Freakonomics (Levitt, Steven D., Dubner, Stephen J) 2005

I was a little worried by the explanatory note which included the following clip from the original NYT article: "Unlike most academics, he [Levitt] is unafraid of using personal observations and curiosities; he is also unafraid of anecdote and storytelling (although he is afraid of calculus). He is an intuitionist." Using anecdote and personal observation to motivate an analysis is one thing, but without good numerical evidence to back it up, anecdote is worse than useless because it can lead to the wrong result, rather than no result at all. The above is immediately followed by "He sifts through a pile of data to find a story that no one else has found. He figures a way to measure an effect that veteran economists had declared unmeasurable." That's better, but let's read the rest of the book to find out whether that's actually true.

Danger in the Introduction! The very first story concerns why crime rates in the US fell in the mid-late 90's after many were predicting an ever-increasing surge. After dismissing some of the popular theories (booming 90's economy, gun control, better policing), Levitt offers his own theory: Roe v. Wade prevented the birth of the very type of children (those of poor, unmarried, teenage mothers) who had been the cause of rising crime rates in the 80's. This is a good theory; better than the popular theories referenced in the book, but unfortunately he offers no real data to back it up.

The body of the book itself is much better. Lots of good "stuff", all well supported by data. Ah, and he even goes further into the data supporting the problematic argument in the Introduction. Patience is a virtue.

Upon completion, I have to admit that I don't understand the "freaky" part of freakonomics. What Levitt is doing is looking at a problem and hand-crafting his statistical models to solve the problem, rather than mashing it into an existing solution in the econometric toolbox. This is what any statistician or econometrician worth their spit does (at least among the ones I know) so Levitt's approach would be refreshing and revolutionary... but only if it actually were. That doesn't change the fact that it's still an entertaining read.