April 13, 2005

The Abortion-Crime Graphs Steven D. Levitt Won't Show You in Freakonomics:

I see that Levitt's book Freakonomics, which argues that legalizing abortion cut crime, is now the #2 bestseller on Amazon.com. I also note that Amazon first posted, then yanked, my reader review poking holes in his theory, presumably to avoid casting a pall upon the sellathon.

So, here are two graphs from my article in the May 9th issue of The American Conservative.

First, Levitt's theory is predicated -- at least publicly -- on abortion reducing the proportion of "unwanted" babies, who are presumed to be more likely to grow up to be criminals. The empirical problem with this is that legalization (which occurred in California, New York, and three other states in 1970 and nationally in 1973), didn't put the slightest dent in the illegitimacy rate, which is, by far, the most obvious objective sign of not being wanted by the mother and father, and has been linked repeatedly with crime:

You'll note that the growth in the illegitimacy rate didn't start to slow down until the mid-1990s when the abortion rate finally went down a considerable amount.

My article offers a simple explanation, drawn from Levitt's own research, of why legal abortion tends to increase illegitimacy.

Second, the acid test of Levitt's theory is that it predicts that the first cohort to survive being culled by legal abortion should have been particularly law-abiding. Instead, they went on the worst teen murder rampage in American history. Here's a graph showing the homicide rate for 14-17 year olds, and below each year is the average birthdate of the 14-17 year old cohort.

For example, the 14-17 year olds in the not particularly murderous year of 1976 were, on average, born about 1960 (i.e., 1976 - 16 years of age = 1960), so they didn't "benefit" from being culled by legalized abortion the way that the 14-17 years olds during the peak murder years of 1993 and 1994 should have benefited, according to Levitt.

In contrast, the homicide rate for the 25 and over cohort (none of whom enjoyed the benefits of legalized abortion) was lower in 1993 than in 1983.

My published articles are archived at iSteve.com -- Steve Sailer

1 comment:

Anonymous said...

Plot them on the same graph. Abortion rates and crime rates. If it exists, the effect should be obvious.