Introduction

This Article builds on an intervention Luke Harris and Uma Narayan made more than two decades ago in the Harvard BlackLetter Law Journal repudiating the conceptualization of affirmative action as a racial preference.^{^¹} The central claim we advance is that affirmative action levels the playing field for all African Americans students, not just those who are class-disadvantaged. Developing this argument is crucial against the backdrop of the argument that affirmative action is both over- and underinclusive.

The underinclusive argument posits that affirmative action does little to remedy the extent to which, across the country, African American students are forced to attend failing primary and secondary schools whose pipelines lead to prisons, not universities. The claim is that most students who attend these schools are simply too disadvantaged to benefit from policies that presume college eligibility at a minimum.^{^²} At best, extending affirmative action to disadvantaged black students would create a “mismatch” problem;^{^³} the policy would put black students in educational contexts that are above their intellectual achievement grade. This mismatch would then cause black students to struggle academically to fit in and succeed.

Importantly, proponents of mismatch do not limit their application of the theory to economically disadvantaged blacks. They apply arguments about mismatch to middle-class African Americans as well. Indeed, in the Fisher v. University of Texas at Austin (Fisher II)⁴ litigation, a number of the amicus briefs specifically draw on the mismatch theory to challenge the constitutionality of the University of Texas’s admissions policy.^{^⁵} Their target in doing so is not blacks who are class-disadvantaged but those who presumptively are not.

Whites, on the other hand, largely escape the mismatch critique. People who argue that affirmative action should focus more on working-class whites, not middle-class blacks, rarely invoke the possibility of mismatch as a concern. The assumption seems to be that, unlike African American beneficiaries of affirmative action, white working-class beneficiaries will not be in over their head.

Cheryl Harris has suggested that the reason arguments about mismatch are almost always rehearsed with reference to African Americans is because the mismatch thesis aligns with preexisting notions of black intellectual deficit.^{^⁶} Put another way, the theory of mismatch is another way of writing intellectual deficiency and inability into race—and more specifically, blackness.^{^⁷} Black intellectual inferiority has long been an important part of the social transcript of American life. Indeed, perhaps the only thing easier in the United States, racially speaking, than questioning black intellectual ability is associating African Americans with crime.^{^⁸} That the mismatch theory at least implicitly relies on longstanding “reasonable doubt” about black intellectual competence and capacity makes it all the more important that scholars and policymakers carefully examine the empirical basis for the theory.^{^⁹}

But that is not our project in this Article. We invoke the theory of mismatch here for a more limited purpose: to reveal how it facilitates the underinclusive argument against affirmative action. This brings us to the overinclusive claim.

The overinclusive critique of affirmative action posits that affirmative action benefits African Americans who are not disadvantaged. Extending this benefit to privileged blacks is unfair, the reasoning goes, particularly because this benefit comes at a cost to poor whites.^{^¹⁰} Why should admissions policies systematically prefer privileged students (read: black middle-class applicants) over disadvantaged students (read: poor whites)?¹¹ That is the question the overinclusive argument against affirmative action encourages us to ask—and the answer that question invites is clear: Admissions policies should not prefer privileged black students over disadvantaged white ones.

Like many of the arguments against affirmative action, the over- and underinclusive claims against the policy are not new. Writing in 1994, Harris and Narayan observed that:

This juxtaposition of the middle-class Black against his poor Black counterpart often has the purpose of setting up an insoluble dilemma between whose horns any possible justification for affirmative action seems to disappear. The middle-class Black does not need or ‘deserve’ any help countering the effects of racism; therefore, affirmative action is not warranted with respect to him or her. By contrast, the poor Black perhaps deserves some sort of help, but is situated so as not to be in a position to benefit from affirmative action policies; thus they are of no practical import to him or her.^{^¹²}

The “insoluble dilemma” to which Harris and Narayan refer is another way of saying that opponents of Affirmative action describe African Americans as either too advantaged to deserve affirmative action or too disadvantaged to benefit from the policy. Privileged or mismatched,¹³ the lose-lose position African Americans occupy in anti-affirmative action discourse places them beyond the remedial reach of the policy.

The remainder of this Article focuses on the privileged side of the “insoluble dilemma.” We do so because, since the publication of Harris and Narayan’s paper more than two decades ago, we have yet to read a full defense of affirmative action that expressly focuses on middle-class black applicants.^{^¹⁴} By and large, proponents of affirmative action treat black students whose experiences do not comfortably fit the K-12 educational disadvantage narrative as the unintended but unavoidably necessary beneficiaries of the policy.

One of the most striking manifestations of this necessary evil defense of affirmative action is the notion that supporting affirmative action for middle-class African Americans is like supporting partial-birth abortion; ordinarily, one might not want to support partial-birth abortion, but one does so nevertheless to ensure women’s reproductive rights in general. Similarly, so the argument goes, one ordinarily might not want to support affirmative action, but one does so to ensure racial diversity and to prevent the resegregation of American colleges and universities.^{^¹⁵}

The failure of proponents of affirmative action to robustly defend the policy for middle-class African Americans strengthens the perception of affirmative action as a racial preference. Put another way, the perception of affirmative as a racial preference has particular traction when its beneficiaries are black but not class-disadvantaged. Think about the matter this way: If people believe that colleges and universities employ affirmative action to admit African Americans who are not economically disadvantaged, the conclusion that affirmative action is a racial preference is easy to reach—black students who are not disadvantaged are getting preferential treatment because of their race. Liberals defend this preference because it advances diversity and contributes to the “robust exchange of ideas.”^{^¹⁶} Conservatives reject it because it violates their understanding of colorblindness and effectuates what they call “reverse discrimination” against whites.^{^¹⁷}

This Article reframes the debate. It does so by explaining why affirmative action for middle-class blacks is neither a racial preference for African Americans nor reverse discrimination against whites. We locate our argument in the context of admissions. Specifically, we identify a number of disadvantages black students—across class—likely experience prior to and in the context of applying to colleges and universities. We argue that these disadvantages can diminish the competitiveness of a black student’s admission file and that affirmative action helps to counteract this negative effect. In advancing this argument, we hope to make clear that racial inequality is not exhausted by class inequality, that affirmative action is not a racial preference but a mechanism to level the admissions playing field, and that the inclusion of middle-class African Americans in affirmative action programs is not an effort to displace working-class or poor blacks but a way of achieving an important and insufficiently acknowledged diversity benefit: namely, intraracial diversity, or diversity among and between black students, including along the class spectrum.^{^¹⁸}

The remainder of the essay is organized as follows. Part I offers a theoretical argument that explains why it is a mistake to frame affirmative action as a racial preference. We identify a number of obstacles African American students across class likely encounter—up to and including the moment of admission—that potentially negatively impact their formal academic performance and the overall competitiveness of their admissions files. These obstacles create what we call an “admissions imbalance” that affirmative action helps to offset. The failure to correct for this imbalance simultaneously hurts black students and benefits students whose educational trajectories do not include the racial obstacles we will describe. To put this point another way, what black students experience as racial disadvantages, white students experience as a “thumb on the scale,”¹⁹ whereby the absence of racial obstacles puts white students at an advantage in the college application process. In this respect, institutions that do not acknowledge and ameliorate the obstacles we outline in Part I end up producing the very thing many attribute to affirmative action: racial preference.

