Risky Shifts, Cautious Shifts

The original experiments on the effects of deliberation are especially relevant to businesses and governments alike. The studies involved risk-taking behavior, with a clear finding that people initially inclined to take risks become still more risk-inclined after they deliberate with one another.⁴ If group members begin with some willingness to engage in risky behavior, groups will engage in more of that behavior as a result of group discussion. Risky decisions include taking a new job, investing in a foreign country, escaping from a prisoner-of-war camp, or running for political office. With respect to many such decisions, members of deliberating groups became significantly more risk-seeking after a brief period of exchanging views. On the basis of this evidence, the prevailing wisdom was that deliberation produced a systematic risky shift.

But later studies drew this conclusion into question—and also created a big puzzle. On many of the same questions on which Americans displayed a risky shift, Taiwanese participants showed a cautious shift. Deliberation led citizens of Taiwan to become a lot less risk-inclined than they were before they started to talk. And it turned out that among American participants, deliberation sometimes produced a cautious shift, as risk-averse people became more averse to certain risks after they talked with one another, depending on the action they were contemplating. The principal examples of cautious shifts were decisions about whether to marry and whether to board a plane despite severe abdominal pain that would possibly require medical attention. In these cases, the members of deliberating groups, even in the United States, shifted not toward risk but toward greater caution.⁵ So much for the idea of a consistent one-directional risky shift!

What explains these unruly findings? A straightforward interpretation reconciles them: the predeliberation median is the best predictor of the direction of the group’s shift.⁶ When group members are initially disposed toward risk taking, a risky shift is likely. Where members are initially disposed toward caution, a cautious shift is likely. Thus, for example, the striking difference between American and Taiwanese subjects is a product not of cultural habits or a different effect from group deliberation, but of a simple difference in where group members started—more precisely, in the predeliberation medians of the different groups on the key questions. Whereas the Americans started out risk-inclined, the Taiwanese started out risk-averse, and that simple fact explained the different directions of the two shifts. The risky shift and the cautious shift are both subsumed under a single rubric: group polarization.

In a finding of special importance for business, group polarization occurs both for matters of fact and for issues of value. Suppose that the question is whether a product will sell a certain number of units in Europe in the next year. If so, group polarization will not be easy to test, simply because the answer is either yes or no, and it is not simple to demonstrate a shift to greater extremism in yes-or-no answers. But suppose that people are asked, on a bounded scale of zero to eight, how likely it is that a product will sell a certain number of units in Europe in the next year, with zero indicating “zero probability,” eight indicating “absolutely certain,” seven indicating “overwhelmingly likely,” six “more probable than not,” and five “fifty-fifty.”

In that event, the answers from a deliberating group will tend to reveal group polarization, as people move toward more-extreme points on the scale depending on their group’s initial median point. If the predeliberation median is six, the group judgment will usually be seven; if the predeliberation median is three, the group judgment will usually be two.⁷

Note here that even federal judges—experts in the law and supposedly neutral—are highly susceptible to group polarization, as both Democratic and Republican appointees show far more ideological voting patterns when sitting with other judges appointed by a president of the same political party.⁸ On a three-judge panel, consisting solely of Democratic appointees, Democratic appointees show very liberal voting patterns in cases involving discrimination on the basis of sex and sexual orientation, environmental protection, and the rights of workers—far more liberal patterns than they show when at least one Republican appointee is on a three-judge panel. The pattern is identical for Republican appointees, who show highly conservative voting patterns on all-Republican panels. You can think of three-judge panels as teams, and judicial voting is affected by the attitudinal composition of the team.

Here’s a vivid way to see the point. If you want to know how an appellate judge will vote in an ideologically contested case, you might want to find out whether she was appointed by a Republican or a Democratic president. It’s a pretty good predictor. But in many areas of the law, here’s an even better predictor of how a given judge will vote: Who appointed the other two judges on the panel?

Juries display group polarization as well. In particular, their punitive-damage awards tend to be far higher than the preferred award of the median member, before deliberation. The result of jury deliberation is to produce an increase in extremism, in the particular form of higher awards.⁹ There is a lesson here about punishment judgments in general. If group members begin with an inclination to be punitive, there is a good chance that the group will end up more punitive still.

