The blog of the Urban Institute
July 27, 2011

Deciphering the Crime Decline

July 27, 2011

In June, the Federal Bureau of Investigation announced that violent crime had declined by 5.5 percent in 2010, repeating a similar decline in the previous year.  Property crime continued to decline as well. Overall, crime in America is at levels not seen since the late 1960s.

Criminologists, sociologists, and economists have been intensely debating the cause of the crime decline. While I have identified 19 published explanations, none is wholly convincing on its own.

The most compelling explanation is that, since 1980, the prison population has quadrupled at the same time crime has declined about 30 percent. During this period, state spending on corrections has tripled. While crime has declined as a result, mass incarceration does not appear to be a particularly cost-effective means of increasing public safety.

The end of the crack epidemic around 1990 is also often cited for the crime decline.  Violence around drug trafficking, crimes committed to get money to buy drugs, and crimes by intoxicated persons all began dropping off at the end of this epidemic. However, data suggest that total drug use in America was lower in 1990 than at any point in the last four decades, and drug use is much higher now than it was in 1990 at the height of the crack epidemic.  Putting these pieces together, we would expect to have seen a more recent bump in the crime rate if the number of drug users was the core explanation.

A bad economy is often proposed as the fuel for changes in crime rates. However, the beginning of the crime decline in 1991 coincided with a deep recession, just as the current deep recession has coincided with an almost unprecedented drop in crime.

Many criminologists have pointed to changes in “routine activities” as responsible for the crime decline. Graham Farrell and colleagues in the United Kingdom have proposed that target hardening—making cars harder to steal, businesses more protected, and personal items easier to recover—explains the decline. Other changes from improvements in technology, such as the Internet and personal entertainment systems, are believed to keep people inside their homes where they are safer. This explains the lower incidence of property crime, and the even larger slide in violent crime is a spillover.

Changes in policing, from the old days of cops sitting in cars reacting to calls for service to the current practice of police walking the beat in neighborhoods talking to risky people in risky places to prevent crime, are also cited.  However, the number of officers per capita has not changed in 30 years, and the rate at which cases are closed has declined slightly even as there are fewer crimes to investigate.

Most observers believe gangs are less prevalent than 20 years ago, and progress has been made controlling organized crime. Yet, researchers have tremendous difficulty testing either of these elements because data are too limited.

Some economists have tried to connect changes in America in the 1970s to changes in criminal offending two decades later. The legalization of abortion, they maintain, led to fewer high- risk teens in the 1990s. Removing lead from gasoline and paint in the ‘70s may have had a similar effect.  Closing  mental institutions in the seventies and eighties may have initially contributed to the crime wave, while improvements in medications in the nineties and beyond may have contributed to its decline.  None of these explanations holds up to close scrutiny.

I would suggest that we are looking at this problem through the wrong lens. Instead of asking why crime has declined, we should be figuring out why crime jumped in the seventies and eighties. Long-term trends suggest that the seventies and eighties were in a fact a crime bubble.

Bubbles burst.  Be it Internet stocks, houses, or gold, bubbles last only as long as new buyers can be enticed into the marketplace. Drug use is socially tolerable as long as it is widespread.  As the number of new drug users declined at the end of the eighties, use became less tolerable and that bubble burst. The same reasoning holds with criminal networks. Changes in routine activities and massive expansions of the prisoner population in the 1970s and ‘80s likely hastened the bursting of the crime bubble.

So, will crime go back up? While crime is very volatile, America seems to be back along its long-term trends, and there's every reason to be optimistic.  That suggests that we may have an opportunity (which I’ll discuss in future posts) to cash in on this different kind of peace dividend, something that would be welcome in the current budget environment.


As an organization, the Urban Institute does not take positions on issues. Experts are independent and empowered to share their evidence-based views and recommendations shaped by research.


Have you considered the aging of the U.S. population as a factor in lower crime rates? Crime is largely a young man's game.
Nice to be optimistic, but there's speculation that the recession will give rise to an expanded group of unemployed urban males, leading to more urban crime whose effects will be felt for a long time to come.
A quick response to the comments. The idea that re-investment in cities led to the crime decline is provocative - certainly that case can be made in cities like New York and DC. However, other places (Philadelphia, Detroit) have not seen much in the way of investment and crime is down there too. But worth thinking about more. As I note, the lead hypothesis has some supporters, although I am skeptical for some technical reasons. To me, the most intriguing part of that paper is the idea that crime declines have been very similar internationally, suggesting that events confined mainly to the US (like mass incarceration) are perhaps not responsible for the crime decline. This is an argument Farrell and colleagues make in defending improved security as an explanation for the crime decline. The argument that the recession should cause crime increases has been made by many notable economists. However, as we approach the third year of this downturn, crime continues to decline at a rapid rate. Perhaps people who lost their jobs in this recession are simply not the type of people who are likely to replace legal income with illegal income because of their age, location and/or social status (married people are much less likely to offend).
Having conducted a 6-year pre-post intervention study with comparison neighborhoods for 4 local neighborhoods and address-specific data, there are clearly a fairly large number of variables that affect the occurrance and extent of criminal behavior. With well defined, consistent, cleaned data over time, econometric analyses can reveal the significant variables that make up a high correlation algorithum for crime. just as all economics is local, so too is crime - being dependent on the particular mix of variables that are significant in each local environment. Don't forget that the combined effects of some variables that are not significant by themselves, may be just what is needed when taken together (don't rule incarceraton rates out just by examining their effect alone). First, when attention is given (the testing-effect) to an issue,people change their behavior. Second, common sense must always be in the mix,if not prevail (pass the reasonable-person test). There is no one magic bullet for the complex and ever fluid topic of crime prevention. I would expect that the great recession will give rise at some point to greater crimes of necessity, if the safety net is frayed more than at present. Interestinig about the 20-year lag mentioned. If you think the impact of drugs and gangs has waned, you are out of touch - talk to some more street Cops on that one.
Another aspect potentially leading to the decline in urban crime is the re-investment in cities. The 1960s and 1970s saw mass dis-investment ... from municipal governments to individual home-buyers. The "white flight" era left our cities in a crunch ... few wanted to be there and even fewer wanted to spend their money on real estate. This can also be said to contribute to the "crime bubble" of the 1970s and 1980s you mention. With re-investment beginning in in the 1980s and growing dramatically in the 1990s, people all of the sudden began re-discovering their urban centers, investing not only with their money, but also with their livelihoods. This brought more interested in improving neighborhoods and getting rid of the crime element. If this is not a part of your continuing study, you may want to consider it as a notable factor.