How Change Happens

by Cass R Sunstein

  26. There are many accounts. For especially good ones, see Kuran, supra note 1; D. Garth Taylor, Pluralistic Ignorance and the Spiral of Silence: A Formal Analysis, 46 Pub. Opinion Q. 311 (1982). A valuable account, with special reference to law, is Richard McAdams, The Expressive Powers of Law 136–162 (2015).

  27. See id.; Sushil Bikhshandani et al., A Theory of Fads, Fashion, Custom, and Cultural Change as Informational Cascades, 100 J. Polit. Econ. 992 (1992).

  28. An intriguing wrinkle is that when a cascade gets going, people might underrate the extent to which those who join it are reacting to the signals of others, and not their own private signals. For that reason, they might see the cascade as containing far more informational content than it actually does. See Erik Eyster & Matthew Rabin, Naïve Herding in Rich-Information Settings, 2 Am. Econ. J.: Microeconomics 221 (2010); Eric Eyster et al., An Experiment on Social Mislearning (2015), https://papers.ssrn.com/sol3/papers.cfm?abstract_id=2704746. Norm entrepreneurs have a strong interest in promoting this mistake.

  29. See Amos Tversky & Daniel Kahneman, Availability: A Heuristic for Judging Frequency and Probability, 5 Cognitive Psychol. 207 (1973).

  30. See Timur Kuran & Cass R. Sunstein, Availability Cascades and Risk Regulation, 51 Stan. L. Rev. 683 (1999).

  31. Gordon Wood, The Radicalism of the American Revolution 29 (rev. ed. 1998).

  32. Id. at 29–30.

  33. Id. at 169.

  34. Id.

  35. Thomas Paine, “Letter to the Abbe Raynal,” in Life and Writings of Thomas Paine, 242 (Daniel Edwin Wheeler ed. 1908). Quoted in Gordon Wood, The Creation of the American Republic, 1776–1787 68 (rev. ed. 1998).

  36. See Lawrence Lessig, The Regulation of Social Meaning, 62 U. Chi. L. Rev. 943 (1995). See also Richard H. McAdams, Cooperation and Conflict: The Economics of Group Status Production and Race Discrimination, 108 Harv. L. Rev. 1003, 1065–1085 (1995).

  37. Lessig, supra note 36, at 966.

  38. See Leonardo Bursztyn et al., From Extreme to Mainstream: How Social Norms Unravel (2017), http://www.nber.org/papers/w23415.

  39. Id.

  40. Helpful discussion can be found in Timur Kuran, Ethnic Norms and Their Transformation through Reputational Cascades, 27 J. Legal Stud. 623 (1998).

  41. George Orwell, Nineteen Eighty-Four (1949).

  2

  The Law of Group Polarization

  Consider the following:

  A group of citizens is concerned about immigration. The citizens think that illegal immigrants are committing serious crimes. They are also fearful that legal immigration has “gone too far” and that it is “taking our jobs.” The group decides to meet every two weeks to focus on common concerns. After a year, is it possible to say what its members are likely to think?

  After a nationally publicized shooting at a high school, a group of people in the community—most of them tentatively in favor of greater gun control—come together to discuss the possibility of imposing new gun control measures. What, if anything, will happen to individual views as a result of this discussion?

  Affirmative action is under attack in the state of Texas. Most professors at a particular branch of the University of Texas are inclined to be supportive of affirmative action; they meet to exchange views and to plan further action, if necessary. What are these professors likely to think and to do after they talk?

  My principal purpose in this chapter is to investigate a striking statistical regularity—group polarization—and to relate that phenomenon to underlying questions about the role of deliberation in the “public sphere” of a heterogeneous democracy. In brief, group polarization means that members of a deliberating group predictably move toward a more extreme point in the direction indicated by the members’ predeliberation tendencies. “Like polarized molecules, group members become even more aligned in the direction they were already tending.”1 Thus, for example, members of the first group will become deeply hostile to immigration; members of the second group will probably end up favoring gun control quite enthusiastically; members of the third group will become more firmly committed to affirmative action.

  Notably, groups consisting of individuals with extremist tendencies are more likely to shift, and likely to shift more; the same is true for groups with some kind of salient shared identity (like conservatives, socialists, Catholics, Jews, and lawyers, but unlike jurors and experimental subjects). When like-minded people are participating in “iterated polarization games”—when they meet regularly, without sustained exposure to competing views—extreme movements are all the more likely.

  Three principal mechanisms underlie group polarization. The first emphasizes the role of information—in particular, the limited “argument pools” within any group and the directions in which those limited pools lead group members. The second points to social influences on behavior and in particular to people’s desire to maintain their reputation and their self-conception. The third emphasizes the relationships among confidence, corroboration, and extremism. The basic idea is that when people find their views corroborated by others, they become confident—and thus more extreme. Radical movements can be fueled in that way.

