Current working papers and manuscripts under review. Drafts are available upon request. For published work, see the Research page.
Under Review
Actually I Don’t: The Impact of Warrantless Arrest Laws on Divorce and Marriage
With Ben Brewer. Under review at Review of Economics of the Household, April 2026.
Between 1976 and 2016, most U.S. states adopted laws authorizing or requiring police to make warrantless arrests at the scene of a domestic violence call. These laws were intended to protect victims, but they may also deter victims from calling police in the first place. We estimate the effect of mandatory, preferred, and discretionary arrest provisions on marriage and divorce using CPS microdata on more than 6 million adults, recently corrected law classifications, and difference-in-differences methods robust to staggered adoption. We find no detectable effect on the probability of being divorced or married at the population level. The null holds across estimators, outcome definitions, the CPS ASEC, the CPS June Fertility and Marital History Supplement, and vital statistics records. A simulation-based power analysis confirms we can rule out effects as large as those found for unilateral divorce reform; supplementary enforcement analysis suggests reporting deterrence as one possible channel.
JEL: J12, K36, K42, I18. Keywords: domestic violence, warrantless arrest, mandatory arrest, divorce, marriage, difference-in-differences.
The Impact of Anti-Bullying Laws on Youth Organ Donation
With Ben Brewer. Under review at Journal of Economics & Human Biology, April 2026.
Since 1999, all 50 U.S. states and D.C. have adopted anti-bullying legislation (ABLs), which have been shown to reduce suicides among 14 to 18 year-old females. Using the universe of deceased organ donors from 1993 to 2024, we employ a difference-in-differences design to estimate ABLs’ spillover effects on organ supply. ABLs reduce per capita suicide-based organ donors aged 14 to 18 among females by 0.007 per 100,000, about 88 percent of the sample mean, with no significant effect for males. The decline is broad-based across major donated organ types and is not accompanied by evidence of lower transplant quality. Though ABLs impose an organ-supply cost, the benefits of prevented suicides clearly dominate; the loss nevertheless underscores the urgency of expanding living donation and increasing organ procurement yields.
Builds on our earlier JPAM paper on anti-bullying laws and adolescent suicidal behaviors. See Research.
JEL: I12, I18. Keywords: anti-bullying laws, organ donation, suicide, teen mental health.
Majoritarian Politics and Violence Against Caste-Based Minorities: Evidence from India
With Devika Hazra. Under review at Economic Development and Cultural Change, April 2026.
We examine whether the 2014 electoral victory of the Bharatiya Janata Party (BJP) led to an increase in crimes against Scheduled Castes in India. Using a difference-in-differences design and administrative data from the National Crime Records Bureau covering 36 states and union territories over 2006 to 2019, we find that states where BJP won majority vote shares in 2014 experienced roughly a 40 percent increase in Scheduled Caste atrocity rates relative to control states in the post-election period. The finding is robust to falsification tests with alternative post-treatment cutoffs, jackknife estimation, and alternative functional forms of the outcome variable. Placebo tests using broader mortality outcomes yield null results, providing little support for a general deterioration in law and order. Mechanism analysis identifies a statistically significant decline in chargesheeting and arrest rates specifically for Scheduled Caste crimes in treated states, consistent with weakened institutional deterrence rather than a spurious reporting effect.
JEL: D72, J15, K42, O17. Keywords: majoritarian politics, caste-based violence, Scheduled Castes, India, difference-in-differences.
Majoritarian Politics and Gender-Based Violence: Evidence from India
With Devika Hazra. Under review at Journal of Development Studies, February 2025.
We examine whether the 2014 electoral victory of the Bharatiya Janata Party (BJP) led to an increase in crimes against women in India. Using a difference-in-differences design on state-level administrative crime data, we find that states where BJP won majority vote shares in 2014 experienced roughly a 49 percent increase in reported crimes against women in the post-election period relative to control states. The finding is validated through event studies and a range of falsification tests. Full abstract available upon request.
JEL: D72, J16, K42, O17. Keywords: majoritarian politics, gender-based violence, crimes against women, India, difference-in-differences.
Working Papers
Are LGB Individuals with Disabilities Underserved? Evidence on Disparities in Social Security Disability Benefit Receipt
With Dhaval Dave and Joseph J. Sabia. March 2026.
This paper examines disparities in Social Security Disability Insurance (SSDI) benefit receipt between LGB and non-LGB households with disabilities, drawing on survey data spanning the COVID-19 pandemic period. Supported by the Social Security Administration (via the UW Madison Retirement and Disability Research Center). Full abstract available upon request.
JEL: H55, I14, J14, J15. Keywords: SSDI, LGB, disability, health disparities, social insurance.
From the Field to the Safety Net: Title IX, Teen Fertility, and Women’s Welfare Program Participation
April 2026.
Title IX’s mandated expansion of girls’ high school athletics is widely credited with raising women’s human capital, but its welfare consequences have not been examined. Using the same identification strategy as Stevenson (2010), I show that the policy’s effects were sharply heterogeneous across the population of compliers. For Black women and women in states with smaller pre-existing boys’ programs, Title IX reduced teen fertility and raised schooling and employment, consistent with the human capital channel. For the larger set of compliers, predominantly white women in states with large pre-Title IX boys’ programs, the policy raised teen birth rates and increased the probability of receiving means-tested transfers in adulthood. A new mechanism test shows the teen-birth effect is concentrated in states with restrictive pre-Roe reproductive policy environments and is indistinguishable from zero in permissive states, consistent with exposure to mixed-sex social environments raising fertility where access to reliable fertility control was limited.
Companion to our JHR paper on Title IX, female athletic participation, and crime. See Research.
JEL: I18, I38, J13, J16, J18. Keywords: Title IX, athletic participation, teen fertility, welfare participation, instrumental variables.
