This is a hard book to review. The first half is economists being genuinely surprised that humans aren't homo economicus, accompanied by bad pop-cogsci for why this might be, and a laughably bad understanding of mathematics (they routinely discount base rates, and call all sorts of things "s-curves", none of which are remotely s-shaped). This section is full of boring anecdotes there presumably to add a human touch to an otherwise dry book, but all that accomplishes is wasting the reader's time. (Un)notably, there are no proposed solutions to the problems described, nor any data that would be helpful to anyone in the First World. Maybe it's interesting if you're an economist, but I'm not.
The second half of the book is where things start to get interesting. The discussion becomes more abstract and starts talking more about how to actually solve some of these problems, about how economics can be used to incentivize some (but not all) solutions in the domain, and about where the standard economic tools will fail. This stuff is fascinating, and presents a lens through which we can inspect our own society---where we succeed and where we do not. The final chapter is on corruption and policies, and I would recommend the chapter to anybody who cares about social change (though I would *not* recommend the book itself.) The description blurred the stark line I had in my head about what constitutes corruption, and why it's not always necessarily a bad thing (eg. inappropriately funneling resources towards people who need them.)
All in all, 3/5. If you're particularly interested in pop-economics, you could do worse than this book, but I'd recommend Yudkowsky's Inadequate Equilibria over it any day.