Drawing principally from research in social psychology, Part II employs empirical evidence to support the theoretical claims Part I articulates. Here we demonstrate that several of the disadvantages black students encounter potentially affect traditional academic performance indicators (such as standardized test scores and grade point averages (GPAs)) and other critical dimensions of an admission file.

Against the backdrop of Parts I and II, one might reasonably ask why academics, policymakers, lawyers, and judges continue to frame affirmative action as a racial preference. That is the question we take up in the conclusion. We suggest that part of the reason people have difficulty seeing affirmative action as a mechanism that levels the playing field for African Americans across class is because the debate over affirmative action overstates or flattens the middle-class status of African Americans. Indeed, in at least some of the anti-affirmative action discourses, opponents of the policy treat middle-class African American youth as being, effectively, as privileged as the children of President Obama.

I. Blacks Who Are Not Class-Disadvantaged Are Truly Disadvantaged: The Theoretical Argument

Although racial diversity, not racial disadvantage, counts as a compelling state interest for affirmative action,^{^²⁰} public support for the policy turns, at least in part, on whether people perceive its beneficiaries to be disadvantaged.²¹ In advancing this claim, we do not mean to suggest that diversity and disadvantage are oppositional concepts or that they are necessarily in tension with each other. Our point is rather that, as a formal doctrinal matter, equal protection doctrine requires that colleges and universities ground their defense of affirmative action on a diversity, not a disadvantage, rationale.^{^²²}

Yet it is clear that for at least some Justices on the U.S. Supreme Court, concerns about disadvantage shape how they think about the constitutionality of affirmative action. One of the clearest examples of this is manifested in the Fisher v. University of Texas at Austin (Fisher I) litigation,^{^²³} in which Abigail Fisher sued the University of Texas, alleging that the school’s affirmative action plan violated her rights to equal protection under the U.S. Constitution. During oral argument, Justice Alito pressed Gregory G. Garre, the attorney representing the university, on what Justice Alito perceived to be a fundamental problem with how the school was administering its affirmative action plan:

Mr. Garre: And I don’t think it’s been seriously disputed in this—this case to this point that, although the percentage plan certainly helps with minority admissions, by and large, the—the minorities who are admitted tend to come from segregated, racially-identifiable schools.

Justice Alito: Well, I thought the whole purpose of affirmative action was to help students who come from underprivileged backgrounds, but you make a very different argument that I don’t think I’ve ever seen before.

The top 10 percent plan admits lots of African Americans—lots of Hispanics and a fair number of African Americans. But you say, well, it’s—it’s faulty because it doesn’t admit enough African Americans and Hispanics who come from privileged backgrounds. And you specifically have the example of the child of successful professionals in Dallas.

Now, that’s your—that’s your argument. If you . . . have an applicant whose parents are—let’s say they’re—one of them is a partner in a law firm in Texas, another one is a . . . corporate lawyer. They have income that puts them in the top 1 percent of earners in the country, and they have—parents both have graduate degrees.

They deserve a leg-up against, let’s say, an Asian or a white applicant whose parents are absolutely average in terms of education and income?

Mr. Garre: No, Your Honor . . . .

Justice Alito: Well, how can the answer to that question be no, because being an African American or being a Hispanic is a plus factor.^{^²⁴}

While there is much that one might say about the foregoing exchange, two points in particular deserve engagement. First, Justice Alito is clearly invoking a version of the black son of a lawyer/white son of a coal miner trope. More specifically, he is contesting the legitimacy of an admissions regime under which “African Americans and Hispanics who come from privileged backgrounds” get a “leg-up against, let’s say, an Asian or a white applicant whose parents are absolutely average in terms of education and income.”

Second, Justice Alito makes clear that, from his perspective, “the whole purpose of affirmative action was to help students who come from underprivileged backgrounds.” One might conclude from this statement, particularly because a Supreme Court Justice articulates it, that a university can defend its affirmative action policy on the ground that it benefits students from “underprivileged backgrounds.” One would be wrong to make this conclusion. As we have already stated and want to repeat here, diversity, not underprivileged backgrounds, serves as compelling justification for affirmative action. That is to say, under current law, diversity “is the whole purpose of affirmative action.”^{^²⁵}

If equal protection law is clear regarding the constitutional basis on which universities can defend their affirmative action policy, why would Justice Alito get the remedial justification for the policy so wrong? The answer, we think, is that like many other people, Justice Alito believes that affirmative action should correct for or counteract the disadvantages people encounter in life, notwithstanding that diversity, and not disadvantage, is the constitutionally legitimate justification for affirmative action.

From Justice Alito’s perspective, the black son of a lawyer is not, to borrow from William Julius Wilson, “truly disadvantaged.”^{^²⁶} The black son of a lawyer son should not, therefore, be a beneficiary of affirmative action.

Justice Alito’s sense that blacks who are not class-disadvantaged are not disadvantaged at all likely shapes his perception of affirmative action as a “leg-up.” Indeed, further along in the oral argument, Justice Alito expressly connects his concerns about blacks who are not class-disadvantaged with his conception of affirmative action as a racial preference. He does so in the following question he directed at the Solicitor General, Donald Verrilli: “Does the United States agree with Mr. Garre that African American and Hispanic applicants from privileged backgrounds deserve a preference?”^{^²⁷} The very framing of Justice Alito’s question reveals not only his insinuation that middle-class blacks are not disadvantaged, but also his conceptualization of affirmative action as a racial preference. The short of it is that Justice Alito’s engagement with both Verilli and Garre reflects the view that because the University of Texas’s admission policy confers a racial preference to students who are not disadvantaged (middle-class blacks), the policy, at the very least, is constitutionally suspect.

In this Part we demonstrate why Justice Alito’s framing of affirmative action is flawed. We do so by interrogating the dominant metaphor people across the ideological spectrum have employed to describe affirmative action—that it is “a thumb on the scale.” We present six schematics to show how the thumb on the scale characterization of affirmative action simultaneously obscures black disadvantages and facilitates the affirmative action-as-preference frame.

We begin with the first schematic, Figure 1:

Figure 1

Figure 1 presents the default admissions regime in the form of a scale. A black person sits on one side of the scale and a white person sits on the other. A critical feature of this scale is that it is perfectly leveled. The black body weighs no more than the white body. The black applicant and the white applicant are in equipoise.

