Just finished reading it yesterday. Overall a fairly interesting read, though J recently tells me that Levitt’s paper which argues for the link between abortion and the crime drop in the 1990’s has been discredited due to a bug in Levitt’s analysis code. I don’t know the details there.
But overall, his knack for picking out seemingly random everyday questions, and finding the data to provide answers for them is easy to appreciate. I feel the book itself loses steam in the latter chapters.. the conclusions about the effect of parents, or the significance of a child’s name seem less solid, probably because the data has less interesting thing to say. The questions he asks are still interesting, but the results seem to contradict less what conventional wisdom would say about the matter.
The one annoying thing I found was that the author goes on and on about the difference between correlation and causality. I’d be fine with a general overview for those readers who aren’t familiar with the concepts, but it seemed that the book almost lectures you every chance it gets. Paragraphs easily skipped, though.