Historical Religious Conflict and the Persistence of Communal Divisions: Evidence from Temple Destruction in Medieval India (Job Market Paper) Latest Version

Can historical institutional destruction create enduring social divisions that persist across centuries? I examine this question by studying the long-run impacts of Hindu temple destruction during medieval Islamic campaigns in India (11th-16th centuries). Using proximity to medieval military campaigns as an instrumental variable, I find that districts exposed to temple destruction experience at least 167 percent more Hindu-Muslim communal violence during 1950-1995. Falsification tests confirm this effect is specific to religious violence, as temple destruction shows no impact on non-communal violence. The persistence operates through lasting social fragmentation: affected districts exhibit higher employment and residential segregation between Hindu and Muslim communities, and greater spatial separation of religious institutions. These findings reveal how destruction of sacred institutions can generate geographically anchored grievances that become embedded in local social structures across nearly a millennium.

Do Large Sporting Events Increase Traffic Accidents? Quasi-Experimental Evidence from the NFL Latest Version

This paper examines the causal impact of major sporting events, which are known to elevate alcohol consumption and traffic congestion, on the incidence of traffic accidents. Utilizing the quasi-random nature of National Football League (NFL) game scheduling and an extensive accident dataset spanning from 2017 to 2022, I exploit within-region variation over time to estimate the effect of gamedays on accident rates. My main results indicate a 15.60% increase in traffic accidents within the 0-5 mile radius of the stadiums on game days. This effect is largely driven by an increase in accidents in the hours immediately preceding and following the games. Additionally, I find no significant changes in accident rates at the county level or in wider distance ranges from the stadiums.

Civil Debt Collection and Bankruptcy (with Nathaniel Pattison, SMU and Richard M. Hynes, UVA Law)

We develop a new dataset using text analysis and classification algorithms to link civil debt collection lawsuits to individual federal bankruptcy filings. Using this data, we examine the impact of civil debt collection on consumer bankruptcy and heterogeneity in this impact by different classes of creditors.

Legal Barriers to Credit: Evidence from Manufactured Housing and Leasehold Land Policies (with Nathaniel Pattison, SMU)

We exploit spatial regression discontinuity design across state borders using multiple large-scale datasets (property records, mortgage origination data, policy documents) to evaluate how property classification laws shape credit access for manufactured housing buyers.