Book Review: Freakonomics (4 of 5)

Freakonomics is an interesting book, especially for anyone who likes numbers. There is no specific purpose, as the book itself admits, except maybe to challenge you not to accept at face value statistics that politicians and the media throw out at you, and to maybe dig a little deeper for the truth. Problem is, I already know not to always accept these statistics, and unless you are an economist who has access to large pools of data from governments, schools, and sumo wrestlers, you really don't have a way to check the numbers you are given.

That said, it's still a very interesting book. Levitt asks random unasked questions and pores over lots of data to find unexpected correlations. While he admits that a correlation between two things cannot determine why, he still concludes or strongly encourages us about the why's. Abortion in the 1970's caused a crime drop in the 1990's. Incompetent school teachers change their students' test answers to correct ones so No Child will get Left Behind and they won't look bad. Real estate agents don't work their hardest at selling your home because the difference between a $300,000 and a $310,000 sale is only an extra $150 in commission for them.

It's all good and interesting, except that as the book progresses, the Stevens offer us less examples of the actual data and resort to simply telling us that when this happened, this tended to happen. They make very convincing arguments, except if we are to take their own advice, we cannot accept the correlations they give us. States with higher abortion rates in the 1970's tended to experience higher crime drops in the 1990's. OK. Was that true for every single high-abortion-rate state or just most of them? Obviously, they can't give us all the raw data for all of their examples, because then the book would be several hundred boring pages of numbers. Although it is unavoidable, this by nature lessens the credibility of his conclusions.

For the most part, however, Freakonomics is a smart book that finds interesting and enlightening patterns in unexpected places. And they even do a good job of attempting to explain and find causes for these patterns. I only wish I had that kind of access to that kind of data...

Edmond the Hun