Quite entertaining, but watch out for those "it turns out that..."s. I would have preferred something that made the research accessible rather than this overly-simplified popularisation. At times they get ahead of themselves, using the technical terms "controlling for", "correlated" and "causal" for most of the book before giving a rough explanation.
Levitt's talk at TED is well worth a watch. Indeed, the economics of a Chicago drug gang is the best story in the book. (The effects of baby name choice bored me senseless.)