Research

This paper examines financial inclusion in the United States, developing new measures of access to financial services and analyzing trends over time. We identify key demographic and geographic determinants of financial inclusion and document recent improvements, particularly in credit access for underserved populations.

We quantify the total stock of balances eligible for the 2022 federal student loan forgiveness policy and explore which groups benefit most. Roughly $441 billion in balances are eligible for forgiveness, which would leave almost 40 percent of federal borrowers with no remaining balance. The borrowers who benefit most, as measured by the ratio of forgiven balances to balances held, are younger, have lower credit scores, and live in lower-income neighborhoods. Compared to other salient fiscal policies, the forgiveness policy distributes less benefit to lower income ZIP codes than the Earned Income Tax Credit, but more benefit to lower income ZIP codes than the 2019 Child Tax Credit and the 2019 education tax credits for higher education. Lastly, we note a recent uptick in credit card and auto loan delinquency for student loan borrowers that may portend more widespread payment difficulties for borrowers if payments resume without relief.

This paper estimates the impact of requiring high school students to complete personal finance education on federal student loan repayment behavior after college. I merge student loan borrowing and repayment data from the College Scorecard with data from the Integrated Postsecondary Education Data System on counts of high school graduates enrolling in college from different states. I estimate the causal effect of personal finance education mandates by relating the change in the share of university students subject to a state mandate to changes in university cohort student loan outcomes. I find only students with higher-income parents respond by adjusting borrowing, reducing median balances by 7%. By contrast, first-generation and low-income borrowers bound by mandates did not significantly adjust borrowing, but were nonetheless more likely to pay down balances. Repayment improvements are in part due to better understanding of the terms governing federal student loans. State mandates that incorporate career research alongside personal finance education are associated with better student loan repayment than those focused only on personal finance education.

Due to the suspension of in-person classes in response to the COVID-19 pandemic, students at universities with earlier spring breaks traveled and returned to campus while those with later spring breaks largely did not. We use variation in academic calendars to study how travel affected the evolution of COVID-19 cases and mortality. Estimates imply that counties with more early spring break students had a higher growth rate of cases than counties with fewer early spring break students. The increase in case growth rates peaked two weeks after spring break. Effects are larger for universities with students more likely to travel through airports, to New York City, and to popular Florida destinations. Consistent with secondary spread to more vulnerable populations, we find a delayed increase in mortality growth rates. Lastly, we present evidence that viral infection transmission due to college student travel also occurred prior to the COVID-19 pandemic.

We exploit the partial deregulation of New York City taxi medallions to provide a causal estimate of the impact of taxi supply on congestion. We employ taxi trip records to measure historical street-level speed. We find that the roll-out of newly authorized taxis caused a local 8-9% decrease in speed. We estimate an empirical congestion elasticity curve from heterogeneous changes in speed and taxi supply, counted from aerial orthoimagery. Additionally, we provide novel urban sensor data to document a substantial traffic slowdown since 2013. Most of the slowdown in midtown Manhattan is accounted for by new supply from ridehail applications.

We examine the impact of financial education on credit decisions during COVID-19. The pandemic presented economic challenges, but policy responses provided opportunities for savvy borrowers. Using variation in state-mandated financial education during high school, we find that mandated borrowers reduced their credit card balances by larger amounts after stimulus checks were distributed and were more likely to buy homes and to refinance mortgages at low rates during the pandemic. The larger credit card balance reduction was driven by middle-income areas and subprime borrowers, while prime borrowers drove mortgage refinancing. Our findings underscore the importance of financial education for economic resilience.

Consumer demand for "Buy Now, Pay Later" (BNPL) has surged, but the specific attributes consumers value remain unclear. We conduct a novel probabilistic stated choice experiment varying BNPL attributes across hypothetical scenarios to estimate consumers' underlying preferences and their willingness to pay (WTP) for each feature. Consumers have a negative WTP for the standard bundle, on average, but younger and lower income consumers have stronger demand. Simulating consumer demand with estimated preference parameters reveals that most shifts away from the standard BNPL bundle reduce demand and create a more negatively selected pool of BNPL users, especially when interest is charged.

Since the 2018 Murphy v. NCAA decision, 38 states have legalized mobile sports betting. We study effects on betting and consumer credit, emphasizing spatial spillovers across state lines. Using consumer spending data and an extended two-way fixed effects framework that separately identifies direct and spillover effects, we find that legalization increases total sportsbook spending roughly tenfold and take-up by 3.1 percentage points. Counties in non-legal states within 15 miles of a legal state experience spillover spending equal to roughly 36% of the direct effect, with these spillovers declining to roughly zero by 60 miles. Using the New York Fed Consumer Credit Panel, we find that median credit scores decline by roughly 1 point and overall delinquency rises 0.3 percentage points from a 10.7% base, with spillover delinquency rising nearly 0.2 percentage points. Under-40 auto loan delinquency increases by half a percentage point and credit card delinquency by one percentage point, driving the overall increase in delinquency. Scaling the population-level delinquency effect by take-up yields implied delinquency increases of roughly 10 percentage points among induced bettors. We conclude with a policy simulation which reveals that spillovers create a fiscal asymmetry: states that have not legalized bear costs from cross-border betting without capturing tax revenue, giving high-exposure states a stronger case for legalization. This incentive is increasing in states that have higher pre-legalization betting activity, population centers near legal states, and a younger population. Methodologically, we show that ignoring spatial spillovers can contribute to attenuated estimates and an under-count of the affected population.

We study self-employment among highly educated, career-oriented business professionals using panel data from prospective graduate management students. Analyzing men and women separately, we assess how cognitive ability, non-cognitive traits, and balanced skills shape self-employment outcomes. For women, non-cognitive traits and balanced skills predict entry. For men, preferences related to work-life balance and job characteristics are more influential. In traditional employment, similar traits raise earnings for both groups. In self-employment, however, returns diverge: women gain more from quantitative skills, while men benefit from balanced non-cognitive traits. Notably, the traits driving entry differ from those linked to later financial success.

This paper examines the impact of a weekend nutrition program on student achievement and attendance in low-income elementary schools in the Mississippi Delta. The intervention provided meals designed to replicate school breakfasts and lunches for weekend consumption. Using a difference-in-differences design, I estimate effects on both the mean and distribution of standardized test scores. Treated students performed better in both language arts and mathematics, with gains of 0.24-0.28 standard deviations in language arts and about 0.16 standard deviations in math. These improvements closed roughly 70-90 percent of the pre-existing test score gap with the state average and were driven by reductions in the share of students at the lowest achievement threshold and corresponding increases at higher levels. Using administrative daily attendance records, I also find significant improvements in attendance: while Friday gains are consistent with the incentive of receiving food bundles, additional increases on Mondays and Tuesdays suggest improved nutrition carried over into the school week. Together, the findings provide causal evidence that weekend feeding programs can meaningfully improve academic outcomes and attendance in food-insecure settings.