coal, 90% or more of the 
sulfur is converted to SO2, as discussed earlier. In addition, as discussed in Chap­
ter 6, China’s actual SO2 removal rates should be significantly lower than 85% 
especially before 2007. Actual thermal efficiencies and capacity factors also vary 
across years.
Sulfur contents are closely related to the costs of SO2 mitigation. Generally 
speaking, higher sulfur contents correspond to lower costs for every ton of SO2 
removed but higher costs for every kilowatt-­
hour of electricity generated. The 
distribution of sulfur contents is shown in Figure 5.10 with the national average 
being about 1.0%. Of the coal-­
fired power plants, 68% burned coal with less 
than 1% sulfur and another 26% between 1% and 2%. The remaining 6% of the 
total capacity was associated with higher than 2%-­
sulfur coal. China not only 
installed SO2 scrubbers not only in coal power plants burning high-­
sulfur coal 
but also in those burning low-­
sulfur coal. China’s distribution of sulfur contents 
had a single peak at around 0.75% (Figure 5.10), which reflected the fact that 
most of China’s coal is mined in one region. For example, two thirds of China’s 
coal production in 2007 came from the seven nearby provinces of Shanxi, Inner 
Mongolia, Shaanxi, Shandong, Anhui, Hebei and Henan (National Bureau of 
Statistics, 1997–2008).
The preceding two factors decide how much SO2 is generated when burning 
a unit quantity of coal. SO2 removal rates are the third factor to reduce the SO2 
emission intensity. Before construction begins, a report of environmental impact 
assessment (EIA) had to be submitted to a governmental authority on environ­
mental protection (NPC, 2002). If the plant was believed to bring unacceptable 
environmental damage – for example, seriously worsen ambient air quality – the 


Policy making  95
EIA report would be rejected. Another policy – “three simultaneities” – required 
pollution control facilities to be designed, constructed and completed at the same 
time as the main project (State Council, 1998). For example, if SO2 scrubbers 
were considered necessary in the EIA report, this policy would demand their 
installation.
3.3  
Technical measures for SO2 removal in the power sector
In order to remove SO2 in electricity generation, coal-­
fired power plants in China 
are required to meet effluent emission standards that are made more stringent 
every six or seven years to reflect growing environmental concerns. In the stand­
ards enacted in 1996, new coal-­
fired power plants that passed EIA after Janu­
ary 1997 should achieve 2,100 mg/Nm3 (if burning coal with ≤1% sulfur) or 1,200 
mg/Nm3 (if burning coal with >1% sulfur; SEPA and AQSIQ, 1996). For coal-­
fired power plants burning bituminous coal with 0.5% sulfur, SO2 concentration 
in the non-­
desulfurized flue gas will generally exceed 1,000 mg/Nm3. Essentially 
the 1996 effluent emission standards meant that coal-­
fired power plants burning 
coal with >1% sulfur should have SO2 scrubbers while those with ≤1% sulfur 
did not need to. In the standards enacted in 2003, for the great majority of coal 
power plants, their effluent SO2 concentration should be kept below 400 mg/Nm3 
0%
5%
10%
15%
20%
25%
Share of SO2
y
t
i
c
a
p
a
c
 
r
e
b
b
u
r
c
s
Sulfur Content
Figure 5.10  
Distribution of sulfur contents in coal power plants in China (with retrofitted 
SO2 scrubbers)
Source: SEPA and NDRC (2008).


96  Policy making
on 1 January 2010 (SEPA and General Administration of Quality Supervision 
Inspection and Quarantine, 2003). In addition to China’s shutting down old, small 
power-­
generating units, the effluent emission standard itself would ensure that a 
dominant share of China’s coal power capacity in 2010 would have SO2 scrubbers 
installed and operate normally.
The standards were updated in 2011 for being effective on 1 January 2012 (Min­
istry of Environmental Protection and General Administration of Quality Supervi­
sion Inspection and Quarantine, 2011). New plants should then reduce their effluent 
SO2 emissions below 100  mg/Nm3 while the standard for existing plants was 
200 mg/Nm3. In southwestern provinces, including Guangxi, Chongqing, Sichuan 
and Guizhou, where local coal contains much higher sulfur contents, the standards 
could be relaxed to 200 mg/Nm3 and 400 mg/Nm3, respectively. Natural gas–fired 
power plants tend to be much cleaner, with the standard being 35 mg/Nm3.
In 2014, a new policy, “Upgrading and Retrofitting Action Plan for Energy Con­
servation and Pollution Mitigation in the Coal-­
Fired Power Sector,” was enacted 
jointly by National Development and Reform Commission, Ministry of Environ­
mental Protection and National Energy Administration (National Development 
and Reform Commission et al., 2014). It required newly constructed coal-­
fired 
power plants in eastern provinces to achieve the standard for natural gas–fired 
power plants, that is, 35 mg/Nm3 for SO2 emissions. Central provinces should 
approach this standard, while western provinces were encouraged to reach the 
level. This much more stringent standard is referred to in China as the ultra-­
low 
emissions. In 2015, another policy mandates the ultra-­
low standard to be achieved 
in most new and existing coal-­
fired power plants with only occasional exceptions 
(Ministry of Environmental Protection et al., 2015).
China’s Law of Standardization and its implementation regulations provide 
legal teeth (NPC, 1988; State Council, 1990). Effluent emission standards are 
clearly stated as “mandatory standards” (State Council, 1990), while products 
not meeting “mandatory standards” are forbidden to produce, sell and import 
(NPC, 1988). In this sense, coal power plants should stop generating electricity 
if the effluent SO2 emissions exceeded corresponding standards. The electric grid 
should not accept the electricity if it were not legally generated.
In order to achieve SO2 removal rates as required by the stringent ultra-­
low efflu­
ent emission standard, coal-­
fired power plants should generally achieve very high 
SO2 removal rates, being 98.5% if Huolinhe lignite or Datong bituminous coals are 
burned or 96.9% for Shenfu bituminous coal (Table 5.2). The SO2 emission intensity 
of electricity generation should also be substantially reduced to about 0.10 to 0.11 
g/kWh. The sulfur contents in these three types of coal, from 0.50% to 0.99%, fall 
within the normal range. For high-­
sulfur coal, especially in southwestern provinces, 
the required SO2 removal rates are much higher, generally beyond 99%. The deep 
reduction can only be achieved through SO2 scrubbers if coal remains as the fuel.
Before China started the large-­
scale deployment of SO2 scrubbers in the early 
2000s, the world in total had installed about 200 GW (Taylor et al., 2005). The 
United States accumulated around 100 GW in a 25-­
year period between 1975 and 
2000 (Taylor et al., 2005). Germany and Japan together took 30% of the world’s 


Policy making  97
market, and the remaining 20% were in other countries (Taylor et al., 2005). The 
scrubber capacity numbers presented in Figure 5.11 were calculated from a pub­
licly available plant-­
level data set (Ministry of Environmental Protection, 2014). 
The dataset includes information on the name and location of coal power plants, 
the serial number and power capacity of generators, the dates that generators and 
Table 5.2 Effluent SO2 emissions and necessary SO2 removal rates
Huolinhe 
 
Datong 
Shenfu 
lignite
bituminous
bituminous
LHV (MJ/kg)
13.9
21.0
21.4
Contents in coal (%)
Sulfur
0.61%
0.99%
0.50%
Carbon
34.1%
55.7%
57.0%
Hydrogen
2.7%
3.4%
3.4%
Oxygen
10.5%
8.3%
8.0%
Nitrogen
0.7%
0.9%
1.1%
Effluent SO2 emissions 
Concentration (mg/Nm3)
2,315
2,291
1,133
(without removal)
Emissions (g/kWh)
6.79
7.27
3.60
For achieving the 
Required SO2 removal 
98.5%
98.5%
96.9%
35 mg/Nm3 standard
rate (%)
Emissions (g/kWh)
0.10
0.11
0.11
Note: Assuming the sulfur retention rate in ash, 90%; thermal efficiency of electricity generation 
(42%, or 293 g of coal equivalent/kWh). Coal quality data are from Shi and Yu (2005).
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
100
200
300
400
500
600
700
800
2000
2002
2004
2006
2008
2010
2012
Share of coal-fired power capacity 
)
W
G
(
 
y
t
i
c
a
p
a
C
Year
Coal-fired power capacity
SO2 scrubber capacity
Share of coal-fired power capacity with SO2 scrubbers (right)
Figure 5.11  
Coal-fired power and SO2 scrubber capacities in China
Source: Ministry of Environmental Protection (2014); EIA (2019); China Electricity Council (2010, 
2006–2015).


98  Policy making
SO2 scrubbers came online, SO2 scrubber technology type and the name of the 
SO2 scrubber company in charge. The SEPA (Ministry of Environmental Pro­
tection after March 2008) established a standard procedure for registering SO2 
scrubbers (SEPA, 2005, 2006b); for example, an SO2 scrubber had to operate 
continuously for 168 hours to test its performance before registration.
China’s share was negligible with only 5.6 GW of SO2 scrubbers at the end of 
2000, or 2.5% of its coal-­
fired power capacity (Figure 5.11). In the 10th Five-­
Year 
Plan (2001–2005), SO2 scrubber capacity rose to 46.7 GW in 2005. The progress 
was noticeable and the proportion of coal-­
fired power capacity with SO2 scrub­
bers increased to 12.5%. However, because total coal-­
fired power capacity esca­
lated from 218.9 GW in 2000 to 360.6 GW in 2005, essentially China had more 
coal-­
fired power plants without SO2 scrubbers to witness a steady increase of the 
power sector’s SO2 emissions. In the 11th Five-­
Year Plan (2006–2010), coal-­
fired 
power capacity grew at a much faster pace to reach 654.3 GW in 2010, while SO2 
scrubber capacity was lifted even faster to 569.3 GW in 2010. Then only 85.0 
GW of coal-­
fired power plants, or 13.0%, did not have SO2 scrubbers. This rate 
had already been much lower than that in 2000. In the following years, the ratio 
further inched higher to 94.4% in 2013. Essentially, in China, nearly all coal-­
fired 
power plants should have SO2 scrubbers to continue operation.
The 11th Five-­
Year Plan witnessed a large-­
scale campaign to retrofit existing 
coal-­
fired power plants (Figure  5.12), besides shutting down many inefficient 
0%
50%
100%
150%
200%
250%
300%
0
30
60
90
120
150
2000
2002
2004
2006
2008
2010
2012
Ratio
)
W
G
(
 
h
t
w
o
r
g
 
y
t
i
c
a
p
a
c
 
l
a
u
n
n
A
Year
Coal-fired power capacity
SO2 scrubber capacity
Ratio between SO2 scrubber and
coal-fired power capacities (right)
Figure 5.12  
The annual growth of coal-fired power and SO2 scrubber capacities in China
Source: Ministry of Environmental Protection (2014); EIA (2019); China Electricity Council (2006–
2015, 2010).


Policy making  99
small units (Xu et al., 2013). The ratios between the annually increased capacities 
of SO2 scrubbers and coal-­
fired power plants were consistently higher than 100% 
in every year of the 11th Five-­
Year Plan. At the retrofitting peak in 2008, SO2 
scrubber capacity grew by 127.4 GW while coal-­
fired power capacity increased 
only by 43.1 GW. After the 12th Five-­
Year Plan, a great majority of SO2 scrubbers 
were either built together with coal-­
fired power plants or further retrofitted for 
meeting more stringent effluent emission standards.
Most SO2 scrubbers fall into three scale categories: 300 MW, 600 MW and 
1000 MW (Figure  5.13). They correspond to several predominant, standard­
ized unit scales that China’s coal-­
fired power units have. Among the 754.9 GW 
of coal-­
fired power units with SO2 scrubbers in 2013, 62.1 GW, 215.1 GW and 
263.9 GW were within the 1,000~1,050-­
MW, 600~650-­
MW and 300~350-­
MW 
ranges, respectively. Two smaller scales have seen their importance fading after 
China focused more on larger and more efficient units. Respectively, 42.3 GW and 
32.7 GW fell within the 200~220-­
MW and 135~150-­
MW ranges. In total, these 
five standardized unit sizes accounted for 618.6 GW or 81.9% of all SO2 scrub­
bers. These size and technology standardization provided one crucial advantage 
in designing and rapidly deploying SO2 scrubbers.
The geographic distribution of SO2 scrubbers reflects that of coal-­
fired power 
plants. Provinces in East China, North China and South China had 249.2 GW, 
213.0 GW and 104.1 GW (or 33.0%, 28.2% and 13.8%) of SO2 scrubbers, 
0
20
40
60
80
100
120
140
Before
2001
2003
2005
2007
2009
2011
2013
)
W
G
(
 
y
t
i
c
a
p
a
C
Year
>=1000 MW
600 MW ~ 999 MW
300 MW ~ 599 MW
200 MW ~ 299 MW
100 MW ~199 MW
< 100 MW
Figure 5.13  
Annually increased SO2 scrubber capacity and unit sizes
Source: Ministry of Environmental Protection (2014).


100  Policy making
respectively, in 2013 (Figure  5.14). Northeast, Southwest and Northwest had 
188.5 GW in total, or 25.0%. Their vast geographic territories indicate that these 
coal-­
fired power plants are scattered at much greater distances from each other 
to potentially enhance difficulties for environmental compliance monitoring and 
enforcement.
Different SO2 scrubber technologies correspond to a wide range of possible 
SO2 removal rates. China had 1589 units of SO2 scrubbers at or above 200 
MW in 2013. The limestone-­
gypsum wet type is the most applied technol­
ogy especially for large coal-­
fired power units, accounting for 93.6% of units 
>=1000 MW, 96.1% of those between 600 MW and 999 MW, 87.6% of those 
between 300 MW and 599 MW and 81.0% of those between 200 MW and 
299 MW (Figure 5.15). The share dropped significantly for units smaller than 
200 MW, being only 30.0% (Figure 5.15). Due to the same consideration of 
economy of scale, seawater type also heavily tilted toward large units (Fig­
ure 5.15). Only 94.5 GW of SO2 scrubbers in 2013 were individually smaller 
than 200 MW (12.5% of all SO2 scrubbers), but they had 2,878 units (64.4% 
of all; Figure 5.15).
0%
10%
20%
30%
40%
50%
60%
0
20
40
60
80
100
120
140
Before
2001
2003
2005
2007
2009
2011
2013
Proportion as retrofit
)
W
G
(
 
y
t
i
c
a
p
a
C
Year
East
South
Southwest
Northwest
North
Northeast
Porportion as retrofit (right)
Figure 5.14  
The annual growth of SO2 scrubber capacity by regions (as categorized by the 
six Regional Supervision Bureaus of the Ministry of Ecology and Environ­
ment; SO2 scrubbers are called “retrofits” when the online dates of SO2 scrub­
bers and coal power units are over one year)
Source: Ministry of Environmental Protection (2014).


Policy making  101
Reference
Ang, B. W. 2005. The LMDI approach to decomposition analysis: A  practical guide. 
Energy Policy, 33, 867–871.
Barron, W. F. & Ng, G. T. L. 1996. An assessment methodology for environmental policy 
instruments: An illustrative application to solid wastes in Hong Kong. Journal of Envi­
ronmental Management, 48, 283–298.
Benkovic, S. R. & Kruger, J. 2001. US sulfur dioxide emissions trading program: Results 
and further applications. Water Air and Soil Pollution, 130, 241–246.
Blackman, A. 2010. Alternative pollution control policies in developing countries. Review 
of Environmental Economics and Policy, 1–20.
BP. 2019. Statistical review of world energy [Online]. Available: www.bp.com/en/global/
corporate/energy-­
economics/statisticalreview-­
of-­
world-­
energy.html.
Cheng, J., Zhou, J. H., Liu, J. Z., Cao, X. Y., Zhou, Z. J., Huang, Z. Y., Zhao, X. & Cen, 
K. 2004. Physicochemical properties of Chinese pulverized coal ash in relation to sulfur 
retention. Powder Technology, 146, 169–175.
China Electricity Council. 2006–2015. Annual report of national power generation. Bei­
jing, China: China Electricity Council.
China Electricity Council. 2010. The basic information of electricity generation units and 
electric grids. Beijing, China: China Electricity Council.
Darwin, C. 1859. On the origin of species by means of natural selection, or the preserva­
tion of favoured races in the struggle for life. London: J. Murray.
0
500
1,000
1,500
2,000
2,500
3,000
0
50
100
150
200
250
300
>=1000
MW
600 MW ~
999 MW
300 MW ~
599 MW
200 MW ~
299 MW
100 MW
~199 MW
< 100 MW
Number of units
)
W
G
(
 
y
t
i
c
a
p
a
C
Unit Size
Limestone-Gypsum Wet Type
Seawater Type
Others
Number of units
Figure 5.15  
SO2 scrubbing technologies by unit sizes
Source: Ministry of Environmental Protection (2014).


102  Policy making
Dong, Z., Ge, C., Gao, S. & Wang, J. 2011. Assessment on the Chinese pollution charges 
system and proposals to the future reform. In: Shen, M., Ge, C., Dong, Z. & Zhang, B. 
(eds.) Progress on environmental economics. Beijing, China: China Environmental Sci­
ence Press.
EIA. 2019. International energy statistics. Washington, DC: U.S. Energy Information 
Administration.
EPA. 1998. AP 42, chapter 1: External combustion sources, fifth ed. Washington, DC: 
EPA.
Feng, T. W., Sun, L. Y. & Zhang, Y. 2009. The relationship between energy consumption 
structure, economic structure and energy intensity in China. Energy Policy, 37, 5475–5483.
Ge, C., Ren, Y., Gao, S., Sun, G. & Long, F. 2011. Wastewater discharge tax design: From 
pollutant discharge levy to environmental tax. In: Shen, M., Ge, C., Dong, Z. & Zhang, 
B. (eds.) Progress on environmental economics. Beijing, China: China Environmental 
Science Press.
Goulder, L. H. & Parry, I. W. H. 2008. Instrument choice in environmental policy. Review 
of Environmental Economics and Policy, 2, 152–174.
IEA. 2017. World energy outlook. Paris, France: IEA.
IEA. 2019. Global EV outlook 2019. Paris, France: IEA.
IMF. 2019. World economic outlook database October 2019. Washington, DC: IMF.
MEP & AQSIQ. 2011. Emission standard of air pollutants for thermal power plants. Bei­
jing, China: MEP & AQSIQ.
Ministry of Environmental Protection. 2014. The list of China’s SO2 scrubbers in coal-­
fired 
power plants. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection & General Administration of Quality Supervision 
Inspection and Quarantine. 2011. Emission standard of air pollutants for thermal power 
plants. GB 13223-­
2011. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection, National Development and Reform Commission & 
National Energy Administration. 2015. Comprehensive implementation of the upgrading 
and retrofitting action plan for energy conservation and pollution mitigation in the coal-­
fired power sector. Beijing, China: Ministry of Environmental Protection.
National Bureau of Statistics. 1997–2008. China energy statistical yearbook. Beijing, 
China: China Statistics Press.
National Development and Reform Commission, Ministry of Environmental Protection & 
National Energy Administration. 2014. Upgrading and retrofitting action plan for energy 
conservation and pollution mitigation in the coal-­
fired power sector (2014–2020). Bei­
jing, China: Ministry of Environmental Protection.
National People’s Congress. 2001. The outline of national 10th five-­
year plan on economic 
and social developments. Beijing, China: The 4th Conference of the 9th National Peo­
ple’s Congress.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. Environmental protection tax law of the people’s repub­
lic of China. Beijing, China: The 4th Conference of the 10th National People’s Congress.
National Statistics Bureau & Ministry of Ecology and Environment. 2019. China statisti­
cal yearbook on environment 2018. Beijing, China: China Statistics Press.


Policy making  103
NDRC. 2001. Key special 10th five-­
year plan on energy development. Beijing, China: 
NDRC.
NDRC. 2004. Technical code for designing flue gas desulfurization plants of fossil fuel 
power plants. DL/T5196-­
2004. Beijing, China: NDRC.
NDRC. 2007. 11th five-­
year plan on energy development. Beijing, China: NDRC.
NPC. 1988. The standardization law of the people’s republic of China. Beijing, China: 7th 
National People’s Congress of the People’s Republic of China.
NPC. 2002. Environmental impact assessment law of the people’s republic of China. Bei­
jing, China: 9th National People’s Congress of the People’s Republic of China.
Ohara, T., Akimoto, H., Kurokawa, J., Horii, N., Yamaji, K., Yan, X. & Hayasaka, T. 2007. 
An Asian emission inventory of anthropogenic emission sources for the period 1980–
2020. Atmospheric Chemistry and Physics, 7, 4419–4444.
Portney, P. R. & Stavins, R. N. 2000. Public policies for environmental protection. Wash­
ington, DC: Resources for the Future.
SEPA. 2001. National 10th five-­
year plan on environmental protection. Beijing, China: 
State Environmental Protection Administration.
SEPA. 2005. Flue gas limestone/lime – gypsum desulfurization project technical specification 
of thermal power plant. Beijing, China: State Environmental Protection Administration.
SEPA. 2006a. Guidelines on calculating SO2 emission quotas. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2006b. Technical guidelines for environmental protection in power plant capital 
construction project for check and accept of completed project. Beijing, China: State 
Environmental Protection Administration.
SEPA. 2007. Detailed methods to verify major pollutants emission reduction in the 11th five-­
year period (on trial). Beijing, China: State Environmental Protection Administration.
SEPA & AQSIQ. 1996. Emission standard of air pollutants for thermal power plants. GB 
13223-­
1996. Beijing, China: SEPA, AQSIQ.
SEPA & NDRC. 2008. National 11th five-­
year plan on acid rain and SO2 pollution control. 
Beijing, China: State Environmental Protection Administration, NDRC.
Sheng, C. D., Xu, M. H., Zhang, J. & Xu, Y. Q. 2000. Comparison of sulphur retention by 
coal ash in different types of combustors. Fuel Processing Technology, 64, 1–11.
Shi, D. & Yu, C. 2005. Optimization design, technical control and coal quality evaluation 
of contemporary coal pre-­
processing. Beijing, China: China Contemporary Audio and 
Video Press.
Singer, J. G. 1981. Combustion, fossil power systems: A reference book on fuel burning and 
steam generation. Windsor, CT: Combustion Engineering.
Smith, S. J., Pitcher, H. & Wigley, T. M. L. 2001. Global and regional anthropogenic sulfur 
dioxide emissions. Global and Planetary Change, 29, 99–119.
State Council. 1990. Regulations for the implementation of the law of standardization. 
Beijing, China: State Council.
State Council. 1998. Administrative rule on environmental protection of construction pro­
jects. Decree No. 253. Beijing, China: State Council.
State Council. 2007a. Notice on distributing implementation plans and methods of statis­
tics, monitoring and assessment on energy conservation and pollutant emission reduc­
tion. Beijing, China: State Council.
State Council. 2007b. Notice on distributing the 11th five-­
year plan on environmental pro­
tection. Beijing, China: State Council.
State Environmental Protection Administration (SEPA) & General Administration of Qual­
ity Supervision Inspection and Quarantine. 2003. Emission standard of air pollutants for 


104  Policy making
thermal power plants. GB 13223-­
2003. Beijing, China: State Environmental Protection 
Administration.
Taylor, M. R., Rubin, E. S. & Hounshell, D. A. 2005. Control of SO2 emissions from power 
plants: A case of induced technological innovation in the U.S. Technological Forecast­
ing and Social Change, 72, 697–718.
Tietenberg, T. H. 1990. Economic instruments for environmental regulation. Oxford 
Review of Economic Policy, 6, 17–33.
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. 
Congress.
Xu, Y. 2011. China’s functioning market for sulfur dioxide scrubbing technologies. Envi­
ronmental Science and Technology, 45, 9161–9167.
Xu, Y., Williams, R. H. & Socolow, R. H. 2009. China’s rapid deployment of SO2 scrub­
bers. Energy & Environmental Science, 459–465.
Xu, Y., Yang, C. J. & Xuan, X. W. 2013. Engineering and optimization approaches to 
enhance the thermal efficiency of coal electricity generation in China. Energy Policy, 
60, 356–363.
Yan, G., Yang, J., Wang, J., Chen, X. & Xu, Y. 2009. Carry out SO2 emission trading 
program to set up long-­
term mechanism for emission reduction. In: Wang, J., Lu, J., 
Jintian, Y. & Li, Y. (eds.) Environmental policy research series. Beijing, China: China 
Environmental Science Press.
Zhang, B., Fei, H., He, P., Xu, Y., Dong, Z. & Young, O. 2016. The indecisive role of 
the market in China’s SO2 and COD emissions trading. Environmental Politics, 25, 
875–898.
Zhang, B., Zhang, H., Liu, B. B. & Bi, J. 2013. Policy interactions and underperform­
ing emission trading markets in China. Environmental Science  & Technology, 47, 
7077–7084.
Zhao, Y., Wang, S., Duan, L., Lei, Y., Cao, P. & Hao, J. 2008. Primary air pollutant emis­
sions of coal-­
fired power plants in China: Current status and future prediction. Atmos­
pheric Environment, 42, 8442–8452.


1  
Goal-­
centered policy implementation
In a country with effective rule of law, law enactment and policy making are the 
most important step for environmental protection, while implementation is more 
or less expected, although some bumps may still exist. China does not have sound 
rule of law, and thus, its policy implementation could be even more important than 
policy making. In the United States, after the Acid Rain Program in the Clean Air 
Act Amendments (1990) was enacted, law enforcement was largely the respon­
sibility of the administrative branch. The rule of law obliges the administration 
to enforce the law. However, in China, no such tradition has been established to 
ensure that laws and policies will be genuinely implemented. Furthermore, Chi­
na’s policy implementation is heavily decentralized to local governments (Chap­
ter 3), while the U.S. federal government has a relatively much stronger capacity 
for implementing their own policies. A key difference between China and the 
United States is that China should first mobilize its decentralized policy imple­
menters before witnessing significant efforts and sulfur dioxide (SO2) mitigation. 
China relies on the goal system to mobilize ministries at the central government 
and, more important, local governments for policy making and implementation, 
as discussed in Chapter 4.
Environmental compliance in China was indeed weak but has been improving 
steadily. China has made much progress in the past 15 years to reverse the ear­
lier poor implementation of environmental policies (Jin et al., 2016). Coal-­
fired 
power plants in China have nearly universally installed SO2 scrubbers, already 
94.4% as of 2013 (Figure 5.11). Although most SO2 scrubbers in China today 
do operate properly to greatly contribute to the deep reduction of SO2 emissions 
(Figure 1.8), evidence of their misreporting and cheating was widely present to 
indicate serious noncompliance problems. A study showed that many factories in 
China were primarily concerned about minimizing operation costs and only oper­
ated their pollutant removal facilities when an inspection was imminent (OECD, 
2006). Official data reported that SO2 emissions from the power sector in 2007 
were 11.5 million tons (Ministry of Environmental Protection, 2006–2009), but 
an independent study estimated that 16.4 million tons were emitted in that year 
(Lu et al., 2010). In addition, official data announced that in 2007, 73.2% of SO2 
6	

Policy implementation1


106  Policy implementation
was removed from coal-­
fired power plants that had SO2 scrubbers (Ministry of 
Environmental Protection, 2009b). In Jiangsu Province, which had a relatively 
good track record on environmental protection, however, the rate was found to 
be only about one third in the first few months of 2007 (SERC, 2009). Especially 
before June  2007, cheating was widespread (Figure  6.1). Although almost all 
coal-­
fired power plants generally reported that their SO2 scrubbers were operating 
normally, later confirmed data found the operating time to be much shorter (Fig­
ure 6.1). For those in operation, their SO2 removal efficiencies were often much 
lower than required (SERC, 2009). However, after July 2007, a great majority 
of their SO2 scrubbers were operating for more than 90% of the time and were 
achieving SO2 removal efficiencies of over 90% (Jiangsu Department of Environ­
mental Protection, 2007–2009). Data from the Ministry of Environmental Protec­
tion reported that SO2 scrubbers had already been removing 78.7% of SO2 from 
associated coal-­
fired power plants in 2008 (Ministry of Environmental Protection, 
2009b), indicating that they were largely operating as they were supposed to do. 
This chapter evaluates such transition and examines how the compliance deci­
sions were reversed.
After SO2 scrubbers are installed, the managers of coal-­
fired power plants decide 
whether to operate them or not. The willingness to install SO2 scrubbers does not 
0%
20%
40%
60%
80%
100%
1/2006
7/2006
1/2007
7/2007
1/2008
7/2008
e
t
a
R
 
n
o
i
t
a
r
e
p
O
Month-Year
Later confirmed rates
Self-reported rates
Figure 6.1  
The operation of SO2 scrubbers in Jiangsu Province, including self-reported 
operation rates and later confirmed operation rates
Source: The Economic & Trade Commission of Jiangsu Province, 2009; Jiangsu Department of Envi­
ronmental Protection, 2007–2009; State Electricity Regulation Commission (Nanjing office), 2009.


Policy implementation  107
necessarily mean that the incentives are strong enough for their proper operation. 
In addition, China’s SO2 scrubbers vary greatly in sizes, technology types, sul­
fur contents, costs of reagents and local environmental governance effectiveness. 
A significant variance should exist in their operation especially across provinces.
Studies have shown that three conditions are favorable to ensure compliance 
with environmental legislation: low compliance costs, high penalties for noncom­
pliance and a high probability of catching noncompliance (Cohen, 1999; Helland, 
1998; Becker, 1968). The latter two conditions are complementary to each other. 
In 2000, Blackman and Harrington judged that in China, “both the probability 
of getting caught for underreporting and the penalty for doing so are quite low” 
(Blackman and Harrington, 2000). A 2009 IEA (International Energy Agency) 
report claimed that the operation of China’s SO2 scrubbers was problematic due 
to high operation costs and ineffective environmental regulation (IEA, 2009). As 
the crucial factor to decide the probability of catching noncompliance, the impor­
tance of an effective compliance monitoring system has been widely recognized 
for implementing environmental policies and achieving their intended objectives 
(Lu et al., 2006; Raufer and Li, 2009). Monitoring and site inspection are essential 
for catching offenders but are subject to the constraints of high costs and limited 
budgets (Arguedas, 2008). The problem is especially serious in developing coun­
tries (McAllister et al., 2010; Blackman and Harrington, 2000). To enforce the 
SO2 allowance trading scheme in the United States, the strategy was to install 
continuous emissions monitoring systems (CEMSs) with “periodic quality con­
trol tests of monitoring devices to maintain the accuracy of emissions data” (The 
U.S. Congress, 1990; Stranlund and Chavez, 2000). Large polluters can attract 
more attention. A study of the U.S. steel industry found that large polluting plants 
attracted more scrutiny than their smaller counterparts, regardless of how good 
their compliance record was (Gray and Deily, 1996).
Similar to policy making, policy implementation in China is also goal-­
centered, 
under which actions from the central and local governments focus more on whether 
they can contribute to goal attainment and less on whether policies are genuinely 
implemented. Heavily decentralized policy implementation facilitates their selec­
tive and goal-­
centered enforcement efforts. Such selective policy implementa­
tion also indicates that many rules in China are not followed or respected even 
by governments, which results in the weak rule of law. Given the political, eco­
nomic, social and technological feasibility of implementation, as well as probable 
capacity constraints, those policies that can lead to more pollution mitigation have 
higher probabilities of being prioritized in implementation. This will trigger the 
policy evolution through implementation selection, as discussed in Chapter 5. For 
implementing a given policy in China’s context of originally low environmental 
compliance rates, those factors that contribute to their enhancement are strength­
ened selectively and sequentially, depending on how progress can be made more 
effectively and efficiently with corresponding efforts. In a certain period, compli­
ance costs are determined by technological statuses and market conditions, which 
are largely not decided directly by the environmental administration. In the longer 
term, technologies may evolve and costs may go down, as examined in detail in 


108  Policy implementation
Chapter 7. The other two factors, penalties for noncompliance and environmental 
compliance monitoring, are primarily examined in this chapter.
This goal-­
centered policy implementation echoes the comparative advantage the­
ory that was originated by David Ricardo to analyze the development of international 
trade (Ricardo, 1817). In a two-­
country, two-­
product model, even though a country 
may have lower productivity or absolute disadvantage in producing both products, it 
still could specialize in and export one product based on comparative advantage and 
import only the other. Heckscher and Ohlin further developed the model to attribute 
the origin of comparative advantage in a country’s factor endowment (Ohlin, 1967). 
It later became the foundation of a development theory that argued that a country 
should base its development on its comparative advantage (Chenery, 1961). Lin 
et al. employed this theory to explain the rapid economic growth of China and other 
countries (Lin et al., 2003). Before the economic reform in 1978, China adopted a 
leap-­
forward strategy to develop capital-­
intensive heavy industries against its com­
parative advantage, and this resulted in slow and unsustainable economic growth, 
whereas after the reform, the comparative advantage of labor was better utilized to 
achieve rapid economic growth and upgrading (Lin et al., 2003).
From a status of prevalent noncompliance, goal-­
centered policy implementa­
tion suggests making progress according to the contingent comparative advantage 
of alternative paths for achieving goals. When policy implementers decide which 
path could better serve the SO2 mitigation goals, the chosen path should follow 
the direction whereby the effort’s “productivity” is comparatively higher. In other 
words, easier measures are taken first before moving to more difficult measures, 
although many problems exist in the process.
This chapter examines the progress of three key measures. Penalties for non­
compliance were first increased. A 2007 policy provided subsidies to coal-­
fired 
power plants for normally operating their SO2 scrubbers, but nonoperation would 
incur a penalty of five times. Managers of those coal-­
fired power plants, mostly 
state-­
owned, would lose their jobs if cheating were caught. These penalties were 
relatively easier to be made available, while the more difficult environmental 
compliance monitoring was strengthened in following two steps to enhance the 
probability of catching noncompliance. First, more resources were made available 
to support the conventional environmental compliance monitoring system that 
features monitoring, reporting and verification (MRV). CEMSs also played a key 
role to signal potential noncompliance. The strategy worked well for coal-­
fired 
power plants that tend to be large, making frequent inspections little constrained 
by the shortage of inspectors. Collusion was also made more difficult for improv­
ing data quality. Furthermore, new technologies for environmental compliance 
monitoring are rapidly emerging and evolving, including sensors, satellites and 
social media. They tend to be much cheaper in monitoring one polluter but less 
accurate for legally confirming compliance statuses and issuing penalties, while 
conventional technologies are much more expensive but, if working smoothly, 
can meet the legal accuracy requirements. In the 2010s and, especially, since 
2015, the Chinese government has been actively developing and integrating these 
big data technologies into governance. In environmental protection with millions 


Policy implementation 109
of polluting sources scattered across China’s vast geographic landscape, environ-
mental compliance monitoring is one primary field to apply these new technolo-
gies for achieving higher compliance rates without demanding more resources.
2  
Compliance on the operation of SO2 scrubbers
2.1  SO2 scrubber technologies
Compliance costs of SO2 scrubbers are mainly for their operation as well as main-
tenance. A comprehension of SO2 scrubber technologies is accordingly essential 
to understand how coal- 
fired power plants may cheat on compliance and how the 
government could catch such noncompliance.
SO2 scrubbers (or flue gas desulfurization) have various technology types. Wet 
scrubbers are the most applied technology with SO2 removal efficiencies nor-
mally over 90% (Figure 5.15). This section introduces major features associated 
with wet scrubbers and briefly compares these with dry scrubbers. After the flue 
gas comes out of a dust- 
removal facility (generally electrostatic precipitator, or 
ESP, in China’s coal- 
fired power plants), it will be directed to an SO2 scrubber 
system. The first step is often to pass the flue gas through fans or boosters, which 
facilitate the flow and adjust the velocity of the flue gas to be in a desirable range 
for the best performance of the SO2 scrubber. Then the flue gas enters an absorber 
tower, where actual SO2 removal happens. Generally speaking, coal-fired 
 
power 
generation units of 300 MW or over should have their own absorber towers and 
two units of 200 MW or less could share one (NDRC, 2004). The flue gas enters 
the absorber tower at the lower- 
middle part and moves upward. Limestone slurry 
mixed with products of chemical reactions fills the lower part of the absorber 
tower and is lifted by several circulation pumps to the upper part. Special nozzles 
are used to spray the slurry for the maximization of SO2 removal efficiency. The 
falling slurry droplets contact the flue gas physically and remove approximately 
90% to 95% of SO2 through chemical reactions. Simply put, the main and over-
all reaction is CaCO3
2
+
+
SO
H O
2
3
→
+
CaSO
CO2
2
+ H O. Air is blown into the 
slurry pool at the lower part of the absorber tower to force the oxidization of SO2-
3   
1
and make gypsum (CaSO 2
4
2
H O): CaSO3
2
+
+
O
2H O
2
2
4
→
↓
CaSO 2H2O . 
Before the flue gas exits from the top of the absorber tower, it passes through 
equipment that removes mist. Because the processing removes much heat from 
the flue gas and the efficient outflow from a chimney (often over 200 m high for 
coal- 
fired power plants) requires the flue gas to be above a certain temperature, a 
gas–gas heat exchanger may be included to heat the outlet flow gas with the inlet 
flue gas to raise its temperature.
Two important side systems are respectively for the preparation of limestone 
slurry and the production of gypsum. Limestone is crushed and mixed with water 
to make limestone slurry. Fresh limestone slurry enters the absorber tower often 
through circulation pumps. The bottom slurry with gypsum is pumped out and 
filtered to separate gypsum. The wastewater is sent to a treatment system.


