Book Review: The Wisdom of Crowds

The Wisdom of Crowds
by James Surowiecki (2004)

Summarised by Lee Yee Fuang

James Surowiecki, the New Yorker financial columnist, provocatively argues that, in many circumstances, the crowds collectively reach better decisions, solve problems more efficiently and even predict the future better than the smartest man or woman alone. This challenges our mental model that valuable knowledge is concentrated in a very few hands, therefore key to solving problems or making good decisions is finding that one right person who will have the answer. His central argument is supported by social scientists who salute "self-organising, decentralized systems," (p. 70).

Surowiecki defines crowds in a broad sense, referring to groups of people from markets to horseracing bettors to game-show audiences and even the Internet. The book was born with the author’s initial writings on markets, therefore crowds’ behaviours in the market are immensely analysed. Notwithstanding, his counterintuitive thesis offers very provoking insights into how businesses operate, how knowledge is advanced, how economies are or should be organized, and how we live our daily lives. With sufficient research done and in a delightfully clear prose, he presents a very persuasive case by probing into fields as diverse as economics, politics, culture, psychology, biology, military and artificial intelligence to show just how this principle operates in the real world.

The book is structured into two parts: the first part, Chapter 1- 6, explains the theory of why crowds may be smart, the three problems crowds usually attempt to solve and conditions necessary for crowds to be wise; the second part, Chapter 7 – 12, explores the different case studies of each type of problem.

Chapter 1 begins with numerous lucid, concrete and compelling examples of how the many are often smarter than the few and what the implications of the phenomena are, revealing that the probability and consistency of crowds making good decisions are much higher than one smartest man in the crowds. Therefore, the crux of the issue is attempting to "chase the expert," looking for one man who will have the answers to an organization's problem, is a waste of time (p. 34).

Surowiecki contends that tapping into the crowd pays off big; citing one of the examples of a naval officer by the name of John Craven, who assembled a team of men with a wide range of knowledge to guess where the U.S submarine Scorpion could have been, after its disappearance in 1968, managed to get it right with the collective intelligence of the team. Yet, he also explains why the group can go wrong; citing the NASA/Columbia disaster as an example where one of the technical advisers convinced from the beginning that foam simply could do no serious damage and kept saying so to anyone who would listen, when there was plenty of evidence to suggest otherwise. Instead of working with the evidence and gathering for more information, the team succumbed to confirmation bias which causes decision makers to unconsciously seek those bits of information that confirm their underlying intuitions.

According to Surowiecki, there is a taxonomy of three types of problems that individuals and groups try to solve; they are:

Cognition (Chapter 7) refers to finding one definitive answer that we try to accurately assess after considering available and missing information. There is not a single right answer, but to which some answers are certainly better than others (e.g. what the best place to build this new public swimming pool? How many copies of this new ink-jet printer will we sell in the next 3 months? What’s a stock worth? Who will win an election? Or what caused a disaster?).

Coordination (Chapter 5) refers to a problem that affects a whole group, and where finding the optimal solution is usually sought by having each individual act in personal self-interest. It requires members of a group (market, subway riders, college students looking for a party) to figure out how to coordinate their behavior with each other, knowing that everyone else is trying to do the same (e.g. how do buyers and sellers find each other and trade at a fair price? How do companies organize their operations? How can you drive safely in heavy traffic? How do you find the best route to work in traffic?)

Cooperation (Chapter 6) refers a problem that affects a whole group, and where finding the optimal solution usually depends on individuals trusting each other and acting fairly and in what they perceive to be collective self-interest rather than just their own. The challenge is to get self-interested, distrustful people to work together, even when narrow self-interest would seem to dictate that no individual should take part (e.g. how to deal with pollution, devise a tax system, or remunerate employees).

Surowiecki believes that crowds can be wise in solving the above problems if they fulfil the four conditions:

Diversity of opinions (Chapter 2): Crucial not in a sociological sense, but a conceptual and cognitive sense to ensure good decisions made. It contributes not just by adding different perspectives to the group but also by making it easier for individuals to say what they really think. Each person uses their own private information, even if it is just an eccentric interpretation of the known facts, so that everyone brings different pieces of information to the table. Essentially, any time most of the people in a group are biased in the same direction, it is probably not going to make good decisions. So when diverse opinions are either frozen out or squelched when they are voiced, the groups tend to be dumb.

This condition is illustrated with an analogy drawn from the bees where the colony sends out a host of scout bees in many directions who would return and does a good waggle dance to attract other forager bees, so that at last the hive will know where the food source is. Ransom Olds, Henry Ford and other would-be automakers who tried and failed are likened to foragers. Finally, they discovered the nectar sources – the gasoline-powered car, mass production, the moving assembly line – and publicity stunts are a kind of equivalent to the waggle dance.

Fostering diversity is more important in small groups and in formal organizations rather than larger collectives like markets as, according to Surowieck, the sheer size of most markets means that a certain level of diversity is guaranteed. Small groups tend to be easier for a few biased individuals to influence the group’s collective decision; on teams or in organizations, cognitive diversity needs to be actively selected.

Research shows that experts’ judgments are neither consistent with the judgments of other experts in the field nor internally consistent, and an expert presented with the same evidence would, half the time, offer a different opinion. Experts are bad at what social scientists called “calibrating” their judgments and they routinely overestimate the likelihood that they are right. But this does not mean that well-informed, sophisticated analysts are of no use in making good decisions, for example, crowds of amateurs trying to collectively perform surgery or fly planes would not be a good idea. What Surowiecki means is that how well-informed and sophisticated an expert is, his advice and predictions should be pooled with those of others to get the most out of him. Diversity makes crowds wise.

