In new research, James Flynn, David Munro, and Caitlin Myers document how sexual behavior has changed in light of total abortion bans enacted after Dobbs vs. Jackson Women’s Health. They find that abortion bans reduced sexual activity and increased contraceptive use, motivated by individuals having an increased awareness of state policy.
In June 2022, the Supreme Court ruled against the constitutional right to an abortion in Dobbs v. Jackson Women’s Health Organization. This landmark decision did more than reshape the legal landscape around abortion access. It may have also changed the sexual behavior of many Americans.
In a new paper, we study whether the total abortion bans enacted in 13 states in the wake of Dobbs induced people to change whether and how they have sex. The question is a natural one: if abortion is no longer available as a backstop against unintended pregnancy, rational individuals facing higher costs from an unwanted pregnancy might respond by having less sex, using contraception more consistently, or both.
The literature on how abortion policy shapes fertility is extensive. Researchers have documented that abortion restrictions reduce abortion rates and raise birth rates, and that post-Dobbs bans in particular have increased births in affected states by roughly 2-3%. However, fewer studies have examined whether the same restrictions shape the decisions that happen upstream of pregnancy: whether to have sex, and whether to use contraception. Whether they actually do is an empirical question that our paper attempts to answer.
Prior work found little evidence of upstream behavioral responses, possibly because people were not aware of abortion restrictions until Dobbs and national coverage brought the restrictions to their attention. Many states had previously tightened access to abortions—including parental involvement laws and clinic closures encouraged by mandating hospital-like standards—but attracted little national attention and had no apparent effect on sexual activity. Surveys from that era confirm as much: most women were unaware of the abortion laws in their own states. This means that many women who might otherwise have sought an abortion may not have discovered how difficult it would be for them to access one until after a pregnancy has already been conceived.
The Dobbs decision was different. When the constitutional right to abortions was stripped, total bans in many states shuttered abortion facilities across a wide swath of the South and Midwest, generating sustained national media coverage and likely increasing awareness of the availability of abortion. If decisions to have sex and what, if any, form of contraception to use depend on perceived access to abortion, then we might expect the post-Dobbs abortion bans to produce larger effects on sexual activity than prior literature found in response to other, less conspicuous shocks to abortion access.
To test this, we assembled three independent sources of evidence.
The first was administrative surveillance data on gonorrhea diagnoses from the Centers for Disease Control and Prevention. Because gonorrhea spreads almost exclusively through unprotected sexual contact, is highly symptomatic, and requires antibiotic treatment, reported infection rates serve as a biological marker of risky sexual behavior. The second source of evidence is retail scanner data from NielsenIQ, which tracks weekly barcode-level sales at grocery stores, drug stores, and mass merchandise retailers nationwide, allowing us to measure condom purchases at the state and county level. Finally, we use the Youth Risk Behavior Survey, which captures self-reported sexual behavior among high school students and is fielded in the spring of odd-numbered years.
Our strategy compared the 12 states that enacted total abortion bans to our 24 control states that protected abortion access, before and after Dobbs. In our synthetic difference-in-differences approach, we adjusted control states to match the pre-treatment trends in treated states, rather than their levels. This matters in our setting because ban states and protected states differed substantially in their baseline sexual activity: gonorrhea rates were about 28% higher in ban states, and condom purchases were roughly 40% lower. Event-study plots confirm that these groups were on parallel trends through 2021 and diverged only after Dobbs.
The results are consistent across all three data sources. Total abortion bans reduced gonorrhea rates among the population aged 15 to 44 by 21% relative to states that protected access, with declines apparent for both men and women and across every five-year age group we examined. Condom purchases rose by about 5% in ban states relative to protected states. These findings are robust across estimators, functional forms, and sample restrictions, and a series of leave-one-out analyses confirmed that no single state is driving any of the three results. We also found that high school students in ban states reported being 21% less likely to have ever had sex, having 32% fewer recent sexual partners, and being 29% less likely to have engaged in unprotected sex, though these findings are largely driven by respondents in Texas, and their precision is somewhat sensitive to our choice of standard errors.
Several additional tests sharpened the interpretation. We found no effect on chlamydia rates, which is consistent with expectations because chlamydia is often asymptomatic and diagnosed through screening rather than self-presentation, making it a noisier marker of short-run behavioral change. We found no effect on syphilis, which is concentrated among men who have sex with men and so should not respond to shifts in the perceived cost of pregnancy. In the NielsenIQ data, we found no effect on purchases of toothpaste or athlete’s foot medication, two products that are unrelated to reproductive behavior, but which might track other underlying changes to consumer behavior.
We also can speak to the mechanism. Work by Daniel Dench, Caitlin Myers, and Mayra Pineda-Torres finds that the fertility effects of Dobbs operate primarily through increases in driving distance to abortion providers: births rose most in counties where the ban produced the largest distance increases. If adjustments to sexual behavior worked the same way, we would expect them to be largest in counties where distance to the nearest provider grew the most.
We tested this directly using county-level models that jointly included a binary indicator for total ban enforcement and a flexible control for distance to the nearest provider. The distance coefficients are small and statistically insignificant for all three of our outcomes, and adding distance to the model barely moved the estimated effect of the ban itself. This suggests that the behavioral responses we document are driven by awareness of the law, not by changes in the travel cost of reaching a provider. This is exactly what we would expect if people are updating their perceived cost of sex in response to a salient legal shock rather than in response to marginal changes in the ease of abortion access.
Along with other studies in the field, our paper has an important implication for how we think about the aggregate fertility consequences of Dobbs. Our calculations suggest that the shift toward more protected or less frequent sex could have reduced pregnancies by somewhere between 0.2-3.4% of the baseline. While other studies have documented a small increase in fertility rates, our paper suggests that fertility rates would have been substantially larger if not for people changing if and how they have sex.
The bulk of the fertility effect reflects abortion access: women who become pregnant, want an abortion, and cannot obtain one because they cannot travel, cannot afford it, or cannot navigate the legal and logistical obstacles of out-of-state or mail-order access. Our accounting exercise implies that roughly 15-20% of women in ban states who become pregnant and seek an abortion are prevented from obtaining one. As some policymakers predicted, some individuals adjusted their behavior to avoid pregnancy. However, many more bore unwanted births.
Authors’ Disclosures: The authors report no conflicts of interest. You can read our disclosure policy here.
Articles represent the opinions of their writers, not necessarily those of the University of Chicago, the Booth School of Business, or its faculty.
Subscribe here for ProMarket’s weekly newsletter, Special Interest, to stay up to date on ProMarket’s coverage of the political economy and other content from the Stigler Center.
