Does Abortion Decrease Abortion?

Alex links to a cool article by Donohue, Grogger, and Levitt, which claims that legalized abortion is responsible for the recent decline in unmarried teen childbearing. This is not as obvious as it sounds, because the drop in unmarried teen childbearing began about 16 years after Roe v. Wade. DG & L attribute the decrease to an abortion-induced drop in the number of teenagers who would be likely to get pregnant. In short, people aborted the fetuses that would have been most likely to get pregnant 13 to 19 years later. Here’s how DG & L put it:

While the link between the current availability of abortion and teen childbearing is straightforward (since abortion is an alternative to carrying a baby to term), the relationship we focus on is far more subtle: legalized abortion in the 1970s led to fewer babies being born under circumstances in which their parents were less willing or able to provide nurturing environments. When these cohorts grew up to be teenagers, their improved childhood environment had the benign effect of reducing the frequency with which they themselves became teen mothers.

Now here’s an interesting possibility: abortions in the 1970s may have had a downward impact on the number of abortions in the 1990s. DG & L’s results show that unmarried teen childbirths decreased in the 1990s as a result of earlier abortions. So far as I can tell (from skimming the article), they do not break down the different routes for the reduction in childbirths – some of the reduction would be from abortions, some from pregnancies that never occurred in the first place. Thus, earlier abortions presumably helped to reduce later abortions.

This is, of course, a partial offsetting effect, not a total effect. Looking at DG & L’s Figure 3, we see that the total number of abortions rose from 1973 to 1981 and then stabilized. But interestingly, the figures are not per capita (unless they’ve been population-adjusted without any note of that fact), which means there must have been a decline in abortions per capita since 1981. I’m not sure where that would put the latest per capita abortion figures relative to pre-1973, but I’m guessing the recent figures would still be higher. Nonetheless, it’s fascinating that abortion could actually produce its own offsetting effect.