Leveraging political incentives for environmental regulation: Evidence from China


Published 07.03.22
Photo credit:
Asian Development Bank/flickr

Leveraging high-powered political incentives for pollution control could distort the decentralised enforcement of environmental regulation

In developing countries such as China and India, billions of people live under extreme pollution every day, while still being economically dependent on pollution-heavy manufacturing industries (Greenstone and Hanna 2014, Ebenstein et al. 2017). As these countries balance economic growth with environmental quality, it is important to understand the economic costs of alleviating pollution. Most research studying this challenge has taken place in developed countries like the United States, thereby shedding limited light on the developing world where the cost of environmental regulation might vary substantially due to differences in industrial structures and factor endowments (Basu and Weil 1998), as well as political institutions and bureaucratic incentives (Acemoglu and Robinson 2013, Greenstone and Jack 2015).

River water quality regulations in China

Our study (He, Wang, and Zhang 2020) fills an important knowledge gap by studying environmental regulations in China, currently the world’s largest emitter and manufacturer. In the late 1990s, after nearly two decades of unprecedented growth in industrial manufacturing, China started to face a variety of pressing environmental challenges, including deteriorating surface water quality (Ebenstein, 2012). According to the World Bank (2007), in 2000 roughly 70% of China’s rivers contained water deemed unsafe for human consumption. 

Seeing the growing social unrest associated with surface water pollution, the Chinese central government began attempts to protect water bodies. They installed several hundred state-controlled water monitoring stations along the major national river trunks. Used mainly for scientific study at first, the efforts hit full steam when in 2003, the central government imposed explicit water quality targets for all the state-controlled monitoring stations, and also adopted a target-based abatement system in which local politicians were promised a chance at promotion only if their targets were met. Through these local officials’ efforts to regulate polluting firms and abate water pollution, China’s surface water quality improved dramatically after 2003. 

Figure 1 Water quality reading

Notes: This figure shows the trend of average water quality readings of national monitoring stations, where 1 represents the highest water quality and 6 represents the lowest water quality. 

However, this political contract between central and local governments was easy to undermine because of imperfect monitoring. Water monitoring stations can only capture upstream emissions which gives local officials the incentive to enforce tighter regulations on polluters immediately upstream of monitoring stations, as compared with their immediately downstream counterparts. 

Research design 

Exploiting this spatial discontinuity in environmental regulation stringency, we compare the productivity and pollution reductions between polluting plants located immediately upstream and downstream from the monitors. The key assumption of the research design is that firms immediately upstream and downstream relative to the monitoring station should be identical before the regulations are implemented (i.e. before 2003), but should differ from each other later as upstream firms face tighter regulations. Since, the water monitoring stations were located based on hydrological factors before water quality readings became a political priority, this assumption seems likely. 

To gather the data for analysis, we use the China Environmental Quality Statistical Yearbooks for information on water quality monitoring stations. We then collect the firm-level production information from the Annual Survey of Industrial Firms (ASIF) and firm-level emission data from the Environmental Survey and Reporting (ESR) database throughout the period 2000–2007. The studied sample comprises 17,726 unique ASIF firms and 9,797 ESR firms located alongside 159 monitoring stations. After collecting the data, we calculate the distance to the nearest station and project the firms to the nearest river basin to estimate their elevation. This is used to determine whether the firms are ‘upstream’ or ‘downstream’ relative to the nearest monitoring station. 

Uneven regulation and productivity losses among firms upstream

We find that polluting firms located immediately upstream are 24% less productive. The productivity loss is primarily driven by upstream polluters investing more in (non-productive) abatement equipment and making costly adjustments to clean up production processes to cope with tighter regulations. The gap observed between upstream and downstream firms cannot be explained by the endogenous locations of monitoring stations or polluting firms alone. Instead, the upstream-downstream gap in productivity existed only in polluting industries. Further, only polluters within a few kilometres upstream are regulated as emissions from farther upstream would dissipate quickly over space and have negligible effect on water quality readings. 

Figure 2 Effects of water quality monitoring on TFP

Notes: Industry and monitoring station fixed effects are absorbed before plotting the regression discontinuities. 

