The impact of environmental regulations on employment: Quasi-experimental evidence from regulated and unregulated industries in China

ABSTRACT

Existing studies on the employment effect of environmental regulations focus exclusively on regulated sectors, but labour demand in unregulated sectors can also be affected in a general equilibrium framework. Using the designation of Key Cities for Air Pollution Control in 2001 in China as a quasi-experiment, we employ PSM and DDD methods and find that the policy lowers employment in emission-intensive polluting firms, but it also increases employment in labour-intensive non-polluting firms, providing evidence of relocation of workers from regulated to unregulated industries. Evidence from city-level aggregate data suggests that the overall employment effect is minimal, and no evidence is found to suggest that workers move from the manufacturing sector to the services sector. The results are robust to a host of robustness checks. The finding implies that the job-killing effect of environmental regulations is over-stated, and non-polluting industries are not suitable comparison units in previous DID literature.

KEYWORDS:

• Environmental regulation
• labor market
• China
• difference-in-difference-in-differences

JEL CLASSIFICATION:

• Q52
• Q58
• J23

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

¹ For the official document (in Chinese), please see http://www.110.com/fagui/law_94153.html (Accessed on May 5, 2023).

² For an example of the end-of-year evaluation of the environmental performance, please see https://www.mee.gov.cn/gkml/zj/bgt/200910/t20091022_173769.htm (Accessed on May 5, 2023).

³ For the official document (in Chinese), please see http://cn.chinagate.cn/economics/2007–11/26/content_9294315_3.htm (Accessed on May 5, 2023).

⁴ Another concern is that the data used in this study dates as far back as 1998, which means that we will not have data for the pre-policy period if we were to include the first 47 cities in the sample.

⁵ The TCZ policy was an environmental policy initiated in 1998 by the central government. The Two Control Zones comprise SO₂ pollution control zones and acid rain control zones. Once a city was labelled as a TCZ city, stricter regulatory measures were put in place. Several papers have taken advantage of this policy as a quasi-experiment to examine the effect of environmental regulations (e.g. Cai et al. 2016; Tanaka 2015).

⁶ For the official document (in Chinese), please see https://www.mee.gov.cn/gkml/zj/bgth/200910/t20091022_174132.htm (Accessed on May 5, 2023).

⁷ For the official document (in Chinese), please see https://www.mee.gov.cn/gkml/zj/bgt/200910/t20091022_173898.htm (Accessed on May 5, 2023).

⁸ Although we do not have a direct measurement of regulation at the firm level, this policy is still unlikely to have a direct impact on enterprises in non-polluting industries for several reasons. First, the central government encourages local governments to focus on high-polluting industries and regularly monitors the emissions from local high-polluting firms. Second, it is not cost-effective to abate emissions from non-polluting industries since marginal abatement cost can be very high given their low emission levels. Third, the policy does not involve a component such as an emission tax in Yamazaki (2017) and Yip (2018) that would affect all industries in the economy. One additional concern is that some firms, despite in non-polluting industries, can be large polluters as well. To address this concern, as discussed later, we exclude firms in non-polluting industries that are also in the Chinese Environmental Statistics Database (CESD), which contains all the large polluters that are responsible for more than 85% of local total emissions. Therefore, the remaining firms in our sample of non-polluting industries should be free of environmental regulations.

⁹ For the official document (in Chinese), please see http://www.gov.cn/zhengce/content/2010–11/22/content_5181.htm (Accessed on May 5, 2023).

¹⁰ The dataset can be accessed at https://disc.gsfc.nasa.gov/datasets/M2TMNXAER_5.12.4/summary (Accessed on May 5, 2023).

¹¹ This is calculated as 95.775 × 0.00034 = 3.26%.

¹² According to the classification by China’s National Bureau of Statistics, in our sample, eastern China includes Hebei, Jiangsu, Zhejiang, Fujian, Shandong, and Guangdong; middle China includes Shanxi, Anhui, Jiangxi, Henan, Hubei, and Hunan; western China comprises Inner Mongolia, Guangxi, Sichuan, Guizhou, Yunnan, Shaanxi, Gansu, and Ningxia; and northeastern China consists of Liaoning, Jilin, and Heilongjiang.

¹³ We thank the reviewer for the advice.

¹⁴ We thank the reviewer for bringing this up.

¹⁵ We thank the reviewer for the comment.

Additional information

Funding

The work was supported by the Northwest University of Political Science and Law Research start-up fund .