I recently obtained a review copy of Freedomnomics, which was economist John R. Lott, Jr.’s reply to Levitt and Dubner’s Freakonomics. As I noted on the Amazon review, I’m conflicted as to whether the reply was necessary, given that Freakonomics itself is more of a puff book than a serious economics treatise. That said, enough people apparently do take Freakonomics’s conclusions seriously enough to make a convincing rebuttal necessary, and Professor Lott has done an excellent job of that. In particular, Lott takes on Levitt and Dubner’s conclusions that campaign spending is a waste of money, abortion reduces crime and realtors are out to scam you. Lott also discusses a number of other interesting topics not raised in Levitt and Dubner’s book, such as the importance of reputations, and the possibility that reputational harm should be considered when determining the sentences for various crimes.
Lott is on the strongest footing with regard to campaign finance, noting generally that for Levitt and Dubner to be right, virtually every politico in the world must be terribly wrong. Levitt and Dubner do have a lesser point, namely that no amount of spending can make up for a candidate’s lack of political acumen or convert an unpopular candidate into a popular one, but by claiming it cannot impact any election by more than a point or two – thereby making all campaign spending in all but the closest races a colossal waste of money – they clearly get carried away, and Lott rightly calls them on that. Similarly, Levitt and Dubner point to the data indicating that realtors make more money selling their own houses by leaving them on the market slightly longer than their clients’ houses stay on the market on average, from which they conclude that realtors just don’t try very hard to get the best deal they can for their customers. That too is rebutted by several of Lott’s arguments, including the equally plausible (or, by Occam’s Razor, more plausible) theory that non-realtors don’t always heed their realtors’ advice, while the realtors themselves obviously do (when was the last time you ignored your own advice to yourself?). And his argument on reputational costs of white collar crime is interesting not so much as a rebuttal to any particular work, but simply as an issue to think about in its own right.
I’m more skeptical of Lott’s position on abortion vs. crime. On the one hand, common sense would appear to support Levitt and Dubner’s basic theory that unwanted kids are more likely to be raised badly, and therefore, more likely to turn to crime, than wanted ones. Lott does not appear to dispute the general connection between unwanted children and crime, but reaches the opposite conclusion with regard to abortion on the more tenuous theory that abortion causes more unwanted children by lowering the incentives for unprotected sex. This makes little sense, as legalized abortion only creates an incentive (or, more accurately, removes a disincentive) for unprotected sex among those willing to obtain abortions. For those who oppose abortions even at a personal level, the legality of abortion should not change the incentive structure at all. The only way it could, in my view, would be if it could be shown that a large number of couples have unprotected sex because they think they can always get an abortion later if need be, only to have a change of heart and suddenly become personally opposed to abortion as soon as the unwanted pregnancy develops. I don’t doubt that this ever happens, but the likelihood of it happening often enough to offset the number of abortions every years appears remote. To the extent that legalized abortion and high numbers of unwanted children correlate (and I have little doubt that they do), it would seem more likely that the sexual revolution caused both the increase in sex outside a stable relationship and the political impetus to legalize or further deregulate abortion.
Like Freakonomics, Freedomnomics is written to make basic points come across clearly to the unwashed masses, not to beat the reader to death with regression analysis. This makes the book much more accessible to the average reader, and more entertaining as well. It also makes the conclusions more suspect, however, as it effectively asks the reader to trust the author’s conclusions rather than demonstrating beyond any possible doubt (as he did in, say, More Guns, Less Crime) that they are correct. For this reason, I’d urge people to read Freedomnomics the way one should read Freakonomics, namely, to make you think about the issues raised, and not necessarily to tell you what to think about them. So read Freedomnomics. I’m confident you’ll enjoy the book, but even if for some reason you don’t, at least you will make Tim Lambert cry.