by Richard H. Thaler

isbn: 0691019347
subject: Nonfiction, General Science, Economics
finished: 1/12/2001

The Winner's Curse is a survey and discussion of the literature relating to economic anomalies -- situations where people's observed behavior differs from the behavior predicted by the rational, self-interested "economic man". The thesis of the book is that since anomalies demonstrably exist, economists would do well to come up with theories that explain and predict people's actual behavior rather than getting too hung up on rationality.

I would divide the anomalies in this book into two categories. Some of them are genuinely surprising and interesting, and they lead to new insights about economics and human behavior. The winner's curse is probably the best example (but there are several other great ones) -- it says that the winner of an auction, who was willing to bid more than anyone else, has probably paid too much. The less interesting anomalies are artifacts of the assumption that people are rational and self-interested, which anyone but an economist can see is not always true.

Something I found fascinating was the variety of techniques that economists use to gather data about anomalies. Broadly speaking, these can be divided into analyzing existing data (for example, stock prices or bets on horse races), and generating new data through experiments. The experiments can be divided into those that use real economic incentives (for example, asking if a person prefers a 50% chance of being given $1 or a 1% chance of winning $40, and then actually giving them the appropriate amount of money), and those that attempt to study preferences by asking hypothetical questions (for example, asking people if they would pay $10,000 to reduce their chances of dying in a car accident by 1%). Each method has different advantages and disadvantages, and while reading this book I was struck a number of times by the level of cleverness and creativity that went into designing the experiments.

The Winner's Curse isn't hard-core economics, but it isn't light reading either. In order to get the most from this book, readers should probably have a basic knowledge of economics or be prepared to look up technical terms as they arise. Although it's a little slow in places, I thought this was a great book and highly recommend it.