110  Policy implementation
SO2 removal efficiency can be controlled by adjusting various factors, includ­
ing the contact time length between the flue gas and the limestone slurry droplets, 
calcium-­
to-­
sulfur ratio (or Ca/S ratio) and liquid-­
to-­
gas ratio (or L/G ratio). The 
contact time depends on the velocity of the flue gas and the path length before its 
leaving the point where limestone slurry is injected. The height of the absorber 
tower is an influential factor determining the path length. Another factor is asso­
ciated with the different injection heights of circulation pumps. Higher injection 
points indicate a longer path for contact and reaction. When the electricity genera­
tion unit is not in full load with less flue gas, not all circulation pumps will have to 
be operated. Then the choice of different circulation pumps could make some dif­
ference in the SO2 removal efficiency. However, higher absorber tower and longer 
contact path correspond to higher electricity consumption to lift limestone slurry.
In the L/G ratio, the liquid refers to the volume of limestone slurry dropping 
from the upper part of the absorber tower, or circulated liquids. The gas is the 
volume of flue gas entering the absorber tower. Higher L/G ratio leads to higher 
SO2 removal efficiency because the chance is higher for an SO2 molecule to be 
absorbed. It is controlled through circulation pumps: if the flue gas volume does 
not change, turning on more pumps indicates a higher L/G ratio. Since the number 
and power of circulation pumps are fixed after an SO2 scrubber comes online, the 
liquid volume has an upper limit, which restrains the maximum contribution of 
enhancing L/G ratio to increase SO2 removal efficiency.
Ca/S ratio is the molar ratio between calcium carbonate (CaCO3) and sulfur 
oxides (SOx, dominantly SO2). In a perfect situation, as predicted in the chemical 
reaction introduced earlier, the ideal Ca/S ratio is 1 to remove all SO2 and use up 
all limestone. But in the actual situation, not all limestone will be consumed and 
not all SO2 will be removed. SO2 wet scrubbers can achieve high efficiencies in 
both aspects. As a result, the actual Ca/S ratio is only a little higher than 1, usually 
around 1.03 for China’s wet scrubbers (Wu and Qian, 2007). Because the inlet 
quantity of SO2 changes with the volume of flue gas and the SO2 concentration, 
even though the Ca/S ratio is maintained stable, the rate of adding limestone to the 
system will still change. On the other hand, the workload of removing SO2 could 
become too heavy when the actual sulfur content exceeds the designed level by 
a significant margin. In this situation, when all circulation pumps are turned on 
and the L/G ratio has reached its maximum, the only major method to maintain 
a required high SO2 removal efficiency is to enhance the Ca/S ratio. However, a 
much higher Ca/S ratio than the designed level will not only add costs but also 
will more likely clog the system and barricade its normal function. The adjustment 
of the Ca/S ratio is through controlling the pH value of the limestone slurry in the 
absorber tower. In daily operation, the pH value should be maintained within a 
range. A pH value above the normal range indicates excessive limestone and that 
the injection rate of fresh limestone slurry should be reduced.
SO2 wet scrubbers consume about 1% of the electricity generated from the cor­
responding power generation units. The rate could be as high as 3.5% when high-­
sulfur coal is burned with a heavy workload of SO2 removal. China’s coal-­
fired 
power plants consumed, on average, 6.79% of the electricity they generated in 


Policy implementation  111
2008 (SERC et al., 2009), in which SO2 scrubbers accounted for a notable share. 
Significant electricity-­
consuming components of SO2 scrubbers include the fans, 
circulation pumps and limestone slurry preparation system.
Two major economies of scale are associated with SO2 scrubbers in construc­
tion. First, the size of an absorber tower is largely determined by the volume of 
flue gas or the scale of the corresponding power generation unit. A larger volume 
of flue gas or a larger scale in megawatts leads to lower average costs for each 
unit of flue gas treated or each megawatt. Because absorber towers are responsi­
ble for a large part of the capital costs, this economy of scale could significantly 
reduce the unit capital costs. Since most nonpower SO2 emission sources consume 
much less coal and do not generate large enough volume of flue gas to provide 
the economy of scale, the unit costs of SO2 scrubbers are often more expensive.
Second, higher SO2 concentration in the inlet flue gas, or higher sulfur input 
rate, raises capital costs for each unit of flue gas treated because they require 
larger systems of limestone preparation and gypsum handling, as well as probably 
higher absorber tower and more circulation pumps. Economy of scale can also 
be realized for these systems to treat each unit of SO2. SO2 concentration in the 
inlet flue gas is mostly determined by two factors: sulfur and thermal contents of 
coal. The link with sulfur contents is quite straightforward: if different types of 
coal only differ in sulfur contents, higher sulfur contents indicate more SO2 in a 
roughly equal amount of flue gas. With the same thermal efficiency, the volume of 
flue gas mainly depends on the thermal input. Then lower thermal contents of coal 
mean that more coal has to be burned for the required thermal input, and accord­
ingly, more SO2 will be generated. Accordingly, sulfur content per unit of energy 
is a better indicator of SO2 concentration in the inlet flue gas of SO2 scrubbers.
The product of SO2 scrubbers is gypsum. Depending partly on the quality, it 
can either be sold in the market or go to landfill. A significant market for gypsum 
is in building materials.
The operation and maintenance (O&M) of SO2 scrubbers are associated with 
costs in materials (including mainly limestone, electricity and water), labor and 
maintenance. The sale of gypsum could earn some revenue but often only at an 
insignificant portion. If the quality of SO2 scrubbers remains about the same, 
maintenance costs are positively related to the capital investment of SO2 scrub­
bers. As a result, larger scales of SO2 scrubbers are linked with lower maintenance 
costs on the bases of megawatt-­
hour or ton SO2 removed. Similarly, economy 
of scale is also relevant to labor costs, but the impact on overall O&M costs is 
constrained by the insignificant share of labor costs; for example, my field trip 
to China’s coal-­
fired power plants found that roughly 15 workers were required 
to run the SO2 scrubber and ESP for a 300-­
MW plant in 2008. Their total annual 
costs could be about 1 million RMB. The average O&M costs of China’s SO2 
scrubbers were about 15 RMB/MWh, indicating that the total O&M costs would 
be approximately 23 million RMB if the capacity factor was about 5,000 hours/
year. Then the share of labor costs was less than 5%.
Materials comprise most of operation costs. In normal operation, the Ca/S ratio 
remains fairly stable, and thus, the limestone consumption is about linearly related 


112  Policy implementation
to sulfur input. Electricity consumption for running SO2 scrubbers can be roughly 
divided into three major parts: in fans that are mainly associated with the flue 
gas volume, in the handling of limestone and gypsum that is affected by sulfur 
input and in circulation pumps connected with both. Most water consumption is 
in the form of evaporation to the flue gas in the absorber tower, and the water is 
released to the atmosphere together with the cleaned flue gas. As a result, water 
consumption is mainly correlated with the volume of flue gas and is also affected 
by economy of scale.
Because of its wide availability and low costs, limestone is the dominant 
absorbing reagent in wet scrubbers. However, other alkaline reagents are some­
times applied, such as seawater and alkaline wastewater. Furthermore, SO2 scrub­
bers can be dry. Lime (CaO) is often used as the absorbing reagent. Because of its 
much lower utilization rate, the Ca/S ratio has to be much higher (e.g., 1.3~1.5). 
Generally, the SO2 removal efficiency is in the range of about 70% to 80%, lower 
than that of wet scrubbers. The capital costs of dry scrubbers are lower, but the 
operation and maintenance costs are higher (EPA, 2003).
2.2  
Noncompliance behaviors
The managers of coal-­
fired power plants have strong incentives to avoid the costs 
of O&M. Data from Jiangsu Province showed that from 2006 to June 2007, the 
self-­
reported operation rates (the percentage of time that an SO2 scrubber is in 
operation alongside the corresponding power generation unit) from coal-­
fired 
power plants were significantly higher than the values that were later confirmed, 
likely through other relevant data such as limestone consumption, gypsum pro­
duction and electricity consumption (respectively, more than 90% and about 60%; 
Figure 6.1). The discrepancy reflects the likely magnitude of misreporting. This 
section discusses several prominent problems that emerged from the author’s 
interviews and the literature. These problems prevented the proper operation of 
SO2 scrubbers and caused very significant uncertainty in estimating SO2 emis­
sions from coal-­
fired power plants.
Typical noncompliance behaviors could include the following: first, SO2 emis­
sions may be underreported and the quality of SO2 scrubbers could be poor. Coal-­
fired power plants underreported SO2 emissions to pay a lower effluent discharge 
fee and to be seen as complying with regulations. In 2007, 98% of coal consumed 
in China’s power plants was raw coal (National Bureau of Statistics, 2008). Coal-­
fired power plants were allowed to pick out coal stones from received raw coal 
to calculate actual coal consumption. In interviews, the author found that coal 
stones were sometimes overreported. This factor could have led to a 1% to 2% 
underestimation of SO2 emissions. Furthermore, China’s coal-­
fired power plants 
usually had to use different coals with sulfur contents that could vary signifi­
cantly. The instability of coal supply was confirmed by Steinfeld et al. (2009). It 
made the underreporting of sulfur contents harder to detect. Interviews in China’s 
SO2 scrubber companies found that many early scrubbers (e.g., before 2005) had 
serious quality problems. In order to reach designed SO2 removal efficiencies, 


Policy implementation  113
besides the replacement of malfunctioning equipment, a few SO2 scrubbers even 
had to have their very expensive absorber towers retrofitted. The main reason 
for the faults in the SO2 scrubbers was that they were designed on the basis of 
underreported sulfur contents. In China’s first public and high-­
profile statement 
to penalize the abnormal operation of SO2 scrubbers, instability and bad quality 
were particularly pointed out, and three power plants were found to use coal with 
much higher sulfur than the designed levels (Ministry of Environmental Protec­
tion, 2008). At the design stage of SO2 scrubbers, if the managers of coal-­
fired 
power plants had been underreporting sulfur contents in the past, they would con­
tinue to do so to conceal their guilt. Some managers did not plan to operate their 
SO2 scrubbers initially and were not concerned about their quality. They installed 
the SO2 scrubbers purely in order to comply with the government’s requirements 
and to qualify for a subsidy for generating desulfurized electricity. The managers 
wanted to minimize capital costs through underreporting sulfur contents. When 
inspections were known in advance, reaching the required SO2 removal efficien­
cies was not a problem because coal-­
fired power plants often kept some low-­
sulfur coal in reserve on-­
site.
Second, illegal bypass ducts may be used to avoid flue gas treatment. Many 
SO2 scrubbers had bypass ducts to allow the flue gas to exit without going through 
the SO2 scrubber systems. The purpose was to enable electricity generation when 
SO2 scrubbers had minor problems and need to be shut down temporarily. In a 
2007 policy, coal-­
fired power plants were not penalized provided that their SO2 
scrubbers were properly operating for at least 90% of the time (NDRC and SEPA, 
2007b). However, bypass ducts also provided opportunities to avoid the operation 
of SO2 scrubbers when they functioned normally. Six coal-­
fired power plants were 
penalized for illegally using bypass ducts and leaving some flue gas untreated in 
2007 and 2008 (Ministry of Environmental Protection, 2008, 2009c).
Third, data from CEMSs may also be inaccurate and manipulated. CEMSs 
could greatly enhance environmental monitoring capacity. China had 60 SO2 
scrubbers at the end of 2004 (Ministry of Environmental Protection, 2010a), while 
a general survey in 2004 found that about 400 CEMSs had been installed in 180 
coal-­
fired power plants (Pan et al., 2005). The author’s site visits and Steinfeld 
et al. also found that CEMSs were being widely used (Steinfeld et al., 2009). As 
far as cost was concerned, there was little reason for coal-­
fired power plants to 
resist the installation of CEMSs. Two CEMSs in Plant 3 in Table 6.1 cost about 
US$132,000, only 0.5% of the capital costs of the plant’s SO2 scrubbers. How­
ever, CEMSs may not report credible and reliable data. One concern was over 
the quality of the equipment used. CEMSs cost much more in the United States: 
according to a cost model from the U.S. Environmental Protection Agency (EPA), 
it generally required more than half a million dollars for one set (The U.S. EPA, 
2007). The 2004 general survey found that only 20% of the CEMSs in China were 
functioning normally (Pan et al., 2005), local environmental protection bureaus 
generally refused to accept data from CEMSs and only one was recognized as a 
credible data source for the purposes of levying the SO2 effluent discharge fee 
(Pan et al., 2005). Later on, CEMSs were officially accepted as data sources after 


114 Policy implementation
 
ge 
US$/MWh
The 
The 
2.5
0.7
6.83 RMB. Upon the request 
 (State Devel-
**
disclosed in Plant 4. Comparing with other plants in the eastern provinces, 1.0% is used here for later analysis. ** 
(State 
standard 
Effluent 
dischar
fee 
2.0
2.8
0.3
0.7
0.4~0.7
2008. 
data 
the 
for 
2
official 
generally 
and 
emium for 
desulfurized 
of US$0.092/kg SO
China’s 
generation; 
Price 
pr
electricity
US$/MWh
2.2
2.2
2.2
2.2
2.2
3.7
estimation 
 
 =
compiling 
gin 
recent 
mar
ofit 
generation
US$/MWh
~ 14.6
> 0
>> 14.6
< 7.3
Pr
of electricity 
most 
their 
reflect 
31, 2008, is used: US$1
 
& 
gin 
 
mar
The fee rate refers to the level 
in 
electricity 
assumed 
as 
efficiencies of 
20%, 
ash, 
thermal 
interviews; 
Operation
maintenance 
(O&M) costs
US$/MWh
~4.1
1.8
2.2
<2.2
(1): >3.7;
(2): <3.7
profit 
and 
 scrubbers. 
in 
3.7
rates 
2
SO
retention 
s 
’
costs 
author
ts
Sulfur content
The 
down 
the 
to 
*
2009. 
shutting 
echnical Supervision, 2007).
s seven coal-fired power plan
sulfur 
are 
%
3.0%
4.0%
3.5%
1.0% 
0.5%
1.0%
0.7~1.1%
T
July 
dollars, the exchange rate on December
and 
assumptions 
according 
95%, 
 
June 
payment if 
 
in 
 
2
t
t
t
t
SO
scrubber 
type
t
t
We
We
We
We
We
We
(1): dry;
(2): wet
additional 
intermediate 
scrubbers, 
interviews 
 scrubbers in China’
wet 
the 
The 
of 
New or 
Retrofit
Retrofit
Retrofit
Retrofit
Retrofit
s 
ofit
’
reflect 
2003). 
s 
Retr
New
New
author
calculated to 
al., 
the 
 
et
efficiencie
in 
2
removal 
Region
Southwest
Southwest
Southwest
collected 
is 
 
2
East
East
East
East
Commission 
Data on SO
fee 
were 
original currency units were in Chinese RMB. In the conversion to U.S. 
ge 
data 
Planning 
2007b); SO
 
The 
opment 
Council, 
Table 6.1
Plant 3
Plant 5
Plant 6
Plant 7
clear information on sulfur contents was 
Note: 
of the interviewees, the names of coal-fired power plants are intentionally not shown.
dischar
Plant No.
No 
effluent 
Plant 1
Plant 2
Plant 4
levels at the corresponding unit scales (Zhejiang Bureau of Quality and 
 
*


Policy implementation  115
their online connection with provincial environmental protection bureaus. How­
ever, the author’s interviewees still said that they did not fully trust data from 
CEMSs. The locations of the sensors could affect the readings of CEMSs, and 
data reporting could also be manipulated. In 2008, three coal-­
fired power plants 
were caught illegally setting up ceilings of outlet SO2 concentrations that could be 
reported (Ministry of Environmental Protection, 2009c).
Fourth, coal-­
fired power plants could either cheat or even collude with environ­
mental compliance inspectors. Data from CEMSs were compared quarterly with 
direct measurements to verify accuracy (State Council, 2007b). A survey in China 
found that multiple inspections per annum tended to deter violation, no matter 
which government level inspectors were from (Lu et al., 2006). However, the 
effectiveness of site inspections could be constrained. According to the author’s 
interviews, many plants were able to raise the removal efficiency of their SO2 
scrubbers from zero to the designed level in half an hour and significantly more 
quickly if from an intermediate level. Some plants therefore kept their scrubbers 
either turned off or on low power and put them in full operation only when an 
inspection was imminent, enabling them to keep costs down while also passing 
the inspection. Even if abnormal operation were caught, a solution could be to 
collude with inspectors through bribes.
3  
Reversing noncompliance: penalty
The proper operation of SO2 scrubbers demands strong enough incentives to over­
come the hurdle of the high O&M costs. In the United States, the average O&M 
costs in 2008 were US$1.55/MWh (EPA and DOE, 2010). These figures could 
hardly be extrapolated for China because of the great differences in the capital 
costs of SO2 scrubbers, labor costs and other items. My interviews collected rel­
evant data from six coal-­
fired power plants, as presented in Table 6.1. The O&M 
costs varied from US$1.8 to 4.1/MWh, all above the average in the United States. 
Sulfur contents were the most influential factor: Plants 1 and 3 burned coals with 
approximately 3% to 4% sulfur content, and their O&M costs were roughly twice 
as high as those in Plants 4, 5, 6 and 7, which burned coals with 1% sulfur content 
or less. The O&M costs could be used as the average marginal costs of operating 
SO2 scrubbers. In generating electricity, several coal-­
fired power plants that the 
author visited had gross profit margins from not much above zero to significantly 
over US$14.4/MWh (including the O&M costs of SO2 scrubbers and the price 
premium for desulfurized electricity).
A survey of China’s inspection authorities and polluting firms found that fines 
for noncompliance were often not high enough to deter potential offenders (Lu 
et  al., 2006). The initial SO2 effluent discharge fee was about 0.20 RMB/kg 
(US$0.031/kg) in most provinces and lower than the marginal abatement costs in 
China’s large plants (Cao et al., 1999; Dasgupta et al., 1997). Polluting firms may 
simply pay to pollute. When facing a penalty, the first reaction of polluting firms 
was to negotiate with environmental protection bureaus or ask for the interference 
of local governments (Lu et al., 2006), which compromised the penalty.


116  Policy implementation
In July 2005, China’s SO2 effluent discharge fee was raised from US$0.031/kg 
in 2003 to US$0.092/kg (State Development Planning Commission et al., 2003). 
However, as converted to US$/MWh in Table 6.1, it was still too low to overcome 
the hurdle of the much higher O&M costs. Another increase was scheduled in 
2007 to reach US$0.18/kg in three years (State Council, 2007a), but the exact 
schedule varied from province to province. In Jiangsu Province, the higher rate 
had been in effect since July 2007 (Jiangsu Department of Environmental Pro­
tection, 2008), but in Henan Province, the lower rate was still being applied in 
the first quarter of 2010 (Henan Department of Environmental Protection, 2010). 
With the higher rate, coal-­
fired power plants burning high-­
sulfur coals would find 
operating SO2 scrubbers cheaper than paying the effluent discharge fee (Plants 1 
and 3 in Table 6.1). However, for others (Plants 4, 5, 6 and 7) when facing only 
this policy, the rational decision was to pay the fee.
Another policy was introduced in 2004. If new coal-­
fired power plants came 
online together with SO2 scrubbers, the desulfurized electricity could enjoy a 
price premium of US$2.2/MWh (NDRC and SEPA, 2007a). In June 2006, the 
policy extended to cover all SO2 scrubbers, including retrofitted ones (NDRC and 
SEPA, 2007a). Some coal-­
fired power plants were awarded higher price premi­
ums, such as Plant 7 in Table 6.1. The price premium and the effluent discharge 
fee together were a little higher than the O&M costs (Table 6.1), but the small 
difference indicated that the proper operation would be a rational decision only 
when most nonoperation cases were caught.
The 11th Five-­
Year Plan witnessed sharp increases in noncompliance penalties. 
In 2007, a harsh penalty measure was associated with the price premium for the 
first time. If the operation rate of an SO2 scrubber were lower than 80%, a penalty 
of US$11.0/MWh would be issued for any additional non-­
desulfurized electric­
ity generation (NDRC and SEPA, 2007b). The required minimum probability of 
catching nonoperation became much lower to induce the proper operation of SO2 
scrubbers. For Plant 3 in Table 6.1 burning high-­
sulfur coal, corresponding to the 
effluent discharge fee of US$0.092/kg SO2, a risk-­
neutral manager would decide 
to operate SO2 scrubbers properly if the probability of catching nonoperation 
exceeded 26% (see Table 6.2 for the calculation formula). For Plants 4 and 5 burn­
ing low-­
 to medium-­
sulfur coals, the minimum probability was about one seventh. 
Furthermore, additional penalties were introduced on the managers of coal-­
fired 
power plants. In China, almost all coal-­
fired power plants were owned by the 
state. The nonoperation of SO2 scrubbers could increase profit and benefit the 
managers’ career and salary. However, according to formal regulations (NDRC 
and SEPA, 2007b) and the author’s interviews, cheating and nonoperation could 
lead to the removal of the managers. They had to calculate the risk for themselves.
The penalties in 2007 also aimed for minimizing potential moral hazard when 
SO2 scrubbers occasionally had to stop operating due to accidents, malfunctions 
or other reasons. While they were out of action, SO2 emissions could be controlled 
either by minimizing the sulfur content of coal or by shutting down electricity 
generation. China would issue no penalty as long as the operation rate were above 
90%, a mild penalty of US$2.2/MWh if the rate were between 80% and 90% and 


Policy implementation 117
Table 6.2 Decision scenarios for the managers of coal-fired power plants
Scenario SO2 scrubbers SO2 scrubbers Electricity Net revenue of a coal-fired power 
functioning
operating
generation plant
(1)
Yes
Yes
Yes
Profit margin
(2)
Yes
No
Yes
Profit margin + O&M costs – C% × 
(Price premium + Discharge fee + 
Penalty)
(3)
No
No
Yes
Profit margin + O&M costs – C% × 
(Price premium + Discharge fee + 
Penalty)
(4)
No
No
No
0
Note: C% is the actual probability of catching the nonoperation of SO2 scrubbers. The 
proper operation of SO2 scrubbers, when they function, requires that the net revenue in sce-
nario (1) is greater than that in scenario (2). The corresponding condition can be calculated as 
O&M costs
C% >
. When SO2 scrubbers do not function, the discon-
Discharge fee
P
+
+
rice premium
Penalty
tinuation of electricity generation becomes a rational decision when the net revenue in scenario 
Profit margin + O&M costs
(4) is greater than that in scenario (3), or C% >
. Because 
Discharge fee
P
+
+
rice premium
Penalt
l y
profit margins are generally positive, it is accordingly easier to push for the proper operation of 
SO2 scrubbers when they function than to ask coal-fired power plants to discontinue electricity gen-
eration when they do not. In order to encourage coal-fired power plants to fix malfunctioning SO2 
scrubbers as soon as possible, the rational decision when SO2 scrubbers function should generate 
greater net revenue than that with malfunctioning SO2 scrubbers. The condition is fulfilled when 
O&M costs
C% >
.
Discharge fee
P
+
+
rice premium
Penalty
a harsh penalty of US$11.0/MWh if the rate were under 80% (NDRC and SEPA, 
2007b). Because it was expensive to restart electricity generation, the O&M costs 
of SO2 scrubbers may not be critical in the decision making when SO2 scrubbers 
could get fixed soon.
When problems have to take much time to fix – for example, several weeks – and 
the penalty of US$11.0/MWh is applied, the economic incentives should make it 
a rational decision to discontinue electricity generation for many coal-fired 
 
power 
plant managers. Electricity generation without operating SO2 scrubbers earned 
a profit margin and avoided the O&M costs of SO2 scrubbers, but if the non-
operation of SO2 scrubbers were caught, coal-fired 
 
power plants would need to 
return the price premium and pay the effluent discharge fee as well as the penalty. 
Many coal- 
fired power plants might continue generating electricity as long as the 
probability of catching the nonoperation of SO2 scrubbers was low enough (see 
Table 6.2 for the specific calculation). For coal-fired 
 
power plants with large profit 
margins (such as Plant 5 in Table 6.1), electricity generation should continue even 
when nonoperation could not be hidden at all. However, when the author visited 
Plant 5, electricity generation in one system had been discontinued for several 
weeks due to its malfunctioning SO2 scrubber. Personal penalties on the manag-
ers could have played a role. Furthermore, even if the decision was to continue 
electricity generation, a high-enough probability 
 
of detection was still necessary 


118  Policy implementation
to encourage coal-­
fired power plants to fix malfunctioning SO2 scrubbers as soon 
as possible (see Table 6.2 for the specific calculation). If the actual probability was 
not expected to reach this level, there would be little concern about the quality of 
SO2 scrubbers, as in the early years.
Furthermore, coal-­
fired power plants should also comply with regulations on 
SO2 removal efficiency and effluent emission standards (NDRC and SEPA, 2007b; 
SEPA and General Administration of Quality Supervision Inspection and Quar­
antine, 2003; MEP and AQSIQ, 2011). Technically in practice, a coal-­
fired power 
plant could choose a designated SO2 removal efficiency. For example, higher 
ratios of Ca/S (the molar ratio between CaCO3 and SOx) or L/G (the liquid-­
to-­
gas 
ratio in volume) would remove more SO2 from the flue gas. Reasonably, if not 
regulated, a coal-­
fired power plant could lower SO2 removal efficiencies to reduce 
costs. On the other hand, because of changing sulfur contents and workload, SO2 
concentration and flue gas volume were not stable. Scrubbers’ capability to track 
the changes – with the same methods of adjusting SO2 removal efficiencies – was 
necessary for their reliable operation.
The Chinese central government mandated minimum SO2 removal efficiencies 
being established (NDRC and SEPA, 2007b) and provincial governments were in 
charge of the details. For example, when SO2 removal efficiencies were lower than 
predetermined levels (generally 90% for wet scrubbers), Henan Province simply 
counted the time as nonoperation (Henan Development and Reform Commission 
and Henan Environmental Protection Bureau, 2007). In normal conditions, the 
incentives were strong enough to make SO2 scrubbers reach the required levels of 
SO2 removal efficiencies. Two actual cases from the author’s field trip could demon­
strate the decisions. In the first case, sulfur contents went up significantly but were 
expected to be a temporary situation. The designed sulfur content for Plant 6’s SO2 
scrubber was 0.84%, but for a period in 2008 when coal supply was constrained, 
the actual sulfur content was higher than 2%. Such a dramatic increase in sulfur 
content became a serious burden. To maintain SO2 removal efficiencies over 90%, 
the solution was to raise the Ca/S ratio from the designed level of 1.03 to 1.3. In the 
second case, when the increased sulfur contents were expected to be long-­
lasting, 
a temporary solution would not be sustainable. One of the eight coal-­
fired power 
plants the author visited had to shut down and modify the original SO2 scrubber to 
handle the much higher sulfur input rate. Particularly, the absorber tower became 
significantly taller by adding another section on the top of the original one. The 
pathway was accordingly longer for the flue gas and limestone slurry to contact and 
react. Additional circulation pumps could also be added to enhance the L/G ratio.
In order to better implement the incentives, responsible government agen­
cies are specified: electric grid corporations were in charge of paying the price 
premium in time; provincial environmental protection bureaus collected effluent 
discharge fees; provincial price agencies were responsible to recover unjustified 
price premium according to actual operation rates (NDRC and SEPA, 2007b). 
Seven coal-­
fired power plants in 2008 and five in 2009 were penalized for cheat­
ing or nonoperation with the US$11.0/MWh penalty applied (Ministry of Envi­
ronmental Protection, 2009c, 2008).


Policy implementation  119
Central and local governments in China are not the only entities that have their 
tasks centered around goals. Because almost all coal-­
fired power plants in China 
were state-­
owned, they were also assigned quota or goals for their total SO2 emis­
sions (SEPA, 2006). Both goals and policies play crucial roles in their compli­
ance decisions on the operation of their SO2 scrubbers. In addition to financial 
penalties, administrative penalties were also applied for noncompliance. In envi­
ronmental enforcement and compliance, decision makers at local governments, 
power corporations and coal-­
fired power plants also kept in mind their SO2 emis­
sion caps or goals. If SO2 removal efficiencies were too low and nonoperation was 
caught, the SO2 emission permits could be used up soon. In addition, seriously 
abnormal operation of SO2 scrubbers was publicly punished (MEP and NDRC, 
2008; Ministry of Environmental Protection, 2009a), which could affect the 
career of the coal-­
fired power plants’ managers. In the words of an interviewee, 
“it is not worthwhile for the managers of a coal-­
fired power plant to risk losing 
the positions to save money for the plant. Anyway, the money is not theirs, but 
the positions are.”
4  
Reversing noncompliance: environmental 
compliance monitoring
The effectiveness of environmental compliance monitoring determines the prob­
ability of catching noncompliance. China’s emission data MRV system is largely 
bottom up, which could potentially suffer from two major challenges. The first 
challenge lies in the system’s high costs. Compliance monitoring resource con­
straints exist in all countries, but the problem is especially daunting in develop­
ing countries, due to the high costs of compliance monitoring, limited resources, 
understaffed environmental agencies, inadequate training and technological sup­
port (Arguedas, 2008; McAllister et al., 2010; Blackman and Harrington, 2000; 
Russell and Vaughan, 2003; Pan et  al., 2005). How to better utilize available 
resources is critical to determine the effectiveness of every domestic policy and 
international environmental treaty. Compliance monitoring is the most resource-­
consuming activity in enforcing environmental policies. For example, an emis­
sion trading scheme should effectively deter cheating and verify actual emission 
levels (Kruger and Egenhofer, 2006), while compliance monitoring was respon­
sible for 69% of transaction costs for German companies in the European Union 
CO2 Emission Trading Scheme (Heindl, 2012). The existence of many small and 
medium-­
sized polluters could seriously attenuate available resources, even in 
developed countries where the rule of law is generally well established. Due to 
the significant economy of scale, large point sources generally have lower com­
pliance monitoring costs on a per-­
ton-­
emission basis and are often prioritized 
(Heindl, 2012; Gray and Deily, 1996). Because of China’s sheer size, the large 
system involves many personnel and occupies substantial resources. In the 12th 
Five-­
Year Plan (2011–2015) alone, the Chinese government planned to invest 
40 billion RMB (~US$5.9 billion) to enhance related environmental regulation 
capacity (MEP, 2013).


