Publications
Window Dressing: Changes in Atmospheric Pollution at Boundaries in Response to Regional Environmental Policy in China
Journal of Environmental Economics and Management, Volume 125 (May 2024)
[Journal link] | [manuscript]
Abstract: In decentralized environmental governance, local governments are likely to adopt the "beggar-thy-neighbor" strategy to relax regulations at boundaries. This study investigates the impact of China's Joint Atmospheric Prevention and Control Policy (JAPCP) in "2+26" cities enforced by the central government on pollution at provincial boundaries. The theoretical model suggests that dual incentives for environmental protection and economic growth may prompt local governments to reduce boundary pollution within the JAPCP-covered area while relocating pollution to uncovered boundaries. Based on Shandong Province data using the difference-in-differences approach, our analysis reveals a 9.6% decline in the air quality at JAPCP-covered boundaries compared to non-boundary areas and a 5.3% increase at JAPCP-uncovered boundaries, which is associated with migration of key regulated industries. Through examining annual work reports, we provide evidence that local governments modify regulatory intensity at various boundaries. These findings indicate that, while regional environmental policies are intended to promote inter-jurisdictional cooperation, the local government responses lead to unintended costs.
Working Papers and Selected Works in Progress
Segmented Empire, Integrated Regions: Multilayered Administrative Boundaries and Labor Market Fragmentation in Late Imperial China
Abstract: This study investigates the impact of administrative boundaries on labor market integration in late imperial China during the 18th and 19th centuries, focusing on the multilayered boundary effects within the Qing Dynasty. Using a novel individual-level dataset derived from homicide court records (Xingke Tiben) and a structural gravity model, the research finds that xiaqu (higher-level administrative divisions) boundaries significantly deterred labor migration, reducing migrant flows by approximately 28%, while province boundaries within the same xiaqu had weaker or even promoting effects. The study also explores the consequences of these boundary effects on labor income convergence and spatial allocation. First, we structurally estimate the relative real income levels across regions and find that the correlation between real incomes of cross-xiaqu location pairs is significantly weaker. Second, our analysis reveals that the size distribution of prefectures within each xiaqu adheres more closely to Zipf's law when compared to the national level distribution, while exhibiting patterns similar to those observed at the provincial level. These findings highlight the hierarchical segmentation of labor markets and the role of administrative divisions in shaping labor mobility, suggesting that such segmentation may have contributed to regional imbalances and hindered national economic integration.
The "Demons" in Demonstration: Unintended Consequences of Superstar-oriented R&D Subsidy Policy in China
– Developed from my Master's thesis, recipient of the "Outstanding Master's Thesis" distinction, Renmin University of China (2025)
– "Outstanding Paper" at The 50th Beijing Camphor Economic Circle (CEC) Seminar (2025)
[SSRN (November 2025)] (New version!)
Abstract: This paper examines the unintended consequences of selective R&D subsidy policies that target leading firms, focusing on the innovation activities of non-subsidized competitors and new entrants. Although such subsidies are typically intended to stimulate innovation, the paper utilizes a Schumpeterian model to demonstrate that they may, in fact, discourage innovation among competitor firms and deter the entry of new firms, thereby undermining aggregate innovation. Empirical analysis of China's "National Technological Innovation Demonstration Firm (NTIDF)" policy supports these predictions: R&D expenditures increase by 21.7% for subsidized firms, but fall by 30% for lagging competitors and private firm entry declines by 7.6%. On average, quality-weighted patent output drops by 8.3% within affected city–industry pairs. Incorporating these findings and potential spatial spillovers into a quantitative trade model, counterfactual analysis reveals that the policy leads to a 0.049% reduction in national real GDP, despite covering only 3.3% of city–industry pairs before 2018. Therefore, this paper highlights the necessity of accounting for competitive dynamics in the design of innovation policies.
Lineage Ties and Migration Dynamics: Clan Strength, Policy, and Demographic Divergence in Qing China
(in progress)