Research

Abstract: This paper examines the effect of sibling sex composition on marriage outcomes in the United States using a comprehensive, newly-constructed dataset from Texas covering individuals born between 1976-1997. Leveraging both singleton and twin analyses, we find robust evidence that having a same-sex sibling significantly increases the likelihood of marriage and accelerates marriage timing. We also reveal substantial heterogeneity across racial groups: White individuals consistently show strong effects, while Black and Hispanic populations sometimes exhibit distinctive patterns, including reduced divorce rates for Black men with same-sex twins and delayed marriage for Hispanic men. Furthermore, we document significant wealth-based variation, with sibling sex effects on marriage predominantly present in affluent counties. This socioeconomic divide suggests that parental resource allocation strategies, sibling relationships, and cultural marriage norms vary substantially by wealth status. Our results illuminate the complex interplay between family structure, economic conditions, and cultural factors in shaping marriage decisions, highlighting the critical importance of considering both racial background and socioeconomic context when studying family formation patterns.

Building wealth over lifetimes became possible for a broader span of the population in developed countries over the 20th century compared to any time in history. This was driven by more people having the capacity to save because of the expansion of middle-class jobs and education, access to highly developed financial markets, and government support for real estate investment. Housing wealth remains the dominant wealth-building vehicle for those outside the top decile of the income distribution. This, coupled with the high and growing level of residential segregation and local allocation of public goods in countries such as the United States, drives the unequal ability of individuals to build wealth depending on neighborhood of origin and residence. Segregated neighborhoods are drawn along racial and class lines, and while much progress has been made, historical and structural factors such as the legacy of slavery have contributed to the difficulty of fully closing the Black–White wealth gap. More generally, children who grow up in lower-status areas are significantly less likely to rise up the wealth and status ladder, and this is driven by a variety of disadvantages in those neighborhoods. These include higher levels of pollution; worse public services, especially education; and fewer prospects for jobs and training. Some of these can be changed by moving individuals and families to better neighborhoods, while the effects of a polluted environment on the development of 0- to 5-year-olds have long-lasting and often irreversible consequences. These factors have kept the “American Dream” of equality of opportunity and the ability to save and build wealth as individuals and households out of reach for significant portions of society. There is renewed interest in infrastructure investments and place-based policies to address this opportunity gap, which, due to its scale, is beginning to be recognized as having negative implications for the aggregate economy. Economists should maintain their focus on these important questions and continue to improve data sets as the range of assets in which people can build and store wealth grows.

Abstract: Using 873 polities spanning 3400 BCE to 1500 CE on a fixed quarter-degree grid, we estimate a centrality gradient with great-circle distance and human mobility cost, showing cells farther from the center of a historical polity were less densely populated. Cross-sections show this gradient steepens with a cell's cumulative years under rule, and a within-cell panel identifies it from plausibly exogenous border movements. As evidence of public good provision and accumulation, Roman road density declines with distance holding population fixed. Distance to the historical center still explains where modern capitals sit, but not modern population density or nighttime lights.

Abstract: This study examines the labor market impact of Korean Free Economic Zones (KFEZs) on regional unemployment and economic inactivity. KFEZs, first introduced in Busan and Incheon in 2003, were designed to attract foreign investment through tax exemptions, tariff reductions, and regulatory incentives. Using monthly panel data from 16 South Korean regions between 2000 and 2020, this research applies the Generalized Synthetic Control (GSC) method to address challenges of staggered treatment adoption and violations of the parallel trends assumption. Contrary to the expectation that KFEZs would stimulate job creation and economic participation, results indicate that treated regions experienced a small but statistically significant increase in unemployment rates (0.15 percentage points) and a substantial rise in the economically inactive population (approximately 89,000 individuals). These findings suggest that, rather than improving labor market outcomes, KFEZs may introduce structural shocks that increase economic inactivity in long term. 

Abstract: Despite decades of efforts to close gender gaps in mathematics, women remain underrepresented in STEM. One prominent explanation is comparative advantage: girls perform relatively better outside STEM, making non-STEM fields more attractive. Yet recent replication evidence finds this channel explains little of the gap. A key limitation is that two-domain comparisons cannot distinguish comparative advantage from unobserved heterogeneity in absolute skill, and lack the variation to characterize relative strengths across fields. We exploit the Texas Academic Decathlon, where the same student is tested across ten academic subjects on the same day, and standardize scores within-person to isolate comparative advantage from absolute ability. Males tilt toward STEM ($+0.21\sigma$ in math, $+0.18\sigma$ in science) while females tilt toward humanities ($-0.20\sigma$ in literature, $-0.13\sigma$ in art). The gap doubles among top-quartile students and translates into differential competitive outcomes. Because all ten subjects are contested under identical conditions, the within-student design rules out competition aversion and isolates domain-specific differences; whether these originate in ability, stereotype threat, or socialized effort allocation, the comparative advantage prediction for field sorting is the same. These findings suggest that two-domain frameworks substantially understate gendered comparative advantage, and that closing average math gaps addresses the wrong margin.

Abstract: This paper examines how Christian identity interacts with political ideology to shape student attitudes toward cancel culture on U.S. college campuses. Using data from the Foundation for Individual Rights and Expression (FIRE) Campus Free Speech Survey, I analyze support for three disruptive tactics: using violence to stop a campus speech, blocking entry to events, and shouting down speakers. Among Protestants, liberal students are more supportive of disruptive tactics than their atheist peers, while conservative Protestants are consistently less supportive. Among Catholics and Orthodox, liberals are more supportive of violence and blocking but less supportive of shouting, while conservatives generally do not differ from atheists. These results show that Christianity’s influence on cancel culture attitudes is not uniform: for Protestants, Christian identity amplifies ideological divides, whereas for Catholics and Orthodox, the patterns are more selective and internally inconsistent. I propose potential mechanisms, including identity conflict and church-based mobilization, to account for these divergent patterns.