120  Policy implementation
The second challenge is intentional data manipulation. Environmental moni­
toring and reporting in China generally must pass through, and be inspected by, 
polluting firms and various levels of local governments and relevant agencies 
before reaching the central government. Most environmental compliance capaci­
ties, such as personnel and governmental expenditure, are in local governments, 
while the central government is mainly in charge of policy making. Emissions 
of CO2, SO2 and NOx are generally calculated via bottom-­
up energy consump­
tion data and emission factors (Liu et al., 2015; Lu et al., 2011; Zhang et al., 
2007). This approach is often subject to the influence of intentional distortions 
for the interest of stakeholders along the path (Tsinghua University, 2010). China 
has been exerting increasingly high pressure on local governments and energy-­
intensive firms to achieve top-­
down energy and emission control goals from the 
central government (Xu, 2011b). In comparison to the technologically challeng­
ing, economically expensive and politically difficult tasks of actual mitigation, it 
would be much more convenient to twist the reported numbers (Jin et al., 2016).
The objective resource constraint and the intentional data manipulation could 
seriously compromise data quality and thus the effectiveness of environmen­
tal compliance monitoring. Facing immense pressure of environmental crises, 
the Chinese government has been actively searching for potential solutions for 
enhancing environmental data quality.
4.1  
Model construction
In order to understand China’s environmental compliance monitoring in greater 
depth, a conceptual, computable model is constructed to simulate the evolution 
of compliance rates under different compliance monitoring strategies and how 
influential factors in three categories – pollution abatement costs, noncompliance 
penalty and, most important, compliance monitoring effectiveness – affect com­
pliance decisions of polluters and thus the compliance rate. Mathematical details 
of the model are provided in the Appendix to this chapter.
This model stands on the shoulders of two pieces of research literature for cre­
ating a theoretical framework. The first well-­
developed economics literature of 
crime and punishment understands crimes as rational choices. Whether a pol­
luter chooses compliance or noncompliance is based on comparing related costs 
and benefits (Polinsky and Shavell, 2000; Becker, 1968; Glaeser, 1999; Xu, 
2011a; Shimshack, 2014; Levitt, 2004). If a polluter pondered not complying 
with an environmental regulation, pollution abatement costs could be saved as its 
expected benefits. However, such behavior would incur expected costs, which is a 
product of (1) penalty on noncompliance and (2) the probability of being caught. 
Risk-­
neutral rational polluters would choose environmental noncompliance if the 
expected benefits were greater than the expected costs. The compliance or non­
compliance decision is assumed to be deliberate but not at random. The second 
mature literature, or a series of related literature, such as on policing, pollution 
control and tax evasion, examines how to enhance the probability of catching non­
compliance. Compliance monitoring could apply various strategies for enhancing 


Policy implementation  121
the probability with a given amount of resources, although the effectiveness is 
mixed. Levitt (2004) found that policing strategies are of only minor signifi­
cance, while the number of police may explain a large proportion of the crime rate 
change. For tax compliance, endogenous audit selection rules screen taxpayers 
for potential auditing, but the impacts on compliance are mixed (Konrad et al., 
2017; Vossler and Gilpatric, 2018). In epidemiology, strategies are developed to 
promote public health and enhance the rate of finding sick patients at early stages 
among a population (Bonita et al., 2006). A population would be first screened, 
and those with positive results would have to go through another round of more 
careful diagnosing for confirming whether they were true or false positive.
These two pieces of literature are integrated together in this study to simulate 
environmental compliance decisions. Two environmental compliance monitoring 
systems are proposed and simulated, as illustrated in Figure 6.2. The conventional 
system that is based on monitoring, reporting and verification is simplified to 
require governmental compliance monitoring resources primarily for site inspec­
tion. Adopting the terminology in epidemiology, the model refers to these activi­
ties as diagnosing. If a polluter were caught as being noncompliant, a penalty 
would be issued. The new compliance monitoring system inserts an additional 
step before diagnosing to actively screen polluters into high-­
risk and low-­
risk 
groups, with higher and lower probabilities of being noncompliant, respectively. 
Diagnosing with higher costs follows with site inspections or other more accurate 
means to confirm noncompliance only in the high-­
risk group. For the convenience 
Polluters in compliance and noncompliance
Screening: cheap but more errors 
High-risk group
Low-risk group
Diagnosing: expensive but accurate
Penalty
No penalty
Polluters making compliance decisions
Compliance 
monitoring 
resources
Screening: c
High-risk grou
Diagnosing: expensive but accurate
Penalty
Compliance 
monitoring
resources
Figure 6.2  
A conceptual model of environmental compliance monitoring
Note: The dash-line arrows indicate the screening system’s flow, while the diagonal-pattern arrows 
refer to the diagnosing system’s flow. Their major difference is the existence/absence of the screen­
ing step with screening technologies. The gray boxes show compliance monitoring resources that not 
only are allocated between screening and diagnosing technologies in the screening system but only to 
diagnosing technologies in the diagnosing system.


122  Policy implementation
of discussion, the conventional system is referred to in this chapter as the diagnos­
ing system, while the new system contains both screening and diagnosing, and it 
will be called the screening system. Numerous studies have applied the economic 
model of crime and punishment for understanding environmental noncompliance 
(Xu, 2011a; Shimshack, 2014; Guo et  al., 2014). The compliance monitoring 
strategy with screening has also been widely applied in multiple fields (Konrad 
et al., 2017; Vossler and Gilpatric, 2018; Bonita et al., 2006).
4.2  
Strengthening the conventional diagnosing system
China has made several prominent improvements in monitoring and site inspec­
tion to address the previously mentioned two challenges for enhancing the prob­
ability of catching the nonoperation of SO2 scrubbers. First, more resources were 
made available for environmental compliance monitoring. The numbers of gov­
ernment employees at all levels increased from 46,984 in 2005 to 52,944 in 2009 
and 61,668 in 2015 for environmental monitoring and from 50,040 in 2005 to 
60,896 in 2009 and 66,379 in 2015 for inspection (Figure 3.1). Although still lim­
ited, the personnel resources had already been enough to have an intensive focus 
on SO2 scrubbers in coal-­
fired power plants. Particularly, only 503 coal-­
fired 
power plants housed 461 GW SO2 scrubbers (1,264 systems) at the end of 2009, 
and the largest 300 had a total capacity share of 82% (Ministry of Environmental 
Protection, 2010a). In 2013, 282 coal-­
fired power plants that were at or greater 
than 1 GW each had 470 GW SO2 scrubbers in total, or 62.3% of all (Ministry of 
Environmental Protection, 2014). Government personnel are sufficient to follow 
these large plants closely and conduct inspections frequently.
The number of polluting sources (N) that require compliance monitoring varies 
dramatically, depending on focused polluter sizes, pollutants and other features. 
China conducted the first census of polluting sources with the census date being 
December 31, 2007, and pollution information for 2007, covering 5,925,576 pol­
luting sources, including 1,575,504 industrial, 2,899,638 agricultural, 1,445,644 
domestic and 4,790 centralized pollution control facilities (Ministry of Environ­
mental Protection et al., 2010). In comparison, China’s annual environmental sta­
tistics report focused on about one tenth of the polluting sources, being 161,598 
industrial sources, 131,837 farms and 7,578 districts for animal husbandry and 
6,910 water treatment plants, 2,315 municipal waste treatment facilities and 866 
hazardous waste treatment facilities in 2015 (Ministry of Environmental Protec­
tion, 2002–2016). Among these sources, 68,121 polluting sources were under spe­
cial supervisory monitoring (Ministry of Environmental Protection, 2002–2016).
Furthermore, in order to make the conventional diagnosing system more effi­
cient, CEMSs have become critical to monitor the operation of SO2 scrubbers 
especially since 2007 (NDRC and SEPA, 2007b). Six plants (all in Table  6.1 
except Plant 2) allowed me to read the computer screens of their CEMSs. The val­
ues of SO2 concentrations changed continuously, and different data were generally 
consistent. Many CEMSs and SO2 scrubbers had been inspected once or twice 
a month. Because CEMSs transmitted data online and in real time, inspections 


Policy implementation  123
often followed abnormal data reporting. Coal-­
fired power plants were informed 
in advance of some inspections, but in many other cases, inspections were unan­
nounced. Inspectors had the right to enter coal-­
fired power plants without being 
delayed. In the plants that I visited, inspection vehicles generally needed just a 
few minutes to drive from the gates to the sites where the SO2 scrubbers were 
installed. China was actively building up its site inspection capacity. The num­
ber of government inspectors at all levels increased steadily (Figure 3.1). China 
focused on monitoring and inspection in its efforts to build capacity. During the 
period between 2006 and 2008, the two functions accounted for 85% of govern­
ment personnel growth for environmental protection (Ministry of Environmental 
Protection, 2006–2009).
Because of the concern about their data accuracy and reliability, as discovered 
in my interviews, CEMSs were not the only data source to track the operation of 
SO2 scrubbers. Other relevant data were collected, including operation and main­
tenance records, load factors of electricity generation, sulfur contents of coal, the 
consumption of limestone and other reagents, electricity consumption, the han­
dling of products from SO2 scrubbers, the opening and closure of bypass dampers 
and records of accidents and responses (NDRC and SEPA, 2007b; SEPA, 2007). 
SO2 concentration in the inlet flue gas corresponds to the sulfur contents within a 
fairly predictable range. The load factors of electricity generation decide the flow 
rate of the flue gas and can check direct measurement with CEMSs. The factors 
together determine the sulfur load to an SO2 scrubber system. For wet scrubbers 
using limestone as the reagent, the molar ratio between CaCO3 and SO2 is nor­
mally quite stable at approximately 1.02 to 1.05 (Ministry of Environmental Pro­
tection, 2010b). Then the sulfur load would decide the consumption of limestone 
and the production of gypsum. The managers of coal-­
fired power plants were 
asked to keep the receipts of limestone purchases, and cheating on receipts was 
considered financial fraud, with harsh penalties on those responsible. Electricity 
is another important input to operate SO2 scrubbers. Because all data should be 
consistent with each other, it became more difficult to cheat.
The problem of collusion appeared under control. Data from CEMSs were sent 
to more than one agency, including environmental protection bureaus and elec­
tric grid corporations. Authorities at China’s four government levels – central, 
provincial, prefectural and county – all inspected SO2 scrubbers. The multiplicity 
of inspection authorities effectively diminished the opportunities of collusion. In 
addition, the pressure to achieve the 10% reduction goal of SO2 emissions in the 
11th Five-­
Year Plan reduced incentives to collude.
4.3  
Building the screening system with big data
The preceding measures to strengthen the diagnosing system indeed worked, 
but for achieving an even deeper reduction of SO2 emissions, China faces much 
more daunting problems in dealing with smaller polluting sources that are a few 
orders of magnitude greater in numbers. New opportunities are emerging with 
newly emerged environmental compliance monitoring technologies (Kitchin, 


124  Policy implementation
2014), which are evolving rapidly in terms of effectiveness in catching noncom­
pliance and efficiency in utilizing compliance monitoring resources. For example, 
CEMSs played a central role in the U.S. Acid Rain Program as well as the Euro­
pean Union Emission Trading Scheme (The U.S. Congress, 1990; Stranlund and 
Chavez, 2000; European Commission, 2012). Remote-­
sensing technologies using 
satellites could provide large-­
scale spatial coverage of multiple pollutants (Streets 
et al., 2013). The measurement extends to areas beyond the current monitoring 
network, although the spatial resolution is coarse (Streets et al., 2013). Social 
media and the prevalent use of smartphones have greatly facilitated and strength­
ened the power of the civil society in monitoring environmental pollution and 
compliance (Stevens and Ochab, 2010; Kay et al., 2015). Various types of sensors, 
in addition to novel carriers such as unmanned aerial vehicles, have been more 
and more widely adopted to measure pollution levels (Snyder et al., 2013; Wang 
and Brauer, 2014).
China has been actively seeking opportunities in big data that can be applied for 
environmental protection. In 2015, State Council formally issued the Action Out­
line for Promoting Big Data Development to encourage the wide integration of 
big data in governance (State Council, 2015). In 2016, the then Ministry of Envi­
ronmental Protection enacted the Comprehensive Plan on Ecological and Envi­
ronmental Big Data Construction (Ministry of Environmental Protection, 2016a). 
It listed a comprehensive plan on how big data could be collected, integrated, 
developed and applied for environmental compliance monitoring, enforcement 
and management.
These new compliance monitoring technologies shed light on new solutions to 
the old challenges. First, in addressing the compliance monitoring resource con­
straint, these technologies could potentially provide a relatively low-­
cost means 
to monitor polluting sources. For example, although one satellite observing the 
Earth’s CO2 and air quality could cost a few hundred million U.S. dollars, such as 
the OCO-­
2 satellite for CO2 monitoring by the National Aeronautical and Space 
Administration with a price tag of US$465 million, its wide spatial and regular 
coverage would substantially reduce the average and, especially, marginal costs 
for one observation (Wall, July 2, 2014; Osterman et al., 2018). Second, many of 
these technologies could circumvent various levels of local governments and pol­
luting sources to provide top-­
down, external and objective data without subjective 
distortions. They are originated from entirely different external sources, not inter­
nal reporting. Satellite or remote-­
sensing data could be gathered in a centralized 
manner without the direct involvement of local governments or polluting sources 
themselves.
Nevertheless, these new technologies also have a critical weakness. Most of 
them generally have not reached the minimum accuracy requirements to legally 
or administratively punish polluters, while conventional technologies currently 
in application (although not all) could fulfill the requirements if intentional data 
manipulation is effectively deterred. Remote-­
sensing data have been successfully 
applied in China to examine the impacts of environmental policies on pollutant 
emissions from coal-­
fired power plants, but the accuracy has not been adequate 


Policy implementation  125
to justify their direct application in legally determining the compliance status of 
individual polluting firms (Zhang et al., 2009; Li et al., 2010).
The trade-­
offs between conventional and new technologies indicate that the 
latter cannot completely replace the former at their current stage, but their clear 
advantages in costs (and objectiveness) are crucial considerations for China’s 
ongoing reform on the conventional diagnosing system to deeply integrate big 
data and other technologies. Section 4.4 mainly focuses on how this reform may 
achieve better efficiency and effectiveness in environmental compliance monitor­
ing. Different technologies are recognized to have different features mainly from 
cost and accuracy perspectives. Their weaknesses and strengths could comple­
ment each other for building a better system than any individual category of tech­
nologies can do alone.
4.4  
Comparing diagnosing and screening systems
Environmental compliance rates are simulated with empirically defined param­
eters as discussed in the Appendix to this chapter. This subsection discusses the 
model simulation and sensitivity analysis results. If any input parameter is not 
targeted in a simulation, it will adopt the empirical value as specified in the current 
scenario as summarized in Table 6.3.
Compliance rates (1−M t) in the screening system depend on their initial levels 
(1
0
−M ; Figure 6.3). For example, if initially with 40,000 inspection staff, the 
screening system results in two equilibrium compliance rates (1−M *) after sev­
eral time steps, about 27% (a very low compliance rate) and 100% (full compli­
ance; Figure 6.3). An equilibrium state is defined as, given the empirical values 
of parameters, the compliance rate remains stable over time and swings back if 
a small disturbance happens (Table 6.3). When the initial compliance rates are 
above a certain level, the final equilibrium compliance rates tend to converge 
to a high level close to full compliance. However, when the initial compliance 
rates are below that level, the available resources would not be adequate to catch 
enough noncompliance cases. Noncompliance will become the dominant choice 
of rational polluters, or the compliance monitoring system falls into a noncompli­
ance trap due to its equilibrium status. The following simulations of the screen­
ing system will primarily report equilibrium compliance rates. In contrast, the 
diagnosing system demonstrates no memory. Its compliance rates at each time-­
step (1−M t) have no relationship with the initial or proceeding levels (1
0
−M  and 
1
1
−
−
M t ). They are decided only by immediately available compliance monitoring 
resources (Rt; Figure 6.3).
The relative effectiveness of the diagnosing and screening systems in enhanc­
ing compliance rates depends heavily on resource availability (Rt; Figure 6.4). 
When resources were too scarce (e.g., less than 30,000 inspection staff or half of 
China’s available personnel in 2015), neither system would be able to result in 
high-­
compliance statuses, although the diagnosing system could achieve slightly 
better outcomes. When resources were abundant (more than 130,000 inspection 
staff or doubling the available personnel in 2015), either system would lead to 


126 Policy implementation
Simulation
6.4
 6.3
 
 6.5(a)
6.5(b)
Figure
Figure
Figure
 
Figure
Empirical ranges in the model 
under special supervisory monitoring 
 
in 2015 (Ministry of Environmental 
Protection, 2002–2016) to 5,925,576 
C
) 
s from 0% (full compliance
s first census of polluting 
simulation
to 100% (complete noncompliance), 
which covers the full range of possible 
compliance rates
number of inspection staff at the 
to 185,108 (the total number of 
environmental officials at all levels 
; 
 varie
0
central and provincial levels in 2015) 
for administration, inspection and 
monitoring in 2015; Ministry of 
Protection, 2002–2016)
Environmental 
in China’
sources with the census date being 
al., 2010)
 scrubbers, 
1a; 
1.6 (very 
31, 2007 (Ministry of 
 
; in the 2007 
P
2
 [Xu, 201
P
C
A, 2007b]) to 
 60% higher than 
for operating SO
s shale-gas development, the 
Environmental Protection et
NDRC and SEP
C
 
 with 
in China’
 was significantly lower than 
 al., 2014])
December
regulation 
 was five times of 
lenient P
P
P
then 
[Guo et
M
From 1,959 inspection staff (the total 
From 68,121 polluting sources that were 
From 0.1 (very harsh 
2018)
s 
, is 
Empirical values in the current 
0
M
scenario
point of the full possible range 
between 0% and 100%.
66,379 inspection staff (in 2015; 
Ministry of Environmental 
Protection, 2002–2016), within 
which 46,800, or 70.5%, were 
environmental inspectors (in 2017) 
as in the “double randomness, one 
publicization” databases (Ministry 
of Ecology and Environment, 
809,500 polluters under compliance 
monitoring as in the “double 
randomness, one publicization” 
databases (Ministry of Ecology and 
Key parameters in the model and their empirical values
Initial noncompliance rate, 
assumed to be 50% as the middle 
Environment, 2018)
 is assumed to be 1.5 times 
P
2/3 (
of the pollution abatement costs, 
being a middle ground in China’
empirical cases as introduced in the 
cell to the right)
 
; 
 
Noncompliance rate (%) at time-step t
the corresponding compliance rate is
 
. Equilibrium noncompliance
, is defined as the level when
When it 
otal available resources for 
compliance monitoring at the time-
, which could be allocated 
between screening and diagnosing, 
 can be changed 
exogenously at a time step. 
.
=
t
R
R
.
0 0
. %
t
R
. 
: the penalty on 
1
P
−
t
d
*
R
t
M
M
rate, 
t
t
−
=
M
M
t
 and 
1 − 
step 
t
Rs
remains a constant: 
 
The number of polluting sources 
under compliance monitoring
 pollution abatement costs (US$/
sources; 
T
C:
ton), which vary across polluting 
noncompliance (US$/ton), which 
is assumed to be fixed for every 
punished polluting source.
R
t
Table 6.3
Parameters
t
 or 
M
R
N
C
P


Policy implementation 127
(b)
(d)
& 
6.5(d)
 
 6.5(c)
 
6.6(a)
6.6(c) & 
Figure
Figure
 
 
Figure
Figure
The impacts of a lognormal distribution 
are also simulated, due to the lack of 
actual information.
Due to inadequate information, the 
ined with a full 
, is exam
s
d
r
r
ratio, 
possible range from 1% to 100%. By 
definition, screening technologies 
must be cheaper than diagnosing 
technologies. Otherwise, the latter will 
be better from both cost and accuracy 
perspectives to make the former 
obsolete.
Due to inadequate information, a full 
range, 0%~100%, is examined.
Due to inadequate information, a full 
range, 0%~100%, is examined.
 normal distribution is assumed with 
a standard deviation of 0.33.
-year per 
 is assumed to be 
 : 0.074 inspector
inspection (see text for empirical 
s
d
r
r
 is equivalently 0.0074 
estimation); 
.
.
s
-year per inspection.
, the corresponding 
10%, or r
, the corresponding 
inspector
1d
 and 
probabilities for screening and 
respectively
2d
K
K
 and 
1s
diagnosing technologies, are 
assumed to be 90% and 99%, 
2s
probabilities for screening and 
diagnosing technologies, are 
assumed to be 70% and 90%, 
respectively
A
d
r
K
K
C
P
 
Cumulative distribution function of 
s
r
vely), which are 
1
Required resources to screen and 
 polluting source (
-
1
K
ype 
 : (T
1
K2
-
one
, respecti
error) the probability that 
diagnose 
r
assumed to remain unchanged over 
d
and 
time
The probability (%) that one 
technology recognizes compliant 
cases as being compliant; 
ype I 
: (T
compliant cases are recognized as 
being noncompliant.
The probability (%) that one 
technology recognizes 
noncompliant cases as being 
noncompliant; 
II error) the probability that 
noncompliant cases are recognized 
as being compliant
•
( )
1
2
Φ
r
K
K


128  Policy implementation
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
5
10
15
20
25
30
35
40
45
50
Inspection staff
e
t
a
r
 
e
c
n
a
i
l
p
m
o
C
Time step
Diagnosing system
Screening system
Campaign
Available inspection
staff (right)
Shock
Figure 6.3  
Model simulation of compliance rates (1−M t) in the diagnosing and screen­
ing systems with available compliance monitoring resources (i.e., exogenously 
determined number of inspection staff in the dashed curve, Rt) and initial com­
pliance rates (1
0
−M , from 0% to 100%)
Note: After compliance rates reach equilibrium levels (1
1
1
0
−
−
−
M
M
M t
*
), a hypothetical envi­
ronmental campaign (temporarily with more inspection staff) is exogenously triggered to run for three 
time-steps and a hypothetical shock (temporarily with fewer inspection staff) for two time-steps. Their 
periods are indicated alongside the dashed curve. All other model parameters adopt the empirical 
values in the current scenario in Table 6.3.
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Inspection staff
Screening system
Diagnosing system
Figure 6.4  
Model simulation of equilibrium compliance rates (1−M *) in the screening and 
diagnosing systems in relation to available inspection staff (R)


Policy implementation 129
nearly full compliance and strategies would not matter much. Most situations in 
the real world, including China in 2015 with 66,379 inspection staff, should fall 
in between: resources are constrained but neither unlimited nor depleted. In these 
situations, the two systems would diverge away from each other and the screening 
system could use available resources much more efficiently to achieve signifi-
cantly higher compliance rates (Figure 6.4).
The number of polluters (N) matters greatly for the relative performance of 
the two compliance monitoring systems. With 66,379 environmental inspection 
staff in the current scenario, the screening system shows significantly higher 
compliance rates than the diagnosing system when the number of polluters is 
between 0.5 million to about 1.2 million (Figure 6.5(a)). Both systems could 
effectively handle fewer than 0.5 million polluters for their nearly full compli-
ance, while neither system could be up for the job with more than 1.2 million 
polluters. As discussed in the Appendix at the end of this chapter, the polluting 
sources under the central government’s special supervisory monitoring, gener-
ally large or hazardous polluters, were 68,121 in 2015. The currently available 
inspection staff would be of little resource constraint to achieve their general 
environmental compliance, as in China’s current situation. The “double ran-
domness, one publicization” scheme covered 809,500 polluters, for which the 
screening system with nearly full compliance tends to have a great advantage 
over the diagnosing system with only about half of polluters under compliance. 
If compliance monitoring does not differentiate the 5,925,576 polluting sources 
in the 2007 census, the overall compliance rate would be very low, being less 
than 5%. As in the Chinese practice, compliance monitoring should strategically 
allocate resources to those bigger and more severe polluters. Otherwise, the 
system would be overwhelmed. From another perspective, the model simulation 
also indicates that small polluters have significantly low environmental compli-
ance rates.
As enlightened in the economic theory of crime and punishment, a penalty 
could enhance compliance rates in a similar way as compliance monitoring. When 
C
the penalty level is ten times of the pollution abatement costs (i.e., 
 being 0.1), 
P
both the screening and the diagnosing systems could yield nearly full compli-
ance (Figure 6.5(b)). For example, in ensuring the normal operation of SO2 scrub-
bers, the penalty for noncompliance was five times the pollution abatement costs 
C
(i.e., 
 being 0.2; Xu, 2011a). China’s compliance monitoring system was closer 
P
to the diagnosing system, but it still effectively brought coal-fired 
 
power plants 
under prevalent compliance as projected by the model (Figure 6.5(b); Xu, 2011a). 
C
When the penalty level barely catches up with the abatement costs (i.e., 
  > 1), 
P
neither system would work although the screening system performs even worse 
(Figure 6.5(b)). This was the case in China’s early days in dealing with water pol-
lution in shale-gas development (Guo et
 
 al., 2014).
A deviation of the statistical distribution of the cost/penalty ratio (Φ( )
• ) 
does not seem to cause much difference for the earlier simulation results 


130 Policy implementation
Figure 6.5  
Model simulation of equilibrium compliance rates (1-M *) in the screening 
and diagnosing systems in relation to (a) the number of polluters (N); (b) the 
 
C 
ratios between pollution abatement costs and noncompliance penalty 


; 


P 


(c) available inspection staff (R), where the pollution abatement cost-to-non-
C 
compliance penalty ratio 


 has a lognormal distribution (Φ( )
• ); and (d) the 


P 


r 
relative resource intensity of screening and diagnosing technologies 
s 




r
d 


0%
20%
40%
60%
80%
100%
0
500,000
1,000,000
1,500,000
2,000,000
2,500,000
s
e
t
a
r
e
c
n
a
i
l
p
m
o
c
m
u
i
r
b
i
l
i
u
q
E
Number of polluters
0%
20%
40%
60%
80%
100%
0.0
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Pollution abatement costs to noncompliance penalty ratio
(a)
(b)


Policy implementation 131
0%
20%
40%
60%
80%
100%
0
20,000
40,000
60,000
80,000
100,000
120,000
140,000
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Inspection staff
0%
20%
40%
60%
80%
100%
0%
10%
20%
30%
40%
50%
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Diagnosing system
Screening system
Unit cost ratio: screening vs diagnosing technologies
(c)
(d)
Figure 6.5 (Continued)
(Figure 6.5(c)). When the pollution abatement cost-to-
 
 
noncompliance penalty 

ratio C 
 is assumed to have a lognormal distribution, the relationship between 



P 



available inspection staff and equilibrium compliance rates is similar to the situ-
ation earlier (Figure 6.5(c)).


132  Policy implementation
Screening technologies should be carefully selected. Otherwise, the screening 
system would not yield higher compliance rates than the diagnosing system. The 
low costs of a screening technology to monitor one polluting source ( rs ) is crucial 
for its better performance (Figure 6.5(d)). It should be no less than 65% cheaper 
than a diagnosing technology (r
d  ; Figure 6.5(d)). Furthermore, in terms of accu­
racy, compliance monitoring technologies for screening and diagnosing have dis­
tinctly different requirements on their Type I and II errors. Screening technologies 
should make fewer Type I errors in wrongly recognizing compliant cases into the 
high-­
risk group (K1s should be generally above 60%; Figure 6.6(a)), while diag­
nosing technologies should make fewer Type II errors in wrongly recognizing 
noncompliant cases as being compliant for them to evade penalties (K2d must be 
generally above 60%; Figure 6.6(d)). The requirements on the other two accuracy 
indicators are much more relaxed. Screening technologies should not put more 
than 80% of noncompliant cases into the low-­
risk group (K2s must be generally 
above 20%; Figure 6.6(c)). The probability of a diagnosing technology to recog­
nize compliant cases as being compliant seems to matter little (K1d; Figure 6.6(b)). 
Although this model assumes only two compliance statuses of a polluter, being 
compliant or noncompliant, polluters do differ in terms of the noncompliance 
severity. In the terminology of this model, compliance monitoring technologies 
should inherently have thresholds on whether to recognize a polluter as being 
compliant or not. The preceding accuracy indicators, especially K2s and K2d, also 
reflect such thresholds. Accordingly, screening technologies only need to catch 
those more severe noncompliant cases or with strong noncompliant signals (due 
to the relaxed requirement of K2s) while diagnosing technologies must convict 
most of these severe noncompliant polluters (K2d). These results could serve as 
the guideline for assessing and selecting screening and diagnosing technologies.
Overall, the screening system in general does show significantly better perfor­
mance than the diagnosing system to achieve higher compliance rates. Depend­
ing on initial compliance rates and available resources, compliance rates in the 
screening system may evolve into two equilibrium levels, being at nearly full 
compliance and prevalent noncompliance. At the 2015 level of inspection staff in 
China, the screening system would be able to yield nearly full compliance for the 
809,500 polluting sources as covered under the “double randomness, one publi­
cization” scheme. However, the diagnosing system that is closer to reality would 
only bring about half of those polluters under compliance.
4.5  
Resilience of screening and diagnosing systems
Campaigns or movements (yundong) are widely used in China’s governance. In 
order to achieve a highly prioritized goal within a short time, the government may 
intensively reallocate unusual amounts of human, financial or political resources 
for certain tasks. These resources are usually “borrowed” from other agencies or 
functions and thus must be “returned” after campaigns conclude. Examples include 
anticrime campaigns, especially “strike hard” (Trevaskes, 2010); anticorruption 
campaigns (Wedeman, 2005); and environmental campaigns (Jahiel, 1998; van 


Policy implementation  133
Figure 6.6  
Model simulation of equilibrium compliance rates (1−M *) in the screening 
and diagnosing systems in relation to the probabilities that (a) the screening 
technology recognizes compliance cases as being compliant (K1s), (b) the diag­
nosing technology recognizes compliance cases as being compliant (K1d), (c) the 
screening technology recognizes noncompliance cases as being noncompliant 
(K2s) and (d) the diagnosing technology recognizes noncompliance cases as 
being noncompliant (K2d)
0%
20%
40%
60%
80%
100%
40%
50%
60%
70%
80%
90%
 
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Probability (K1s) 
0%
20%
40%
60%
80%
100%
0%
20%
40%
60%
80%
100%
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Probability (K1d)
(a)
(b)


134  Policy implementation
0%
20%
40%
60%
80%
100%
10%
20%
30%
40%
50%
60%
70%
80%
90%
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Probability (K 2s)
0%
20%
40%
60%
80%
100%
50%
60%
70%
80%
90%
100%
s
e
t
a
r
 
e
c
n
a
i
l
p
m
o
c
 
m
u
i
r
b
i
l
i
u
q
E
Diagnosing system
Screening system
Probability (K2d)
Figure 6.6  
(Continued)
(d)
(c)


Policy implementation  135
Rooij, 2006). Opposite to environmental campaigns, compliance monitoring might 
also experience shocks as observed in the author’s fieldwork in China, for exam­
ple, when inspection staff in one region or for one environmental task are tempo­
rarily “borrowed” for launching campaigns in another region or for other tasks.
Campaigns can achieve rapid progress on the targeted tasks. However, when 
the temporarily available resources are retreated, such campaign-­
style compliance 
monitoring and enforcement often fail to reach sustained compliance. One nota­
ble example of a largely short-­
lived environmental enforcement campaign is the 
“midnight action” for solving the unacceptable water pollution in the Huai River in 
1997 that shut down about 5,000 small polluting factories (Bai and Shi, 2006; Liu, 
1998). Improvements were achieved in the short term with significantly reduced 
water pollutant emissions and cleaner water quality (Liu, 1998). However, pollu­
tion rebounded quickly after the campaign was over (Bai and Shi, 2006).
The compliance monitoring model as constructed in this study provides an under­
standing of the short-­
lived impacts of environmental compliance monitoring cam­
paigns and shocks in the diagnosing system. The diagnosing system has no memory, 
and its compliance rate at a given time directly corresponds to the immediately avail­
able enforcement resources (Figure 6.3). In contrast, the screening system has a mem­
ory and this feature suggests that short-­
term environmental campaigns might be more 
strategically utilized to establish the screening system for compliance monitoring and 
achieve high compliance rates. As illustrated in Figure 6.3, although compliance moni­
toring resources are kept at the same level before and after environmental campaigns, 
the equilibrium compliance rate will be fundamentally lifted from a low to a high sta­
tus. The compliance rate evolution could be explained with the compliance monitoring 
model. Before a campaign starts, the prevalent noncompliance indicates that a great 
majority of compliance monitoring resources should be spent in the diagnosing step to 
convict polluters. A small proportion of resources will be enough to screen noncompli­
ant cases for the relatively expensive diagnosing. When the environmental campaign is 
launched, with more and more noncompliant polluters being caught in noncompliance, 
their rational decisions will result in higher compliance rates. Then fewer polluting 
sources will be screened into the high-­
risk group in the following time-­
step, which 
requires less resource for diagnosing. In addition, the simulation also suggests that if 
transformed into the screening system, China might reduce the number of environ­
mental inspection staff from the current level but still maintain high compliance rates.
Environmental compliance monitoring shocks have opposite impacts as cam­
paigns. A temporary shortage of inspection staff could destabilize high equilibrium 
compliance rates back to low levels (Figure 6.3). As explained in the model construc­
tion, the compliance rate in the screening system at a time-­
step is only affected by 
that in the previous time-­
step. This short memory leads to the screening system’s 
limited resilience when facing environmental compliance monitoring shocks.
If compliance rates with a longer past contribute to compliance decisions at a 
current time-­
step, the screening system of compliance monitoring will become 
more resilient. A longer memory shows that environmental campaigns should run 
longer for elevating the compliance rate to a higher equilibrium, while temporary 


136  Policy implementation
environmental shocks would be less damaging, with the dipped compliance rate 
quickly rebounding afterward.
Environmental campaigns with temporary increases in inspection staff or 
shocks with their temporary reduction could destabilize the equilibrium rates in 
the screening system with longer-­
term impacts, while the impacts in the diag­
nosing system would be short-­
lived as seen in empirical cases. Environmental 
campaigns might be especially utilized to pull the system out of a possible non­
compliance trap. If polluters have longer memories and their current compliance 
rate is directly determined by those in the past multiple periods, the screening 
system will demonstrate more resilience against the short-­
term campaigns and 
shocks.
Note
	
1	 Adapted with permission from XU, Y. 2011. Improvements in the operation of SO2 scrub­
bers in China’s coal power plants. Environmental Science & Technology, 45, 380–385. 
Copyright (2011) American Chemical Society. Much has been revised and expanded on.
References
Arguedas, C. 2008. To comply or not to comply? Pollution standard setting under costly 
monitoring and sanctioning. Environmental & Resource Economics, 41, 155–168.
Bai, X. & Shi, P. 2006. Pollution control: In China’s Huai river basin: What lessons for sus­
tainability? Environment: Science and Policy for Sustainable Development, 48, 22–38.
Becker, G. S. 1968. Crime and punishment – economic approach. Journal of Political 
Economy, 76, 169–217.
Blackman, A. & Harrington, W. 2000. The use of economic incentives in developing coun­
tries: Lessons From international experience with industrial air pollution. The Journal of 
Environment Development, 9, 5–44.
Bonita, R., Beaglehole, R. & Kjellström, T. 2006. Basic epidemiology. Washington, DC: 
WHO.
Cao, D., Yang, J. & Ge, C. 1999. SO2 charge and tax policies in China: Experiment and 
reform. In: Environmental taxes: Recent developments in China and OECD countries. 
Paris, France: OECD.
Chenery, H. B. 1961. Comparative advantage and development policy. American Eco­
nomic Review, 51, 18–51.
Cohen, M. A. 1999. Monitoring and enforcement of environmental policy. In: Folmer, 
H. & Tietenberg, T. (eds.) The international yearbook of environmental and resource 
economics 1999/2000. Cheltenham: Edward Elgar Publishing Limited.
Dasgupta, S., Wang, H. & Wheeler, D. 1997. Surviving success: Policy reform and the 
future of industrial pollution in China. World Bank Policy Research Working Paper 
1856. Washington, DC: World Bank Group.
The Economic & Trade Commission of Jiangsu Province. 2009. Quarterly thermal effi­
ciencies of electricity generation units. Nanjing, China.
EPA. 2003. Air pollution control technology fact sheet – flue gas desulfurization. EPA-­
452/F-­
03-­
034. Washington, DC: EPA.
EPA & DOE. 2010. Electric power annual 2008. Washington, DC: EPA.


