Measuring abortion

Public Interest, Wntr, 2005 by Charles Murray

SEX and Consequences: Abortion, Public Policy and the Economics of Fertility [dagger] is a model of contemporary social science discourse, revealing in one book both how the enterprise should be conducted and its vulnerability to tunnel vision on the big issues.

Phillip B. Levine, a professor of economics at Wellesley College, sets out in Sex and Consequences to explore the thesis that the role of abortion is akin to the role of insurance. Legal abortion provides protection from a risk (having an unwanted child), just as auto insurance provides financial protection against the risk of an accident. Legalizing abortion has a main effect of reducing unwanted births, just as auto insurance has a main effect of reducing individuals' losses from auto accidents.

But abortion faces the same problems of moral hazard as other kinds of insurance. Just as a driver with complete insurance may be more likely to have an accident, a woman who has completely free access to abortion may be more likely to have an accidental pregnancy. Levine hypothesizes that legislated restrictions on abortion might serve the same purpose as deductibles do on auto insurance--they alter behavior without having much effect on net outcomes. Thus a state with some restrictions on abortion may have no more unwanted births than a state without restrictions, even though the number of abortions is smaller in the restrictive state. The restrictions raise the costs of abortion, and women moderate their behavior to reduce the odds of an unwanted pregnancy.

Levine develops his model carefully and with nuance, and eventually wends his way back to conclusions about its empirical validity (it is broadly consistent with the evidence). But the chapters between the presentation of the model and the conclusions about it are not limited to the insurance thesis. They constitute a comprehensive survey of the quantitative work that has been done on the behavioral effects of abortion, incorporating analysis of the abortion experience worldwide as well as in the United States.

THE book's virtues are formidable. Levine writes clearly, avoids jargon

(or explains what the jargon means when he can't avoid it), and is unfailingly civil in characterizing the positions in the abortion debate. He is judicious, giving the reader confidence that he is not playing favorites when the data are inconclusive or contradictory. The breadth and detail of the literature review are exemplary. The book is filled with convenient summaries of material that could take a researcher weeks to assemble--a table showing the differences in abortion policy across European countries plus Canada and Japan, for example. Levine also gets high marks for one of the most challenging problems for any social scientist who is modeling complex human behavior: making the model simple enough to be testable while not losing sight of the ways in which it oversimplifies the underlying messiness of human behavior.

The book's inadequacies reflect not so much Levine's failings as the nature of contemporary social science. Abortion policy is one of the great moral conundrums of our time. Anyone who is not the purest of the pure on one side or the other has had to wrestle with the moral difference (or whether there even is one) between destroying an embryo when it is a small collection of cells and when it is unmistakably a human fetus. None of the tools in Levine's toolkit can speak to this problem. Levine is aware of this, and makes the sensible point that more argumentation on the philosophical issues is not going to get us anywhere. He has picked a corner of the topic where his tools are useful, he says, and that's a step in the right direction. Still, as I read his dispassionate review of the effects of abortion policy on the pregnancy rate, I could not help muttering to myself occasionally, "Aside from that, Mrs. Lincoln, how was the play?" *

EVEN granting the legitimacy of looking where the light is good, Sex and Consequences may be faulted for sheering away from acknowledging how much scholars could do to inform the larger issues if they were so inclined. Here is Levine discussing the non-monetary costs of abortion:

   The procedure may be physically unpleasant for the patient. She may
   need to take time off from work and spend time traveling to an
   abortion provider that may not be local. When she gets to the
   provider's location, there may be protesters outside the clinic,
   making her feel intimidated or even scared. If her family and/or
   friends find out about it, she may feel some stigma. Finally, it
   should not be overlooked that the procedure may be very difficult
   psychologically for a woman in a multitude of ways that cannot be
   easily expressed.

"Cannot be easily expressed"? The woman is destroying what would, if left alone, have become her baby. That's easy enough to express. That Levine could not bring himself to spit out this simple reason why "the procedure may be very difficult psychologically" is emblematic of the tunnel vision that besets contemporary social science.