Presented at the workshop "Smoke and Mirrors: Air Pollution as a Social and Political Artifact," Center for Global, International, and Regional Studies, UC Santa Cruz, Santa Cruz, CA, 11-12 January 2002

DETERMINANTS OF AIR POLLUTION MANAGEMENT IN URBAN CHINA Guodong Sun 1 and H. Keith Florig Department of Engineering and Public Policy Carnegie Mellon University Pittsburgh, Pennsylvania 15213

Abstract Most of current literature on air pollution policy focuses on the industrialized world. Thus, theories explaining air pollution behavior have been espoused and examined mostly within a narrow range of political, social, cultural, and economic systems. To contribute to a broader understanding of the forces that drive air pollution policy, we report here on a study of the factors that have influenced air pollution in urban China from the 1970s to 1990s. Our analysis is based on a series of interviews conducted with officials of local environment bureaus in five cities with a wide range of population size, per capita income, heating demands, and fuel availability. Air pollution in urban China did not receive much official attention until the mid-1970s, when the consuming political campaigns of the Cultural Revolution had subsided and public health authorities could give proper study to the health impacts of various pollutants. At that time, ambient TSP levels exceeding 1000 µg/m3 were not uncommon in Chinese cities, a consequence of using coal to meet 80% of primary energy needs, with virtually no pre- or postcombustion treatment to reduce emissions. Among the earliest actions taken to curb emissions were the addition of particulate removal devices to boilers used for residential heating and industrial processes. To reduce emissions from household cook stoves, officials encouraged the use of briquettes made from processed coal containing less sulfur and producing less fly ash than raw coal. In contrast to particulate control programs in western countries, we find that impetus for these changes came largely through top-down government-sponsored campaigns, with little or no involvement of citizen groups or the news media. Evidence suggests that particulate controls in China were implemented at significantly lower levels of per capita income than in countries with less authoritarian political systems. As China continues its transition to a market economy, 1

Address correspondence to Guodong Sun, Dept. of Engineering and Public Policy, Carnegie Mellon University, Pittsburgh, PA 15213, Phone: 412-268-4640, Fax: 412-268-3757, Email: [email protected]

1

new forces for particulate control are coming into play. These include cost-minimizing behavior by industry, pollution levies, non-subsidized fuel prices, and a transition to gaseous household fuels made possible by increasing wealth.

1. Introduction Chinese environmental and public health authorities are well aware of the serious health consequences of polluted air. In the past two decades, they have persuaded the Chinese leadership to invest an increasing proportion of GDP in environmental controls. But China is not a wealthy country, and political forces for economic development are a strong counterweight to proponents of better environmental quality. Since the beginning of economic reform in 1978, many Chinese cities have experienced a period of deteriorating air quality followed more recently by a stabilization, or in some cases, improvement in air pollution levels. Understanding the various economic, political, and institutional forces that guided China down this path is important, both to advance the science of environment and development, and to illuminate future Chinese decision-making on urban air pollution management. There is a substantial literature on environmental policy development and implementation in both industrialized and less-industrialized countries. This literature describes a number of theoretical frameworks for explaining the evolution of environmental behavior. In this paper, the extent to which these existing theories can explain patterns of air pollution management in China is explored. The primary focus is to explain how and why local environmental authorities undertake urban air pollution control. This question is addressed through a series of case studies on control of particulate matter (PM) pollution in five Chinese cities. 1.1 Literature review The environmental economics literature is replete with theories concerning the forces that drive environmental policy and the relationship between environmental quality and economic development. These theories differ in the emphasis that they place on economic, political, and organizational dimensions. Here, some of the most prominent theories are reviewed, and the extent to which each might apply to air pollution management in China is discussed. A number of descriptive studies have noted an inverted U-shaped relationship between measures of environmental quality and per capita income (Selden and Song 1994; Stern, Common et al. 1996; Agras and Chapman 1999). The explanation most often advanced for these "environmental Kuznets curves (EKC)" is that economic activities in the early stages of economic development increase pollution levels. Environmental quality continues to deteriorate with growing per capita income up to the point at which growing preferences for environmental quality shift marginal societal resources to environmental protection rather than to economic

2

growth. Most of the studies that have examined this behavior report turning points in the EKC at per capita incomes of a few thousand U.S. dollars. Such sweeping explanations, however, are not particularly edifying. They provide little insight into the political and organizational forces that drive particular policies, especially at the local level. There are several political models that have been developed to explain the generation of environmental legislation and regulation. The "pluralist/interest group framework" emphasizes the impact of interest groups on policy making. Within this framework, legislation is regarded as the product of bargaining among interest groups, and public policy reflects the equilibrium reached after struggles among interest groups. (Bailey 1998). But this framework overlooks the influence from government officials in determining legislative outcomes. A major revision of the pluralist/interest group framework is to incorporate governmental officials as political actors. However, this revision still fails to recognize fully the unique features of government officials as political actors. Government institutions offer their officials with endowment and power to make authoritative decisions (Bailey 1998). This distinguishes government officials from other political actors. The "institutional rational choice framework" places government institutions at center stage. This theory posits that politicians and bureaucrats are the key actors in determining the supply of legislation, and that their decisions are based on maximizing the power and survivability of the institutions in their charge. Mayhew (1974) depicted U.S. Congressmen as "single-minded seekers of re-election." Pursuit of re-election will, based on information from "position-taking," drive legislators to support or supply legislation that conforms with the expressed concern of the "attentive public" or from the potential preferences of the "inattentive public." Although reelection theory clearly explains some short-term phenomena, the theory cannot fully explain the introduction of legislation with long-term benefits but short-term costs (e.g., air pollution legislation, under which most of the costs are incurred immediately but the benefits accrue over many years), nor does reelection theory consider the "opportunity costs associated with policy leadership," i.e. legislators pursuing reelection are not willing to invest their time and energy into issues of less public concern. Bailey (Bailey 1998) proposed another conceptual framework, which "treats law-making as a form of production," to explain the supply and content of air pollution control policies from the U.S. Congress. “Legislative entrepreneurship" is determined by the costs and benefits of producing legislation. The benefits encompass a variety of rewards to legislators' goals ranging

