公 法 评 论 惟愿公平如大水滚滚,使公义如江河滔滔
et revelabitur quasi aqua iudicium et iustitia quasi torrens fortis




Sanford Levinson[*]


I am deeply grateful to have been asked to deliver the 1999 Roberts Lecture. It is both an immense honor and, of course, an immense responsibility even to attempt to live up to the distinguished speakers who have preceded me. At the very least, though, I believe that I have chosen a topic commensurate with the importance of the occasion. As Princeton political theorist Stephen Macedo says at the very beginning of his important book, "[d]iversity is the great issue of our time: nationalism, religious sectarianism; a heightened consciousness of gender, race, and ethnicity; a greater assertiveness with respect to sexual orientation; and a reassertion of the religious voice in the public square are but a few of the forms of particularity"[1] that we confront daily under the general rubric of "diversity."

My own interest in the topic may truly be described as overdetermined. As a teacher of constitutional law for a full quarter-century, I have invariably assigned and discussed various cases and materials involving affirmative action or the toleration that is due particular religious sects whose behavioral norms are radically at odds with those of most of their fellow Americans. Indeed, I have offered seminars on "multiculturalism" and the Constitution. I might add that I also address some of these subjects in a second-year course that I often teach on the particular role that the Constitution plays in structuring the contemporary American welfare state, for many of the constitutional struggles about diversity are strongly interlaced with the realities of the modern welfare state. Were there, for example, no state universities providing education to its students at significantly below-market cost, then many of the most volatile debates about affirmative action--i.e., the use of racial or ethnic preferences to select those who shall receive such benefits--would be off the table. Similarly, as Chief Justice Rehnquist has argued, we would not be debating whether the state must pay unemployment compensation to someone who has left a job for religious reasons--another issue that tests one's views on the practical meaning of diversity and multiculturalism--if the modern state were not, in fact, supplying such compensation.[2] Thus my teaching interests alone could adequately explain my interest in the topic of "diversity."

It is scarcely irrelevant, though, that my home setting, the University of Texas Law School, has the unique distinction (if that is the right word) of being the defendant in two of the three most important cases involving the particular topic that probably most often comes to mind upon hearing the word "diversity"--the use of racial classifications in higher education. What to many of my colleagues elsewhere is quite literally only an "academic interest" has, for us at the University of Texas, become extraordinarily important in shaping the circumstances of our daily lives as teachers. Whatever one's teaching interests, it is impossible for anyone at the University of Texas to avoid grappling with the implications of diversity--or its absence--as a reality of contemporary American life.


I begin with the three cases to which I have previously alluded, not because I intend in this discussion to break new doctrinal ground or, for that matter, even offer suitably elaborate explication of contemporary doctrine, but, rather, simply to set the stage for the more theoretical issues that are my principal focus. The first case is Sweatt v. Painter,[3] in which Heman Sweatt, who was denied admission to the University of Texas Law School for no reason other than his being African-American, successfully challenged the school's admissions policy on constitutional grounds. Texas had, in a desperate attempt to come under the "separate but equal" doctrine that had not yet been invalidated in Brown v. Board of Education,[4] established a so-called "downtown law school" that non-whites could attend. The Supreme Court aptly cut through any arguments that this facility was in fact equal because courses were taught by University of Texas faculty in classes with a better ratio of students to teachers than existed across town at the "real" University of Texas. As Chief Justice Vinson wrote, "[t]he law school, the proving ground for legal learning and practice, cannot be effective in isolation from the individuals and institutions with which the law interacts."[5] He noted that Texas's policy excluded African-Americans from contact with "the racial groups which number 85% of the population of the State and include most of the lawyers, witnesses, jurors, judges and other officials with whom petitioner will inevitably be dealing when he becomes a member of the Texas Bar."[6]

Although the word was not then in common use, I do not believe it is an undue stretch to interpret the Court as pointing out that legal education, practically speaking, demands that students be exposed to the diversity of groups within the state if they are to be effectively prepared for the various tasks of the practicing lawyer. Although Vinson made no argument that white students were significantly harmed by being deprived of access to the remaining fifteen percent of the population, it seems impossible to believe that the Court then, or anyone now, would question the presence of such harm, even if it was, as a practical matter, far less damaging to white students' future effective ability to practice law than to African-American law students deprived of an integrated educational setting.

The second great case is, of course, Regents of the University of California v. Bakke,[7] which tested the legitimacy of an admissions program established by the Regents in regard to the medical school at the University of California Davis. The Court was bitterly divided and incapable of producing a majority opinion, as has often been true of cases involving racial preferences. Four "liberal" justices would have upheld the Davis program, which set aside sixteen places to which only members of certain racial and ethnic minorities--in particular, "Blacks, Chicanos, Asians, and American Indians"--could apply.[8] Four others would have flatly rejected it, though they relied on a federal statute forbidding the taking of race into account rather than on the Constitution itself.[9] The "swing" opinion was that of Justice Powell.[10] Like the four "liberals," Powell held that the statute meant only that race could not be taken into account in any way that would violate the Equal Protection Clause of the Fourteenth Amendment.[11] One must, therefore, determine not whether California had taken race or ethnicity into account, which it obviously had, but, rather, whether it had sufficiently good (and constitutional) reasons for doing so. Although Powell agreed with the "conservatives" that the particular program was indeed illegal because it operated as a hard-and-fast quota, he nonetheless agreed with the "liberals" that race and ethnicity could be taken into account by universities in the admissions process so long as it wasn't part of a process that included rigid quotas (as distinguished from "goals").[12]

Most importantly, Powell justified the possibility of racial or ethnic preferences on the grounds that they represented a reasonable way to achieve diversity within a student body, a goal that he thought legitimate.[13] Earlier in his opinion, he had explicitly rejected the legitimacy of such preferences as a way, for example, of responding to (and thus seeking to remedy) the past history of American racial discrimination[14] or of providing "role models" of achievement by members of the benefited minorities that would, presumably, both inspire others of their own group and serve to dispel invidious stereotypes on the part of the majority population.[15]

In discussing the diversity rationale, Powell quoted, with seeming endorsement, a Harvard statement about its own admissions criteria. The drive to produce "diversity" within the student body had led the Harvard admissions committee to pay special (and, presumably, favorable) attention to the race of applicants. "`[T]he race of an applicant may tip the balance in his favor just as geographic origin or a life spent on a farm may tip the balance in other candidates' cases.'"[16] Furthermore, Harvard explained that the desire to "provide a truly heterogeneous environment that reflects the rich diversity of the United States" required paying some attention to the numbers of applicants admitted.[17]

It would not make sense, for example, to have [only] 10 or 20 students out of 1,100 whose homes are west of the Mississippi. Comparably, 10 or 20 black students could not begin to bring to their classmates and to each other the variety of points of view, backgrounds and experiences of blacks in the United States.[18]

One way of interpreting Harvard's policy is simply the recognition that no sane person could in fact believe that there is a singular "black point-of-view" or "black experience," anymore than there is a singular identity binding together persons from "west-of-the-Mississippi." That is, even if the addition of persons from the category of African-Americans, relative to a homogeneously white population, represents a net gain in diversity, along at least one dimension, it is also the case that one must pay attention to the diversitywithin any of the relevant racial or ethnic groups. Critics of affirmative action, including members of the United States Supreme Court,[19] sometimes write as if supporters of affirmative action ignore the presence of intra-group diversity, but this is clearly false. I know of no one who is so stupid as to believe that all (or even most) members of any given group necessarily have similar opinions on a variety of important issues.

In any event, because of Justice Powell's emphasis on the almost unique legitimacy of "diversity" as a constitutional value, it has become the favorite catchword--indeed, it would not be an exaggeration to say "mantra"--of those defending the use of racial or ethnic preferences. As Eugene Lowe has written, "[c]elebrating the value of racial and ethnic diversity has become routine in educational circles,"[20] not least, it should be obvious, because such celebrations seem licensed and, indeed, encouraged by the Supreme Court. Whatever the actual efficacy of the Supreme Court in changing the behavior of American institutions,[21] it seems indisputable that the Court sometimes fulfills the function of the French Academy in establishing the conventions of "law talk,"[22] so that all properly socialized lawyers, and many non-lawyers as well, adopt certain conventions of argument because the Court leads the way. It is a version of the old children's game of "Simon Says." If Simon says, "Stop talking about the difference between commerce and manufacture,"[23] then a mode of analysis that had been constitutive of law talk only a few years before[24] disappears almost overnight. More to the present point, if Simon says, "Start talking about diversity--and downplay any talk about rectification of past social injustice," then the conversation proceeds exactly in that direction.

"Diversity" is thus a ubiquitous topic of contemporary discourse. Indeed, it has joined "family values" and "good medical care" as something that everyone is for, as demonstrated by the fact that it is becoming ever more difficult to find anyone who is willing to say, in public, that institutional or social homogeneity is a positive good and diversity a substantive harm.[25] Opponents of affirmative action almost never attack the merits of diversity per se, but, rather, the specific means thought necessary to assure the achievement of a desired degree of diversity.[26] Were that degree attainable by non-objectionable means, most opponents of affirmative action insist, they would be utterly delighted.

"Diversity" is, however, not a self-interpreting word. Political theorists have for several decades now posited the notion of "essentially contested concepts,"[27] i.e., notions that are extremely important but, nonetheless, without truly definite meaning. Consider the crucial American value of "freedom;" one is foolish indeed to believe that Americans have ever agreed on precisely what that term entails, although all have agreed that it is a term worth fighting (and killing) for.[28] Perhaps my favorite example of "essential contestedness" can be found in the very title of a marvelous book, Equalities, where Yale professor Douglas Rae and his co-authors elaborate no fewer than 128 logically coherent notions of "equality."[29] One reason why debates about, say, the "equal protection" clause are so bitter is because one person's cogent notion of "equality" may differ drastically from another's equally cogent notion, though each prefers to believe, falsely, that only his or her particular notion represents "real" equality. Would that it were that easy! The same, I fear, is increasingly true of those who raise the banner of "diversity" and then argue bitterly about its meanings, especially in a context where one seeks not only, if at all, the agreement of the trained philosopher but also, and as a practical matter far more importantly, the imprimatur of a court trained to think that some legal magic resides in a program's being successfully described as contributing to diversity.

Consider now the third of the key cases--and the second one to involve the University of Texas Law School--Hopwood v. Texas,[30] in which Cheryl Hopwood successfully challenged the admissions program at the University of Texas Law School insofar as it took race and ethnicity into account.[31] Although the federal district judge had upheld the admissions process in operation at the time of the judgment,[32] a three-judge panel of the Fifth Circuit Court of Appeals reversed the judge and held that the Law School must design an entirely race-neutral admissions process.[33] But what about Bakke and its toleration of a race- and ethnic-sensitive, diversity-seeking admissions policy? A majority of the panel held that Bakke no longer stated the effective view of the Supreme Court and, therefore, the operative meaning of the Equal Protection Clause.[34]

The majority noted the undoubted fact that Justice Powell spoke for himself alone, as well as perhaps the even more embarrassing fact, at least in retrospect, that the four justices who would have upheld the Davis program did so not by reference to the value of a diverse student body--a notion wholly absent from their opinions--but, rather, by emphasizing the program's utility in overcoming an egregious heritage of past discrimination.[35] The majority held that subsequent decisions had so undercut Justice Powell's pro-diversity rationale that the panel, however "inferior" it might be within the structure of federal courts,[36] was no longer bound by it. Instead, it interpreted post-Bakke cases as holding "that the use of ethnic diversity simply to achieve racial heterogeneity . . . is unconstitutional."[37] The Hopwood panel came as close as any court has yet done to reading the Fourteenth Amendment as indeed requiring the "color-blind[ness]" of which Justice Harlan spoke in his canonical dissent in Plessy v. Ferguson.[38]

Does this mean that university admissions must be based on a single metric from which no deviation is allowed? Not at all. The court points to "a host of factors" that can legitimately be considered in the admissions process, including "ability to play the cello, make a downfield tackle, or understand chaos theory. An admissions process may also consider an applicant's home state or relationship to school alumni."[39] Which means, for example, that the most important affirmative action program at the University of Texas Law School--the rigid setting aside, as a quota, of a full eighty percent of seats at the school for residents of Texas, whatever their "merit" may be when compared with non-resident applicants--is apparently safe from any kind of attack.[40] Moreover, law schools "specifically may look at things such as unusual or substantial extracurricular activities in college, which may be atypical factors affecting undergraduate grades."[41] Finally, schools were given permission to "consider factors such as whether an applicant's parents attended college or the applicant's economic and social background."[42] It should be obvious that the court cannot fairly be described as hostile to diversity as such (but only to the specific kind of "diversity" that is at the heart of the contemporary debate).

