The Wisdom of Crowds: Why the Many Are Smarter Than the Few and How Collective Wisdom Shapes Business, Economies, Societies and Nations

Surowiecki argues that aggregated judgments from many people frequently outperform the judgment of experts, and specifies the four conditions required for this to hold: diversity of opinion, meaning members hold different private information; independence, meaning their judgments are not determined by others; decentralisation, meaning they can draw on local and specific knowledge; and aggregation, meaning a mechanism exists for combining the judgments into a collective one.

His opening example is Galton's 1907 observation that the median guess of nearly eight hundred people at a cattle weight was within one pound of the correct figure. He extends this to markets as aggregation mechanisms, to the search for the missing submarine Scorpion, and to prediction markets.

He divides problems into cognition problems with a definite answer, coordination problems where each participant must anticipate others, and cooperation problems requiring self-interested people to work together, and analyses each separately. The second half examines failures. When independence is lost, information cascades produce herding, in which each participant rationally follows the previous ones and the aggregate information disappears. He covers the Columbia shuttle disaster, stock market bubbles and committee dynamics as cases where the conditions failed.

The conditions are the book's contribution, since they specify when crowd judgment works and when it fails, which is the part most often dropped in popular use. The book is journalism rather than research and the cases are selected. The independence condition has become harder to satisfy as social platforms have made judgments visible before they are formed, which strengthens the book's warning rather than its optimism.