3

from personal money income (corruption for instance) and security of current position to pursuit of prestige and personal satisfaction. Time and resources invested in developing a policy constitute the major parts of the cost. According to this theory, legislators endorse a specific policy only when the policy will yield a positive "profit." However, to define and measure the "profit" is a significant challenge. Public involvement is the major ingredient of many of these political models. Given the limited role of public involvement in China, the applicability of these theories to China may be limited. A number of historical studies of the air pollution control in the United States provide a broader perspective. Before environmental control was federalized in the early 1970s, there were mainly two players in the battlefield along with a hesitant government that favored economic growth: the public that demanded a better environment and the resisting polluters that were reluctant to sacrifice economic benefit. Until the mid to late 20th century, the federal government played a very limited role in protecting environmental quality since they viewed environmental quality as a local problem (Andrews 1999). Meanwhile, the pursuit of economic growth and strong preferences for acquisition of the material benefits of industrialization provided little incentive for local governments to take effective measures to solve smoke problems. The generation and implementation of environmental policies were mainly driven by the demand for good air quality from the public, especially from some highly concerned and involved groups (women's groups, civic groups, and some engineering societies). Each groups' contribution to the anti-smoke movement was different. Women's groups mounted campaigns to raise public awareness. Civic groups (such as Chambers of Commerce) helped in the formulation of an antismoke ordinance. Engineers provided technical options to solve smoke problems. (Grinder 1980; Tarr and Zimring 1997) It wasn't until 1963 that the U.S. federal government finally moved to develop national air pollution legislation and establish institutions tasked with managing the national air pollution problem. Several factors have fostered this change. One of the most important factors was the demand from key industries for moderate and uniform federal standards to avoid more stringent and inconsistent state and local standards, which were pushed by many local civic and conservation groups. Furthermore, public interest in a national air pollution policy was generated by the rise of middle-class families with stronger environmental awareness, and the realities that state and local governments were incapable of solving deteriorating air pollution problems. (Andrews 1999) Local flexibility in setting up environmental goals and institutions is important because of diversity in social, economic, and physical circumstances (Ringquist 1993; Sabatier and JenkinsSmith 1993). But the full benefit of local flexibility can be realized only if the policies of local

4

government are determined by a democratic process and the public's right of free speech is respected and protected by law. An open and competitive electoral process pressures elected officials to respect voters' interests. It further requires that voters (the public) have good awareness of environmental damage and have the willingness to organize advocacy groups to lobby government. Support from higher levels of government could play an essential role in helping local authorities to overcome some barriers, for example lack of technical and organizational capacity, or limited political clout in facing down major polluters. 1.2 Research questions Many of the favorable factors for the generation and enforcement of environmental regulations in the industrialized democracies, such as the United States, do not exist in China. On a national basis, China has yet to reach the level of economic development at which the environmental Kuznets curve literature would predict that environmental degradation will be reversed. China's investment in environmental protection did not exceed 1% of GDP until 1997 (China Environment Yearbook Committee 1998). Economic growth has always had higher priority than environmental pollution control issues on China's development agenda (Edmonds 1998). In the last two decades, economic growth has been one of the most important factors determining local governmental leaders’ performance evaluations, which are conducted by upperlevels of government and influence their future promotion. Furthermore, local government leaders control the budget, personnel, and more importantly, the appointment and promotion of the director of local environmental authority. The result is that local environmental authorities have not been willing to openly challenge economic growth (Chan et al. 1995). The development of environmental regulations in the United States has often been the product of intense public involvement. Such a tradition does not exist in China (Tang, Lo et al. 1997). China's political system does not accommodate public involvement, public awareness of environmental damages is low, and the public largely defers to government "experts." All these factors work against any organized and meaningful public involvement in formulating environmental regulations along the lines of an industrialized democracy. The adverse effect of low public involvement is amplified by the considerable flexibility that local governments have in defining local implementation plans under China's top-down management system. Designing policies to suit local conditions is a common practice in China. A widely known dictum is "to take such measures as are suitable to local conditions." The lack of democratic processes can lead local governments to abuse this privilege by taking a soft position against violators in a way that best serves either bureaucratic goals (e.g., budget growth) or the personal goals of officials (e.g., career advancement, bribery).

5

The existence of these unfavorable factors run counter to the fact that ambient PM levels have declined in most Chinese cities since the late 1980s. It raises a number of interesting and provocative questions about how a society protects the environment when many of the factors that have promoted the emergence of control in western democracies are absent. It is not the intention of this study to develop a new theoretical framework to explain China's air pollution control history. However, several guiding questions are explored in the analysis: •

What was the influence of a top-down environmental management system on policy implementation at local levels?

•

How have local environmental regulations and enforcement evolved with economic growth?

•

What are the major institutional and technological factors that have influenced policy implementation at local levels, and how?

2. Research design and data collection 2.1 Case-study approach A case-study approach was employed in this study. Five cities are chosen for detailed analysis. The five are chosen to span a wide range of economic development, population size, and energy (e.g., coal) availability. As a practical matter, the personal connections of one of the authors (Sun) with local Environmental Protection Bureaus (EPBs) also determined which cities were selected for study. The five cities studied are Hohhot, Benxi, Binzhou, Nantong and Taiyuan. Table I gives a comparison of these five cities. China's efforts to control urban air pollution have targeted three problems: particulate matter pollution from coal combustion; pollution caused by vehicle emissions, and SO2/acid rain. However, the allocation of resources to these three problems has been quite different in both quantity and timing. Particulate matter (PM) pollution was the first to get national and local attention, and has over time attracted the most resources. The major strategies adopted to solve the PM problem have included: •

retrofitting coal combustion facilities, and using/upgrading particle control devices;

•

promoting diffusion of coal briquettes;

•

establishing a local "Control Zone for Combustion Smoke" (CZCS or Control Zone hereafter);

•

developing district and central heating systems; and

•

substituting coal with gaseous fuel for cooking.