Policy implementation  137
European Commission. 2012. The monitoring and reporting regulation – general guid­
ance for installations [Online]. Available: https://ec.europa.eu/clima/sites/clima/files/
ets/monitoring/docs/gd1_guidance_installations_en.pdf.
Glaeser, E. L. 1999. An overview of crime and punishment. Cambridge, MA: Harvard 
University and NBER.
Gray, W. B. & Deily, M. E. 1996. Compliance and enforcement: Air pollution regulation 
in the US steel industry. Journal of Environmental Economics and Management, 31, 
96–111.
Guo, M. Y., Xu, Y. & Chen, Y. Q. D. 2014. Fracking and pollution: Can China rescue its 
environment in time? Environmental Science & Technology, 48, 891–892.
Heindl, P.  2012. Transaction costs and tradable permits: Empirical evidence from the 
EU emissions trading scheme. ZEW Discussion Paper No. 12-­
021 [Online]. Available: 
https://www.econstor.eu/handle/10419/56029.
Helland, E. 1998. The enforcement of pollution control laws: Inspections, violations, and 
self-­
reporting. Review of Economics and Statistics, 80, 141–153.
Henan Department of Environmental Protection. 2010. Information on the collection of 
SO2 effluent discharge fee from province-­
regulated coal-­
fired power plants in the first 
quarter of 2010. Zhenzhou, China: Henan Department of Environmental Protection.
Henan Development and Reform Commission & Henan Environmental Protection Bureau. 
2007. A notice to transfer NDRC and SEPA’s policy: Management measures on desulfur­
ized electricity price and the operation of desulfurization facilities of coal-­
fired power 
generators (on trial). Zhenzhou, China: NDRC and SEPA.
IEA. 2009. Cleaner coal in China. Paris, France: IEA.
Jahiel, A. R. 1998. The organization of environmental protection in China. The China 
Quarterly, 156, 757–787.
Jiangsu Department of Environmental Protection. 2007–2009. Monthly report on the oper­
ation of SO2 scrubbers at coal-­
fired power plants. Nanjing, China: Jiangsu Department 
of Environmental Protection.
Jiangsu Department of Environmental Protection. 2008. On strengthening the collection of 
effluent discharge fee from coal-­
fired power plants. Nanjing, China: Jiangsu Department 
of Environmental Protection.
Jin, Y. N., Andersson, H. & Zhang, S. Q. 2016. Air pollution control policies in China: 
A retrospective and prospects. International Journal of Environmental Research and 
Public Health, 13.
Kay, S., Zhao, B. & Sui, D. 2015. Can social media clear the air? A case study of the air 
pollution problem in Chinese cities. Professional Geographer, 67, 351–363.
Kitchin, R. 2014. The data revolution. Thousand Oaks, CA: Sage Publications.
Konrad, K. A., Lohse, T. & Qari, S. 2017. Compliance with endogenous audit probabilities. 
Scandinavian Journal of Economics, 119, 821–850.
Kruger, J. & Egenhofer, C. 2006. Confidence through compliance in emissions trading 
markets. Sustainable Development Law & Policy, 6, 2–13, 63–64.
Levitt, S. D. 2004. Understanding why crime fell in the 1990s: Four factors that explain the 
decline and six that do not. Journal of Economic Perspectives, 18, 163–190.
Li, C., Zhang, Q., Krotkov, N. A., Streets, D. G., He, K. B., Tsay, S. C. & Gleason, J. F. 
2010. Recent large reduction in sulfur dioxide emissions from Chinese power plants 
observed by the ozone monitoring instrument. Geophysical Research Letters, 37.
Lin, J. Y., Cai, F. & Li, Z. 2003. The China miracle: Development strategy and economic 
reform. Sha Tin: Published for the Hong Kong Centre for Economic Research and the 
International Center for Economic Growth by the Chinese University Press.


138  Policy implementation
Liu, H. 1998. An overview of water pollution prevention and control in the Huai River val­
ley. Enviornmental Management in China, 5–8.
Liu, Z., Guan, D. B., Wei, W., Davis, S. J., Ciais, P., Bai, J., Peng, S. S., Zhang, Q., Hubacek, 
K., Marland, G., Andres, R. J., Crawford-­
Brown, D., Lin, J. T., Zhao, H. Y., Hong, C. 
P., Boden, T. A., Feng, K. S., Peters, G. P., Xi, F. M., Liu, J. G., Li, Y., Zhao, Y., Zeng, 
N. & He, K. B. 2015. Reduced carbon emission estimates from fossil fuel combustion 
and cement production in China. Nature, 524, 335.
Lu, X., Dudek, D. J., Qin, H., Zhang, J., Lin, H., Yang, Z. & Wang, Y. 2006. Survey on the 
capacity of environmental administrative enforcement in China. Research of Environ­
mental Sciences, 19, 1–11.
Lu, Z., Streets, D. G., Zhang, Q., Wang, S., Carmichael, G. R., Cheng, Y. F., Wei, C., Chin, 
M., Diehl, T. & Tan, Q. 2010. Sulfur dioxide emissions in China and sulfur trends in East 
Asia since 2000. Atmospheric Chemistry and Physics, 10, 6311–6331.
Lu, Z., Zhang, Q. & Streets, D. G. 2011. Sulfur dioxide and primary carbonaceous aero­
sol emissions in China and India, 1996–2010. Atmospheric Chemistry and Physics, 11, 
9839–9864.
McAllister, L. K., Van Rooij, B. & Kagan, R. A. 2010. Reorienting regulation: Pollution 
enforcement in industrializing countries. Law & Policy, 32, 1–13.
MEP. 2013. The 12th five-­
year plan on capacity building of environmental regulations. 
Beijing, China: MEP.
MEP & AQSIQ. 2011. Emission standard of air pollutants for thermal power plants. Bei­
jing, China: MEP, AQSIQ.
MEP & NDRC. 2008. Announcement on punishing coal-­
fired power plants with abnormal 
operation of SO2 scrubbers in 2007. Beijing, China: MEP, NDRC.
Ministry of Ecology and Environment. 2018. The comprehensive implementation of ‘dou­
ble randomnesses, one publicization’ for environmental inspection. Beijing, China: Min­
istry of Ecology and Environment.
Ministry of Environmental Protection. 2002–2016. Annual statistical report on the envi­
ronment in China. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2006–2009. Annual statistical report on the envi­
ronment in China. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2008. Statement to penalize coal-­
fired power plants 
for the abnormal operation of their SO2 scrubbers in 2007. Beijing, China: Ministry of 
Environmental Protection.
Ministry of Environmental Protection. 2009a. 2008 assessment report on provincial major 
pollutant emissions. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2009b. Information on the mitigation of major pol­
lutants in 2008. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2009c. Statement to penalize five coal-­
fired power 
plants for the abnormal operation of their SO2 scrubbers in 2008. Beijing, China: Min­
istry of Environmental Protection.
Ministry of Environmental Protection. 2010a. China’s capacities of water treatment plants 
and SO2 scrubbers at coal-­
fired power plants. Beijing, China: Ministry of Environmen­
tal Protection.
Ministry of Environmental Protection. 2010b. Guideline on best available technologies of 
pollution prevention and control for coal-­
fired power plant industry (on trial). Beijing, 
China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2016a. Comprehensive plan on ecological and envi­
ronmental big data construction. Beijing, China: Ministry of Environmental Protection.


Policy implementation  139
Ministry of Environmental Protection. 2016b. final accounts of ministry of environmental 
protection. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection, National Statistics Bureau  & Ministry of Agri­
culture. 2010. Public report on the first national census of polluting sources. Beijing, 
China: Ministry of Environmental Protection.
National Bureau of Statistics. 2008. China energy statistical yearbook. Beijing, China: 
China Statistics Press.
NDRC. 2004. Technical code for designing flue gas desulfurization plants of fossil fuel 
power plants. DL/T5196-­
2004. Beijing, China: NDRC.
NDRC & SEPA. 2007a. The 11th five-­
year plan on SO2 control in existing coal-­
fired power 
plants. Beijing, China: State Environmental Protection Administration, NDRC.
NDRC & SEPA. 2007b. Management measures on desulfurized electricity price permium 
and the operation of desulfurization facilities in coal-­
fired power generators (on trial). 
Beijing, China: State Environmental Protection Administration, NDRC.
OECD. 2006. Environmental compliance and enforcement in China – an assessment of 
current practices and ways forward. Paris, France: OECD.
Ohlin, B. 1967. Interregional and international trade. Cambridge, MA: Harvard Univer­
sity Press.
Osterman, G., Eldering, A., Avis, C., Chafin, B., O’Dell, C., Frankenberg, C., Fisher, B., 
Mandrake, L., Wunch, D., Granat, R. & Crisp, D. 2018. Orbiting carbon observatory – 2 
(OCO-­
2) data product user’s guide, operational L1 and L2 data versions 8 and lite file 
version 9. Pasadena, CA: Jet Propulsion Laboratory, California Institute of Technology.
Pan, L., Wang, Z. & Wang, Z. 2005. Present status and countermeasure suggestion for 
thermal power plants CEMS in China. Research of Environmental Sciences, 18, 42–45.
Polinsky, A. M. & Shavell, S. 2000. The economic theory of public enforcement of law. 
Journal of Economic Literature, 38, 45–76.
Raufer, R. & Li, S. Y. 2009. Emissions trading in China: A conceptual ‘leapfrog’ approach? 
Energy, 34, 904–912.
Ricardo, D. 1817. On the principles of political economy and taxation. London: J. Murray.
Russell, C. S. & Vaughan, W. J. 2003. The choice of pollution control policy instruments 
in developing countries: Arguments, evidence and suggestions. In: Folmer, H. & Tieten­
berg, T. (eds.) International yearbook of environmental and resource economics. Chel­
tenham: Edward Elgar.
SEPA. 2006. Guidelines on calculating SO2 emission quotas. Beijing, China: State Envi­
ronmental Protection Administration.
SEPA. 2007. Verification of major pollutants emission reduction in the 11th five-­
year 
period (on trial). Beijing, China: State Environmental Protection Administration.
SERC. 2009. New credit to CEMSs at SO2 scrubbers in Jiangsu province. Nanjing, China: 
SERC.
SERC, NDRC, National Energy Bureau & MEP. 2009. Report on electric industry’s energy 
conservation and pollutant mitigation in 2008. Beijing, China: SERC, NDRC, MEP.
Shimshack, J. P.  2014. The economics of environmental monitoring and enforcement. 
Annual Review of Resource Economics, 6, 339–360.
Snyder, E. G., Watkins, T. H., Solomon, P. A., Thoma, E. D., Williams, R. W., Hagler, G. S. 
W., Shelow, D., Hindin, D. A., Kilaru, V. J. & Preuss, P. W. 2013. The changing paradigm 
of air pollution monitoring. Environmental Science & Technology, 47, 11369–11377.
State Commission Office for Public Sector Reform. 2018. Regulations on the authorities, 
organization and personnel of the ministry of ecology and environment. Beijing, China: 
State Commission Office.


140  Policy implementation
State Council. 2007a. Notice on distributing composite working plan on energy conserva­
tion and pollutant emission reduction. Beijing, China: State Council.
State Council. 2007b. Notice on distributing implementation plans and methods of statis­
tics, monitoring and assessment on energy conservation and pollutant emission reduc­
tion. Beijing, China: State Council.
State Council. 2015. Action outline for promoting big data development. Beijing, China: 
State Council.
State Council. 2019. Advice on the comprehensive implementation of joint ‘double ran­
domnesses, one publicization’. In: Market regulation. Beijing, China: State Council.
State Development Planning Commission, Ministry of Finance, SEPA & State Economic 
and Trade Commission. 2003. Measures for the administration of the charging rates for 
pollutant discharge fees. Decree No. 31 [Online]. Available: http://www.lawinfochina.
com/display.aspx?id=2705&lib=law&EncodingName=big5.
State Electricity Regulation Commission (Nanjing Office). 2009. Information on the oper­
ation of SO2 scrubbers in coal-­
fired power plants in Jiangsu Province. Nanjing, China: 
SERC.
State Environmental Protection Administration  & General Administration of Quality 
Supervision Inspection and Quarantine. 2003. Emission standard of air pollutants for 
thermal power plants. GB 13223-­
2003. Beijing, China: State Environmental Protection 
Administration.
Steinfeld, E. S., Lester, R. K. & Cunningham, E. A. 2009. Greener plants, grayer skies? 
A report from the front lines of China’s energy sector. Energy Policy, 37, 1809–1824.
Stevens, M.  & Ochab, B. 2010. Participatory noise pollution monitoring using mobile 
phones. Information Polity: The International Journal of Government & Democracy in 
the Information Age, 15, 51–71.
Stranlund, J. K. & Chavez, C. A. 2000. Effective enforcement of a transferable emissions 
permit system with a self-­
reporting requirement. Journal of Regulatory Economics, 18, 
113–131.
Streets, D. G., Canty, T., Carmichael, G. R., De Foy, B., Dickerson, R. R., Duncan, B. N., 
Edwards, D. P., Haynes, J. A., Henze, D. K., Houyoux, M. R., Jacobi, D. J., Krotkov, 
N. A., Lamsal, L. N., Liu, Y., Lu, Z. F., Martini, R. V., Pfister, G. G., Pinder, R. W., 
Salawitch, R. J. & Wechti, K. J. 2013. Emissions estimation from satellite retrievals: 
A review of current capability. Atmospheric Environment, 77, 1011–1042.
Trevaskes, S. 2010. Policing serious crime in China: From ‘strike hard’ to ‘kill fewer’. 
London and New York: Routledge.
Tsinghua University. 2010. A study on the management system of environmental pollution 
data collection in China. Beijing, China: Tsinghua University Press.
The U.S. Congress. 1990. Clean air act amendments 1990. Washington, DC: The U.S. Congress.
The U.S. EPA. 2007. CEMS cost model. Washington, DC [Online]. Available: www.epa.
gov/ttn/emc/cem/cems.xls.
Van Rooij, B. 2006. Implementation of Chinese environmental law: Regular enforcement 
and political campaigns. Development and Change, 37, 57–74.
Vossler, C. A. & Gilpatric, S. M. 2018. Endogenous audits, uncertainty, and taxpayer assis­
tance services: Theory and experiments. Journal of Public Economics, 165, 217–229.
Wall, M. 2014. NASA launches satellite to monitor carbon dioxide, July 2 [Online]. Avail­
able: www.space.com/26403-­
nasa-­
oco2-­
carbon-­
dioxide-­
satellite-­
launch.html.
Wang, A. & Brauer, M. 2014. Review of next generation air monitors for air pollution [Online]. 
Available: https://www.semanticscholar.org/paper/Review-­
of-­
Next-­
Generation-­
Air-
­
Monitors-­
for-­
Air-­
Wang-­
Brauer/ca660dbe9fa087b2bea4080064c251914155c177.


Policy implementation  141
Wedeman, A. 2005. Anticorruption campaigns and the intensification of corruption in 
China. Journal of Contemporary China, 14, 93–116.
Wu, J. & Qian, Y. 2007. The construction of FGD at Beilun power plant on its 5 × 600 MWe 
units. Electrical Equipment, 8, 105–107.
Xu, Y. 2011a. Improvements in the operation of SO2 scrubbers in China’s coal-­
fired power 
plants. Environmental Science & Technology, 45, 380–385.
Xu, Y. 2011b. The use of a goal for SO2 mitigation planning and management in China’s 
11th five-­
year plan. Journal of Environmental Planning and Management, 54, 769–783.
Zhang, Q., Streets, D. G., Carmichael, G. R., He, K. B., Huo, H., Kannari, A., Klimont, 
Z., Park, I. S., Reddy, S., Fu, J. S., Chen, D., Duan, L., Lei, Y., Wang, L. T. & Yao, Z. L. 
2009. Asian emissions in 2006 for the NASA INTEX-­
B mission. Atmospheric Chemis­
try and Physics, 9, 5131–5153.
Zhang, Q., Streets, D. G., He, K., Wang, Y., Richter, A., Burrows, J. P., Uno, I., Jang, 
C. J., Chen, D., Yao, Z. & Lei, Y. 2007. NOx emission trends for China, 1995–2004: 
The view from the ground and the view from space. Journal of Geophysical Research-­
Atmospheres, 112.
Zhejiang Bureau of Quality and Technical Supervision. 2007. The quota & calculation 
method of coal consumption for generating station. Hangzhou, China: Zhejiang Bureau 
of Quality and Technical Supervision.


­
Appendix
Modeling environmental compliance 
monitoring systems
Key parameters
This model’s primary output is the compliance rate of polluters under compli-
ance monitoring: M t is the noncompliance rate at the end of the time-step 
 
t, 
while 1-M
 
t is the corresponding compliance rate. For simulating the evolution 
of compliance rates over time, the model is designed to follow time-steps. In
 
 
each time-step, 
 
enforcement activities are first conducted to comprise compli-
ance monitoring and penalty on noncompliance. The compliance rate at the end 
of the previous time-step 
 
could affect the subsequent performance of environ-
mental compliance monitoring, that is, the probability of catching noncompli-
ance. This probability is assumed to be commonly available information for all 
polluters. Based on the expected penalty and compliance costs, polluters make 
compliance decisions to yield an overall compliance rate at the end of the cur-
rent time-step.
As listed in Table 6.3, the model has a series of input parameters, whose values 
are given exogenously. They fall into several major categories: (1) environmental 
compliance monitoring system, including initial noncompliance rate (M 0), the 
total available resources for compliance monitoring (Rt) and the number of pol-
luting sources (N); (2) the ratio between pollution abatement costs and penalty on 

noncompliance C 
 as well as its distribution (

Φ

( )
• ); (3) compliance monitoring 




technologies, including 
P 
required resources for monitoring one polluting source 
(r), the probability that one technology recognizes compliant cases as being com-
pliant (K1) and the probability that one technology recognizes noncompliant cases 
as being noncompliant (K2). Screening and diagnosing technologies are further 
distinguished with subscripts s and d, respectively. The four parameters (K1s, K1d, 
K2s and K2d) are assumed to be specific for a given compliance monitoring technol-
ogy and do not change over time and cases. Compliance-monitor
 
ing technologies 
and systems could make two types of errors in identifying noncompliance (Polin-
sky and Shavell, 2000; Bonita et al., 2006). We assume H0: a polluter is under 
environmental compliance. A Type I error indicates that a polluting firm is under 
compliance, but the environmental compliance monitoring wrongly identifies the 
case as noncompliance to mistakenly punish it. A Type II error refers to the situ-
ation that although a polluter is not complying, the system wrongly recognizes it 
as being compliant. Accordingly, the illegal polluter walks away without penalty. 


Policy implementation 143
Both errors consequently lower the deterrence effect, which might lead to lower 
compliance rates.
The diagnosing system
In this diagnosing-only system, the probability 
 
of noncompliant polluting sources 
Rt
that are rightfully punished is: 
d
1
´
´K
r
N
2d, while the probability of compliant 
polluting sources that are mistakenly punished is 
d
Rt
d
1
×
×(
)
1−K
r
N
1d . A polluter 
Rt
d
Rt
will choose compliance when C
P
+
×
d
1
×
×(
)
1−
<
K
P ×
×
d
1
d
×K
r
N
1
2
r
N
d, or 
C
Rt
d
d
<
×
d
1 ×
+
(
)
K
K
2
1
d
d −1 . Because all resources are devoted to diagnosing, 
P
r
N
d
R
R
t =
t
d. One polluting source could be diagnosed more than once to potentially 
incur a penalty every time that it is caught noncompliance. Corresponding to 
available enforcement resources, the noncompliance rate will be
Rt
M t
1
=
−
1
Φ(
(
×
× K
K
r
N
2
1
d
d
+
−1)) 
Equation 6.1
d
Because M t is not related to M t-1, the noncompliance rate under the diagnosing 
system will not show dynamic evolution over time when other factors remain 
unchanged.
The screening system
When resources are inadequate, some polluting sources may be neither screened 
nor diagnosed, while the optimal allocation of resources will make sure that 
all screened-out polluting 
 
sources in the high- 
risk group are diagnosed and no 
available resource is wasted. Due to the existence of Type I and II errors, each 
group contains compliant and noncompliant sources. The high- 
risk group in the 
Rt
time-step t will comprise 
s ×
×
M
K
t−1
2s noncompliant polluting sources and 
Rt
r
s
s ×
−
(
)
1
1
M
K
t−1 ×
−
(
)
1s  compliant polluting sources. Accordingly, the noncom-
r
pliance rate in the high-
s
M
K
t−1
risk group is 
 
M t
×
h =
2s
t−
−
1
t
1
.
M
K
×
+(
)
1
1
−
×
M
K
(
)
−
The low- 
risk group will contain all remaining polluting 
2s
sources, including 
1s
those 
Rt
Rt
screened out and those not screened, 
N −
×
s
M
K
t−
−
1 ×
−
s ×
−M
K
t
1
2s
(
)
1
1
×
−
(
)
1s , 
r
s
r
Rt
Rt
Rt
s
or (
)
N −
+
s
s ×
×
M
K
t−
−
1
(
)
1
1
−
+
s ×
−
(
)
M
K
t
1
2s
×
1s. The number of noncom-
r
s
r
s
r
s Rt
pliant polluting sources is N
M
×
−
t−
−
1
1
s ×
×
M
K
t
2s. Then the noncompliance 
r
s
Rt
N
M
×
−
t−
−
1
1
s ×
×
M
K
t
r
2s
rate in the low-risk group is 
 
M t =
s
l
Rt
N
s
M
K
t
1
Rt
.
−
×
−
−
×
−
s
2s
×
−
(
)
1
M t
1 ×(1−K
r
s
r
1s)
s
­


144 Policy implementation
After diagnosing, the number of noncompliant polluting sources that are 
Rt
rightfully punished is 
d ´
´
M
K
t
r
h
2d. The probability of noncompliant pollut-
t
d
Rd ×
×
M
K
t
r
h
2d
d
Rt
M t
ing sources that are rightfully punished is 
d
1
=
×
×
×
h
K
N
M
×
t−
−
1
1
r
N
M t
2d. 
The number of compliant polluting sources that are mistakenly pun
d
-
ished is 
Rt
d
M
K
r ´ -
(
)
1
1
t
h ´ -
(
)
1d C : \ wspath\ WS5551\ Math_Preference\ Equat
tion\ pref\ Euclid.eqp
d
, and the corresponding probability is 
Rt
d ×
−
(
)
1
1
M
K
t
r
h ×
−
(
)
1d
Rt
1
M t
d
=
×
d
1−
×
h ×
−
(1
K ). We assume that the two 
N
M
×
−
(
)
1
t−
−
1
1
r
N
1−M t
1d
d
probabilities are known to all polluting sources for their following compliance 
decisions.
Rt
−
Thus, the expected compliance cost is C
P
d
1
1
M t
+
×
×
×
h ×
−
(
)
1
K
r
N
1−M t−1
1d , 
d
Rt
1
M t
while the expected penalty on noncompliance is P ×
×
d
×
×
h
K . 
r
N
d
d
M t−1
2  
For a decision of compliance, the former should be lower than the latter: 
Rt
1
1−M t
Rt
1
M t
t
d
h
1
C
P
+
×
×
×
×
−
(
)
1
K
P
d
h
C
R
1d <
×
×
×
×K
<
×
1
M t−
−
1
r
N
t
1
2d , or 
d
×
r
N
d
−
d
M
P
r
N
d
K
K
2
2
s
d
×
−(
)
1
1
−
×
K
K
1
1
s
d
(
)
−
r
N
M
K
t−
−
1 ×
+
s
1
1
M
K
t
1
.
2
(
)
−
×( −
1s)
Rt
screened 
min  is further defined as a threshold when all polluting sources have just been 
(R
N
t
s =
×r
s ), all polluting sources in the high-risk group are 
 
diagnosed 
Rt
t
(Rt =
×
(
(
s
M
K
t−
−
1
Rs
d
×
+
s
1
1
M
K
t
2
×
−
1)
(
×
−
1s
d
))×r ) and all resources are uti-
r
s
r
lized 
s
R
R
t =
+
t
s
Rt . Then R
N
t
t−
−
1
t
1
d
min =
×(
(
r
M
s +
×K
M
2s +
−
(
)
1
1
×
−
(
)
K
r
1s
d
)
)
×
.
When R
R
t
t
Rt
£
Rt =
min , 
d
. The com-
1
(
+1)
r
d ×
×
(
(
M
K
t−
−
1
+
−
1
2
×
−
r
s
1
1
M
K
t
)
(
1s))
s
C
Rt
K
K
2
2
s
d
×
−(
)
1
1
−
×
K
K
(
)
pliance condition is 
1
1
−
<
×
s
d
r
M t
.
P
N
(
(
s +
×
−
−
1
K
M
s
1
1
t
1
2 +
−
(
)×
−
(
K1s
d
))×r )
Additional compliance monitoring resources beyond Rt
min will be devoted 
to diagnosing those polluting sources in the high- 
risk group. These sources 
could be diagnosed and punished once or multiple times. In this situation, 
R
R
t
t
d =
−R
R
t
t
s =
−N
r
× s.
Then corresponding to available enforcement resources, the noncompliance 
rate at the end of time-step 
 
t will be

t

If R
R
t
t

−
£
t
×
−
, 
R
K
K
(
)
1
1
−
×
K
K
(
)

min M =
−
1
Φ
×
2
2
s
d
1
1


s
d


N
(
(
r
M
+
×
t−
−
1
K
M
+
−
(
)
1
1
t
1
; Equation 6.2

s
2s
×( −
×
K
r 

1s
d
))
)
K
K
M
K
s
d
t
s
×
−
×
+
−
2
2
1
2
(
(
C
P
Rd
t
d
<
×
×
1


Policy implementation 145

If 
t

t
t , 
t

R
N
−
×r
K
K
×
−(
)
1
1
−
×
K
K
(
)
−

R
R
>
min M =
−
1
Φ

s ×
2
2
s
d
1
1
s
d

.  
Equation 6.3




N
r
×
t
d
M
K
−
−
1 ×
+
s
(
)
M t
1
2
1−
×
−
(
)
1
K1 

s
Rt  will be greater if the noncompliance rate at the end of time-step 
 
t-1, M t-1
is higher or screening and diagnosing are more resource-
min
, 
 
intensive with greater 
r
s and r
individual 
d. Given a certain amount of total emissions under regulation, smaller 
polluting sources will result in a greater number of polluting sources, 
N, for compliance monitoring and thus higher demand for resources. Because the 
noncompliance rate, M t, changes over time, Rt  will change accordingly.
More accurate compliance monitoring technologies (
min
K
K
1
1
s
d
,
,K
K
2
2
s
d
,
® 1) with 
lower costs for an average polluting source (r r
,
® 0) tend to induce higher com-
pliance rates. Various factors could affect the availability 
s
d
of enforcement resources 
Rt
per polluting source (
). The economy of scale in compliance monitoring could 
N
have two folds. On one hand, larger polluting sources could lead to an internal 
economy of scale because the required enforcement resources are more related to 
the number of sources. More enforcement resources, larger polluting sources and 
a smaller amount of total emissions will increase the resource availability indica-
tor. Even if with the screening step or effective compliance monitoring strategy, 
the probability of catching enough noncompliance cannot be enhanced to a high 
enough level without sufficient enforcement resources. On the other hand, the 
geographical proximity of polluting sources could provide an external economy 
of scale. The sources could then be equivalently bundled and reduce the compli-
ance monitoring costs for one polluting source.
Corresponding to their required features, screening technologies are less accu-
rate but also less expensive than diagnosing technologies. They must have such 
trade- 
offs to fit in the expected complementary roles. If one technology were both 
cheaper and more accurate than the other, the latter technology would be entirely 
replaced by the former.
Input parameters in China’s empirical case
In order to empirically illustrate and analyze the model, the input parameters will 
adopt empirical values from the Chinese context. A current scenario and the range 
of parameters are defined with the best available empirical data in China’s current 
situation. They are briefly summarized in Table 6.3, and this subsection provides 
a more detailed explanation.
Available resources for compliance monitoring (Rt) are a key input parameter 
that this model focuses on. For simplicity, compliance-monitoring 
 
resources (Rt) 
and costs of screening and diagnosing technologies (r
s and r
d) are counted as the 
number of environmental inspection staff. China has been gradually increasing 
governmental employees for environmental inspection. The resource availabil-
ity still faces constraints, but it does not fall into the situation of extreme scar-
city. From 2001 to 2015, staff for environmental inspection grew from 37,934 to 
66,379 (Ministry of Environmental Protection, 2002–2016). More important, with 
­


146 Policy implementation
the full establishment of regional supervisory centers/bureaus in 2008 by the then 
Ministry of Environmental Protection, the central government has significantly 
strengthened its capacity of environmental inspection, accounting for 0.48% (294 
employees) of inspection staff at all four levels in 2009 and 0.82% (542 employ-
ees) in 2015, up from 0.07% in 2008 (41 employees; Ministry of Environmental 
Protection, 2002–2016). Six regional Supervision Bureaus were allowed to have, 
in total, 240 formal employees for taking charge of supervision tasks within their 
jurisdictions (State Commission Office for Public Sector Reform, 2018). Not all 
staff employed in the inspection section are environmental inspectors, for exam-
ple, to play supporting roles such as office work. In 2017, China had 46,800 envi-
ronmental inspectors in the databases for “double randomness, one publicization” 
(Ministry of Ecology and Environment, 2018). The closest year with available 
data on inspection staff was 2015. Accordingly, about 70.5% of inspection staff
 
were environmental inspectors. The empirical model simulation adopts this ratio 
to examine the impacts of resource availability on environmental compliance 
rates. The current scenario thus has 66,379 inspection staff, or 46,800 environ-
mental inspectors. If not specified, they will remain unchanged over time.
The number of polluting sources (N) was been briefly described in Section 4.2. 
The current scenario takes the intermediate number, 809,500 polluting sources as 
targeted in 2017 under the “double randomness, one publicization” scheme.
Pollution abatement costs and the associated penalty for noncompliance range 
across sectors, technologies and severity of noncompliance. The ratio between 

compliance costs and penalty C 
 is a key variable in this compliance monitoring 



P 



model. In 2007, in order to tackle the long-term 
 
problem of weak environmental 
policy enforcement, China not only subsidized those coal-fired 
 
power plants to 
normally operate their SO2 scrubbers but, more important, also issued a penalty, 
being five times of the subsidy/costs on a per- 
kilowatt-hour 
 
basis (Xu, 2011a; 
NDRC and SEPA, 2007b). In dealing with potential noncompliance on water pol-
lution and withdrawal, however, China’s penalty was barely able to catch up with 
the pollution abatement costs (Guo et al., 2014). In the current scenario, the cost/
penalty ratio is assumed to be 2/3. Furthermore, pollution abatement costs are not 
identical across polluting firms due to, for example, economy of scale, the sulfur 
content of coal and whether the pollution removal facility is a retrofit or built 
together with the main equipment. In compiling China’s SO2 emission inventory, 
Lu et al. (2011) assumed that the sulfur content had a normal distribution. The 
current scenario follows, due to the key influence of sulfur contents on SO2 abate-

ment costs, to assume that the cost/penalty ratio C 
 has a normal distribution 


) among the polluting sources.

P 


(Φ( )
•

The costs of screening and diagnosing technologies are accounted as the 
required number of inspectors in a year per environmental observation, either 
screening or diagnosing inspection (inspector-year 
 
per observation, being noted 
as r
randomness, one publicization” 
s and r
d, respectively). China has comprehensively established the “double 
method for governmental, including environ-
mental and other, inspections on firms (State Council, 2019). For environmental 


Policy implementation  147
inspections, the method had been well established in 2017 (Ministry of Ecology 
and Environment, 2018). Under this method, polluting firms and environmental 
inspectors will both be randomly selected from databases, while the information 
will be publicized to the public. In 2017, 809,500 polluting firms and 46,800 envi­
ronmental inspectors were included in the databases, while 632,600 environmen­
tal inspections were conducted (Ministry of Ecology and Environment, 2018). 
Accordingly, 27 inspections were conducted by an average inspector in 2017. 
According to the author’s earlier fieldwork in China (Guo et al., 2014; Xu, 2011a), 
one inspection generally involves two inspectors. Thus, the cost of environmental 
inspection or diagnosing technology (r
d) was 0.074 inspector-­
year per inspection. 
It is adopted in the current scenario.
Different screening and diagnosing technologies have different cost structures. 
For example, a sophisticated satellite-­
based technology has very high initial 
capital costs, but its marginal costs of monitoring one more pixel are negligi­
ble. For example, OCO-­
2 cost US$465 million to set up, but with more than 
100,000 measurements of column CO2 concentrations each day (Osterman et al., 
2018), each measurement since its launch in July 2014 cost merely about US$2 
to US$3 per measurement, considering neither operation and maintenance costs 
that will raise the unit cost nor expected longer lifetime that will reduce the 
unit cost. According to the author’s fieldwork in China’s coal-­
fired power plants, 
continuous emissions monitoring system (CEMS) costs about 500,000 RMB/set 
around 2010. China has been publishing hourly data from CEMSs in key pollut­
ing sources. With an expected lifetime of approximately 5 to 10 years, the unit 
cost would also be about US$1 to US$2 per published data point. Screening often 
requires multiple observations. OCO-­
2 has a 16-­
day ground-­
track repeat cycle 
to result in about 23 repeated observations per year for one pixel, or at a cost of 
roughly US$50 per year. CEMSs in China could provide more than 8,000 hourly 
observations per year and have an annual cost of about US$8,000 to US$16,000. 
Accordingly, the costs of an average screening technology are assumed to be 
in the range of several hundred U.S. dollars per year for one polluting source. 
In contrast, compliance monitoring by environmental inspectors is cheaper to 
set up but more expensive to operate. For example, China in 2015 at the cen­
tral level had 542 employees for environmental inspections (Figure 3.1), with 
a total cost of 63.5 million RMB (~US$10.2 million in 2015 exchange rate, or 
US$18,800/person-­
year; Ministry of Environmental Protection, 2016b). Accord­
ingly, the average cost of one inspection was about 0.074 inspector-­
year/inspec­
tion / 70.5% × US$18,800/person-­
year, or US$2,000/inspection. In the current 
scenario, the unit cost of a screening technology (rs) is then assumed to be one 
order of magnitude cheaper than that of screening technology, or equivalently 
0.0074 inspector-­
year per screening round.
The compliance monitoring accuracy of one technology is hard to exactly 
measure, because only data on observed compliance and noncompliance are avail­
able but not those on absolute truth. Furthermore, the dichotomy of compliance 
and noncompliance does not measure the severity of noncompliance, while more 
severe cases, due to their stronger signal-­
to-­
noise ratios, tend to be easier to catch. 