The point is best demonstrated by Judge Jerry Smith's own gratuitous comment in the majority opinion that

Plaintiff Hopwood is a fair example of an applicant with a unique background. She is the now-thirty-two-year-old wife of a member of the Armed Forces stationed in San Antonio and, more significantly, is raising a severely handicapped child. Her circumstance would bring a different perspective to the law school.[43]

Judge Smith, then, clearly seems to endorse a self-conscious search for "different perspective[s]" within the law school's applicant pool as he offers his own completely unverified (but commonsensical and quite possibly accurate) belief that raising a severely handicapped child would generate sufficiently interesting opinions about legally relevant issues as to justify a law school in giving special weight to someone in Hopwood's position as against, presumably, a parent of a non-handicapped child. The central point, though, is that the construction of "diverse" classes remains legitimate even for the Fifth Circuit, which scarcely wished to endorse homogeneity as such. Instead, it "simply" forbade the use of race or ethnicity as a proxy for the kinds of diverse backgrounds that can legitimately be sought by admissions committees.

Hopwood generated a national debate, as did the decision of the Regents of the University of California to order a race-neutral admissions process, which was, of course, followed by the passage of Proposition 209 via state-wide initiative-and-referendum, which attempts to bring an end to any and all race-preferential programs within the State.[44] Whatever the status of Hopwood in regard to structuring the University of Texas admissions program, the Supreme Court's refusal to take the case assured that its legal reach would be limited to the three states--Texas, Louisiana, and Mississippi--covered by the Fifth Circuit,[45] and that the national debate would continue. That debate centers on the meaning of "diversity," its ostensible desirability, and whether the Constitution, properly understood, erects barriers to achieving it.[46]


How, then, do we talk about diversity? I begin by offering several examples of what might be termed "diversity-talk," all quoted in an important and widely-reviewed recent book defending affirmative action, by the former presidents of Harvard and Princeton.[47] It is probably not surprising to read that the presidents of the sixty-two leading universities that comprise the Association of American Universities "believe that our students benefit significantly from education that takes place within a diverse setting," which means the opportunity to "encounter and learn from others who have backgrounds and characteristics very different from their own."[48] More surprising to many may be statements by the CEO's of Coca-Cola and Chrysler. Coca-Cola, described as "focused on taking actions that serve us best over the long run," is committed to "building . . . a diverse workforce. As a company that operates in nearly 200 countries, we see diversity in the background and talent of our associates as a competitive advantage and as a commitment that is a daily responsibility."[49] Finally, what was then only the Chrysler Corporation--though one assumes that the new Chrysler-Daimler would say much the same thing--"believe[s] that workforce diversity is a competitive advantage. Our success as a global community is as dependent on utilizing the wealth of backgrounds, skills and opinions that a diverse workforce offers as it is on raw materials, technology and processes."[50]

Thus, celebrations of racial and ethnic diversity extend well beyond the world of education.[51] Indeed, I would even hazard the guess that one could hear substantially more public opposition to "diversity-based" affirmative action on most university campuses than in most gatherings of business people. There seem to be few corporate analogues to Harvard's Harvey Mansfield, Jr.,[52] or my own colleague Lino Graglia.[53] Indeed, no one trying to understand the contemporary politics of affirmative action can fail to note that major industry has been notably unwilling to join those, like California's Ward Connerly,[54] who are leading the fight against it. Whatever explains the current backlash against the practice, it cannot plausibly be viewed as part of a campaign by big business. That being said, it is worth making the effort to understand what corporations mean by "diversity."

One meaning--I suspect a quite common one--is simply that one's workforce reflects in some important sense the demographic composition of the surrounding society. As noted by David A. Thomas and Robin J. Ely in their article in the Harvard Business Review, such a definition says literally nothing about the consequence of any such diversity for the way that work is carried out.[55] They note, for example, that from at least one perspective, the most successfully "diverse" organization in the United States today is the United States Army.[56] Few persons have been so bold as to suggest, though, that General Colin Powell brought a "black perspective" to his vision of military strategy. The whole point of the Army's vision of diversity is to demonstrate that all sorts of persons can be successfully assimilated into the organization without consequence for the definition of its mission or performance of its operational goals.

I have written elsewhere that one way of defining the "professional project," vividly illustrated within most American law schools, is to "bleach out" idiosyncratic personal aspects of our selves, including those that might relate to sex, race, gender, ethnicity, sexual orientation, and religion and to adopt instead a rigorously impersonal professional stance defined as appropriate for one who is behaving "as a lawyer."[57] This is obviously an even more significant, and widely accepted, claim with regard to the military. Those who advocate, as I do, welcoming gays and lesbians into the military do not, at least publicly, claim that they will bring a distinct set of viewpoints that will lead to likely changes in the way that war is actually conducted. From this perspective, demographic diversity serves only to provide access to important social institutions, but it does not convey any information as to how these institutions actually operate. Thomas and Ely refer to this as a "discrimination-and-fairness paradigm," where even those who accept the validity of racial or other preferences are, nonetheless, ultimately committed to an internal organizational notion of "color-blindness" or "gender-blindness."[58] "Under this paradigm, it is not desirable for diversification of the workforce to influence the organization's work or culture."[59] There will be relatively little interest in, if not hostility to, "explor[ing] how people's differences generate a potential diversity of effective ways of working, leading, viewing the market, managing people, and learning."[60] Although there is clearly a desire to draw the applicant pool from a widely diverse population, and pride taken in the heterogeneity of background of those within the organization, it is hard to describe the overall view as one that values "diversity" as such.

I confess that I am more interested in defenses of diversity that indeed focus on the potential contribution that a diverse workforce can make to an institution by enhancing the quality of the work done within a given company or institution.[61] One often finds arguments that the presence of some factor X enhances the quality of the product identified with the institution. The use of very good paint or of a very good microchip, it might be argued, will make one car better than its competitor, which uses an inferior paint or microchip. Similarly, the hiring of especially able engineers or product designers would similarly enhance the product. In contrast, one might freely admit that some other factor Y has nothing to do with product enhancement, but, nonetheless, assert its desirability because of its relevance to achieving other important objectives. Running a major athletic program, for example, has almost nothing to do with enhancing the education of college students, but it may be effective as a means of generating alumni contributions or, if a public university, of attracting support from certain state legislators. Alumni "legacies" seem similarly dubious product enhancers, though one assumes that university fundraisers would bewail any significant diminution in number of such legacies as likely to harm their efforts.

For me, "product" has a very broad definition, ranging from automobiles and soft drinks to scholarly knowledge or to the dispositions linked with being an "educated person." An important question, then, is how often proponents of "diversity" argue that the attainment of a suitably mixed population will lead to a higher quality of the product identified with his or her institution. This question is especially important in relation to corporate proponents of "diversity."

Recall that the CEO of Coca-Cola declares that "a diverse workforce" will "serve us best over the long run."[62] One notes the no-nonsense, self-interested thrust of this argument. Coca-Cola is not doing something merely because the country (or even the world) as a whole will be better off, even if, perhaps, at a cost to the economic interests of Coca-Cola's shareholders. That would be a rather tragic conclusion, suggesting a tension between what is good for Coke and what is good for the nation. Instead, the CEO displays a far more comedic view of the world, presenting the happy conclusion that the shareholders (and, presumably, consumers) of Coca-Cola will themselves be better off with "a diverse workforce." And why is this the case? The rationale offered has something to do with the fact that Coca-Cola "operates in nearly 200 countries." That is why Coke's management regards "diversity in the background and talent of our associates as a competitive advantage" and proclaims the achievement of diversity to be "a commitment that is a daily responsibility."[63]

Wherein comes this "competitive advantage"? First, can it plausibly be said that a diverse workforce will produce a better-tasting product or that it will even produce the same product in a better manner, e.g., more efficiently? It is obvious how the answer might be affirmative. Hiring people with a variety of talents in formulating potential soft drinks would certainly work to Coke's advantage, but one has to be quite imaginative to believe that such talents correlate in any robust way with the racial or ethnic demography of the workforce. I suppose it is possible that one could learn from a diverse workforce that Greeks, say, prefer sweeter drinks than do Danes, and thus formulate Coke differently in the two countries. This assumes, of course, that Coca-Cola (or any other company pronouncing a similar commitment to "diversity") chooses to formulate its drinks differently around the world, rather than adopting a marketing policy that relies on its product tasting just the same whether one buys Coke in Armenia or Zanzibar.

There is at least one circumstance where a diverse workforce would indeed be a more efficient one: if the prior, non-diverse, workforce, was generated by the presence of discriminatory barriers to employment. What "discrimination" means in this context is that a business deprives itself of what would, to a non-discriminatory observer, be the most meritorious employees as a result of a commitment to bigotry. The lifting of discriminatory barriers will allow such employees to be hired and, therefore, bring their skills to the enhancement of the product, whether directly, as by inventing new formulas or indirectly, as by producing it more efficiently.

As already suggested, though, this is not an argument for diversity per se; rather, it is an argument for meritocracy, with the assumption being that a non-discriminatory hiring process will in fact generate a heterogeneous workforce. The presence of a homogeneous workforce is less an occasion for first-order criticism than for second-order suspicion as to the process by which it has been achieved. It is worth pointing out that one should be wary of confusing suspicion with proof. Assumptions that nothing other than discrimination explains any instance of variation in the demographic composition of any given segment of the workforce seem to call into question at least one tenet of "diversity," especially in its "multiculturalist" form, which emphasizes the importance of a person's "culture" as a basic shaping force. If variations in culture are as significant as they are sometimes alleged to be, then one would expect the multiculturalism of American society (and of the wider global society) to be reflected, at least to some extent, in "tastes" or propensities for different kinds of jobs and other life experiences. The key point, though, is that the highest good of an "anti-discrimination" focus is not the achievement of a diverse workforce per se, but, rather, the achievement of a bias-free selection process that guarantees to those rejected that they were not victimized by the use of illegitimate criteria and to the outside public, including shareholders, that the products are being produced in the most efficient manner possible. But if a non-discriminatory process generates a non-diverse workforce, for whatever reason, it may still be the case that the particular product is the best it can be, and being produced as efficiently as it could be.

So what of Coke's worldwide empire? How does this become relevant? One might well surmise that persons will be more likely to order Coke from their fellow nationals, who will assuage concerns that drinking Coca-Cola is simply a way of supporting American imperialism. The Italian branch of Coca-Cola thus benefits from the presence of Italian nationals in the workforce, as bottlers, delivery-persons, and sales agents, just as the Thai branch should be eager to display Thai nationals, and so on throughout the world. Similarly, the advertising agencies preparing Coke's copy might want to include in the management team people from the targeted populations, believing that they will be especially able both to think up culturally-attractive jingles or, just as importantly, spot materials to which local audiences might take umbrage.[64] One need not go abroad, of course, to offer such justifications. One would be surprised if Coca-Cola is entirely indifferent to the ethnic composition of those it sends to solicit sales in, say, the Rio Grande Valley area of Texas or in Chinatowns in New York or San Francisco. Given the particular kind of consumer good that Coca-Cola is, one would expect the company's management to be at least as aware of, and sensitive to, the increasingly multicultural nature of American society as are most academic sociologists.

What Coca-Cola may be doing is making a version of a very traditional argument, which is simply that a company, to maximize profits, must be attentive to the preferences of its customers, including a desire to be served by "people like them."[65] Here, Thomas and Ely posit an "access-and-legitimacy paradigm."[66] "Where this paradigm has taken hold, organizations have pushed for access to--and legitimacy with--a more diverse clientele by matching the demographics of the organization to those of critical consumer or constituent groups."[67] Thus, "[m]any consumer-products companies that have used market segmentation based on gender, racial, and other demographic differences have also frequently created dedicated marketing positions for each segment,"[68] with attendant new opportunities for members of the given groups.[69] It is worth noting that "customer preference" arguments carry with them obvious dangers; for many years they served as justification for hiring only whites in the South on the basis that the majority of white customers (who were, of course, the majority of all customers) just wouldn't accept being served by members of racial or ethnic minority groups. It is refreshing--perhaps an especially appropriate word to use in regard to Coca-Cola!--to see the customer preference argument made in a way that generates a diverse workforce, though one might still issue a cautionary note that devotees of diversity accept at their peril the priority of customer preferences as a criterion for workforce hiring.

