The Effects of State Abortion Bans on Reproductive Health and Mortality: A Bayesian Hierarchical Low-Rank Difference-in-Differences Approach

Abstract

The 2022 Supreme Court decision in Dobbs v. Jackson Women’s Health Organization eliminated federal constitutional protections for abortion, prompting 13 states to implement total or six-week abortion bans. These bans created an abrupt reduction in reproductive autonomy for tens of millions of women, concentrated in states with already fragile maternal and infant health infrastructure. This dissertation investigates the population-level consequences of these policies for fertility, infant mortality, and maternal mortality using national vital statistics data and a Bayesian hierarchical low-rank difference-in-differences framework. I first establish that abortion bans produced a sustained increase in births. Using birth certificate and Census data through 2024, I estimate that the fertility rate in states with bans remained approximately 2% above expected levels through the second full post-implementation year, with increases concentrated among women of color and women whose highest educational attainment was a high school diploma. The persistence of this elevation despite expanded interstate travel and telehealth access suggests that financial, logistical, and legal barriers continue to limit abortion access for disadvantaged populations. I then ask whether these additional births were accompanied by worsening infant health. The neonatal mortality rate in states with bans was 6–7% above expected levels in both 2023 and 2024, an excess driven by deaths from congenital anomalies, which were 13–15% above expected. The infant mortality rate among Black infants was 9–10% above expected, while no significant differences were observed among other racial and ethnic groups. These patterns are consistent with two mechanisms: the forced continuation of pregnancies with life-limiting fetal diagnoses, and the disproportionate inability of Black women to access alternative pathways to abortion care. Finally, I examine whether abortion bans were associated with increased maternal mortality. Using restricted-access mortality data through 2023, I estimate that the maternal mortality rate in states with bans was approximately 9% above expected levels, though credible intervals included zero across all analyses. This may reflect insufficient statistical power given the rarity of maternal death and limited post-implementation follow-up, or a lack of an underlying effect. Continued surveillance and analyses of more prevalent outcomes such as severe maternal morbidity are needed to clarify whether abortion bans affect maternal mortality at the population level. Together, these findings reveal a chain of consequences: abortion bans increase births among the most disadvantaged populations, and these additional births are accompanied by elevated neonatal mortality, particularly from congenital anomalies. Evidence on maternal mortality remains inconclusive. The concentration of fertility and infant mortality effects among racial and ethnic minority groups and socioeconomically disadvantaged women suggests that these policies widen existing disparities in reproductive and infant health.