148  Policy implementation
In theory, the screening strategy would only work when the noncompliance rate 
in the high-­
risk group is higher than that in the low-­
risk group. The more different 
their noncompliance rates between these groups gap are, the better the screening 
strategy will be. In the current scenario, K1s, K1d, K2s and K2d are assumed to be 
90%, 99%, 70% and 90%, respectively.


7	

Environmental technology 
and industry1
1  
Goal-­
centered SO2 mitigation path
Besides other critical measures, pollution mitigation often involves facilities 
such as those installed in coal-­
fired power plants to remove sulfur oxide (SO2), 
nitrogen oxide (NOx), particles, mercury and carbon dioxide (CO2), together with 
renewable-­
energy facilities for reducing coal consumption such as wind turbines 
and solar panels and hybrid and electric vehicles. Two major factors determine 
how rapidly a country could utilize these facilities for pollution mitigation. First, 
there must be a strong demand for their rapid deployment and normal operation, as 
examined in detail in Chapters 5 and 6. Second, if the demand is put in place, 
enough supply capacity should be established to meet the demand. A develop­
ing country could take the latecomer’s advantage to utilize the supply capacity 
in developed countries. However, because of China’s sheer size, the rest of the 
world might not be able to accommodate its huge demand. With constrained sup­
ply capacity but significantly greater demand, the international price of pollution 
control facilities could rise sharply, and this would discourage their utilization and 
slow the pollution mitigation process. Rapid pollution mitigation in China relies 
greatly on the rapid establishment of a domestic industry.
SO2 mitigation achieved rapid progress over the past two decades from low 
starting positions. On the supply side, in the late 1990s, China had few domestic 
firms and barely any commercialized technologies. The Chinese markets were 
dominated by foreign firms and foreign technologies. After a decade, a large num­
ber of firms entered the market to meet the newly emerged huge demand for SO2 
scrubbers to even drive down prices substantially.
As an illustration of the differences between goal-­
centered and rule-­
based gov­
ernance, the progressive paths in China and the United States have been dra­
matically different in reaching the wide deployment of SO2 scrubbers in coal-­
fired 
power plants and their normal operation of high SO2 removal rates (Figure 7.1). 
From the very beginning, the normal operation of SO2 scrubbers in the United 
States with rule-­
based governance has been achieved while the progress went 
mainly through the deployment dimension. In contrast, China deployed SO2 
scrubbers with poor operation in the early stage and then proceeded simultane­
ously in the dimensions of deployment and operation until the technical limits of 


150  Environmental technology and industry
SO2 removal rates were roughly reached. Accordingly, the requirements on the 
quality of SO2 scrubbers were initially low in the Chinese market and became 
increasingly higher only later, while in the U.S. market, quality was important 
from the beginning.
In China, under goal-­
centered governance, at the early stage of deployment with 
few SO2 scrubbers and incapable policy implementation, more SO2 mitigation 
would be achieved if the focus were on further deployment rather than on opera­
tional improvement. With more and more SO2 scrubbers in place, any improve­
ment in the operation of the growing stock would lead to a greater reduction in 
SO2 emissions. For achieving their SO2 mitigation goals, the rational choice of 
the Chinese central and local governments led to a path in which initial progress 
was made mainly in deploying more SO2 scrubbers, and it was only afterward that 
their level of operation caught up.
Implementing policies on the deployment and normal operation of SO2 scrub­
bers require different amounts of resources for compliance monitoring. On one 
hand, the compliance monitoring on the physical existence of SO2 scrubbers is 
straightforward and the huge sizes – for example, an absorbing tower is generally 
several meters in diameter and tens of meters high – make them easily visible. The 
one-­
by-­
one inspection indicates that the corresponding compliance monitoring 
0
100
200
300
400
500
600
0%
20%
40%
60%
80%
100%
)
W
M
0
0
0
,
1
(
t
n
e
m
y
o
l
p
e
d
e
v
i
t
a
l
u
m
u
C
SO2 removal rate
China 
(2006–2008, 2010)
United States
(1973–2010)
Figure 7.1  
The progressive paths on the deployment and operation of SO2 scrubbers in 
China and the United States
Source: Lefohn et al. (1999); Xu (2011b); Ministry of Environmental Protection (2011b, 2008–2012); 
EIA (1986–2006, 2007–2011); Xu (2013).


Environmental technology and industry  151
costs for each SO2 scrubber do not greatly differ, regardless of how many have 
been deployed. On the other hand, the compliance monitoring on the installation 
is just a onetime event, but when in operation they demand significantly more 
resources on a day-­
by-­
day basis. A well-­
functioning environmental compliance 
monitoring system has significant initial costs of establishment. A  significant 
proportion of additional costs for monitoring one more SO2 scrubber are largely 
borne by the polluting firms because they are responsible for installing their own 
monitoring equipment. For policy enforcers, the compliance monitoring costs 
have a great economy of scale and they increase relatively modestly with wider 
deployment of SO2 scrubbers.
The political resistance against the deployment and against the normal opera­
tion of SO2 scrubbers also differs. The normal operation and maintenance (O&M) 
costs are significantly higher than the annualized capital costs, especially for SO2 
scrubbers with compromised quality (Xu, 2011b). Data on the capital costs of SO2 
scrubbers were retrieved from two sources to report a dramatic reduction together 
with an expanding domestic market of SO2 scrubbers (Figure 7.2). From February 
to August 2006, China’s Association of Environmental Protection Industries sur­
veyed SO2 scrubber projects in operation or under construction at the end of 2005 
(Xu et al., 2006). One hundred thirteen projects (223 coal-­
fired power units) with 
0
20
40
60
80
100
120
140
160
180
200
0
20
40
60
80
100
120
140
)
W
k
/
$
S
U
(
s
t
s
o
c
l
a
t
i
p
a
c
t
i
n
U
Annual installation of SO2 scrubbers (1,000 MW)
United States, 2000–2009
China, 2000–2008
Figure 7.2  
Annual average unit capital costs of SO2 scrubbers in China and the United States
Source: Xu et al. (2006); EIA (2012–2013); Ministry of Environmental Protection (2008–2012); Xu 
(2013).
Note: China’s average unit capital costs refer to limestone-gypsum wet scrubbers. Annual average 
exchange rates were used for currency conversion. Data from 2000 to the peak year of deployment 
are shown.


152  Environmental technology and industry
a total capacity of 83,850 MW applied limestone-­
gypsum wet scrubber technol­
ogy and had cost information available. Data on projects using other technologies 
are much less continuous to provide longitudinal insights. They had already been 
or were expected to be in operation over the period from 2002 to 2008. Their 
expected time in operation could partly reflect when the contracts were signed and 
accordingly the then market situation. Furthermore, the author’s interviews pro­
vided an independent source to cross-­
check the survey data and to shed light on 
their more recent changes. Data for the United States came from the U.S. Energy 
Information Administration (EIA; 2012–2013).
Quality has a great impact on the capital costs of SO2 scrubbers. For example, 
SO2 scrubbers in Hong Kong’s two coal-­
fired power plants were contracted with 
firms from mainland China, and the unit capital costs were three to four times 
those of similar projects in mainland China, although still at about half of the 
comparable costs in the United States. Hong Kong’s SO2 scrubbers require high-­
quality equipment, engineering and construction and enough redundancy, and 
they take about twice the amount of time from contract to completion. Beyond 
higher labor costs, the higher price in Hong Kong above “The China Price” could 
be mainly explained as a quality premium.
Considering the reduction of capital costs in the Chinese market (Figure 7.2), the 
investment for one more SO2 scrubber would decrease to indicate that the politi­
cal resistance dwindles when many SO2 scrubbers had been deployed. The O&M 
costs for each SO2 scrubber varied less along the deployment dimension because 
of the necessary consumption of electricity, limestone, and water (Table  6.1). 
More SO2 scrubbers led to greater overall O&M costs, and this increased the over­
all political resistance. However, installing SO2 scrubbers without normal opera­
tion wasted financial resources, and it conflicted with environmental policies. The 
associated political pressure for each SO2 scrubber from the civil society, despite 
its underdeveloped status in China, and from within the government increased 
when more SO2 scrubbers were deployed to make the problem more visible. The 
overall net political resistance against the normal operation of existing SO2 scrub­
bers could increase at the very early stage of deployment and then shrink when 
more SO2 scrubbers are in place.
Given China’s then poor record of implementing environmental policies, the 
evolving quality requirements contributed to goal attainment with a rapid path 
that could be theoretically understood. The Chinese government can make a cer­
tain amount of effort to work for pollution mitigation with two choices, either to 
deploy more pollution control facilities or to enhance the operational performance 
of the existing stock. The goal is to maximize the impacts of efforts on pollution 
mitigation at every step. After a certain amount of pollution control facilities have 
been deployed, the net political resistance against the deployment of one more 
facility and against the enhancement of operational performance by 1% could be 
roughly taken as unchanged with the level of deployment. Accordingly, a given 
amount of effort could either raise the deployment rate by α (in the two cases of 
SO2 scrubbers, the unit is megawatts, MW) or the performance of existing facili­
ties by β% (in the SO2 scrubber case, the unit is percentage points of SO2 removal 


Environmental technology and industry 153
rates). The initially deployed facilities have a total capacity of A, and the initial 
performance is B%. Then the initial pollution mitigation effect of the facilities is 
roughly proportional to A × B%. The performance has a technical upper limit, 
B*%.
The option of devoting the efforts to the deployment could raise the pollution 
mitigation effect to (A + α) × B%, and the other option to work on the opera-
tion would have an effect of A × (B% + β%). If there is no constraint, a rational 
decision maker to maximize the impact of his or her efforts will choose the first 
A + α
β
B%
%
+
option when (A + α) × B% > A × (B% + β%), or when 
>
 or 
A
B%
α
β%
α
β%
>
. The second option will be taken when 
<
, and the two options 
A
B%
A
B%
α
β%
are no different when 
=
. With the progress on the deployment and opera-
A
B%
tion, the choice could change. This is what goal-centered 
 
governance would indi-
cate. If adding one constraint that the choice should prioritize policy enforcement, 
the progress should be first made to improve the operation. Only when B% has 
reached B*%, more facilities are allowed to be deployed. This could illustrate 
rule-based governance.
One more constraint could be added to describe the situation on the supply side. 
As examined below with more details, the goal-centered 
 
governance strategy low-
ers technological barriers of market entry to facilitate the rapid establishment of 
a large-enough supply capacity
 
, while the rule-based governance 
 
strategy would 
correspond to higher market- 
entry barriers and discounted supply capacity in the 
Chinese context. To simplify the model, the supply capacity under rule-based 
 
governance is η% less than that in goal-centered
 
 governance, and thus, the same 
amount of efforts could only raise the deployment rate by α
η
×
−
(
%
1
).
The SO2 scrubber case is simulated here to exemplify the usefulness of this 
very simple model. Here are the assumptions of the earlier parameters: (1) A0: the 
initial capacity of SO2 scrubbers, 7,000 MW, equivalent to the level in 2000 (Min-
istry of Environmental Protection, 2008–2012); (2) B0%: the initial SO2 removal 
rate in coal- 
fired power plants with SO2 scrubbers, 31.3%, equivalent to the level 
in Jiangsu Province in 2006 (Xu, 2011b); (3) B*%: 79%, the highest SO2 removal 
α
4,500 MW
rate China achieved in 2010 (Figure 6.1); (4) 
: 
, or the required 
β%
1%
effort from decision-makers 
 
was the same to deploy 4,500 MW of SO2 scrubbers 
and to increase the SO2 removal rate of the existing stock by 1%. The number is 
assumed to fit China’s actual data; (5) η%: 50%, assumed to indicate the impacts of 
higher market-entry barriers in the rule-
 
of-
 
 
law strategy. As illustrated in Figure 7.3, 
the projection with the goal-centered 
 
governance strategy fits well into China’s SO2 
mitigation path for coal-fired 
 
power plants with SO2 scrubbers. If considering no 
constraint from the supply side, rule-based 
 
governance mainly would differ from 
goal- 
centered governance at the early stage of progress. However, if considering 
the potential supply constraints due to higher market-entry barriers, pollution miti
 
-
gation under rule-based governance would proceed at a much slower pace.
 
­


154  Environmental technology and industry
Figure 7.3  
Model projection of the SO2 mitigation path in China’s coal-fired power plants: 
(a) deployment and operation of SO2 scrubbers under goal-centered govern­
ance (the dots refer to actual data); (b) avoided SO2 emissions under goal-
centered and rule-based governance
Source: Xu (2013).
0
100
200
300
400
500
600
0%
20%
40%
60%
80%
100%
SO2
)
W
M
 
0
0
0
,
1
(
 
y
t
i
c
a
p
a
c
 
r
e
b
b
u
r
c
s
SO2 removal rate
2006
2007
2008
2010
(a)
Avoided SO2 emissions
Cumulative efforts
Actual path
Goal-centered governance
Rule-based governance without supply constraint
Rule-based governance with supply constraint
2006
2007
2008
2010
(b)


Environmental technology and industry  155
2  
Technology licensing under goal-­
centered SO2 
mitigation path
The international technology market provided opportunities for China’s domestic 
firms to license foreign technologies and to quickly ramp up their technological 
capabilities, although at a cost. Functioning markets for transferring technologies 
to developing countries not only are important for their economic development 
and upgrading along the value chain but also have critical implications for the 
environment. Due to China’s huge and steadily growing emissions, how fast and 
effective environmentally friendly technologies were adopted was a key determi­
nant for its environmental cleanup. Technology transfer from developed to devel­
oping countries has long been recognized as a key measure in addressing CO2 
mitigation (United Nations, 1992). One important method of technology transfer 
is through technology licensing. With available markets for technologies, a tech­
nology owner could choose between licensing its product or directly investing 
in the client country, and a firm that needs technology could either license in or 
innovate indigenously (Arora et al., 2001a; Teece, 1988; Arora et al., 2001b). In 
international negotiation on transferring low-­
carbon technologies from developed 
to developing countries, developed countries generally argue for market-­
based 
solutions and adequate protection of intellectual property rights (IPR), while 
developing countries often demand nonmarket solutions at lower than market 
rates (Ockwell et al., 2010). The differing positions become an obstacle to the 
agreement of new and effective climate treaties (Ockwell et al., 2010).
Despite unfavorable conditions, the global market for technology has been sig­
nificant, amounting to about US$35 to US$50 billion in the mid-­
1990s (Arora 
et al., 2001b) and roughly US$100 billion in 2002 (Arora and Gambardella, 2010). 
However, only a small portion – less than one third for the United States – of 
technological transactions were between unaffiliated organizations and thus true 
market transactions (Arora and Gambardella, 2010; Saggi, 2002). Most cross-­
border technology licensing happens among developed countries and that from 
developed to developing countries is much rarer (Arora and Gambardella, 2010). 
Product markets in most developing countries are not large enough to attract many 
potential technology licensors. Developing countries generally lag behind devel­
oped countries in human and technological capacities that enable them to effec­
tively absorb licensed foreign technologies and exploit their full value (Metz et al., 
2000). Additionally, effective IPR protection could help address the problems of 
unauthorized use of intellectual property (Gans and Stern, 2010), but developing 
countries often do not have well-­
developed systems of IPR protection and thus are 
placed in relatively disadvantageous positions in creating an attractive market for 
technology (Strokova, 2010). However, large developing countries like China are 
able to access foreign low-­
carbon technologies, although not those at the cutting 
edge (Ockwell et al., 2010; Lewis, 2007). China’s rapid development of many 
industries had roots partly in the importation of foreign technologies, including, 
for example, wind turbines (Lewis, 2007), large hydroelectric turbines (Liang, 
2001) and high-­
speed railways (Chan and Aldhaban, 2009).


156  Environmental technology and industry
An especially prominent case was that of SO2 scrubbers. SO2 scrubber tech­
nologies have been commercially deployed since the mid-­
1970s, mainly in devel­
oped countries. Up until 1998 (expressed in terms of generating capacity of power 
stations thus equipped), the pace of deployment was about 10 GW per year in 
the world and 4 GW per year in the United States (Srivastava et al., 2001). Many 
international firms had established their technological and engineering reputations 
in this field. China began to significantly deploy SO2 scrubbers about three dec­
ades later than developed countries, with a deployment rate of over 100 GW per 
year in the 11th Five-­
Year Plan (Chapter 5). Because of their high SO2 removal 
efficiencies – generally over 90% with wet-­
type technologies – SO2 scrubbers 
became the most vital technology in achieving China’s goal of a 10% reduction 
in SO2 emissions in the 11th Five-­
Year Plan (2006–2010; Xu, 2011b, 2011c). 
Among the more than 500 GW of SO2 scrubbers in China at the end of 2010, more 
than 90% were installed by Chinese firms using licensed foreign technologies 
(Ministry of Environmental Protection, 2011a). Major Chinese firms universally 
licensed foreign technologies and relied heavily on them. Conversely, fewer than 
5% were installed by foreign firms or under joint ventures (Ministry of Environ­
mental Protection, 2011a). Domestic firms dominated the market, in spite of their 
initial lack of proven technologies and experience.
The goal-­
centered SO2 mitigation path created three characteristics of Chi­
na’s SO2 scrubber demand in the early stage. The difficult SO2 mitigation goals 
together with China’s colossal size required more than 100 GW SO2 scrubbers 
annually, which was multiple times as big as the world together had experienced 
before (Figure 5.12). Their initial poor operation significantly relaxed actual qual­
ity requirements (Figure 7.1). The initial one-­
sided emphasis on the deployment 
of SO2 scrubbers indicated that the huge demand for SO2 scrubbers would be cre­
ated swiftly from a low level in the 10th Five-­
Year Plan, which led to stringent 
time constraints for SO2 scrubber firms (Figure 5.12). They played key roles in 
shaping the strategies of domestic technology licensees and foreign technology 
licensors for tapping into the market.
2.1  
The strategy of domestic technology licensees
China’s domestic firms as technology licensees could fall into the three follow­
ing categories: state-­
owned, university-­
established and nonstate. “State-­
owned” 
firms refer to those controlled by state-­
owned power corporations, which could 
have faced less fierce competition to win SO2 scrubber projects because of their 
special “internal” relationship. Indigenous SO2 scrubber technologies had been 
developed by a few research institutes and universities to directly transfer their 
human and technological capabilities to state-­
owned and university-­
established 
firms. Nonstate firms could behave differently due to their relative lack of such 
initial capabilities. In addition, although most of China’s major firms relied heav­
ily on licensed technologies, some concentrated on applying their own. China 
had five large state-­
owned power corporations at the national level in the late 
2010s, four having major SO2 scrubber firms, and two were selected for interview. 


Environmental technology and industry  157
In the available SO2 scrubbers at the end of 2011 with unit scales not smaller 
than 100 MW, the two firms had market shares of 11.9% and 3.3%, respectively. 
Another smaller firm owned by one of the five power corporations was also vis­
ited, and its market share was 0.4%. The special relationship with their parent 
corporations put them in relatively advantageous positions in market competition. 
Eight firms that had no association with power corporations were interviewed. 
Their market shares ranged from 0.7% to 6.0%, being 22.8% in total. In addition, 
two foreign firms and their Chinese representative offices as technology licensors 
were also interviewed to provide an external perspective.
Domestic firms’ decisions to license in SO2 scrubber technologies were heavily 
influenced by the three demand characteristics under the goal-­
centered SO2 miti­
gation path. First, the sheer size of China’s demand for SO2 scrubbers challenged 
the supply capacity. One concern was whether China had enough engineers. This 
condition was met partly through rapidly training many more university students 
(Figure 7.4). In 2000, 496,000 undergraduate students graduated from full-­
time 
four-­
year undergraduate programs, including 213,000 in engineering. In 2010, the 
numbers had grown to 2,591,000 and 813,000, respectively. In 2018, the numbers 
further climbed to 3,868,000 and 1,269,000, respectively. In 2018, about the same 
number of undergraduate students (3,665,000) graduated from other full-­
time pro­
grams with shorter study periods of two or three years. The age group, 20 to 24, 
comprised 5.95% of China’s population in 2018, or 16.6 million for each yearly 
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
1998
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
)
e
l
p
o
e
p
 
0
0
0
,
1
(
 
s
e
t
a
u
d
a
r
G
Year
Science
Engineering
Agriculture
Medicine
Others
Figure 7.4  
Yearly university graduates in China from four-year undergraduate programs 
by subjects
Source: Ministry of Education (1999–2019).


158  Environmental technology and industry
age (National Bureau of Statistics, 1996–2019). Accordingly, in 2018, about half 
of China’s newly available labor force had a received formal university educa­
tion. Other part-­
time or Internet-­
based undergraduate programs trained another 
4.1 million graduates in that year. These enhanced human resources provide a 
crucial foundation for China’s rapid deployment of pollution-­
removal industrial 
facilities.
The huge market also helps diminish one concern that licensors might not trans­
fer technologies completely after receiving payments (Arora et al., 2001b). In the 
case of SO2 scrubber technology, royalties dominated the revenue stream in tech­
nology licensing and effectively deterred such a moral hazard. By way of exam­
ple, an American firm charged one licensee US$652,118 as the up-­
front lump-­
sum 
fee (Table 7.1): interviews discovered that a license’s approximate royalty rate 
should be 2% of SO2 scrubber contract values. Between 2004 and 2010, the firm’s 
income from royalties was nearly 40 times as much as the up-­
front lump-­
sum fee 
(the licensee completed 34,900-­
MW wet SO2 scrubbers in that period; Ministry 
of Environmental Protection, 2011a), and the national average contract value was 
about US$35/kW (Xu et al., 2006)). From another perspective, as demonstrated 
in the case of a Japanese licensor, a licensee’s loss was limited to approximately 
the up-­
front lump-­
sum fee when the technology transfer was not satisfactory. In 
addition, if a licensor gained a bad reputation, this could limit its future business 
opportunities in the huge and rapidly growing Chinese market.
Second, the quality requirements for SO2 scrubbers were initially low. The 
deployment of SO2 scrubbers took off around 2002, but the normal operation was 
improved significantly only in about 2007 (Xu, 2011b; Xu et al., 2009). In the five 
gap years, many managers of installed SO2 scrubbers did not plan to operate them 
normally and cared very little about the quality, while quality was closely associated 
with the technological advancement of a supply firm. In addition, China’s reform 
in the power sector in 2002 created multiple independent power corporations to 
Table 7.1 Up-front 
 
lump-sum fees of SO2 scrubber technology licenses (the Chinese 
licensees here are all listed on stock markets and the data are from their annual 
reports)
Chinese licensee
Country origin Lump-sum fee*
Year
Technology type
of the foreign 
licensor
Wuhan Kaidi
Germany
US$277,304
1998
Dry type
Fujian Longjing
Germany
US$3,989,234
2001
Wet type
Circulation fluidized bed
Wuhan Kaidi
United States
US$652,118
2002
Wet type
Zhejiang Feida
United States
US$1,250,000
2002
Wet type
Jiulong Electric
Japan
US$1,126,563
2002
Wet type
Jiulong Electric
Austria
US$1,423,765
2004
Wet type
Insigma Technology
France
US$1,200,000
2004
Wet type
* Exchange rates on December 31, 2010 were used: 1 US$ = 6.62 RMB = 0.75 euro.


Environmental technology and industry  159
encourage competition – this was even though all of these were state-­
owned. The 
rapid construction of new power plants strained their available financial resources 
to create strong incentives to minimize capital investment for each new project, 
while the poor quality of SO2 scrubbers could substantially reduce capital costs. 
Furthermore, the low requirement for quality was strengthened by the largely sepa­
rate decisions of capital investment and daily operation and by the different incen­
tives of respective decision-­
makers. Managers of coal-­
fired power plants should 
have an incentive to install high-­
quality SO2 scrubbers while capital investment 
was within the authority of the upper levels of management in power corporations. 
The low requirement for quality and technological advancement substantially low­
ered the technological market-­
entry barrier not just for the SO2 scrubber firms but 
also along the entire supply chain. In contrast, the quality requirement and techno­
logical market-­
entry barrier in the U.S. market were much higher.
China’s regulators also paid attention to the quality requirements, especially with 
the knowledge of domestic firms’ initially unsatisfactory technological statuses. 
Technologies could come from international transfer or in-­
house innovation. Vari­
ous factors could affect the choice of a country or a firm between these two technol­
ogy strategies. China used to focus almost entirely on in-­
house innovation under 
the rule of Chairman Mao when China segregated itself from the world. The “Not 
Invented Here” syndrome – that internally developed technologies are preferred – 
was found to be a barrier to technology licensing (Arora and Gambardella, 2010), but 
it does not seem to be deeply rooted in China in the economic reform era. Secondary 
innovation based on imported technologies, coupled with original and integrated 
innovation, had been established as three cornerstones of China’s indigenous inno­
vation strategy (State Council, 2006). With regard to the installation of SO2 scrub­
bers, China stipulated in tendering documents that established technologies were 
required. As late as 2005, bidders were clearly asked to specify a foreign technology 
provider that had installed SO2 scrubbers of the same or greater scale (Guizhou 
Qiandong Power Station, 2005). Interviews also confirmed the general requirement 
for foreign, commercialized technologies in the early years when almost no Chinese 
firms had any proven experience. This requirement was relaxed only in later years 
after many firms in the market had completed enough projects.
Third, time was a serious constraint. In the late 1990s and early 2000s, few 
domestic firms were capable of designing SO2 scrubbers. The sudden appearance 
of a huge market led to the creation of many new firms and the reorientation of 
existing ones from other industries. Because few firms had any prior experience 
and the market was large enough to accommodate many, most – except those 
owned by coal-­
fired power corporations – were placed on a more or less equal 
footing. Firms would achieve distinction if they could establish engineering and 
management teams and develop their technological capability faster than others. 
Another time constraint was the short period from the issue of tendering docu­
ments to completion of the bidding process; this typically lasted only one to four 
weeks. Additionally, the design process could not take more than a few months 
if the construction was to begin on schedule. Successful firms had to respond 
quickly and provide acceptable quality.


160  Environmental technology and industry
These time constraints helped push domestic firms toward technology licens­
ing, due to their weak technological foundations. When demand for SO2 scrub­
bers started to surge, domestic technologies were generally not able to satisfy 
the time constraints because of their immaturity. Domestic research and devel­
opment generated “naked” technologies, to quote the word of one interviewee. 
Demonstration projects on a commercial scale should be followed by multiple 
projects to make the technology mature and ready for wide commercial deploy­
ment. The commercialization of these “naked” technologies would require at least 
a few years plus significant financial resources and the willingness of coal-­
fired 
power plants to take risks by trying them. The expected short-­
term peak in Chi­
na’s scrubber market diminished the potential return on investment in indigenous 
technology. The easy prospect of licensing foreign technologies also reduced the 
incentive to take risks with indigenous innovations. All the major Chinese firms 
in the market licensed foreign technologies in order to acquire and substantiate 
their technological capabilities. No clear difference could be found among state-­
owned, university-­
established and nonstate firms. Even the nonstate firm that 
mainly applied its own technology had to initially license from abroad.
As tacit knowledge cannot be so easily transferred as codified knowledge, kno­
whow played a positive role in establishing a sound market for technology. The 
contractual acquisition of know-­
how presents more problems than licensing pat­
ents (Arora et al., 2001b). However, in a developing country like China with poor 
IPR protection, the licensing of patents might be unnecessary in the absence of 
know-­
how as the knowledge contained in the patents have already entered the 
public domain. Chinese firms had generally chosen to legally license, rather than 
to illegally acquire, SO2 scrubber technologies. Legal licensing secured a com­
plete package including systematic training, technical documentation and trade 
secrets in a relatively short timescale, without exposing the licensees to legal 
disputes. One alternative option was to recruit experts from foreign firms, but 
the legal risks were not insignificant and the received technologies may not be 
complete because it would be difficult to recruit an entire team. It would also take 
much longer for the acquiring firms to comprehend a technology by this means 
than they would through technology licensing. The associated costs would not be 
low either, because foreign experts generally had to be paid considerably more 
than standard Chinese salaries. Furthermore, illegal acquisition did not provide a 
technological guarantee from a trusted provider, while this guarantee was stipu­
lated by coal-­
fired power plants in their tendering documents.
China’s domestic firms could quickly absorb licensed technologies to meet the 
time constraints. From as early as the 1970s, China had, through its own research 
and development on SO2 scrubbers, built up vital capabilities to establish domes­
tic firms and assimilate imported technology (Shu, 2003). From the mid-­
1970s 
to the mid-­
1980s, China appraised several technologies, although on scales that 
were at least one or two orders of magnitude smaller than any commercial pro­
ject. For example, a 300-­
MW unit corresponds to a flue gas flow rate of about 
1,000,000 Nm3/hour (cubic meter at standard temperature and pressure per hour), 
while the largest Chinese experiment at the time had a flow rate of 70,000 Nm3/hour 


Environmental technology and industry  161
0.00%
0.25%
0.50%
0.75%
1.00%
1.25%
1.50%
1.75%
2.00%
2.25%
0
500
1,000
1,500
2,000
2,500
3,000
3,500
4,000
4,500
1995
2000
2005
2010
2015
Ratio
l
l
u
f
 
0
0
0
1
(
 
l
e
n
n
o
s
r
e
p
 
D
&
R
-
 
)
e
l
p
o
e
p
 
t
n
e
l
a
v
i
u
q
e
 
e
m
i
t
)
B
M
R
 
8
1
0
2
 
n
o
i
l
l
i
b
(
 
e
r
u
t
i
d
n
e
p
x
e
 
d
n
a
Year
R&D Personnel (Full-time Equivalent)
R&D expenditure
Technology market transaction value
R&D expenditure vs. GDP (%)
Technology market transaction value vs. GDP (%)
Figure 7.5  
R&D personnel, expenditure and market value (in 2018 RMB) in China
Source: National Bureau of Statistics (1996–2019).
(Shu, 2003). From the mid-­
1980s to 2000, foreign technologies were demonstrated 
on a commercial scale (Gu, 2004; Shu, 2003). In 2000, having resulted in a consid­
erable fund of domestic human and technological capability, foreign technologies 
were officially recognized as the basis for further development of SO2 scrubber 
technologies in China (National Economic and Trade Commission, 2000). China’s 
absorptive capacities were effectively distributed to all major firms including non­
state ones through a free labor market of engineers and managers.
Recognizing the constraints of technology licensing such as on expansion 
beyond China, in the past two decades, China has put a much heavier empha­
sis on research and development (R&D). In 2000, China had 922,000 full-­
time 
equivalent personnel on R&D and this number rapidly grew by 375% to 4.4 mil­
lion in 2018. R&D expenditures were raised from 0.60% of gross domestic 
product (GDP) in 1995 to 2.19% in 2018 (Figure 7.5). A much more vibrant mar­
ket for technology emerged and the transaction value increased from 0.46% of 
GDP in 1995 to 1.97% in 2018 (Figure 7.5). Together with the rapid growth of 
China’s GDP, the R&D expenditures and technology market transaction values 
had become 1084% and 1365% greater in 2018 from the levels in 2000 in real 
terms (Figure 7.5). This R&D boom strengthened China’s capacity to absorb for­
eign technologies and innovate domestic intellectual property. In the category of 
environmental technology, China’s residents and nonresidents were granted 103 
and 69 patents, respectively, in 2000 in China’s patent filing office, which were 
about 10% of those in the United States. They grew to 7,459 and 881 patents, 


162  Environmental technology and industry
respectively, in 2018, while the figures in the United States were correspondingly 
1,258 and 1,369 patents (Figure 7.6).
2.2  
The strategy of foreign technology licensors
The strategy of potential foreign technology licensors was also shaped by the pre­
viously mentioned three characteristics of China’s SO2 scrubber demand under a 
goal-­
centered SO2 mitigation path. First, the huge demand for SO2 scrubbers created 
profitable business opportunities. Their decision of technology licensing involves 
the revenue effect (i.e., payments received from licensing) and rent-­
dissipation effect 
(i.e., revenue loss due to a new or strengthened competitor in the product market; 
Arora and Fosfuri, 2003). A stronger revenue effect promotes the decision to license, 
while a stronger rent-­
dissipation effect discourages licensing. For major foreign 
firms that held intellectual property of SO2 scrubber technologies, the option to do 
nothing was rarely attractive because of the temptation of the huge emergent Chi­
nese market. The revenue effect was indeed significant. Technology licensing only 
required a small office in China to monitor licensees and to “service” the partnership. 
For example, each of the two interviewed American firms had an office in Beijing 
with about five staff members, whereas their licensees were in charge of contracts 
worth several hundred million dollars annually. The initial cost in transferring tech­
nologies was covered by up-­
front lump-­
sum fees paid by licensees (Table 7.1). The 
0
3,000
6,000
9,000
12,000
15,000
18,000
21,000
1980
1985
1990
1995
2000
2005
2010
2015
)
y
g
o
l
o
n
h
c
e
t
 
l
a
t
n
e
m
n
o
r
i
v
n
e
(
 
e
c
i
f
f
o
 
g
n
i
l
i
f
 
y
b
 
s
t
n
a
r
g
 
t
n
e
t
a
P
Year
China: Resident
China: Nonresident
U.S.: Resident
U.S.: Nonresident
Others: Resident
Others: Nonresident
Figure 7.6  
Patents on environmental technology by filing office in the world
Source: WIPO (2019).


