Article Review

BY ROBERT J. BARRO

DOES ABORTION LOWER THE CRIME RATE?

C rime in the U.S. has fallen dramati- can also explain why the national crime rate cally since 1991. By 1998, the homi- continued to fall through 1998 and is project- cide rate was down by more than one- ed to keep falling for a further 10 to 20 years.

third, and the rates for all violent crime and In fact, the authors estimate that the number property crime were each down by around of abortions and the consequent reduction in one-quarter. crime would have been significantly greater

Many explanations have been offered for had not the 1976 Hyde Amendment, which re- the decline, including increased expenditures stricted federal funding of abortions through on prisons and police, better policing strate- Medicaid, been in effect in various forms. gies, the strong U.S. economy, and the di- Second, a few states, including New York minished role of crack cocaine. Spending on and California, legalized abortion by 1970, prisons and police has been increasing since three years before the U. S. Supreme Court’s the 1970s however, so this factor therefore Roe vs. Wade decision in 1973. As the theory does not explain why crime rose until 1991 implies, the early legalizers experienced falling and then fell sharply. Better policing methods, crime rates sooner than the rest of the coun- as promoted by Mayor Rudolph W. Giuliani in try. Finally, abortion rates responded to the New York, may be making a difference. But legalization differently across states, and those

EXPLOSIVE: crime rates also fell substantially in Los An- with the highest rates of abortion in the 1970s geles and the District of Columbia, which are experienced the sharpest drops in crime in

A new study not renowned for their policing skills. As to the 1990s. the strong economy, it is hard to show gen- WEEDED OUT. The researchers estimate that, erally that growth in income and employment for every 1,000 extra abortions in 1973 to

contends that leads to less crime. For example, the economy 1976, there were 380 fewer property crimes, falling crime

performed well from 1983 to 1989, while na- 50 fewer violent crimes, and 0.6 fewer mur- tional crime rates rose. Finally, the fall in ders in 1997. Overall, the abortion effect ac- crime rates after 1991 occurred in places counted for one-half of the drop in crime from

in the 1990s is a factor.where crack cocaine had never been much of 1991 to 1997. The rest is explained by in-creases in prisons, police, and other factors, or due to the rise A recent study, “Legalized Abortion and goes unexplained.Crime,” by Professors John Donohue of the About 20% of the abortion-related drop in in abortions Stanford University Law School and Steven crime arose because of the reduced populationLevitt of the University of Chicago, proposes of 15-to-24-year-olds (the high-crime age after Roe vs. a new causal factor: the legalization of abor- group) in the 1990s. The main effect, however,tion in the early 1970s. The idea is that the is the reduced frequency of crime among Wade children who were not born would have been those in the 15-to-24 age group who did man-disproportionately likely to grow up in pover- age to be born and grow up. Apparently,

ty and on welfare with a young and poorly abortion particularly weeded out the children educated single parent. Because these fac- who would have been likely to follow criminal tors are known to breed crime, the children careers. not born would have been prime candidates to Donohue and Levitt argue reasonably that be criminals 15 to 25 years later. Hence, the they are carrying out objective scientific re- absence of these children contributes to the search about the determinants of crime and drop in crime rates since 1991. My reactions that the policy implications can be left to oth- to this idea: (1) This is a surprising hypothe- ers. The effect on crime, even if confirmed by sis. (2) It may well be correct. (3) The impli- further study, would probably not moderate cations are explosive politically and are likely the views of pro-lifers, who view abortion as to cause a major ruckus. murder. Similarly, the evidence would have lit- TIMING Donohue and Levitt present three tle influence on pro-choice advocates, who al- types of evidence to support their theory. ready view a woman’s right to an abortion as First, the sharp rise in legal abortions-from a fundamental liberty. But for people with fewer than 750,000 in 1973 to a plateau of less extreme views, including me, the policy around 1.5 million a year since the early implications could be important. If abortion

Robert J. Barro is a professor of eco- 1980s-fits with the timing of the drop in na- rights turn out to be a strong crime fighter, nomics at Harvard University and a tional crime rates since 1991. The cumulative senior fellow of the Hoover lnstitution

then opinion is likely to move in favor of ([email protected]). effect from abortions 15 to 25 years earlier these rights.

30 BUSINESS WEEK / SEPTEMBER 27. 1999