Skip to content
mentua.org

"The Wisdom of Crowds" by James Surowiecki: Summary and Review

Line drawing of a row of separate cards, each with its own written guess, linked by dotted lines to one ballot box with a filled circleLine drawing of a row of separate cards, each with its own written guess, linked by dotted lines to one ballot box with a filled circle

9 min read≈ 2,060 words5 questions · up to 5 points

What this is about

In 2004 a business journalist claimed that a crowd of ordinary people often decides better than the experts in it. You learn the four conditions James Surowiecki says a group needs, the three kinds of problems he applies them to, and why the same crowd can become a herd. You also learn what critics and later research found, and whether the book is still worth reading.

The book in three sentences· skip ahead

Surowiecki argues that a group of people often judges better than its smartest member, as long as the people think differently, decide on their own, use their local knowledge and have a fair way to combine their views. He tests this idea on guessing games, markets, teams, companies and governments. The book explains well why groups fail when people copy each other, but it overstates how often the right conditions exist.

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

Author
James Surowiecki
Published
2004
Publisher
Anchor Books (paperback, 2005)
Length
336 pages
Audiobook
9h 29m, read by Grover Gardner
Best for
Readers who make decisions in teams and want to know when a group's judgment can be trusted.

James Surowiecki’s book makes a claim that most of us find hard to believe: a crowd of ordinary people, combined in the right way, often judges better than the experts in it. Surowiecki, who wrote the business column “The Financial Page” for The New Yorker, published The Wisdom of Crowds in 2004. Its title answers Charles Mackay’s Extraordinary Popular Delusions and the Madness of Crowds of 1841, and that contrast explains why the book mattered. Mackay and, later, Gustave Le Bon had treated crowds as foolish or dangerous. Surowiecki argued that the same crowd can be wise or foolish, and that the way a group is organized decides which.

The book has two halves. The first sets out the idea and the conditions it needs. The second applies it to particular settings: small teams and committees, business, markets, science and democracy. This review follows the main ideas in order of their weight in the book.

Groups are often “smarter than the smartest people in them”

The central claim of the book is that a group can know more than any of its members. Surowiecki states it in the introduction.

Under the right circumstances, groups are remarkably intelligent, and are often smarter than the smartest people in them.

James Surowiecki, The Wisdom of Crowds, Introduction

He opens with the most famous case. In the autumn of 1906, the British scientist Francis Galton visited a country fair where about 800 people paid to guess the weight of an ox. Many of the guessers were butchers and farmers, but many others had no special knowledge of cattle, so Galton expected the average guess to be far off. Surowiecki reports that the crowd’s average was 1,197 pounds and the ox weighed 1,198. In his words, “the crowd’s judgment was essentially perfect”.

The mentua Idea What Is the Wisdom of the Crowd? explains why averaging works. It also adds a detail that the book leaves out. In his own article, Galton used the middle guess of 1,207 pounds, and he gave the average only later, after a reader asked for it.

Surowiecki adds modern examples. On the quiz show Who Wants to Be a Millionaire?, contestants who were stuck could phone a friend or ask the studio audience. The friends gave the right answer 65 percent of the time, while the audience vote was right 91 percent of the time. The Iowa Electronic Markets, where people bet real money on election results, did better than the polls. In the week before each of the four presidential elections that preceded the book, the market was on average 1.5 percentage points off the final result, while Gallup’s final polls were 2.1 points off.

The lesson Surowiecki draws is practical: instead of searching for the one expert who knows best, ask the group and combine its answers.

Crowds solve three kinds of problems: cognition, coordination and cooperation

Surowiecki sorts the problems that groups face into three types. Cognition problems have a right answer, such as the weight of an ox or the winner of an election. Coordination problems ask people to fit their actions together, such as walking through a busy street without collisions. Cooperation problems ask self-interested people to work together, for example to pay taxes or to trust strangers in trade.

The book is famous mostly for its cognition examples, but the other two types also get a fair amount of space. For coordination, he retells an experiment by the economist Thomas Schelling. In 1958 Schelling asked Yale law students where and when they would try to meet a stranger in New York if they could not agree on a place in advance. Almost all of them said noon, and more than half chose the information booth at Grand Central. For cooperation, Surowiecki borrows Margaret Levi’s term “contingent cooperators”: most people are willing to contribute as long as they believe that others do the same.

A wise crowd needs four conditions

The core of the book is a list of four conditions under which a crowd judges well. Each condition protects the private information that different people hold.

  1. Diversity of opinion. Each person should bring some private information or a different way of looking at the facts, even an unusual one.
  2. Independence. People’s views should not be decided by the views of those around them.
  3. Decentralization. People should be able to specialize and use their own local knowledge.
  4. Aggregation. Some method, such as an average, a vote or a market price, must turn the private judgments into one collective answer.