6

This study focuses only on the first three programs. This is because the role of local environmental authorities in the last two programs is no more than assisting the leading agencies such as the Planning Commission and/or Urban Development and Construction Bureau. Table I. Comparison of the five case-study cities Hohhot

Benxi

Taiyuan

Nantong

Binzhou

Urban Residents (Thousands, 1996)

700

810

1,700

430

190

Per capita non-agriculture GDP (1000 Yuan, 1996)a

9.4

11.3

11.4

20.6

12.9

Easy

Easy

Easy

Difficult

Medium

Heating period (months per year)

~6

~6

~6

0

3-4

Share of secondary industries in GDP (%)

47

66

54

65

41

Average annual PM level in 1997 (TSP, µg/m3)

448

550

504

219

348

Terrain & weather for dispersion of air pollutants c

Poor

Poor

Poor

Good

Medium

Coal Availability b

a: 1

Yuan ~ $0.12 USD

b:

based on the distance from the city to coal mines; easy = all demands for coal can be met locally, medium = no local coal mine while needs can be met by mines within a few hundred kilometers, difficult = coal is mainly supplied by mines that are located beyond a thousand kilometers away.

c:

Taiyuan, Hohhot and Benxi don't have much precipitation, and are surrounded by mountains on three sides. Nantong and Binzhou are on open plains, but Nantong has much more precipitation than Binzhou.

2.2 Data collection Interviews with EPB staff and reviews of archived documents were the primary methods used to collect information. The interviews followed a four-part protocol. Each part focused on a different aspect of the history of air pollution management in the case-study city. The first part of the interview reviewed the decision-making context and the EPB's capacity to diagnose air pollution problems. The second part probed past activities to control air pollution problems. The third part explored the three targeted control policies (retrofit of coal combustion facilities, promoting coal briquettes, and the development of Control Zone) in detail, with a focus on decision-making processes and their implementation. In the last part, engineering and cost data

7

were collected on the technical measures that were used to comply with the targeted control policies. Twenty-two EPB employees were interviewed. Because the target policies go back almost 20 years in some cases, interviewees were selected, in part, for their long tenure with their organization. At least one interviewee in each city had been involved in the earliest efforts of air pollution control. Based on the information collected, the history of air pollution management in each city was reconstructed, with a focus on the period after the promulgation of the 1979 Environmental Protection Law (trial version).

1,400

Taiyuan Benxi

1,200

TSP pollution levels (µg/m^3)

Hohhot Nantong 1,000

Binzhou

800

600

400

200

0 1980

1985

1990

1995

The five case-study cities experienced a deterioration of air quality in the 1980s followed by subsequent stabilization or even improvement in the 1990s. The Chinese ambient standard (annual average) for TSP is 200 µg/m3 in residential areas, and 300 µg/m3 in industrial areas.

3. Dynamics of local air pollution management Since both the content and form of environmental regulations are influenced by the forces underlying the formulation process, it is important to have a better understanding of these forces. A number of analysts have proposed that the major forces promoting air pollution control in China are elite awareness and, surprisingly, public protests (Ross 1992; Wang and Blomquist 1992). The proposition about the role of elite awareness is well studied. But analysis on the role of public involvement has been hampered by the fact that information on these events is usually not released to the public, and may sometimes be classified.

8

This study collected direct evidence regarding this issue from the case-study cities. The analysis shows that air pollution control at local levels has been mainly driven by upper-levels of government, and that local residents played a very limited role in both policy formulation and implementation. 3.1 Public participation at the local level: forms and impacts As representatives to the People's Congress are nominated by local leaders of the Communist Party (CP) and are not freely elected, they rarely raise issues that challenge the local party line. China has a Political Consultation Conference through which representatives can raise environmental concerns of their constituents. Since the beginning of economic reform in the late 1970s, economic development has been the main goal of the CP. In this political setting, representatives in the People’s Congress have had little incentive to be advocates for environmental protection. One of the interviewees in Taiyuan had served on the Environmental Committee of the Municipal People's Congress for four terms. She noted that residents are more likely to direct their complaints to the municipal government than to the People's Congress, and that People's Congress does not have effective control of either the EPB budget or the appointment and promotion of EPB directors. An interviewee in Benxi simply denied the role of People's Congress by saying that "......they (representatives) listen to the mayors." Legal support for public participation turned out to be nominal too. Article 6 of the national Environmental Protection Law (EPL) stipulates that "all organizations and individuals shall have the obligation to protect the environment and shall have the right to report on or file charges against organizations or individuals that cause or damage the environment." The Air Pollution Prevention and Control Law (APPCL) has a similar article. However, the implementation guidelines developed for the APPCL do not include any provision for accepting citizen complaints. Interviewees in case-study cities were asked about the occurrence of protests over air pollution issues, but no examples were offered. No matter what the motivation, protest (kang-yi) in China has been regarded as a serious political event against the government, and could result in severe consequences. However, appeal to the authorities for help (shang-fang) has been acceptable, as long as those appealling for help do not stir a mass movement (ji-ti shang-fang). Interviewees in all the case-study cities said that a number of appeals for help with environmental problems had been filed in their cities. This most often took the form of letters to local environment bureaus. However, the actions taken by environmental authorities in response to those appeals were usually superficial. The interviewees could not recall any substantial actions taken against polluters in response to citizen complaints. Typical responses by local