Moreover, the black person and the white person look exactly the same. The only difference between the two is that one has a black face and the other’s face is white. Race, under this view, marks neither disadvantage nor privilege. It is nothing more than skin color. This decidedly thin conception of race invites us to conclude that, in the context of admissions, race does not—and should not—matter.^{^²⁸}

In sum, Figure 1’s perfectly balanced scales and similarly situated representation of the black and white applicants communicate the idea that the admissions process is colorblind and is one in which applicants are treated perfectly equally. Under the admission system that Figure 1 depicts, neither the white applicant nor the black applicant is favored. Nor is either one disadvantaged. The admissions process Figure 1 depicts is balanced, racially neutral, and fair.

Now consider Figure 2:

Figure 2

Figure 2 depicts a perception of how affirmative action changes the picture of neutrality that Figure 1 presents. Appearing in the form of a thumb, affirmative action creates an imbalance. The weight of the thumb tips the scales in favor of the African American applicant.

Figure 2 illustrates two additional asymmetries that help to shore up the perception of affirmative action as a preference. First, the black applicant is doing nothing to achieve the competitive advantage he enjoys. It is not, in other words, an advantage he earned. He is in the favorable admissions position Figure 2 presents (the scales tilt in his favor) because of the preference affirmative action accords to him simply based on his race. The white applicant, meanwhile, struggles mightily to pull the scales down to its original—and presumptively racially neutral—position. But no matter how hard he works, the scales remain tipped in favor of the black applicant. The weight of the so-called “black bonus” of affirmative action is too heavy for the white applicant to counteract.^{^²⁹}

Second, the black applicant is not the “underprivileged” black person who, for Justice Alito, is the appropriate beneficiary of affirmative action. Professionally dressed, the black applicant stands in for those privileged and undeserving black beneficiaries of affirmative action whom opponents of affirmative action regularly conjure up. The white applicant, on the other hand, is more casually dressed, suggesting a working-class or poor-white identity. With respect to economic resources, this applicant is, at best, the white person who, in Justice Alito’s account, affirmative action disadvantages—a white applicant who comes from a family that is “absolutely average in terms of education and income”—and at worst, economically far below that average. In short, under Figure 2, the admissions system prefers the black applicant (who is presumptively privileged and exerting no effort) over the white applicant (who is presumptively disadvantaged and hard-working). Here, the admissions process is unbalanced, racially biased, and unfair. It represents an instance of “reverse discrimination.”^{^³⁰}

It is Figure 2 that opponents of affirmative action have in mind when they ask: Why should an admissions policy extend a preference to the black child of a lawyer but not the white child of coal miner?^{^³¹} When asked, this question is not necessarily about facilitating the upward mobility of the coal miner’s child.^{^³²} It is often about criticizing affirmative action as a policy that focuses on race, not disadvantage.^{^³³}

Importantly, both liberals and conservatives acquiesce in the image of affirmative action Figure 2 depicts. Both conservatives and liberals regularly refer to affirmative action as a thumb on the scale and both conceptualize the policy as a preference.^{^³⁴} As noted earlier, the basic difference between conservative and liberal positions on affirmative action is that whereas liberals believe that the costs of affirmative action are outweighed by the benefits (including diversity), conservatives perceive the costs of the policy (including “reverse discrimination”) to be too high.

Consider now Figure 3:

Figure 3

Like Figure 2, Figure 3 depicts an admissions regime that is not level. Here, too, the scales are uneven. But this time, they tip in favor of the white applicant. Without further explanation, and against the well-established backdrop of (1) stereotypes about the poor work ethic and intellectual deficit of African Americans,^{^³⁵} (2) discourses about the achievement gap between black and white students,^{^³⁶} (3) the assumption that whites do not benefit from affirmative action,^{^³⁷} and (4) the unlikeliness that any institution would intentionally discriminate in favor of white applicants or against black applicants,^{^³⁸} the conclusion one might draw from Figure 3 is twofold. First, the white applicant deserves to be in the competitive position he enjoys; and second, the black applicant deserves to be at a disadvantage. Figure 3, in other words, represents what some might call a “racially natural disequilibrium,” one that derives from the “natural” fact that the white applicant is smarter and/or has worked harder than the black applicant. Under this view, merit explains why the black applicant is disadvantaged and the white applicant is advantaged. No remedial intervention is therefore necessary. What is needed is what Dinesh D’Souza would call a “cultural reconstruction.”³⁹ Black parents, community leaders, and political figures should change the cultural habits of black teenagers,^{^⁴⁰} encourage them to study hard and stay in school, and persuade them to not associate the pursuit of academics with “acting white.”^{^⁴¹} Hard work, greater effort, and a stronger commitment to education on the part of black students would tip the scales in their favor, make them more competitive, and remove the asymmetry Figure 3 illustrates. The solution for the imbalance Figure 3 presents, in sum, is for black students to fix themselves and become, like Asian Americans, a “model minority.”^{^⁴²}

Figure 4 contextualizes the imbalance Figure 3 presents. It puts into focus two ways in which race tilts the scales in favor of whites, both of which are invisible in Figure 3:

Figure 4

Under Figure 4 (and unlike Figure 3), the asymmetry in the picture is not a function of merit.⁴³ The white applicant is advantaged not because he is smarter or works harder than the black applicant. The white applicant benefits from two thumbs on the scale: (1) the thumb of black racial disadvantage (the disadvantages that black—but not white—applicants experience), and (2) the thumb of white racial advantage (the advantages that white—but not black—applicants experience). We discuss them in turn.

The black disadvantage thumb is shorthand for a number of structural racial disadvantages black applicants bring to or experience in the context of the admissions process. We discuss these disadvantages more fully in Part II and provide a summary articulation of some of them below:

Deficient Standardized Tests (the extent to which standardized tests have a disparate impact on African Americans and do not accurately assess their abilities);
Explicit racial biases (consciously held negative views about a group);
Implicit racial biases (unconsciously held negative views about a group);
Stereotype threat (the perception that one’s performance in a particular domain will confirm negative stereotypes about one’s group);
Racial isolation (the sense of alienation and marginalization that derives from the underrepresentation of one’s group within a particular institutional setting);
Negative institutional cultures (environments whose norms and practices render some groups insiders and others outsiders).

Because the foregoing factors put blacks as a group, relative to whites as a group, at a competitive disadvantage, we might think of these factors, cumulatively, as constituting a thumb on the scale for whites.

Given our earlier observations about race and class, we should be clear to note that the racial disadvantages we set forth above are not class-dependent. Conspicuously absent from our list is K–12 segregation and inequality.^{^⁴⁴} We excluded that factor to avoid getting bogged down in a debate about whether K–12 inequality is a class-based problem or a racial one (we think it is both). Instead, we focus on a set of factors whose impact unequivocally transcends class and refer to them, cumulatively, as the black racial disadvantage thumb.

