Background & Objectives: Nearly half of all pregnancies in the United States (US) are unintended (i.e., mis-timed or unwanted), and roughly 5% of US women experience an unintended pregnancy each year, suggesting the population-level need for contraceptives is not being met. Further, these pregnancies are experienced disproportionately by women who are younger, women of color, and women of lower socioeconomic status - indicating these groups are particularly underserved. The Patient Protection and Affordable Care Act (ACA) had the potential to improve our ability to meet this population-level need by increasing access to and affordability of contraceptive products and services. There is evidence that the ACA increased health insurance coverage and is associated with an increase in the use of highly effective long-acting reversible contraceptives, but it is unclear whether these effects translated into fewer unintended pregnancies. Further, it is unknown whether these effects were equitably distributed across race and ethnicity. Thus, the objectives of this dissertation are to: 1) estimate the overall impact of the ACA on unintended pregnancy, and if evidence of an impact exists, describe the timing of this impact, 2) explore three mechanisms of the ACA by investigating the impact of three major provisions (i.e., the dependent coverage provision, Marketplace subsidies, and ACA insurance expansions), and 3) assess the impact of the ACA on racial/ethnic disparities in unintended pregnancy. Methods: Data from multiple cross-sectional cycles of the National Survey of Family Growth (NSFG) were used. NSFG uses a multistage probabilistic sampling methodology to survey non-institutionalized civilian men and women regarding family planning, marriage, divorce, and both general and reproductive health. I included sexually active female respondents aged 18-44 interviewed between 2006 and 2019 (n=25,426). To address objectives (1) and (2) I used a difference-in-differences approach to compare trends in unintended pregnancy between women who were eligible to benefit from the intervention (the overall ACA or one of the listed components), to that of women who were ineligible to benefit. Eligibility was determined by respondent age and income. To address objective (3), I used a pre/post analysis to explore how racial/ethnic disparities in unintended pregnancy differed prior to and following enactment of the overall ACA and its components. Results: There was evidence that: 1) the overall ACA was associated with a 2.1 percentage point (ppt) decrease in unintended pregnancy among eligible women, and this decrease was fairly consistent during and following the ACA's implementation period, 2) the dependent coverage provision was associated with a large (8.2 ppt) decrease in unintended pregnancy among lower income young women, and 3) the disparities in unintended pregnancy between Hispanic and non-Hispanic (NH) White women and between NH Black and NH White women decreased by 2.9 ppt and 4.1 ppt, respectively, among eligible women following full implementation of the ACA. There was insufficient evidence that the Marketplace subsidies or insurance expansions were associated with unintended pregnancy, or that the dependent coverage provision, Marketplace subsidies, or insurance expansions were associated with racial/ethnic disparities in unintended pregnancy. Conclusions: The overall ACA and the dependent coverage provision may be associated with reductions in unintended pregnancy, and the magnitude of these associations appear to differ across sociodemographic subgroups (i.e., income, race/ethnicity) - holding implications for health equity. These findings provide insight regarding how the ACA works to influence reproductive health, and for whom - which is critical information for both researchers and public policy makers who seek to improve reproductive health and health equity.
Read