Environmental technology and industry  163
commercial success of licensees would result in considerable royalties to the licensor 
if the contracts were honored. After the know-­
how and trade secrets were transferred, 
the intellectual property rights were at risk of misuse or infringement, possibly with 
the royalties not being fully paid. Despite this, most foreign firms decided to take this 
risk in order to avoid the much greater risk inherent in direct investment.
After technologies are transferred, one primary concern of technology licensors 
arose on whether licensees paid royalties honestly. Both licensors and licensees 
reported in interviews that major Chinese firms were paying royalties regularly. Also, 
several expiring licenses had been renewed, indicating a good record of royalty pay­
ments. As a preventative measure, design software was encrypted and only specially 
prepared computers could install it with annual reregistration. Several interviewees 
in the Chinese firms said that, after a few years, they had figured out what was inside 
the black box but still chose to pay royalties. It was not very difficult to keep track of 
licensees. The huge size of SO2 scrubbers often made local news and the Ministry of 
Environmental Protection annually published details of every SO2 scrubber and its 
contractor (Ministry of Environmental Protection, 2011a). Besides, a good partner­
ship with licensors suited the long-­
term interests of licensees. Technological sophis­
tication had increased step by step in the Chinese SO2 scrubber market as reflected 
in the unit scales: the 300-­
MW scale was dominant before 2005, but after 2006, the 
600-­
MW scale became crucial and then the 1,000-­
MW scale or greater (Ministry 
of Environmental Protection, 2014). Every significant increase in scale indicated a 
new technical advance. Accordingly, the licensing of scrubber technologies was a 
continuous operation and not a one-­
off process. Good partnerships, strengthened 
by honest royalty payments, could also help licensees expand into new markets 
through future technology licensing. In addition, a partnership may generate busi­
ness opportunities for both sides. For example, when a large coal-­
fired power plant 
in Hong Kong decided to install SO2 scrubbers, it first approached several interna­
tional firms, including one from the United States. But the American firm was fully 
committed in the domestic market and was not willing to take the financial risk of 
an Engineering, Procurement, and Construction (EPC) project in Hong Kong. Its 
Chinese licensee was introduced and finally won the contract.
Royalty rates may decrease over time to reduce the costs of honoring licensing 
contracts. For example, one license divided the ten-­
year contract period into three 
phases with declining royalty rates. In several other cases, the royalty rate was rene­
gotiated when competition in the market became much too fierce to significantly 
shrink the profit margin. Excessively high royalty rates could damage licensees’ 
competitiveness. The final result might be a reduced income from royalties and 
an increased risk of no payment being made at all. The renegotiation strengthened 
the partnerships between licensors and licensees and thus worked for the inter­
ests of both sides. In one licensing contract signed in 1998, the level of royalties 
was originally associated with the volume of flue gases. Because China’s capital 
costs of installing SO2 scrubbers had dropped substantially since then (Figure 7.2), 
the royalty rate would increase significantly as a percentage of the contract value. 
Renegotiation took place to lower the royalty rate. The partnership remained strong 
with both the licensor and the licensee maintaining market success.


164  Environmental technology and industry
Lawsuits, particularly those resolved outside China, were also a deterrent to 
potential infringement, which maintained the strong revenue effect. For example, 
Insigma Technology is a Chinese firm listed on the Shanghai Stock Exchange, and 
it releases information regularly. It signed a technology licensing contract with 
a French firm in December 2004 (Table 7.1). However, in April 2006, Insigma 
declared that it would cancel the contract and thereafter stop using the licensed 
technology. Royalties were paid for six projects in 2005 and 2006 with a total 
capacity of 7,450 MW (Sina Finance, 2010). The firm later signed a new contract 
with an Italian firm in September 2006, which was for one year and was to be 
automatically renewed if no objections were received from either side. The fee 
for royalties was a fixed sum of €20,000 (US$26,600) for every project regardless 
of the contract value (Sina Finance, 2010). The French firm later sued Insigma in 
Singapore (where disputes should be resolved according to the licensing contract). 
The court made a decision in February 2010 and Insigma was ordered to pay com­
pensation of US$2,085,737 for the loss of royalties in 2005 and US$24,566,684 
for the loss afterward (Sina Finance, 2010). The lawsuit may have helped deter 
other significant licensees from not honoring their licensing contracts.
Second, low-­
quality requirements and correspondingly low technological 
market-­
entry barriers led to active market entry of new firms to contain the rent-­
dissipation effect for technology licensors. If the downstream operations of a firm 
are small or the downstream market is in fierce competition, the rent-­
dissipation 
effect will be limited and technology licensing becomes more likely (Arora and 
Gambardella, 2010). Indeed, the Chinese downstream SO2 scrubber market was 
newly created and in fierce competition (Ministry of Environmental Protection, 
2011a). In addition, market evolution also demonstrated that the rent-­
dissipation 
effect should be minimal. Foreign firms tended to lag behind domestic ones in 
understanding the market’s real demand, especially in the early period. Among all 
the foreign firms, the examined Japanese firm ought to be the best prepared for 
the Chinese market. It owned more Chinese patents on flue gas desulfurization 
than any other firm (State Intellectual Property Office, 2010) and, between the late 
1980s to 1990s, had won contracts to install China’s first-­
ever commercial wet 
SO2 scrubbers (four units of 360-­
MW capacity; Gu, 2004). However, up to the 
end of 2010, its technology was only applied to a further 3,300 MW, with the final 
project in 2006 (Mitsubishi Heavy Industries, 2011). Interviews in China revealed 
that many foreign firms generally licensed design software together with other 
know-­
how in order to enable their Chinese licensees to compete independently, 
but this Japanese firm was reluctant to hand over design software and wanted to 
participate more actively. Thus, the technology transfer of know-­
how was not 
complete. The decision could have been influenced by the expectedly significant 
rent-­
dissipation effect due to potentially high rents as a result of its favorable 
position in granted patents. However, partly because the relationship made them 
slower in responding to the market and hampered their competitiveness, its Chi­
nese licensees decided instead to do business with other technology licensors. For 
example, according to the annual reports from a firm listed on the Shanghai Stock 
Exchange – Jiulong Electric, the holding firm of Yuanda Environmental Protection 


Environmental technology and industry  165
Engineering – although US$1.1 million was paid to the Japanese firm as the up-­
front lump-­
sum fee, just two years later it decided to sign another licensing con­
tract with a European firm and gave up the Japanese technology (Table 7.1). Even 
with the tight control of technology licensing, the Japanese firm earned little profit 
or rent from the Chinese market, an indication of a small rent-­
dissipation effect. 
The existence of many technology licensors diminished the rent-­
dissipation effect 
because no single licensor had significant market power.
Third, time constraints discouraged direct participation of foreign technology 
licensors in the Chinese market. Two interviewed American firms each had a 
small representative office in Beijing, but their licensing strategies were notably 
different. They reported that the Chinese government put no restrictions on allow­
ing foreign firms to bid for SO2 scrubber projects, but many foreign firms did not 
expect that they would earn significant profits by establishing subsidiaries or joint 
ventures in China. One major American firm expected the Chinese market to peak 
for only a few years before it began shrinking; this expectation proved prescient 
(Figure 5.12). The initial investment of capital and human resources to establish a 
subsidiary in China would therefore only be of temporary benefit. The firm’s past 
experience in other countries suggested that direct investment could not be freely 
withdrawn, and accordingly, it was not justified in this particular Chinese market. 
In addition, the lack of adequate human resources also constrained some foreign 
firms from choosing direct investment, particularly due to the revived U.S. market 
for SO2 scrubbers (U.S. Energy Information Administration, 2011).
2.3  
Why technology market can emerge in China?
Even in developed countries – as Gans and Stern argue – an effective market for 
technology is difficult to establish because it often fails to satisfy the three criteria 
of effective market design as specified by Roth that successful marketplaces must 
be “thick, uncongested and safe” (Gans and Stern, 2010; Roth, 2008). The Roth 
criteria were proposed to fix broken markets or build new ones if they are missing, 
which could be especially useful for environmental protection as market failure 
is often the cause. First, an efficient market requires many potential buyers and 
sellers, or market thickness, to enhance the chances of effective matching. How­
ever, many ideas are not independent but reliant on other complementary ideas and 
assets to achieve their full value, with notable examples in low-­
carbon technolo­
gies (Harvey, 2008). This problem makes the licensing of a single idea less desir­
able. If the ideas belong to different entities, ineffective coordination could limit 
the willingness of potential buyers and sellers to participate in the market. Second, 
the market should overcome Roth’s “congestion” criterion, whereby buyers and 
sellers should be able to negotiate with a number of possible trading partners and 
have sufficient time to make effective selections. In a congested market, competi­
tion is not sufficient and the price does not reach market equilibrium. Because nec­
essary information disclosure for buyers to assess a technology’s value might lead 
to unwanted diffusion, the information is often kept secret between buyers and sell­
ers to constrain open market competition, thus failing the “congestion” criterion. 


166  Environmental technology and industry
Third, market transactions should be “safe”; that is, conducted in good faith and 
with safeguards that allow the expression of real intention and information and 
result in mutual satisfaction. A drawback on this point is that, after licensors have 
disclosed information, licensees might be able to exploit it independently, without 
signing licensing contracts, creating issues over misuse of intellectual property.
The Chinese market for SO2 scrubber technologies satisfied all three Roth crite­
ria. Key contributing factors could include China’s large market size, the maturity 
of available technologies and goal-­
centered governance. First, because the size of 
the Chinese market for SO2 scrubbers as a downstream market for the technolo­
gies is far greater than any other country, major foreign SO2 scrubber firms, as 
potential licensors, could hardly overlook the potential business opportunities. 
The large market and low technological barriers facilitated by technology licens­
ing have created many domestic firms as potential licensees. Multiple sellers from 
the United States, Europe and Japan actively licensed out their technologies (Xu 
et al., 2009, 2006). In addition, in the Chinese market up to 2010, 16 firms – all 
Chinese – had completed at least 10 GW of SO2 scrubbers, all using licensed-
­
in foreign technologies (Xu et al., 2006; Ministry of Environmental Protection, 
2011a). The three types of Chinese firms – state-­
owned, university-­
established 
and nonstate – did not show significantly different behavior in the market for 
technology. Fierce competition drove down costs and diminished expected profit 
from direct investment, but revenue from technology licensing was significant. 
The rent-­
dissipation effect was overwhelmed by the revenue effect of technology 
licensing, which accordingly became a dominant choice of foreign firms. As a 
large country, China has a strong capacity to absorb new technology due to its 
previous R&D, and this capacity was effectively distributed to all three types of 
firms through a free labor market. Licensors and licensees held multiple bilateral 
negotiations simultaneously to help solve the market congestion problem. Fur­
thermore, the safety of technology licensing also benefited from China’s large 
market size. As a result of the large market, there were significant revenues from 
royalties that encouraged licensors to transfer complete packages of technolo­
gies. The market for SO2 scrubbers at every unit scale was substantial and the 
unit scales escalated over time to require continuous technological support from 
licensors. Such dynamism favored long-­
term partnerships between licensors and 
licensees for their mutual benefit and fostered honest royalty payments.
Second, the maturity of SO2 scrubber technologies played a crucial facilitating 
role. After several decades of commercial deployment in developed countries, 
many firms had acquired complete technology packages. Personal and corporate 
expertise, or know-­
how as tacit knowledge, was a vital part of the technology 
package. Acquiring knowhow raised costs and contracting problems, but given 
the inadequate standard of IPR protection in China, technology licensing became 
necessary in order to acquire complete packages of technologies. Many foreign 
firms had become independent technology holders, and a potential licensee only 
needed to negotiate with one licensor for a complete technology package. When 
deciding whether to license out technologies or set up direct subsidiaries in devel­
oping countries or even just do nothing, firms from developed countries needed 


Environmental technology and industry  167
to compare the expected profits of each market option. The dominant business 
reality in the market was technology licensing. For a potential licensee, the tech­
nology could either be developed internally or acquired externally. Favorable con­
ditions created the demand for foreign technologies in the Chinese market.
The Chinese market also met the second Roth criterion on the lack of con­
gestion. The maturity and wide deployment of SO2 scrubbing technologies also 
enabled a fairly accurate estimation of the technology’s value to facilitate mar­
ket transactions. Interviews revealed that, although the negotiation of technology 
licensing was generally bilateral, without disclosing information to third parties, 
licensors and licensees often negotiated with several entities on the other side at 
the same time for most suitable licensing contracts. IPR protection is recognized 
as a key means to ensure market safety and satisfy the third Roth criterion (Gans 
and Stern, 2010). As examined earlier, know-­
how and credible threat of lawsuits 
ensured the general satisfaction of this criterion. The disclosure of the necessary 
information for value assessment in negotiations caused fewer problems because 
knowhow could not be easily acquired.
The existence of many potential licensees enabled licensors to design their 
strategies to maximize profit. At least three clear strategies emerged among three 
licensors. A major American firm licensed to only two Chinese firms and built up 
long-­
term partnerships through full technical support. One license was restricted 
to the licensee’s home province for a certain period and the other covered the 
whole of mainland China. The licensees had a near monopoly to use the specific 
technology in their assigned market territories. Another significant American firm 
had about eight licensees in China; the strategy was to increase the market share 
of its technology as well as its royalties, but the licensees were still selected so as 
to prevent unqualified ones from ruining the technology’s reputation. In addition, 
as mentioned earlier, a Japanese firm licensed its technology to a few Chinese 
firms but, unlike the two American firms, refused to transfer design software. The 
two American firms had their technologies widely applied but the Japanese tech­
nology was abandoned without much deployment. From the perspective of the 
level of royalties, the two American strategies were clear winners.
An effective market for cutting-­
edge technologies is understandably more 
difficult to establish. It is probable that not many organizations have acquired 
intellectual property as potential licensors. The value of a particular cutting-­
edge 
technology is harder to assess and the accumulation of know-­
how may still be in 
progress with a consequently high price of the final product which will limit its 
deployment. These unfavorable conditions discourage the emergence of potential 
licensees. Information disclosure to facilitate licensing will also raise more con­
cerns on the part of technology owners. As a result, the Roth criteria of effective 
market design will be harder to meet for cutting-­
edge than for mature technologies.
Third, goal-­
centered governance resulted in a path of SO2 mitigation to signifi­
cantly lower market-­
entry barriers for domestic firms. The previous two factors 
are mainly given, while governance strategy could be more deliberately taken. 
For developing countries that have not established a sound rule of law and strong 
domestic industries for pollution removal, goal-­
centered governance may induce 


168  Environmental technology and industry
a feasible path for improvement. In order to meet time constraints and technologi­
cal requirements, major Chinese firms universally licensed in foreign technolo­
gies to quickly build technological strength. In the early period, China had not 
established a system to well implement environmental policies and thus many 
SO2 scrubbers were not operating normally. For meeting governmental regula­
tions, coal-­
fired power plants chose to install the cheapest SO2 scrubbers but did 
not expect to run them. For domestic firms that had no technological advantages, 
this initially low but escalating requirements on the quality of SO2 scrubbers pro­
vided helpful stepping-­
stones to enter the market.
The utilization of wind energy followed a comparable path under goal-­
centered 
governance, which also helped to lower market-­
entry barriers for the establish­
ment of a domestic wind turbine industry. Similar to the SO2 mitigation case, the 
initial stage of wind energy development also focused more on the deployment 
to follow the goal-­
driven demand. In China’s 11th Five-­
Year Plan for Renew­
able Energy Development, the major goal for wind electricity referred to gen­
eration capacity whereas actual electricity generation served as a supplementary 
goal (NDRC, 2008). One average kilowatt-­
hour of wind capacity consistently 
generated much less electricity in a year in China than in the United States, and 
this partly indicated poorer operating conditions in China (Figure 7.7). When the 
deployment of wind turbines became sufficiently wide, the Chinese government 
started to pay more attention to their operation. Problems in the quality and opera­
tion of wind turbines emerged with their deployment to threaten not just wind 
0
20
40
60
80
100
120
140
160
180
200
0
50
100
150
200
250
300
350
400
2000
2002
2004
2006
2008
2010
2012
2014
2016
2018
Wind capacity (1,000 MW)
)
h
W
T
(
 
y
t
i
c
i
r
t
c
e
l
e
 
d
n
i
W
Year
Wind electricity: China (left)
Wind electricity: United States (left)
Wind capacity: China (right)
Wind capacity: United States (right)
Figure 7.7  
Wind energy development in China and the United States
Source: BP (2019).


Environmental technology and industry  169
electricity generation but, more important, also the safety of the electric grid and 
to push for greater focus and higher requirements (SERC, 2011). In 2010, the 
National Energy Administration published a plan to enact 247 technical stand­
ards for wind energy development, including several which were already in force 
(National Energy Administration, 2010). Lower technological market-­
entry bar­
riers played a positive role to encourage new firms. In 2006, the Chinese market 
had 12 firms that supplied wind turbines, and the number rose to 29 in 2012 (Shi, 
2007; China Wind Energy Association, 2012). Many component suppliers along 
the supply chain also actively entered the market (Chinese Wind Energy Equip­
ment Association, 2011). Compared to wind turbine manufacturers, market-­
entry 
barriers were even lower and the technologies were less complex for component 
suppliers, and this resulted in fiercer competition and thinner profit margins.
Furthermore, unlike SO2 scrubber firms, the Chinese firms in the wind industry 
licensed their technologies from a very different category of foreign firms. Foreign 
licensors of SO2 scrubber technologies were generally major firms that were closely 
involved in the downstream business of installing SO2 scrubbers (Xu, 2011a). In 
contrast, major foreign wind turbine manufacturers were largely reluctant to license 
technologies to Chinese firms, and most foreign licensors were design firms or small 
manufacturers that focused more on upstream technological development. This 
phenomenon is explained in the theory of markets for technology as the rational 
choice based on the respective industrial structure (Arora and Gambardella, 2010). 
The good-­
enough quality, lower price and no geographic constraints of technology 
licenses made the Chinese domestic wind industry potentially competitive.
The market for technology might also work for other large developing coun­
tries, such as India. They may also have potentially large markets through which 
to spawn many domestic operators and fierce competition. Many other low-­
carbon 
and pollution-­
control technologies have been commercialized with much know-­
how. A caveat is that these large developing countries may not necessarily always 
have large domestic markets for pollution mitigation. These are partly determined 
by government policies and not just by the overall sizes of their economies. Their 
abilities to take on board foreign technologies might not be consistently strong. 
However, there is great potential for large developing countries to make use of mar­
kets for technology to build their industrial prowess with mature technologies. Goal-­
centered governance may provide more feasible pathways for domestic industries in 
these developing countries to take roots and further grow from weak starting points.
3  
Environmental industry under goal-­
centered SO2 
mitigation path
3.1  Market entry and competition
Considering both firms that pollute the environment and others that provide pol­
lution removal facilities, the impacts of the goal-­
centered SO2 mitigation path in 
China may not be straightforward. On one hand, although empirical studies gen­
erated mixed results on the “pollution haven hypothesis” in the Chinese context 


170  Environmental technology and industry
(Levinson and Taylor, 2008; He, 2006; Shen, 2008), its key root cause – poor 
environmental regulation, including weak policies and poor enforcement  – is 
argued to potentially benefit polluting firms for not acting on, delaying or comply 
only partially with pollution control (Harney, 2008). In China, policies on envi­
ronmental protection and business standards were recognized by polluting firms 
as less important barriers to market entry (Niu et al., 2012). Relative to the Euro4 
fuel quality standards, the poorer Euro2 standards in China could reduce costs by 
1.1 and 1.9 U.S. cents per gallon for gasoline and diesel, respectively (Liu et al., 
2008). The cost burden also acts as a political and regulatory hurdle to bring pol­
luting firms under full compliance. On the other hand, from the perspective of 
supplying pollutant removal facilities, weak regulation could lower market-­
entry 
barriers to encourage competition, innovation and the establishment of industrial 
capacities for pollution control (Stigler, 1971; Dean and Brown, 1995).
Two important barriers on the supply side could slow down the deployment 
of SO2 scrubbers in China. No existing supply capacity could meet the unprec­
edented peak demand of over 100 GW a year (Figure 5.12). The capital costs of 
about US$65 to 90/kW (Figure 7.2) were initially too high, being over 10% of 
the costs of building new coal-­
fired power plants (SERC, 2006). If the large labor 
force and industrial base in China could be effectively mobilized for the deploy­
ment of SO2 scrubbers, the supply capacity would not have a major problem in 
meeting the rapidly growing demand. The lack of significant restrictions on for­
eign direct investment indicates that both foreign and domestic firms could tap 
into the labor force.
The huge Chinese market can easily accommodate many SO2 scrubber firms 
without losing economies of scale. Whether the supply potential could be released 
depends on whether existing firms could expand their capacity and (more impor­
tantly) whether new firms could emerge. Although the U.S. market had only 
about ten firms, and with new firms rarely entering, the Chinese market had over 
60 firms – almost all of which were newly established, most being domestic but 
some being foreign – thereby indicating much lower market-­
entry barriers (Fig­
ure 7.8). In the past decade, the annually added capacity of SO2 scrubbers increased 
significantly both in China and the United States, but the evolution of unit capital 
costs showed a rapid cost reduction in China and a cost spike in the United States 
(Figure 7.2). In China, the rapidly rising demand triggered intensive market entry 
to create fierce competition followed by a cost reduction whereas competition in 
the United States was rather limited, and this constrained the expansion of the sup­
ply capacity. When the demand for SO2 scrubbers grew, the price was pushed up.
As discussed earlier, domestic firms did not have technological advantages, 
especially in the early period. Nevertheless, because of the existence of many 
potential licensors in the technology market, no foreign firm was able to prevent 
others from licensing technologies to China. Technologies therefore could not be 
used as a barrier to exclude Chinese firms from competing. The crowded market 
enabled fierce competition not just for providing SO2 scrubbers. Competition also 
took place between foreign firms for licensing to especially promising Chinese 
firms that were expected to win many projects and return significant revenues 


Environmental technology and industry  171
0
10
20
30
40
50
60
70
0
2
4
6
8
10
12
14
2000 2002 2004 2006 2008 2010
2000 2002 2004 2006 2008 2010
Number of FGD companies in the Chinese market
t
e
k
r
a
m
 
.
S
.
U
 
e
h
t
 
n
i
 
s
e
i
n
a
p
m
o
c
 
D
G
F
 
f
o
 
r
e
b
m
u
N
Year
New entry
Existing
United States
China
Figure 7.8  
Firms in the Chinese and U.S. markets installing 100-MW-scale or greater SO2 
scrubbers
Source: Ministry of Environmental Protection (2008–2012); EIA (2007–2011); Xu (2013).
Note: “Existing”: firms have been in the market in the past. “New entry”: firms entering the market for 
the first time. The U.S. numbers use the left axis, and the Chinese numbers use the right axis.
from royalties. Those potential licensees were mainly established by coal-­
fired 
power producers. Interviews showed that financial payments were the most criti­
cal aspect of negotiating licenses, although other aspects were also important, 
such as the suitability of technologies and the scope of licenses. The willingness 
to accept lower up-­
front lump-­
sum fees and lower royalty rates made a licensor 
more competitive. After the significant variance of early contracts, the up-­
front 
lump-­
sum fee stabilized to be about US$1.2 million for wet scrubbers (Table 7.1).
3.2  
International competitiveness of China’s SO2 scrubber industry
Due to specific features in various environmental fields, goal-­
centered govern­
ance may present very different impacts on different environmental industries. 
One significant difference is on the international competitiveness of China’s SO2 
scrubber and wind turbine industries, as could clearly be seen from the reaction 
of the United States to China’s rising industrial prowess. Over the same period 
as China’s rapid growth was taking place, the United States also witnessed sig­
nificantly wider deployment. From 2004 to 2010, its SO2 scrubber capacity grew 
from 100 GW to 181 MW and its wind capacity from 6.8 GW to 40.3 GW (EIA, 


172  Environmental technology and industry
2012–2013). “The China Price” was a critical reason for trade disputes between 
China and the United States. In 2010, the price tag of SO2 scrubbers in China was 
about US$20/kW as revealed in the author’s fieldwork, whereas in the United 
States, it was US$206/kW (EIA, 2012–2013). For wind turbines, the average 
price in 2010 was US$700/kW in China and US$1,460/kW in the United States 
(Figure 7.9). However, China’s SO2 scrubbers barely made any news in trade dis­
putes between the two countries while those of wind turbines were highly visible 
(Cooper, September 28, 2012). From another perspective, the Chinese SO2 scrub­
ber industry did not contribute to international SO2 mitigation whereas its wind 
industry strengthened the global CO2 mitigation capability.
Despite the success in building up the supply capacity and achieving cost 
reduction, China’s large SO2 scrubber industry did not become competitive in 
the international market as indicated by the nearly tenfold price difference in the 
segregated Chinese and U.S. markets (Figure 7.2). Many SO2 scrubbers were of 
low quality, and this increased the operation and maintenance costs and shortened 
their lifetimes. Although the delayed improvement of the operation of SO2 scrub­
bers was critical for lowering the initial quality requirement and technological 
barriers to market entry, after 2007 when the normal operation of SO2 scrubbers 
was largely expected, the prices stayed low. The gap between 2002 and 2007 
was too long and China was trapped in a low-­
quality bottom. The huge quality 
0
200
400
600
800
1,000
1,200
1,400
1,600
1,800
0
5
10
15
20
)
W
k
/
$
(
 
e
c
i
r
p
 
e
n
i
b
r
u
t
 
d
n
i
W
Annual installation of wind turbines (1,000 MW)
United States (2004–2010)
China (2004–2010)
Figure 7.9  
Average prices of wind turbines in China and the United States
Source: IEA and ERI (2011); Wiser and Bolinger (2012); BP (2019); Xu (2013).


Environmental technology and industry  173
premium presented serious financial challenges to power corporations. In addi­
tion, the quality of SO2 scrubbers was quite opaque to investors, and only the SO2 
scrubber firms had the best knowledge of the product. In the five gap years, a race 
to the bottom had pushed the quality and price of SO2 scrubbers to reach a mini­
mum and stable level. Because no SO2 scrubber firm had established a reputation 
for quality, any significant price increase would put the firm in a disadvantageous 
position in competition. Even when China started to allow BOT (Build, Operate, 
Transfer) contracts for SO2 scrubbers to better integrate the decisions of capi­
tal investment and daily operation (NDRC and SEPA, 2007), the trap remained 
a difficult one to escape from. Another important reason for the segregation of 
the Chinese and U.S. SO2 scrubber markets lay in the restriction of technology 
licensors. Almost every major Chinese SO2 scrubber firm licensed and relied on 
foreign technologies that felt themselves constrained in the Chinese market (Xu, 
2011a). Even projects in Hong Kong required special permission from technology 
licensors.
However, the lower market-­
entry barrier at the early stage of wind energy 
development was still much higher compared to that of SO2 scrubbers. Although 
costs were much lower in China than in the United States, a race to the bottom on 
quality and price did not happen and the price of China’s wind turbines remained 
stable (Figure 7.9). The operational requirement never dropped to a bottom as in 
the SO2 scrubber case. One critical reason lied in their different regulatory foun­
dations. Although the enforcement capacity for the deployment and operation of 
SO2 scrubbers could be built on the existing regulatory system, the weak environ­
mental policy enforcement indicated that such a system had not been satisfactorily 
established in China. In comparison, the compliance monitoring system for wind 
electricity delivery had been largely established despite wind energy being a new 
energy type for electricity supply. Furthermore, because electricity generation has 
direct and significant economic benefits to local governments, the political will 
for greater demand and better management was much stronger than in the case 
of SO2 scrubbers. Because the poor operation or quality of wind turbines would 
affect wind electricity generation and thus the revenue, investors in wind farms 
value quality substantially more than those investing in SO2 scrubbers.
Despite the highly visible trade disputes between China and the United States, 
the actual trade in wind turbines was minimal. In 2011, the total capacity of 
exported wind turbines was equivalent to only 1.3% of that installed domesti­
cally (China Wind Energy Association, 2012). Although four Chinese wind tur­
bine manufacturers had been ranked among the largest ten in the world, unlike the 
other six as regional or global suppliers, they remained largely domestic (Li et al., 
2011). Besides other influential factors, one important reason could be the quality 
gap that made the Chinese wind turbines fail to reach the technological market-­
entry barriers in developed countries. However, the Chinese wind industry could 
have a promising future. If the price difference between China and the United 
States were taken as the upper limit of the quality premium or the depth of the 
quality trap, the wind industry would be much more likely to escape the trap than 
the SO2 scrubber industry.


174  Environmental technology and industry
As demonstrated in the two comparative case studies, the depth of the low-­
quality trap could be determined by how long the operational improvement of 
pollution control facilities is delayed. The delay should be long enough for the 
domestic supply capacity to become established but short enough to prevent a 
race to the bottom on quality and price. Another influential factor on the depth 
of the trap is how strong the initial enforcement capacity is. Because electricity 
generation corresponds to much stronger enforcement capacity than the mitiga­
tion of conventional pollutants, China could have a better chance to build inter­
nationally competitive industries for renewable energy that generally has to be 
converted into electricity. Low market-­
entry barriers for quality and technological 
advancement are a key factor to make the Chinese market and industrial develop­
ment vibrant. In the later upgrading, China could focus more on raising the corre­
sponding requirements but on keeping other barriers low to minimize the negative 
impacts of such enhancement.
4  
Inter-­
goal coordination under goal-­
centered governance
China’s Five-­
Year Plans feature multiple goals in several fields, including econ­
omy, social development, environmental protection and resource conservation. 
Goals on economic growth rates are always the first one in the goal table in each 
Five-­
Year Plan, while they have been listed as “expecting” since the 11th Five-­
Year Plan when goals were first differentiated between “expecting” and “binding” 
(National People’s Congress, 2001, 2006, 2011, 2016, 1996). Although goals on 
environmental protection have been gaining importance and become “binding,” 
the relationship between economic development and environmental protection is 
still crucial to profoundly affect the sustainability of the environmental political 
will and the achievement of environmental goals. One pivotal concern is how to 
coordinate various goals for maximizing their potential synergies and minimiz­
ing conflicts. SO2 mitigation and economic development have two-­
way impacts. 
First, SO2 mitigation is one constraint for economic development. Energy con­
sumption and economic growth are fundamental drivers of SO2 emissions, whose 
mitigation thus reversely becomes a limiting factor. Second, SO2 mitigation also 
relies on the emergence and development of a pollution removal industry to fea­
sibly provide the technological means of SO2 mitigation, which could create new 
jobs and economic opportunities.
Over the past four decades, central economic planning has also gradually 
shifted toward decentralized market evolution. Various local governments are also 
actively competing with each other in establishing local industries that can serve 
the huge national market. One key feature of the four-­
decade economic reform 
has been the gradual peeling of constraints on the market. The state-­
owned sec­
tor has been generally retreating and those remaining ones are more profit-­
driven 
than like governmental agencies. China’s economic reform has created many mar­
kets from a negligible basis after the Cultural Revolution and greatly enhanced 
the importance of the markets. The boundary between the state and the market has 
also become clearer.


Environmental technology and industry  175
China’s SO2 mitigation path as examined earlier surely has contributed to its 
SO2 mitigation goals. At the same time, new economic opportunities emerged and 
were generally seized, which should also have facilitated the advancement of eco­
nomic goals. In comparison with rule-­
based governance, goal-­
centered govern­
ance has resulted in much lower requirements on inter-­
goal coordination. Local 
governments in China are the primary, decentralized entities to bear the respon­
sibilities and incentives for achieving both environmental and economic goals. 
They can have greater flexibility in adapting their policies and actions to take the 
best advantage of changing situations.
These goals are also crucial indicators of how the Chinese central government 
balances between environmental protection and economic development. When 
economic goals were emphasized while environmental goals were not, local gov­
ernments primarily focused on achieving economic goals. These goals are not 
fully coordinated but generally are independently implemented in a bottom-­
up 
manner. They do not demand centrally planned coordination either, as shown pre­
viously in China’s surprising emergence of the SO2 scrubber industry. They will 
seek appropriate ways for balancing how they achieve both goals. Decentralized 
policies and market evolution may utilize unexpected opportunities and circum­
vent unexpected difficulties in a much better way than any intelligent central plan­
ner can foresee in advance. Goal-­
centered governance thus can better maximize 
synergies and minimize conflicts among various goals and government tasks.
Note
	
1	 Adapted with permission from Xu, Y. 2011. China’s functioning market for sulfur diox­
ide scrubbing technologies. Environmental Science & Technology, 45, 9161–9167. Cop­
yright (2011) American Chemical Society; and Xu, Y. 2013. Comparative advantage 
strategy for rapid pollution mitigation in China. Environmental Science & Technology, 
47, 9596–9603. Copyright (2013) American Chemical Society. Much has been revised 
and expanded on.
References
Arora, A. & Fosfuri, A. 2003. Licensing the market for technology. Journal of Economic 
Behavior & Organization, 52, 277–295.
Arora, A., Fosfuri, A. & Gambardella, A. 2001a. Markets for technology and their implica­
tions for corporate strategy. Industrial and Corporate Change, 10, 419–451.
Arora, A., Fosfuri, A. & Gambardella, A. 2001b. Markets for technology: The economics 
of innovation and corporate strategy. Cambridge, MA: MIT Press.
Arora, A. & Gambardella, A. 2010. Ideas for rent: An overview of markets for technology. 
Industrial and Corporate Change, 19, 775–803.
BP. 2019. Statistical review of world energy [Online]. Available: www.bp.com/en/global/
corporate/energy-­
economics/statisticalreview-­
of-­
world-­
energy.html.
Chan, L. & Aldhaban, F. 2019. Technology transfer to China: With case studies in the high-­
speed rail industry. PICMET 2009 Proceedings, Portland, OR, August 2–6, 2858–2867.
China Wind Energy Association. 2012. Statistics on China’s wind turbine installation in 
2011. Beijing, China: China Wind Energy Association.