Think now about the white racial advantage thumb. Two advantages, together, might be thought of as a thumb on the scale. First, white applicants benefit from what we call the “intergenerational value of whiteness.” Every white person benefits from the intergenerational value of whiteness. Our argument here is not principally about the fact that some whites have been able to transfer resources and wealth (including cultural capital) across generations in ways that African Americans generally have not. If our claim about white advantage centered on transfer of resources and wealth, many whites—including the son of a coal miner previously discussed—would be excluded from the benefits we attribute to whiteness. Our argument about white advantage is grounded by the observation that to be white in the United States (irrespective of one’s genealogical relationship to slavery or Jim Crow) is necessarily to inherit the historical badge of honor, privilege, respectability and positive social meanings associated with whiteness and white people.^{^⁴⁵} While class mediates the degree to which whites benefit from this advantage,^{^⁴⁶} the phenomenon transcends class.

The second white advantage Figure 4 means to capture is in-group favoritism. Racial inequality is not just a function of out-group derogation and exclusion, but also preferential treatment toward the in-group. Although positive in-group bias and negative out-group bias are often discussed as two sides of the same coin, research shows that in-group favoritism and out-group prejudice are in fact distinct.^{^⁴⁷} Though each operates to maintain systems of inequality, discrimination can be motivated by in-group favoritism alone, without any negative intent or hostility toward the out-group, and vice versa.^{^⁴⁸} In-group favoritism is ubiquitous, but has the greatest effect when in-group members are in positions of power and can thus preferentially bestow a range of benefits, access to resources, and opportunities to other in-group members. Because of the political and economic power whites have (and historically have had) as a group, whites are more likely to benefit from in-group favoritism than any other racial group, including African Americans. Researchers have proposed that in-group favoritism plays a key role in the differential advantage of whites in a large swath of life, from general helping behavior,^{^⁴⁹} to tipping,^{^⁵⁰} to evaluation of employment, and housing applications and interviews.^{^⁵¹} The cumulative advantages that can accrue from being the repeat beneficiary of in-group preferences are likely significant.

To sum up, under Figure 4, the default admissions regime is uneven. The scales are unbalanced. Two thumbs on the scale tip the process in favor of whites to the disadvantage of African Americans.

Figure 5 reintroduces the thumb of affirmative action into the analysis:

Figure 5

In this image, affirmative action is a part of the picture but it does not create the imbalance. Indeed, even with the thumb of affirmative action, the scales are still uneven. They continue to lean in a direction that disadvantages African Americans. The combined weight of the black racial disadvantage thumb and the white racial advantage thumb exceeds that of the affirmative action thumb. Under Figure 5, while affirmative action is ameliorative, the policy does not completely counteract the combined effects of white advantage and black disadvantage.

At this point in the analysis, an important caveat is in order: Notwithstanding the very specific narrative Figure 5 invites, we should be clear to state that we do not present Figure 5 as a strong empirical claim. It is hard to know, for example, whether, with the thumb of affirmative action back on the scales, the scales remain tilted in favor of whites. In some instances, the effect of affirmative action might be to level out the scales; in other instances, the policy might have an overcorrection effect, tilting the scales in favor of African Americans. To put the point as Harris and Narayan might, we are not arguing that “affirmative action policies are, or can be, magical formulas that help us determine with perfect precision in every case the exact weights that must be accorded a person’s class background, gender, and minority status, so as to afford him or her perfect equality of opportunity.”⁵² Because context will surely matter, the specific picture Figure 5 paints is best viewed as a soft default. The more important takeaway from Figure 5 is that affirmative action does not disrupt an otherwise racially neutral baseline at which applicants are similarly situated. The policy attempts to correct for the thumbs of black racial disadvantage and white racial advantage, both of which tilt the admissions scales in favor of white applicants.

This brings us back to Figure 2, the representation of affirmative action as a racial preference. The image, again, is this:

Figure 2

Part of what this image communicates is that, but for affirmative action, the white applicant would be in a more competitive position than he currently occupies. A stronger version of this argument frames affirmative action as an admissions barrier for whites. More precisely, the claim is that affirmative action causes admissions officers to deny admissions to white students those officers would otherwise admit.

That argument is more spin than empiricism. Indeed, as Kimberly West-Faulcon has argued, drawing on the work of Goodwin Liu, there is “causation fallacy” in the argument that asserts that but for affirmative action the University of Texas would have admitted Abigail Fisher. As West-Faulcon explains, there are simply too few black students in the admissions pool of elite colleges and universities for affirmative action to have the causation effect Abigail Fisher and others attribute to it.^{^⁵³}

In some ways, even the figures we employ to challenge the racial preference framing of affirmative action invite the “causation fallacy” argument in that they paint a picture of affirmative action in which—throughout the admissions process—a white student finds himself in contestation with a black student over a particular admissions spot. We know of no school that would describe its admissions process in that way. Most, if not all, schools would say that they employ some version of “holistic review” under which admissions decisions resemble something like a totality of the circumstances analysis. Of course, standardized test scores and GPA typically weigh more heavily than other variables. The point is that while most elite colleges and universities employ affirmative action in the context of admissions, they do not therefore necessarily have affirmative action slots per se.

Second, even if a school aspired to conduct its admissions process so that each spot was a tournament between a white student and a black student, it simply could not do so. Given, as we have said, the relatively small number of African Americans in the applicant pool of elite colleges and universities, and the fact that not all black applicants are beneficiaries of affirmative action, it is hard to see how there could ever be an admissions system in which every white student finds himself in competition with a black beneficiary of affirmative action for a particular admissions spot.

Significantly, a relatively recent study on race and admissions describes the effect of affirmative action on white applicants as negligible.⁵⁴ Focusing specifically on Harvard University, the study concludes that removing all African Americans and Latinos from the admissions process at Harvard would increase the likelihood of white applicants being admitted by only 1 percent.⁵⁵ Assuming, arguendo, that admissions regimes include tournaments between black and white applicants and that affirmative action is a racial preference—this study indicates that the worst-case scenario for white applicants looks something like Figure 6:

Figure 6

In only one of the hundred admissions scales—the first one in the upper-left corner—is a white student in a tournament with a black applicant who is benefitting from the affirmative action thumb on the scale. The remaining ninety-nine scales do not involve blacks, Latinos, or affirmative action.^{^⁵⁶}

To repeat: We do not think that admissions systems operate as white vs. black tournaments; we do not believe that affirmative action is a racial preference; and it is not our view that, if one thinks of admissions processes as scales, the scales are otherwise balanced but for affirmative action. We employ Figure 6 not to acquiesce in any of the foregoing ideas but to visually contextualize the strongest case against affirmative action. In context, that case, as Figure 6 suggests, is decidedly weak.