Note some other implications of Coke's policy, at least if one views it as motivated by a desire to protect its business interests in two hundred countries or even its market share in the United States. All countries or groups may be created equal in some abstract sense, but they are most certainly not created in equal numbers. It may be desirable, for example, to hire some Norwegians to take care of the relatively small Norwegian market. But if market preference is the basis of the hiring decision, then a Norwegian (or Norwegian-American) applicant might face special problems if the company already has hired relatively few Norwegians, but a sufficient number to saturate the Norwegian market. After all, few people outside of Norway would be affirmatively impressed by the presence of a Norwegian representative, unless, of course, she had talents having nothing to do with her ethnic identity.

Moreover, it is vital to recognize that Coca-Cola does not argue that its workforce in each country needs to be diverse; rather, only its overall workforce must be so. I cannot help thinking of some of the debates about school desegregation, particularly with regard to the difference it makes if one looks at the racial demographics of a given school as a whole or if one looks at individual classroom figures. It is all too easy to demonstrate that diversity (or "racial balance") exists in the former even if there is almost complete separation of the races in the latter.[70] Similarly, a multinational company could easily have a strikingly diverse international workforce while having equally striking homogeneous workforces in any given country or locale. (How many non-Thai nationals work for Coca-Cola in Thailand?)

One could, of course, say much the same about the assertion by Chrysler's then-CEO that "workforce diversity is a competitive advantage," especially insofar as he, too, points to Chrysler's presence in "a global community" and says that the company's success depends as much "on utilizing the wealth of backgrounds, skills and opinions that a diverse workforce offers as it [does] on raw materials, technology and processes."[71] I cannot forbear from noting a particular paradox in Chrysler's offering diversity of "opinions" as a corporation-embraced good. One immediately wants to ask, "Opinions about what?" I am confident that the answer for Chrysler, or any other corporation for that matter, is only opinions about the best way to build and sell cars or whatever other good is being produced by the given business. It would be astonishing, for example, if Chrysler is affirmatively interested in whether its workforce is homogeneous or heterogeneous regarding opinions about abortion, foreign policy, the merits of modern art, or, perhaps most to the point, labor policy or general economic concerns, such as the relationship between globalization and job security or the future of the welfare state.

I do not necessarily mean to pick on business corporations. They may, indeed, be altogether typical of any large-scale institution. As University of California law Professor Robert Post has emphasized, all "managerial" enterprises, whether private or public, define themselves in terms of specific work tasks.[72] The United States Postal Service is defined by its expertise in delivering the mail, not by its achievement of an on-the-job chautauqua featuring lively interchanges among the workforce or between the workforce and its customers. Indeed, I dare say that most of us would be appalled if our mail deliverer or postal clerk selling us stamps began inserting his or her own opinions about our choice of reading materials. Imagine a postal worker who, while delivering your mail, urged you to stop subscribing to Playboy or the National Review or pressed upon you material suggesting how you can achieve eternal salvation.

One argument for "diversity" within an academic setting (or on television talk shows) is precisely that it generates an "uninhibited, robust, and wide-open" debate[73] that an institution presumably desires (and that Justice Powell emphasized in his own embrace of "diversity" in Bakke[74]). For better and worse, few institutions are defined by their commitment to such debates. Certainly, this is not true of any business institution of which I am aware, and, as we shall see presently, this is scarcely an accurate description of a properly functioning university.[75]

One's suspicion about diversity as practiced by Coca-Cola, and many other companies, is that it relates primarily to the marketing of products rather than their actual production. Not surprisingly, there is a debate about the relevance of diverse backgrounds to the quality of what is produced. One authoritative review of the evidence states:

Under ideal conditions increased diversity may have the positive effects predicted by information and decision theories [i.e., generation of varied ideas and approaches, greater creativity]. However, . . . the preponderance of empirical evidence suggests that diversity is most likely to impede group functioning. Unless steps are taken to actively counteract these effects, the evidence suggests that, by itself, diversity is more likely to have negative than positive effects on group performance.[76]

Consider in this context a study supporting the unsurprising proposition that persons coming from different backgrounds often misread each other's social signals.[77] Different people were asked to respond to the statement: "In order to have efficient work relationships, it is often necessary to bypass the hierarchical line."[78] In one example, systematic differences emerged between Swedes and Italians. Swedish managers are described as believing "that a perfect hierarchy--in which one's boss knows everything--is impossible."[79] It is, therefore, perfectly proper to bypass superiors and go directly "to the person most likely to have the needed information and expertise."[80] "Most Italians," however, "consider bypassing the boss as an act of insubordination," so that mutual frustration rapidly builds up between Swedish workers and Italian superiors.[81] On the other hand, Swedish superiors tend to view Italian subordinates as "lack[ing] initiative and . . . unwilling both to use personal judgment and to take risks," because the Italians engage in "constant requests for permission and information."[82]

Thomas and Ely note that "many attempts to increase diversity in the workplace have backfired, sometimes even heightening tensions among employees and hindering a company's performance."[83] By way of response, they devote their article primarily to articulating the merits of an "emerging paradigm: connecting diversity to work perspectives,"[84] by which they mean strategies for incorporating "employees' perspectives into the main work of the organization and to enhance work by rethinking primary tasks and redefining markets, products, strategies, missions, business practices, and even cultures."[85]

I want very much to believe that examples of organizational disruption caused by workforce diversity are limited only to companies that are organizationally inept at taking advantage of diversity, perhaps because they are insufficiently committed to the ostensible emerging paradigm. One might ask, though, if it matters. That is, does one's support for "diversity" turn on the answers to such empirical questions, or is it based on certain normative commitments that are, in some important senses, independent of empirical evidence? There may be good reasons to support affirmative action besides enhancement of organizational performance. One may believe, as I do, that society as a whole is likely to be better off if its workplaces are diverse, a point I shall return to shortly. But this is a different argument, patently public-regarding, in contrast to the purportedly self-regarding arguments proffered by the CEOs I have discussed. Self-regarding arguments have the advantage of appearing more hard-headed and less idealistic; they may, for better and worse, however, be subject to more stringent empirical tests than are public-regarding arguments that forthrightly admit that costs may have to be paid in order to achieve desirable social goals.


Let us, at last, return to educational institutions, presumably very different from Coca-Cola or Chrysler. Still, educational institutions can be said to produce certain products, including well-educated graduates and scholarship. They may also be in the business of producing good citizens, a civic task that may relate, in certain complex ways, to "pure" education or scholarly attainment, but is, at the end of the day, quite different.

One could give an entire lecture, at the very least, about the notion of the "well-educated" person. Let me offer only two possible criteria of being well-educated. One involves what might be termed "disciplinarity;" for example, chemistry courses should produce persons knowledgeable about chemistry; economics courses should, similarly, produce persons proficient in performing the tasks expected of a competent economist. As a law professor, I am interested in doing what I can to turn our students into competent analysts of law and legal structures by way of preparing them to engage effectively in the role of lawyer. A second notion might focus less on disciplinary prowess and more on other attainments. Thus, among other things, the well-educated person might be expected to possess a certain level of curiosity about the world in general, reflected perhaps in manifesting the social virtues needed to relate to a wide variety of different persons with different interests. Truly well-educated persons should be able, if sitting next to someone at a dinner party with whom they apparently have "nothing in common," to engage in conversation and even learn something from the perspectives and experiences brought to bear by the putative stranger. Beyond this is the distinctively civic virtue by which persons learn to treat social strangers as their fellow citizens, to whom they owe obligations and loyalty, even if they seemingly have little in common other than shared political identity. And beyond this is the empathic identification with "complete" strangers, such as, say, Kosovars or Chechens, who have nothing in common with most of us other than the shared status of being human and the ability to suffer. How does "diversity" work in regard to these various tasks of education?

I begin with the most obvious tasks of educational institutions: disciplinary education and the encouragement of disciplined scholarship. What does "diversity" contribute to these goals? My answer is vigorous and unequivocal: "It depends." In regard to the classes that I teach, on American constitutional development and on the American legal profession, I strongly believe that the quality of these classes is indeed a function of the classroom demographics. To take the most obvious example, I find it intensely frustrating to discuss the role of race in American history before a post-Hopwood class that includes few or no African-Americans. It would be equally troublesome to try to discuss gender-related issues in an all-male (or, for that matter, all-female) class. Racially or gender-mixed classes can provide the same benefits provided by the presence of "a former policeman in [a] criminal procedure [class], a former business executive in corporations, and a former bank examiner in commercial law," all of whom "are sure to enrich those classes in ways that" other students, without such backgrounds, cannot.[86] To the extent that Hopwood functions to make much more difficult the attainment of a sufficient degree of heterogeneity in my own classrooms, I believe that the quality of legal education at the University of Texas Law School has been diminished and, therefore, the students significantly disserved.[87]

I should acknowledge, though, that my views are, to some extent, prompted by the specific courses that I teach. I am not at all certain that classroom demographics, at least with regard to race or gender, would be relevant if I taught courses, say, on the taxation of international businesses, though one can certainly imagine that it might be helpful to have students from some of the "host" countries of multinational enterprises. Looking around at the rest of the university, I am far more confident that certain courses in the humanities and social sciences benefit from a "diverse" student body than do courses in the so-called "hard" sciences or mathematics. To the extent that I support the use of racial or gender preferences in regard, say, to admitting students to a graduate program in mathematics,[88] the reasons are necessarily other than a belief that the quality of the disciplinary education itself depends on classroom demographics.

It is worth noting some skepticism about whether classroom demographics matter in any course that operates only via lectures and does not include opportunity for some genuine give-and-take response to the material presented. Even among courses that depend on discussion, some subject matters more plausibly draw on students' own "personal" insights than do others. To the extent that students interact with one another outside of courses there could be real benefits from "diversity." Presumably, however, universities have more interest in (and competence to evaluate) what goes on inside their classrooms than in potential late night conversations in dormitories. And most proponents of "diversity" generally focus on that part of university education that is classroom-specific.[89]

Moreover--and here I think we come to the heart of the matter--one should acknowledge that Judge Weiner, in his concurring opinion in Hopwood, was entirely fair in noting that my law school in fact has a somewhat limited notion of diversity. "Focusing as it does on blacks and Mexican Americans only, the law school's . . . admissions process misconceived the concept of diversity . . . . [B]lacks and Mexican Americans are but two among any number of racial or ethnic groups that could and presumably should contribute to genuine diversity."[90] He is absolutely right. Indeed, it is his opinion, far more than the obtuse majority opinion penned by Judge Smith, that offers the greatest challenge to those who proclaim their devotion to "diversity" as a "compelling need in higher education." The reason is that he accepts the central proposition, but then goes on to show that the arguments that universities offer with regard to ostensibly diversity-oriented admissions policies are under-inclusive.

I begin with a personal example: I have become especially aware of the limited nature of University of Texas's diversity because of a developing interest of mine in the constitutional implications of American expansionism, including the seizure and dominion over Puerto Rico as part of the aftermath of the Spanish-American War in 1898. Many fascinating and delicate constitutional problems were raised (and continue) as a result of our venture into imperialism. Indeed, Puerto Rico today is described by many as the world's largest remaining colony,[91] if we mean by "colony" a political entity that enjoys no formal political sovereignty and is under the ultimate control of a polity in which its own members do not have the right to participate politically. Although Puerto Ricans are American citizens, they cannot, of course, vote for the President nor do they have voting representation in Congress. And, as President Clinton's recent pardon of Puerto Rican terrorists reminded us (even if he did a notably inept job of defending his altogether defensible decision), there is a significant, albeit small, movement in Puerto Rico that is borrowing a page from America's own revolutionary past and seeking independence from those it deems its colonial oppressors.

I think it is relevant to note that my new-found interest had its genesis almost two years ago at a Yale Law School conference on Puerto Rico organized by a remarkable third-year law student from Puerto Rico. She was determined to "bring to the mainland," as it were, the issues that so passionately concerned her. Not only were the issues brought to the mainland, she also was able to bring to New Haven a number of representatives of the decidedly different points of view, including the Governor. These views ranged from the Governor's desire for statehood to those advocating for independence, with those wishing to maintain the present "commonwealth" status in between.