  An understanding of the three mechanisms provides many insights into social change and democratic institutions. It illuminates a great deal about likely processes within political parties, legislatures, and multimember courts—not to mention ethnic and religious groups, extremist organizations, terrorists, criminal conspiracies, student associations, faculties, institutions engaged in feuds or “turf battles,” workplaces, and families.

  One of my largest purposes is to evaluate the social role of enclave deliberation, understood as deliberation within small or not-so-small groups of like-minded people. I suggest that enclave deliberation is, simultaneously, a potential danger to social stability, a source of social fragmentation, and a safeguard against social injustice and unreasonableness. As we will see, group polarization helps explain an old point, with clear foundations in constitutional law in many nations, to the effect that social homogeneity can be quite damaging to good deliberation. An understanding of group polarization thus illuminates social practices designed to reduce the risks of deliberation limited to like-minded people.

  How and Why Groups Polarize

  Group polarization is among the most robust patterns found in deliberating bodies, and it has been found in many diverse tasks. As a result, groups often make more extreme decisions than would the typical or average individual in the group (where extreme is defined solely internally, by reference to the group’s initial dispositions). There is a clear relationship between group polarization and cascade effects. As we will see, the former, like the latter, has a great deal to do with both informational and reputational influences. A key difference is that group polarization involves the effects of deliberation.

  Although standard within psychology, the term group polarization is somewhat misleading. It is not meant to suggest that group members will shift to two poles. Instead, the term refers to a predictable shift within a group discussing a case or problem. As the shift occurs, groups and group members move and coalesce not toward the middle of antecedent dispositions, but toward a more extreme position in the direction indicated by those dispositions. The effect of deliberation is both to decrease variance among group members, as individual differences diminish, and to produce convergence on a relatively more extreme point among predeliberation judgments.

  Consider a few examples of the basic phenomenon, which has been found in over a dozen nations:2

  1. A group of moderately profeminist women will become more strongly profeminist after discussion.3 (This finding is obviously relevant to the dynamics of #MeToo.)

  2. After discussion, citizens of France become more critical of the United States and its intentions with respect to economic aid.4 (This finding is obviously relevant to skepticism about the United States in various nations.)

  3. After discussion, whites
predisposed to show racial prejudice offer more negative responses to the question whether white racism is responsible for conditions faced by African-Americans in American cities.5 (This finding is obviously relevant to increases in racial antagonism.)

  4. After discussion, whites predisposed not to show racial prejudice offer more positive responses to the same question.6 (This finding is obviously relevant to the softening of ethnic and racial divisions.)

  As statistical regularities, it follows, for example, that that those moderately critical of an ongoing war effort will sharply oppose the war after discussion; that those who believe that climate change is a serious problem are likely to hold that belief with considerable confidence after discussion; that people tending to believe in the inferiority of a certain racial group will become more entrenched in this belief as a result of discussion.

  The phenomenon of group polarization has conspicuous relevance to social media and the communications market—where groups with distinctive views and identities often engage in within-group discussion—and also to the operation of many deliberating bodies of relevance to law and politics, including legislatures, commissions, multimember courts, and juries. I will return to this point shortly; for now, notice a few obvious possibilities. If the public is sharply divided, and if different groups design their own preferred communications packages, the consequence may be further division, as group members move one another toward more extreme points in line with their initial tendencies. Different deliberating groups, each consisting of like-minded people, may be driven increasingly far apart, simply because most of their discussions are with one another.

  In a similar vein, members of a political party, or of the principal political parties, may polarize as a result of internal discussions; party-line voting is sometimes explicable partly on this ground. Extremist groups will often become more extreme. A set of judges with similar predilections on a three-judge panel may well produce a more extreme ruling than any individual member would write if he were judging on his own. As we will soon see, the largest group polarization typically occurs with individuals already inclined toward extremes.

  Risky Shifts and Cautious Shifts

  Group polarization was first found in a series of experiments involving risk-taking decisions. Before 1961, conventional wisdom held that as compared with the individuals who compose it, a group of decision-makers—for example, a committee or board—would be likely to favor a compromise and thus to avoid risks. But the relevant experiments, originally conducted by James Stoner, found otherwise; they identified what has become known as the risky shift.7 Deliberation tended to shift group members in the direction of greater risk-taking, and deliberating groups, asked to reach a unanimous decision, were generally more risk-inclined—sometimes far more risk-inclined—than the mean individual member predeliberation.

  It is important to distinguish at this point between two aspects of these findings, both of relevance to law and policy. The first involves the movement of deliberating groups, when a group decision is necessary, toward the group’s extreme end; this is sometimes described as a choice shift. This means that if a group decision is required, the group will tend toward an extreme point, understood against the background set by the original distribution of individual views. In other words, the group will be more extreme than the median or average member. Undoubtedly, the group’s decision rule will matter here; majority rule might produce a different outcome from a requirement of unanimity. If those with the most extreme views are least tractable and most confident, a unanimity requirement might produce a shift toward the most extreme points.