176  Environmental technology and industry
Chinese Wind Energy Equipment Association. 2011. Chinese wind turbine generator sys­
tem selection manual (edition 2011). Beijing, China: China Wind Energy Association.
Cooper, H. 2012. Obama orders Chinese company to end investment at sites near Drone 
Base. New York Times, September 28.
Dean, T. J. & Brown, R. L. 1995. Pollution regulation as a barrier to new firm entry – initial 
evidence and implications for future-­
research. Academy of Management Journal, 38, 
288–303.
EIA. 1986–2006. Annual steam-­
electric plant operation and design data (Form EIA-­
767). 
Washington, DC: U.S. Department of Energy.
EIA. 2007–2011. Electric generator report data (form EIA-­
860). Washington, DC: U.S. 
Department of Energy.
EIA. 2012–2013. Electric power annual 2010–2011. Washington, DC: U.S. Department 
of Energy.
Gans, J. S. & Stern, S. 2010. Is there a market for ideas? Industrial and Corporate Change, 
19, 805–837.
Gu, X. 2004. A summary of installing SO2 scrubbers in Luohuang power plant and the 
impact on the society and environment. Annual conference of the Chinese Association 
of Science, Hainan, China.
Guizhou Qiandong Power Station. 2005. Tendering document for the flue gas desulfuri­
zation island. Zhenyuan, Guizhou [Online]. Available: www.in-­
en.com/power/html/
power-­
2006200604145584.html [Accessed January 27, 2011].
Harney, A. 2008. The China price: The true cost of Chinese competitive advantage. New 
York: Penguin Press.
Harvey, I. 2008. Intellectual property rights: The catalyst to deliver low carbon technolo­
gies. Breaking the Climate Deadlock, Briefing Paper. London: The Climate Group.
He, J. 2006. Pollution haven hypothesis and environmental impacts of foreign direct invest­
ment: The case of industrial emission of sulfur dioxide (SO(2)) in Chinese provinces. 
Ecological Economics, 60, 228–245.
IEA & ERI. 2011. China wind energy development roadmap 2050. Paris, France: IEA, ERI.
Lefohn, A. S., Husar, J. D.  & Husar, R. B. 1999. Estimating historical anthropogenic 
global sulfur emission patterns for the period 1850–1990. Atmospheric Environment, 
33, 3435–3444.
Levinson, A. & Taylor, M. S. 2008. Unmasking the pollution haven effect. International 
Economic Review, 49, 223–254.
Lewis, J. I. 2007. Technology acquisition and innovation in the developing world: Wind 
turbine development in China and India. Studies in Comparative International Develop­
ment, 42, 208–232.
Li, J., Cai, F., Tang, W., Xie, H., Gao, H., Ma, L., Chang, Y. & Dong, L. 2011. China wind 
power outlook 2011. Beijing, China: China Environmental Science Press.
Liang, W. 2001. Power equipment of the gigantic three Gorges project. Proceedings of the 
Fifth International Conference on Electrical Machines and Systems, 1, 676–678.
Liu, H. A., He, K. B., He, D. Q., Fu, L. X., Zhou, Y., Walsh, M. P. & Blumberg, K. O. 
2008. Analysis of the impacts of fuel sulfur on vehicle emissions in China. Fuel, 87, 
3147–3154.
Metz, B., Turkson, J. K. & Intergovernmental Panel on Climate Change, Working Group 
III. 2000. Methodological and technological issues in technology transfer. Cambridge 
and New York: Cambridge University Press.
Ministry of Education. 1999–2019. Education statistics yearbook of China. Beijing, China: 
Ministry of Education.


Environmental technology and industry  177
Ministry of Environmental Protection. 2008–2012. The list of China’s SO2 scrubbers in 
coal-­
fired power plants. Beijing, China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2011a. China’s capacities of water treatment plants, 
SO2 scrubbers and NOx removal systems at coal power plants. Beijing, China: Ministry 
of Environmental Protection.
Ministry of Environmental Protection. 2011b. Statistical data on the environment. Beijing, 
China: Ministry of Environmental Protection.
Ministry of Environmental Protection. 2014. The list of China’s SO2 scrubbers in coal-­
fired 
power plants. Beijing, China: Ministry of Environmental Protection.
Mitsubishi Heavy Industries. 2011. Delivery record [Online]. Available: www.mhi.co.jp/
en/products/pdf/delivery_record.pdf [Accessed January 27, 2011].
National Bureau of Statistics. 1996–2019. China statistical yearbook. Beijing, China: 
China Statistics Press.
National Economic and Trade Commission. 2000. Key planning points on flue gas des­
ulfurization technologies and their localization (2000–2010). Beijing, China: National 
Economic and Trade Commission.
National Energy Administration. 2010. A set of standards for wind energy in the energy 
sector. Beijing, China: National Energy Administration.
National People’s Congress. 1996. Outlines of the 9th five-­
year plan and long-­
term goals 
in 2010 for economic and social development of the people’s republic of China. Beijing, 
China: The 4th Conference of the 10th National People’s Congress.
National People’s Congress. 2001. The outline of national 10th five-­
year plan on economic 
and social developments. Beijing, China: The 4th Conference of the 9th National Peo­
ple’s Congress.
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. The outline of the 13th five-­
year plan on economic and 
social development. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
NDRC. 2008. The 11th five-­
year plan on renewable energy. Beijing, China: NDRC.
NDRC & SEPA. 2007. Working plan on experimenting BOT management of flue gas desul­
furization in coal power plants. Beijing, China: State Environmental Protection Admin­
istration, NDRC.
Niu, Y., Dong, L. C. & Chen, R. 2012. Market entry barriers in China. Journal of Business 
Research, 65, 68–76.
Ockwell, D. G., Haum, R., Mallett, A. & Watson, J. 2010. Intellectual property rights and 
low carbon technology transfer: Conflicting discourses of diffusion and development. 
Global Environmental Change-­
Human and Policy Dimensions, 20, 729–738.
Roth, A. E. 2008. What have we learned from market design? Economic Journal, 118, 
285–310.
Saggi, K. 2002. Trade, foreign direct investment, and international technology transfer: 
A survey. World Bank Research Observer, 17, 191–235.
SERC. 2006. Capital costs of electric projects completed in the 10th five-­
year plan. Bei­
jing, China: SERC.
SERC. 2011. Supervision report on wind electricity safety. Beijing, China: SERC.


178  Environmental technology and industry
Shen, J. 2008. Trade liberalization and environmental degradation in China. Applied Eco­
nomics, 40, 997–1004.
Shi, P. 2007. Statistics on China’s wind turbine installation in 2006. Beijing, China: China 
Wind Energy Association.
Shu, H. 2003. SO2 emission control for coal fired power plant. Electrical Equipment, 4, 
4–8.
Sina Finance. 2010. Information on Insigma Technology Co., Ltd. [Online]. Available: 
http://finance.sina.com.cn [Accessed February 9, 2011].
Srivastava, R. K., Jozewicz, W. & Singer, C. 2001. SO2 scrubbing technologies: A review. 
Environmental Progress, 20, 219–228.
State Council. 2006. The outline of national science and technology development plan in 
the middle and long term. Beijing, China: State Council.
State Intellectual Property Office. 2010. Database of patents granted in China. Beijing, 
China [Online]. Available: www.sipo.gov.cn/sipo2008/zljs/ [Accessed July 2, 2010].
Stigler, G. J. 1971. The theory of economic regulation. The Bell Journal of Economics and 
Management Science, 2, 3–21.
Strokova, V. 2010. International property rights index – 2010 report. Washington, DC: 
Americans for Tax Reform Foundation; Property Rights Alliance.
Teece, D. J. 1988. Capturing value from technological innovation – integration, strategic 
partnering, and licensing decisions. Interfaces, 18, 46–61.
United Nations. 1992. United nations framework convention on climate change. New 
York: United Nations.
U.S. Energy Information Administration. 2011. Official energy statistics from the U.S. gov­
ernment. Washington, DC [Online]. Available: www.eia.doe.gov/ [Accessed May 18, 
2011].
WIPO. 2019. WIPO statistics database [Online]. Available: https://www.wipo.int/edocs/
pubdocs/en/wipo_pub_943_2019.pdf.
Wiser, R. & Bolinger, M. 2012. 2011 wind technologies market report. Washington, DC: 
U.S. Department of Energy.
Xu, F., Yi, B., Zhuang, D., Yang, M., Yan, J. & Yan, Z. 2006. Survey report on the construc­
tion and operation of SO2 scrubbers at coal power plants in the 10th five-­
year plan. 
Beijing, China: The 4th Conference on Flue Gas Desulfurization Technologies.
Xu, Y. 2011a. China’s functioning market for sulfur dioxide scrubbing technologies. Envi­
ronmental Science and Technology, 45, 9161–9167.
Xu, Y. 2011b. Improvements in the operation of SO2 scrubbers in China’s coal power 
plants. Environmental Science & Technology, 45, 380–385.
Xu, Y. 2011c. The use of a goal for SO2 mitigation planning and management in China’s 
11th five-­
year plan. Journal of Environmental Planning and Management, 54, 769–783.
Xu, Y. 2013. Comparative advantage strategy for rapid pollution mitigation in China. Envi­
ronmental Science & Technology, 47, 9596–9603.
Xu, Y., Williams, R. H. & Socolow, R. H. 2009. China’s rapid deployment of SO2 scrub­
bers. Energy & Environmental Science, 459–465.


8	

Goal-­
centered governance
1  
Alternative governance models
China is experiencing very serious environmental damage. Nevertheless, the 
country in the past decade has achieved probably the fastest sulfur dioxide (SO2) 
mitigation pace for a large country. Significant progress has been made to clean up 
air and water. Its energy system has been gaining momentum to transition away 
from coal and toward renewables. With the economy more than 30 times bigger, 
SO2 emissions within one decade dropped to a level that was seen only before the 
economic reform era began in the late 1970s. Strong political will was formed to 
increasingly prioritize environmental protection among governmental affairs. The 
entire Chinese government across the central, provincial, municipality and county 
levels has been much better mobilized and committed. Policies are constantly 
enacted by various central and local authorities. The conventional poor policy 
implementation has been more effectively addressed and rapidly evolving to gain 
greater efficiency. In the coal-­
fired power sector, China managed to achieve essen­
tially universal coverage of SO2 scrubbers. More important, the original nonoper­
ation of SO2 scrubbers was also reversed to reach high SO2 removal rates. On the 
other hand, China established the largest SO2 scrubber industry, which provided 
employment and economic outputs. However, two decades ago at the early stage 
of China’s SO2 mitigation, few domestic firms existed with barely any domestic 
commercialized technologies. Although China has not been widely recognized 
by developed countries as a market economy, new firms were actively formed 
and swarmed into the new market to seek profitable opportunities. China’s inad­
equate protection of intellectual property rights did not seem to have prevented 
widespread market-­
based technology licensing from firms in developed countries. 
Despite numerous problems, China can claim great success in SO2 mitigation in 
the past two decades. These different components of environmental governance 
must work together to witness a favorable outcome. This book assesses the out­
come and, most important, aims to explain the trajectory.
Conventional wisdom can easily explain China’s environmental crises but has 
serious difficulties in understanding the cleanup process. Democracy and the 
rule of law are believed to be crucial contributors to forming strong political will 
and enabling the means to achieve pollution mitigation. However, China is not a 


180  Goal-­
centered governance
democracy and often ranked much behind developed countries in the rule-­
of-­
law 
index. Accordingly, we expect that China’s rapid economic growth will result in 
environmental crises and unacceptably high SO2 emissions, but the later, even 
faster SO2 mitigation is surprising because it defies the original expectations. 
China has not been fundamentally changed from the perspective of democracy 
and rule of law. The Chinese Communist Party is still the ruling political party in 
China. Governmental officials at various levels are still appointed but not demo­
cratically elected. Although certain progress has been made, Chinese society is 
still far from reaching the similar rule-­
based status as developed countries.
In one common conventional impression, the Chinese government is authori­
tarian and highly centralized with forceful central planning. Accordingly, in this 
theory, China’s environmental cleanup in the past two decades would be explained 
from the perspective of central planning. The central government might have 
designed the trajectory and its unchallenged authority could then implement such 
a design. This logic goes that the Chinese government does not have the checks 
and balances as in those democratic, developed countries, which enables China’s 
central planners to design an optimized path with good coordination among vari­
ous policy makers and implementers. When few domestic firms existed, the Chi­
nese government did not require the good operation of SO2 scrubbers to enable 
low technological market-­
entry barriers, provide and localize necessary supply 
capacities and reduce costs of SO2 mitigation. When many firms have been well 
established in the market, effluent emission standards and other regulations were 
made more stringent with better implementation for more effective SO2 mitiga­
tion. These newly emerged environmental industries provide economic opportu­
nities and cushion the negative impacts of stringent environmental protection on 
economic growth.
However, this explanation must assume that China’s central planners were 
extremely intelligent and well informed, but little evidence shows that such high-­
quality central planning has ever existed. As a developing country, China’s data 
collection system is less advanced than that in developed countries, especially two 
decades ago, to provide adequate data support for central planning. China’s com­
plexity and scale also make such high-­
level central planning intelligence impossi­
ble to achieve. The extreme centralization under Chairman Mao resulted in social, 
political and economic chaos with disastrous consequences. It is hardly convinc­
ing that central planning can lead to either rapid SO2 mitigation amid momentous 
economic growth or the establishment of a large SO2 scrubber industry.
Furthermore, the rule-­
based environmental governance that accounts for the 
trajectories in developed countries can also experience difficulties if applied to 
provide a primary explanation. As indicated in the World Bank’s governance 
indicators as well as in general impression, China’s performance has not been 
remarkable. China is still unable to make rules as important as developed coun­
tries prevalently do for environmental governance. In addition, under rule-­
based 
governance, although individual entities make their own decisions based on the 
rules, the rules are often centrally enacted by legislatures and/or courts as laws 
and the executive branch as regulations. Even if the rule of law is well established 


Goal-­
centered governance  181
in a society, whether rule-­
based governance can produce good outcomes depends 
on the quality of rulemaking. Rule-­
based governance alone is not a guarantee of a 
good outcome. Poorly designed rules and effective implementation may turn out 
to be undesirable, while policy making in China has not gained a decent reputa­
tion on its soundness, and consultation has also been much less thorough than 
that in developed countries. For example, before 1997, market speculation was a 
serious crime in China that was written into the Criminal Law. The intention was 
to maintain the order of a planned economy.
This book provides a different account of China’s environmental cleanup. 
China today has abandoned the Soviet-­
style central planning that was featured in 
the first three decades of the People’s Republic under the leadership of Chairman 
Mao. However, rule-­
based governance has not been well established. New laws 
and policies take a considerable amount of time to form and settle. For example, 
the Civil Code had just been enacted in May 2020 after many decades of grad­
ual formation. Instead, a new governance strategy has been tried and gradually 
become mature, with various goals taking the central stage. This goal-­
centered 
governance model is a mixture of centralization and decentralization to explain 
China’s SO2 mitigation trajectory much better than the central planning approach 
or rule-­
based governance can.
2  
Goal-­
centered governance
Readings of China are polarized, especially when China becomes bigger and 
more influential. One side profoundly denounces China and accuses the country 
of being messy, of not being a democracy, of having a rubber-­
stamp legislature 
and of being authoritarian without adequate respect to the rule of law. The Chinese 
government has been heavily criticized for breaking many rules that are highly 
valued in liberal democracies, such as those related to political liberty. Freedom 
of speech and civil society are constrained. Rising income inequality and privi­
leges of the wealthy and the powerful add social tensions. However, another side 
supports the Chinese government as they see many positive outcomes in China’s 
development. Together with rapid and sustained economic growth, the social wel­
fare system has been expanded dramatically to widen health care coverage even 
in rural communities, increase retirement pension and alleviate poverty. The Chi­
nese people can now enjoy living standards that were unimaginable one genera­
tion ago. They can largely choose where to live, work or travel as well as what to 
buy and sell. A great majority of the population has received significant returns 
of the economic development, although the distribution is uneven. Both views on 
China seem to have strong evidence to validate their claims. Then how can we 
understand China with these two sharply polarized readings? Are they connected? 
How China may further reform to embrace a better future?
For evidence-­
based researchers, the negative views on China could be mainly 
about rules and their implementation, while the positive views could be primarily 
shaped by outcomes. Although not all arguments on either side are sound, both 
views can find enough evidence to back them up. SO2 mitigation, or environmental 


182  Goal-­
centered governance
protection in general, is one government affair that exemplified such situations. 
The rapid mitigation was surprising but has been verified from multiple independ­
ent data sources, including external satellite data. Although active policy making 
and effective implementation were pivotal for achieving SO2 mitigation goals, 
many policies failed or were not implemented well. Initially, a large fleet of SO2 
scrubbers were built but not normally operating. In any understanding of China’s 
governance, a theoretical explanation should be able to accommodate both sides 
but not ignore the evidence of the other side. Furthermore, how are the two sides 
connected? In China’s case, does the favorable outcome have to be accompa­
nied by numerous policy blunders? If the rules were required to be well designed 
and implementable before putting into practice, would that affect the favorable 
outcomes?
This book explains China’s puzzles into a goal-­
centered governance model. As 
this book has examined in individual chapters on China’s SO2 mitigation, goal-­
centered governance has two foci, including goals and policies. Goals direct poli­
cies and policies achieve goals. Rule-­
based governance also has such two foci, but 
goals become secondary. The decisions in governance are mainly about enacting 
rules that are expected to be genuinely implemented. Fewer policies (or regula­
tions and laws) are enacted and the policy making might be more centralized, but 
they tend to be more carefully drafted. The outcome is an implicit product of such 
rules but not in the form of explicit, binding goals.
The goal-­
centered governance model can be understood from its organization 
mechanisms, features and applicability.
2.1  
Organization mechanisms
China has two hands in environmental governance, one visible and the other 
invisible. SO2 mitigation and environmental cleanup were achieved when the 
two hands cooperated. As a visible hand, the top leadership sets up prioritized 
goals with neither full-­
fledged deliberation nor stringent requirements on the path 
selection. The path results from bottom-­
up efforts of decentralized stakeholders 
as directed by an invisible hand of governance. The invisible hand of the market 
has been widely recognized and utilized. Rational market participants maximize 
their self-­
interests or profits, while this decentralized process also leads to the 
maximization of a society’s overall economic interest. Goal-­
centered governance 
could resemble and enable such an invisible hand to guide the central and local 
governments toward goal attainment. When their self-­
interests are served with 
various incentives for goal attainment, the overall goal will be achieved to sat­
isfy society’s overall interest. If more stringent goals are enacted, the incentives 
should also be strengthened. In order to finally achieve environmental cleanup, 
environmental goals must be prioritized with increasing stringency over a long 
period. If goals are changed, the invisible hand will direct the system away from 
the original goals and toward new ones.
As illustrated in Figure 8.1, goal-­
centered governance comprises three pillars: 
centralized goal setting, decentralized goal attainment and decentralized policy 


Goal-­
centered governance  183
making and implementation. First, the process for setting up goals of nationwide 
priority is highly centralized. The top leadership, with the Political Bureau of the 
Chinese Communist Party and its Standing Committee at the core, is in charge 
of supplying the country with goals as they deem crucial, especially in Five-­
Year 
Plans. The relationship among different goals could be balanced at this stage. 
Some goals could be prioritized that correspond to higher ratings in the perfor­
mance assessment of local leaders. In the case of SO2 mitigation, the Chinese top 
leadership did generally respond to what society wants, although the process was 
not democratic. The goals on SO2 mitigation and environmental protection were 
revised more stringent when such demand escalated.
Second, for decentralized goal attainment, national goals are distributed to 
provincial governments and then lower-­
level local governments, as in the case 
of SO2 mitigation and environmental protection goals. These individualized, 
quantitative goals guide the efforts of local governments and related ministries. 
Strong enough incentives are put into place to reward goal attainment and pun­
ish failures. Because China’s local leaders are appointed but not elected, their 
jobs are explicitly linked to the performance of achieving various goals with 
different priorities. The Chinese Communist Party’s organization plays a crucial 
role in establishing such a crucial personnel relationship between the central and 
provincial governments and their further subsidiaries. In addition, the central 
government receives much greater revenues than it spends, while the situation 
for local governments is generally the opposite: to demand a significant fiscal 
transfer from the central government. If local governments failed their individual 
goals, their leaders would face grim opportunities of promotion and could even 
be removed. Those who outperform others are distinguished for better promotion 
opportunities.
Third, policy making and implementation are heavily decentralized. With the 
responsibility of achieving goals, local governments have sufficient flexibility, 
authority and capacity for policy making and especially implementation, while 
the central government is especially weak in policy implementation. Require­
ments are significantly lowered on the quality of policy making, the optimal 
choice of policy instrument and coordination among policies. As a developing 
Policy Implementation
Policy Making
Local 
Governments
Pollution Control Firms
Polluting Firms
Localized Goals 
& Incentives
Society & Economy
Central Government
National Goals
Top Leadership
Centralized goal setting
Decentralized goal 
attainment
Decentralized policy 
making & implementation
Figure 8.1  
An illustration of the goal-centered governance model


184  Goal-­
centered governance
country, China has not acquired enough strengths from these perspectives despite 
continuous improvement. The weak rule of law indicates that the system neither 
requires nor ensures their genuine implementation. Policies compete with each 
other and evolve with implementation selection.
2.2  
Features
Under goal-­
centered governance, several key features could emerge.
First, not all goals are important and prioritized goals are few. The mobilization 
of the entire Chinese government, from central to local levels, depends on the cred­
ible incentives for their goal attainment performance. Any additional goal could 
dilute the effectiveness of existing ones. Accordingly, the number of nationally 
prioritized goals should be constrained, while provincial governments and central 
ministries may have their second-­
tier goals with lower priorities. Governmental 
efforts are highly concentrated on those goals of high priority, while in areas with 
lesser or no goals, the performance could be significantly compromised.
Second, policy making is active and each makes an incremental contribution 
to goal attainment. Local governments and central ministries are mandated to 
achieve their individualized goals. The incentives are mainly associated with the 
goals’ attainment, while any mistakes in policy making and implementation are 
much more leniently accommodated. Furthermore, they also have great authority 
and flexibility in policy making, adoption, innovation and learning in the decen­
tralized arrangement. These favorable conditions encourage active policy making, 
as witnessed in the case of SO2 mitigation. Because it is local governments but 
not their environmental protection bureaus that bear the responsibility of achiev­
ing goals, they often involve multiple bureaus in making their specialized policies 
that may contribute to SO2 mitigation. Ministry of Ecology and Environment, its 
predecessors and its composing departments, as well as other central ministries, 
have also been actively trying new policy tools. Unlike the situation in the United 
States that the Acid Rain Program in the Clean Air Act Amendments (1990) and 
its previous versions may claim a lion’s share of credits, China does not feature 
any pivotal policy of similar importance for SO2 mitigation, while SO2 mitigation 
goals were achieved through numerous policies and each contributed a small and 
accumulative share.
Third, more policy failures exist and policy implementation is selective. These 
may be seen as the necessary costs of the goal-­
centered governance model, espe­
cially when China is still in the process of strengthening its policy-­
making quality 
and policy implementation effectiveness. Policies in China may fail from multi­
ple perspectives. The design itself may be less mature and flawed. Decentralized 
policy making indicates that not all policy makers, especially those in local gov­
ernments, have adequate intellectual support. Policy implementation may have 
unexpectedly high obstacles from various interest groups or weak enforcement 
capacity. To ensure the faithful implementation of individual policies is only a 
secondary priority for local governments. When good implementation of a certain 
policy contributes significantly to goals, more efforts will be directed to this issue. 


Goal-­
centered governance  185
For SO2 mitigation, those policies on installing SO2 scrubbers were first targeted 
in implementation, while their operation was only made a priority later when the 
significant and growing fleet of SO2 scrubbers increased the impacts of such pol­
icy on reducing SO2 emissions. Environmental policy implementation capacity 
was strengthened, and new environmental compliance monitoring technologies 
were actively adopted with SO2 mitigation goals in primary focus.
Fourth, requirements on goal coordination are lower. With impacts on SO2 mitiga­
tion, industrial, energy and environmental policies are enacted generally indepen­
dently from each other for achieving their specific goals. Various policies for one or 
multiple goals could have synergies and/or conflicts. In goal-­
centered governance 
for SO2 mitigation, policy coordination largely is not centrally organized. Conflicting 
policies may not be implemented well to positively contribute to goal attainment, 
and thus, they would dwindle. Those compatible policies that have synergies will be 
expanded from local to national levels or adopted from one region to another. In other 
words, such policy coordination is not achieved primarily through intentional intel­
ligent design but via bottom-­
up evolution through implementation selection.
Fifth, requirements on information availability and measurability are lower 
with moral hazards better contained. Policy making and implementation are much 
more data-­
intensive than the assessment of goal attainment. Significant uncertain­
ties exist and many potential factors could affect the final outcome, such as in 
the case of SO2 mitigation. Because the efforts of local governments are difficult 
to accurately measure and sometimes hardly observable, local leaders in China 
may simply pay frequent lip service, emphasize constraints and external factors 
other than their own efforts but behave differently in reality. Comparison across 
regions then faces high hurdles to disable effective competition among local gov­
ernments. However, under goal-­
centered governance, goals are primarily on those 
measurable outcome indicators, such as SO2 emissions and air quality, which sig­
nificantly reduce the required information. Lip service is much less helpful than 
actual efforts for achieving goals.
Corresponding to the questions that are raised in the book, the coexistence of 
favorable outcomes and unfavorable policy pathways is only puzzling because 
they cannot be properly explained by the rule-­
based or central planning govern­
ance models, while a decent theoretical understanding can be reached with the 
goal-­
centered governance model. If the system has a very low tolerance for prob­
lems in policy making and implementation, especially for China as a developing 
country, the favorable outcomes might indeed be seriously compromised. Nev­
ertheless, the costs of policy deficiencies can be reduced when China gradually 
acquires the capability and capacity for high-­
quality policy making and effective 
policy implementation.
2.3  
Applicability
Since the Qin dynasty (221–207 BCE) first established centralized rule in China, 
local governments have always been crucial in Chinese governance to distinguish 
the importance of the central–local relationship. The vast territory and population, 


186  Goal-­
centered governance
as well as huge regional differences, weaken direct ruling by the emperors or 
prime ministers who reside in the distant capital. Although China has long been 
enacting laws and policies in texts, such as those by Shang Yang in a major reform 
in the 4th century BCE that led to the rise of the Qin Kingdom, the modern sense 
of the rule of law has never been well established to occupy the central stage of 
governance.
Corresponding to the organization mechanisms of goal-­
centered governance, 
the system may fail under three situations. First, the achievement of governmental 
goals does not lead to outcomes that the society wants. The supply of goals by 
the top leadership may have a lag or lead from the demand, but the gap should 
not be too wide to let the system fail. This concern is closely related to arguments 
in China’s context without democracy. When China was much poorer and the 
public prioritized economic growth and jobs over environmental protection, envi­
ronmental goals were ranked much lower than economic goals. When the public 
started to pay more attention to life quality and clean environment, environmental 
goals should then be ranked high among governmental affairs. It is not neces­
sary that the goals are exactly identical as what the society desires. For example, 
the maximization of long-­
term tax revenues may be compatible with improving 
the living standard of the public. After the Mongol empire under Genghis Khan 
occupied North China in early 13th century, one high-­
ranking official suggested 
eliminating all Han Chinese and using the land for grazing because Han Chinese’s 
primary economic activities were not raising animals. His goal was for the land to 
generate more tax revenues. Another key advisor to Genghis Khan, Yelv Chucai, 
proposed that if the Han Chinese could be left alive to still engage in agriculture 
and business, they would contribute much more tax. His advice was taken, and the 
outcome was favorable to both the Mongol court and the people.
Second, the decentralized goal attainment fails. The central government may 
not be able to impose their prioritized goals onto local governments. A frequent 
complaint in the Chinese government was that “policies and orders cannot go 
beyond Zhongnanhai.” Zhongnanhai, or “Central and Southern Seas,” is a com­
pound in Beijing where the central government of the People’s Republic of China 
is located. This sentence generally means that the central government cannot 
smoothly impose their policies and orders onto local governments. Even Chair­
man Mao complained before the Cultural Revolution that the Beijing municipal 
government was “penetrable by neither water nor needles.” Local governments 
and central ministries may malfunction or no effective incentives are available to 
incentivise or force them to work for their assigned goals. A long-­
lasting ques­
tion in the Chinese history is the collapse of the Ming dynasty (1368–1644) in 
early 17th century. Historians pointed out one crucial reason in Emperor Wanli 
(r. 1572–1620) when he left many key positions vacant and the government could 
not function (Huang, 1981). In the later decades of the Tang dynasty (618–907), 
local leaders had exclusive power over military, civil affairs and fiscal revenue. 
They could also pass their titles to heirs who were chosen by themselves. Essen­
tially, local governments were semi-­
independent kingdoms, which eventually led 
to the collapse of the Tang dynasty.


Goal-­
centered governance  187
Third, policy making and implementation are overcentralized, and local gov­
ernments have very limited flexibility or capability in choosing their own paths 
for achieving goals. One-­
size-­
fits-­
all rules from Beijing may be at a great distance 
from diverging regional realities to undermine their effectiveness and efficiencies. 
Active policy making, innovation and learning could be suppressed with overcen­
tralization or when mistakes were much less accommodated. When policy-­
making 
authorities, fiscal revenues/expenditures and capable officials are concentrated 
into the central government, local governments may be too weak to perform their 
jobs well. Local governments in wealthy regions may experience little difficulty 
in attracting capable employees or building enough capacity in policy making and 
implementation for achieving their goals. However, China has significant regional 
disparity in economic development. If left alone, poor regions would not be able 
to utilize the policy and technological tools effectively and efficiently.
Goal-­
centered governance is mainly for new and evolving governmental affairs 
without well-­
established policies. In comparison to two decades ago, China has 
designed, enacted and implemented many policies for SO2 mitigation and other 
environmental goals. Many will last to make SO2 mitigation a routine governmen­
tal affair, such as the effluent emission standards of thermal power plants. These 
tested policies and correspondingly strengthened implementation systems will 
form an escalating base for the continuous advancement of environmental protec­
tion until reaching fundamental solutions. Then goal-­
centered governance could 
gradually give way to rule-­
based governance and other governmental affairs may 
receive more attention with prioritized goals. In the past two decades, key envi­
ronmental goals in China’s Five-­
Year Plans have been extended from SO2 and 
chemical oxygen demand (COD) in the 11th Five-­
Year Plan (2006–2010), plus 
ammonia-­
nitrogen (NH3–N) and nitrogen oxide (NOx) in the 12th Five-­
Year Plan 
(2011–2015), plus water quality grade, the Air Quality Index and fine particulate 
matter (PM2.5) concentrations in the 13th Five-­
Year Plan (2016–2020; National 
People’s Congress, 2011, 2006, 2016). With the continuous progress, it will not be 
surprising to see that SO2 mitigation goal removed and an ozone (O3) goal added 
in the future, if not in the upcoming 14th Five-­
Year Plan (2021–2025).
This goal-­
centered governance has been tested as an effective strategy for China 
to make rapid advancement from unfavorable situations and to significantly lower 
key requirements on policy making as in a rule-­
based governance system. From 
one perspective, it is an effective and efficient path-­
finding strategy for China 
to reach a more sustainable, rule-­
based future. With new problems continuously 
emerging, it should and will be the crucial strategy in China’s future governance 
even when China reaches the stage of a developed country.
The goal-­
centered governance model may be best utilized in countries with 
the following characteristics: (1) newly prioritized governmental affairs or others 
with rapid evolution to require continuous focus; (2) developing countries where 
policies have not been maturely established and policy making has not achieved 
adequate quality and acquired enough data and intellectual support; (3) being 
large in scale with genuine necessity of multiple governmental levels and where 
the central government can impose adequate incentives on local governments to 


188  Goal-­
centered governance
encourage policy innovation, while goal evaluation is largely fair with good data 
support and rewards are issued based mainly on meritocracy; (4) where the system 
is more tolerant to mistakes in policy making and implementation and pays pri­
mary attention to outcomes and only secondarily on paths; and (5) local govern­
ments are capable of policy innovation and resourceful for policy implementation.
Countries in federal systems may not find this governance model applicable 
because incentives very likely are neither adequately available nor strong enough 
for the federal government to incentivize state governments. Small countries may 
not need this governance strategy as the central government is much closer to the 
society and local governments are not as important as those in large countries. For 
countries that have established sound rule of law, goal-­
centered governance may 
not occupy center stage either because the system is less tolerant of mistakes in 
policy making and implementation, while active policy innovation may indeed 
encounter more mistakes and failures. Highly centralized countries in policy mak­
ing and implementation may constrain such bottom-­
up efforts as well. This goal-­
centered governance model is not necessarily inapplicable in democracies, but 
the application nevertheless may be much constrained if competition across local 
governments may not have enough impetus and incentives.
Despite the constraints of its applicability, governments at various levels across 
countries with different institutional and developmental contexts may still be able 
to draw helpful insights from the goal-­
centered governance model and explicitly 
apply goals in organizing their governance. Decentralized policy innovation and 
competition can be encouraged in countries with sound rule of law, despite vari­
ous constraints of existing rules.
2.4  
Comparison with other theories
This study’s development of the goal-­
centered governance model not only ben­
efits immensely from earlier theoretical explorations but also demonstrates sig­
nificant differences.
Goal-­
setting theory in social psychology is one key intellectual source (Latham 
et al., 2008; Latham and Yukl, 1975; Locke and Latham, 1990, 2002; Locke et al., 
1981). The goal-­
setting theory mainly emphasizes on how goals could enhance 
task performance of individuals, while goal-­
centered governance pays primary 
attention to the performance of local governments, central ministries and other 
governmental agencies. In addition, the latter has a heavy focus on the flexibility 
of those decentralized stakeholders in utilizing policies for achieving those goals.
The goal-­
centered governance model can be regarded as a specific application 
of pragmatism with clear directions (Alford and Hughes, 2008), while it places 
goals at the center and makes policies instrumental. The criteria of assessing pol­
icies are based on whether they contribute, undermine or have no impacts on 
goal attainment in actual contexts but not on prior selection. Policy innovation, 
competition, revision, learning and expansion are common, and specific policies 
will rarely be unequivocally relied on. This governance model is a theoretical 
extension of Deng Xiaoping’s cat theory. Deng Xiaoping was officially accredited 


Goal-­
centered governance  189
as the “chief architect of China’s reform and open-­
up” by the Chinese Commu­
nist Party. However, he did not have a clear long-­
term blueprint on how China’s 
economic reform should proceed when China just got out of the devastation of 
the Cultural Revolution, but many doctrines remained strong. As summarized in 
his famous quote, “regardless of whether the cat is a white cat or a black cat, as 
long as it can catch mice, it is a good cat.” He was less interested in the debate 
about whether China’s economic reform may contain too much capitalism but 
mainly focused on whether the country can prosper at a faster pace. This strategy 
was sharply different from Chairman Mao’s, under whose leadership China had a 
stringent restriction on the choice of paths or “cats.” Another famous quote could 
summarize his main idea: “we would rather have socialistic grass than capitalistic 
grain.” This goal-­
centered governance has clear directions as specified in goals, 
but the pathfinding is much less constrained.
It also echoes adaptive and polycentric governance to address complexity and 
uncertainty that emphasize localized solutions (Dietz et al., 2003; Chaffin et al., 
2014; Ostrom, 2010). This goal-­
centered governance emphasizes more on how 
these solutions could evolve in decentralized and bottom-­
up manners with moti­
vated local governments under the centralized direction of goals. In comparison 
to the comparative advantage strategy that advocates good, incremental improve­
ments but not perfect, once-­
and-­
for-­
all solutions to environmental problems (Xu, 
2013), this goal-­
centered governance model is more incorporative to explain in 
what conditions the comparative advantage strategy will be taken, why it can 
work and what impacts it may exert on governance. The competition among local 
governments and other goal bearers borrows the idea from the Tiebout model 
(Tiebout, 1956), but they are also quite different. The incentives for the competi­
tion are not bottom up from local citizens but are top down from imposed goals. 
For explaining the development of China’s environmental industries, the ecologi­
cal modernization theory may provide an alternative understanding that connects 
environmental protection with economic modernization (Hajer, 1995; Zhang 
et al., 2007). The goal-­
centered governance model, in comparison, explains that 
the impacts on environmental industries were not intentionally planned, and envi­
ronmental and economic policies were not deliberately coordinated for new firms 
in a developing country like China to actively enter the market and grow up.
Incrementalism is another crucial intellectual source to build the goal-­
centered 
governance model (Lindblom, 1959; Lindblom, 1979). Neither emphasizes on 
key, deliberately designed policies with maximized impacts on achieving objec­
tives, but each policy should make incremental but accumulative contributions. 
However, goal-­
centered governance does have explicit goals at the center as ends, 
while policy making is not centralized for finding optimized means. Instead, the 
incremental improvement was made by decentralized local governments, not by 
centralized policy makers. Goal-­
centered governance is compatible with Joseph 
Stigler’s economic theory of regulation (Stigler, 1971). It understands the demand 
for regulations with an additional key source from goals, while the supply of 
regulations is decentralized to witness active policy making, innovation and 
competition.


