Research

Main Findings: Counties with greater recognized wartime sacrifice show larger postwar gains in basic schooling. A one-unit increase in the IHS of native-county martyrs per 100,000 residents in 1953 is associated with a 1.130 percentage-point increase in formal schooling and a 0.813 percentage-point increase in primary completion. The formal-schooling result remains positive under flexible cohort controls.

Contributions: Links 63,739 War of Resistance profiles from the national martyr registry to 6.63 million respondents in the 1982, 1990, and 2000 censuses. The cohort design uses local war timing to study educational recovery within a national expansion. Dated school names and memorial facilities document commemoration. Facility availability during childhood is positively associated with formal schooling.

Previously circulated as War and Human Capital Accumulation: Evidence from the Second Sino-Japanese War.

Main Findings: Famine survivors who later joined the CPC spend significantly more on face-related consumption in adulthood, with larger effects among post-reform entrants, men, rural residents, and those with low trust in government.

Contributions: Using a triple-differences (DDD) strategy to identify how historical famine exposure with political affiliation jointly shape long-term consumption behavior.

Main Findings: Local fiscal incentives tied to high-tax firms increase R&D investment but discourage innovation quality, as firms shift toward strategic rather than substantive innovation.

Contributions: Develops a theoretical framework linking tax-sharing arrangements to innovation misallocation, and implements a generalized DID strategy based on China’s VAT reform. The analysis highlights how tax-base protection and reduced external competition distort firms’ innovation choices.

Main Findings: Using a staggered DID model, we find that the Rural E-Commerce Demonstration County (REDC) policy significantly improves local economic outcomes, increasing per capita income by 10.9% across Chinese counties.

Contributions: Our study provides novel empirical evidence on the role of e-commerce in rural revitalization in a developing-country context. We further apply machine learning models, including random forest, SVM, and ML, to identify and predict the policy's underlying transmission channels.