9

environmental authorities were to ask polluters to improve their operations or to impose a small fine. The polluters’ lack of technical and/or financial capacity was frequently cited by environmental authorities as a reason to justify their approach. However, the real reason seems to be that industrial production has such a high priority that no one, including the affected residents, has the power to take substantial action against the polluters. Especially in the 1970s, neither the environmental authorities (the "Three-waste" Office at that time) nor the residents wanted to be blamed for "destroying or disturbing industrial production," which could bring them serious consequences including being branded as having "political problems." 3.2 Involvement of upper-levels of government Downs (1967) argues that government officials maximize their own utility by pursuing multiple goals including power, monetary income, prestige, convenience, and security. Lieberthal and Oksenberg provided empirical evidence for Chinese officials' pursuit of these bureaucratic goals, especially promotion, security of current position, and winning approval from subordinates (Lieberthal and Oksenberg 1988). Stimulating local environmental authorities to act is an important ingredient of China's topdown environmental management system. A variety of mechanisms have been adopted by upper levels of government to ensure that environmental goals, either set by themselves or passed down from even higher governmental agencies, are transferred to the city level. This study finds that two mechanisms have been widely used: changes in the officials' power (rank, scope of authority, staff size, etc), and financial mechanisms (grants, low interest loans, loan write-offs, tax breaks, favorable price policy, etc). Associating performance with promotion or demotion Within China's multi-level political system, each official is assigned a bureaucratic rank. Rewarding good performance by advancing rank has been effective in stimulating officials to carry out their duties. This mechanism also applies to government officials nominally elected by the People's Congress, for example the mayor, because they are "appointed" in fact by an upper level of government. A vice mayor in Benxi for environmental protection affairs was promoted to NEPA because his good performance was recognized by Qu Geping, Director of NEPA at that time. It was regarded as a big reward among local bureaucrats, not only because his rank rose, but also because officials in the central government usually have more opportunities to get further promotions. A document (Unknown 1995) collected from Hohhot explicitly suggests that the mayor take performance measurement very seriously. Hohhot is one of a group of 37 cities in which the Central government first implemented a performance measurement system for environmental

10

management, called the Comprehensive Ratings for Urban Environmental Quality Improvement and Management (CR). Hohhot was ranked the lowest of the 37 cities in two consecutive years (1993-94). Invigorated by this bad record, the mayor allocated new funds for the development of a Control Zone for Combustion Smoke and launched new initiatives to promote the use of coal briquettes. In 1995, about 36 million Yuan was invested in the development of a CZCS, of which 2 million Yuan was used to induce about eight thousand households to switch from raw coal to coal briquettes for cooking, and 4 million Yuan was invested in particulate control devices (PCDs) for boilers. Several interviewees privately mentioned that the post of EPB director is lucrative and attractive (fei-que, or fei--chai) for bureaucrats. A direct reason is that the EPB director has the authority to judge polluter's degree of compliance with environmental regulations, and decide the negotiable pollution fee. To keep this position, the EPB director certainly has to satisfy the mayor's expectations on the matter of environmental protection. EPB directors in Benxi, Taiyuan and Hohhot tended to have shorter tenures than their counterparts in Binzhou and Nantong. No direct evidence shows that EPB directors in Binzhou and Nantong have done better work. At least, however, there have been fewer air pollution problems in these two cities. An interviewee in Benxi once mentioned that a couple of EPB directors in his city were quickly removed because of “little knowledge of environmental protection and unsatisfactory performance.” Shorter tenure might also be a consequence of higher competition among government officials because of its potential to bring them more power and bribes. It implies that EPB directors are likely to be more willing to follow the mayor’s line in order to keep their positions. Since the late 1970s, economic growth has been the highest priority for the Chinese government as a whole. Maintaining healthy economic growth is one of the most powerful justifications for the promotion of a local mayor. This has stimulated mayors to give much higher priority to economic growth than to environmental protection. However, worrying that bad air quality would scare away foreign investors has driven local leaders to control air pollution, albeit with the ultimate goal of boosting the economy. Interviewees in Benxi and Nantong especially stressed this phenomenon. Financial mechanisms Financial mechanisms have also been widely used as important levers by upper-level EPBs in enforcing their policies. Major forms of financial aid include grants, low interest loans, writeoffs of unpaid loans, tax breaks and favorable pricing policies. Interviewees in all case-study cities except Binzhou mentioned the importance of financial aid, especially free grants (including international aid), from the provincial or even the central government. These financial mechanisms not only offer local governments financial assistance, but also serve to influence agenda setting in localities.

11

A good example is the "Seven-year Campaign to Control Environmental Pollution" in Benxi, which was implemented at the order and with the support of the National Environmental Protection Agency (NEPA). A total of 483 million Yuan was invested from 1989 to 1995 in environmental protection projects. About 40% of the total investment was in the form of grants from the Central and provincial governments. Another significant part of the investment2 was raised from a favorable pricing policy granted from the Central government to Benxi Steel & Iron Company (BSIC). This price policy allowed BSIC to sell a specified amount of its product at market price, which was much higher than the regulated price, and to use the revenue from the price difference to pay for environmental control programs. Less than 20% of the total investment was from the accumulated pollution fee, which had been collected regularly from polluters that violate emission standards. Air pollution control projects funded by this program included eight electrostatic precipitators in BSIC. (Environmental Protection Bureau of Benxi 1996) Hohhot also received significant financial support from the Central government. An example is the first central heating project in this city, which was constructed in the late 1980s. This project received an 11.5 million Yuan grant and a 7.5 million Yuan low-interest loan from the State Planning Commission (SPC). SPC eventually wrote off this loan because Hohhot did not pay it back. (Hohhot Heat Supply Company 1992)

4. Policy implementation: variations in approaches and progress 4.1 Policy implementation Retrofitting coal boilers with particle control devices (PCDs) Retrofit of coal boilers and the use of particle control devices was the first policy to control particulate matter pollution. Black smoke from boilers without PCDs, which is sometimes called "smoke dragons (yan-long)" in policy campaigns, is a very apparent and visible source of particle pollution. When this policy was formulated in 1973 and at the early stage of implementation, China's administrative and legal settings for environmental protection were still in their fledgling stages, with very limited monitoring and diagnosing capacities. Boilers were first targeted because black smoke from stacks was so obvious. This policy was implemented before emissions or design standards for boilers, which weren't issued until 1983 and 1980, respectively. In these early days, progress could only be evaluated based on two criteria: a) whether or not a boiler had been retrofitted with a new combustion chamber; and b) whether or not a particle control device had been installed.

2

The specific portion is not clear, but it is less than 40%.