190  Goal-­
centered governance
3  
Implications
In the past two centuries, China has tried, voluntarily or involuntarily, many dif­
ferent governance models. When one model was proved ineffective, reforms were 
attempted, and frequently, revolutions were started. Even under the rule of the 
Chinese Communist Party since 1949, China has tried sharply different models. 
Under Chairman Mao, the Chinese government was much more centralized. His 
goals significantly deviated away from what the society wanted, but no effective 
checks could counterbalance and prevent his goals from becoming the nation’s. 
The results were disastrous.
Through trial and error and with tremendous costs, China should have found 
an effective model to govern the vast, complex, developing country with a deep 
institutional history. The goal-­
centered governance model has demonstrated its 
effectiveness and efficiency in fundamentally reversing the rising trend of SO2 
emissions as well as China’s multifaceted environmental crises. Nevertheless, the 
governance model has two potentially highly damaging risks. First, goals may 
not be formed to satisfy society’s demands, like what happened under Chairman 
Mao. The current focus on environmental protection could have a chance to be 
disrupted, and thus, the entire governance system would be directed in another 
direction. Second, overcentralization and low tolerance to policy mistakes may 
undermine the system’s effectiveness and efficiency. Local governments and 
other governmental agencies may be weakened on the incentives, authorities and 
capacities of policy making and implementation. One indicator would be whether 
policy innovation and learning are still active.
A famous quote from Voltaire, a French writer, is that “the perfect is the enemy 
of the good.” The goal-­
centered governance model is far from being perfect. Even 
when it achieves great success, the process is full of stumbles, policy deficien­
cies, unsatisfactory policy implementation and even frequent abuse of govern­
mental authorities. However, as China has tried, alternative governance models 
may hardly provide better outcomes due to difficulties from uncertainties, com­
plexities and data inadequacy in China’s contexts, although they may work well 
in another country’s contexts. Rule-­
based governance demands high requirements 
on policy making quality, optimal choice of policy instruments and inter-­
policy 
coordination, but these were not China’s strengths especially in the early stages of 
dealing with major issues such as SO2 mitigation and environmental cleanup. This 
goal-­
centered governance is a “good” model but certainly not a “perfect” one due 
to its numerous weaknesses. Especially for developing countries with many diffi­
culties in policy making and implementation, this proven “good” model provides 
a promising way to organize governance for achieving what the society deems 
significantly desirable, while a “perfect” governance model may be unreachable. 
The pursuit of being perfect should not stop a country from becoming better.
Environmental crises that have accumulated over a few decades cannot be 
solved within a few years. Efforts should be sustained even when the govern­
ment changes after elections or leadership reshuffle. In developed countries, the 
rule-­
based governance model has been effective to achieve economic prosperity 


Goal-­
centered governance  191
and later sustained reduction of pollution with laws at the center. The gradually 
formed and tested goal-­
centered governance model offers a feasible method for 
China to fundamentally solve environmental degradation problems. The SO2 mit­
igation has transcended multiple Five-­
Year Plans since the 9th Five-­
Year Plan 
(1996–2000) under three top leaderships. The demand for environmental quality 
has grown stronger among the public, and China’s top leadership has also been 
largely supplying national goals to match the demand. It is expected that environ­
mental goals will remain highly prioritized among governmental affairs in China.
Climate change is a much greater environmental problem than any conven­
tional air or water pollution. This goal-­
centered governance model has also been 
used in tackling the mitigation of China’s greenhouse gas emissions since the 12th 
Five-­
Year Plan (2011–2015) when a goal to reduce carbon dioxide (CO2) intensity 
by 17% over the five years was first written into the national plan (National Peo­
ple’s Congress, 2011). Goal attainment, policy making and implementation have 
also been heavily decentralized. The market has been actively taking advantage 
of economic opportunities from CO2 mitigation to develop, deploy and innovate 
technologies, such as renewable energy, electric vehicles and energy efficiency. 
Similar to SO2 mitigation, CO2 mitigation has centralized goals, but its actual 
attainment is largely decentralized. It is expected that this goal-­
centered govern­
ance model will also lead to China’s eventual transition of climate mitigation.
References
Alford, J. & Hughes, O. 2008. Public value pragmatism as the next phase of public man­
agement. American Review of Public Administration, 38, 130–148.
Chaffin, B. C., Gosnell, H. & Cosens, B. A. 2014. A decade of adaptive governance schol­
arship: Synthesis and future directions. Ecology and Society, 19.
Dietz, T., Ostrom, E. & Stern, P. C. 2003. The struggle to govern the commons. Science, 
302, 1907–1912.
Hajer, M. A. 1995. The politics of environmental discourse: Ecological modernization and 
the policy process. Oxford and New York: Clarendon Press, Oxford University Press.
Huang, R. 1981. 1587, a year of no significance: The Ming dynasty in decline. New Haven: 
Yale University Press.
Latham, G. P., Borgogni, L. & Petitta, L. 2008. Goal setting and performance management 
in the public sector. International Public Management Journal, 11, 385–403, 113.
Latham, G. P. & Yukl, G. A. 1975. Review of research on application of goal setting in 
organizations. Academy of Management Journal, 18, 824–845.
Lindblom, C. E. 1959. The science of muddling through. Public Administration Review, 
19, 79–88.
Lindblom, C. E. 1979. Still muddling, not yet through. Public Administration Review, 39, 
517–526.
Locke, E. A. & Latham, G. P. 1990. A theory of goal setting & task performance. Engle­
wood Cliffs, NJ: Prentice Hall.
Locke, E. A. & Latham, G. P. 2002. Building a practically useful theory of goal setting and 
task motivation – A 35-­
year odyssey. American Psychologist, 57, 705–717.
Locke, E. A., Saari, L. M., Shaw, K. N. & Latham, G. P. 1981. Goal setting and task-­
performance – 1969–1980. Psychological Bulletin, 90, 125–152.


192  Goal-­
centered governance
National People’s Congress. 2006. The outline of the national 11th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2011. The outline of the national 12th five-­
year plan on eco­
nomic and social development. Beijing, China: The 4th Conference of the 10th National 
People’s Congress.
National People’s Congress. 2016. The outline of the 13th five-­
year plan on economic and 
social development. Beijing, China: The 4th Conference of the 10th National People’s 
Congress.
Ostrom, E. 2010. Beyond markets and states: Polycentric governance of complex eco­
nomic systems. American Economic Review, 100, 641–672.
Stigler, G. J. 1971. The theory of economic regulation. The Bell Journal of Economics and 
Management Science, 2, 3–21.
Tiebout, C. M. 1956. A pure theory of local expenditures. Journal of Political Economy, 
64, 416–424.
Xu, Y. 2013. Comparative advantage strategy for rapid pollution mitigation in China. Envi­
ronmental Science & Technology, 47, 9596–9603.
Zhang, L., Mol, A. P. J. & Sonnenfeld, D. A. 2007. The interpretation of ecological mod­
ernisation in China. Environmental Politics, 16, 659–668.


Page numbers in italic indicate a figure and page numbers in bold indicate a table on the 
corresponding page.
Index
3rd Five-Year Plan (1966–1970) 44
6th Five-Year Plan (1981–1985) 43
9th Five-Year Plan (1996–2000) 18, 62
10th Five-Year Plan (2001–2005) 45 – 48, 
62, 66, 92, 98
11th Five-Year Plan (2006–2010) 20, 
43 – 49, 57, 62, 64, 66, 67, 78, 92, 
98 – 99, 116, 168, 187
12th Five-Year Plan (2011–2015) 119, 
187, 191
13th Five-Year Plan (2016–2020) 187
Academy of Environmental Planning 29
Academy of Environmental Sciences 29
accountability 64
acid rain and SO2 pollution control, 11th 
Five-Year Plan on 43, 46, 93
Acid Rain Program, U.S. 49, 78, 79, 105, 
124, 184
Action Outline for Promoting Big Data 
Development 124
administration 28 – 29, 29, 30, 33
Africa, energy consumption and 
electrification rate in 89
agricultural pollution, under Ministry of 
Agriculture 27
air and water pollution: in China 1; 
controlling 65; DALYs in China due 3, 
3; mitigation 23; premature deaths due 
to 1, 2, 3
Air Quality Index (AQI) 65, 66, 71
ambient particulate matter (PM) pollution: 
cause of 86; in China 1 – 4, 2, 3; in India 
4 – 5, 4; PM2.5 goals 66
Asian financial crisis of 1997 18
Association of Environmental Protection 
Industries, China’s 151
autocracy, democracy and 6
Basic Thoughts of the National 11th Five-
Year Plan, The 45, 46
Beijing, AQI in 71, 71
Blackman, A. 107
“blue sky” 66
BOT (Build, Operate, Transfer) contracts, 
for SO2 scrubbers 173
budget balance: of central and local 
governments 34, 35; by provinces 35, 36
calcium/sulfur (Ca/S) molar ratio in coal 
93, 110
campaigns/movements (yundong) 
135 – 136
carbon dioxide (CO2); emissions 6; 
mitigation, goal of 43, 191
Central Department of Organization 40
central government: budget balance 34, 
35, 36; in charge of policy making 28, 
35; environmental authorities at 29, 33; 
governmental revenue and expenditure 
to GDP ratios by 33 – 35, 33; reforms 
at 27; shares of expenditures (2018) 
36 – 38, 37 – 38
centralized: and decentralized personnel 
management 39 – 40; goal setting 44 – 49, 
183; political will 17 – 18
China 25, 43; administrative reform 27; 
air quality 1, 2; average prices of wind 
turbines in 172 – 173, 172; central–local 
fiscal relationship (1994) 34 – 35; central 
planning 180; challenges in policy 
making 77 – 80; coal consumption in 
10, 11, 84, 87, 90; coal-fired power and 
SO2 scrubber capacities in 96 – 99, 97, 
98; DALYs in 3, 3, 4; deployment and 
operation of SO2 scrubbers in 149 – 150, 
150; economic growth in 85 – 87, 85; 


194  Index
economy 9, 18, 22; electricity 
generation by fuels in 89 – 90, 90; 
employees on environmental protection 
(2015) 33; energy consumption 43, 78, 
86 – 90, 86, 88, 92, 174; energy intensity 
goal 43, 78, 84, 92; environmental 
agencies in 28; environmental 
compliance in 105; environmental 
crises 1 – 5, 180; environmental 
policies and laws in 29, 31 – 32, 39; 
environmental/renewable energy 
industries 23; financial sector 18; firms 
in 170, 171; GDP in 7 – 8, 7, 18, 19, 22; 
goal-centered governance in 80 – 84; 
goal-centered policy implementation 
105 – 109; goals in Five-Year Plans 
42 – 44; governance indicators of 
8 – 9, 8; governmental income/
expenditure-to-GDP ratio in 33 – 35, 
33; international competitiveness of 
SO2 scrubber industry 171 – 174; job 
and demographic structures 18 – 20, 19, 
22; Law of Environmental Protection 
47; Law of Standardization 47, 96; 
laws in 31; leadership change 44 – 45; 
market-incentive policies 79; mobilizing 
government 42 – 72; NGOs in 17; 
patents on environmental technology 
161 – 162, 162; policies in 83; polity 
democracy index for 5 – 6, 5; pollution 
mitigation in 149; power sector shares 
10, 11; premature deaths 1, 2; provincial 
environmental authorities 30; R&D 
expenditures in 161, 161; shares in 
governmental expenditure for 36 – 39; 
strategies on environmental protection 
65; sulfur contents distribution in coal 
power plants 93 – 94, 95; unit capital 
costs of SO2 scrubbers in 151 – 152, 151; 
weak rule of law 32, 39, 107; 
wind energy development in 168 – 169, 
168; yearly university graduates in 
157 – 158, 157
“China Price, The” 172
Chinese Communist Party 13, 17, 25, 39, 
180, 183
Civil Code 181
Clean Air Act Amendments (CAAA, U.S.) 
12, 31, 49, 78, 79, 105, 184
climate change 26, 191
coal: consumption 10, 11, 84, 87, 90, 149; 
lower sulfur contents in 93; prices of 87, 
87; share in electricity generation 89 – 90
coal-fired power: annual growth of 98 – 99, 
98; decision scenarios for managers of 
plants 117; plants 95 – 100, 105 – 106, 
112 – 119, 123; and SO2 scrubber 
capacities 96 – 97, 97
competition, market entry and 169 – 171
compliance: costs 107, 109; monitoring 
119, 120 – 121; see also environmental 
compliance monitoring
compliance on SO2 scrubbers operation: 
noncompliance behaviors 112 – 115; SO2 
scrubber technologies 109 – 112
Comprehensive Plan on Ecological 
and Environmental Big Data 
Construction 124
Congleton, R. D. 6
continuous emissions monitoring systems 
(CEMSs) 107, 108, 113, 115, 122
corruption 6, 9
Cultural Revolution (1966–1976) 25
Darwin, C. 83
decentralization 27 – 28; in economic 
reform 36; fiscal revenue and 
expenditure 33 – 39; goal attainment 
61 – 65, 183; of governmental affairs 35, 
40; human resources 32 – 33; personnel 
management 39 – 40; of policy making 
31 – 32, 42, 82 – 83, 183 – 184
Decisions on Realizing Scientific View 
of Development and Strengthening 
Environmental Protection (2005) 46
deforestation 6, 7
democracy: and environment 5 – 7; and 
political will 17; and rule of law 13, 
179 – 180
Deng Xiaoping 188 – 189
Department of Organization of the Chinese 
Communist Party 27, 39
disability-adjusted life years (DALYs) 3; 
in China, due to air and water pollution 
3, 3, 4; in India 4, 4; premature deaths 
and 3
division of labor, for policy making and 
implementation 28 – 30
domestic technology licensees, strategy of 
156 – 162, 158
“double randomness, one publicization” 
scheme 129, 132
eco-compensation policy 31
ecological civilization 23
economic development: and energy 
conservation 84; and environmental 
protection 25, 27, 28, 174; and 
environmental quality 7, 51; Five-Year 
Plans 43; and SO2 mitigation 174


Index  195
effluent emissions: and SO2 removal rates 
97, 118; standards 31 – 32, 51, 66, 79, 
83, 95 – 96, 180, 187
electricity generation: annual growth of 90, 
91; energy consumption/transition for 
10, 89, 89, 92; by fuels in China 89 – 90, 
90; provincial 52; SO2 scrubbers 95, 
112, 117; wind 173
electrostatic precipitator (ESP) 109
employment and population structures, in 
China 18 – 20, 19, 22
energy consumption: annual growth of 87, 
88; economic conditions and 18, 20, 78, 
84, 174; electrification of 10, 89, 89, 92; 
and energy efficiency 86, 86; by fuel 87, 
88; reduction of 43
energy intensity goal 43, 78, 84, 92
energy transition effect 85, 92
Engineering, Procurement, and 
Construction (EPC) project, in Hong 
Kong 163
environment: and democracy 5 – 7; income 
and 6
environmental campaigns 135 – 136
environmental capacity 47, 51
environmental compliance monitoring 
105, 108 – 109, 119 – 136; building 
screening system with big data 123 – 125; 
comparing diagnosing and screening 
systems 125 – 135; conceptual model 
of 121; model construction 120 – 122; 
resilience of screening and diagnosing 
systems 135 – 136; strengthening 
conventional diagnosing system 122 – 123
environmental crises, in China 1 – 5, 180
environmental enforcement 27, 119
environmental governance 8; centralized/
decentralized personnel management 
39 – 40; evolution of environmental 
administration 25 – 27; for implementing 
political will 25 – 40; policy making 
and implementation 28 – 39; see also 
environmental protection
environmental impact assessment (EIA) 
reports 64, 94 – 95
environmental industry under goal-
centered SO2 mitigation path: 
international competitiveness of China’s 
SO2 scrubber industry 171 – 174; market 
entry and competition 169 – 171
Environmental Kuznets Curve 6, 7
Environmental Performance Index 1, 2
environmental policies 23
environmental protection: 11th Five-Year 
Plan for 43, 45; administration 27; 
authority of 25; as Basic National Policy 
25; as budgetary item 36, 37, 38; chain 
of command for 27 – 28; in China 25; 
economic development and 25, 27, 28, 
174; economic growth and 22, 23; goals 
on 43, 174; implementing 28; importance 
in new ideology establishment 46; 
personnel at governmental levels 28 – 29, 
29; political will for 17 – 24; prioritized 
42 – 72; provincial personnel 30; 
recognized as governmental affair 25, 
26; regulations on 31; SARS and 20 – 22; 
share in governmental expenditures 
36 – 38, 38; south–north water diversion 
project 27; strategies on 65; tax law 83; 
urban air quality and 23
Environmental Protection Agency (EPA), 
U.S. 113
environmental protection bureaus (EPBs) 
27 – 28
Environmental Protection Law 31
environmental quality 7, 47, 51, 65, 191
European Union Emission Trading Scheme 
119, 124
expenditures/revenue, of central and local 
governments 33 – 35, 33
financial sector 18
First National Conference on 
Environmental Protection (1973) 25
fiscal revenue and expenditure 33 – 39, 33
Five-Year Plans, goals in 42 – 44; see also 
individual plans
flue gas desulfurization see SO2 scrubbers
fluidized bed combustion (FBC) 93
foreign affairs and national defense 36
foreign technology licensors, strategy of 
162 – 165
fossil-fuel-fired power plants 65
fossil fuels 87, 89
fractions of sulfur retained in ash 93, 93
GDP (gross domestic product): capital 
investment and 64; in China, South 
Korea, Japan and US 7 – 8, 7; 
governmental revenue and 33 – 35, 33; 
growth rates of 18, 19, 22, 43, 85; R&D 
expenditures and 161
Genghis Khan 186
Gerlagh, R. 6
Global Burden of Disease study: China’s 
premature deaths due to air and water 
pollution in 1, 2; DALYs in China due 
to air and water pollution 3, 3
global financial crisis of 2008 20


196  Index
goal(s): in China’s Five-Year Plans 
42 – 44; in environmental protection 
43; evolution 65 – 72; implementation 
49 – 50; types of 65
goal attainment 43, 183; criteria for 
61 – 63; incentives for 63 – 65, 184
goal-centered governance: alternative 
governance models 179 – 181; 
applicability 185 – 188; characteristics 
187 – 188; in China 80 – 84; comparison 
with other theories 188 – 189; features 
184 – 185; illustration of model 
183; implications 190 – 191; inter-
goal coordination under 174 – 175; 
organization mechanisms 182 – 184
goal-centered policy implementation and 
supply 105 – 109; enabling 80 – 81; 
policy evolution by implementation 
selection 81 – 84
goal-centered SO2 mitigation path 
149 – 154; environmental industry under 
169 – 174; technology licensing under 
155 – 169, 158
goal distribution 43; from central to 
provincial governments 50 – 57; 
correlation coefficients of key factors 
for provinces 52, 53 – 54; provincial goal 
57, 61; from provincial to municipality 
governments 58 – 61; regression results 
to provinces/municipalities 56, 59
goal setting 43, 183; methods of 46 – 49; 
setting up national goal 44 – 46; in social 
psychology 188
governance indicators 8 – 9, 8
governmental revenue and expenditure to 
GDP ratios 33 – 35, 33
government effectiveness 9
grain storage 36
Great West Development 52
greenhouse gas concentrations, 
stabilization of 42, 50, 191
groundwater pollution, under Ministry of 
Land and Resources 26 – 27
Guangdong Province, distributing goals to 
municipality 59 – 60, 59
Guatemala 7
gypsum 111
Hainan Province 52
Harrington, W. 107
health care 36
Hebei Province 23, 59, 59, 68
Henan Province 116, 118
household air pollution from solid fuels 
1 – 4, 2, 3
Hu Jintao 18, 20, 21, 46
Human Environment, UN Conference on 
(1972) 25
human resources and fiscal expenditures 
32 – 39
IEA (International Energy Agency) report 
107
income and environment 6
incrementalism 189
India: ambient PM pollution in 4; energy 
consumption and electrification rate 89; 
governance indicators of 8 – 9, 8; market 
for technology 169; polity democracy 
index for 5, 6
indoor air pollution 1, 3
industrial and residential sectors 10
Insigma Technology 164
inspection 29 – 30, 29, 33
inter-goal coordination under goal-
centered governance 174 – 175
international competitiveness of China’s 
SO2 scrubber industry 171 – 174
International Monetary Fund 8
IPE (Institute of Public & Environmental 
Affairs) 21 – 22
Japan: economic growth in 85, 85; GDP in 7
Jiangsu Province 48, 59, 60, 106, 112, 116, 
153
Jiang Zemin 18, 44
Jiulong Electric 164 – 165
job creation 18 – 20, 22
Kenya 7
Kyoto Protocol 52
Law of Atmospheric Pollution Prevention 
and Control 31
Law of Environmental Protection 31, 47
Law of Standardization 47, 96
Law of Water Pollution Prevention and 
Control 31
Levitt, S. D. 121
Li Keqiang 17, 18, 22
limestone 109 – 112
liquid-to-gas ratio (L/G ratio) 110
local governments: achieving top-down 
goals 82; budget balance 34, 35, 36; 
environmental agencies in 28, 32; in 
era of Reform and Open-up 28; goal 
distribution 50 – 61; governmental 
revenue and expenditure to GDP 
ratios by 33 – 35, 33; implementing 
environmental policies 27; mobilization 


Index  197
of 43, 82; at provincial/municipality 
levels 32; responsibility for 
environmental quality 47, 65 – 66; shares 
of expenditures (2018) 36 – 38, 37 – 38
Locke, E. A. 42
major pollutants 77 – 78
Management Methods of Environmental 
Statistics 62
market: entry and competition 169 – 171; 
-oriented economic reforms 43; 
speculation 181; state and 26, 39, 174
Midlarsky, M. I. 6
Ming dynasty (1368–1644) 186
Ministry of Agriculture 27
Ministry of Ecology and Environment 
(MEE) 26 – 27, 28, 31, 44, 64, 77, 81, 184
Ministry of Environmental Protection 
(MEP) 26, 64, 67, 96, 106, 124, 163
Ministry of Land and Resources 26 – 27
Ministry of Water Resources 27
mitigation effect 85, 92
monitoring 29 – 30, 29, 33
National Acid Precipitation Assessment 
Program 49
National Aeronautical and Space 
Administration 124
National Development and Reform 
Commission (NDRC) 26, 45, 96
National Energy Administration 96, 169
National Environmental Protection 
Administration 47
National Party’s Congress 44
National People’s Congress 13, 31, 32, 44, 
45, 77
Neumayer, E. 6
noncompliance, reversing: environmental 
compliance monitoring 119 – 136; 
penalty 115 – 119
noncompliance behaviors, on SO2 
scrubbers operation 112 – 115
nonfossil fuels 90
nongovernmental organizations (NGOs), 
in China 17
nonhydro renewables 89
non-power-sector emissions 51
nonstate firms 156
“Not Invented Here” syndrome 159
Obama, B. 42
ocean environment, under State Oceanic 
Administration 27
oil and natural gas 87, 89
Open-up policy 9, 25
organization mechanisms, of goal-centered 
governance 182 – 184
Outline of the National 11th Five-
Year Plan on Economic and Social 
Development, The 45, 46, 77 – 78
ozone pollution 69, 70, 71
patents on environmental technology 
161 – 162, 162
Payne, R. A. 17
Pellegrini, L. 6
penalties for noncompliance 108, 115 – 119
People’s Republic of China; see China
personnel management, centralized/
decentralized 39 – 40
policing strategies 121
policy making and implementation: 
challenges in 77 – 80; compliance on 
operation of SO2 scrubbers 109 – 115; 
decentralized 31 – 39, 183; division 
of labor for 28 – 30; environmental 
compliance monitoring 119 – 136; goal-
centered 105 – 109; lower barriers 81 – 82; 
overcentralized 187; penalty 115 – 119; 
reversing noncompliance 115 – 136
political stability and absence of violence/
terrorism 9
political will 6; centralized 17 – 18; 
economy/jobs/environment (1998–2002) 
18 – 20, 19; for environmental protection 
17 – 24; SARS and prioritization of 
environmental protection (2003–2012) 
20 – 22; sustainability of (2013–present) 
22 – 24
pollution: abatement costs 120; health 
impact, measurement of 3; mitigation 
149; ozone 69, 70, 71; see also air and 
water pollution; ambient particulate 
matter (PM) pollution
power sector: shares of coal consumption 
and SO2 emissions 10, 11 – 12, 12, 59, 
60, 90; technological factors for effluent 
SO2 emissions in 92 – 95, 97
premature deaths: DALYs and 3; reduction 
and causes of 1, 2
provincial governments, on policy making 28
public: in democracy 6; health, goal for 
protecting 65
pulverized coal (PC) combustion 93
Qin dynasty (221–207 BCE) 185
Qinghai Province 52
Rebels (zao fan pai) 25
Red Guards (hong wei bin) 25


198  Index
Regional Supervision Bureaus 30
regulation, economic theory of 189
rent-dissipation effect 162
research and development (R&D) 
expenditures 161, 161
revenue: and expenditures, of central and 
local governments 33 – 35, 33; effect 162
Ricardo, D. 108
rule-based environmental governance 
180 – 181
SARS and environmental protection 
(2003–2012) 20 – 22
science and technology 36
Scientific View of Development 21, 63
sectoral employment changes and GDP 
growth rates, across China 18, 19
Shanghai: revenue–expenditure gap for 35; 
SO2 emissions 48, 52, 55
Shang Yang 186
Shanxi Province, distributing goals to 
municipality 59, 60 – 61
Shenzhen, AQI in 71, 72
Shijiazhuang: daily O3 concentrations in 
69, 70, 71; daily PM2.5 concentrations in 
68 – 69, 69; daily SO2 concentrations in 
68, 68; monthly average AQI in 
69, 70
Singapore, polity democracy index for 5, 6
SO2 (sulfur dioxide) emissions: in 9th 
Five-Year Plan 18, 62, 90, 92; in China 
9 – 10, 10 – 12, 12; controlling 48; daily 
SO2 concentrations, in Shijiazhuang 
68, 68; decomposition of 90, 91; 
designated intensity, in coal power 
plants 58, 58; economic growth and 84; 
emission mitigation goals of 67 – 69, 
83; environmental capacity for 47; goal 
implementation 50, 58, 63; intensity of 
electricity generation 96; key factors 
for 84 – 92; mitigation of 10, 12, 20, 
43, 51, 65, 67 – 68, 72, 77 – 78, 84, 
181 – 182, 185, 191; policy scope for 
achieving mitigation goals 84 – 100; 
in power and nonpower categories 51, 
63; reduction goal of 43 – 44, 45 – 46, 
48 – 49, 59 – 60; regulations 44, 66, 118; 
removal efficiencies/rates 51, 57, 94, 
118; by sector 10, 11; setting up goals 
47; technical measures for 95 – 100; 
technological factors for 92 – 95; 
underestimation of 112; in United States 
12, 12
SO2 mitigation path: and economic 
development 174; environmental 
industry under goal-centered 169 – 174; 
goal-centered 149 – 154; model 
projection of 154; technology licensing 
under goal-centered 155 – 169, 158
SO2 scrubbers: BOT contracts for 173; 
capacities 95, 97 – 99, 97 – 100, 171; 
capital costs of 112, 115, 151 – 152, 151; 
categories 99; coal-fired power and 79, 
97, 97, 98, 99, 106; compliance costs of 
109; compliance on operation 109 – 115; 
data, in China’s coal-fired power plants 
114; deployment and operation of 
149 – 151, 150; designing 93; economies 
of scale and 111 – 112; effluent discharge 
fee 115 – 116; electricity-consuming 
components of 111; firms 156; 
geographic distribution of 99; goals 
and policies in compliance decisions 
on operation 119; installation 94 – 96, 
95, 106 – 107, 151, 159, 171, 185; 
international competitiveness of industry 
171 – 174; noncompliance behaviors on 
operation 112 – 115; nonoperation of 
116 – 117; O&M costs of 111 – 112, 115, 
151, 152; operation in Jiangsu Province 
106, 106; planning 51, 55; product of 
111; reduction of emissions through 
51, 85, 96, 105; technologies 100, 
101, 109 – 112, 156; see also reversing 
noncompliance
“Socialistic Thoughts with Chinese 
Characteristics in the Xi Jinping Era” 23
social psychology 42
social welfare system 181
South Korea: GDP in 7; polity democracy 
index for 5, 6
south–north water diversion project’s 
environmental protection 27
Standing Committee of the Political 
Bureaus 21
state and market 26, 39, 174
State Council, 1998 reform of 26
State Environmental Protection 
Administration (SEPA) 26, 44, 45 – 46, 
50, 58, 64, 78, 98
State Environmental Protection Agency 
(1984) 25, 26
State Oceanic Administration 27
state-owned enterprises/firms 18, 156
Steinfeld, E. S. 112, 113
Stigler, J. 189


Index  199
Suggestions on Designing the National 
11th Five-Year Plan, The 45, 46
sulfur contents 94, 111; coal consumption 
and 63, 78; control of 85; distribution in 
coal power plants 93 – 94, 95; see also 
SO2 entries
suspension policy 64
sustainability of environmental political 
will (2013–present) 22 – 24
Tang dynasty (618–907) 186
tax compliance 121
technology licensing under goal-centered 
SO2 mitigation path 155 – 169; criteria 
of effective market design 165 – 167; 
strategy of domestic technology 
licensees 156 – 162, 158; strategy of 
foreign technology licensors 162 – 165; 
technology market emerging in China, 
reasons for 165 – 169
thermal contents of coal 111
“three representativeness” 20
Tibet: governmental revenue/expenditures 
35; SO2 emissions 48, 52
top-down goal distribution 49 – 61
Total Emission Control regime 66
unemployment 20
UNFCCC (United Nations Framework 
Convention on Climate Change) 42, 50
United States: average prices of wind 
turbines in 172 – 173, 172; CEMSs cost 
in 113; Clean Air Act Amendments 
(1990) 12, 31, 49, 78, 79, 105, 184; 
coal consumption 12; deployment and 
operation of SO2 scrubbers in 149 – 150, 
150; economic growth in 85, 85; energy 
consumption and electrification rate 89; 
Energy Information Administration 152; 
Environmental Protection Agency (EPA) 
113; firms in 170, 171; GDP in 7; 
governance indicators of 8 – 9, 8; patents 
on environmental technology 161 – 162, 
162; polity democracy index for 5, 6; 
power sector shares of coal consumption 
and SO2 emissions 10, 11; SO2 emissions/
intensities in 12, 12, 49; unit capital costs 
of SO2 scrubbers in 151 – 152, 151; wind 
energy development in 168 – 169, 168
university-established firms 156
unsafe water/sanitation/handwashing 1, 
2, 3, 4
“Upgrading and Retrofitting Action Plan 
for Energy Conservation and Pollution 
Mitigation in the Coal-Fired Power 
Sector” policy 96
veto 64
Wang Xinfang 45 – 46
Wanli (Emperor) 186
water: consumption 112; environment 
management 27; pollution, health 
impacts of 3; see also air and water 
pollution
Wen Jiabao 18, 20, 21, 46
wind: energy development 168 – 169, 168, 
173; and solar energy 89; turbines, 
average prices of 172 – 173, 172
Winslow, M. 6
World Bank 8, 31, 180
World Trade Organization in 2001 20
Xie Zhenghua 64
Xi Jinping 17, 18, 22
Yelv Chucai 186
Zhejiang Province 62
Zhongnanhai (Central and Southern 
Seas) 186
Zhu Rongji 18