Andrew Spicer's Weblog - Current - Archive Index - Email

Freakonomics
16/05/2005

Three weeks ago Sunday I bought the book Freakonomics; A rogue economist explores the hidden side of everything by University of Chicago economist Steven D. Levitt and New York author Stephen J. Dubner. It sounded interesting, was advertised in the New Yorker and endorsed by Malcolm Gladwell, so I thought I would give it a try. (Reading the book was a much faster than writing this review.)

On the book's dust jacket, Levitt is described as "a much heralded scholar who studies the stuff and riddles of everyday life -- from cheating and crime to sports and child rearing -- and whose conclusions regularly turn the conventional wisdom on its head." Dubner, on the other hand, sounds like he is essentially the smart ghost writer who wrote the book Levitt was too busy to write.

Overall, I thought it was entertaining book, though not quite up to my expectations. Critic Felix Salmon is essentially right when he complains that the book has no unifying theme, and that the short book has an unfortunate amount of filler (in the form of lists, tables, and data examples). What was worse for me was that a few of the discoveries left me unconvinced or even in disagreement.

In Chapter 2, for example, real estate agents are essentially accused of failing to serve their clients' needs as a result of the incentive structure that prevails. Agents are paid a commission based on the selling price of a house. If an agent holds out for an extra $10,000 on the sale of a $300,000 house, that's a big deal for the seller, but a mere $150 for the agent. (Commissions may be different here in Canada; I don't know.) As a result, the Steves suggests this incentive structure drives real estate agents to go for churn instead of the highest possible price.

As evidence, they point to a study that shows real estate agents keep their own homes on the market an average of 10 days longer and sell for 3% more.

While this may point to greater effort on the part of real estate agents when serving themselves -- and it may not, as there could be other explanations -- Levitt and Dubner must be wrong if they mean to suggest that the market is misperforming here as a result of perverse incentives driving real estate agents to sell out their clients in favour of volume. What... are we to believe that all houses could sell for 3% more if only agents were more dedicated to homeowners? Besides the questionable desirability of such an outcome, what about the equilibrium where supply meets demand?

I'm not the economist, but these guys seem to be suggesting that houses across the nation cost 3% less than standard economic theory would suggest. If that is the case, we should thank real estate agents for their miraculous work.

The Steves then go on to accuse agents of using coded real estate descriptions to signal their colleagues that they are looking to sell cheap on particular properties. They come to this conclusion in part by observing which words in home listings are correlated to high selling price vs. those words correlated to low selling price. Supposedly agents use low-end words as a hint to other agents that they should go ahead and bid low, a quick sale is desired. However, nowhere do they show that most houses described as "fantastic" (which is correlated to low selling price) could alternatively have been described as containing "state-of-the-art" this or "granite" that (both of which correlate to high selling price). More likely, agents are doing their best to describe the house they are trying to sell. After all, claiming a "gourmet" kitchen or "maple" floors won't really help you get a higher price unless you've actually got them. Especially when agents have a reputation to maintain among the other agents they sell to.

Words like "fantastic" and "charming" are probably correlated to low selling prices because those house have nothing better to be said about them.

The most notable assertion that has been made by Steven Levitt is his claim that the evidence shows that a major cause of the falling American crime rates in the 1990s was the legalization of abortion a generation before.

This shocking idea is summarized rather simply in the Introduction:

As far as crime is concerned, it turns out that not all children are born equal. Not even close. Decades of studies have shown that a child born into an adverse family environment is far more likely than other children to become a criminal. And the millions of women most likely to have an abortion in the wake of Roe v. Wade -- poor, unmarried, and teenage mothers for whom illegal abortions had been too expensive or too hard to get -- were often models of adversity. They were the very women whose children, if born, would have been much more likely than average to become criminals. But because of Roe v. Wade, these children weren't being born. This powerful cause would have a drastic, distant effect: years later, just as these unborn children would have entered their criminal primes, the rate of crime began to plummet.

More specifically they explain that women who choose to have an abortion usually have a good reason to not want to have a child. Whether it be the fact she's unmarried, or poor, or an addict, or too young, too unstable, or whatever, she "may feel that she cannot provide a home environment that is conducive to raising a healthy and productive child." The Steves suggest that, on the average of cases, these pregnant women were more likely to be right.

The argument is not a pro-abortion argument per se. The argument is based on studies that show typical children that went unborn when abortion was legalized were much more likely to come from impoverished households, single parents, teenage mothers, and undereducated mothers. They tell us that all of these factors have been shown by research to be strong predictors that a child is more likely to have a criminal future.

Levitt and Dubner review several of the other factors that have been credited with the sharp reduction in crime, such as innovative policing strategies, the strong economy, tougher gun laws and the aging of the population. Studies and anecdotes are referenced that suggest that these factors had marginal impact, at best. Enlarged police forces, increased imprisonment, and changes in the drug market were given some credit for the drop (10%, 33%, and 15%, respectively). But the book suggests that the biggest factor is the aforementioned legalization of abortion in the United States.

After getting used to the idea, it began to interest me. It was obvious that the Steves weren't saying that "easier access to abortion --> less crime", but something more like "easier access to abortion --> fewer children raised in poor home environments --> less crime". That's interesting because we can also focus on other ways to create that effect... "x --> fewer children raised in poor home environments --> less crime". Then I noticed that, at least in this book, there isn't much evidence presented that shows that the formula of "fewer children raised in poor home environments --> less crime" actually took place in the wake of Roe v. Wade.

I tried to research it myself, but wasn't terribly successful given that I don't have a great deal of time to spend. USA teen birth rates seem to have declined after Roe v. Wade in 1973 (and climbed again after the late 80s crack epidemic). But, there would seem to be no decline in the percentage of families with children under 18 years living below the poverty level. (Actually, there was a steady increase from a low in 1969 to a high in 1993.) And, the theory's chief critic, Steve Sailer, has a graph that seems to show no impact on the rate of births to unwed mothers, either. (Again, actually a steady increase.) So, it is not obvious exactly how the legalization of abortion in the United States is supposed to have led to the dramatic drop in crime rates. We may be left to credit this to a decrease in non-quantifiable "unwantedness".

On the whole I enjoyed Freakonomics, but I don't feel that it really lived up to the hype. I was not "dazzled", as Malcom Gladwell promised I would be. I found some of the analyses to be rather interesting, a couple left me a bit doubtful, but none blew me away. It was entertaining, but rather light. I'd recommend waiting for the paperback, or borrowing it from the library.

The same day that I bought Freakonomics, I also bought The Rebel Sell by Joseph Heath and Andrew Potter. Both books take, as a fundamental building block, the idea that "incentives are the cornerstone of modern life". I'm only about 100 pages into The Rebel Sell and it is already an order of magnitude deeper and more insightful than Freakonomics.


-permalink-

Links

.

Index has been moved to http://www.andrewspicer.com/toc.html