Tuesday, September 14, 2010

Crime and Punishment

"It isn't what we don't know that gives us trouble, it's what we know that ain't so."

- Will Rogers

Today, let's have a look at a glitch in the matrix. We'll consider a simple proposition which seems true, but is in complete disrepute among academics in the relevant field, and very much on the fringe of polite opinion. Then we'll demonstrate how contrary are official opinion and reality. Finally, we'll reflect on what implications the discrepancy has on the reputations of academia, our effective level of freedom of conscience, and the health of our nation.

The proposition is this: Increasing the length of sentences served by criminals necessarily results in fewer crimes being committed. To this, the layman says: Duh. Criminology professors disagree and will often claim that decreasing sentences is the true key to reducing crime. Many in the mainstream media are happy to echo this. Dan Gardner, who I harass today because he is often perceptive on other issues and really should know better, is a frequent champion of this belief. Sadly, most of his columns have been taken off-line, but here is a good example that I dredged up.

I'd like to offer a simple a priori proof that longer sentences reduce crime. There is a large and growing body of academic attempts to empirically answer this question, but in the absence of controlled experiments - as is so often the case in the social sciences - such studies are close to useless. Even if experiments did exist, one logical proof beats a thousand regression analyses - science, being a subset of reason, is trumped by it every time.

First, simply note that violent offenders have much, much higher average rates of subsequent violent crime than the people who have not been committed of a violent crime. In other words, violent criminal rates of recidivism are higher than the population average. This point is trivially obvious.

If we were to effectively imprison all violent offenders for life, they would have a zero percent recidivism rate. Since they would have a non-zero recidivism rate if we released them after, say, three years, it stands to reason that life imprisonment would reduce the number of violent crimes committed in subsequent years by the number that would have been committed by this set of previously-imprisoned violent offenders.

For it to be true that increasing their sentences would not reduce crime, it would be necessary to claim that new violent criminals arise to take the place of those that are imprisoned for life. This sort of behaviour is possible (and, IMO, likely) for certain crimes, such as the market-demanded crimes of drugs and flesh. But for muggings? For rapes? It is completely untenable to claim that extending sentences for such crimes would not result in a dramatic decline in their occurrence.

Also note that we have not mentioned the additional effects of deterrence. Criminologists generally scoff at the ability of stiff sentences to deter would-be criminals, dismissing such theories, as Gardner does, as a blind adherence to a "homo economicus" model of human behaviour. But the only escape from the logic presented above of recidivism-based crime reduction is to posit criminals that take up their dark trades as a result of a shortage, a response that strikes me as quite - wait for it - rational.

And really, if none of that convinces you, how about this? Crime in England, once the most civilized nation on Earth, has increased by several orders of magnitude over the past hundred years, despite increasing material wealth and improvements in law enforcement technology. Possibly these figures are biased by higher rates of reporting, new definitions of what is and is not a crime, and so on. The homicide rate however, surely the most consistent of crime-related statistics, has doubled since the 1950's.

So in conclusion: Putting criminals in jail results in fewer innocent victims of crime. A shocking result, unless you happen to live outside the American Empire between 1964 and 2010 (and counting), in which case it is painfully, blindingly obvious. I'm quite confident that the trustworthy history books of the 22nd century will devote a chapter or five to this odd phenomenon that beset the western democracies in the late 20th century, wherein they en masse reduced sentences, built and staffed endless appeal courts, and took away many of the powers police require to effectively do their jobs, and subsequently paid lots of smart men to sit around and wonder aloud why crime was such a problem.