Weighing Imprisonment and Crime

In the early 1990s, with violent crime at record levels and public alarm growing, federal and state lawmakers responded with new policies that sent more offenders to prison for longer periods. The federal Violent Crime Control and Law Enforcement Act of 1994, in particular, made sweeping changes to U.S. correctional policy by imposing longer prison sentences for federal crimes and encouraging states to implement similar penalties. Two decades later, the nation’s prison population has soared and crime has fallen to levels not seen since the 1960s.

The Pew Charitable Trusts’ public safety performance project interviewed nine leading scholars about the relationship between imprisonment and crime. Although they tackle the question from different perspectives, these experts agree that the relationship between imprisonment and crime is complex and that a number of other factors are at work in the nation’s crime decline. The nine scholars:

Q: What does research tell us about why the crime rate has fallen?

Richard Rosenfeld: I'd say the research has returned a fairly consistent message. The record peacetime expansion of the economy in the 1990s clearly played an important role. The enormous growth in imprisonment also played a role, at least in the short run, although there are lots of qualifications to be mentioned there. And the aging of the country’s population certainly was a factor. Beyond that, things are a bit more speculative and subject to debate.

William Spelman: When you get down to it, nobody really knows. A bunch of us tried to examine the crime drop of the ’90s that extended into the last decade, but with a few exceptions, we could not find any policy levers that had a clear and consistent effect on the crime rate. Potential explanations include demographic shifts, better economic opportunities, changes in drug markets, new police strategies, and increased imprisonment. But most attempts to parcel out responsibility among these explanations are unpersuasive.

Alfred Blumstein: Two major factors were the declining demand for crack and police aggressiveness in capturing guns from young people. The growing demand for crack in the 1980s led to major recruitment of young people to accommodate the demand and to replace others who had been sent to prison. It was those young people, with their handguns, who were largely responsible for the dramatic rise in homicides— largely young black males against young black males—from 1985 to 1993. The demand for crack began to decline after 1993, and so young people were no longer needed as sellers, and the robust economy could readily absorb them. They no longer needed their guns, and police aggressiveness in finding and confiscating the guns provided an incentive to put the guns away, which further reduced violence.

Philip Cook: There’s a lot of evidence that putting more police on the streets, like we did in many cities through the COPS [Community Oriented Policing Services] program in the mid-1990s, affected crime rates. We don’t know what all those police were doing, but I think one lesson was that if you give modern police chiefs more resources, they can reduce crime. Beyond that, using police in a more targeted, systematic way certainly had an effect. Frank Zimring’s work on New York City is a case in point. But the extent of that influence on overall crime rates is not clear.

Q: What do we know about the specific impact of increased incarceration on the crime rate?

Spelman: If the question is, “Has the dramatic expansion of our incarcerated population had an effect on crime?” then the answer is unequivocally yes, of course. If you put more bad guys in jail so they can’t get at the rest of us, then you will experience a crime reduction. How much? Our research concluded that about 25 percent of the 1990s crime drop can be linked to increased incarceration. Incarceration rates flattened out in the 2000s, but crime rates kept dropping. So incarceration can’t account for the last 15 years of reductions.

Rosenfeld: The best we can say is that mass incarceration had a significant but not overwhelming impact on crime, accounting for perhaps 10 to 25 percent of the reduction.

Daniel Nagin: These claims about what role the increase in our incarceration rate has played in reducing crime are complicated. I also don’t find it that useful because the effect of imprisonment on crime rates depends upon what policy we’re talking about. Incarceration expanded because of a host of different policy changes affecting who goes to prison and for how long. So it’s not possible to make a blanket statement about all the very, very different policies that influenced this trend.

Jeremy Travis: When you quadruple the incarceration rate, as we did in this country over a period of four decades, the key question is whether this big investment in prison resulted in a significant decline in crime. The conclusion of our [National Academy of Sciences] committee was that there is no clear answer from a scientific point of view. So states and the federal government invest more than $80 billion a year in this system with the expectation of a big payoff in terms of crime rates, and most studies show that the impact has been quite modest.

Blumstein: The impact of increased incarceration was limited because such a large fraction of that growth was associated with drug offenses. About 20 percent of state prisoners and 50 percent of federal prisoners were sentenced for drug offenses, and the ability of incarceration to impact drug offending is inherently limited. When a rapist is locked up, you remove his rapes from the street; when a drug seller is locked up, you recruit a replacement, and there is no impact on the transactions, which are driven largely by demand.

