American families changed a lot starting in the 1960s and 1970s. Two years stand out in particular: 1960, when the birth-control pill entered the market, and 1973, when the Supreme Court decided Roe v. Wade. A new study makes the provocative argument that the latter, not the former, is what really prompted Americans to get married and start having children at older ages than they used to. (Hat tip to Tyler Cowen; free draft of the paper here.)
The Pill had two different effects. One, obviously, it allowed its users to have sex with a hugely reduced chance of pregnancy. But two, precisely because it had this effect, it lowered the cost of sex and spurred people to have more of it. Because the Pill is far from perfect as a contraceptive — mainly thanks to human error, about 9 percent of Pill users become pregnant during the first year — this cuts into the reduction in unintended pregnancy we would otherwise expect.
Indeed, according to the study, the two effects roughly canceled each other out. Surveys show that sexual activity increased when the Pill was introduced and gained acceptance, but marriage and childbirth didn’t follow the pattern one would expect. For instance, “There is a downward trend [in fertility] that begins with the 1940 cohort, which was 20 years old when the pill was introduced, but this is largely arrested between the 1950 and 1955 cohorts, who experienced rapid diffusion of the pill.” A more sophisticated analysis — looking at state-by-state variation in whether the Pill was legal and available to young, unmarried women – shows basically no effect when it comes to marriage, “shotgun” marriage, and childbirth among women 16 to 22.
Abortion is a different story. Marriage and first-childbirth ages did change for women who were 16 or younger when it was decided. And another state-by-state analysis — this one looking at variation in whether states had legalized abortion before Roe or allowed minors to consent to medical care (which had an unexpected result after the decision) — confirms the finding. “Policy environments in which abortion was legal and readily accessible by young women are estimated to have caused a 34 percent reduction in first births, a 19 percent reduction in first marriages, and a 63 percent reduction in ‘shotgun marriages’ prior to age 19,” the author writes.
These conclusions don’t have strong policy ramifications; they just help to explain history. The results on contraception certainly undermine efforts to shove birth-control pills at teens in the hopes it will reduce pregnancy — but these days, that activism is shifting toward “long-acting, reversible” forms of contraception such as the IUD and implants, which don’t have the same user-error problem. The abortion results are even less politically salient; we pro-lifers object to the procedure because it takes an innocent life, which will overshadow just about any other effect, good or bad, including those put forth in the study. In addition, though abortion rates soared after Roe, they’re actually slightly lower today than they were in 1973, so abortion’s effect on marriage and childbearing today is probably different from what it once was.
I also expect to see some pushback on the results. It’s surprisingly difficult to suss out which states changed their abortion and contraception policies in which years; the new study disagrees with previous work on up to 27 (!) of the states’ contraception policies, for example. The author seems to have done incredibly thorough research, looking at everything from statutes to opinions from state attorneys general, but some of these decisions are subjective. And of course there are a variety of other methodological choices involved as well.
In general, though, this is a fascinating study that moves the literature forward. Half a century after the sexual revolution, we’re still trying to figure out what hit us, so every little bit helps.