As a direct result of my introduction to the issue, I now assign one of the central 1901 Insular Cases[92] addressing the status of Puerto Rico to my introductory courses on constitutional law and have added long excerpts from that case to a casebook on constitutional law that I co-edit.[93] That being said, I have no doubt whatsoever that the discussions in classes using these materials would benefit significantly if they included students from Puerto Rico. There are, to my knowledge, no Puerto Rican students at the University of Texas Law School, and I suspect this is true at a number of--perhaps most--American law schools. There are, no doubt, a number of explanations. Concerning Texas in particular, the State has not been a historical site of Puerto Rican migration, unlike, say, New York or New Jersey. That in itself does not, however, fully explain why Puerto Rican-Americans are not sought out by the University of Texas Law School. Perhaps one might argue that one consequence of the meager migration to Texas is that, concomitantly, there is little or no history of past discrimination by Texas against Puerto Ricans, unlike the cases with African- and Mexican-Americans. Even if this is true, recall that it is illegitimate for the University to ground its affirmative action programs on the presence of past discrimination in the wider society. Most relevant, almost certainly, is the fact that Puerto Ricans, by being an insignificant part of the Texas population, do not count as a significant political constituency for any Texas politician. No credible political threat is posed to those who ignore the specific interests of Puerto Ricans.[94] I deeply regret this limit in our own range of diversity, for the absence of Puerto Rican students, at least on the days that we are discussing the issues raised by the Insular Cases, means that my course is of lesser quality than it might otherwise be.

As suggested earlier, it would by no means be enough to have one Puerto Rican in the classroom, given the bitter divisions among Puerto Rican proponents of statehood, independence, and maintenance of the constitutionally uncertain "commonwealth" status. It would, no doubt, be as illuminating to my students to hear heartfelt expressions of the various positions as it was to me. Of course, if the only point were hearing heartfelt expressions, that could be attained by playing tapes of the New Haven conference. More important, presumably, would be the opportunity to participate in a genuine encounter with the speakers, asking them questions and engaging them in real conversation. One might still ask if such encounters depend on mutual presence at the same university, as opposed to using contemporary technology to bring together remarkably diverse groups in the common ground of "distance learning" or cyberspace. I am thinking, for example, of a course on nuclear arms policy, organized well over a decade ago by Martin Sherwin at Tufts University, that consisted of his American students in Medford, Massachusetts, and a class of students at the Moscow State University who met together at frequent intervals via satellite. It would be easy enough to hook up similar connections with students at the University of Puerto Rico, and that might in fact provide a richer kind of diversity of experience than that provided by relatively atypical persons who leave their homes to attend schools thousands of miles. We should increasingly rethink our assumptions that university education necessarily takes place in fixed geographical settings. Thus, writes Arthur Levine, the president of Teachers College, Columbia University:

It is possible right now for a professor to give a lecture in Cairo, for me to attend that lecture at Teachers College and for another student to attend it in Tokyo. It's possible for all of us to feel we're sitting in the same classroom . . . . It's possible for the professor to point to me and my Japanese colleague and say, "I want you to prepare a project for next week's class." If we can do all of that, and the demographics of higher education are changing so greatly, why do we need the physical plant called the college?[95]

To be sure, one might still lament the lack of an opportunity to go to a coffeehouse or bar afterward and continue the arguments provoked by the relatively minimal classroom discussions, but, perhaps, intensive Internet discussions among the various students spread across the world would provide a more than adequate substitute at least with regard to the specific variable of diversity.

I most definitely do not believe that only Puerto Ricans are capable of thinking cogently about Puerto Rico. What I do believe is that persons from Puerto Rico are, as an empirical matter, far more likely to have thought about issues relevant to people in Puerto Rico, such as the political status of the island, than are non-Puerto Ricans. Though I may know enough to raise the most basic issues for my students, I have no sense of the nuances that come from extended immersion within an ongoing culture, especially, a culture of bitter argument. Nor, as a matter of blunt fact, do I have very great personal incentive to educate myself in-depth about the issues facing Puerto Ricans. Life is short, and the demands for my limited time and attention span are overwhelming. An intellectual interest in Puerto Rico, which I now most certainly have, is not the equivalent of truly caring deeply enough about the fate of that island to invest the ever-scarcer resources of time and energy in ways that would allow me to become truly expert. It is a sad truth that most of us care most deeply about, and invest most significantly in, those things (and people) who are closest to us, both literally and metaphorically.

Race and ethnicity, therefore, at least on occasion, may act as proxies, not so much for holding specific views, but for the probability of being deeply interested (and at least somewhat knowledgeable) at all in certain issues, i.e., those issues most germane to the group in question. (One of these issues might be whether there are specific issues that are necessarily germane to the group!) I think it simply undeniable that African-Americans are more likely to be concerned with the problems facing African-Americans--and, for that matter, more aware of the complexities and divisions within the group of those comprising the community of African-Americans--than are non-African-Americans. This is not at all the same thing as saying that African-Americans bring some privileged understanding of African-Americanness or that whites are without the capacity to think cogently about African-Americans (or that some whites do not care very deeply about the welfare of African-Americans).

These latter assertions indeed raise the most profound difficulties and perhaps even fall victim to the Fifth Circuit's statement that "government may not allocate benefits or burdens among individuals based on the assumption that race or ethnicity determines how they act or think."[96] But it seems foolhardy to deny the presence of a correlation between racial and ethnic status and the propensity to think about related problems. To deny this is indeed to accept the Fifth Circuit's assertion that race is no more rational a basis for decision-making "than would be choices based upon the . . . blood type of applicants."[97] Most of us almost never think of our blood type; as an empirical matter, though, it is foolish beyond belief to say that we think as rarely about the implications of our racial and ethnic identities as about our blood type. And, with regard to at least certain groups that do spend a lot of time thinking about blood types--i.e., hemophiliacs and cancer patients interested in receiving bone-marrow transplants--one wonders if the Fifth Circuit, given its solicitude for parents of disabled children as presenters of important perspectives, would find it unthinkable to grant a preference to these particular bloodtype-obsessed persons because of the insights they might contribute in courses on health policy and law.

Similarly, if I taught courses on, say, the American role abroad, I think classroom discussion, and therefore the production of educated students, would benefit from the presence of members of some of the various foreign countries that are often the objects of American policy. As I listened last spring to National Public Radio dispatches about the NATO bombing of Yugoslavia, I could not help but wonder how many Americans have met a single person from Kosovo or Macedonia or how, concomitantly, our views about the propriety of our policies might differ if we placed human faces on the abstractions called "Serbs," "Kosovars," or "Croats."[98] Thus Professor Loewy, in his own reflections on diversity and law school admissions, notes that "[i]n a class of 200, I would tend to choose the first Iranian or the first Sikh over the twenty-third African-American."[99] If one finds Professor Loewy's choice troublesome, then I suggest that one has not sufficiently thought through what it means to take diversity seriously. Truly to take diversity seriously might well require revolutionizing current admissions practices.

Consider only the fact that the University of Texas Law School operates under a very rigid admissions quota by which the Law School may not admit more than twenty percent of non-Texans. Could a university claiming to be seriously committed to "diversity" justify such a quota? If Heman Swett was ill-served by being channeled to a law school that lacked any representatives of eighty-five percent of the population, are Texan students much better served by attending a contemporary school that may lack significant representation of the vast majority of the country that is non-Texan and, indeed, of the overwhelming majority of increasingly globalized world that is non-American? The devastating parochialism of the "downtown law school" was absolutely obvious; other parochialisms may be less so, but no less present and, ultimately, only a little less stifling, both as to the Americans charged with (or volunteering for) extraordinary influence in structuring the global order, and as to the foreigners who are on the receiving end, for good and ill, of American power.

It is perhaps relevant here to mention a conversation I had this past year with the president of a major American university who mentioned his ongoing struggles with his university's own admissions office. He strongly believes that the university, which is truly "world class" by any conceivable measure, should strive to admit at least ten percent of its entering class from abroad. The admissions office is unwilling to go beyond five percent, in part, one gathers, because of fears of adverse reaction from disgruntled "local" applicants, some of whom, inevitably, will be children of alumni whose continued financial support is thought vital to the university. (Recall the Harvard statement quoted by Justice Powell emphasized Harvard's interest only in "the rich diversity of the United States"[100] rather than in the entire world. There is no particular reason why Harvard's students need exposure only to their fellow Americans.)

But to play this kind of "diversity card" is highly unfair. No university could genuinely try to maximize the possible diversity of its student body. Consider the fact, for example, that the editors of the Harvard Encyclopedia of American Ethnic Groups in 1980 listed no fewer than 106 separate ethnic groups within the United States alone.[101] As Miranda Oshige Mcgowan has shown, the category of "Asian" or "Asian-American" used by Davis and other schools ignores to an almost grotesque degree the profound differences among the many groups collapsed into one, empirically-false, identity as "Asians." "[P]eople categorized racially as Asian often do not view themselves as such, nor do they necessarily feel a sense of identity or kinship with others categorized as Asian. Instead, many Asian Americans define themselves primarily in terms of national origin and feel an affinity with others of the same national origin."[102] Consider a school that offered preference to an invented category called "Balkans" and treated Serbs, Croats, and Albanians as a single undifferentiated group. This may be all-too-analogous to collapsing persons of Japanese, Chinese, Vietnamese, and Korean background into one group called "Asians."[103] Similarly, to refer blandly to "American-Indians" masks the fact that there are at least 170 separate Indian tribes whose members, no doubt, would vigorously resist being assimilated into a single undifferentiated category. (Are there no significant differences between Navahos and Mohawks or, indeed, between the Navahos and their "next-door neighbors," the Hopis, with whom they seem to be engaged in almost endless dispute?) In this ever-more-complex multicultural world of ours, there are always going to be many more distinct groups making their claims for succor than there are spaces available. There is, by necessity, an "economy of diversity;" it is simply an existential reality of having to live under conditions of scarcity.

It is also worth noting, though full development of the point would certainly be worth an essay (or book) of its own, that much of the preceding discussion has assumed that we in fact know what we are talking about when we (or I) refer to people as "African-American," "Mexican-American," "Navajo," "Vietnamese," or whatever. But it is increasingly obvious beyond any reasonable doubt that such categories are truly social constructions, subject to remarkable instability upon close analysis. As more and more Americans (and people throughout the world) are of "mixed" parentage, deciding how to categorize people is itself a subject of major controversy.[104] Indeed, the eminent Harvard sociologist Orlando Patterson has suggested that if W.E.B. DuBois was at least half right in suggesting that the twentieth century would be the century of the color line, then the twenty first century will see the substantial eradication of the color (and ethnic) line, "made obsolete by migratory, sociological, and biotechnological developments that are already under way."[105] It may be that to maintain racially- or ethnically-oriented "diversity" programs will require even more than is currently the case that one engage in highly questionable, indeed demeaning, conversations about whether some person X is a "real" member of group Y given that he/she has the wrong last name, grew up in the wrong locale, etc. It is the necessity to engage in such conversations that gives weight to Justice Stevens pointed dissent in Fullilove v. Klutznick,[106] when he remarks that supporters of racial preference programs should look to the Nuremberg codes (or, at the very least, the elaborate racial codes of the Old South) for guidance. (Does a "single drop of blood" suffice, or must one have at least three grandparents of the relevant ancestry, assuming, of course, that we don't ask embarrassing questions about the "purity" of the grandparents themselves?) One must note that Stevens himself has ended up a relative supporter of such programs even as the majority of the Court has adopted the position he once seemed to endorse,[107] and, of course, for all of my ambivalence amply reflected here, I still wish to praise, rather than to bury, such programs myself. Still, I think there is much merit in Patterson's position and its suggestion that the demise of "diversity" preferences will come more from the practical pressures generated by the increasing "hybridity"[108] of the American (and world) population than from the adoption of a "principled" position of color-blindness.

All of this being said, I have increasingly come to believe that "diversity" is an idea of relatively limited utility with regard to understanding the actualities of such programs even if we put to one side questions like those raised in the last paragraph. Given that any and all "diversity"-oriented programs will necessarily be limited in their scope, preferences for only certain racial and ethnic groups must be defended on the basis of some argument other than a striving for diversity as such. One must always assert, as a practical matter, that the diversity provided by group A is more important, along some relevant dimension, than that provided by some groups B, C, . . . . To fail to rank order will always open oneself up to the devastating riposte suggested by Judge Wiener's opinion in Hopwood:[109] "Why did you stop with members of groups A-G, rather than go on to seek out members of putatively absent groups H-Z?" It is always possible to achieve even more diversity than one has now, if only one has the will (and makes a possibly lunatic decision to prefer diversity against any competing value). So why do we prefer the specific groups we do? Answers must always lie in group-specific reasons, and only rarely will these reasons track those that would be suggested by someone for whom diversity per se was the primary goal.

A "tracking" reason might be something like the following: "We have examined all of the contending groups, and we believe that the most truly unusual perspectives are those associated with group A. If one purpose of `diversity' is to provoke what some literary theorists call `defamiliarization,' the calling into question, or `problematizing,' our most basic assumptions of how the world is ordered, then encounter with someone from group A will best fill the bill." It may be that there is some university or organization that does operate under such criteria, but I am not familiar with it.