Franklin Zimring: Prison building was generally helpful in terms of the crime rate in the ’90s, and not so much after that. Also, New York City is a very interesting story we can’t ignore. Although the incarceration rate went up in the United States by 65 percent between 1990 and 2007, apparently explaining the crime drop, the incarceration rate declined 28 percent in New York. And guess which city led all others in crime reduction during that time? You got it, New York. I talk about it as defying gravity. New York had major, unprecedented crime declines from the peak rates in 1990. Homicide down 82 percent. Robberies down 84 percent. Burglaries down 86 percent. Auto theft down 94 percent. If these were animal breeds, they’d be on the endangered species list.

As you increase the number of people you put behind bars, you dip into a population that is increasingly less criminally active, so the benefits of incarceration diminish.Steven Raphael, University of California, Berkeley

Q: Has the incarceration rate reached a point of diminishing public safety returns?

Steven Raphael: As you increase the number of people you put behind bars, you dip into a population that is increasingly less criminally active, so the benefits of incarceration diminish. We have expanded the scope of what is punishable by incarceration and, in the process, are netting less and less dangerous people on average. You see the effect of this in a number of ways. The prison population has shifted toward less severe offenses. And the average inmate admitted is five years older than he or she was 15 or 20 years ago.

Mark A.R. Kleiman: With respect to prison building and crime control, the half-million or million beds we added were probably helpful. Everything beyond that—the second million additional inmates—was pretty much a waste. Because as you increase the incarceration rate, you start to send away people for less-severe offenses, like a sixth bad check or a third auto theft. These guys are not exactly Al Capone. So over time, the average nastiness of the prisoners goes down and the public safety benefit is diminished.

Spelman: It costs something like $30,000 per year to house, feed, guard, and provide medical care for a single prisoner, and of course incarcerating them takes them out of the economy, where the vast majority had a legitimate job and were only stealing and dealing in their spare time. So when you get to these marginal offenders, the cost of the crime you prevent by pulling them out of society is considerably less than that $30,000 a year.

Rosenfeld: There is good evidence that after a certain point, as prison populations continue to grow, the benefit of incarceration declines and reverses, and you even see crime increase. That seems to me to be where we are now. There is also the question of diminished deterrence. In neighborhoods where half the young men go to prison, and going to prison is seen as a rite of passage, then any credible threat of deterrence is severely weakened. The last thing we want to do is make imprisonment so normal in some communities that it loses its deterrent value.

In neighborhoods where half the young men go to prison, and going to prison is seen as a rite of passage, then any credible threat of deterrence is severely weakened.Richard Rosenfeld, University of Missouri, St. Louis

Q: How effective is increasing the length of prison terms in deterring future crime?

Nagin: In the context of the American criminal justice system, there is very little evidence that increasing already lengthy prison sentences has much of a deterrent effect. And there is a growing body of evidence that it scars prisoners, their families, and their communities.

Raphael: The evidence suggests that what seems to deter people from committing crime is a high likelihood of getting caught or punished, and that the severity of punishment—say 10 years versus seven years—has very little impact on people’s thought process. The idea that you have a certain and immediate punishment for a transgression is far more effective than a threat to make someone’s life much more miserable in the distant future. And of course violence, when it is not for acquisitive purposes, doesn’t seem to be responsive to sentence severity. It’s often driven by emotion, it’s impulsive, so the thought of punishment is not associated with the act.

Cook: Increasing a prison term from 15 years to 20 years or from 20 to 30 years will have very little effect because people discount the future. We all do, not just criminals. Will those extra 10 years tacked on to the 20-year sentence, coupled with the 1 in 20 chance of getting caught, affect someone’s thinking about whether to commit a robbery? Well, maybe sometimes. But how much are we getting out of that given the great expense of prison? Also, any benefit you get from incapacitation is eliminated by the time you get that far out in the life of the prisoner because they age out [of criminal behavior]. So the two big effects prison is supposed to have on crime—deterrence and incapacitation—both favor certainty over severity.

Kleiman: If offenders were perfectly rational, crime control would be easy. You’d ratchet up the severity of punishment to the point where even a small probability of being caught would mean it’s not worth it. That’s what we’ve been trying for the last 30 years, and it basically hasn’t worked. Everybody is more sensitive to an immediate consequence than to future consequences, like an additional five years of prison sometime in the distant future.

Spelman: There are three essential elements of punishment: speed, certainty, and severity. Certainty matters a lot, and although there’s less evidence, speed matters, too. If we were able to identify, adjudicate, and punish bad guys really fast, like within a few weeks, there would be a lot more deterrent value to punishment. Severity doesn’t matter a bit. The average criminal has a very short time horizon. He is leading a hand-to-mouth existence and isn’t thinking in terms of a long career track or what life will be like five or 10 years downstream. He’s trying to find a place to sleep and money for his next crack hit, so these 10-year federal sentences for crack dealing? Totally worthless.

