Research

This paper examines how the quality of emission disclosures influences public beliefs about pollution and trust in reported data. I focus on a nationwide platform that provides daily emission records for all municipal Waste-to-Energy (WtE) plants in China. I combine emission data from 2020 to 2022 with a second-hand housing dataset spanning 92 cities. I estimate pollution beliefs using a residential sorting model where households base housing choices on perceived pollution exposure and the reliability of disclosed information. I find that, in most cities, pollution is perceived as higher when emission records are flawed, compared to when they are valid.  I then explore possible sources of the prevalent "over-perception," including plant manipulation and city-specific characteristics. Such divergence, however, can be mitigated by enhancing long-run information quality, defined as the valid rate of emission records over the three years. Finally, I estimate an average marginal willingness-to-pay of $4.13 for a 1% improvement in information quality within a reference period. The substantial substitution effect between information quality and distance to the plant reveals that better-informed households are more willing to live closer to WtE plants, suggesting reduced local opposition to these facilities.

Multi-level environmental governance often encounters coordination challenges, particularly in transboundary regions. This study investigates whether the introduction of the River Chief System (RCS) improves water quality by assigning water management responsibilities to top local government officials. Using firm-level water pollution data and the personal backgrounds of provincial and city River Chiefs (RCs) in the Pearl River Basin from 2015-2019, I find that the RCS reduces water pollution, with a stronger effect near jurisdictional boundaries. Further analysis of informal connections among RCs shows that the policy effect is more pronounced when a city RC has prior overlapping work experience with a provincial RC, due to stronger political alignment. In contrast, the effect weakens when neighboring city RCs share the same birthplace or alma mater, due to competition for promotion and diminished cross-border cooperation. However, the influence of these horizontal connections decreases when one of the connected city RCs is part of the provincial RC's network. These findings suggest that informal networks can facilitate coordination and collaboration in collective actions.

This paper investigates how informal networks shape intergovernmental fiscal distribution in four Southeastern provinces of China. We use newly collected data that contain background information on allocation-relevant officials and focus on the Urban Maintenance and Construction Funds. We find that provincial top leaders allocate significantly more funds to cities governed by officials with the same birthplace or overlapping work histories. However, cities with top leaders from the same universities as any directors of allocation-relevant provincial departments receive less funding. We analyze heterogeneity across cities and connection types and discuss two mechanisms behind our results – trust and competition.  Finally, we investigate how these network-based allocations affect the promotion of officials and the efficiency of fund use. 

Increasing sample size is an intuitive way to cope with weak identification caused for instance by a variable with low variability or the presence of many fixed-effect parameters capturing unobserved heterogeneity. Computing nonlinear estimators can become tedious with large samples but subsampling aggregation (subagging) can restore scalability as proposed by Politis (2024). This paper studies weighted versions of this procedure with a special focus on GLS-type weights.