Justice Powell insisted that a diverse student body is a constitutionally acceptable goal for higher education.56 The central reason is that universities should be allowed to ensure a “robust exchange of ideas,” an interest connected with the first amendment itself.57 Justice Powell acknowledged that this interest seemed strongest in the context of undergraduate education, where views are formed on a large number of topics. But even in a medical school, “the contribution of diversity is substantial.”58 A medical student having a particular background, including a particular racial background, “may bring to a professional school of medicine experiences, outlooks, and ideas that enrich the training of its student body and better equip its graduates to render with understanding their vital service to humanity.”59 Justice Powell also emphasized that doctors “serve a heterogeneous population” and suggested that graduate admissions decisions are concerned with contributions that follow formal education.60
Thus Justice Powell concluded that the crucial question was whether a race-conscious admission program, giving benefits to people because they are members of racial minority groups, was a necessary means of promoting the legitimate goal of diversity. Here he reached his famous conclusion that racial or ethnic background could be a “plus” in the admissions decision, though quotas would not be allowed.61 For Justice Powell, a legitimate admissions program should be “flexible enough to consider all pertinent elements of diversity in light of the particular qualifications of the applicant, and to place them on the same footing for consideration, although not necessarily according to them the same weight.”62 Thus it would be acceptable to promote “beneficial educational pluralism” by considering a range of factors, including “demonstrated compassion, a history of overcoming disadvantage, ability to communicate with the poor, or other qualifications deemed important.”63
Justice Powell’s arguments have carried the day, even decades later.64 At least in education, the court has held that while racial quotas are unacceptable, race can be considered as a “plus,” at least for the benefit of African American applicants. To be sure, some members of the court believe that the Constitution requires race neutrality, and it is possible that before long, the court will forbid race to be considered in any way. Should it?
My central concern here is the principal basis for Justice Powell’s conclusion: the value of ensuring a “robust exchange of ideas” in the classroom and the legitimacy of promoting racial diversity in order to ensure that exchange. To understand the contemporary relevance of Justice Powell’s opinion, it is necessary to outline the constitutional principles governing affirmative action programs. The court has settled on the view that affirmative action programs, like all other programs embodying racial discrimination, should be subject to “strict scrutiny” from courts and invalidated unless they are the least restrictive means of achieving a compelling state interest. It is also clear that past “societal discrimination,” meaning discrimination in the nation’s past, is not a legitimate basis for discrimination against whites.65 At least for the moment, it is equally clear that narrow, remedial affirmative action programs are acceptable if they are specifically designed to correct for proven past discrimination by the institution that is acting affirmatively.66
What remains unclear is when, if ever, a public institution is permitted to justify affirmative action by reference to “forward-looking” justifications, not involving a remedy for past discrimination.67 A state might, for example, try to defend affirmative action in hiring police by urging that a police force will simply be more effective if it contains African Americans among others—especially in a community that contains people of multiple races. Justice Powell was really offering a similar claim about higher education: whether or not a college or university has itself discriminated against African Americans or others, it should be permitted to discriminate in favor of them if it is doing so as a means of ensuring a “robust exchange of ideas.” But the court offered a general pronouncement about forward-looking justifications.
As we have seen, there is no doubt that universities are permitted to promote diversity and dissent by seeking a mix of faculty members and students. They can seek people with different backgrounds, different talents, and different opinions. Efforts of this kind are pervasive; this is what many admissions offices try to do. True, some serious free speech issues might be raised if an admissions office discriminates in favor of, or against, particular points of view. But even if public institutions are barred from pursuing diversity of ideas by discriminating directly against some points of view, such institutions are surely permitted, without offending the first amendment, to seek a variety of backgrounds and experiences in the hope that better discussions will result.
To be sure, race is different, and if an institution discriminates against people because of their skin color, it faces a heavy burden of justification, even if the people against whom it discriminates are white. But if Justice Powell is right, affirmative action programs can be similarly justified. The simple idea here is that racially diverse populations are likely to increase the range of thoughts and perspectives and to reduce the risks of conformity, cascades, and polarization associated with group influences. We have seen that on the judiciary, judges with diverse views can act as “whistleblowers,” correcting ill-considered views of the law. In educational institutions, a high degree of diversity, including racial diversity, might have the same effect. On some issues, a racially uniform class might polarize to an unjustified position, simply because students do not hear important perspectives.
