Did abortion make the crime figures fall?
In the days when I ran an American magazine of ideas, there were certain topics that were always guaranteed to have copies flying off the news stands, readers cancelling subscriptions and half the staff not talking to the editor. Anything on race, anything on genetics, anything on abortion, anything on crime. So I sympathise with all those American editors who have yet to publish anything on a paper floating around the think tanks in America and written by two respected academics from Stanford University law school and the University of Chicago.
The paper makes one simple, gripping claim. It argues that the precipitous drop in crime rates in the 1990s might have something to do with the precipitous increase in abortions in the 1970s. In 1973, when the United States Supreme Court legalised abortion, vast numbers of poor and disproportionately black women who had previously had unwanted children were suddenly able to abort them.

Thus a generation of poor kids who, research tells us, are the most likely group to become criminals, were snuffed out before they were born. Twenty years later, when they would have been in their peak crime-committing years, they are not around. Hence, in part, the plummeting crime rate.

Needless to say, nobody wants to believe the paper. Anti-abortion conservatives are horrified that the phenomenon they abhor - abortion - could be responsible for a development they applaud - falling crime. Pro-abortion liberals are equally horrified that they might be identified with the evil of eugenics, or indeed with anything that might suggest that criminality can be predicted from birth, or that associates crime with the poor and black.

As the social scientist Charles Murray puts it: "It throws a gigantic mudball into the whole debate. Pro-lifers are upset because it offers a rationalisation for abortion and the pro-choice lobby doesn't want to hear that if you stop poor people having children, you will help to solve the crime problem."

So there has so far been a remarkable public silence about the paper, despite its obviously fascinating implications for the debates about abortion, crime and race.

For what it's worth, the paper, which I have been able to read, is a serious and scholarly one. Its authors, John Donohue and Steven Levitt, are respected academics. The methodology is sound; the arguments tight. The authors are able to show a clear and close statistical linkage between otherwise unaccountable drops in crime and increases in abortion rates some 20 years before.

They take into account other factors - such as an unprecedentedly good economy, record imprisonment rates, and a simple decline in the numbers of adolescents in the mid-1990s. But even when all these things are considered, there is still a steep drop in crime that is unaccounted for. And the increase in abortion rates in the 1970s statistically mirrors this decline in almost uncanny fashion.

It may be hard for the British to appreciate it, but in the 1960s abortion was illegal in most American states. It was only at the very end of that decade that a handful of states - New York and California foremost among them - legalised abortion. Then in 1973, the Supreme Court's ruling in Roe v Wade struck down all state prohibitions on abortion. The result was dramatic. The number of abortions in America went from 750,000 in 1972 to 1.6m in 1980. The rate of abortions per 1,000 women jumped from about 15 to about 30 in the same period. The relatively simple premise that Levitt and Donohue make is that unwanted children are disproportionately likely to be brought up by absent or resentful parents and are, therefore, statistically more likely to commit crime in later life.

A whole slew of uncontroversial studies shows this to be the case. What the sudden availability of abortion did, Levitt and Donohue posit, was to remove this pool of unhappy, unwanted children from the population, and so result in a generation of adolescents 15-20 years later who were better adjusted and less criminally inclined.

Makes sense, doesn't it? The explosive nature of the data, however, comes in its analysis that the legalisation of abortion also disproportionately affected the poor and racial minorities.

Racial minorities were already twice as likely to have abortions in 1973 as whites, and after Roe v Wade, their abortion rate soared from about 30 per 1,000 women to about 60 within five years. It peaked in 1978. Removing poor black, unwanted, children would have an especially powerful impact on crime rates 20 years later, Levitt and Donohue predicted. And so, apparently, it has.

The most powerful evidence they provide is related to the fact that New York and California legalised abortion a few years ahead of the rest of the United States. So they experienced a sharp increase in abortion rates four years ahead of the rest of the country.

If Donohue and Levitt are correct, then California and New York should have seen their crime rates drop sooner and faster than the rest of the country in the 1990s. And indeed, that is exactly what happened. Naturally enough, Rudy Giuliani, the mayor of New York, would rather put this success down to his policy of zero-tolerance but New York's early and particularly sharp drop in crime may have had more to do with the abortion factor than with his brilliant political stewardship.

There's mercifully no evidence in the paper that the authors in any way support eugenics; or advocate forced abortions or sterilisation for the poor in order to reduce crime. Quite the contrary. As social scientists, they are simply pointing out a statistical correlation that may help explain an otherwise mystifying piece of very good news. Their argument is also more nuanced than simply saying that future criminals were aborted. They also argue that the children who were not aborted were more likely to be wanted, and therefore may have had a better upbringing than previous generations. So not only were troubled children more likely to be absent, happy children were more likely to be present. The result was a far better environment for rearing, and a far worse environment for criminality.

The hostility to the research in the American culture war - and the remarkable public silence over it - may be because the paper hasn't yet been peer-reviewed or published in an academic journal. We'll see if the silence lasts. It's irrational, whatever your views on the subject of abortion. If you believe, as I do, that abortion is morally wrong in almost every circumstance, it matters not a jot if it leads to a more crime-free society. The ends still don't justify the means. And if you believe that a woman has an absolute right to choose what to do with her own body, then she should have that right, regardless of the consequences.

And as long as nobody advocates using abortion as a tool of eugenic social policy, then the implications of this argument are extremely limited. You might sell a few magazines, and annoy a few readers. And eventually, heaven forfend, you might even stumble onto the truth.

~ Andrew Sullivan, "The Sunday Times", 15 Aug 1999.


Return to Politics index, or Site homepage.