Heeding The Herd Instinct


Why the Many Are Smarter than the Few and How Collective Wisdom Shapes Business, Economies, Societies, and Nations

By James Surowiecki

Doubleday -- 296pp -- $24.95

Crowd behavior is usually considered an ugly phenomenon. Think lynchings, riots, Nazism, China's cultural revolution -- or even financial bubbles and panics. Unreason and prejudice seem to rule. Chances are most people would agree with the Frenchman Gustave Le Bon, who wrote in 1895: "In crowds, it is stupidity and not mother-wit that is accumulated."

Financial journalist James Surowiecki begs to differ. In The Wisdom of Crowds, the New Yorker staff writer provocatively argues that, in many circumstances, the group collectively reaches better decisions -- and solves problems more efficiently -- than the smartest man or woman alone. In this, the author is supported by social scientists who salute "decentralized self-organizing systems," such as Adam Smith's invisible hand. Crowd wisdom can be seen in the superiority of American market capitalism over Soviet central planning. Surowiecki provides numerous examples of how the many are often smarter than the few, and he explores the implications of the phenomenon. He discusses when tapping into the crowd pays off big and why the group can go wrong. He makes a strong case that society should take advantage of crowd insights rather than depending on experts. In essence, he suggests, Le Bon got it backwards.

The capital markets provide the classic example of Surowiecki's thesis: Even if they are sometimes prone to bouts of enthusiasm or depression, they're an amazing social and economic institution for communicating all kinds of data and knowledge through price changes. The more pervasive the financial markets, the more investors will find and fund profitable ideas and, at the same time, flee from failed management strategies. Sound collective judgment shows up elsewhere in the economy, too. Linux, the open-source operating system created by Finnish programmer Linus Torvalds in 1991 but effectively owned by no one, is now the major rival to Microsoft Corp.'s (MSFT ) Windows. Independent computer programmers from around the world contribute to improving the operating system, and solving the problems that intrigue them, although Torvalds and his peers keep a tight rein on what changes are acceptable.

The collective beats the individual expert in smaller groups, too. Surowiecki offers a number of examples, including the fascinating case of the May, 1968, disappearance of the submarine USS Scorpion on its way to Newport News, Va. The U.S. Navy had a general idea where the sub sank, but it was an area 20 miles wide and many thousands of feet deep. Naval officer John Craven hit upon a solution. He gathered a group of diverse experts and asked for their best guesses on why the sub ran into trouble, its speed as it fell to the ocean floor, the slant of its descent, and so on. Craven took all the speculations, ran them through a sophisticated mathematical formula, and ended up with the team's overall guesstimate. The Navy found the ship 220 yards from where Craven's group had predicted it would be, yet not one individual had picked that spot. "The final estimate was a genuinely collective judgment that the group as a whole had made," says Surowiecki. "It was also a genuinely brilliant judgment."

Of course, the crowd can go spectacularly wrong, as some of the best parts of this book reveal. For instance, the stock market works well most of the time. It is a decentralized mix of enthusiasm and skepticism, longs and shorts -- a global conflict of opinion and judgment that keeps prices within a reasonable approximation of value. Yet speculation at times spirals out of control, and the market becomes a single-minded mob. Witness the latter days of the 1990s, when a critical mass of dazzled investors began thinking that a price-earnings ratio of 100 was conservative and a price-earnings ratio of infinity alluring.

Among the necessary conditions for wise counsel from the many are diversity and independence, as well as mechanisms for aggregating and producing collective judgment. Certain "decision markets" fit that bill, including the Iowa Electronic Market for betting on election results, which has proved superior to polling, and the Hollywood Stock Exchange for popular wagering on box-office returns, which beats other Tinseltown forecasting tools. So, Surowiecki wonders, why do companies rely so much on the judgment of one person -- namely the CEO? Despite genuflections to decentralization, most fail to tap the insights of their employees, suppliers, and others by using well-designed decision markets for everything from forecasting demand to deciding which products to create. One reason: Such an approach might limit CEOs' ability to justify their gargantuan compensation packages.

Surowiecki isn't an ideologue, doesn't shy from evidence that counters his theory -- and isn't always persuasive. Still, he musters ample proof that the payoff from heeding collective intelligence is greater than many of us imagine.

By Christopher Farrell

    Before it's here, it's on the Bloomberg Terminal.