Far more representative, I am confident, is the answer provided by Professor Jack Balkin, who forthrightly states that, "[i]n the context of educational affirmative action, I understand `diversity' to be a code word for representation in enjoyment of social goods by major ethnic groups who have some claim to past mistreatment."[110] Thus, according to Balkin, inasmuch as "human capital (obtained through education) is one of the most important methods of wealth transmission in our age, it makes sense that Blacks and Hispanics should be given a larger share of this capital to make up for their relative deprivations due to social subordination."[111] He notes that "the groups that would receive a boost in cultural capital in Texas might be different from those in Washington State or Florida, for example."[112] But, if we are to "be honest about what `diversity' is really about," he says, we should admit that "[i]t is primarily about distributive justice of human capital, and [only] secondarily about creating a civic space in which the major ethnic groups in a given society can live relatively peaceably."[113]

It is very difficult, as a practical matter, to disagree with Balkin's observations, but it should be obvious that there is also a chasm between such arguments and those that view "diversity" as a positive good per se. Balkin's arguments focus on the past traumas of our polity that in fact deprived certain groups of a "fair chance" to compete for the "human capital" represented by education or access to the employment market. Someone more genuinely committed to the positive values of diversity should be far less interested in the historical explanation for its lack and more committed to assuring a desirable mix in the future. To return to my own example above, I do not in the least need to believe that Texas has denied Puerto Ricans any human capital to believe that the law school (and my classes) would benefit from the presence of persons from Puerto Rico. And, if I were to support the vigorous recruitment by the University of Texas of persons from, say, Eastern Europe, I would less focus on any purported wrongs done the Eastern Europeans by Texas (or even the United States) than on the positive benefits of breaking down our own parochialism. To adopt the valuable language of Stanford Law School Dean Kathleen Sullivan, "diversity" should not be viewed as a penalty we pay to rectify our past sins, but, rather, a policy warmly embraced because of its service to the present and future interests of the relevant institutions.[114]

The differences among the various approaches to "diversity" are especially well illuminated if we consider one additional example of a group that might claim some consideration in an admissions process ostensibly committed to the premise that "diversity" is relevant to the quality of education received. I refer to people who have religious views that are out of the "mainstream," at least as defined by the basically secular culture dominant in most American colleges and universities or, at least, those usually viewed as "elite" institutions. It is an open secret that the religious identities of students at the relatively few academic institutions that do not practice de facto open admissions[115] are scarcely representative of the genuine range of views found in a remarkably pluralistic American society. Indeed, former university Presidents Bowen and Bok caution that it is "easy to forget the importance of differences in religion as well as race and culture,"[116] an odd phrasing if one views religion as an important part of "culture." As someone who regularly teaches in my courses about the interaction of state and religion,[117] I believe that the discussion (and education of my students) would benefit as much from the active presence of, say, fundamentalist Christians as do discussions about race or ethnicity benefit from the presence of African- or Mexican-Americans.

So why don't more elite colleges and universities take religion into account in their admissions process, at least to the extent of awarding scarce (and valuable) places to those who come from "non-mainstream" religions?[118] One quick answer, of course, is that this would be prohibited, at least for state universities, because of the First Amendment's prohibition of the "establishment of religion."[119] That answer is scarcely satisfactory, though; indeed, I am increasingly tempted to describe it as equal in obtuseness to the Fifth Circuit's decision in Hopwood. It is, after all, now a commonplace of legal doctrine that racial classifications, because presumptively prohibited, can be used only if justified by some "compelling interest," and most proponents of racial and ethnic preferences (including myself) purport to believe that the striving for "diversity" meets this high standard. If that is true, then it is, to say the least, difficult to see how that interest would be less "compelling" in regard to, say, Christian Pentacostalists, Hassidic Jews, Mormons, Eastern Orthodox, or Muslims, all of whom may well be "underrepresented" at America's elite institutions.[120] To be sure, there would be many classes in which the religious identity of the student body would be absolutely irrelevant, but that, as we have already seen, is also likely to be the case with regard to racial, ethnic, or gender identity.

Balkin rejects the case for bringing religion within the diversity fold, though his explanation underscores the differences among forms of the arguments for diversity. Thus, he writes:

If "diversity" is a code word, as I have suggested, which has little to do with ideological differences, it makes perfect sense that conservative Christians would not and should not be beneficiaries of affirmative action even if they would make the student body more diverse in an ideological sense. That is because there is no sustained history of de jure persecution and social and economic oppression of conservative Christians in the United States that compares with anything suffered by Blacks and Hispanics. To be sure, there are forms of class oppression that conservative Christians suffer, but they do not really differentiate conservative Christians from other Americans in the same class position. And although conservative Christians may suffer from considerable stigma in the mass media, it is hard to claim that they have been denied the ability to amass human capital as Christians in the way that other groups have been denied these opportunities in the past.[121]

Even if one believes that Balkin's assertions are empirically correct, one need not deny, first, that the classrooms of many of our greatest secular universities pay a cost for the absence of distinctly religious voices and, second, that Balkin's reasons for rejecting religious preferences sound more in rectifying past injustice than in designing the optimal educational surroundings for students lucky enough to be admitted to elite institutions.

So far I have been discussing universities as if they consist almost entirely of students attending classes. What about faculty members who are, at least within research-oriented universities, expected to produce scholarship? In regard to the production of scholarship, "diversity" may be even more problematic than in regard to producing well-educated students. I start with the point that in the overall mix of disciplinary areas found in the contemporary college or university, there may be relatively few where one can plausibly argue that the kinds of scholarship produced are a function of the demographics of the scholars producing it. But there is a far more basic point that must be addressed, which involves a central paradox in regard to any and all disciplines. As Timothy Hall contends, the very point of a given "discipline" is to limit the number of arguments that will be taken seriously by those who are well-educated members of it.[122] Thus, Justice Powell's emphasis on the importance attached by the First Amendment to the notion of "robust debate,"[123] which is presumably enhanced by a diverse student body, turns out to be interestingly problematic within an academic context.

Whatever the genuine importance of vigorous argument to a liberal education, it is also true that the academy properly places limits on arguments that would be intolerable in the non-academic public square.[124] One is absolutely protected, for example, if he or she wishes to pass out leaflets on the street or in a public park denouncing Darwinian theories of evolution or endorsing the importance of an astrological understanding of the world. However, it would be a grave mistake in a college geology course to expect to be taken seriously if one proclaimed that the world was created only some six thousand years ago, as is thought by some biblical fundamentalists, and that ostensible tests indicating an older planet are misleading. A student would properly receive a failing grade if he proffered the view, on a final examination, that few Jews were killed by Nazi Germany, and that the numbers have been wildly exaggerated by a Zionist conspiracy. Every student may have a legal right to her own opinion in the public square, but there is certainly no entitlement that that opinion be granted any respect within the classroom.

Far more to the point, no academic department seeks maximum diversity when engaging in hiring. No political science department, for example, would hire a "political astrologer;" indeed, no reputable department would award a Ph.D. to a candidate whose dissertation sought to prove that Geminis tend to engage in interestingly different political behavior than Libras. Academic "disciplines" are just that--highly structured ways of perceiving, and then teaching about, the world--and woe unto those who try to break free of disciplinary bonds by rejecting fundamental presuppositions. The range of "diverse" arguments acceptable within a classroom, around a seminar table, or at an academic convention is always far more restricted than what is acceptable within the quite literally "undisciplined" public square, where even "craziness" has its rights. The point of disciplined education is precisely that whatever one's legal right to hold any opinion he or she wishes, there is no such right within an academic setting, where opinions beyond the pale are properly subject to the sanction of flunking and ultimately coerced withdrawal from the academic community. Chairman Mao made famous, in the 1960s, his ostensible desire to see "100 flowers bloom" within the Chinese garden of political debate. No academic department with which I am familiar would endorse Mao's call (nor, of course, did he take his own slogan seriously). Even a mythical department that did want to present even a dozen points of view would be limited by practical budgetary considerations, and it would always be the case that one would rank order the particular points or perspectives one thought most important.

We cannot leave this broad tour of academic institutions without paying attention to some important goals beyond "pure" education. There are important civic values attached to participation in academic life. Bowen and Bok, for example, emphasize the importance of diverse student bodies to produce "greater `cultural awareness across racial lines . . .' and stronger commitments to improving racial understandings."[125] It is hard to overestimate the importance of "improving racial understanding" in contemporary American society. But is it only greater "racial understanding" that we need, or do we also need, say, greater understanding of those with what are deemed unusual religious views? Indeed, to be fair to Bowen and Bok, it is precisely on the page from which I am quoting that they remind their readers of the "importance of differences in religion."[126]

It is at this point that we can return full circle to Coca-Cola and Chrysler, for it should be obvious, in a country where even now millions of people do not go to college at all, and most go for only four years, that the reputed social benefits of "diversity" in bringing about "racial understanding" could presumably be gained from enhanced "diversity" in other institutions as well. Cynthia Estlund has recently suggested just such an argument.[127] She begins with the obvious fact that the setting within which most people spend most of their out-of-home life is the workplace. She quotes Akhil Reed Amar's and Neal Katyal's eloquent reminder that "[i]f a far-flung democratic republic as diverse--and at times divided--as late twentieth century [sic] America is to survive and flourish, it must cultivate some common spaces where citizens from every corner of society can come together to learn how others live, how others think, how others feel."[128] They conclude, "If not in public universities, where? If not in young adulthood, when?"[129] Although she gladly supports diversity within universities, she notes that the university is not the "only such space"[130] in which diversity is important. Indeed, it is almost certainly not even the most important such space, at least once we move beyond "young adulthood." It is the workplace that encourages (or forces) people to interact with others significantly different from themselves, certainly more so than, say, churches, bowling leagues, or other institutions beloved by latter-day celebrants of "civil society." That is, the principal defense of affirmative action in the Coca-Cola's and Chrysler's of the American economy is more the importance to America's civic health than the not-altogether-plausible argument that it necessarily enhances the quality of the products produced by these companies.[131]

Moreover, as noted earlier, Coke's genuine diversity interests, in terms of its worldwide business empire, might still yield a quite different actual workforce mix than an interest in achieving a more just, less socially- and politically-divided America.

To the extent that we adopt the Bowen-Bok (or Estlund) civic-education argument for diversity, we should note, at the very least, that its ramifications go well beyond institutions of higher education. Indeed, if the consequences are as important as they suggest, then policies promoting "strong diversity" should become far more pervasive than is now the case. There is no good reason to think that limiting "diversity" to a relatively few elite institutions of higher education will suffice to generate the promised benefits. Still, one must always ask about how much the instrumental goals of any given institution--whether it is producing automobiles or producing education--must be sacrificed to the political goals of civic education. And one must recognize that particular arguments for diversity offered for institution X may not at all be the same as those offered in regard to institution Y.


My conclusions are really quite modest. First, "diversity" is one of those words, like "equality," "democracy," and "freedom," whose meaning, if not entirely a construct of the speaker, is, nonetheless, significantly ambiguous. Moreover, like those words, "diversity" as a general notion is thought to be a "good thing," though, concomitantly, someone who doesn't share one's own views about the concrete meaning of this good is often subject to dismissive contempt. This is, to put it mildly, unfortunate. One should be able to accept that there are diverse notions of diversity, each of them likely to offer us one valuable perspective on the elephant we are trying to visualize even if none captures the whole of that complex animal. Secondly, because of the very diversity of the world within which we find ourselves, our commitment to diversity is necessarily limited, subject to constraints posed first of all by inevitable scarcity of resources, but imposed as well by a reasonable commitment to other goods besides maximizing the good that may well be available in diversity.



[ * ]This piece represents a written and updated version of the Owen J. Roberts Memorial Lecture presented at the University of Pennsylvania Law School on October 14, 1999. Established in 1956 by the Order of the Coif of the University of Pennsylvania Law School, this lecture series memorializes Owen J. Roberts, Associate Justice of the Supreme Court of the United States from 1930-1945 and Dean of University of Pennsylvania Law School from 1948-1951. The Order of the Coif, the Law Alumni Society, and the Law School currently sponsor the Roberts Lecture with financial support from Montgomery, McCracken, Walker & Rhoads, LLP.

The Journal of Constitutional Law appreciates the opportunity to publish this year's Owen J. Roberts Memorial Lecture. We owe a debt of gratitude to interim Dean Charles W. Mooney, Jr., the University of Pennsylvania Law Review, and of course to Professor Levinson for their participation in this grand tradition.

