Charles Murray is not only a friend; he is also among the most important intellectual figures in the modern conservative movement. So I’m hesitant to dissent from his views, particularly when it comes to explaining social trends. But at a recent lunch, we had a good-natured disagreement about incarceration and crime, which he elaborated on in a blog post.
The core of our disagreement is the role incarceration has played in the enormous drop in crime we’ve seen during the last 15 years or so. Charles argues that “simple incapacitation … plus a substantial deterrent effect is a plausible explanation for why violent crime dropped at all.” He harbors the belief that “without the massive increases in incarceration after the mid-1970s, crime rates wouldn’t have turned around at all. Higher imprisonment was the necessary condition for 100 percent of the reduction in violent crime.”
My view is that incarceration has played a crucial role in the drop in crime, for reasons Murray cites. But so have other factors, including target hardening and private security measures, the end of the crack cocaine epidemic, demographic changes, an increase in police per capita, and improvements in policing techniques (including “hot-spot” policing). As William Bratton, the very successful former police chief in Los Angeles and New York, has said, “We’ve gotten better at spotting crime trends more quickly. We can respond much more quickly.” Others argue the decrease in crime is attributable, at least in part, to a cultural re-norming, gun laws, the economy, advancements in medicine and public health, and more. Whatever the precise cause-and-effect, it’s implausible to believe these non-incarceration factors have had no impact whatsoever on crime rates.
Murray offers up this counterfact to consider: “Suppose we had maintained imprisonment for violent crime at the rate that applied in 1974. In that case, we would have had 276,769 state and federal prisoners in 2010 instead of the 1,518,104 we actually had. Suppose tomorrow we freed 1.2 million inmates from state and federal prisons. Do we really think violent crime would continue to drop at a somewhat slower pace?”
Set aside the fact that this calculation embodies some questionable assumptions (over this time period, hundreds of thousands of prisoners entered and exited prisons each year, average sentence lengths for most violent crimes increased substantially, and so did the fraction of all prisoners admitted as a result of parole violations and the fraction of all prisoners whose latest conviction was a non-violent drug crime). What’s important to keep in mind is that while America will not release 1.2 million prisoners between now and the end of this week, we have, in fact, released around 7.5 million state and federal prisoners between 2000-2010 and are currently releasing more than 700,000 prisoners each year, or roughly 1,900 each day (see Table 2 in this Department of Justice report). And yet, crime rates have continued to spiral downward.
For what it’s worth — and to me it’s worth quite a lot — the finest living thinker on crime and society, James Q. Wilson, believes greater incarceration can explain about one-quarter to 30 percent of the decrease in crime. John J. DiIulio, who studied under Wilson and is himself a respected analyst of crime trends, more or less agrees. (A recent DiIulio essay which deals with the causes of drops in crime can be found here.) This puts Messrs. Wilson and Dilulio on the high side among criminologists when it comes to the impact of incarceration on crime. (Even criminologists who accept the highest published estimates of the median number of felony crimes averted each year by incarcerating violent offenders — more than a dozen per year per prisoner — find no basis for attributing even half the decrease in either violent crime rates or in overall crime rates strictly to increased incarceration rates.)
One final observation: I tend to be skeptical of mono-causal explanations when it comes to social phenomena and behavioral trends. These tend to be the result of a complex, and sometimes mysterious, interplay of events. For examples, since the early-to-mid-1990s, out-of-wedlock births have increased from less than 30 percent of all births to more than 40 percent of all births today. Yet during that period almost every other social indicator — including crime, drug use, welfare, education test scores, teen suicides, divorce, and abortion –improved. In some areas, like crime and welfare, the progress has the dimensions of a sea change. This is a remarkable, unexpected and encouraging development. It also reaffirms the conservative belief that modesty and caveats are in order when it comes to our ability to understand, let alone predict, social trends and human behavior.