When Crime Drops, Eventually the Prison Population Does Too
It's common—but entirely unsupported by the evidence—to argue that bad economic times lead to higher crime rates and higher prison populations. In the Washington Post this weekend, Mike Konczal airs the contrarian position that maybe our most recent recession led to a drop in the prison population. Keith Humphreys argues that this is entirely unsupported by the evidence too:
The prison population started rising during the mid-1970s oil shock and kept right on rising during the recessions of 1980, 1981-1982, 1990-1991 and 2001. If we want to explain a historic reversal of a multi-decade trend, we cannot logically do it by pointing to a factor that occurred repeatedly — a lousy economy — while that trend was underway (and p.s. the rate of incarceration also rose during the Great Depression).
Look instead for more novel factors to explain why the incarceration rate is finally falling, such as the lowest crime rates we have in generations, lower fear of crime than in generations, the emergence of effective alternatives to incarceration, and/or, if Kevin Drum is right, the dramatic reduction in lead in the environment.
I appreciate the shout out on lead, but I want to register a small semantic complaint: in what way would falling crime rates be a novel explanation for falling incarceration rates? Seems like ham and eggs to me. Based solely on the dramatic drop in crime rates over the past two decades, I'm willing to bet that prison populations will continue to drop for a good long time.
Konczal marshals some fairly unconventional arguments for and against the idea that recessions are related to incarceration rates, but never mentions the massive U.S. decline in crime rates since 1991. But you really can't do that. There have indeed been changes in the way we think about incarceration over the past few years—on both left and right—but those changes have themselves been driven by lower crime rates that everyone now agrees are permanent. That's where it all starts.