12

The five case-study cities adopted quite different strategies in enforcing this policy. Financial shortages were the major problems facing the Hohhot EPB. A large portion of boilers in this city are heating boilers. Most of them were operated and maintained at government agency (rather than state-owned industry) expense. Because Hohhot's level of economic development was low, the Hohhot government was unable to allocate funds to implement this policy. In its early stage, the implementation was supported by an energy conservation fund. But the "leaking basket" nature of this free grant without a powerful audit mechanism seriously weakened its effectiveness. Many grantees invested this money in projects that had nothing to do with air pollution control. Although this flaw was corrected by changing grants to loans in the early 1980s, the situation was not significantly improved because the total budget was still very small. Most boilers were not retrofitted or equipped with PCDs until the late 1980s. Benxi and Taiyuan had better situations than Hohhot. This policy was not only well funded in these two cities but also had relatively good technical support. Technology experts in the Taiyuan EPB had the capability to design retrofits for boilers with capacities of less than 20 t/h. Their major problems were the retrofit of large industrial boilers, which were more challenging technologically and required much larger investments. Binzhou's administrative status was promoted from township level to municipal level in 1982. Its EPB was formed in 1983. A survey in 1989 showed that all industrial and heating boilers had been equipped with PCDs, which were predominantly cyclone collectors. From the middle 1980s to the early 1990s, central heating boilers replaced most household heating stoves. These boilers were manufactured according to new standards for air pollutant emission, and already had PCDs installed. Although the Nantong municipal government offered full financial support to boiler owners that had financial difficulties, Nantong's progress in 1970s still fell behind that of Taiyuan and Benxi. Nantong made faster progress in the following years as a result of increasing pressure from upper levels of government. Substituting raw coal with coal briquettes for cooking Boiler retrofit and the use of ash collectors are effective in controlling PM emissions from large point sources. However, even after boilers were modified, TSP concentration levels were still very high and even kept rising in some cities. One of the major reasons is that small coalburning household stoves had not been targeted by any environmental policy. Partially for this reason and partially for energy conservation, coal briquettes are recommended in "Technology Policy for Prevention and Control of Air Pollution from Coal Combustion" issued by Environmental Protection Committee (EPC) of the State Council in 1984. At the time, coal briquettes as household fuel were widely regarded as a temporary measure

13

pending the availability of gaseous fuels, which was not available to every resident because of a lack of supply. Hohhot, Benxi and Taiyuan are located in China’s three major coal producing regions, and have an abundant supply of quality coal. The coal briquette programs in these three cities encountered several difficulties. The first problem arose from the potential users. Residents in these three cities had easy access to high quality coal at a very low price. Some could even get free coal from their work units. In the early 1980s, most employers in Hohhot delivered free coal from mines to employees’ homes and charged a nominal transportation fee of 7 Yuan ($1) per ton (Yue 1999). A similar policy was also found in Taiyuan (Ge 1999). As a result, coal briquettes could not compete with raw coal on price. Another problem related to user preferences was a difference in convenience between briquette-fired and coal-fired stoves. A test of five stoves designed for coal briquettes in Hohhot in 1994 showed that they needed 90-150 minutes for the fire to reach a state appropriate for cooking. Stoves burning raw coal required only 30 minutes. Another problem with the briquette program was on the supply side. Producing coal briquettes was not profitable because of price controls. Institutional factors also caused many problems. The production and distribution of coal briquettes were usually controlled by an energy supplier that was owned by the municipal government, which was the sole recipient of governmental subsidies for coal briquettes. But local environmental authorities had no control of the subsidies. This situation was worsened by a policy that forbid private companies to enter this market. Different strategies were adopted in Hohhot, Taiyuan and Benxi to deal with these difficulties. The Hohhot EPB succeeded in lobbying the municipal government in 1994 to change the subsidy policy from subsidizing the monopolist producer to subsidizing the households that switched to coal briquettes. More importantly, the EPB gained the authority to distribute the subsidies. From 1997, the Hohhot municipal government increased the subsidy to a level that could roughly offset residents' increase in expenses. Sales and delivery of raw coal were banned in Hohhot's urban area. Taiyuan focused attention not only on the supply side but also on the users. A regulation from the municipal government banned the use of raw coal in households that are located in a CZCS. A joint task force consisting of police and officials from the EPB was sent out to confiscate raw coal in residents’ yards. By the end of 1995, over 80% of households who had no access to gaseous fuel had switched to coal briquettes.

14

Nantong and Binzhou had different stories. Coal briquettes had been the traditional cooking fuel in Nantong for many years. Although households in Binzhou had no tradition of using coal briquettes, it replaced raw coal very quickly after the government started to encourage their use for purpose of energy saving in the middle of 1980s. High coal prices due to lack of supply was one of the major factors that stimulated the rapid penetration of coal briquettes in these two cities. Control Zone for Combustion Smoke The Control Zone for Combustion Smoke (CZCS) was first introduced into practice in Shanghai in 1982, and was recommended nationwide by NEPA in 1985. This policy required municipal governments to define regions with their cities to be CZCS that would receive special attention for particulate emissions control, and finally expand to the whole city. EPBs were to define the zones, but implementation and enforcement was to be performed at the district level (i.e., one level below municipal). "Guidelines for Development and Management of Control Zone for Combustion Smoke," issued by the Environmental Protection Committee in 1987, clarifies the definition of a CZCS, recommends technical measures for developing a CZCS, and specifies the responsibilities of the governmental agencies involved. According to this document, a CZCS must meet the following requirements:3 a) at least 80% of the boilers, kilns, and stoves in cafeterias and restaurants within the zone must meet the emission standards in Ringelman Number; and b) at least 70% of the industrial boilers and facilities must meet the emission standards which regulate the concentration of particulate matter in flue gas. The national guidelines for establishing CZCS did not specify a minimum size for the zone, but progress toward zone expansion was made an explicit part of the Comprehensive Rating system, under which the job performance of local EPB directors and even mayors were evaluated. A major motivation underlying CZCS is to take advantage of China's administrative systems, which are organized according to geographic region, and to motivate lower governmental institutions to participate in air pollution control. Another significant feature of the CZCS program is that it provides incentives to control PM pollution from sources that are small in terms of individual emissions but are large in number. That is because, according to the development guidelines, a large hot-water boiler and a small restaurant stove are each counted as “one source". A study by Sun shows that the CZCS policy is implicitly an exposure-based environmental regulation, which would allocate more regulatory efforts to those sources that are in closest proximity to people (i.e, to household and neighborhood sources rather than to industrial sources) (Sun 2001). The development guidelines 3

Stricter requirements applied to cities in the non-heating zone or specially identified cities.

