Wilson's criticisms of Pew study miss mark, ignore contrary research

A reader points me to an article in the LA Times by James Q. Wilson ("Do the time, lower the crime," March 30) responding to the recent study by the Pew Center on the States regarding America's high incarceration rate, which now exceeds one out of every 100 US adults. Wilson is one of the chief ideologists promoting a "get tough" approach nationally, so it's worth taking a moment to identify a few of the overstatements and obfuscations that to me typify his public pronouncements on crime.

For example, on the all-important subject of whether high imprisonment rates reduce crime, Wilson writes:

nowhere in the [Pew] report is there any discussion of the effect of prison on crime, and the argument about costs seems based on the false assumption that we are locking people up at high rates for the wrong reasons.

In the last 10 years, the effect of prison on crime rates has been studied by many scholars. The Pew report doesn't mention any of them. Among them is Steven Levitt, coauthor of "Freakonomics." He and others have shown that states that sent a higher fraction of convicts to prison had lower rates of crime, even after controlling for all of the other ways (poverty, urbanization and the proportion of young men in the population) that the states differed. A high risk of punishment reduces crime. Deterrence works.

For starters, his comment about incarceration vs. safety results in states cannot survive a comparison between Texas and New York, for example, so I'd like to see the research backing up that statement. By relying on Mr. Levitt's often controversial work, he's identified a scholar whose estimates of the effectiveness of imprisonment fall on the high end of those produced in the last decade. Levitt thinks that imprisonment accounted for as much as 32% of the reduction in crime in the 1990s (See "Understanding why crime fell in the 1990s").

Other econometric estimates - including one by UT-Austin's Bill Spelman - found that expanding the prison population accounted for about a quarter of the crime reduction in the '90s. (Bill and I have enjoyed a friendly dispute about this in the past, because I think some of his assumptions overstate incarceration's effectiveness and understate its harms). Overall, according to a recent paper by the Vera Institute, Levitt and Spelman "produced a fairly consistent finding, associating a 10 percent higher incarceration rate with a 2 to 4 percent lower crime rate."

But if we are to be honest about the state of empirical research on the topic, one cannot declare emphatically, as Wilson does, that "deterrence works" or that expanded incarceration "reduces crime." According to the Vera Institute, "One could use available research to argue that a 10 percent increase in incarceration is associated with no difference in crime rates, a 22 percent lower index crime rate, or a decrease only in the rate of property crime."

What's more, even the highest estimates, like Mr. Levitt's, still contend that 2/3 of the crime 0reduction had nothing to do with incarceration. So the decline in crime, according to these sources, mostly wasn't because of putting more people in prison.

Wilson says as much when he writes that, "Several scholars have separately estimated that the increase in the size of our prison population has driven down crime rates by 25%." But crime has declined much more than that since the early '90s, and Texas' prison population tripled since then, for example, so if it takes a 300% increase in prison capacity to get a 25% reduction in crime, how far can we really take that strategy?

Wilson similarly ignores research that suggests real, immediate limits to the benefits of incarceration in states that have large prison systems, again from the Vera Institute (p. 7):

Raymond Liedka, Anne Piehl, and Bert Useem have confirmed, moreover, that increases in prison populations in states with already large prison populations have less impact on crime than increases in states with smaller prison populations. States experience “accelerating declining marginal returns, that is, a percent reduction in crime that gets ever smaller with ever larger prison populations,” they argue. Thus, increases in incarceration rates are associated with lower crime rates at low levels of imprisonment, but the size of that association shrinks as incarceration rates get bigger.

So for states like California and Texas, the prospect of reducing crime by 2-4% by adding 10% to the prison population (in both states around 160,000 prisoners, give or take), represents massive additional expenditures that a) probably aren't worth the bang for the buck and b) are practically impossible in the real world because of limits on the state's ability to staff existing facilities.

Given that context, I audibly chortled upon reading Wilson's complaint that the Pew study compared increased state spending for prisons with that for universities, chiding Pew for failing to consider "whether society gets as much from universities as it does from prisons." Given the low marginal returns on prison spending and the much higher return rates on investment in higher education, we can have that debate, sir, but I doubt it will turn out quite the way you seem to think.

Professor Wilson offers a throwaway line that "except for some minor drug offenders" we don't imprison people needlessly. But in Texas (where Spelman's study was done), that's a LOT of folks. E.g., drug offenders are far and away the largest category of Texas probationers revoked to prison, and make up nearly a third of the prison population. Some "minor drug offenders" get awesomely long sentences - does Wilson think that improves safety, or not? He implies "no," but he certainly downplays the issue considering drug addicts are taking up so many prison beds.

I was happy to see Wilson advocating stronger probation programs, but they seemed almost an afterthought to his larger, erroneous thesis that mass incarceration is responsible for reduced crime. Nobody actually knows for sure what caused the crime drop in the '90s, though no one - even the research Wilson cites - claims that expanded incarceration caused most of the reduction. There's no doubt that some folks need to be incarcerated, but there's equally little doubt we incarcerate many more than is necessary to ensure public safety.

h/t to Bill Bush for forwarding the link.