The chapter on diversity is called “The Difference Difference Makes”. Surowiecki’s point is that a group of smart people who all think alike is weaker than a mixed group, because similar people make similar mistakes. His example of failure is the Bay of Pigs invasion of 1961. President Kennedy’s advisers kept their private doubts to themselves and agreed on a plan to overthrow Fidel Castro’s government in Cuba with a small force of exiles, and the invasion failed.

Copying turns a crowd into a herd (“information cascades”)

The chapter on independence, “Monkey See, Monkey Do”, is the most useful in the book. It explains how crowds fail through information cascades. A cascade starts when people decide one after another and can see the earlier choices. Each new person gives more weight to what others did than to what they know themselves. After a few steps, everyone follows the first choices, and the group stops adding new information, even if the first choices were wrong.

Imagine two restaurants side by side. The first guest picks one at random, and the second guest sees one table taken and follows. Every later guest does the same, so the busy restaurant gets busier while the empty one might be better. Surowiecki’s point is that a hundred people who copy each other are one opinion, not a hundred.

In a 2006 interview with CBS News, he applied this to the dot-com bubble of the late 1990s. Investors stopped asking what Amazon.com was really worth and asked what others would pay for its stock tomorrow. He described the result as “this kind of vicious cycle, where everyone is really just looking at what everybody else is doing”. In this reading, a bubble shows what happens to a crowd that has lost its independence, which is a narrower claim than saying that crowds are stupid.

Decentralization works only if someone puts the pieces together

Surowiecki likes decentralized systems, in which people close to a problem act on their own knowledge. Among his examples are markets and science. He points to the research on the SARS virus in 2003 as a case where scientists who shared their information freely made fast progress.

But he insists that decentralization has a cost. Local knowledge helps only if it reaches the people who decide. His example is the American intelligence services before the attacks of 11 September 2001. The 9/11 Commission found that one agency held information that another needed, and nobody combined the two. So a decentralized system needs a way to combine what its parts know, or the knowledge stays where it is and nobody uses it.

Small groups and markets can be wise or foolish

The second half of the book applies these ideas to settings that most readers know from work. On small groups, Surowiecki discusses the decisions of NASA’s managers before the space shuttle Columbia broke apart in 2003. In his account, a hierarchy that valued agreement did not take the doubts of its own engineers seriously. He also reports on investment clubs: the worst performers were close-knit groups that decided by consensus, while the best welcomed dissent. The practical lesson that follows from the four conditions is to ask people for their own view before the discussion starts, so that the leader’s opinion does not set everyone else’s.

On markets, he praises prediction markets, where people trade bets on future events and the price works as a forecast. The Hollywood Stock Exchange, an online market for films, correctly predicted 35 of the 40 Oscar nominees in 2002. The last chapter turns to democracy, which he sees less as a way to find correct answers and more as a way to solve coordination and cooperation problems.

How it differs from other books about crowds

Three neighbors show what is new in the book.

  • Charles Mackay’s Extraordinary Popular Delusions and the Madness of Crowds (1841) collects stories of bubbles and fads in which crowds go wrong. Surowiecki’s title answers Mackay directly, and his chapter on cascades explains the mechanism behind Mackay’s stories. Our review is “Extraordinary Popular Delusions and the Madness of Crowds” by Charles Mackay: Summary and Review.
  • Gustave Le Bon’s The Crowd (1895) argues that people in a crowd lose their judgment and form one emotional mind. Surowiecki’s crowds are different in kind: they are people who judge separately and whose views are combined, not people who shout together in a square. The two books describe two different situations rather than contradicting each other. Our review is “The Crowd” by Gustave Le Bon: Summary and Review.
  • Daniel Kahneman’s Thinking, Fast and Slow (2011) explains why single people make predictable errors. Surowiecki explains how a group can cancel those errors, as long as they point in different directions. Kahneman’s advice to collect each person’s written view before a meeting fits Surowiecki’s condition of independence. Our review is “Thinking, Fast and Slow” by Daniel Kahneman: Summary and Review.

The book’s distinct strength is its range. Surowiecki moves from an ox at a fair to quiz shows, betting markets, intelligence agencies, science and government, and he keeps showing the same four conditions at work.

Where it falls short

The book has three main weaknesses.

The first is that its conditions are less precise than they sound. The legal scholar Cass Sunstein, in a long and friendly review in The New Republic, argued that the conditions are neither necessary nor sufficient. They fit cognition problems, but coordination often works better when people are similar and follow each other. Sunstein also pointed out that a crowd can only be wise if its members know something. The French mathematician Condorcet showed in 1785 that when each voter is more likely to be wrong than right, a bigger group becomes more likely to be wrong. Sunstein called the book’s silence on this result “a significant omission”.