NAFTA and US Crime: Disentangling the Trade and China Shocks
With Devika Hazra. April 2026.
Did NAFTA exposure raise local property crime in the United States? The effect has been hard to pin down because the tariff phase-in overlapped with a larger second shock, the rise of Chinese imports culminating in Permanent Normal Trade Relations in 2001, and the two shocks hit largely the same manufacturing-heavy counties. We construct a county-level NAFTA exposure measure from 1990 industry employment shares and the Mexican tariff reductions phased in between 1993 and 2001, and we include the Autor-Dorn-Hanson China shock as an explicit control. A one-standard-deviation increase in NAFTA exposure raises the county-level burglary rate by 0.62 per 1,000 population, roughly 5 percent of the sample mean, and the larceny rate by 0.85 per 1,000. Rates of aggravated assault, robbery, murder, and every weapon-specific violent crime we examine are statistically unchanged. Together the pattern is consistent with an economic-distress interpretation in which the response appears in property rather than violent offenses. The results survive event-study pre-trend tests, shift-share randomization inference with demeaned shocks, leave-one-industry-out checks, and outlier exclusion. A coarser specification that uses only state-level manufacturing share yields a coefficient approximately equal to the sum of the disentangled NAFTA and China effects, suggesting that earlier estimates attributing the full increase to NAFTA were conflating the two trade shocks.
JEL: F14, F16, K42, R11. Keywords: NAFTA, trade liberalization, property crime, China shock, shift-share, local labor markets.
The Structure of Suicide Mortality in NAFTA-Exposed U.S. Commuting Zones, 1990–2016
April 2026.
I estimate the effect of the 1993–2001 NAFTA tariff phase-in on U.S. suicide mortality using commuting-zone variation in exposure to Mexican tariff cuts. A one-standard-deviation increase in exposure raises the commuting-zone suicide rate by 0.64 per 100,000 over 1990–2007 (5 percent of baseline), rising to 0.67 through 2016. The effect loads on heavy manufacturing rather than textile and apparel, and within suicide on firearm and intentional-poisoning methods. It concentrates in prime-age 25–44 cohorts. Accidental drug overdose is null in every window, separating the NAFTA mortality channel from the post-2001 opioid-supply shock. Event-study pre-period coefficients are flat and jointly consistent with parallel trends.
Companion to the NAFTA crime paper above; looks inside the trade-shock mortality response.
JEL: F14, F16, I12, I18, J61. Keywords: NAFTA, trade shocks, deaths of despair, suicide, mental health, shift-share.
Did Recreational Marijuana Lower High School Graduation? Evidence from Colorado
With Zachary Fone. April 2026.
Has recreational marijuana legalization lowered high school graduation? We estimate the effect of retail exposure on four-year graduation rates in Colorado, the first U.S. state to legalize. The standard single-state dose-response specification on the full 2010–2022 change yields a large positive coefficient. Re-running the same specification on the 2010–2013 pre-treatment window, before any retail store had opened, recovers about 83 percent of the post-period slope, and a within-district school-level proximity DiD across 425 Colorado public schools matched to NCES geocoded coordinates returns a tight null. Taken together the results suggest the published positive single-state effect can pick up urban-rural convergence rather than a treatment effect; we recommend the pre-treatment placebo as a routine robustness test for single-state RML–education designs.
JEL: I21, I28, K42, H75. Keywords: recreational marijuana legalization, graduation rates, pre-treatment placebo, difference-in-differences, Colorado.
Up in Smoke: The Impact of Medical Marijuana Laws on Organ Donation
With Ben Brewer. April 2026.
This paper estimates the effect of medical marijuana laws (MMLs) on organ donation using a state-year panel from 1993 to 2024. Exploiting the staggered adoption of MMLs across U.S. states within a two-way fixed effects framework, we find no statistically significant effect of MMLs on deceased organ donation. Event study evidence shows no pre-trends and no post-treatment shift for all-cause donors. The main result is robust to multiple heterogeneity-robust DiD estimators designed for staggered adoption, sensitivity bounds for omitted variables, and alternative timing conventions. Cause-specific decompositions are similarly null for MVA, suicide, homicide, and cerebrovascular donors individually. The findings suggest that medical marijuana legalization has not generated large changes in the organ supply.
JEL: I12, I18, K32. Keywords: medical marijuana, organ donation, traffic fatalities, dispensaries, difference-in-differences.
Pediatric Vehicular Heatstroke: Heat Exposure, Cabin Biophysics, and the Limits of State-Level Policy Identification
May 2026.
Using the universe of U.S. pediatric vehicular heatstroke deaths (970 children under five, 1998 to 2025) and a case-crossover design with a purpose-built cabin-temperature index, I estimate the heat-mortality dose-response on closed-vehicle exposure and evaluate two state-level policy responses: caregiver criminal liability and Good Samaritan civil immunity. Caregiver criminal liability shows no detectable effect on heat-day excess risk. Good Samaritan immunity is directionally negative on the bystander-rescue subset under the primary case-crossover specification but does not consistently replicate across designs and data sources. Full abstract available upon request.
JEL: I18, K42, Q53, Q54. Keywords: pediatric mortality, case-crossover, heat exposure, hot car laws, Good Samaritan, identification.
Work in Progress
“Trade Shocks, Birth, and Mortality: The Population-Health Response to NAFTA and USMCA in U.S. Commuting Zones, 1985-2023.”
A companion project to the NAFTA crime paper, examining population-health outcomes across U.S. commuting zones over nearly four decades of trade regime change. JEL: F14, F16, I10, J13.
“Anti-Hazing Laws, Mental Health, and Student Success.” Examines whether state anti-hazing laws affect adolescent and young-adult mental health and educational outcomes. JEL: I12, I18, I21, K36.
