Regulations about abortion are often wildly controversial. But what effects to they actually have? Caitlin Myers addresses these issues in “From Roe to Dobbs: 50 Years of Cause and Effect of US State Abortion Regulations” (Annual Review of Public Health 2025, pp. 433-446).
As a starting point, consider the years before and after the 1973 US Supreme Court decision in Roe v. Wade that struck down existing abortion restrictions across the country. The left-hand panel shows the states which has repealed the bans on abortion before Roe in purple, those that had relaxed but not eliminated their ban before Roe in pink, and those in which abortion was legalized by Roe in gray. In the purple states that had already repealed their ban on abortion, the number of abortions had risen in the years before Roe, but had then started declining–and the decline continued after the passage of Roe. Part of the reason for the decline in the early-legalization states is that, after Roe, women no longer had to travel from other states where abortion was illegal. In the other groups of states, the number of abortions rose.
As Myers argues, the effects on abortion levels of states that repealed their abortion bans before 1973 is very large–probably larger than the increase in abortion following the Roe decision. She writes:
Of the three broad policy changes liberalizing abortion access—early reforms, early repeal, and repeal with Roe—it is early repeal that results in the greatest effects on national abortion and birth rates. As Joyce et al. (51) conclude following a detailed analysis of the effects of distance to early repeal states, “The story that emerges from these data is that…Roe v. Wade was arguably less important for unintended childbearing than was access to services in California, the District of Columbia and especially New York in the years before Roe” (pp. 813–14) because so many people were able to travel to these early repeal states even if their state of residence had not yet legalized abortion.
States then tested the limits of what the Supreme Court would allow with a variety of restrictions: mandatory waiting periods before an abortion, mandatory counseling before an abortion, different types of content that might be involved in that counseling, parental permission for teenager and/or spousal permission for wives, whether Medicaid funding could be used to pay for abortions, whether abortions needed to be performed in or near hospitals, what doctors were allowed to perform abortions, and others. This array of rules–as they were proposed, passed or failed in legislatures, and were upheld or not by courts–provides a rich set of contexts for researchers.
Here’s one example. In North Carolina in the 1980s and into the 1990s, there was a state fund to pay for abortions for low-income women: in this way, the state did not draw on federal Medicaid funds to pay for abortions. But the state fund sometimes ran out of money. Myers writes: “Cook et al. (25) exploit a natural experiment that took place within North Carolina between 1980 and 1994 when the state abortion fund ran out of money on five different occasions. Comparing changes in outcomes among women seeking abortions and eligible for funding, the authors conclude that when funding is unavailable, about one-third of pregnancies that would have been terminated are instead carried to term …”
This kind of study is referred to as a “natural experiment”–that is, there was no plan for the North Carolina fund to run out of money. It seems unlikely that sexual activity in North Carolina was being adjusted according to the state of the fund. Instead, some North Carolina women seeking abortions found that funding was available, and others didn’t, and this had an effect on their deicsions.
Myers goes into detail in considering the array of natural experiments that have been analyzed. For example, when a state altered its abortion laws, then women who lived relatively close to that state were also affected, because it was relatively easy for them to travel to that state, while women living farther from that state were less affected, because their costs of travelling to that state were higher. As another example, those interested in, for example, the application of difference-in-differences statistical methods may want to check out the paper.
Here, I’ll mention some of the bottom lines of this survey of the evidence (citations omitted here, but appear in the article iteself): When and where abortion is more restricted, birth rates are higher. Higher birth rates, especially for women at younger ages, are associated with lower levels of educational achievement, and thus with lasting effects on employment outcomes. These effect are typically larger for black women then for white women.
What about the period since the 2022 US Supreme Court decision in Dobbs v. Jackson, which struck down Roe v. Wade and thus gave states much wider latitude in setting abortion laws? Of course, the evidence on this point is still evolving, and the setting for abortion is now rather different than it was before 1973. Myers notes:
- “Abortion prior to 12 weeks’ gestation remains legal in 34 states (65) and many states have bolstered their protections (22), providing many more destinations than existed in 1971, when abortion was legal in only 6 jurisdictions.”
- “The delivery of abortion services has also evolved, with a major shift occurring in 2000 when the US Food and Drug Administration (FDA) approved the drug mifepristone for the termination of pregnancies. The proportion of medication abortions grew rapidly, from 6% of all abortions in 2001 to 39% in 2017.”
- “[I]n December 2021 the FDA lifted the restriction permanently (55), allowing health care providers to dispense abortion medications directly to patients via mail without requiring the patient to receive in-person consultation or tests (85). This expanded abortion access in the 32 states that did not restrict telehealth abortion (5), likely fueling the rise in medication abortions to 63% of all abortions by 2023 … By the end of 2023, telehealth accounted for nearly 1 in 5 abortions in the United States (83), and national abortions had actually risen relative to pre-Dobbs levels …”
- “Yet not everyone seeking an abortion can find a way to drive hundreds of miles to reach facilities in nonban states or will find telehealth medication abortion an acceptable option. Near-total abortion bans enforced in the first 6 months after Dobbs are estimated to have increased births in ban states by an average of 2.3% relative to if no ban had been enforced (26). The estimated effects of bans on fertility are greater in states where distances are greatest, reaching 4.4% in Mississippi and 5.0% in Texas …”
In addition, teenage birth rates have fallen dramatically over the last three decades for an array of reasons not directly related to availability of abortion: less sexual activity, greater use of contraception, and more broadly, a larger share of young women viewing their early adulthood as a time for education and job experience, with later ages for marriage and childbearing.