An Experiment in Colorado

To cast light on these questions, the two of us created (along with our friend and colleague David Schkade) a distinctive experiment in group deliberation—one that, we believe, accurately reflects much deliberation in the real world. In this experiment, we recruited citizens from two cities and assembled them into small groups (usually six people), all from the same city. The groups were asked to deliberate on three of the most contested issues of the time: climate change, affirmative action, and same-sex civil unions. The two cities were Boulder, known by its voting patterns to be predominantly liberal (aka “The People’s Republic of Boulder”), and Colorado Springs, known by its voting patterns to be predominantly conservative (and nicknamed “The Citadel”). We did a reality check on the participants before the experiment started, ensuring that the Boulder residents were in fact left of center and that the Colorado residents were in fact right of center. The citizens were first asked to record their views individually and anonymously—and then to deliberate together in an effort to reach a group decision. After deliberation, the individual participants were asked to record their postdeliberation views individually and anonymously.

There were three effects from group deliberation:

1. People from Boulder became a lot more liberal on all three issues. By contrast, people from Colorado Springs became a lot more conservative. The effect of group deliberation was to shift individual opinions toward extremism. This shift happened in two ways. First, group “verdicts” on climate change, affirmative action, and same-sex civil unions were more extreme than the predeliberation average of the group members. Second, the anonymous views of individual members became more extreme, after deliberation, than were their anonymous views before the participants started to talk. There’s a big lesson here. Group deliberation often makes not only groups but also individuals more extreme, so much so that they will state more-extreme views privately and anonymously.
2. Deliberation increased consensus within groups. Before people started to deliberate, many of the groups showed a lot of internal diversity, in the sense that there was considerable divergence in people’s individual opinions. Sure, people in Boulder are generally liberal, but in our groups, they did not always agree with one another on the particular issues that we selected. Discussion brought liberals into line with each other, and the same thing happened with conservatives. As a result of a brief period of discussion, group members showed a lot more agreement and less variation in their anonymous postdeliberation expressions of their private views.
3. Deliberation sharply increased the disparities between the views of the largely liberal citizens of Boulder and the largely conservative citizens of Colorado Springs. Before deliberation, there was considerable overlap between many individuals in the two cities. After deliberation, the overlap was a lot smaller. Liberals and conservatives became more sharply divided.

The implications are clear. As a general rule, deliberating groups tend to end up adopting a more extreme position in line with their inclinations before they started to talk, and a major effect of deliberation is to squelch internal diversity—and thus to push different groups apart.

Why Polarization?

Why does group polarization occur? There are three principal reasons.¹⁰

The first and most important explanation involves the now-familiar idea of informational influence, but with a few twists. Group members pay attention to the arguments made by other group members. The arguments in any group with some initial predisposition will inevitably be skewed in the direction of that predisposition. Suppose, for example, that most group members begin by thinking that a new venture is likely to succeed. If so, there will be a lot of arguments to that effect. As a statistical matter, the arguments favoring the initial position will be more numerous than those pointing in the other direction. Individuals will have thought of or heard of some, but not all, of the arguments that emerge from group deliberation. As a result of the arguments that are made by group members, deliberation will naturally lead people toward a more extreme point in line with what group members initially believed.

The second explanation, also familiar, involves social influences. As we have seen, people want to be perceived favorably by other group members. People’s publicly stated views are sometimes a function of how they want to present themselves. Once they hear what others believe, some group members will adjust their positions at least slightly in the direction of the dominant position to preserve their preferred self-presentation. They shift accordingly.¹¹ If a group’s leader or if most group members favor a particular course of action, those who challenge that view do so at their own risk. Here is yet another reason that wise leaders often speak tentatively or not at all, inviting group members to speak their mind.

The third explanation of group polarization is more subtle. It stresses the close links among three factors: confidence, extremism, and corroboration by others.¹² When people lack confidence, they tend to be tentative and therefore moderate, knowing that their own views may be wrong. The great American judge Learned Hand once said that “the spirit of liberty is that spirit which is not too sure that it is right.” Tentative people respect the spirit of liberty. But as people gain confidence, they usually become more extreme in their beliefs. The reason is that a significant moderating factor—their own uncertainty about whether they are right—has been eliminated. With respect to group polarization, the key point is that agreement from others tends to increase confidence and, through that route, to increase extremism.