For example, we can easily imagine all-white classrooms discussing the issue of racial profiling; the absence of racial diversity could be a serious problem. Those who have not had bad experiences, as a result of such profiling, will lack crucial information. Justice Sandra Day O’Connor offered these comments on African American justice Thurgood Marshall: “Justice Marshall brought a special perspective. . . . His was the mouth of a man who knew the anguish of the silenced and gave them a voice. . . . I have been perhaps most personally affected by Justice Marshall as a raconteur. . . . Occasionally, at Conference meetings, I still catch myself looking expectantly for his raised brow and his twinkling eye, hoping to hear, just once more, another story that would, by and by, perhaps change the way I see the world.”68
What was true for Justice O’Connor is true for white students in many educational settings. In the context of racial profiling, and in many other imaginable cases, a degree of racial diversity is likely to introduce valuable information and perspectives. These may change how the group sees the world, whether or not it leads to a different conclusion on the merits.
To say this is not to make the absurd claim that white people all agree with one another about racial profiling or that African Americans have the same experiences and opinions about that complex issue. And in light of the fact that members of all races contain people with a range of both favorable and unfavorable views about racial profiling, it would be possible to respond that any problem, if it exists, is not because the group is all white but if and because its members begin with a uniform view of racial profiling. What matters is diversity of ideas, not racial diversity. And if this is so, what, if anything, is added by promoting diversity not of views but of racial background?
The answer must be that African Americans, by virtue of their experience, are able to add something to the discussion as such. That seems far from implausible. If students need to know something about the magnitude and the experience of racial profiling, those who have been subject to such profiling will be able to offer novel insights. And if African Americans do, in fact, have an unusually high degree of hostility to racial profiling, that is by itself a point worthwhile to know and try to understand. So too if they do not show such hostility. Of course supplemental readings could be used to expose people to diverse views. The value of diversity lies not simply in learning about facts but also in seeing a range of perspectives, including the emotions attached to the
m—and in being in the actual physical presence of those who have those perspectives and perhaps cannot be easily dismissed.
These points might be used by a fair-minded institution to defend a set of policies designed to ensure a reasonable diversity of view in classroom discussions. Because of the importance of a wide range of ideas to the educational enterprise, the goal seems both legitimate and compelling. Are affirmative action programs the least restrictive means of promoting that goal? The answer depends on the nature of those programs. It is easy to imagine cautious efforts, using race as a factor, in which the “least restrictive means” test is indeed satisfied. And that point is sufficient to suggest that Justice Powell’s approach is essentially correct.
To be sure, the same arguments about the importance of diverse views might be enlisted very broadly and in circumstances that might seem less attractive. Imagine, for example, an effort by a mostly African American university to point to the need for diversity as a way of defending discrimination against African Americans and in favor of whites. Such a university might claim that it wants significant representation by whites in order to reduce the risks from group influences and to improve the quality of discussion. It does indeed follow, from what I have said thus far, that this argument is legitimate. A classroom that is entirely African American might well suffer from conformity effects and polarization, and an educational institution might want to correct the situation.
If courts should be suspicious of the argument in this context, it is because they do not trust the sincerity of those who make it. Courts might believe that the reference to diversity is actually a pretext for an illicit discriminatory motive. But it is easy to imagine cases in which diversity is the real concern and no pretext is involved. The argument I am making is narrow and modest. In some cases, racial diversity is important for improving the educational process within the relevant school. But in some cases, the claim seems extremely weak. Would a mathematics class, or a course in physics, be improved if it contained a degree of racial diversity? This seems unlikely. If courts are going to scrutinize affirmative action programs, they should not offer a blanket ruling for or against a diversity rationale in higher education. They should accept that rationale in the context of undergraduate education, but not for programs for which racial diversity is not necessary to promote a “robust exchange of ideas.” In the context of law school, the centrality of racial issues to important aspects of legal education should be enough to allow narrowly tailored affirmative action programs to survive constitutional scrutiny.
Conclusion
Conformity and Its Discontents
Human beings pay close attention to the informational and reputational signals sent by others. These signals produce conformity, even in cases in which many people have reason to believe, on the basis of their private information, that others are mistaken or worse. Informational and reputational influences also produce cascades, in which people do not rely on, and fail to disclose, the information they themselves have. Cascades and errors occur spontaneously when people take account of the decisions and statements of their predecessors. Errors are magnified when people are rewarded for conformity—and minimized when people are rewarded not for correct individual decisions but for correct decisions at the group level.
