Crime Doesn't Pay as Well as It Did

Commentary by Pete du Pont

If you have the impression that fewer people these days are being murdered, raped, robbed and assaulted, and that not as many homes are being burglarized, you're right. There has been a documented decrease since 1993 in every category of violent crime plus burglary.

The murder rate across the country has dropped by 23 percent, rape by 12 percent, robbery by 21 percent, aggravated assault by 11 percent and burglary by 15 percent.

Not by coincidence, it would appear, another significant occurrence in recent years is that more felons are going to prison, and are staying there longer. For a long time now, researchers have been pointing out that a small number of repeat criminals commit a good proportion of all serious crimes. It seems to stand to reason that the more of that small number we keep behind bars, the less crime we are going to have. And that is apparently what has been happening.

The probability of going to prison has increased for every one of those crimes since 1993, and so has the median prison sentence.

If all that information doesn't leave you feeling completely safe, you're right again. Despite all this improvement, you're still three times more likely to be a crime victim than in the 1950s. But the additional good news is that the trend toward sending more criminals to prison with longer sentences continues.

About 50 years ago, a popular monthly comic book was entitled "Crime Does Not Pay." Economists tend to measure crime, like many other human activities, in economic terms, and they maintain that the reason we had such an increase in crime in the 1970s and 1980s was that crime did pay. Further, they maintain that the rate is falling because the cost has been rising.

The National Center for Policy Analysis has developed a measure of the cost of crime to a potential criminal called "expected punishment." That's the amount of prison time a criminal can expect to spend, given the probabilities of being apprehended, prosecuted, convicted and going to prison, and the median sentence for a crime. The index of expected punishment shows just about what you would expect.

Expected punishment for each violent crime and burglary plunged from about 50 days in 1950 to only 10 days in 1970, and didn't rise appreciably until about the mid-1980s. Even today, expected punishment is only a little over 22 days - less than half of what it was in 1950.

Criminals may not measure expected punishment in the same sophisticated way, but they have a rough idea of the likelihood of going to prison and how long people are staying in prison. Sure enough, there were 5 violent crimes and burglaries per 1,000 population in 1950 - and by 1970, with expected punishment having plummeted, there were three times as many. When expected punishment started increasing in the mid-1980s, the crime rate began going in the other direction.

Why did expected punishment fall so drastically? There were the usual "criminals are really victims of society and we need to get rid of the root causes of crime" groups, but again economics played the primary role. Courts ordered an end to prison overcrowding and states for the most part declined to spend the money to build more prisons. There were actually fewer commitments to prison in 1970 than in 1960, even though the number of violent crimes and burglaries nearly tripled.

States finally started building more prisons in the 1990s, and despite the cost of building prison cells and keeping prisoners in them, society is benefiting. Last year economist Steven Levitt found that incarcerating one additional prisoner reduces the number of crimes by about 15 per year, the majority of them property crimes, and yields a social benefit of at least $53,900 per year.

Thus spending $25,000 a year to keep a criminal in prison is money well spent. Residents of places like New York City, where the murder rate is now as low as it was in the 1960s would no doubt agree.