It is partly for this reason, then, like-minded people, having deliberated with one another, become surer that they are right and thus more extreme. In many contexts, people’s opinions turn extreme simply because their views have been corroborated and because they become more confident after learning that others share their views.¹³ Groups can badly blunder in this way. They might conclude that a new hire is terrific, that a proposed investment is bound to fail, or that a new policy is going to work really well. They end up disappointed, regretful, and often surprised, above all because group dynamics increased their confidence and their conviction.

In Groups, Out Groups, and Shared Identity

A great deal of work suggests that group polarization is more likely and is heightened when people have a sense of shared identity and belong to a tight-knit group or club. This point casts further light on the reasons for group polarization.¹⁴

In short, people may become polarized because they are attempting to conform to the position that they see as typical within their own group. If their group’s identity is made especially salient, the in-group norms are likely to become more extreme.¹⁵ (So if a leader says, “We are all [name a religion],” or “We all built this company together,” expect polarization.) And if arguments come from a member of an in-group, they are especially likely to be persuasive. Such arguments just feel right.

People will also be especially fearful of the social pressures that come from rejecting what an in-group member has to say. In a family, a religious organization, a small business, or a company with a sense of identity and hence special insiders, group polarization is particularly likely unless people take special steps to combat it. There is a lesson here for social networks, including those online, where shared identities can be breeding grounds for both confidence and extremism. As a corrective, Alex Pentland of MIT’s Media Lab points to the importance not only of engagement, which produces high levels of communication, but also of an appetite for exploration, which promotes seeking out and sharing new information, new opinions, and new solutions.¹⁶

It should not be a big surprise that the views of out-group members often have little force and might be irrelevant or even suggest that the opposite is true.¹⁷ Should you really listen to someone who belongs to a group that you believe to be silly, bad, or systematically wrong? If your adversaries or competitors say something, you might be more inclined to believe that just the opposite is true.

Consider a test of whether apparently credible media corrections alter the belief, supported and pressed by former Alaska governor Sarah Palin, that the Affordable Care Act would create “death panels.”¹⁸ Among those who viewed Palin favorably but had limited political knowledge, the correction turned out to succeed in changing beliefs. The correction also succeeded among those who viewed Palin unfavorably. But the correction actually backfired among Palin supporters with a high degree of political knowledge. After receiving the correction, they became more likely to believe that the Affordable Care Act contained death panels. Ironically, the correction intensified their original belief. The study suggests that if members of an out-group support some proposition, their very support might entrench the preexisting beliefs of the in-group.

When out-group members speak out, the reputational pressure is not likely to be strong: Do you really have to worry about the social consequences of rejecting the views of someone usually thought to be wrongheaded, or badly motivated, by people like you? If a self-identified conservative hears arguments from someone who is known to be left of center, those arguments are likely to fall on deaf ears. The clear lesson is that when a group is highly cohesive, and when its members are closely identified with it, polarization is especially likely—and likely to be especially large.

We have clear evidence of this phenomenon among groups of investors; tightly knit investment clubs, unified by social ties, lose a lot of money.¹⁹ Investment clubs do a lot better when their members do not socialize and instead see each other as colleagues rather than friends.

Accuracy

Does group polarization lead to accurate or inaccurate answers? Do deliberating groups err when they polarize?

No simple answer would make sense. Everything depends on the relationship between the correct answer and the group’s predeliberation tendencies. If the group is leaning toward the right answer, polarization will lead its members directly to the truth. It follows that group polarization might not produce errors or failures at all.

But there are no guarantees here. As a result of group influences, some people will fail to disclose what they know. When individuals are leaning in a direction that is mistaken, the mistake will be amplified by group deliberation. We have encountered troubling examples. When individuals are inclined to commit the planning fallacy, groups are even more likely to commit that fallacy. And when most people are prone to make logical errors, group processes lead to more errors rather than fewer. No one profits from the planning fallacy or logical errors, and group polarization helps to produce a lot of fallacies and plenty of errors.