The monitoring data show that polluting upstream industries generate significantly lower chemical oxygen demand (COD) emissions than their downstream counterparts. In addition, even though upstream firms emit less, they pay higher amounts of emission fees than downstream firms, implying that local officials hold double standards in regulation enforcement.

Estimating the productivity cost of reducing water pollution

When the central government started to link water quality readings to political promotions in 2003, the productivity of monitored firms dropped significantly compared to years immediately prior. We find that upstream firms were just as productive as downstream firms from 2000 to 2002. However, when the government began to link water quality readings to political promotions since 2003, the productivity of upstream firms dropped significantly.

The environmental regulations led to significant economic losses to China. As the more heavily regulated firms injected more money into cleaner equipment and other abatement actions, these capital inputs did not increase their output. As a result, a 10% reduction in pollution led to a 3% drop in productivity for China’s polluting industries. Taken together, this means that China’s efforts towards reducing water pollution led to a total loss in industrial output of more than 800 billion Chinese yuan between 2000 and 2007, or more than 110 billion Chinese yuan per year. 

Political incentives distorting enforcement 

The higher the political incentive to local officials, the more significant the gap in productivity between upstream and downstream firms. In China, prefecture-level leaders cannot get promoted to the provincial level once they exceed 57 years of age (Wang 2016). Using this policy to separate leaders who have a strong political incentive to meet the targets as compared to those who are less incentivised, we discovered that there was an even larger gap in the productivity of upstream and downstream firms in the prefectures where the local leaders were 56 years old or younger.

The gap in productivity between upstream and downstream firms is largest when it is more difficult for local officials to manipulate the data. The gap in productivity between upstream and downstream firms gets particularly large when monitoring stations are ‘automated’ – the data are fed directly to the central government, and therefore less susceptible to direct data manipulation. This suggests that local officials used to manipulate water quality readings for those traditional ‘manual’ stations, instead of actually regulating upstream polluters.

Policy implications 

The study demonstrates a flaw in the political centralisation of decentralised policies. Under political centralisation, when the central government wants to mobilise local governments to implement decentralised policies, it often adopts a target-based incentive scheme where political rewards are promised contingent on meeting certain performance criteria. However, if the central government is unable to perfectly monitor all aspects of decentralised programme enforcement, local government officials will focus on only the aspects the central government can monitor while shirking on the rest. 

In the context of improving water quality, the central government leveraged high-powered political incentives to improve surface water quality but could only observe water quality readings of the state-controlled monitoring stations, which reflect emissions upstream but not downstream from them. Local government officials in turn, ensure the information the central government receives meets the targets by imposing significantly tighter regulation on upstream firms – prioritising ‘water quality readings’ over ‘actual water quality’. As a result, the water being monitored became cleaner, while a whole host of firms escaped regulation altogether. A well-intended central programme deviated from its original intent under decentralised enforcement. This example offers lessons on how the centralised government can rethink its incentive and target-based enforcement structure for decentralised policies. 

Editors' note: This column is based on a PEDL project.


Acemoglu, D and J A Robinson (2013), "Economics versus politics: Pitfalls of policy advice", Journal of Economic Perspectives 27(2): 173-92.

Basu, S and D N Weil (1998), "Appropriate technology and growth", The Quarterly Journal of Economics 113(4): 1025-1054.

Ebenstein, A (2012), "The consequences of industrialization: Evidence from water pollution and digestive cancers in China", Review of Economics and Statistics 94(1): 186-201.

Ebenstein, A , M Fan, M Greenstone, G He, and M Zhou (2017), "New evidence on the impact of sustained exposure to air pollution on life expectancy from China’s Huai River Policy", Proceedings of the National Academy of Sciences 114(39): 10384-10389.

Greenstone, M and R Hanna (2014), "Environmental regulations, air and water pollution, and infant mortality in India", American Economic Review 104(10): 3038-72.

Greenstone, M, and B K Jack (2015), "Envirodevonomics: A research agenda for an emerging field", Journal of Economic Literature 53(1): 5-42.

He, G, S Wang, and B Zhang (2020), "Watering down environmental regulation in China", The Quarterly Journal of Economics 135(4): 2135-2185.

Wang, S (2016), “Fiscal competition and coordination: Evidence from China”, UC Berkeley Working Paper.

World Bank (2007), "Cost of pollution in China: Economic estimates of physical damages".