ENGO Participation and Firm Emissions: Evidence from China’s PITI Program

Description

This dataset is constructed to examine whether environmental non-governmental organization (ENGO) participation can mitigate governance distortions under China’s decentralized environmental regulatory system. Local officials may prioritize economic growth over pollution control, creating a principal–agent problem between central and local governments. The dataset therefore evaluates whether third-party environmental information disclosure can supplement formal regulation and improve firm-level environmental performance. Using the introduction of the Pollution Information Transparency Index (PITI) as a quasi-natural experiment, the dataset supports analysis of its effects on firm-level chemical oxygen demand (COD) and sulfur dioxide (SO₂) emissions. The dataset combines firm-level, city-level, and policy-level information to construct a firm-year panel. Firm-level variables include pollutant emissions, production and financial characteristics, firm size, administrative rank, pollution intensity, and tax contributions. City-level variables capture local economic, industrial, fiscal, and environmental conditions. Policy indicators identify the implementation of PITI and other relevant pilot programs. Empirical results show that ENGO participation significantly reduces firm-level COD emissions. The estimated SO₂ effect is also negative, but smaller and less precisely estimated. The effect operates mainly through strengthened enforcement incentives, with suggestive evidence of more stringent regulatory instruments. The reductions are concentrated among more visible firms. For both pollutants, stronger effects are found among firms with higher pollution intensity and greater tax contributions. For COD, the effects are also more pronounced among larger firms and firms with higher administrative rank. Decomposition results indicate that the COD reduction mainly reflects enhanced end-of-pipe treatment, whereas the weaker SO₂ response is associated with lower pollution-generation intensity. The dataset is compiled from the AESPF, ASIF, CSMAR, WIND, Chinese statistical yearbooks, and official government documents. It can be used to replicate the associated empirical analyses and support further research on environmental regulation, information disclosure, decentralized governance, and firm environmental behavior.

Steps to reproduce

Software Requirement: The replication package is developed using Stata (version 18 or higher). Package Installation: Ensure that community-contributed commands (e.g., 'reghdfe', 'ivreghdfe') are installed via 'ssc install'. Execution: Set the working directory to the folder containing the files. Run '03_main.do' to replicate all empirical results, including tables and figures presented in the manuscript.

Institutions

• Beijing Institute of Technology

  Beijing, Beijing
• Zhejiang University

  Zhejiang, Hangzhou

Categories

Licence