Abortion and crime
I have not (yet) read Freakonomics, but I like the concept of using economics (and science in general) to address practical issues.
But it appears that Steven Levitt may have screwed up when he concluded that legalizing abortion substantially reduced the crime rate. Apparently, he and co-author John Donahue committed an “inadvertent but serious computer programming error” which meant they never actually did one of the controls they claimed.
Other researchers have had trouble substantiating a link between the legalization of abortion in 1973 and the subsequent drop in crime as children born subseqently (or not) came of age.
The article is short (and like most things in the Economist) well-written and interesting; it has the clever title "Oops-onomics" to boot. Unlike most Economist articles, however, it is available online.
I should note I don't find this particular issue particularly compelling. My views on abortion, that it is not the government's business, would be the same even if legalizing abortion increased crime.