[**]W. St. John Garwood and W. St. John Garwood Jr. Regents Chair in Law, University of Texas Law School; Visiting Professor of Law (1999-2000), New York University School of Law. An earlier version of this Lecture was delivered at Skidmore College, on April 8, 1999, as the Fiscus Lecture, which memorializes Ronald J. Fiscus, the author of a very interesting book, THE CONSTITUTIONAL LOGIC OF AFFIRMATIVE ACTION (1992), the topic of which is obviously relevant to the topic of this essay. I am especially grateful to Professor Beau Breslin for that invitation and for his hospitality on the occasion. The present, considerably rewritten, version (though it retains some of the informality of the lecture venue within which it was initially presented) reflects further thinking provoked by questions and comments offered by the Skidmore and Penn audiences, as well as by responses from Jack Balkin, Cindy Estlund, Sam Issacharoff, Rick Pildes, Robert Post, Scot Powe, John Rosenberg, Gerald Torres, and Eugene Volokh to earlier drafts of this essay. I am grateful to all who have tried to help me think through this difficult topic. Given the volatility of the topic, it is more important than ever to issue a blanket exculpation to anyone named above with regard to any errors (or simply dubious propositions) that readers may find in what follows.

I should also note that I prepared the text of the Roberts Lecture before having the opportunity to read George Sher, Diversity 28 PHIL. & PUB. AFF. 85 (1999), and I thank William Ewald, who was a wonderful host during my visit to Penn, for bringing the essay to my attention. There are obvious overlaps in our arguments. In particular, I agree with Professor Sher, and will argue below that most "affirmative action" programs, especially in university settings, are motivated far more by a desire "to benefit members of previously wronged groups" than by a desire to achieve "diversity" per se. Id. at 104. (Or it may simply be, and this is much the same thing, that there is no coherently articulated notion of "diversity" that justifies the actual practices of universities).


[2] See Thomas v. Review Bd. of the Ind. Employment Sec. Div., 450 U.S. 707, 721 (1981) (Rehnquist, J. dissenting) (arguing that one of the main causes of the tension between the Establishment and Free Exercise Clauses is the existence of social welfare benefits, such as unemployment compensation).

[3] 339 U.S. 629 (1950). See also Sweatt v. Painter Archival and Textual Sources (last modified April 3, 1999) <http://ccwf.cc.utexas.edu/~russell/seminar/sweatt/sweattindex.html> (providing a comprehensive set of materials on the case).

[4] 347 U.S. 483 (1954).

[5] Sweatt, 339 U.S. at 634.

[6] Id.

[7] 438 U.S. 265 (1978).

[8] Id. at 274, 279 (discussing the racial criteria for the special admissions programs).

[9] See id. at 412 (Stevens, J. dissenting) (relying on Title VI of the Civil Rights Act of 1964).

[10] See id. at 269 (Powell, J.).

[11] See id. at 287 (noting the legislative intent behind the Fourteenth Amendment was to give blacks the same rights and opportunities as whites).

[12] This is not the occasion for a discussion of whether the "goals" versus "quota" distinction makes all that much sense. I am inclined to think not.

[13] See id. at 311-12 ("[Diversity] clearly is a constitutionally permissible goal for an institution of higher education.").

[14] See id. at 307-08.

[15] See id. at 310 (rejecting the State's claim that its program would better serve under-served communities).

[16] Id. at 316 (quoting the Harvard College Admissions program).

[17] Id. at 323.

[18] Id.

[19] See, e.g., Shaw v. Reno, 509 U.S. 630 (1993) (O'Connor, J.). Justice O'Connor states:

A reapportionment plan that includes in one district individuals who belong to the same race, but who are otherwise widely separated by geographical and political boundaries, and who may have little in common with one another but the color of their skin, bears an uncomfortable resemblance to political apartheid. It reinforces the perception that members of the same racial group--regardless of their age, education, economic status, or the community in which they live--think alike, share the same political interests, and will prefer the same candidates at the polls. We have rejected such perceptions elsewhere as impermissible racial stereotypes.

Id. at 647. I have no quarrel with Justice O'Connor's reminder that "members of the same racial group" do not "think alike" or "prefer the same candidates at the polls." But, of course, she cites no one who does believe this. I do have a bit of a quarrel with her notion that members of racial groups do not even share common "political interests," though agreement or disagreement undoubtedly depends on precisely what one means by shared "political interests." It may be, for example, that both the insurance industry and medical patients have a strongly shared "interest" in the future of American health policy, though it would be surprising indeed if the industry and patients necessarily shared specific opinions as to how best to structure the medical services industry. If Justice O'Connor's notion of a "share[d]" political interest is the latter, then, again, I know of no one who suggests that all members of given racial or ethnic groups agree with one another on how to resolve pressing political issues, so she is criticizing a non-existent opinion. If, on the other hand, she is suggesting that there is no "share[d]" interest in the former sense, then, I respectfully suggest, as argued below, that her view would be extraordinarily foolish. See generally discussion infra pp. 20-22.


[21] See, e.g., GERALD N. ROSENBERG, THE HOLLOW HOPE: CAN COURTS BRING ABOUT SOCIAL CHANGE? 39 (1991) (espousing a skeptical view on this matter).

[22] See generally J. M. Balkin & Sanford Levinson, Constitutional Grammar, 72 TEX. L. REV. 1771 (1994).

[23] See, e.g., NLRB v. Jones & Laughlin Steel Corp., 301 U.S. 1 (1937); United States v. Darby, 312 U.S. 100 (1941).

[24] See, e.g., Carter v. Carter Coal Co., 298 U.S. 238 (1936).

[25] But see PETER BRIMELOW, ALIEN NATION: COMMON SENSE ABOUT AMERICA'S IMMIGRATION DISASTER (1995) (expressing the view that the United States is fundamentally changing for the worse because of changes in immigration law that are letting in far too many non-Anglo-Saxons). One has no doubt that Brimelow would oppose the Immigration and Naturalization Service's policy of "diversity admission" to the United States, see infra note 51.

[26] See, e.g., Peter T. Kilborn, Jeb Bush Roils Florida on Affirmative Action, N.Y. TIMES, Feb. 4, 2000, at A1 (detailing a tumultuous public hearing generated by the Governor's attempt to end race- and ethnic-based preferences). "`We are embracing diversity, not rejecting it,' Mr. Bush, the first speaker at today's hearing, told the mostly black audience. `This plan will create more opportunity for people.'" Id. According to the article,

There is widespread support among whites for Mr. Bush's program, which would end preferences for businesses owned by women and minorities in bidding for state contracts. And it would end college admissions preferences based on race, substituting a program guaranteeing admission to at least 1 of the 10 state universities for high school students who graduate in the top 20 percent of their class.

. . .

Some black leaders see promise in the One Florida plan, in that it might create opportunities for the poor, disproportionate numbers of whom are black. But they bristle at the governor's abolition of affirmative action . . . .



[28] See, e.g., ERIC FONER, THE STORY OF AMERICAN FREEDOM (1998) (limning the quite different, and often conflicting, meanings assigned to the notion of "freedom" throughout American history).

[29] DOUGLAS RAE ET AL., EQUALITIES (1981). See also PETER WESTEN, SPEAKING OF EQUALITY: AN ANALYSIS OF THE RHETORICAL FORCE OF `EQUALITY' IN MORAL AND LEGAL DISCOURSE (1990) (arguing that equality is basically an "empty" concept, taking meaning only from a prior theory of rights or fairness that allows one to assert that certain obvious differences among human beings (skin color, height, weight, age, beauty, LSAT scores, etc., etc.) may, or may not, be taken into account in distributing the burdens and benefits of living in American society). I am not sure whether Professor Westen, a member of the distinguished law faculty at the University of Michigan that is defending its own affirmative action program against constitutional attack on the ground of its contribution to diversity, would argue that "diversity" is a similarly "empty" concept.

I note that one of Professor Westen's colleagues has also written scholarly articles on diversity that display considerable skepticism about the analytical premises behind at least some versions of the concept. See Deborah C. Malamud, Affirmative Action, Diversity, and the Black Middle Class, 68 U. COLO. L. REV. 939 (1997); see also Deborah C. Malamud, Values, Symbols, and the Facts in the Affirmative Action Debate, 95 MICH. L. REV. 1668 (1997) (book review) [hereinafter Malamud, Values, Symbols, and the Facts]. Again, I do not know what particular position Professor Malamud takes with regard to the Law School's assertion of diversity as the primary defense of its admissions policy.

[30] 78 F.3d 932 (5th Cir. 1996). See also generally Reva B. Siegel, The Racial Rhetorics of Colorblind Constitutionalism: The Case of Hopwood v. Texas, in RACE AND REPRESENTATION: AFFIRMATIVE ACTION 29 (Robert Post and Michael Rogin, eds., 1998).

[31] This is in fact a somewhat oversimplified description, but it does not affect the central argument.

[32] See Hopwood v. Texas, 861 F. Supp. 551 (W.D. Tex. 1994). This process replaced the one under which Ms. Hopwood herself applied and which the University was no longer willing to defend. See id. at 560.

[33] See Hopwood, 78 F.3d at 962.

[34] See id. at 945-46.

[35] See id. at 951.

[36] See U.S. CONST. art. III, § 1 (distinguishing between the Supreme Court and such "inferior" courts as Congress may choose to establish). I want to make it clear that I mean nothing disrespectful by reference to "inferior" courts, for I have consistently been critical of many of the claims to supremacy by the "supreme" court. See, e.g., SANFORD LEVINSON, CONSTITUTIONAL FAITH 27-53 (1988) (describing and defending an institutionally "protestant" approach to the Constitution that rejects the claim of the Supreme Court to "ultimate" authority over constitutional meaning); see also Sanford Levinson, On Positivism and Potted Plants: `Inferior' Judges and the Task of Constitutional Interpretation, 25 CONN. L. REV. 843 (1993)(defending the propriety of "inferior" judges thinking for themselves when engaging in constitutional interpretation); Sanford Levinson, Hopwood: Some Reflections on Constitutional Interpretation by an Inferior Court, 2 TEX. F. ON CIV. LIBERTIES & CIV. RTS. 113-122 (1996) (refusing to criticize Fifth Circuit for departure from Supreme Court precedent).

[37] Hopwood, 78 F.3d at 945 (emphasis added).

[38] 163 U.S. 537, 559 (1896) (Harlan, J. dissenting) ("Our Constitution is color-blind, and neither knows nor tolerates classes among citizens.").

[39] Hopwood, 78 F.3d at 946.

[40] It is worth noting that such a policy would be "unconstitutional" in the new Europe that is structured by the norms of the Treaty of Rome and subsequent treaties. Nationals of any of the countries within the European Union have a right of equal access to the public universities operated by any of the members. To this extent, at least, Europe in only fifty years has become a more genuinely "united" entity than has the United States of America after 210 years. See generally David S. Clark, Transnational Legal Practice: The Need for Global Law Schools, 46 AM. J. COMP. L. 261, 265 (Supp. 1998) (discussing student mobility in the European Union as a result of programs such as ERASMUS and SOCRATES).

[41] Hopwood, 78 F.3d at 946.

[42] Id.

[43] Id.

[44] See LYDIA CHáVEZ, THE COLOR BIND: CALIFORNIA'S BATTLE TO END AFFIRMATIVE ACTION 2 (1998) ("[Proposition 209] promised to end the use of race and gender preferences in state employment, contracting, and education.").

[45] As a matter of empirical fact, it is not clear that Hopwood has had anything near the impact in Louisiana and Mississippi that it has had in Texas. Possible reasons for this differential impact are beyond the scope of this essay.

Moreover, the issue is currently before a United States district court within the Sixth Circuit, in which the University of Michigan Law School is attempting to defend its race- and ethnicity-sensitive admissions procedures against statutory and constitutional attack. See Gratz v. Bollinger, 183 F.R.D. 209 (E.D. Mich. 1998) (deciding technical procedural issues); Grutter, v. Bollinger, 16 F. Supp.2d 797 (E.D. Mich. 1998) (same). Not surprisingly, the essence of the University's defense is the importance of "diversity," as evidenced by the title the University gave to a compilation of reports by various experts submitted to the Court, "The Compelling Need for Diversity in Higher Education." The compilation and all other legal documents are available at University of Michigan, Information on Admissions Lawsuits (visited Jan. 20, 2000) <http://www.umich.edu/~urel/admissions/legal>.

[46] A bibliography search of legal journals in spring 1999 produced ninety-one articles since 1990 with the word "diversity" in the title (and this did not include articles on "biodiversity"). No doubt a search now, in January 2000, would produce many more articles and a general literature review would almost certainly produce thousands of titles.


[48] Id. at 252.

[49] Id. at 12 (quoting M. Douglas Ivester, Chairman and CEO of The Coca-Cola Company).

Given the role that Coca-Cola plays in this lecture, it is probably advisable to note a recent article, Constance L. Hays, New Coke Chief Says Diversity Is High Priority, N.Y. TIMES, Mar. 10, 2000, at C2.

In a striking move intended to counter accusations that the Coca-Cola Company has not cared enough about building a diverse work force, Coke's chairman and chief executive said yesterday that management compensation at the company, including his own, would be closely tied to diversity goals that have yet to be announced.

Id. Ms. Hays writes, "white men control most of the power in the company" and suggests that the announcement was in response to a recent suit filed by "eight current and former black employees who say they were unfairly treated because of their race." Id. "A lawyer for the plaintiffs called the announcement `a spectacular victory' and said it would improve conditions at the company. `This is the first step toward moving Coca-Cola from a laggard to a leader on diversity.'" Id. And, according to a stock analyst with Salomon Smith Barney, "[t]he most important thing Coke needs to do is mend its global public image. And if this helps in that respect, the shareholders are going to benefit." Id.

[50] BOWEN & BOK, supra note 47, at 12.

[51] Perhaps the most intriguing example that I discovered in the course of my inquiry into the topic is the presence within American immigration law of explicitly titled "diversity admissions" of immigrants who would not otherwise qualify for admission (but do not possess attributes that would disqualify them for admission). The apparent origin of the program is the congressional reaction to some unforeseen consequences of the demise of the national-origins-quota system of immigration in 1965, which had been justifiably criticized for trying to preserve a too-homogeneous America. See THOMAS ALEXANDER ALEINIKOFF ET AL., IMMIGRATION AND CITIZENSHIP: PROCESS AND POLICY 290-92 (4th ed. 1998) (discussing legislative attempts to increase diversity in immigration). (I am grateful to David Martin, one of the co-editors of this excellent casebook, for making me aware of this.) Among these consequences was a "steep reduction" in those granted permission to immigrate from Europe, particularly from Ireland, whose nationals had been strong beneficiaries of the national-origins system established in 1924. See id. at 291. Thus Congress in 1986 provided for 10,000 new admissions over two years "for persons from countries `adversely affected' by the 1965 changes." Id. Not surprisingly, "[t]he list of 36 such countries was disproportionately, but not exclusively, European." Id. Even though the lucky admittees were selected by lottery, "Irish nationals turned out to be the biggest winners." Id. The system was changed again in 1988, this time preferring persons from "`underrepresented countries,'" with winners this time being natives of Bangladesh, Pakistan, Poland, Turkey, and Egypt. Id. Congress in 1990 changed the Act again, this time settling on "an extraordinarily intricate formula" to allocate the 55,000 annual "diversity admissions;" a random lottery selects the winners from among those who are eligible. Id. Current winners appear to be persons from Bangladesh, Ghana, Nigeria, Bulgaria, Sierra Leone, Romania, Ukraine, and Albania. See id. at 292. As can easily be seen, the most recent version of the "diversity admissions" program was adopted during the Reagan-Bush era and cannot therefore be easily ascribed, as one might have suspected, to the return to Washington of Democrats.

I trust I am not alone in finding this entire program, and its political origins, fascinating, not least because the operative theory of the program indeed appears to be that national diversity, with regard to those immigrating to the United States, is itself a positive good, at least within the limited numbers. One can imagine a number of different reasons why this might be so, ranging from the assumptions that Bulgarians bring with them importantly different attributes from, say, Poles or Hungarians (who are not favored by the "diversity admissions" program) to the more self-interested view that the United States benefits from having within its borders members of all countries with which the United States interacts so that they can transmit to their friends and relatives back home a desirable picture of the United States and of its motives as an international actor.

[52] Professor of Government, Harvard University.

[53] A. Dalton Cross Professor of Law, University of Texas Law School.

[54] For a description of Mr. Connerly's achievements in campaigning against affirmative action see generally Edward W. Lempinen, Connerly Widens Anti-Affirmative Action Campaign, THE S.F. CHRON., Jan. 16, 1997, at A17 ("Connerly formally opened a national campaign against race- and gender-based affirmative action."); Tony Perry, California and the West; Connerly to Lead GOP Fund-Raising, L.A. TIMES, June 25, 1998, at A3 ("Ward Connerly [is] the controversial leader of the drive that ended racial preferences in state hiring and college admissions.").

[55] See David A. Thomas & Robin J. Ely, Making Differences Matter: A New Paradigm for Managing Diversity, HARV. BUS. REV., Sept.-Oct. 1996, at 79, 79-90.

[56] See id. at 81.

[57] See generally Sanford Levinson, Identifying the Jewish Lawyer: Reflections on The Construction of Professional Identity, 14 CARDOZO L. REV. 1577 (1993) [hereinafter Levinson, Identifying the Jewish Lawyer].

[58] Thomas & Ely, supra note 55, at 81.

[59] Id.

[60] Id.

[61] See Samuel Issacharoff, Bakke in the Admissions Office and the Courts: Can Affirmative Action Be Defended?, 59 OHIO ST. L.J. 669, 677 (1998) (arguing that diversity as a criterion in hiring or admissions decisions must be intended to promote institutional goals, rather than aiming for "racial balance for its own sake," in order to be held constitutional).

[62] See supra note 49 and accompanying text.

[63] BOWEN & BOK, supra note 47, at 12 (quoting M. Douglas Ivester, Chairman and CEO of The Coca-Cola Company).

[64] A website illustrating corporate mistakes in international presentation of products includes the following wonderful example:

When Coca-Cola first shipped to China, they named the product something that when pronounced sounded like "Coca-Cola". The only problem was that the characters used meant "Bite the wax tadpole", or "female horse stuffed with wax", depending on the dialect. They later changed to a set of characters, with a close phonetic equivalence, that mean "Happiness in the mouth".

Ronald West Frye, Bite the Wax Tadpole, (visited Jan. 29, 2000) <http://www.geocities.com /Heartland/Valley /8522/tadpole.htm>. I am grateful to Tom Russell for informing me of this website.

[65] An especially important area in which this might be true is medicine. See Kenneth DeVille, Defending Diversity: Affirmative Action and Medical Education, 89 AM J. PUB. HEALTH 1256, 1259 (1999) (citing inter alia M. Garb, Like Doctor Like Patient: More and More Patients Choose Physicians of Same Ethnic Group, Sexual Preference, 15 AM. MED. NEWS 17 (Sept. 1992)). DeVille writes that "[p]hysicians are less effective at treating patients who do not trust them," id. at 1260, and it appears to be the case that African-Americans in particular are mistrustful of the quality of care received from white doctors.

For example, African Americans are the patient group most likely to distrust the care they will receive at the end of life. By influencing such factors as patterns of compliance, preventive care, and patient disclosure of information and choice of therapies, such distrust can have a substantial impact on the care that minority patients receive. This suspicion is so deep-seated and widespread that, in the short term, the only remedy is to provide minority patients with physicians with whom they feel safe and comfortable.

Id. (citations omitted).

[66] Thomas & Ely, supra note 55, at 83.

[67] Id.

[68] Id.

[69] See generally Stuart Elliott, Ads Speak to Asian-Americans, N.Y. TIMES, Mar. 6, 2000, at C1. "So-called diversity marketing--pursuing customers based on differentiations like race, language, sexual orientation and age--has become a lucrative strategy for mainstream advertisers." Id. Many companies are having difficulty penetrating a seemingly lucrative niche. "`Frankly, many marketers really don't think about this market because they don't have many Asian-Americans on their staffs,' said Alfred L. Schreiber, president at Diversity Imperatives . . . `That lack of representation in corporate America puts this group at a disadvantage.'" Id.

[70] This seems to be especially true if classes are selected by the use of certain kinds of "tracking" programs, but it can also co-exist with free choice in classes.

[71] BOWEN & BOK, supra note 47, at 12 (quoting Robert J. Eaton, Chairman and CEO of Chrysler Corporation).


[73] See New York Times Co. v. Sullivan, 376 U.S. 254, 270 (1965).

[74] See Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 311 (1978).

75] Several people who responded to this lecture as delivered in Philadelphia described me as (or accused me of being) very "conservative" insofar as I seemed to accept the validity of organizational imperatives against the expressive interests of the individuals who constitute the organization. There is, I suppose, a certain core of truth in the charge, though I am not sure that "conservative" is the mot juste. It seems to me that almost anyone with a sociological bent both recognizes, as a descriptive matter, that we live our lives playing certain kinds of varying social roles and accepts, normatively, the legitimacy of the role constraints on many, if not all, occasions. I am more than happy to define certain role conceptions or organizational strictures as too confining and, therefore, to support opening up space for the expression of added facets of the personalities of the organization's actors. I am much too much a child of the 1960's to believe that we ought necessarily to accept organizational roles, whether public or private, without question. That being said, I also cannot imagine what life would be like without a sense of some institutional constraint, and I am certainly strongly supportive in many instances of organizational actors "knowing their place," as it were, and realizing that behavior that would be perfectly appropriate outside the organizational structure is completely inappropriate within it. Taking the example in the text, I would indeed be dismayed if postal workers attempted to convert me or otherwise engage in religious conversation when I was trying to buy stamps, though I have no dismay, only disagreement, if a Jehovah's Witness or an Evangelical Christian tries to do the same as I walk through Washington Square. I simply would not take anyone seriously who rejected in toto, descriptively or normatively, a role-oriented analysis of human behavior. And I certainly would not want to spend much time around such a person.

[76] Katherine Y. Williams & Charles A. O'Reilly, III, Demography & Diversity in Organzations: A Review of 40 Years of Research, 20 RES. IN ORGANIZATIONAL BEHAV. 77, 120 (1998). (I am indebted to Cynthia Estlund for this reference.) Similar skepticism is expressed in Eugene Volokh, Diversity, Race as Proxy, and Religion as Proxy, 43 UCLA L. REV. 2059, 2060-2063 (1996) (citing, inter alia, Stephen M. Bainbridge, Community and Statism: A Conservative Contractarian Critique of Progressive Corporate Law Scholaraship, 82 CORNELL L. REV. 856 (1997); Robert Drago, Share Schemes, Participatory Management and Work Norms, 23 REV. OF RADICAL POL. ECON. 55, 58-59 (1991) (discussing gender homogeneity); Donald McCloskey, Bourgeois Virtue, AM. SCHOLAR 177, 183-84 (1994) (discussing ethnic and religious homogeneity)).

[77] See NANCY J. ADLER, INTERNATIONAL DIMENSIONS OF ORGANIZATIONAL BEHAVIOR 39-62 (2d ed. 1991) (noting that work behaviors of individuals vary across cultures).

[78] Id. at 43-45 (citing André Laurent, The Cultural Diversity of Western Conceptions of Management, 13 INT'L STUD. OF MGMT. & ORG. 75, 86 (1983)).

[79]Id. at 44.

[80] Id. at 43-44.

[81] Id. at 44.

[82] Id.

[83] Thomas & Ely,supra note 55, at 80.

[84] Id. at 85.

[85] Id.

[86] Arnold H. Loewy, Taking Bakke Seriously: Distinguishing Diversity from Affirmative Action in the Law School Admissions Process, 77 N.C. L. REV. 1479, 1488 (1999).

[87] One must always be wary, though, of assuming that members of given minorities will conform to some pre-existing stereotype that will provide the desired "diversity." Thus Deborah Malamud writes:

The most conventional arguments for diversity are arguments that each member of a group is a representative of that group's cultural characteristics and viewpoints, and that the institution (or polity) is enriched by bringing these different cultures under the institutional umbrella. The problem with these arguments is that . . . they script the lives of diversity hires [or students admitted under "diversity preferences"]. For example, the black college student who is a jazz musician but switches to classical piano can be seen as ceasing to do his "job" for the institution . . . .

Deborah C. Malamud, Values, Symbols, and the Facts, supra note 29, at 1691.

[88] I take it that very few undergraduate admissions preferences are program specific, if for no other reason than the admissions department has only the slightest idea of what the admitted students will ultimately study. In addition, much of the most important education at the undergraduate level goes on outside the classroom, in long conversations that may have little to do with one's formal studies. The situation is quite different with graduate programs. Although one assumes that graduate students will indeed have conversations about what Justice Cardozo called "life in all its fullness," that is not what is truly bringing them to graduate study, nor is it (often) an important admissions criterion on the part of the admitting department. One is interested above all in who will make the best physicist, historian, zoologist, or logician.