15

for CZCS offer the municipal government full authority in choosing a candidate region within each city. Naturally, municipal governments would first choose the “easy” regions, which are regions having fewer large industrial sources. Retrofitting large industrial sources has much higher demand for both investment and technical capacity. A residence region with mainly heating boilers and restaurant stoves can more easily meet the CZCS requirements than an industrial region with mainly industrial boilers and kilns. The first control zones in all the five case-study cities were very small and located in downtown areas, which usually have a much smaller number of pollution sources, and no large sources. In 1988, the central government instituted a comprehensive rating (CR) system for urban environmental performance (Environmental Protection Committee of State Council 1988). This system included as one attribute the ratio of the area of the CZCS to the total urban area. The government currently uses this system to evaluate the environmental performance of 37 cities annually, including Taiyuan and Hohhot. These ratings feed into the overall performance evaluations of both the mayor and EPB's of these cities, which therefore affect promotion opportunities. In spite of these favorable aspects of the CZCS policy, its establishment still met a variety of problems in the case-study cities. CZCS was supposed to be an effective measure in Hohhot and Binzhou as they have a relatively small portion of large sources like industrial boilers. But their progress was the slowest among the five cities. After the first CZCS was established in Hohhot in 1986, further development was suspended due to financial difficulties. No more CZCS was developed in the following eight years until the effort was resumed in 1994. By 1998, CZCS covered only 45% of the urban area. Binzhou started to establish a CZCS in 1992 in response to mounting pressure from the provincial government. Binzhou experienced significant changes in energy mix from the middle 1980s to the early 1990s: cooking fuel switched from raw coal to coal briquettes and then to gaseous fuel, and household heating stoves were replaced by neighborhood heating boilers. Although these changes provided Binzhou a better basis to develop CZCS, the municipal EPB still had trouble making the boilers comply with emission standards due to lack of financial resources and administrative capacity. The first and second CZCS totaling 7.8 sq. km was finished in 1992, and a third one of 4.5 sq. km in 1995. The total area of these three CZCS zones is only about 60% of the urban area. For two reasons, the Benxi EPB had no enthusiasm for developing CZCS. First, they did not believe that CZCS would work in their city. Second, Benxi is not among the 37 cities being subject to CR evaluation. The major sources of particulate matter in this city are a small number of large sources that belong to Benxi Steel & Iron Corporation (BSIC) and several other super

16

large enterprises. 4 CZCS policy requires only 70% of industrial boilers and kilns meet emission standards. An interviewee thought that BSIC's political power was so strong that their stacks would be exempted from this policy, and that CZCS policy would not solve Benxi's air pollution problems. Development of CZCS in Nantong was a consequence of a hard push from the provincial government. The first CZCS of 1.3 sq. km in Nantong, located in the downtown area, was finished in 1988. After that, no further progress was made in Nantong in the following couple of years. The Nantong EPB was criticized by the provincial government for this reason and other unsatisfactory performance. This pushed the then EPB director to personally lead the development of a CZCS, and prompted the municipal government to list CZCS in the government's campaign "to do a few real things for residents." CZCSs were expanded to the entire urban area except two remote townships by 1998. Nantong’s high level of economic development compared to other case-study cities is reflected in the range of measures used to create a CZCS. Nantong employed several technological measures having high implementation costs. These included relocating polluters from the urban area to the suburbs, and replacing small boilers with a CHP station. Such measures were not prevalent in Hohhot and Benxi due to high implementation cost. 4.2 Institutional factors influencing policy enforcement The institutional context for urban air pollution management has been changing in the past 20 years as China's economic and political settings change. The municipal EPBs of these five case-study cities were first placed under the Economic Commission and then the Urban Construction and Development Bureau. Because their first priority was industrial production, environmental protection was given short shrift. In 1984, the National Environmental Protection Agency (NEPA) was established as an independent government organ. The local environment authorities were moved under NEPA command, which strengthened environmental policy implementation. Still, however, there were a variety of institutional problems that limited EPB effectiveness. Problems with policy coordination The dispersal of responsibilities, authorities and benefits among different agencies has been one of the primary causes of deficiencies. For example, boiler operation is managed by at least three governmental agencies. The Labor Bureau issues an operation license regulating the safety aspects of operation, the Material Bureau supplies the coal, and the EPB is responsible for 4

In 1997, 28 out of Benxi’s 31 major industrial pollution polluters were located in the urban area. They included the 5 largest plants of BSIC. These 31 largest industrial polluters contributed 75% of the total flue gas emission in Benxi, paid over 85% of the pollution fee, and employed about 89,000 people (13% of Benxi's population). The 28 major pollution sources that are located in the urban area had 215 boilers and 109 stacks. (Benxi Environmental Monitoring Central Station 1998).

