November 2, 2006

Levitt is skeptical of the pornography-cuts-rape theory, too

On his Freakonomics blog, Steven D. Levitt writes:

Everybody and their brother is sending me links to Steven Landsburg’s most recent Slate column that reports on studies by economists that suggest internet porn reduces rape and the release of blockbuster violent movies reduces violence.

While the idea might strike non-economists as crazy, the theory makes sense. When you lower the price of a good that is a substitute for a second good, the quantity of the second good should fall. It is not obvious that internet porn and rape are substitutes (they may very well be the opposite, what economists call complements), but it is not impossible.

I have to confess that Halloween with four young kids (a fairy, a witch, batman, and Dorothy from the Wizard of Oz) has been a full time job for the last week. So I haven’t read the academic articles. The paper or porn and rape was actually written by my former student Todd Kendall, now at Clemson. It is not really fair (especially to your former student) to be skeptical without reading the paper, but I have to say I am skeptical of the empirical result. The kind of variation in the data that gives the result is that states that are quicker to adopt the internet saw bigger declines in rape. He then does a nice thing in the paper, going beyond just this one prediction to test other hypotheses, like do crimes other than rape fall with the internet (he says no) and does other sexual behavior change with the internet (he says yes). The concern is always, with this kind of approach, that there are other factors that might be driving both the adoption of the internet and the decline in rape. The challenge to those who want to refute Todd Kendall’s argument is to identify those variables. The challenge for Todd is to find other kinds of “natural experiments” that support his hypothesis.

The most obvious other natural experiment was the huge increase in pornography in late 1960s and early 1970s that coincided with a big increase in the rape rate.

It's just like how the abortion-cuts-crime theory was popular a priori in the 1970s and 1980s, but then faded out of discussion in the early 1990s when the first generation born after legalization of abortion went on a murder spree. Then, after the crime rate dropped in the mid and late 1990s, Levitt and Donohue revived the theory in 1999, and by then, most of the chattering classes couldn't remember why it had dropped out of favor earlier, so they found it immediately convincing. Similarly, nobody remembers today that four decades ago pornography and rape increased together, so this popular old a priori theory is ready for relaunch by a naive young professor into an intellectual world that doesn't remember much of anything.

The chattering class especially tries to forget everything it learned about crime, since that is so inextricably linked with race. Everybody is supposed to forget everything they know about crime and race from thinking about their own real estate decisions when it comes time to discuss public affairs. And remembering the race-rape connection is one of the worst thought crimes possible.

P.S. I just got a very nice email from Dr. Kendall saying that he hoped his study shed some new light on the topic, but he definitely didn't think his study should be the last word on the subject.

My published articles are archived at -- Steve Sailer

No comments: