On June 23, 2016, the Supreme Court announced its much-anticipated decision in Fisher v. University of Texas at Austin, 1 136 S. Ct. 2198 (2016). allowing affirmative action in college admissions to continue. No single feature of Fisher surprised court watchers more than its author, Justice Anthony Kennedy. 2 Id. at 2205. As Richard Primus wrote in the New York Times: “[T]he most deceptive thing about it is its first words: ‘Justice Kennedy delivered the opinion of the court.’” 3 Richard Primus, Affirmative Action in College Admissions, Here to Stay, N.Y. Times (June 23, 2016), http://nyti.ms/28PLjIN.
No wonder. Until Fisher, Justice Kennedy had never voted to uphold race-conscious affirmative action. In his 2003 dissent in Grutter v. Bollinger, 4 539 U.S. 306 (2003). he derided the majority for accepting the University of Michigan Law School’s use of the term “critical mass” to justify race-based affirmative action in admissions. 5 Id. at 389 (Kennedy, J., dissenting). No one would have predicted then that he would go on to deliver an opinion upholding affirmative action that invokes Grutter no less than a dozen times. How did Justice Kennedy go from affirmative action dissenter to defender?
Bakke is the key to understanding Justice Kennedy’s supposed reversal in Fisher. Commentators have missed this point, likely because Justice Kennedy’s Fisher opinion does not directly cite Bakke even once. 6 Fisher, 136 S. Ct. at 2205-15. But Justice Kennedy’s Grutter dissent adhered to Justice Powell’s opinion in the 1978 decision Regents of the University of California v. Bakke. 7 438 U.S. 265 (1978). In Bakke, Justice Powell allowed limited use of racial preferences in admissions decisions in the pursuit of a diverse student body so long as it satisfied strict scrutiny. 8 Id. at 320. “The opinion by Justice Powell, in my view, states the correct rule for resolving this case,” Kennedy wrote. 9 Grutter, 539 U.S. at 387 (Kennedy, J., dissenting).
Justice Kennedy admired Justice Powell’s rule in Bakke yet detested its application in Grutter. He cautioned that Grutter diverged from Bakke in ways that would “perpetuate the hostilities that proper consideration of race is designed to avoid.” 10 Id. at 394-95. Fisher presented him with an opportunity to reset the shape and trajectory of affirmative action in line with Bakke.
This Essay demonstrates that we cannot understand the ruling in Fisher without looking at Bakke. Part I surveys the Bakke–Grutter–Fisher line of cases to explain how Grutter diverged from Bakke and why this bothered Justice Kennedy. Part II demonstrates how Fisher aims to preserve an understanding of affirmative action that is informed by Bakke and untainted by Grutter. Not only does Fisher describe affirmative action programs and precedent in ways that maintain fidelity to Bakke, but it also disregards aspects of Grutter that diverged from Bakke, including the use of critical mass as the measure of diversity. Part III identifies the concerns of social cohesion animating Justice Kennedy’s return to Bakke and traces the evolution of those concerns since Bakke. While Bakke’s concern for social cohesion focused on resentment among whites likely to arise from any use of racial preferences, Fisher is not limited in this way. Part IV concludes that universities interested in enrolling a diverse student body would do well to read Justice Powell’s opinion in Bakke, reconsider the use of critical mass to justify race-based affirmative action, and recognize how concerns of social cohesion shape the form of constitutionally permissible affirmative action.
I. Bakke to Fisher
In 1978, the Supreme Court in Bakke invalidated an admissions program at the University of California Davis Medical School that reserved 16 of 100 places in each entering class for “qualified” minorities. 11 Bakke, 438 U.S. at 271, 274-75. Justice Powell approved a university’s use of race in admissions, but only to further one interest: “the attainment of a diverse student body.” 12 Id. at 311-12. He concluded, however, that setting aside a specified number of seats was not an appropriate means to achieve the goal of diversity, 13 Id. at 316. because it failed to “consider all pertinent elements of diversity in light of the particular qualifications of each applicant” 14 Id. at 317. and “treat each applicant as an individual in the admissions process.” 15 Id. at 318.
Twenty-five years later, the Court in Grutter upheld the race-conscious admissions program of the University of Michigan Law School. 16 Grutter v. Bollinger, 539 U.S. 306, 343 (2003). But even though the Court in Grutter endorsed Justice Powell’s opinion in Bakke, Grutter diverged from Bakke in at least three important ways.
First, Grutter declared that the Law School’s policy of admitting a “critical mass” of minority students was a “narrowly tailored use of race.” 17 Id. Justice Powell’s Bakke opinion had endorsed Harvard College’s admissions plan, which acknowledged “some relationship between numbers and achieving the benefits to be derived from a diverse student body” 18 Bakke, 438 U.S. at 323 (emphasis added) (quoting Brief for Columbia Univ. et al. as Amici Curiae at app. 3, Bakke, 438 U.S. 265 (No. 76-811), 1977 WL 188007, at app. *3). but did not give a name or form to that relationship. By contrast, the Court in Grutter approved the Law School’s openly declared interest in “critical mass,” defined as “a number that encourages underrepresented minority students to participate in the classroom and not feel isolated.” 19 Grutter, 539 U.S. at 318.
Second, Grutter predicted that “25 years from now, the use of racial preferences will no longer be necessary to further the interest approved today.” 20 Id. at 343. Writing for the majority, Justice O’Connor introduced this “sunset provision” to avoid “[e]nshrining a permanent justification” for race-conscious admissions policies that “must be limited in time.” 21 Id. at 342. She forecast a term of twenty-five years after noting that “the number of minority applicants with high grades and test scores ha[d] . . . increased” 22 Id. at 343. in the twenty-five years since Justice Powell’s opinion in Bakke.
Third, Grutter expressed the educational benefits of diversity in more explicitly racial terms. Justice Powell’s opinion in Bakke had cast the educational benefits of diversity in largely universal terms, declaring that “the ‘nation’s future depends upon leaders trained through wide exposure’ to the ideas and mores of students as diverse as this Nation of many peoples.” 23 Bakke, 438 U.S. at 313 (quoting Keyishian v. Bd. of Regents, 385 U.S. 589, 603 (1967)). By contrast, Justice O’Connor’s opinion in Grutter expounded the educational benefits of racial diversity in particular, including “promot[ing] ‘cross-racial understand¬ing,’ help[ing] to break down racial stereotypes, and ‘enabl[ing] students to better understand persons of different races.’” 24 Grutter, 539 U.S. at 330 (quoting Grutter v. Bollinger, 137 F. Supp. 2d 821, 850 (E.D. Mich. 2001)).
Justice Kennedy’s Grutter dissent decried the first two divergences from Bakke. He rejected the term “critical mass” as “a delusion used by the Law School to mask its attempt to make race an automatic factor in most instances and to achieve numerical goals indistinguishable from quotas.” 25 Id. at 389 (Kennedy, J., dissenting). He appealed to Justice Powell’s rule in Bakke when he argued that “[w]hether the objective of critical mass ‘is described as a quota or a goal, it is a line drawn on the basis of race and ethnic status,’ and so risks compromising individual assessment.” 26 Id. at 391 (quoting Bakke, 438 U.S. at 289). Nor was Justice Kennedy convinced of Grutter’s predicted end date for affirmative action, finding it “difficult to assess the Court’s pronouncement that race-conscious admissions programs will be unnecessary 25 years from now.” 27 Grutter, 539 U.S. at 394 (Kennedy, J., dissenting). He resisted Grutter’s twenty-five-year clock insofar as “the basic protection put in place by Justice Powell will be suspended for a full quarter of a century.” 28 Id.
Despite Justice Kennedy’s impassioned dissent in Grutter, his opinion in Fisher invokes the majority opinion from Grutter roughly a dozen times. 29 Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198, 2205-15 (2016). How, then, did Justice Kennedy come to terms with Grutter? The answer lies with Bakke.
II. Fisher to Bakke
On first impression, Bakke’s significance for Fisher is far from obvious. After all, Justice Kennedy’s opinion in Fisher does not directly cite Bakke even once. 30 Id. But first impressions can be deceiving. Although Fisher omits direct references to Bakke, it casts affirmative action programs and precedent in ways that maintain fidelity to Bakke. Fisher’s adherence to Bakke becomes clear once we attend to the ways Justice Kennedy describes UT Austin’s admissions program: he showcases elements of the program that follow Justice Powell’s rule in Bakke and downplays those aspects that he condemned in his Grutter dissent.
Justice Kennedy’s Grutter dissent described Justice Powell’s opinion in Bakke as “based on the principle that a university admissions program may take account of race as one, nonpredominant factor in a system designed to consider each applicant as an individual, provided the program can meet the test of strict scrutiny by the judiciary.” 31 Grutter, 539 U.S. at 387 (Kennedy, J., dissenting). For Justice Kennedy, the partial, indirect reliance on race in admissions was of the essence because it allowed each applicant to be considered as an individual. In describing the Fisher program, Kennedy emphasizes that “race is but a ‘factor of a factor of a factor’ in the holistic-review calculus,” 32 Fisher, 136 S. Ct. at 2207 (quoting Fisher v. Univ. of Tex. at Austin, 645 F. Supp. 2d 587, 608 (W.D. Tex. 2009)). and “race, in this indirect fashion, considered with all of the other factors . . . can make a difference to whether an application is accepted or rejected,” 33 Id. (emphasis added). thus allowing individualized consideration.
Even as Justice Kennedy admits that the admissions program considered in Fisher is “similar to the one in Grutter,” 34 Id. at 2206 (emphasis added). his opinion highlights those aspects of the Fisher program that distinguish it from the Grutter program. One difference Justice Kennedy accentuates is the procedures used in the final stages of the two review processes. In Grutter, he charged that “[t]he consultation of daily reports during the last stages in the admissions process suggests there was no further attempt at individual review save for race itself.” 35 Grutter, 539 U.S. at 392 (Kennedy, J., dissenting). By contrast, he stresses in Fisher that “[t]he admissions officers who make the final decision as to whether a particular applicant will be admitted make that decision without knowing the applicant’s race.” 36 Fisher, 136 S. Ct. at 2207.
At the same time, Justice Kennedy brushes aside the ways in which the programs in Fisher and Grutter are similar, most notably in their reliance on critical mass as the central measure for diversity. Justice Kennedy’s Fisher opinion does not ridicule the concept of critical mass as his Grutter dissent did. In fact, the term does not appear until the final section of the Fisher opinion—and then only to respond to Fisher’s critique of the concept. 37 Id. at 2210.
Fisher claimed that UT Austin had failed to define the level of minority enrollment that would constitute a critical mass. 38 Id. Instead of tackling the definition of critical mass head on, Justice Kennedy responds, “this Court’s cases have made clear . . . the compelling interest” justifying university affirmative action programs “is not an interest in enrolling a certain number of minority students.” 39 Id. “Rather, a university may institute a race-conscious admissions program as a means of obtaining ‘the educational benefits that flow from stu¬dent body diversity.’” 40 Id. (quoting Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411, 2419 (2013)) (citing Grutter, 539 U.S. at 328). Although Justice Kennedy cites the Court’s 2013 ruling in Fisher’s case and Grutter as authorities for these words, they actually distill Justice Powell’s rule in Bakke.
Fisher also claimed that the university had “already ‘achieved critical mass’ . . . using the Top Ten Percent Plan and race-neutral holistic review.” 41 Id. at 2211 (quoting Brief for Petitioner at 46, Fisher, 136 S. Ct. 2198 (No. 11-345), 2015 WL 5261568, at *46). Once again, Kennedy glosses over the question of what it means to “achieve critical mass,” 42 See id. at 2210-12 (eliding any clear test for determining “critical mass”). instead merely detailing that “the University conducted ‘months of study and deliberation, including retreats, interviews, [and] review of data,’ and concluded that ‘[t]he use of race-neutral policies and programs ha[d] not been successful in achieving’ sufficient racial diversity at the University.” 43 Id. at 2211 (alterations in original) (citations omitted).
A puzzling aspect of the Fisher opinion is that Justice Kennedy emphasizes minority enrollment statistics and “nuanced quantitative data” on classroom diversity to support the university’s assessment that race-neutral policies were not enough to achieve “sufficient racial diversity.” 44 Id. at 2211-12. In so doing, Justice Kennedy seems to allow that universities may employ race-conscious measures to enroll enough minority students to achieve the “educational benefits of diversity,” which sounds rather like employment of the “critical mass” standard that Justice Kennedy disclaimed in Grutter. 45 Grutter, 539 U.S. at 389 (Kennedy, J., dissenting). To understand this aspect of Justice Kennedy’s opinion, we must once again turn to Bakke. As I discuss elsewhere, 46 See Yuvraj Joshi, Measuring Diversity, 116 Colum. L. Rev. Online (forthcoming 2017) (analyzing uses of data and metrics in race-conscious admissions and why the concept of critical mass is controversial). Justice Powell’s endorsement of Harvard’s admissions plan as “[a]n illuminating example” of “[the] kind of program [that] treats each applicant as an individual in the admissions process” 47 Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 316-18 (1978). implied an acceptance of “some relationship between numbers and achieving the benefits to be derived from a diverse student body.” 48 Id. at 323 (emphasis added); see also Grutter, 539 U.S. at 336. By relying on minority enrollment numbers to demonstrate that race-neutral alternatives were insufficient, Justice Kennedy also seems to recognize a numerical component to the educational benefits of diversity, so long as that numerical component is implicit and imprecise and does not (as Justice Kennedy believes critical mass does) “attempt to make race an automatic factor in most instances and to achieve numerical goals indistinguishable from quotas.” 49 Grutter, 539 U.S. at 389 (Kennedy, J., dissenting).
Not all novel aspects of Grutter are absent from Fisher. While Justice O’Connor’s opinion in Grutter endorsed Justice Powell’s opinion in Bakke, it cast the benefits of a racially diverse student body in more explicitly racial terms, including “promot[ing] ‘cross-racial understand¬ing,’ help[ing] to break down racial stereotypes, and ‘enabl[ing] students to better understand persons of different races.’” 50 Id. at 330 (quoting Grutter v. Bollinger, 137 F. Supp. 2d 821, 850 (E.D. Mich. 2001). This aspect of Grutter reappears in Fisher: Justice Kennedy directly invokes Grutter’s more explicitly racial understanding of the educational benefits of diversity. 51 Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198, 2210 (2016) (quoting Grutter, 539 U.S. at 330). This language acquires additional importance in light of the racial justice protests unfolding across college campuses. Universities reluctant to institute measures to combat racial bias and insensitivity out of fear of opposition or litigation may take some comfort in the avowed importance of “cross-racial understanding” to realizing the educational benefits of diversity. 52 Grutter, 539 U.S. at 330.
Still, it is striking that most conspicuous in Fisher are those aspects of Grutter that maintain fidelity to Bakke and largely absent from Fisher are those aspects of Grutter that diverged from Bakke. The remainder of this Essay considers why this is and what it means for the future of affirmative action.
III. Justice Kennedy’s Concerns
Justice Kennedy appears motivated by two sets of concerns in trying to mold Fisher in the image of Bakke. One concern is with precedent. In his Grutter dissent, Justice Kennedy cautioned that “the numerical concept of critical mass has the real potential to compromise individual review,” which would be contrary to Bakke. 53 Id. at 389 (Kennedy, J., dissenting). Thus, Fisher allowed him to undo aspects of Grutter that he believes diverged from the “correct rule” of Bakke. 54 Id. at 387.
But Justice Kennedy’s preoccupation with precedent is rooted in his second concern: social cohesion. Reva Siegel has convincingly shown that in both preserving and limiting affirmative action measures, the Justices in the political middle of the Court (like Kennedy) have reasoned from an “antibalkanization” perspective that is “more concerned with social cohesion than with colorblindness.” 55 Reva B. Siegel, From Colorblindness to Antibalkanization: An Emerging Ground of Decision in Race Equality Cases, 120 Yale L.J. 1278, 1281 (2011); see also id. at 1302 (“Proponents of antibalkanization recognize that, to get beyond race, it may be necessary to take race into account; but, for them, taking race into account means crafting interventions that ameliorate racial wrongs without unduly aggravating racial resentments.”). For example, Justice Powell (another Justice in the center of the Court) was mindful of the threat racial preferences pose to social cohesion when he wrote in a footnote in Bakke: “All state-imposed classifications that rearrange burdens and benefits on the basis of race are likely to be viewed with deep resentment by the individuals burdened. The denial to innocent persons of equal rights and opportunities may outrage those so deprived and therefore may be perceived as invidious.” 56 Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 294 n.34 (1978). To mitigate the “deep resentment” likely to be felt by “innocent persons,” Justice Powell affirmed race-conscious diversity measures as a way to achieve racial inclusion by quieter and less determinate means than racial quotas.
Echoing Justice Powell’s Bakke opinion, Justice Kennedy’s Grutter dissent argued: “Preferment by race, when resorted to by the State, can be the most divisive of all policies, containing within it the potential to destroy confidence in the Constitution and in the idea of equality.” 57 Grutter, 539 U.S. at 388 (Kennedy, J., dissenting). Justice Kennedy worried that because admissions programs based on critical mass were (in his view) “tantamount to quotas,” they would “perpetuate the hostilities that proper consideration of race is designed to avoid,” and that “perpetuation, of course, would be the worst of all outcomes.” 58 Id. at 394.
Justice Kennedy’s Fisher opinion reveals that he remains concerned about social cohesion, but his understanding of social cohesion may be evolving. There is a subtle but important difference between references to “divisiveness” in his Grutter and Fisher opinions. In Grutter, Justice Kennedy seemed preoccupied with the threat to social cohesion posed by the use of racial classifications itself: “Preferment by race, when resorted to by the State, can be the most divisive of all policies.” 59 Id. at 388 (emphasis added). By contrast, Justice Kennedy in Fisher seems less concerned with whether racial classifications are used and more concerned with how they are used. As he writes: “Formalistic racial classifications may sometimes fail to capture diver¬sity in all of its dimensions and, when used in a divisive manner, could undermine the educational benefits the University values.” 60 Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198, 2210 (2016) (emphasis added). Fisher’s rendering of “divisiveness” suggests that it is the form, not merely the fact, of racial classification that poses a threat to social cohesion. 61 Constitutional scholars have debated whether the Equal Protection Clause is properly interpreted through a color-blind, anticlassification principle concerned with individual rights to equal treatment or a race-conscious, antisubordination principle concerned with group inequalities. An important strand of this literature considers how these two principles overlap and interact in shaping the form of equal protection law. See generally Jack M. Balkin & Reva B. Siegel, The American Civil Rights Tradition: Anticlassification or Antisubordination?, 58 U. Miami L. Rev. 9, 13 (2003) (“[A]ntisubordination values have played and continue to play a key role in shaping what the anticlassification principle means in practice.”); Reva B. Siegel, Equality Talk: Antisubordination and Anticlassification Values in Constitutional Struggles Over Brown, 117 Harv. L. Rev. 1470, 1477 (2004) (“[A]ntisubordination values live at the root of the anticlassification principle . . . .”).
The Court’s jurisprudence on affirmative action in higher education bears out this understanding. 62 But cf. Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 724 (2007) (noting that cases involving race-conscious measures in high schools are not governed by Grutter, since “[i]n upholding the admissions plan in Grutter . . . this Court relied upon considerations unique to institutions of higher education”). Bakke and Grutter modulated the form of constitutionally permissible affirmative action in institutions of higher education, at least partly in response to concerns of social cohesion. 63 See Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 294 n.34 (1978) (“All state-imposed classifications that rearrange burdens and benefits on the basis of race are likely to be viewed with deep resentment by the individuals burdened.”); Grutter, 539 U.S. at 341 (“As Justice Powell recognized in Bakke, so long as a race-conscious admissions program uses race as a ‘plus’ factor in the context of individualized consideration, a rejected applicant ‘will not have been foreclosed from all consideration for that seat simply because he was not the right color or had the wrong surname.’” (quoting Bakke, 438 U.S. at 318)); see also Robert C. Post, Foreword: Fashioning the Legal Constitution: Culture, Courts, and Law, 117 Harv. L. Rev. 4, 74-75 (2003) (observing how “the Court in Grutter and Gratz constructs doctrine that in effect demands obscurity” out of concern for “the likelihood of racial balkanization”); Neil S. Siegel, Race-Conscious Student Assignment Plans: Balkanization, Integration, and Individualized Consideration, 56 Duke L.J. 781, 781-82 (2006) (analyzing how the requirement of individualized consideration responds to concerns about balkanization). But whereas Justice Powell appeared to center his concern for social cohesion on resentment among whites, Justice Kennedy’s opinion in Fisher is not limited in this way. Fisher was decided during a year marked by intense racial conflict across America’s universities, with minority students recounting experiences of racism and isolation and calling for race-conscious responses to these problems. 64 See, e.g., Anemona Hartocollis & Jess Bidgood, Racial Discrimination Protests Ignite at Colleges Across the U.S., N.Y. Times (Nov. 11, 2015), http://nyti.ms/1kNYq2z; Katherine Long, What It’s Like to Be Black on Campus: Isolated, Exhausted, Calling for Change, Seattle Times (Apr. 9, 2016, 8:00 AM), http://www.seattletimes.com/seattle-news/education/what-its-like-to-be-black-on-campus-isolating-exhausting-calling-for-change. Justice Kennedy is likely attuned to and routinely challenged to reckon with this reality. 65 See, e.g., Schuette v. Coal. To Defend Affirmative Action, 134 S. Ct. 1623, 1676 (2014) (Sotomayor, J., dissenting) (explaining why “race matters” in American social life). As he wrote in a 2007 opinion: “The enduring hope is that race should not matter; the reality is that too often it does.” 66 Parents Involved, 551 U.S. at 787 (Kennedy, J., concurring in part and concurring in the judgment). Fisher admits of the understanding that in an American society where race matters, a rigid adherence to colorblindness may itself pose a threat to social cohesion.
IV. After Fisher
To understand constitutionally permissible affirmative action after Fisher, we must go back to Bakke. Universities interested in enrolling a diverse student body while operating within constitutional constraints would be wise to read Justice Powell’s opinion.
Grutter’s most significant divergence from Bakke—the concept of critical mass—remains controversial. When the court last considered Fisher’s case in 2013, Justice Scalia accentuated the amorphousness of the term when he quipped: “We should probably stop calling it critical mass then, because mass, you know, assumes numbers, either in size or a certain weight. . . . Call it a cloud or something like that.” 67 Transcript of Oral Argument at 71-72, Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (No. 11-345), https://www.supremecourt.gov/oral_arguments/argument_transcripts/11-345.pdf. Justice Alito’s Fisher dissent is just as unrelenting: “[W]ithout knowing in reasonably specific terms what critical mass is or how it can be measured, a review¬ing court cannot conduct the requisite ‘careful judicial inquiry’ into whether the use of race was ‘necessary.’” 68 Fisher, 136 S. Ct. at 2222 (Alito, J., dissenting) (quoting Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411, 2420 (2013)). While Justice Kennedy in Fisher does not repudiate the concept of critical mass as he did in Grutter, he does not endorse it either. Given the intense scrutiny of critical mass as the measure of diversity, universities reassessing their affirmative action programs post-Fisher should consider adhering to Bakke’s formulation of the diversity interest as expanded in Grutter and endorsed in Fisher.
Finally, universities would do well to recognize that the Court’s jurisprudence on affirmative action in higher education is concerned with social cohesion, but concerns of social cohesion run both ways. While Justice Powell in Bakke expressed concern for the “deep resentment” likely to be felt by “innocent persons” who do not gain admission, 69 Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 294 n.34 (2003). Justice Kennedy in Fisher does not limit his concern for social cohesion in this way. Fisher admits of the understanding that the concerns of minority communities also matter in healing social divisions and realizing the educational benefits of diversity.
* Yale Law School Peter and Patricia Gruber Fellow in Global Justice. This Essay benefited from conversations with Reva Siegel, Paul Kahn and Anouchka Appanah and from the many thoughtful comments of Jordan Laris Cohen and the editors of the Stanford Law Review. Views expressed in this Essay are mine and not those of any organization with which I am affiliated.