17

compliance with emission standards. The EPB has the least authority among the three in influencing the operation of boilers, and requires cooperation – coordinated by the Environmental Protection Committee – from the other two bureaus. It took several years in Binzhou to get the Labor Bureau's cooperation in adding instruction on environmental matters to their training programs for boiler operators and in issuing a joint license on safety and emissions. The EPBs have the responsibility to promote the use of coal briquettes, but didn't have the authority to control their production and distribution, as well as the allocation of the government subsidy. This was cited by interviewees in Hohhot as the major obstacle to implementing the coal briquette program in their city. The change in subsidy policy for coal briquettes in Hohhot occurred in 1995 after several years of hard struggle between the EPB and the Material Bureau to obtain the authority to allocate the governmental subsidy. Under the new subsidy policy, EPB gained the allocation authority, and redirected subsidies to coal briquettes users rather than producers. A direct effect of this policy change was the clearance of the entry barrier for private coal briquettes producers. Competition among producers has resulted in improvement in quality and lower prices. Although the involvement of multiple agencies makes policy implementation more difficult, there are examples in which interagency coordination has lead to successful PM control. One example is Nantong's policy of suspending coal supply to boilers that violate emission standards. For this policy, the EPB obtained complete cooperation from the Material Bureau. In a city that lacks access to local coal supply, this policy was very successful in pushing boiler owners to comply with emission standards. A successful interagency coordination usually requires significant involvement of the municipal government. Conflicting policies are another impediment to successful PM reduction. These have been very common in the case-study cities. Taiyuan's EPB once found that it was very cost-effective in mitigating PM emission to substitute ordinary coal with coal of low ash content. Low-ash coal is produced from mines in an adjacent region, which is not in Taiyuan’s jurisdiction. However, the substitution could not be implemented because of a policy in Taiyuan forbidding the import of coal into Taiyuan in order to protect the interests of local coal mines. Conflicting interests Environmental protection authorities have acquired many functions, including environmental planning, regulation, and more recently, for-profit environmental consulting. Some of these functions are incompatible. For example, part of the pollution levies and fines that EPBs collect had been retained to fund EPB operations. This encouraged EPBs to keep polluters polluting to maintain the revenue stream from levies and fines. Thus, when an EPB established a CZCS with a mandate from NEPA of making 70% of industrial polluters comply with regulations, a common

18

practice was to put less pressure on those polluters in the zone that pay the most in levies and fines, or even simply put them in the 30% category. 4.3 Significant variation under a top-down administrative system Notwithstanding the top-down administrative and legal systems for environmental protection in China, significant variations from city to city in approaches and progress have been very common. Analysis of the survey results indicates that the following factors have contributed to these variations: economic development level, air pollution situation, energy availability, and industrial structures. Economic development level Economic development level is one of the most important factors. Both longitudinal comparison in a specific city and cross-city comparison reveal its impacts. The pollution fee has been the primary source of financing for EPBs. According to the law, no more than 20% of the pollution fees can be allocated to EPB operations, with the remainder going back to polluters in the form of loans for pollution control. However, it has been a common practice in case-study cities for the EPB to keep more than 20% of the fees it collects. A document from Benxi shows that over 40% of pollution fees were used for EPB operation in the 1980s. Thus, an EPB's ability to enforce environmental policies has been constrained by the amount of pollution fee that is collected from polluters. This, in turn, is determined by both the number of pollution sources out of compliance with regulations and the financial situation of the polluters. EPBs in poorer cities such as Binzhou and Hohhot have a more difficult time extracting pollution fees from polluters. Such cities have fewer profitable industrial enterprises from which to collect pollution fees. In a society in which most workers still depend on their work unit to provide a host of social services, closing plants that cannot afford pollution fees is hardly an option. That is evidenced in the case-study cities. This survey shows that a larger portion of pollution fee was allocated to support EPB's operation and capacity building in poorer cities than in wealthier cities. It is difficult to compare the cost-effectiveness of this investment against investing the pollution fee in pollution control projects. However, among the five case-study cities, those that have used more pollution fee for EPB operation usually made slower progress. Better financial situations have given wealthier cities greater capacity to implement environmental policies. However, there are still gaps between what they pursue and what they could afford. Financial difficulties exist in different forms. For example, Nantong is still struggling for more funds to move more industrial polluters out of the urban area. Binzhou is waiting for additional financial support to make the entire urban area meet CZCS requirements. Hohhot is struggling for money to subsidize the use of coal briquettes in the remaining 40-50 thousand households that need them. Although both Taiyuan and Hohhot are subject to the

19

Comprehensive Rating evaluation, over 95% of Taiyuan's urban area met CZCS standards in 1995, whereas only 45% of Hohhot met CZCS standards in 1998. The most plausible explanation for this difference seems to be the difference in wealth of these two cities. Air pollution situation Taiyuan, Hohhot and Benxi have among the highest ambient TSP levels in China. Interviews with officials in these cities reveal that the Central government has cited these high TSP levels to exert pressure on local government to improve air quality. For those cities with the worst particulate problems, the central and provincial governments have responded with grants to be used for air pollution control. The central government contributed over RMB 70 million to Benxi's "Seven-year Campaign to Control Environmental Pollution." Both Taiyuan and Hohhot have also received special grants from their provincial governments for air pollution control. Energy availability Extensive coal use is the most direct cause of the air pollution problems in all five case-study cities. Therefore, many strategies to reduce air pollution involve changes in energy supply and energy efficiency. The easy availability of raw coal at low prices has seriously hindered the application of clean energy technology, especially in Taiyuan, Hohhot and Benxi. Compared to Nantong and Binzhou, people in the other three cities have developed a very energy intensive life style, and have less incentive to save energy. For example, promoting the use of coal briquettes met problems in all the three cities, and forced the local governments to take stronger efforts to enforce the replacement of raw coal. In Binzhou, by contrast, coal briquettes replaced raw coal in a short period with very little governmental support. Residents in Nantong had a tradition of using coal briquettes for cooking because of the high price of raw coal. Furthermore, nearly all residents of Binzhou and Nantong now use gaseous fuel for cooking. But the fraction of gas users is much less in Hohhot and Benxi. Hohhot is a very poor city, and residents cannot afford to buy gaseous fuel. Residents of Taiyuan and Benxi, however, are as wealthy as residents of Binzhou, where gas use is common. The major reason that gas has limited penetration of the Taiyuan and Benxi markets seems to be the continuing availability of raw coal at low prices. Industrial structure The easy availability of cheap coal is one of the major reasons that heavy industries dominate the economies of Benxi and Taiyuan. This predominance of heavy industry poses several problems for air pollution management. Firstly, the pollution intensity of heavy industry is greater than that of other economic activities, so the investment requirements for pollution control per unit economic output are highest in cities with lots of heavy industry. Secondly, because the local economy is dominated by a few giant enterprises, those polluters wield great

20

political power, and so can avoid compliance with environmental regulation. Finally, heavy industries usually have large boilers, which pose additional financial and technical challenges. In cities dominated by heavy industry, the EPB's progress depends greatly on the amount of cooperation the EPB can obtain from the largest enterprises.

