This observation of John's is quite amusing:
"The most interesting finding from this experiment, however, was that when it was carried out amongst students, the most selfish behaviour, the behaviour regarded as most rational by classical economics, was exhibited by those students studying Economics. In other words, being taught to perceive the world through the goggles of classical economic theory, altered the way students conceptualized their relationship with other humans and with the world as a whole."
I'm surprised, though, that behavoural economics being seen as a new thing (it is being touted as a new thing, isn't it? That's the impression I'm getting....).
I remember - years ago - reading another pop-economics book (bought seconds before a train was leaving, as usual) - Paul Ormerod's 'Death of Economics' (which is widely reviewed on this link). I thought it was very good, by the way. Then (from memory - my copy was long-since loaned out) Ormerod made the point that econometrics were almost worthless from any practical point of view and that studying behaviour without forced templates was where it was all at.
I'd question one thing though. Behavioural Economics seems to offer something fairly interesting as far as I can see. Sure, some versions are better than others. I tried with Freakonomics and was irritated by about five things in the first five minutes reading it, so I put it down. But I'm only just getting into Nudge at the moment, and I've been dipping into Yes! a fair bit, and they both look quite good so far. Working out how to motivate people to do things is of interest to people other than ad agencies and marketeers, surely?