I hadn't read reviews of Think Like a Freak, but Levitt and Dubner's first book (Freakonomics) was very good, especially where they were writing about Levitt's own research, which often makes use of cute natural experiments. However, the supply of cute natural experiments is pretty limited, and I don't think their second book (SuperFreakonomics) was nearly as good, especially where they had to rely much more on packaging other people's research than on Levitt's own contributions. The second book was able to pull some excellent content gathered from their blog though. So, given the trajectory the series was on it's fair to say I had reasonably low expectations of the book.
When it comes to pop-economics books, I can usually judge them on the number of contributions they will eventually make to my first-year teaching (i.e. how many cool new examples, or quirky applications, or improved ways of explaining economic concepts can I draw from them?). So, I was disappointed that there was little contribution there for me. There is not a great deal in the way of applied economics at all, and much of the book came across to me as re-hashed parts from the first two books, or re-packaged stories that I have read elsewhere. If you've read the first two books, you may feel the same way.
Having said that, the book wasn't intended to follow the same formula as the first two. It's pitched as more of a self-help book. As the authors note on page 9:
The first two books were rarely prescriptive. For the most part, we simply used data to tell stories we found interesting, shining a light on parts of society that often lay in shadow. This book steps out of the shadows and tries to offer some advice that may occasionally be useful, whether you are interested in minor lifehacks or major global reforms.However, if you're already attuned to an economic way of thinking about the world, the self-help bits come across as largely common sense and the value of the book will be lost on you. At that point you're probably better switching to some Tim Harford instead.
It wasn't all bad though. Some of the stories were still useful or interesting - for instance, I always like new examples of the unintended consequences created by incentives and the book provided me with a few, and the discussion of why Nigerian email scammers admit they are from Nigeria. Being let off the hook on why SuperFreakonomics gave away the secrets of the algorithm that uses bank data to identify terrorists (from SuperFreakonomics) was good too (see also this Levitt paper from 2012 (gated)).
Overall the book wasn't the worst I have read over the last year (I won't name and shame that title), but far from the best either. Looking forward, unless Levitt has identified some new cute natural experiments to share, I have to agree mostly with the conclusion to Nick Summers' review on the Bloomberg BusinessWeek site:
This is how brands end, with the sound of scraping the bottom of a barrel Malcolm Gladwell has already rummaged through. In the book’s last chapter, about the virtues of quitting, Levitt and Dubner wonder if they too should throw in the towel. “After three Freakonomics books, can we possibly have more to say—and will anyone care?” I think they know the answer.Other reviewers share Nick's pessimism about the 'Freakonomics brand' (see David Runciman at The Guardian as one example), or for a couple of less negative reviews see Mahmoud Kassem at The National or Philip Roscoe at Times Higher Education.