Just Proven: Prisons Do Keep Down Crime

KEEPING criminals in prison reduces crime rates by two or three times more than conventional wisdom holds.

That conclusion of a new study doesn't mean even more people should be popped in jail, says the author, Harvard University economist Steven Levitt. ''Somehow the criminal-justice system has fumbled its way to the right balance.''

The fact that taking criminals off the streets reduces crime may seem like common sense. But it has been hard to prove using statistics. Looking at nationwide numbers, criminologists have found little correlation between crime rates and the number of people in jail. The incarceration rate in the United States has more than tripled in the last two decades. At year-end 1994, the prison population exceeded 1 million. One in every 175 US residents was in prison, compared with 1 in 450 at the end of 1980. Annual government outlays on prisons are roughly $40 billion. Among industrial nations, only Russia has a higher incarceration rate.

Despite the growth in the number of prisoners, the crime rate continued to rise until recently. The reported rate of violent crime per capita in those two decades almost doubled; property crimes per capita rose 25 percent. Such facts led some commentators to label an increasing reliance on imprisonment a policy failure. They urged the use of alternative correctional programs, decriminalization of drug offenses, or a moratorium on new prisons.

''We cannot build our way out of the crime problem,'' Massachusetts Chief Justice Paul Liacos said last week of new prisons.

But if more criminals hadn't been tucked away in prison, the crime rate might have risen even faster because of underlying crime determinants such as gang involvement, more single-parent families, and a lack of jobs for teenagers, Mr. Levitt says.

Levitt has sorted out the impact of prisons on crime from other factors by looking at crime rates and prison populations in states where prisoners' rights groups, especially the national prison project of the American Civil Liberties Union (ACLU), have succeeded in winning civil suits alleging that overcrowding in prisons was ''cruel and unusual punishment'' and thus unconstitutional. In 12 states, the entire prison system has been under court order concerning overcrowding (rather than only a portion of the prison facilities). Levitt uses statistics from these 12 states for his analysis in a National Bureau of Economic Research paper.

In these states, prior to filing of the litigation, the growth in the number of prisoners per year was 4.2 percent above the national mean. After filing, prisoner growth rates fell below the national average in nine of the 10 states.

Levitt then looks at what happens to crime rates in the states where litigation has forced action on overcrowding. Courts do not usually tell the criminal-justice system how to solve an overcrowding problem, just that it has to be eliminated. This leaves to the state the decision as to whether to build new prisons, shuffle prisoners among prisons or place them in another state, slow the number of individuals put into the prison, or release prisoners. Few did the latter. Levitt finds that three years after a final court decision, violent-crime rates in these states are 11.2 to 11.6 percent higher than they would have been otherwise, and property-crime rates are 6.2 percent to 6.4 percent higher.

Levitt concludes that for each one-prisoner reduction induced by prison overcrowding litigation, the total number of crimes committed in the state increases on average by 15 per year. This closely accords with the level of criminal activity reported on average by prisoners when asked in surveys. These results, Levitt says, stand up well, across all the crime categories he examines, ranging from murder and rape to burglary and automobile theft.

He estimates the cost to individuals and the nation of 15 crimes to be about $45,000. Such an analysis allows a price to be put on violent as well as property crimes. That compares with the per-prisoner incarceration costs of about $30,000 a year. The social benefit of putting a small-time criminal in jail is likely below $45,000, he figures. For them, another penalty may be better.

''The finding that prisons appear cost-beneficial does nothing to reduce the importance of identifying and correcting those factors that lie at the source of criminal behavior,'' he states.

Jenni Gainsborough, an ACLU spokeswoman, says crime needs real solutions in more drug treatment, education, and job training - programs proven to work, rather than politicians' ''knee-jerk, tough-on-crime sound bites.''