Poster Paper: Heterogeneous Effects of Sex Education on U.S. Teen Birth Rates

Thursday, November 7, 2019
Plaza Building: Concourse Level, Plaza Exhibits (Sheraton Denver Downtown)

*Names in bold indicate Presenter

Nicholas D.E. Mark and Lawrence Wu, New York University

The United States has one of the highest teen birth rates among rich countries (Sedgh et al., 2015), a distinction that has sustained the interest of researchers, politicians, and the public for decades (Murcott, 1980; Luker, 1997; Edin and Kefalas, 2005). This interest has corresponded with an extremely rapid drop in the teen birth rate since its peak in 1991 (Santelli and Melnikas, 2010), and the investment of billions of federal dollars in funding for sex education. This investment has been divided between support for ”comprehensive” programs that cover contraception and medically accurate information on pregnancy and sexually transmitted infections, and ”abstinence-only” programs that rely on a message that abstinence until marriage is the only acceptable strategy for a healthy adolescence (Santelli et al., 2006). While a substantial body of research has examined the effects of abstinence-only education, less is known about the effectiveness of comprehensive sex education at scale. Continuing federal and state support for both types of sex education, and recent reallocations of federal funding away from comprehensive programs toward abstinence-only ones, make this a pressing current policy issue.
Both comprehensive and abstinence-only programs have garnered significant attention from researchers interested in evaluating their effectiveness. Abstinence-only programs and state mandates have been rigorously evaluated at the level of individual schools and students (e.g. Trenholm et al., 2008) and at the population level (Cannonier, 2012; Kearney and Levine, 2012; Carr and Packham, 2016), and have generally been found to be ineffective at reducing the teen birth rate. Comprehensive sex education has been evaluated using local randomized control trials (e.g. Farb and Margolis, 2016) but not at the population level.
This paper aims to fill that gap in the literature by providing causal estimates of the effect of comprehensive sex education on teen birth rates. Using data from restricted birth certificates, county-level population estimates, and federal funding, we employ two-way fixed effects models that adjust for county-level and state-year variation to estimate the effect of comprehensive and abstinence-only sex education on county-level teen birth rates. This effectively compares counties in the years they received funding to other counties in the same state and year that did not receive funding. We also conduct two sets of robustness checks and report a range of possible effect sizes.
We replicate previous null findings for the effects of federal funding for abstinence-only sex education, we find that funding for comprehensive programs decreased the teen birth rate in affected counties by between 3 and 10%, depending on the choice of assumptions. Second, we introduce new estimates by single-year age groups which indicate that the effects of sex education on birth rates were more than twice as large for women aged 14-16 (5-12%) than for women aged 17-19 (2.5-4%). These contributions highlight the greater effectiveness of comprehensive as compared to abstinence-only sex education, and place renewed emphasis on the role of public policy in shaping demographic outcomes. Before the conference, we will also estimate heterogeneous effects by race/ethnicity.