Freshman Group Think
By A.S. Erickson | Monday, September 22, 2008
BOOK REVIEW THE WISDOM OF CROWDS
James Surowiecki
Doubleday, 2004 Editor’s Note: The following book was assigned to be read by the class of 2012 by the First-Year Office. Dean Lacey, Professor of Government, chose the book and will give an address on it to the ‘12s upon their arrival.
The Wisdom of Crowds revolves around one idea and one idea only: crowds are better at making decisions than individuals. A startling claim at first. The author James Surowiecki waters his thesis down with a set of conditions that the group must meet in order for it to be intelligent—but more on those later.
Surowiecki, who pens the New Yorker’s business column as a day job, begins his book with a story from the life of the nineteenth century polymath Francis Galton. In his old age Galton was preoccupied with genetics, and, for that reason, found himself at a country fair one day looking at the effects that breeding can have on livestock. Surowiecki writes, “Breeding mattered to Galton because he believed that only a very few people had the characteristics necessary to keep societies healthy. He had devoted much of his career to measuring those characteristics, in fact, in order to prove that the vast majority of people did not have them.” In short, he was a eugenicist, and he was always on the lookout for more examples that would support his theory.
As he was wandering through the fair, he came upon a contest for the fairgoers: an ox was placed on display, and (for sixpence) passersby could guess how much the ox would weigh after it had been slaughtered and dressed. The closest guesses would receive prizes. Clearly, many of the guessers were experts: that is, farmers or butchers, but many were also non-experts, or laypeople out enjoying the fair. In his later paper about the incident, published in Nature, Galton indicated the parallels between the contest and democratic elections: “The average competitor was probably as well fitted for making a just estimate of the dressed weight of the ox, as an average voter is of judging the merits of most political issues on which he votes.” At the end of his book Surowiecki, too, attempts to tie his findings to politics.
After the contest was over and the winners announced, Galton collected all of the guesses in order to prove his theory; namely, that the average voter was genetically capable of very little when compared to the experts. The results stunned him. When Galton averaged all of the votes (nearly 800 of them) it came out at 1,198 pounds: the ox weighed 1,197 pounds. The average guess was closer than the best guess. Galton was flummoxed, and Surowiecki found the foundation for his book.
The insight behind Galton’s initial experiment drives The Wisdom of Crowds. “What Francis Galton stumbled on that day in Plymouth was the simple, but powerful, truth that is at the heart of this book: under the right circumstances, groups are remarkably intelligent, and are often smarter than the smartest people in them.” It is true that, in experiments like the ox contest or a ‘guess the number of jelly beans in a jar contest,’ one or several people will often guess better than the average. Yet, Surowiecki contends, the chances that a single person can consistently outperform the average are slim to none. The idea that the masses can outperform the ‘experts’ most of the time is so utterly counterintuitive that it makes for a fascinating argument, and Surowiecki pulls it off well.
Part of what makes Surowiecki’s claim seem so counterintuitive is the many examples society has seen of a mob mentality, where the whole group takes on the character of its most extreme (and in the example of mobs, most violent) members. So what is it, exactly, that makes a mass of people a ‘crowd’ and not a ‘mob.’ The author outlines three criteria that a group must meet in order for it to have the ‘wisdom of crowds’: independence, diversity, and decentralization.
All three are closely intertwined with one another. There must be independent and overlapping sources of information; without it groups fall prey to groupthink. The best way to guarantee this is to have diversity in the group—not diversity in the Dartmouth sense of the word, but diversity in backgrounds, assumptions, intellects, and beliefs. The best way to maintain this diversity and independence is by keeping its constituents decentralized.
Surowiecki focuses on the types of problems that collective intelligence can help to solve, and to make things simple, there are three of these as well: cognition problems, coordination problems, and cooperation problems. Cognition problems have definite and defined right answers. Sometimes there is no single right answer but better answers than others; these, too, are cognition problems. The ox contest, for example, was a cognition problem because the right answer was 1,197 pounds, and the best answer was whichever guess was closest to that number. Another example of a cognition problem is in sports gambling; at the end of the game there is a clear winner. Bookies are constantly fine-tuning their spreads based on the bets that are placed—at least, they are if they want to stay in business. The bookie is unconsciously listening to the wisdom of crowds when he does so. Corporations face these kinds of questions all the time: ‘How many SUVs will we sell next quarter?’ or ‘Will the FDA approve our new drug, if so, when?’
The group dimension to cognition problems is not immediately obvious; indeed, oftentimes questions about corporate strategy are answered by one man alone, the CEO. Coordination problems are quite clearly group problems; they are the problems that focus on how a group can coordinate their behavior with one another. For example, how buyers and sellers in a market are able to find each other and trade at a price that both agree on, or how the dynamics of traffic on a busy highway work.
The last problem, the cooperation problem, is often called the ‘free-rider’ problem in philosophy and other disciplines. How is it that most people pay taxes even though on an individual basis it makes more sense not to? this question asks. In other words, why do most people cooperate rather than free-ride off of others? The shortest answer is that if everyone free-rode than the system would collapse. These kinds of problems include not only tax paying, but also pollution control and wage-setting.
Throughout the book Surowiecki generously attributes some of his ideas to Friedrich Hayek, and any close reader of Hayek will already be familiar with many of the arguments in this book: that, for example, the aggregate of all the individualized and particular knowledge in a situation is more accurate than a central power’s knowledge could ever be. But while Hayek is focused primarily on markets, Surowiecki is interested in extending this insight as far as it will go.
His crusade could not be aimed at a more interesting subject; Surowiecki’s book is illuminated with countless engrossing case studies that either confirm his thesis or point out a situation in which the ‘wisdom of crowds’ would be of use. As to the latter examples, his most intriguing and exciting suggestion has to do with the use of internal markets for traditionally non-market-oriented decision-making.
One of the better-known examples of this is the Iowa Electronic Markets (IEM), which allows one, with a modest sum of money, to buy and sell futures on any presidential, congressional, or gubernatorial election in the U.S. The IEM outperforms the major national polls about seventy-five percent of the time. Taking into account the future’s volatile nature, markets tend to outperform other methods when it comes to predicting how the future will look.
Similar results have been documented when corporations turn to artificial markets for forecasting. In the 1990s, for example, Hewlett-Packard had a market set up to forecast printer sales. The market was small (only twenty to thirty people) but diverse: each of its members were taken from throughout the company. Each member would buy or sell shares according to how he thought printer sales would be next month, or next quarter. These markets ran for three years, and over that time they outperformed the company’s internal predictions three quarters of the time.
The government itself, recognizing the power of these decision markets, tried to install a two-tiered decision making market within the Department of Defense. The first part would have, like Hewlett-Packard, consisted of a small number of insiders from diverse parts of the DoD and other agencies. The second would have allowed the public at large to trade shares based on what they believe the future holds. Unfortunately this interesting project never got off the ground thanks to low-brow, populist extraordinaire Byron Dorgan, the senator from North Dakota. His understanding of the basics of supply and demand leaves much to be desired. As the nation moves forward, it will be interesting to see who chooses to harness the power of decision markets.
The author claims the book is divided in two, theory in the first half and case studies in the second. In reality, the book is one long string of interesting case studies after another—much to its benefit. Surowiecki writes briskly and clearly, conveying his thesis with the proper amount of verve. The class of 2012 should feel blessed for not being burdened with something tedious or fantastical this summer, as has been the case with the required reading of the past—but with something that is genuinely interesting: the wisdom of crowds.
|
