The Impact of Roe v. Wade on American Fertility (UPDATED)

The Issue:

The reversal of Roe v. Wade returns the legal status of abortion to the states. A careful accounting of the impact of reversing Roe requires understanding the effects Roe had in the first place. A starting point for any such discussion is to consider the impact of abortion legalization on U.S. fertility that occurred in the early 1970s.

The legalization of abortion reduced births in the U.S., with larger reductions in some regions and demographic groups. As a result, it altered the pattern of living circumstances and subsequent life outcomes of the population of children born after abortion became legal.

The Facts:

• By the time Roe came along, millions of women already had access to legal abortion in five states. Before the Roe v. Wade decision, abortion had already been legalized three years earlier in New York, Washington, Alaska, and Hawaii through legislation, and in California abortions were legal as a result of a state Supreme Court decision. Women in those states had access to legal abortions as did the large number of women who crossed state lines to obtain legal abortions prior to the Roe decision. (I cover this history and its impact in greater detail in Sex and Consequences:  Abortion, Public Policy, and the Economics of Fertility.)
• Access to legal abortion reduced the number of births per woman in the United States. In order to determine the impact of legalizing abortion on number of births, it is important to differentiate its effect from other factors that were contributing to declines in fertility — such as changes in the availability of contraception, shifts in social attitudes and improved labor market opportunities for women. In past research, I have compared patterns in birth rates across the states that legalized abortion early relative to those that legalized via the Roe decision to estimate its impact. Differences in birth rates between those two groups of states were relatively stable in the 1960s, when abortion was largely illegal in the entire country (see chart). In 1971, however, birth rates dropped by around 5 percent in those five states that had legalized abortion relative to the other states. In 1974, after the Roe decision was handed down, that difference was reversed as birth rates fell in states where abortion had just become legal (perhaps with a bit of a time lag as abortion facilities did not appear instantly) and then it stabilized once again. Recognizing the nine-month lag between conception and birth, this pattern supports the conclusion that abortion legalization reduced birth rates in the United States in the 1970s. Based on these calculations, we estimate that roughly 125,000 fewer births occurred annually because of the Roe decision.
• The impact on births differed by demographic group. This estimate masks tremendous variation in the impact of Roe across demographic groups. Births to teens and to black women fell by around 12 percent (data are not available to study Hispanic women separately during that period). Births to women between 35 and 44 fell at a rate twice the national average and the impact on unmarried women was twice as great as it was for married women.
• Interstate travel to obtain an abortion was common prior to Roe. Using similar research methods, I found evidence that births among women living in states near those that offered legal abortions declined much less in response to Roe because, in effect, many of them already had access to legal abortion via travel. If abortion were legal in New York, then access to abortion would be available to many women in New Jersey and Connecticut, limiting the impact of the Roe decision in those states. In Texas, however, travel to a legal abortion provider prior to Roe would have been far more difficult.
• Changes in abortion policy today may not have the same effect. Certainly, the world is a different place today than it was 45 years ago, and this makes it difficult to draw sharp parallels with past experience. For instance, long-acting contraception did not exist and it is now becoming more popular. Air travel is also much cheaper (after adjusting for inflation) than it was in the era of a regulated airline industry.
• If Roe v. Wade were reversed, the legality of abortions would vary from state to state. Some states are more likely to outlaw abortion but several others would likely maintain its legality. The Center for Reproductive Rights categorizes states according to their likelihood of outlawing abortion if Roe v. Wade were overturned. They report that the legal status of abortion is less likely to change in 20 states (including DC; the others are mainly in the Northeast, Florida, and the West), is moderately likely to change in 8 states (scattered – Illinois, North Carolina, Colorado, New Hampshire, and others), and highly likely to change in 24 states (mainly in the South, Midwest, Central, and Mountain states). The 20 states that would likely preserve access to legal abortions reflect a disproportionate share of the country’s population. Thus, even if all the other 31 states outlawed abortion, the estimated increase in births extrapolating from the pre-Roe v. Wade experience would likely be well under 100,000 per year.
• If lessons from the past hold, we would expect that reversing Roe v. Wade would have differential effects by region and demographic group. Geographically, the change would have a greater effect on women who live furthest from a legal abortion state and those of lower economic means in those states would be particularly affected because of the greater cost of travel. Demographically, if the factors at play in the 1970s are still current, we would expect the same groups who experienced larger reductions in births when abortion was legalized would face larger increases in births if it were outlawed. This would include teens, older women of childbearing age, unmarried women, and black women. Lower-income women would also likely experience a greater impact from the policy change.
• The characteristics of children born would likely change. Births averted in the presence of legal abortion are also not representative of all births. Evidence shows that those children who were not born as a result of abortion legalization in the early 1970s would have been more likely to die as an infant, grow up in a lower income, single parent household, or receive public assistance (see here). When those children grew up, they would have acquired less education and been more likely to receive welfare and be single parents (see here). Presumably outlawing abortion now would increase the incidence of these outcomes.