Cascades, like conformity, are not a problem in themselves. Sometimes cascades produce good outcomes, at least compared to a situation in which people rely solely on their own information. The real problem is that when cascades are occurring, people do not disclose information from which others would benefit. The result is that both individuals and private and public groups can blunder, sometimes catastrophically. Institutions involved in making, enforcing, and interpreting the law are subject to conformity and cascade effects. Government has often blundered as a result. We have seen that within courts, precedential cascades are highly likely, especially in complex areas; and in such areas, cascades tend to be both self-perpetuating and self-insulating.
The general lesson is clear. It is extremely important to devise institutions that promote disclosure of private views and private information. Institutions that instead reward conformity are prone to failure; institutions are far more likely to prosper if they create a norm of openness and dissent. The point very much bears on the risks of group polarization. Groups of like-minded people are likely to go to extremes, simply because of limited argument pools and reputational considerations. The danger is that the resulting movements in opinion will be unjustified. It is extremely important to create “circuit breakers” and to devise institutional arrangements that will serve to counteract movements that could not be supported if people had a wider range of information.
These points have implications for numerous issues in law and policy. I have focused on a small subset of those issues here. We have seen that an appreciation of social influences casts new light on the expressive function of law. Simply by virtue of what it says, and even if it is rarely enforced, law can affect the behavior of those who are unsure whether to engage in certain conduct—and also the behavior of those who are unsure whether to challenge those who engage in that conduct. Bans on smoking in public and on sexual harassment are cases in point. Law’s effectiveness, in this regard, lies in its power to give a signal about appropriate behavior, and also to dissipate pluralistic ignorance, by providing information about what other people think is appropriate behavior. A legal enactment can operate in the same fashion as Solomon Asch’s confederates and Stanley Milgram’s experimenter. Because people care about the reactions of others, law’s expressive function will be heightened if the relevant conduct is visible; bans on smoking in public places are an obvious example.
For the same reason, that function will be weakened if prospective lawbreakers live in a supportive subcommunity; consider bans on the use of narcotics. With an understanding of social influences, we can therefore make some predictions about when law is likely to be effective merely by virtue of what it says—and about when law will be ineffective unless it is accompanied by vigorous enforcement activity.
I have suggested that many of the institutions in the U.S. Constitution serve to reduce the likelihood of bad consequences from conformity, cascades, and group polarization. Such institutions increase the likelihood that important information, and alternative points of view, will receive a public airing. The system of bicameralism is the most obvious example, producing a system in which lawmaking is done by two institutions with different cultures, thus creating a potential check on unjustified movements. I have also urged that the framers’ most distinctive contribution consisted not in their endorsement of deliberative democracy, which was uncontroversial, but in their commitment to heterogeneity in government, seeing (in Alexander Hamilton’s words) the “jarring of parties” as a method for “promoting deliberation.”
More controversially, I have suggested that an understanding of social influences suggests the importance of ensuring a high degree of diversity on the federal bench. It is foolish to pretend that Republican appointees do not, as a class, differ from Democratic appointees; and we have reason to appreciate the value, on any panel, of having a potential “whistleblower,” in the form of one judge of a different party from the other two. Of course judges are rarely lawless, but a group of like-minded judges is prone to go to extremes. An appreciation of social influences on belief and behavior also supports the legitimacy of efforts to promote racial diversity in higher education, at least where such diversity is likely to improve learning.
Even if occasionally alarming, much of the behavior discussed here attests to the reasonableness and good sense of ordinary people. In the face of doubt, we do well to pay attention to the views of others. After all, they might know better than we do. It is prudent to be cautious about challenging other people, not only because they might be right but also because people do not always like to be challenged. Even in the most freedom-loving societies, people dissent at their peril. A reluctance to disagree is not merely prud
ent; it is often courteous too. But conformity creates severe dangers. Behavior that is sensible, prudent, and courteous is likely to lead individuals and societies to blunder, simply because people fail to learn about facts or opinions from which they would benefit.
It is usual to think that those who conform are serving the general interest and that dissenters are antisocial, even selfish. In a way this is true. In some settings, conformists strengthen social bonds, whereas dissenters imperil them or at least introduce tension. But in an important respect, the usual thought has things backward. Much of the time, it is in the interest of the individual to follow the crowd but in the social interest for individuals to say and do what they think best. Well-functioning institutions take steps to discourage conformity and to promote dissent, partly to protect the rights of dissenters but mostly to protect interests of their own.
Notes
Preface
1 Anna Collar provides a valuable discussion in Religious Networks in the Roman Empire (2014).
2 The classic treatment, emphasizing game theory and complementary to some of the discussion here, is Edna Ullmann-Margalit, The Emergence of Norms (1976).
3 Whitney v. California, 274 US 357, 376 (1927) (Brandeis, J., concurring).
Conformity Page 13