* * *

Central to our thesis thus far is the claim that affirmative action is not a racial preference because, among other things, it corrects for a number of disadvantages African Americans experience up to and during the admissions process. The question now is whether we can support that thesis empirically. Is there evidence, in other words, to support our contention that African Americans across class are vulnerable to the racial disadvantages this Part discussed—and evidence linking those disadvantages to critical aspects of an admissions file? The short answer to both questions is “yes.” Part II elaborates on these questions—and their answers.

II. Quantifying Black Racial Disadvantage: The Empirical Evidence

This Part provides empirical evidence to support our claim that black students across class, and not just class-disadvantaged black students, experience multiple disadvantages that likely affect their academic performance and the overall competitiveness of their admissions files. As a starting point, we ask you to consider the following scenario.

Imagine that the son of a black lawyer starts his first day at a predominantly white high school. What concerns and anxieties does his parent have as she waves goodbye to her teenager? Some concerns—whether her child will eat lunch at the appropriate time (if at all) or will spend too much time on social media—are universal parenthood worries. But beyond these concerns, the black parent will have a number of very specific worries. The black parent will wonder whether her son made it safely to school without being mistakenly profiled as a criminal suspect—and whether at school so-called Resource Officers (local police officers assigned to work at the school) will intentionally profile him as a criminal suspect. She will wonder whether the teachers understand that educators hold lower expectations (both consciously and unconsciously) of black students. She will ask herself whether this will be another year in which teachers patronize her child, offer less constructive feedback on his work, and provide few, if any, mentorship opportunities. The black parent will also hope that her advice to her son “not to argue with the teachers and do exactly as they tell you” will prevent teachers from perceiving him as “a boy with an attitude.”

The black parent will wonder whether this year her son will be cajoled—again—into giving the annual Black History Month speech, pressured into joining the basketball team (though he would rather play tennis), and tokenized and marginalized as the holder of the most insignificant position in student government. The black parent will also think about the racial demographics of the school in a broader institutional sense. Will her son, finally, have a black teacher? A teacher of color? Will there be a few more black students at the school this year? (She knows there will not be many more than the last time she researched the matter.) Will some of the black students at the school be in her son’s class so that he feels some measure of racial comfort—or will her son have to supply racial comfort to his classmates to make them feel less anxious and nervous around him?^{^⁵⁷}

Will her son be tracked into the less demanding classes? Will he be counseled out of thinking about competitive colleges and universities as a future possibility? Will any of her son’s teachers evidence a concern about and be attentive to his overall racial well-being?

The black parent’s concerns put into sharp relief some of the racial obstacles her son will face throughout his high school career, notwithstanding that he is middle-class. This Part explains how these obstacles can have a negative effect on the admissions file of this student five years downstream, shaping not only “hard” evaluative measures, such as standardized test scores and grades, but also “soft” evaluative measures, such as leadership experience, awards, extracurricular opportunities, internships, and letters of recommendations. We focus our attention on four specific racial disadvantages—stereotype threat, implicit biases, explicit biases, and negative institutional culture—and show that they are independent of economic class status.

A. Stereotype Threat

Stereotype threat is one of the most studied, and most prevalent, forms of psychological strain faced by black students.^{^⁵⁸} This Subpart explains how stereotypes can negatively affect a black student’s performance on standardized tests, classroom engagement, and whether and to what extent the student seeks help outside of the classroom or visits a teacher during office hours.⁵⁹ Each of the foregoing effects potentially diminishes the competitiveness of a student’s admissions file.

Roughly, stereotype threat refers to a scenario in which (a) because of one’s membership in a particular social group, negative stereotypes exist about one’s ability to perform in some specific domain, (b) one is consciously or unconsciously concerned about confirming those stereotypes, and (c) that concern undermines the quality of one’s performance. For example, pervasive stereotypes that blacks are less intelligent or less capable⁶⁰ may cause black students to fear that their performance in school will confirm, to themselves or to others,⁶¹ that these negative stereotypes are true. These worries raise the stakes of school performance, adding an additional layer of pressure to achieve that increases stress,⁶² undermines learning and engagement,⁶³ taxes cognitive resources,⁶⁴ and impairs academic performance.⁶⁵

The effect of stereotype threat on black students has been observed in hundreds of psychological studies employing rigorous experimental methods, both in controlled laboratories and in the field.⁶⁶ For example, an early laboratory experiment conducted by Claude Steele and Joshua Aronson randomly assigned black and white participants to complete verbal questions from the graduate record examinations (GRE) under one of two conditions: threat or no threat.⁶⁷ In the threat condition, participants were told that the questions were a test that would measure their intellectual ability, a statement that for black students activates known racial stereotypes about intellectual inferiority and triggers the concern that they may confirm this negative stereotype by performing poorly. In the no threat condition, participants were told that the questions were simply a laboratory problem-solving task that would not evaluate their ability. The authors found that black students performed significantly worse than white students in the threat condition, but equivalent to white students in the lower-stakes no threat condition. In other words, when racial stereotypes of intellectual inferiority were made salient—simply by stating that the test was diagnostic of intellectual ability—black students faced the psychological strain of stereotype threat, which disrupted their performance and produced a racial performance gap; in contrast, when the questions were framed as nonevaluative (which is very rarely how tests are presented in the real world), the racial performance gap disappeared.

This finding has been replicated many times, not just in the laboratory, but also in real-world settings. For example, California administers two sets of exams with similar content but different psychological effects, closely mirroring the threat and no threat conditions outlined above. The state-mandated high school exit exam represents a high-stakes testing environment (threat) that students must pass to graduate from high school, whereas state achievement exams are low-stakes because students’ performance on these tests does not affect their ability to graduate, grades, or other academic outcomes (no threat). In a study of these exams, Sean Reardon and colleagues found that black and Latino students tended to perform just as well as white students on the low-stakes achievement exams, but significantly worse on the high-stakes exit exams.⁶⁸

The effects of stereotype threat have also been documented experimentally many times through randomized controlled trials, where half of students were assigned to participate in brief exercises designed to reduce stereotype threat.⁶⁹ By administering these exercises at the outset, psychologists have directly measured the effect of stereotype threat on black students’ academic achievement by observing what happens when this psychological strain is mitigated as a factor.

Meta-analyses⁷⁰ examining the performance of the thousands of students who have participated in stereotype threat-reduction field experiments over the past decade estimate that the difference between academic achievement when stereotype threat has been experimentally reduced versus when the psychological environment has been left in its natural state is the equivalent of about a sixty-three point difference on the SAT.⁷¹ In fact, these meta-analyses find that, after threat is reduced, students contending with negative stereotypes about their performance (black students) actually outperform their non-stereotyped peers (white students).⁷² These results show the clear depressive effect of stereotype threat on black students’ academic performance; indeed, instead of reflecting academic ability, test scores and grades often reflect the presence of race-related psychological threat in students’ environment. Does a sixty-three point difference on the SAT matter? In the highly competitive world of college admissions, every point matters.

Importantly, this process of stereotype threat is not unique to black students; rather, it is a normal and ubiquitous response to an environment where negative stereotypes and bias about any group loom large. For example, awareness of the stereotype that older adults have problems with memory can cause older individuals to perform worse on memory tasks when the tasks are framed as diagnostic of memory capacity.⁷³ After activation of stereotypes that women are worse than men at math⁷⁴ and driving,⁷⁵ women underperform compared to men on math and driving tests, whereas no gender differences are observed when these stereotypes are not made salient. Similarly, when an athletic task is framed as a test of natural athletic ability, white men perform worse than black men;⁷⁶ when white men think that they are being compared in math ability to Asian men, their performance on math tests declines.⁷⁷ Stereotype threat results in the greatest performance deficits for those who identify most with the threatened group identity (for instance, their race) and the domain in which they are stereotyped (for example, school performance),⁷⁸ but can affect members of any group that is negatively stereotyped, regardless of whether they believe that the stereotype is true.

In the extensive body of research on race-based stereotype threat (or any other kind of stereotype threat), there is no empirical nor theoretical evidence that relative economic advantage shields black students from the effects of racial stereotype threat. A black student who is economically advantaged is aware of and contends with negative racial stereotypes of intellectual ability and therefore is just as susceptible as a black student who is economically disadvantaged to the harmful effects of stereotype threat due to his or her race.

In fact, to the extent that economically advantaged black students are participating in activities associated with higher socioeconomic status of which blacks have not typically been a part (including ballet, playing the violin, and attending private schools), these students might experience more stereotype threat. Research shows that context cues, such as a small number of ingroup members present, low minority representation in brochure photographs, and even physical objects typically associated with the dominant, non-stereotyped group can trigger stereotype threat.⁷⁹ Based on this work, middle-class blacks who tend to have more exposure to mostly-white environments than economically disadvantaged blacks may face more consistent and stronger effects of stereotype threat because they contend with greater exposure to such cues. There is some evidence to support this proposition. Something as simple as taking an exam in a mostly white classroom can depress black students’ scores more than taking the identical exam in a room with a higher concentration of same-race students.⁸⁰

The effects of stereotype threat go beyond performance suppression. Stereotype threat can negatively affect classroom learning, prevent ability building, and undermine academic engagement.⁸¹ For example, because of stereotype threat, black students may be reluctant to join study groups, ask questions in class, or visit professors’ office hours. Avoiding each of the preceding interactions reduces the risk that the black students will say or do something that will confirm negative stereotypes of their intelligence or competence.⁸² In this respect, stereotype threat is best understood as a “double jeopardy” phenomenon, affecting not only the back-end educational dynamics (performance on tests) but front-end dynamics as well (classroom learning and study groups).⁸³

Note that the stereotype threat dynamics we have discussed could have an interactive effect in ways that compound the black student’s level of disadvantage. For example, the teacher may assume that the black student who is not speaking in class did not do the assignment or, if he did, must not have understood what he read. She may assume, further, that the black student did not seek guidance because that student is lazy, unmotivated, or disinterested in his future career. Her default with respect to why the student neither speaks in class nor visits her during office hours is unlikely to be that stereotype threat (or some other racial phenomenon, for example, racial alienation or isolation) is playing a role. The problem here, then, is not only that the teacher is unlikely to intervene to diminish the effect of stereotype threat. It is also that her reaction to the student, based on racial stereotypes (for example, that the student is lazy or unmotivated), could potentially heighten stereotype threat for the student, further decreasing the likelihood that the student will participate in class or visit the teacher during office hours.

Significantly, the consequences of this stereotype threat do not end here. The teacher in our hypothetical is unlikely to write a positive letter of recommendation for the student. The fact that the student will not have visited her during office hours and does not speak in class means that the teacher will have to rely on the very thing that is most negatively affected by stereotype threat—the student’s formal academic performance.

B. Implicit and Explicit Biases

While explicit and implicit biases are very different social phenomena, both can have a negative effect on admissions-relevant aspects of a student’s educational experience. Specifically, both forms of biases can negatively affect (1) a teacher’s substantive evaluation of a student’s written work; (2) the level of enthusiasm a teacher expresses in letters of recommendation; (3) teachers’ and administrators’ facilitation of internships and other educational opportunities (such as conferences); (4) teachers’ and administrators’ recommendations for awards; and (5) a student’s leadership opportunities. We begin with a discussion of explicit biases.

1. Explicit Biases

Explicit biases refer to biases about which we are consciously aware. They can take the form of stereotypes (such as, “black people are lazy”)⁸⁴ or attitudes (such as, “I dislike black people”).⁸⁵ For example, in 2008, 48 percent of Americans surveyed reported explicit negative attitudes towards blacks.⁸⁶ By 2013, that statistic increased to 51 percent. Stereotypes and attitudes do not always travel in the same normative direction. One might like African Americans (a positive attitude) but think they are lazy (a negative stereotype); one might dislike Asian Americans (a negative attitude) but think they are smart (a positive stereotype).

Nowadays, explicit bias is seldom stated as directly as our previous examples may imply. In other words, people will rarely say “black people are lazy” or “I dislike black people.” Instead, explicit bias is often more surreptitiously conveyed under the cover of what social psychologists call symbolic racism. Central to symbolic racism is the view that people who are consciously racist will often avoid speaking in terms that reveal their racial views.⁸⁷ In lieu of using language that directly implicates race, these individuals often employ formally race-neutral language that masks underlying antiblack attitudes or stereotypes regarding African Americans.

According to the theory of symbolic racism, white Americans’ puzzlement over the persistence of racial disparities despite an apparent end to de jure discrimination during the civil rights era⁸⁸ facilitated the development of a rhetoric that attributed black racial inequality to deficiencies in black American values and culture. In other words, under the belief system of symbolic racism, racial discrimination is no longer a barrier to blacks in education, employment, and other contexts. To the extent that blacks are underrepresented at elite colleges and universities, and overrepresented in prisons, the problem is not bad laws or governance practices but bad individual behavior or culture.⁸⁹ Because symbolic racism eschews the biological racial inferiority rhetoric of the Jim Crow era, and because it harnesses socially endorsed, traditional American values such as individualism, self-reliance, and hard work, the explanation it offers for ongoing racial disparities is not obviously in tension with civil rights norms of racial egalitarianism.⁹⁰ Symbolic racism thus captures a phenomenon in which one can argue against various forms of racial remediation, including affirmative action, using traditional American principles and values.⁹¹ The invocation of these values provides a kind of rhetorical cover for people’s residual antiblack affect.⁹²