  The second involves the movement of (even private) individual judgments as a result of group influence; this is the standard meaning of the term group polarization. To the extent that the private judgments of individuals are moved by discussion, the movement will be toward a more extreme point in the direction set by the original distribution of views. It is possible to have one kind of movement without the other, though ordinarily the two accompany one another.

  A possible (and contemporaneous) reading of Stoner’s early studies would be that group dynamics usually move people—both groups and individuals within them—in the direction of greater risk-taking. But this conclusion would be much too simple. Later studies showed that under certain conditions, it was possible, even easy, to induce a cautious shift as well. Indeed, certain problems reliably produced cautious shifts.8 The principal examples involved the decision whether to marry and the decision whether to board a plane despite severe abdominal pain possibly requiring medical attention. In these cases, deliberating groups moved toward caution, as did the members who composed them. Burglars, in fact, show cautious shifts in discussions with one another, though when they work together, the tendency is toward greater risk-taking.9

  Later researchers noticed that in Stoner’s original data, the largest risky shifts could be found when group members “had a quite extreme risky initial position,” in the sense that the predeliberation votes were weighted toward the risky end, whereas the items “that shifted a little or not at all started out near the middle of the scale.”10 Thus the direction of the shift seemed to turn on the location of the original disposition, and the size of the shift depended on the extremeness of that original disposition. A group of very cautious individuals would produce a significant shift toward greater caution; a group of individuals inclined toward risk-taking would produce a significant shift toward greater risk-taking; and groups of individuals in the middle would produce smaller shifts in the direction indicated by their original disposition.

  Similar results have been found in many contexts with relevance to law and democracy, involving, for example, questions about economic aid, architecture, political leaders, race, feminism, and judgments of guilt or innocence. Polarization has been found for questions of obscure fact (e.g., how far Sodom on the Dead Sea is below sea level), as well as for evaluative questions, including political and legal issues and even the attractiveness of people in slides.

  The Outrage Heuristic

  A number of years ago, I was involved in a series of studies of outrage, punitive intentions, and monetary punishments. Our basic finding was that when ordinary people are thinking about how much to punish people, they use the outrage heuristic.11 They begin by deciding how outrageous the underlying conduct was, and their judgments about punishment build on that decision. We found that people’s outrage judgments, on a bounded numerical scale, almost exactly predicted their punitive intentions on the same scale. That means that people are intuitive retributivists. They believe that people should be punished for wrongdoing, as a way of reflecting the outrage of the community. Unless prompted, they do not think about optimal deterrence (and even when prompted, they resist the idea).

  One of our studies tested the effects of deliberation on both punitive intentions and monetary judgments.12 The study involved about three thousand jury-eligible citizens; its major purpose was to determine how individuals would be influenced by seeing and discussing the punitive intentions of others. Our central goal was to explore how social interactions heighten outrage.

  People initially were asked to record their individual judgments privately, on a bounded scale, and then asked to join six-member groups to generate unanimous “punishment verdicts.” Subjects were asked to record, in advance of deliberation, a “punishment judgment” on a scale of 0 to 8, where 0 indicated that the defendant should not be punished at all and 8 indicated that the defendant should be punished extremely severely. (Recall that outrage judgments on such scales are mirrored by punishment judgments, so we were essentially measuring outrage.) After the individual judgments were recorded, jurors were asked to deliberate to reach a unanimous punishment verdict. It would be reasonable to predict that the verdicts of juries would be the median of punishment judgments of jurors—but that prediction would be badly wrong.

  The finding that I want to emphasize here is that deliberation made the lower punishment
ratings decrease when compared to the median of predeliberation judgments of individual jurors—whereas deliberation made the higher punishment ratings increase when compared to that same median. When the individual jurors favored little punishment, the group showed a leniency shift, meaning a rating that was systematically lower than the median predeliberation rating of individual members. This means that when people began with low levels of outrage, deliberation produced lower levels still. But when individual jurors favored strong punishment, the group as a whole produced a severity shift, meaning a rating that was systematically higher than the median predeliberation rating of individual members. In groups, outrage grows—a reflection of group polarization in action.

  Mechanisms

  There are three main explanations for group polarization.13 Significant support has been found for all of them.

  Information

  The first explanation, emphasizing the role of information, starts with a simple claim: any individual’s position will be affected by what information she ends up hearing, and by which arguments presented within the group seem most convincing. People’s positions therefore move to fit with the information and arguments shared within the group, taken as a whole. Because (and this is the critical point) a group whose members are already inclined in a certain direction will have a disproportionate number of arguments supporting that same direction, the result of discussion will be to move individuals further in the direction of their initial inclinations.

  The key is the existence of limited information and a limited argument pool, one that is skewed (speaking purely descriptively) in a particular direction. Members of a group will have thought of some, but not all, of the arguments that justify their initial inclination. In discussion, the arguments of different people may be stated and heard, but the total argument pool will be tilted in one direction or another, depending on the predispositions of the people who compose the group. Hence there will be a shift in the direction of the original tilt.