I finally got round to reading the book Freakonomics - I'm afraid it wasn't as good as I wanted it to be at all.
It's basically about applying economics ideas (incentives and disincentives, etc) to everyday things, like what might make a schoolteacher rig the results of her children's exams. And while it is full of interesting titbits (like the correlations between abortion laws and general crime statistics), it all feels very superficial and also very very self-aggrandising.
Every chapter begins with a quote from the New York Times telling us how insightful and mould-breaking Steven Levitt is, by telling some story about how he wowed some intellectuals at Stanford or something. But then the chapters themselves don't really help you to get to know him, or how or why he does his work. They just go through a fairly routine process of the following form: "You might have thought the statistics for X would be affected by Y, and that's what the so-called experts were saying in the papers. But actually that's not true, because I found a big set of data and checked it out." (Or the reverse, where a surprising correlation is found.)
There's no flavour of the working methods, and no way for the reader to come away from the book with an understanding of how to apply these tools to new situations. Lots of interesting facts, but not very satisfying.