Following its introduction in the mid-1980s, symbolic racism quickly emerged as a powerful predictor of policy attitudes regarding racialized issues such as affirmative action⁹³ as well as voter preference in elections of black candidates.⁹⁴ More importantly for our purposes, a broad set of studies has shown that symbolic racism is rooted in antiblack animosity. For example, in a study by P. J. Henry and David Sears examining the validity of the phenomenon, symbolic racism was assessed in five survey samples alongside measures of political party identification and ideology, of racial affect (whether, for example, people express warmth to African Americans and claim to like them), and of traditional black stereotype endorsement (for example, the belief that blacks are less intellectually able).⁹⁵ Using factor analyses, a statistical method designed to identify the variables to which a particular phenomenon like symbolic racism relates, the authors found that symbolic racism mapped equally onto racial animosity (in other words, traditional prejudice and antiblack affect) and conservative political predisposition (in other words, conservative ideology and conservative Republican party identification). Interestingly, the researchers found that racial animosity and political predispositions were not strongly correlated with one another, leading them to conclude that “[t]he symbolic racism belief system is the glue that joins these two elements.”⁹⁶

Although symbolic racism is rooted in residual antiblack animosity, it is demonstrably distinct. Two indications of this are that more whites support symbolic racism than old-fashioned racism, and symbolic racism has a greater effect on race-related policy outcomes than old-fashioned racism.⁹⁷

To be clear: We are not saying everyone who opposes affirmative action or other forms of civil rights-oriented policy interventions is racist, symbolically or otherwise. Our suggestion is decidedly more modest—that people who are consciously and intentionally racist typically will not express themselves in ways that betray their racial commitments. Instead, they will employ race-neutral language that aligns with traditional American values or otherwise conceals their biases. The phenomenon of symbolic racism, in short, provides at least some empirical support for the proposition that intentional and explicit biases are not a thing of the past.

Self-reports by blacks provide further support for the argument that explicit bias shapes the experience of African Americans. If hundreds of studies show that blacks report experiencing a particular social phenomenon, our collective default should be to take those reports seriously, including reports of racism. Research reveals that, across age, gender, and class, blacks consistently report more experiences of unfair treatment and discrimination than whites.⁹⁸ In one large-scale national survey, researchers found that nearly 49 percent of blacks reported encountering some form of discrimination (for example, not given a promotion, hassled by police, denied/received inferior service) in their lifetime.⁹⁹ Of these respondents, the vast majority (89.7 percent) reported race as a reason for this discrimination.

Findings of the foregoing sort apply to black students as well. One study, for example, corroborated black students’ self-reports of explicit racial discrimination by conducting field observations of the daily interactions among students, teachers, and administrators, along with interviews of students and their families. More precisely, the study included an ethnographic examination of a collection of middle-class black male students at a suburban high school.¹⁰⁰ These students and their families reported multiple experiences with racial profiling, stereotypes, and differential treatment based on race. Moreover, the researcher observed repeated differential treatment in discipline, assumptions made by teachers and administrators about the deviance and lack of intelligence of black students, and frequent racist discourse about black students’ performance.¹⁰¹ A particularly significant finding in the study is that “class privilege could not shield [the middle-class black students] from” the foregoing experiences. Indeed, students across class often experienced the institutional culture of the school as a “racial ‘witch-hunt.’”¹⁰²

Scholarship on graduate students similarly reveals robust self-reports of racism.¹⁰³ One study engaged seventy-four black doctoral and graduate students from over seventy U.S. colleges and universities. The researchers measured students’ experiences of discrimination (as recorded in student diaries) every day for fourteen days. They found that black students reported experiencing racial discrimination approximately every 3.5 days on university campuses. Furthermore, blacks who reported racial discrimination were also likely to report anxiety, feelings of negativity, and depression. There was no indication in the study that middle-class status inoculated black students from the foregoing experiences.

A final context in which to analyze black students’ self-reports of racism is undergraduate education. Recall the wave of student protest across the country this past year. Black students employed this protest to highlight their sense of isolation and alienation at predominantly white colleges and universities.¹⁰⁴ By the end of the calendar year, African Americans at Harvard, Princeton, Yale, Columbia, Indiana University, University of Missouri, and many other schools (ranging widely in tuition cost and admissions requirements) littered the media with reports of subtle and blatant forms racism. Focusing specifically on the University of Missouri, the Huffington Post ran an editorial entitled, “It shouldn’t be so hard to accept that racism is a problem at Mizzou.”¹⁰⁵ The article chronicled some of the racist acts faced by black college students at the University of Missouri, including these.

On the morning of 26, 2010, in the final days of Black History Month, students woke up to find cotton balls spread across the grounds in front of the Black Culture Center on campus—a scene evoking slavery.
A year after the cotton ball incident, also during Black History Month, a racist slur was spray-painted on a statue outside a
On the night of 5, 2015, members of the Legion of Black Collegians, a historic black student government group, were rehearsing for a homecoming performance on stage. Some members had a heated exchange with what they later described as an “obviously intoxicated” young white male who interrupted their rehearsal to question why they were there. When he stumbled off the stage and fell he was heard saying into his cellphone: “These n****rs are getting aggressive with me.”

The New York Times published a similar list that included instances of racial profiling by campus police officers, offensive Halloween costumes, racial slurs in school newspapers and classroom discussions, and theme parties at which students appear in blackface and wear modes of dress and makeup stereotypically associated with African Americans. This list likely includes only the most public and unequivocal manifestations of racist conduct, not the entire universe of such events.¹⁰⁶ The above examples, and our broader discussion of black self-reports of racism, suggest that it is a mistake to think that explicit racial bias on college campuses is a thing of the past.¹⁰⁷

Still, we have not linked our discussion of explicit biases to the domain of student experiences with which we began: admissions-relevant student experiences. In other words, we have not pointed to a moment in which a particular teacher in a particular institutional setting evaluated a particular student negatively because of that student’s race. But because very few teachers are likely to announce their racism, the absence of that empirical showing should hardly surprise us. This bring us to the implicit bias literature. That body of work provides an evidence-based way of linking racial bias to teachers’ and admissions officers’ evaluation of students.

2. Implicit Biases

Implicit bias refers to the attitudes and/or stereotypes that exist in the absence of a person’s intention, awareness, deliberation, or effort. Unlike explicit bias, which reflects people’s attitudes and beliefs that they consciously endorse, implicit bias results from cognitive processes that operate at a level below conscious awareness and without intentional control.¹⁰⁸

Most implicit attitude measures rely on reaction times,¹⁰⁹ and the Implicit Associations Test (IAT) is perhaps the most well-known reaction time task. The IAT gauges differences in how easy or difficult it is for people to associate individual exemplars of various social categories (whites vs. blacks, rich vs. poor, gay vs. straight, and so on) with abstract words and categories with evaluative implications (such as good vs. bad, pleasant vs. unpleasant). The IAT is based on the assumption that any exemplar (for instance, black) is cognitively associated with a corresponding evaluation (for instance, good or bad) and that pairing the exemplar with the corresponding evaluative words (for instance, black = bad) results in faster reaction times than pairing the exemplar with unrelated or incongruent evaluative words (for instance, black = good). The IAT is commonly used in laboratory and field research and its reliability¹¹⁰ and validity¹¹¹ are well documented. Considerable research has indicated that most Americans display a pro-white/antiblack bias on the IAT.¹¹² This bias has been demonstrated in children as young as six years old.¹¹³ In short, implicit bias is an invisible but pervasive reality of American life.

