Let me start by saying I really enjoyed Freakonomics and after seeing the slew of bad reviews for the sequel, my expectations going in were pretty low. Overall, the books is not a terrible read - it's quick and easy and Levitt and Dubner do tell some interesting stories. My problem with the book is that the arguments are intellectually inconsistent and the book essentially refutes itself.

Take, for instance, the point the authors make about drunk walking being more dangerous than drunk driving. The model they use has been well-refuted around the blogosphere (check out Ezra Klein or Tom Vanderbilt for more on the specifics). The problem with the authors' back of the envelope calculation is actually described in the third chapter of their book, which is all about the law of unintended consequences! Did the authors even consider what might happen when they tell thousands of people that you can statistically drive across the country and back, intoxicated, before getting caught, hurt or killed? Or that drunk driving is several magnitudes less dangerous than drunk walking?.. something most people probably already perceived as perfectly safe?

The book really starts to refute itself in the last chapter, the infamous global warming chapter. There's been tons of criticism that the authors got the science and the facts wrong. My issue is with the argument they make against climatologists' models. After spending a hundred odd pages making claims built on statistical models with dubious assumptions and drawing conclusions that seem absolute, the authors attack climate change models for, of all things, questionable statistics, dubious assumptions and claims that the results as absolute. If there is good reason to question global warming models, then there is equally good reason to question just about every one of the models described in SuperFreakonomics.

In the end you have to wonder what the motivation was to publish this book. Levitt was a generally well-respected economist and co-editor of a prominent economic journal. But enough of the book fails to meet the standards of academic rigorousness that any journal would call for that I can't stop coming back to the question, "why"?

0 comments: