Mon Sep 16, 2002

Incarceration and Inequality, Again

Iain Murray responds to my comments about incarceration rates and the African-American population. At the end of his post, Iain says “The JPI study was, in my opinion, simplistic, flawed and inexact. In its choice of headlines it also showed poor judgment. It deserved the treatment it got.” The casual reader might think I had ridden in on my horse to defend the JPI. Not at all. (See my earlier post for fuller context.)

The same disclaimer applies as before: I am not an expert on criminal justice policy. Iain writes:

I’m not sure whether Kieran actually read my article, rather than just Virginia’s comments, but the article had three points. First, the JPI’s numbers were objectively wrong. Second, the analogy they used was flawed, and that there was a better analogy available which implied something different. And third, in order to make a case that education spending suffering in comparison to corrections spending disproportionately affected African-American men, they ignored the data about the substantial increase in African-American women in college, something that seems to indicate that educational opportunities were made available to that community, and taken up.

We don’t disagree at all on the first two points. As I said in my original entry, “Debunking false claims is the right thing to do, of course, and the JPI people should know better than to try to make political hay out of flawed analogies.”

On the third point, the comparison with education is still a bad analogy. Iain changes the groups being compared, but not the basic, flawed comparison (size of jail population vs college population). That was my original point: both he and the JPP end up arguing over a comparison that doesn’t tell us much about trends in incarceration. The JPP are wrong; Murray wins, but everyone is led off in an unproductive direction.

Finally, Iain makes the following argument:

In a just society, and I happen to think the American experiment is one of the closest to “justness” that there has ever been, the incarceration rate reflects the level of criminality in communities. Spending on corrections is therefore reactive. The plain fact is—and the New York Times got this right—the increase in incarceration in the 90s was driven by an increase in violent crime. Bill Spelman of the University of Texas at Austin built a model to assess the effect of prison expansion since the 70s on the violent crime rate. His conclusion was that by the late 90s, if we had not expanded our prisons, America would be suffering about 1200 violent crimes per 100,000 population. The actual rate was less than 600. The expansion of prisons was necessary, it seems to me.

Well, yes—- in a just society incarceration would reflect criminality. But there are at least three difficulties. First, we need to decide what’s illegal: the things you can be locked up for have a tendency to vary over time and across societies. Second, while you may “happen to think” that U.S. society is “one of the closest to ‘justness’ that there has ever been”, I think there’s some room for disagreement on that point—- at least enough to make one think twice about using it as a premise in an argument like this. Third, I question whether it is a “plain fact” that the prison boom was “driven by an increase in violent crime.” It’s not obvious to me from crime data that this is so.

For example, here’s a time series of the homicide rate from 1950 to 1999). As you can see, it really takes off in the 1970s. It declines precipitously in the mid-1990s. In between, there is a rise in the late ‘80s and early ‘90s, but that peaks below the all-time high in 1981. The trend suggests that the relationship between crime and incarceration is more complex than Iain Murray wants to allow. (If incarceration rates reflect violent crime rates, why did the prison boom not begin in, say, 1974?) Rather than assume that the incarceration rate is a simple reflection of levels of crime, or that it reflects the basic “justness” of U.S. society, I’d suggest it’s more profitable to examine the role that other forces, most notably state policy, play in this process.