The second weakness is the choice of examples. Most of the evidence consists of stories told after the fact, and the reader cannot tell how many cases that did not fit were left out. Sunstein also noted that prediction markets have special features, such as voluntary trading by a few well-informed people, that do not carry over to crowds in general. Later research has also added limits. In a 2011 experiment by Jan Lorenz and colleagues, students who saw each other’s estimates became more confident but not more accurate, which supports the book’s warning but shows how easily independence is lost.

The third weakness is age. The book appeared in 2004, before social media became part of daily life and before the financial crisis of 2008. Its examples of wise crowds online, such as Google ranking web pages by the links that many people make, look less convincing in a world where most people see what their friends think before they form their own view. Surowiecki himself gave the reason in the CBS interview: “Crowds go wrong when diversity really breaks down”.

Which edition and format to choose

The facts on this page describe the American paperback from Anchor Books (2005); in Britain the book appeared with Abacus. Surowiecki argues mainly with stories, not with charts or mathematics, so the book works well as an audiobook. The unabridged audiobook, read by Grover Gardner, lasts about nine and a half hours; choose it over the shorter abridged version that Random House also published in 2004.

If you read in print, read the introduction and the chapters on diversity, independence and decentralization carefully, because they contain the theory. The later chapters on business and science can be read more quickly, while the chapters on small groups and markets deserve a slower reading.

Quiz

5 questions · up to 5 points

Open a question and pick one answer. Only your first answer counts: +1 if it is right, −1 if it is wrong.

  1. What is Surowiecki's central claim in The Wisdom of Crowds?
    Show answer

    Under the right conditions, a group often judges better than even its smartest member. This is the thesis of the introduction. The words "under the right conditions" matter, because most of the book is about those conditions.

  2. Which four conditions does Surowiecki say a wise crowd needs?
    Show answer

    Diversity of opinion, independence, decentralization and aggregation. These are the four conditions. Each one protects the group's private information from being lost or copied.

  3. Restaurant A is empty and restaurant B next door has a few guests, so every new visitor picks B. What does Surowiecki call this pattern?
    Show answer

    An information cascade, in which people follow earlier choices instead of their own information. Each visitor copies the visible choices of those before. The group stops adding new information, even if A is better.

  4. What does the book say about committees and small teams, such as NASA's managers before the Columbia disaster?
    Show answer

    They can be wise, but they often fail when they value agreement more than dissent. Surowiecki blames the Columbia decisions partly on a hierarchy that did not use engineers' doubts. The Bay of Pigs fits the same pattern.

  5. How does this review suggest reading The Wisdom of Crowds today?
    Show answer

    As a clear guide to when groups judge well, read next to its critics and later research. The conditions and the analysis of failures still hold. The book's optimism about markets and online crowds needs the later evidence beside it.

The takeaway

James Surowiecki's "The Wisdom of Crowds" (2004) argues that a group often judges better than its smartest member if four conditions hold: diversity, independence, decentralization and a fair way to combine views. He shows this with Galton's ox, quiz shows and prediction markets, and he explains how teams and markets fail when people copy each other. Critics say the conditions are rare and that the members must know something. Before a group discusses a question, collect everyone's own answer first.

Key quotes

“Under the right circumstances, groups are remarkably intelligent, and are often smarter than the smartest people in them.”
James Surowiecki, The Wisdom of Crowds, Introduction
“Crowds go wrong when diversity really breaks down.”
James Surowiecki, Interview with CBS News, 8 January 2006

Should you read it?

Read it if

  • You run meetings, estimates or forecasts and want to know why collecting views separately works better than discussing first.
  • You want a readable account of prediction markets, information cascades and groupthink in one book.
  • You have read Le Bon or Mackay on mad crowds and want the strongest case for the other side.

Skip it if

  • You already know the idea well; the Idea on the wisdom of the crowd covers its core in a few minutes.
  • You want rigorous evidence or mathematics; the book argues mostly with stories and a few studies.
  • You want an account of online crowds and social media, which came after the book.

As an Amazon Associate, mentua.org earns from qualifying purchases. It costs you nothing and never shapes the verdict.

Ideas from this book

References

  1. The Wisdom of Crowds. James Surowiecki, Penguin Random House

    The publisher's page for the book, with editions and audiobooks.

  2. Mobbed Up. Cass R. Sunstein, The New Republic, 2004

    A long and fair review by a legal scholar, who praises the book but argues that its conditions are neither necessary nor sufficient.

  3. How Social Influence Can Undermine the Wisdom of Crowd Effect. Jan Lorenz, Heiko Rauhut, Frank Schweitzer and Dirk Helbing, Proceedings of the National Academy of Sciences, 2011

    An experiment that confirms the book's warning: when people see each other's estimates, they grow more confident but not more accurate.

the-wisdom-of-crowdsjames-surowieckiwisdom-of-the-crowdcollective-intelligencefrancis-galtonprediction-marketsiowa-electronic-marketsinformation-cascadesgroupthinkbay-of-pigscolumbia-disasterdiversityindependencedecentralizationaggregationherd-behaviorbook-reviewsummary