Independence of members from one another (Chapter 3): People in the crowd need to be independent with their opinions not determined by the opinions of those around them, so that they pay attention mostly to their own information or private information, and not worrying about what everyone around them thinks. When people start paying too much attention to what others in the group think, it usually spells disaster.

Independence is important to intelligent decision making for two reasons: first, it keeps the mistakes that people make from becoming correlated. Errors in individual judgment will not wreck the group’s collective judgment as long as those errors are not systematically pointing to the same direction. Second, independent individuals are more likely to have new information rather than the same old data everyone is familiar with. The smartest groups are made up of people with diverse perspectives who are able to stay independent of each other. You can be biased or irrational, but as long as you are independent, you will not make the group dumber.

Psychologists carrying out a study of men’s influence found out that the more people put on a street corner and looked at an empty sky for sixty seconds, the more passersby stopped, tilted their heads and looked up. This explains why the crowd becomes more influential as it becomes bigger. The governing assumption is when things are uncertain, the best thing to do is just to follow along. It is not an unreasonable assumption. But the catch is that if too many people adopt that strategy, it stops being sensible and the group stops being smart.

This also happens to fund managers, football coaches, or corporate executives who demonstrate the symptom of herding and mimicking to protect themselves because that is where it is safest, but they destroy whatever information advantage they might have had, since mimicking managers are not trading on their own information, but relying on others’ information. This also explains the phenomena of stock-market bubbles which are a classic example of group stupidity: instead of worrying about how much a company is really worth, investors start worrying about how much other people will think the company is worth.

Therefore, one key to successful group decisions is getting people to pay much less attention to what everyone else is saying (p. 65) and the paradox of the wisdom of crowds is that the best group decisions come from lots of independent individual decisions.

Decentralization (Chapter 4): The idea of the wisdom of crowds needs to be decentralized, so that no one at the top is dictating the crowd's answer; people are able to specialize and draw on local knowledge. Flocks of birds being able to break up, reform, break up and reform seem to be working on one mind to protect themselves from the mercy of their prey, but they are actually acting on its own and following certain rules. No bird can command another bird to do anything, but they work on what they know (Ch. 5, p. 101). The success of Linux, owned by no one, gathering individual contributions, on their own, from all over the world to fix its problems and make it ever-more reliable and robust is another good example to illustrate a strong case for a decentralized system.

Yet, Surowiecki cautions that it is easy to make decentralization disorganisation. It is hard to make real decentralization work and keep it going. The greatest weakness is that there is no guarantee that valuable information which is uncovered in one part of the system will find its way through the rest of the system. Some valuable information never gets disseminated, making it less useful than it otherwise would be. The challenge is to find a balance between making individual knowledge globally and collectively useful, while still allowing it to remain resolutely specific and local.

The U.S intelligence community is cited as a decentralised system which did not work out well because an aggregator is missing. There is no mechanism to tap into the collective wisdom of the agency nerds, CIA spooks and FBI agents. Interestingly, Singapore government’s way of solving its traffic jam problems by having electronic smart card attached to the dashboard to make the cost of the decision to drive immediately obvious to the driver is cited as decentralization in action and it forms a mechanism to tap into the collective wisdom of its people (p. 148).

A good method for aggregating opinions (Chapter 4): Some mechanism needs to exist for turning private judgments into a collective decision. The crowd needs a way of summarizing people's opinions into one collective verdict. Surowiecki proposes that a "stock-market" type model could be used to as an aggregator to better predict outcomes of future events and/or to share information within and among members of an organization. He uses the Iowa Electronic Market (IEM) as an example of diverse people with partial knowledge accurately predicting the outcomes of elections. Best decisions are the product of disagreement and contest, not consensus or compromise.

The final chapter, Chapter 12, questions whether democracy is actually an excellent vehicle for making intelligent decisions and uncovering the truth? Surowiecki cites the political scientist James Fishkin who claims that Americans are not being given either the information or the opportunity to make intelligent political choices because they are not given enough info and the chance to talk. Fishkin believes that ordinary people are more than capable of understanding complex issues and making meaningful choices about them. He suggests deliberative polling whereby a national holiday is announced for everyone to come together to talk and cast the preliminary votes before the actual election. Surowiecki ends with that the “decisions that democracies make may not demonstrate the wisdom of the crowd. The decision to make them democratically does” (p. 271), which depicts that he believes the wisdom of crowds for democracy lies in the process, and not the decision. This does not seem to have as much flow to the rest of the book; it is also not as convincing as the other chapters.

Overall, Surowiecki makes a strong case that society should take advantage of crowd insights rather than depending on experts. In essence, he suggests, Mackay and Le Bon got it backwards. Though there is some skepticism about the premise of the book, I am open to the idea and would recommend it to anyone who is interested in being stimulated by new ideas, interested in group dynamics, interested in communicating information within organisations, or interested in trying to make better sense of daily life. This book is not a must have, but it sure challenges you to think about things in a different way. It is easy to read and entertaining, while mentally stimulating at the same time.

Contributed by Lee Yee Fuang

Comments

[Claire]

IMHO, I think the "inherent danger" that Ivan refers to occurs when people are making decisions based relationally on what others are deciding - hence the risk of information cascades.

I took a different slant from Surowiecki in that crowd value is most useful when the group is diverse and the individuals are not all connected which reduces the risk of information cascade.

[Ivan Chew]

The book "Six Degrees: The Science of a Connected Age" by Duncan J. Watts cited a social experiment by this researcher called Solomon Asch, where he proved that an individual starts having doubts when s/he is the minority in the decision. To be more exact, the experiment showed how social group pressure
lead to "conformity modification and distortion".

I found this URL which explains the experiment in detail (http://www.age-of-the-sage.org/psychology/social/asch_conformity.

So there's an inherent danger in "taking advantage of crowd insights".