Travis: As we know, the increase in incarceration was caused by two types of policy choices, almost equally divided between choices to make long sentences longer and choices to put people in prison who would not otherwise have been sentenced to prison. The evidence shows that neither policy choice has had a big crime reduction effect. Keeping someone in prison for 20 years instead of 15 years is not going to have a big crime reduction payoff because that person is already in prison. That’s common sense. Meanwhile, the adverse consequences for people serving these long sentences and the ripple effect of that on children and society are profound.

Q: Why hasn’t crime increased as a result of the economic recession?

Cook: Do recessions cause crime? I looked at 13 business cycles going back to 1933 to see if crime rates turn up during recessions. Some crimes do. Burglary and robbery increased in every recession except one, but the effect was not huge, a few percentage points. Homicide does not seem to be affected by business cycles at all. So the effect of the business cycle on crime is surprisingly small. One likely reason is that recessions are complex events that include some changes that are likely to curtail crime, such as reduced drug and alcohol use and more time spent at home.

Rosenfeld: It would be astounding if a $13 trillion economy did not have an effect on street crime, but we need to dig deeper to understand the impact of particular economic indicators. For decades the focus was on unemployment. But I think a much better indicator is consumer confidence. When consumers are more optimistic, crime declines; when they’re pessimistic, it increases. One indicator distinguishing this most recent recession from prior ones is the historically low rate of inflation. For the first time in over 50 years, prices actually went negative in 2009, meaning they decreased in absolute terms and rates were lower than at any time since the Great Depression. Rising prices send consumers in search of lower cost goods. The cardinal characteristic of stolen goods is that they are cheap. So, high rates of inflation strengthen the incentives for robbers, burglars, and thieves to supply underground markets, and crime rates increase. The process operates in reverse as inflation ebbs.

Zimring: This recent period is a mystery. There are no experts. We had a big decline in crime nationally in the ’90s, then a sort of trendless period between 2000 and 2007. And then after 2007, there were two reasons to think that the bananas should have hit the fan—the meltdown of the economy and the turnaround in imprisonment, meaning it finally stopped growing. But instead we get a sort of a man-bites-dog period of really surprising and pretty substantial crime declines. And nobody is coming up with any plausible explanations for it.

In the context of the American criminal justice system, there is very little evidence that increasing already lengthy prison sentences has much of a deterrent effect.Daniel Nagin, Carnegie Mellon University

Q: Are there some other causes of the crime decline that seem plausible but require more evidence?

Cook: What’s been left out of the entire conversation is a discussion of private action on crime prevention. Americans spend an enormous amount on private crime prevention and avoidance. There are more people in private security now—around 1 million—than there are uniformed officers. Technology also has had a profound effect on the volume and nature of crime. So, for example, nobody carries cash anymore, and thus the theft of cash has been replaced by credit card fraud. And if a cellphone is stolen there’s an app that lets you track down the thief. And why did auto theft go down so much since the early 1990s? Well, perhaps it’s the new devices built into vehicles that make it damn hard to steal a car these days. The other hugely important private contribution is public cooperation with the police. We’ve seen this remarkable increase in reporting of crime, and overall people are more trusting of police than they were 25 years ago.

Nagin: There is good evidence that hot spots policing has been effective in the intermediate term in preventing crime. If a policeman is standing or sitting in a police car in front of a liquor store, nobody in their right mind is going to try to rob that store. It’s about using police so that potential offenders see them and respond to the certainty of apprehension. Then, you’ve stopped that crime from happening. In cities where this sort of policing pressure has been applied, it’s been effective. And while we don’t have any silver bullets, there is enough that’s promising to justify reorienting our crime control strategy away from an emphasis on ever-more-severe prison sentences to the smarter use of the police.

Kleiman: The lead issue deserves attention. We know the biological processes by which lead impairs self-command. We have good individual-level case-control data showing [childhood] lead exposure strongly correlated with criminal activity. We have both cross-national and local analyses showing that places with bigger changes in lead, up or down, have corresponding changes in crime, up or down, about 18 years later. The cohort data cast some doubt on how much of the crime decline can be explained by the lead decline, since reduced lead exposure should have affected the younger cohorts only, but crime went down in all cohorts. But my bet would be that lower lead levels made a substantial contribution to crime reduction.

Raphael: There’s a lot of mystery. You’ve got the abatement of lead paint. Some people credit immigration, saying that immigrants are less likely to offend than others. In addition, our population is aging and we have an overall higher level of educational attainment, all of which militate toward lower crime rates.