From lenient to stringent: Environmental policy target adjustment and the decline in mortality

Leveraging China’s most stringent air pollution control legislation implemented between 2013 and 2017-the Air Pollution Prevention and Control Action Plan (APPCAP)-we provide direct causal evidence that variation in emission reduction targets led to reductions in mortality. Exploiting both temporal and spatial differences in PM2.5 reduction targets, we develop a causal inference framework using county-level and individual-level data.We find that PM2.5 reduction targets reduced the mortality rate in treated counties by 1.392‰, rather than PM10. To address issues of self-selection and non-random assignment, we adopted the dual machine learning (DML) method for robust estimation. The reduction in mortality is strongly associated with fiscal autonomy, economic development, and the stringency of environmental performance evaluations, but shows no correlation with geographical distance. We further explore the mechanisms underlying the decline in mortality, including reduced disease risk, fewer traffic accidents, and improved health awareness and behavior. However, despite the policy’s rigor, local governments exhibited a race-to-the-bottom dynamic rather than regional coordination. Our findings underscore the importance of differentiated emission reduction strategies and integrated regional cooperation as critical pathways for achieving sustained improvements in air quality and public health in the next stage of pollution control.