Abortion Politics

The Wall Street Journal and the American Spectator have sunk to an embarassing low with the publication of an article arguing that abortion is reducing the number of Democrats voters - believe it or not, in the literal sense that more of the "mssing voters" would have been Democrats than Republicans.

There are 19,748,000 Democrats who are not with us today.

There are 13,900,000 Republican who are not with us today.

By comparison, then, the Democrats have lost 5,848,000 more voters than the Republicans have.

These Missing Americans--and particularly the millions of Missing Voters--when compounded over time are of enormous political consequence.

I feel sad for people who care so little about ideas that their response to every issue is, will it help us win? Washington is full of people like that. It's been said that politics degrades all debate into an anti-intellectual counting of noses. Now we are counting missing noses.

In addition to being pathetic the article is wrong. Responding may dignify it more than it deserves but the article does at least get the numbers wrong in an interesting way so here goes.

The reduction in population from abortion is far smaller than the number of abortions. How can this be? The relationship between abortion and birth is not mechanical but depends on the choices that women (and men) make among sexual frequency, contraceptive use, fertility, child spacing and other variables. An unmarried, poor teenager who has an abortion may give birth to a child when she is older, married and financially more secure, that she would not have had if she had not had the abortion. The abortion changes the timing of birth but not the total number of births.

Similarly, if abortion were made illegal, contraceptives would be used more often and sexual frequency would decrease. As a result, it is theoretically possible that restricting abortion could reduce the number of births.

In a careful and elegant paper, Levine, Trainor and Zimmerman (1996), find just this. LTZ estimate that restrictions on Medicaid funding of abortions reduced the number of abortions but the number of pregnancies fell even further so the number of births actually went down not up.

Putting things the other way, compensating behavior means that abortion liberalization will reduce the number of births by less than the number of abortions. Five states legalized abortion in 1970, prior to Roe v. Wade (Alaska, California, Hawaii, New York, and Washington). Levine, Staiger, Kane and Zimmerman (1999) estimate that births in these states fell by 5% more than in states that had not legalized abortion. Applying this number to today's rates they estimate "a complete recriminalization of abortion would result in 320,000 additional births per year." Since there are about 1.3 million abortions a year, only about a quarter of all abortions represent a net reduction in births.

The reduction in births, even though considerably smaller than than the number of abortions, is not distributed randomly across the population so abortion policy can have an impact on things like crime and teenage pregnancy but the number of Democrats and Republicans has got to be one of the least interesting consequences.

Hat tip to MemeFirst for alerting me to the article.