[89] See, e.g., University of Michigan, Expert Report of Patrcia Gurin: Summary and Conclusions (last visited January 30, 2000) <http://www.umich.edu/~urel/admissions/legal/expert/summ.html>. Gurin states:

It is clear . . . that interaction with peers from diverse racial backgrounds, both in the classroom and informally, is positively associated with a host of what I call "learning outcomes." Students who experienced the most racial and ethnic diversity in classroom settings and in informal interactions with peers showed the greatest engagement in active thinking processes, growth in intellectual engagement and motivation, and growth in intellectual and academic skills.

Id. (emphasis added).

[90] Hopwood v. Texas, 78 F.3d 932, 965-66 (5th Cir. 1996) (Wiener, J., concurring).

[91] Others might name Tibet as the largest colony, though China has attempted to absorb Tibet into the legal structure of the Peoples' Republic of China, in contrast to the decided non-absorption of Puerto Rico into the legal entity called The United States of America.

[92] See Downes v. Bidwell, 182 U.S. 244 (1901).

[93] See PROCESS OF CONSTITUTIONAL DECISION MAKING: CASES AND MATERIALS 297 (Paul Brest et al., eds., 4th ed. 2000).

[94] The same is true, incidentally, with regard to Hispanics from Central or South America.

[95] Arthur Levine, The Soul of A New University, N.Y. TIMES, Mar. 13, 2000, at A21.

[96] Hopwood v. Texas, 78 F.3d 932, 946 (5th Cir. 1996) (quoting Metro Broad., Inc. v. FCC, 497 U.S. 547, 602 (1990) (O'Connor, J., dissenting)) (emphasis added).

[97] Id. at 945.

[98] During the discussion following my presentation of the Roberts Lecture, Professor Regina Austin took vigorous exception to my (then) implicit premise that students in effect serve as their professors' educational "instruments," to be used to develop points the professor believes germane to the topic of the course. I plead guilty to this charge and am happy to make the point explicit. I believe that teachers often, and properly, do attempt to draw on the experiences of their students in order to illuminate matters under discussion. If I were teaching a course on the Vietnam War, for example, I think it would certainly be beneficial to have within the classroom, say, veterans from the United States armed forces, persons from both North and South Vietnam, and an anti-war activist. Or if I were teaching a course on sports law, I think it would be useful to have as a member of the class a professional tennis or basketball player, and so on. I do agree with Professor Austin that, generally speaking, students come to universities to be educated by their professors rather than to serve as educational resources for their professors, but I do not see these as dichotomous alternatives. One need not believe that students should be appointed full professors in order to recognize that even the most untutored first-year student may, because of his or her own life experiences, have something valuable to contribute to a particular classroom discussion. A good professor will, of course, warn students about the fallacy of extrapolating from individual events, or "anecdotes," but a good professor will also scarcely ignore the valuable information that can be provided by looking at specific examples.

[99] Loewy, supra note 86, at 1489.

[100] Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 323 (1977) (emphasis added).

[101] See HARVARD ENCYCLOPEDIA OF AMERICAN ETHNIC GROUPS vi (Stephen Thernstrom, ed. 1980). Some of these groups, to be sure,

now exist almost exclusively in the recollections of their descendants . . . . On the other hand, some groups have not yet been in the United States long enough to establish generational continuity or to develop the array of institutions conventionally associated with ethnic groups, but they seem to be in the process of formation.

Id. One suspects that a new edition of the Encyclopedia would contain entries on the Hmong, Vietnamese, Iraqis, and Palestinians, only four of the groups absent from the original edition that are present in significant numbers in one or another major American city.

[102] Miranda Oshige McGowan, Diversity of What?, in RACE AND REPRESENTATION: AFFIRMATIVE ACTION, supra note 30, at 237, 242.

[103] There are other notorious problems with the category "Asian," of course, given the extension of the Asian land mass from Istanbul to Vladivostock. Are Iranians or Siberian Russians "Asian"? Any negative answer must rest on something more than a geographical notion of identity.

[104] See generally DAVID HOLLINGER, POSTETHNIC AMERICA: BEYOND MULTICULTURALISM (1995); Lawrence Wright, One Drop of Blood, THE NEW YORKER, July 25, 1994, at 46; Christopher A. Ford, Administering Identity: The Determination of "Race" in Race-Conscious Law, 82 CAL. L. REV. 1231 (1994).

[105] See Orlando Patterson, Race Over, THE NEW REPUBLIC, Jan. 10, 2000, at 6.

[106] 448 U.S. 448, 532 (1980) (Stevens, J. dissenting).

[107] See, e.g., Metro Broad., Inc. v. FCC, 497 U.S. 547, 601 (1990) (Stevens, J., concurring). Stevens states:

Specifically, the reason for the classification--the recognized interest in broadcast diversity--is clearly identified and does not imply any judgement concerning the abilities of owners of different races or the merits of different kinds of programming . . . . The public interest in broadcast diversity--like the interest in an integrated police force, diversity in the composition of a public school faculty or diversity in the student body of a professional school--is in my view unquestionably legitimate.

Id. at 601-02

[108] See Patterson, supra note 105, at 6.

[109] See Hopwood v. Texas, 78 F.3d 932, 962 (5th Cir. 1996) (Wiener, J. concurring).

110] Letter from Jack Balkin, Professor, Yale Law School, to Sanford Levinson, Professor, University of Texas Law School, (Mar. 18, 1999) (electronic mail responding to an earlier version of the present Lecture, on file with University of Pennsylvania Journal of Constitutional Law).

[111] Id.


[113] Id.

[114] See Kathleen M. Sullivan, Sins of Discrimination: Last Term's Affirmative Action Cases, 100 HARV. L. REV. 78 (1986).

[115] Bowen and Bok

estimate that only about 20 to 30 percent of all four-year colleges and universities [have enough applicants to be able to pick and choose among them.] Nationally, the vast majority of undergraduate institutions accept all qualified candidates and thus do not award special status to any group of applicants, defined by race or on the basis of any other criterion.

BOWEN & BOK, supra note 47, at 15.

[116] Id. at 228.

[117] This topic most naturally arises in any constitutional law course, but I should note that I also raise the issue of "personal," as opposed to "professional," identity in my course on the legal profession. See Levinson, Identifying the Jewish Lawyer, supra note 57, at 1577.

[118] See generally Volokh, supra note 76, at 2070-76.

[119] U.S. CONST. amend. I.

[120] I note that Detroit has a very large concentration of Arab-Americans (which may itself be an "overly-inclusive" categorization collapsing important differences between, say, Iraqis and Lebanese). And, of course, not all Arab-Americans are in fact Muslim. Still, one might consider a recent article by Gary Lee in the Travel section of the Washington Post, which had the following headline: Not Your Father's Detroit; The Motor City has become the unlikely capital of Arab America. Visit its streets, lined with ethnic shops, restaurants and houses of worship, for a stateside taste of the Middle East, WASH. POST, Jan. 9, 2000, at E1. Lee states that:

According to Zogby International, a national pollster, 275,000 Middle Easterners--mostly Arab Americans have settled in Detroit, Dearborn and other surrounding towns. It's the biggest concentration of Arab Americans in the country, and is tied with the Hispanic population as the second-largest ethnic group in southern Michigan, after African Americans.

Id. I wonder how much effort the University of Michigan has made to assure a "representative" number of students from this group, including, of course, "representatives" of very important religious strains within it. It would obviously be anomalous if all Arab-American students at the University happened to be secular and, perhaps, willing to reinforce anti-sectarian stereotypes directed at "Islamic fundamentalists" in the same say, for example, that "mainstream" Christians or non-Orthodox Jews may gladly demean Christian fundamentalists or Orthodox (and, certainly, Hassidic) Jews.

I cannot claim to have read all of the Michigan evidence, but I note that Professor Gurin's submission emphasizes "racially/ethnically different student populations (African American, White, and Latino Students)." See University of Michigan, Expert Report of Patricia Gurin: Empirical Results from the Analyses Conducted for This Litigation (visited Feb. 20, 2000) <http://www.umich.edu/~urel/admissions/legal/expert/empir.html>. Perhaps even more relevant is the extensive statement prepared by Professor Thomas J. Sugrue, who notes in his "statement of qualifications" that his book, THE ORIGINS OF THE URBAN CRISIS: RACE AND INEQUALITY IN POSTWAR DETROIT (1996) won four major awards, including the extremely prestigious Bancroft Prize in American History. See University of Michigan, Expert Report of Thomas J. Sugrue (visited Feb. 20, 2000) <http://www.umich.edu/~urel/admissions /legal/expert/sugrutoc.html> ("My book and a number of my articles discuss race relations and inequality in Michigan, with close attention to metropolitan Detroit."). It is, therefore, especially startling to realize that Professor Sugrue apparently does not regard Arab-Americans as an "ethnic group" that should be discussed. See, e.g., id. at Table 2: Michigan Population by Race/Ethnicity, at 21 (dividing the population into "White; Black; American Indian/Eskimo/Aleut/; Asian/Pacific Islander; Hispanic, Mexican; Puerto Rican; Cuban; Other Hispanic;" and, finally, "Other Race."). Professor Sugrue is obviously adopting the classification schema of the United States Census Bureau, which, at best, is highly problematic if not outright incoherent (or worse). See, e.g., HOLLINGER, supra note 104.

I note that Professor Sugrue begins his one-paragraph "conclusion" to his statement by stating, "[i]n an increasingly diverse country, deep divisions persist between whites, blacks, Hispanics, and American Indians." University of Michigan, Expert Report of Thomas J. Sugrue (visited Feb. 20, 2000) <http://www.umich.edu/ ~urel/admissions/legal/expert/sugrutoc.html>, at 47. Yes, this is true. But is he wholly unaware of the deep strains of anti-Arabism and anti-Islamicism that run through American society? I suspect, for example, that at least as many Arab-Americans are the subjects of invasive searches and hostile questioning when attempting to engage in their "right to travel" as, say, American Indians. (Would anyone like to predict what this country would be like had Timothy McVeigh and Terry Nichols turned out to inhabit the identity originally assigned to them, i.e., "Islamic fundamentalists"?)

One sometimes gets the feeling that ostensible defenders of "diversity" and "multiculturalism" have no real idea of how truly diverse and multicultural the United States has become, fixated as they are on the "traditional" racial and ethnic cleavages within this country.

[121] Letter from Jack Balkin, Professor, Yale Law School, to Sanford Levinson, Professor, University of Texas Law School, (Mar. 18, 1999) (electronic mail responding to an earlier version of the present Lecture, on file with University of Pennsylvania Journal of Constitutional Law).

[122] Timothy L. Hall, Educational Diversity: Viewpoints and Proxies, 59 OHIO ST. L.J. 551 (1998).

[123] See, e.g., Dun & Bradstreet, Inc. v. Greenmoss Builders, Inc., 472 U.S. 749, 757 n.4 (1985) (noting "robust debate" to be a "central First Amendment value") (citation omitted).

[124] And my examples have nothing to do with "hate speech" or other standard examples of contemporary controversy in regard to academic speech.

[125] BOWEN & BOK, supra note 47, at 228.

[126] Id.

[127] See Cynthia L. Estlund, Working Together: The Workplace in Civil Society (Feb. 17, 2000) (unpublished manuscript, on file with the University of Pennsylvania Journal of Constitutional Law).

[128] Id. at n. 277 (citing Akhil Reed Amar & Neal Kumar Katyal, Bakke's Fate, 43 UCLA L. REV. 1745, 1749 (1996)).

[129]Id. (citing Akhil Reed Amar & Neal Kumar Katyal, Bakke's Fate, 43 UCLA L. REV. 1745, 1749 (1996)).


[131] Estlund's manuscript offers a treasure trove of citations to the literature about the efficacy of "contact" with those different from oneself in leading to greater tolerance, cross-group friendships, etc. See, e.g., Lee Sigelman, et al., Making Contact? Black-White Social Interaction in an Urban Setting, 101 AM J. OF SOC. 1306, 1307 (1996) ("The idea that familiarity breeds positivity has usually been sustained, particularly when people interact under conditions of relative equality."). To be sure, the evidence is mixed. As Estlund writes, "[e]vidence of the persistence of intergroup bias and friction is nonetheless instructive. It suggests that even constructive interracial contact is unlikely to eliminate prejudice, especially unconscious bias (and that the law needs to reckon with that fact)." Estlund, supra note 127, at 23.