5. Summary Comparing China's history of PM control with that in industrialized countries, this research finds that both the factors that had caused air pollution problems in the first place and the strategies adopted in response (technical improvement and energy substitution) are similar. However, the dynamics of air pollution control and the factors influencing policy implementation are very different. This difference is explained by several factors. China has a unique environmental management system, which is characterized as a top-down administrative system lacking organized and meaningful public involvement. Some agents and forces that have played essential roles in industrialized countries do not exist in China. China is also different from industrialized countries in its legislative structure and in the relationship between central and local governments. Elite awareness along with the pursuit of political and social stability has motivated the central government to take action on particulate matter pollution. The central government pursued its goals at the local level by designing a system of positive and negative incentives that induced local governments to pursue bureaucratic goals. However, the effectiveness of this system has fallen short of its potential, largely because of institutional factors such as poorly designed incentive structures. The implementation of control strategies for particulate matter pollution has been influenced by many additional factors including economic development levels, energy availability, air pollution levels, and industrial structure. These factors have caused significant variations in practice from city to city. A surprising finding of this five-city study is that China's central authorities took action to control urban air quality at an earlier stage of economic development than has been observed elsewhere. This probably reflects the relatively strong control that the Chinese central government had over local conditions during the period covered by this study, Caution should be exercised in extending the findings of this study to other domains of environmental policy. Control of particulate matter, especially the three policies studied, poses less threat to economic growth than other more costly environmental controls such as reduction of sulfur air pollution. The benefits of particulate controls are local, quick to be realized, and easy to monitor by both residents and upper levels of government.

21

Acknowledgments This research was made possible through support from the Center for Integrated Study of the Human Dimensions of Global Change. This Center has been created through a cooperative agreement between the National Science Foundation (SBR-9521914) and Carnegie Mellon University, and has been generously supported by additional grants from the Electric Power Research Institute, the ExxonMobil Foundation, and the American Petroleum Institute.

References Agras, J. and D. Chapman (1999). “A Dynamic Approach to the Environmental Kuznets Curve Hypothesis.” Ecological Economics 28: 267-277. Andrews, R. N. L. (1999). Nationalizing Pollution Control. Managing the Environment, Managing Ourselves: A History of American Environmental Policy. New Haven, Yale University Press: 227-254. Bailey, C. J. (1998). Congress and Air Pollution: Environmental Policies in the USA. Manchester, Manchester University Press. Benxi Environmental Monitoring Central Station (1998). The Major Industrial Pollution Sources in Benxi in 1997. Benxi. Chan, H. S., et al. (1995). “The Implementation Gap in Environmental Management in China: The Case of Guangzhou, Zhengzhou, and Nanjing.” Public Administration Review 55(4): 333-340. China Environment Yearbook Committee, Ed. (1998). China Environment Yearbook: 1997. China Environment Yearbook. Beijing, China Environment Yearbook Publishing House. Downs, A. (1967). Inside Bureaucracy. Boston, Little Brown. Edmonds, R. L. (1998). “Studies on China's Environment.” The China Quarterly(156): 725-732. Environmental Protection Bureau of Benxi (1996). A Review of the Seven Year Campain to Control the Environmental Pollution in Benxi. Benxi: 80. Environmental Protection Committee of State Council (1988). Implementation Methods for Comprehensive Ratings for Urban Environmental Quality Improvement and Management. Ge, X. (1999). Air Pollution Control History in Taiyuan. Taiyuan. Grinder, R. D. (1980). The Battle for Clean Air: The Smoke Problem in Post-Civil War America. Pollution and Reform in American Cities, 1870-1930. M. V. Melosi, University of Texas Press: 83-103. Hohhot Heat Supply Company (1992). Central Heating in Hohhot. Hohhot.

22

Lieberthal, K. and M. Oksenberg (1988). Salient Characteristics of The Structure of Power. Policy Making in China. Princeton, Princeton University Press: 135-168. Mayhew, D. R. (1974). Congress: The Electoral Connection. New Haven, Yale University Press. Ringquist, E. J. (1993). Environmental Protection at the State Level: Politics and Progress in Controlling Pollution. Armonk, Sharpe. Ross, L. (1992). “The Politics of Environmental Policy in the People's Republic of China.” Policy Studies Journal 20(4): 628-642. Sabatier, P. A. and H. C. Jenkins-Smith, Eds. (1993). Policy Change and Learning: An Advocacy Coalition Approach. Boulder, CO, Westview Press. Selden, T. M. and D. Song (1994). “Environmental Quality and Development: Is There a Kuznets Curve for Air Pollution Emissions?” Journal of Environmental Economics and Management 27: 147-162. Stern, D. I., M. S. Common, et al. (1996). “Economic Growth and Economic Degradation: The Environmental Kuznets Curve and Sustainable Development.” World Development 24(7): 1151-1160. Sun, G. (2001). An integrated study of China's air pollution management: effectiveness, efficiency, and governance. Department of Engineering and Public Policy. Pittsburgh, Carnegie Mellon University: 173. Tang, S.-Y., C. W.-h. Lo, et al. (1997). “Institutional Constraints on Environmental Management in Urban China: Environmental Impact Assessment in Guangzhou and Shanghai.” China Quarterly(152): 863-874. Tarr, J. A. and C. Zimring (1997). The Struggle For Smoke Control in St. Louis: Achievement and Emulation. Common Fields: An environmental history of St. Louis. A. Hurley. Saint Louis, Missouri Historical Society Press: 199-220. Unknown (1995). Summary Report on Comprehensive Ratings for Urban Environmental Quality Improvement and Management in Hohhot. Hohhot. Wang, X. and R. F. Blomquist (1992). “The Developing Environmental Law and Policy of the People's Republic of China: An Introduction and Appraisal.” The Georgetown International Environmental Law Review 5: 25-75. Yue, X. (1999). History of Air Pollution Control in Hohhot. Hohhot.

23