Two widely respected scholars studying the causes of the declining U.S. crime rate, one of the intriguing social puzzles of the decade, have reached a provocative conclusion: Legalizing abortion in the early 1970s eliminated many of the potential criminals of the 1990s.

The research, which has been circulating among economists and criminal-law experts, suggests that those who would have been at greatest risk of criminal activity during the peak crime years of young adulthood - the unwanted offspring of teenage, poor and minority women - were aborted at disproportionately high rates more than two decades ago.

Steven Levitt, a University of Chicago economist, and John Donohue III, a Stanford University Law School professor, conclude that legalized abortion may explain half of the overall crime reduction the nation experienced from 1991 to 1997.

Levitt said the findings support the idea that legalized abortion "provides a way for the would-be mothers of those kids who are going to lead really tough lives to avoid bringing them into the world. They're the ones who are most likely to have been unloved by their mothers, to have faced intense poverty, to have had tough lives."

A copy of the controversial paper, "Legalized Abortion and Crime," was provided to the Chicago Tribune, although it has not been submitted for publication in an academic journal.

The findings have been the subject of three academic workshops, at Harvard University, the University of Chicago and Stanford University.

The authors emphasize that their findings do not constitute an endorsement of abortion, and say their research was motivated by a desire to discover the forces responsible for reducing crime.

In particular, they said, they hoped that research into the reasons for the decline in crime would avert needless public spending on ineffective programs.

But they concede their paper may be attacked as suggesting that abortion has a beneficial social effect or that certain groups should be encouraged to have abortions, an idea they insist they do not advocate.

Levitt acknowledged the possibility that "no one will like it."

But, he added: "I don't think it's our job as economists or scientists to withhold truth because some people are not going to like it. I just think it's important to understand the impact of social policies."

The reactions

When told of the paper, David O'Steen, executive director of the National Right to Life Committee in Washington, D.C., called the thesis bizarre.

"You mean killing unborn babies in the '70s led people in the '90s to do less shoplifting?" O'Steen asked. "I can't believe that any significant percent of the population would argue that we should kill unborn babies to affect whatever they say is being affected."

A spokesman for the National Abortion and Reproductive Rights Action League declined to comment until members of the organization had the opportunity to study the paper.

One of those who has read the paper is Aaron Edlin, professor of economics and law at the University of California at Berkeley, who called it "a convincing case for a very surprising result."

Richard Posner, chief judge of the U.S. Court of Appeals for the Seventh Circuit in Chicago, called it "a striking, original, rigorous and persuasive - although not conclusive - demonstration of the commonsensical point that unwanted children are quite likely not to turn out to be the best citizens."

John Monahan, a law professor at the University of Virginia School of Law who has read the paper, called it one of the most provocative pieces of scholarship he had seen.

"Their thesis is so strikingly original that people are at first taken aback," he said. "After the findings are explained, however, people then shift to the political implications of the article. . . . There's something here for everybody to be upset about."

What they found

In their 45-page analysis, the authors detail the following findings:

• The timing of the crime drop in the 1990s coincides with the period roughly 20 years after the Supreme Court's landmark 1973 decision in Roe vs. Wade legalized abortion nationwide. Thus, the children who would have been born if the pregnancies had not been terminated would have reached the peak ages for criminal activity, roughly ages 18 to 24, in this decade.

• The five states that legalized abortion in the three years before the Supreme Court decision experienced drops in property crimes, violent crimes and murder before the other states.

• Places with high abortion rates in the 1970s experienced big drops in crime in the 1990s, even when accounting for a wide variety of forces that influence crime, such as income, racial composition and incarceration levels. Both individual states and multistate regions with higher abortion rates in the first three years after Roe vs. Wade later saw greater decreases in crime. The subsequent fall in crime was about 15 percent greater for regions with high abortion rates than for regions with low abortion rates. Every 10 percent increase in abortion in the years they studied later led to about a 1 percent drop in crime, the authors found.

• The drop in crime goes beyond what might have been expected simply because abortion led to fewer births of males who reach the peak crime years in young adulthood.

As a result, the authors conclude that the women who chose abortion were those at greatest risk for bearing children who would have been most likely to commit crimes as young adults. These women are teenagers, minorities and the poor - all groups of women who have abortions at higher rates than the overall population of women of childbearing age.

". . . the effect of abortion legalization is still to lower crime even when those women who had previously delayed having children (by resorting to abortion) subsequently increased their childbearing," Donohue and Levitt wrote. "This suggests that it is not simply who has the abortion that leads to the lower crime rate . . . but the ability of the woman to choose better timing for child rearing that lowers criminality."

Why the study was undertaken

Both authors said they began to consider the possible link between abortion and crime years ago, largely because they were astounded at the high rate of abortions.

Nearly one-fourth of pregnancies in the U.S. end in abortion, a rate that is high among the world's developed nations, according to the Alan Guttmacher Institute, a New York City-based organization that conducts research on reproductive issues and advocates reproductive choice for women.

The rate of legal abortions, which reached a high of nearly one in three pregnancies in 1980, has since declined to its lowest level in two decades. The decline is largely because rates of unintended pregnancy have declined as well, according to a recent Guttmacher Institute study.

After abortion was legalized, 8.1 million abortions were legally performed in the 1970s. Since legalization, more than 34 million legal abortions have been performed, 1.37 million of them in 1996, the latest year for which statistics are available.

"I was just stunned at the magnitude of the abortions relative to births," Donohue said. "It's such a huge number that it has to have had some big impact somewhere."

Certain groups of women are roughly twice as likely as the overall population of women of childbearing age to have abortions, according to the Guttmacher Institute. These women are under 25, separated or never married, poor and minorities.

Although white women obtain 60 percent of all abortions, black and Hispanic women have much higher rates of abortion.

However, Donohue downplayed the significance of race. "I don't think it is a racial story. I think it's much more about people who are born under very unfortunate circumstances who will suffer a lot more, and I think neglect, abuse and the attendant anger that occurs because of that can really be a stimulus to crime."

"The notion that there is a correlation between being unwanted and having problems in life is not a new or startling notion," said Cory Richards, vice president of public policy for the Guttmacher Institute.

Although he said he would want to review the paper's findings closely before commenting extensively, he added: "This is not an argument for abortion per se. This is an argument for women not being forced to have children they don't want to have. This is making the point that it's not only bad for the women, but for children and society."

Other possible explanations

The reasons for the rapid fall of crime in the 1990s, the so-called crime "bust," have been intensely debated. Other possible explanations include the increasing use of prisons, more police, improved policing strategies such as those adopted in New York City, declines in crack-cocaine trade, the strong economy and the growing use of security guards and alarms.

"It's a great puzzle," said Daniel Nagin, a public-policy professor who studies crime at Carnegie Mellon University in Pittsburgh. "I haven't heard any convincing explanation for the reasons for the decline."

Nagin and Alfred Blumstein, a criminologist at Carnegie Mellon University and the director of the National Consortium on Violence Research, said the authors of the new study are well-respected. "They've got to be taken seriously," Blumstein said.

"Donohue and Levitt don't have an ax to grind regarding abortion," said Monahan of the University of Virginia. "They just want to see whether it has had an effect on crime rates."

Donohue and Levitt acknowledged their conclusions are speculative. "It would be hard to ever prove this relationship to the degree of certainty that, say, a scientist might want," Levitt said.

Nonetheless, Donohue added, "I think we've amassed enough evidence to make people take the issue seriously."