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

In Chapter 8, Surowiecki discusses how cooperation allows the scientific community to regularly solve great world problems. It was the collaborative work of 11 research laboratories across France, Germany, the Netherlands, Japan, the United States, Hong Kong, Singapore, Canada, the United Kingdom, and China that allowed for the discovery of the specific coronavirus that caused severe acute respiratory syndrome (SARS). Although the project was initiated by the World Health Organization, the individual laboratories were allowed to search for answers independently. The only rule was that they would all take part in daily teleconferences to share their work and discuss future plans. The information available to them, such as virus samples, were shared across all members, allowing for them to work on the same evidence simultaneously.

This collaborative system, without the pressure of top-down direction, allowed the scientists working at the laboratories to use their own private information and personal judgment to search in different directions. This maximized diversity and exhausted an extensive range of possibilities, allowing the labs to isolate and conclude that the coronavirus they suspected was the cause of SARS was indeed the culprit. They arrived at this conclusion within a month of the outbreak.

Although popular imagination still thinks of science as the product of geniuses working alone in a lab, the modern field of science is collaborative, with papers often co-authored by more than 10 people. This collaboration allows science to advance quickly because it promotes a division of cognitive labor. On the micro scale, individual scientists become increasingly knowledgeable in a narrow field where they gain tacit knowledge and operate sophisticated machinery; on the macro scale, this diversifies expertise across the whole field of science. This division guarantees diversity in people’s expertise and perspective, and when their knowledge is aggregated, it makes everyone more productive.

Collaboration is possible because the scientific community’s organization encourages it: The system promotes open access to information, which benefits everyone. Similarly, scientists who make discoveries are willing to publish their findings because it is how they gain public recognition. Most importantly, their legitimacy must first be recognized from peers within the field, meaning that they must be accepted by their competitors. This encourages scientists to remain team players.

There are two major challenges to this system. The first is the growing commercialization of the field. Since great scientific discoveries, such as the creation of a new effective drug, can generate great profit, there are economic forces that can skew the direction of scientific research. The second challenge is reputation bias. Although scientists mostly work collaboratively, statistically, most published papers are not widely read, and the few that are widely read overshadow the rest. Similarly, scientists who are reputable in the field will often find that their collaborative works are more widely cited than a comparable study by an unknown peer. Surowiecki cautions that, though reputation does confer a certain sense of proven trustworthiness, it must not develop into a scientific hierarchy.

Small groups are often inefficient at decision-making, but when used well, they can become greater than the sum of their parts. Surowiecki uses the final flight of the space shuttle Columbia as an example of the mistakes to avoid when wanting to tap into the wisdom of small groups.

The Columbia’s 28th launch was marred by a piece of foam detaching from its left bipod and striking its fuel tank on its ascent into orbit. The Debris Assessment Team immediately announced their concerns to NASA, but the Mission Management Team (MMT), in charge of coordinating the shuttle’s voyage, dismissed them. The Columbia mission ended in a disastrous crash that killed its crew upon returning to orbit.

The Columbia Accident Investigation Board that reviewed their team meetings found that the leaders of the MMT had come to the discussion convinced that there was no problem to fix and that even if there was, little could be done now. They asked few questions and encouraged only consenting voices, guiding the flow of conversation to agree with their preconceived conclusion.

As shown in Chapter 2, small groups can become more than the sum of their parts. However, this is hard to achieve, and most of the time, members of a group subtract from its value. This is the case of the Columbia incident: A group of individually intelligent engineers came to the wrong conclusion by parroting each other.

First, the group identity overshadowed individual judgment. Members of the MMT team thought of themselves as part of a group, which made it incredibly difficult to detach their judgment from the group identity. Second, they came to the discussion table having already taken a position and believing that they understood more than they did. This meant that their decisions were verdict based. They did not look at evidence and arrive at a conclusion; they favored a position from the start and weighed evidence in favor of their position while dismissing the contrary. This fallacy is called confirmation bias.

Small groups that develop their own identities tend to favor verdict-based judgment, and their efforts are pooled into convincing others that their judgment is correct and that investigation is pointless. This is dangerous because it emphasizes consensus over dissent, sometimes exerting social pressure to encourage individual members to identify with the group rather than rely on their private judgment. In doing so, small groups prevent a diversity of options from being explored and take solace in dismissing members’ doubts in favor of maintaining the illusion of certainty.

In the case of the Columbia incident, there were alternatives to save the crew members’ lives, but they were not explored due to this tendency of small groups to converge opinions around a preconceived notion. Minority opinion was completely absent at the January 24th meeting.

Another factor that prevents small groups from being wise is group polarization, a phenomenon that is not yet well understood but is proven to exist, in which deliberation radicalizes rather than moderates people’s opinions. Group polarization is speculated to arise because people compare themselves to others when they debate, meaning that those who are moderate in their opinion want to retain their relative position in the group, so if everyone moves toward one radical end, they will move the same relative distance to remain in the center. Another important factor that contributes to group polarization is information cascades. When convinced and vocal members of the group form the majority, those who are uncertain are more likely to be swayed in that direction because it is easy and safe to imitate the group. Finally, people who have not held decision-making positions often defer to the opinions of those who appear more confident even if they have more expertise, whereas people who have held decision-making positions tend to be more confident even in fields they have little understanding of.

To ensure that small groups are efficient and stronger than the sum of their parts, they must become depolarized. This can be achieved by encouraging dissenting opinions and avoiding hierarchy. Contrary to popular belief, Surowiecki argues that group decisions are not inherently inefficient because effective aggregation of the independent opinions of its members can often be more accurate than the conclusion that any one individual in the group can arrive at; small groups can foster collective wisdom.