Monday, July 20, 2009

Crime is dropping in cities across the United States; experts baffled

The Washington Post reports:

Violent crime has plummeted in the Washington area and in major cities across the country, a trend criminologists describe as baffling and unexpected.

[Washington, DC], New York and Los Angeles are on track for fewer killings this year than in any other year in at least four decades. Boston, San Francisco, Minneapolis and other cities are also seeing notable reductions in homicides.

"Experts did not see this coming at all," said Andrew Karmen, a criminologist and professor of sociology at the John Jay College of Criminal Justice in New York.
Maybe the criminologists are systematically ruling out certain possible explanations.

Maybe if the facts are so baffling to them, they should consider changing their theories.

I keep reading about how the United States has too many people incarcerated. Isn't it possible that our policy hasn't been totally irrational, but is actually working?

Another post for my "experts" tag.


LemmusLemmus said...

Sorry, but you're being disingenious here:

1. Incarceration rates move slowly. Hence they can hardly explain something like a 17% drop from one year to the next.

2. Do you really think students of crime are so stupid as not to consider that incarceration might have crime-reducing effects? Well, they're not. Example (pdf).

John Althouse Cohen said...

I mentioned the incarceration rate as one possible factor among many. I don't think that's a disingenuous suggestion. I could have also mentioned other possible explanations for the trend, but the incarceration rate sprang to mind because it's gotten a lot of media attention lately.

I was also trying to illustrate a broader problem: criminologists who purport to be experts yet can't even approximately predict future trends based on current circumstances/policies. And that's part of a broader theme on this blog of supposed experts who conceal the fact that what they're doing is more art than science. If most criminologists are in a liberal academic milieu where it's in vogue to lament America's high rate of incarceration, and if they can bet their colleagues would look at them askance if they spoke favorably of it, that's a seriously problematic bias.

As for Steven Levitt (whose paper you linked to), I'd call him an economist, not a criminologist. (I'm planning to do a blog post relating to this distinction in the future -- specifically the disparity between economists' and criminologists' views on the death penalty as a deterrent.) Also, that Levitt paper supports the view that incarceration rates are partly responsible for America's declining crime rate. ("Four factors appear to explain the drop in crime [in the 1990s]: increased incarceration, more police, the decline of crack and legalized abortion.") Also, I'm not sure if the Levitt paper is a good example of garden-variety academic opinion -- that last factor was wildly incendiary!

LemmusLemmus said...

"I mentioned the incarceration rate as one possible factor among many." That wasn't clear from the post (although in retrospect I have to say it's a possible reading).

"criminologists who purport to be experts yet can't even approximately predict future trends based on current circumstances/policies." The limited predictive power of the state of criminological knowledge is indeed a problem - it would be desirable for it to be better. The 1990's drop that Levitt writes about in the paper I linked to is the most recent example. (Let's wait and see whether 2009 will actually produce such a large decrease across America.) The problem here might be that the Washington Post is, by and large, right: that changes in local police tactics cause noteworthy changes in crime rates. This is something that's not easily looked into using the kind of data students of crime usually work with.

Idiological bias, to the extent that it exists, is always a problem.

"Criminologist" is a tricky label because it is sometimes used for people who work in criminology departments (excluding the economist Levitt) and sometimes for academics who study crime (including Levitt). I hence consciously avoided the label in my comment.

Yes, the paper supports the view that incarceration reduces crime. It hence shows that not only have academics thought of that factor, some actually agree that "our policy hasn't been totally irrational, but is actually working". That's why I linked to it.

Jason (the commenter) said...

Maybe it's all the guns people are buying because of Obama?