As Rachel Godsil observes, “The power of implicit bias to undermine educational opportunities of students of color are obvious—their contributions may fail to be recognized for their merit, they may well experience incidents in which they are treated differently by teachers, peers, and administration, or even assumed not to be students at all.”¹¹⁴ Below we explain why Godsil is right. More precisely, we discuss the relationship between implicit bias and educational outcomes. We begin with a 2010 study by Dr. van den Bergh and colleagues that explores the relationship between the implicit biases of teachers and the overall academic performance of students.¹¹⁵ We then explore how implicit biases affect both the evaluation of writing and the substantive content of letters of recommendations. As we will explain, each of the foregoing contexts in which implicit biases plausibly operate can impact the perceived competitiveness of a student’s admissions file.

a. Implicit Biases and Overall Academic Performance

The study by Dr. van den Bergh and colleagues focused on both implicit and explicit biases; however, our analysis will highlight the implicit bias dimensions of their findings. In short, the researchers found that teachers who scored high in implicit racial bias produced racial minority students with poor academic performance.¹¹⁶ Their study compared teachers’ implicit racial attitudes (as measured by the IAT) with actual students’ standardized test scores.¹¹⁷ They found that racial minority students in classes of teachers with negative implicit attitudes showed lower standardized test scores than racial minority students of teachers with more positive implicit attitudes. The study also found that the achievement gap between racial minority and racial majority students was higher among classes whose teachers had more negative implicit attitudes.

At the same time, there are limitations to the van den Bergh study. The correlational nature of the findings, the fact that the overall observed impact is modest (7–8 percent of the variance), and that the study was conducted with Dutch and Turkish/Moroccan students in a different educational context all suggest that we should interpret the results of the study with caution. On the other hand, the findings are consistent with studies in the United States showing that test scores of girls in math classes taught by teachers with negative implicit biases are lower than the test scores of girls from teachers with more positive implicit attitudes.¹¹⁸ Black students could be vulnerable to a similar dynamic, particularly because implicit bias is so widespread. In short, there is reason to believe that, in classrooms all across America, the implicit biases of teachers are negatively impacting black students’ test scores—the very test scores admissions officers use to screen students for admissions.

b. Implicit Biases and the Evaluation of Writing

There are other ways in which implicit biases can impact teachers’ evaluation of black students. Consider, for example, a relatively recent study in which evaluators rate the writing skills of blacks and whites.¹¹⁹ Reeves and colleagues simulated an evaluation of a legal memo with actual law firm partners. This study was not conducted on college campuses, but its high external validity offers a strong parallel to teachers’ and advisors’ evaluation of the work of black and white students. In this mock evaluation of legal memos, the researchers manipulated whether partners from several law firms in the United States reviewed and rated a legal memo ostensibly written by Thomas Meyer, either a black or a white associate. The study found that partners who reviewed the memo written by the black Thomas Meyer identified more errors, had more critical comments, and rated the memo worse than the white Thomas Meyer.

Interestingly, bias was found at the data gathering stage, when partners were searching for errors and critiquing the memo, and not at the rating stage, when partners were rating the overall quality of the writing effectively by adding up all of the errors. In other words, when Meyer was “black,” partners searched for and found more errors in the memo than when Meyer was “white.” Blackness seemed to have engendered a more critical eye. Then once the errors and negative comments were tabulated, the black Thomas Meyer was scored lower than the white Thomas Meyer.

A critical dimension of the study is that it suggests that it matters precisely when and where we look for bias. The partners discriminated on the front-end of the evaluation (when they were assigning errors), not on the back-end (when they aggregated those errors and made an overall evaluation of the writing). The fact that the partners followed fair procedures (counting the number of errors) to rate the memos did not cure the prior act of discrimination (the allocation of more errors to the memos they thought black attorneys wrote).

Think about the implications of the foregoing study for black students. Students write multiple essays—across the curriculum—in high school. The cumulative effect of the writing evaluation bias that Reeves and colleagues found could be quite significant.

c. Implicit Biases and Letters of Recommendations and Mentoring

Consider another implicit bias problem that a black student might encounter: letters of recommendation that coaches, teachers, and religious leaders write to champion the student’s abilities and promise. Implicit bias has been shown to affect how people write letters of recommendation, with implications for the success of the applicant in a competitive applicant pool.

In one study,¹²⁰ researchers conducted an analysis of three hundred recommendation letters for women and men applying for medical school faculty positions and found that: (a) letters written for women were shorter; (b) they contained fewer assurances of competence and achievements but more assurances of compassion; and (c) they portrayed women as students and teachers rather than researchers and professionals. These differences led evaluators to perceive women as less competent than men, likely contributing to gender gaps in hiring, advancement, and receiving grants.¹²¹ Here, similar to the essay-writing study, bias altered how individuals constructed the abilities of women compared to men.

Studies of letters of recommendation show precisely how powerful implicit bias can be. Letter writers typically prepare letters as a personal favor to the applicant and often spend hours of their personal time to champion their student’s cause. Letter writers’ motivation to be racially biased in this context is low. If anything, their motivation to be biased in favor of the letter recipient is high. Yet even in a context where decisionmakers—here, letter writers—are motivated to do the right thing, implicit bias can constrain their capacity to do so.

d. Implicit Biases and Evaluations of Resumes

At least two studies suggest that implicit biases can shape how decisionmakers evaluate resumes. In one study, researchers¹²² asked male and female participants to evaluate the same resume with a randomly assigned male or female name. They found that both male and female evaluators gave the male applicants better evaluations in teaching, service, and research and were more likely to hire male over female applicants. This study is relevant to our analysis not only because black women, as women, likely experience this dynamic,¹²³ but also because they likely face a similar dynamic with respect to race. Indeed, in a comparative study of job applications with African American-sounding names and white-sounding names,¹²⁴ researchers found that applicants with African American-sounding names had to send an average of fifteen resumes to get one call back as opposed to ten resumes for applicants with white-sounding names.¹²⁵ Moreover, the study found that, with respect to callback rate, having a white sounding names was equivalent to having eight additional years of experience.¹²⁶

These resume studies are relevant to our analysis in at least two ways. First, faculty typically ask students for resumes as a predicate to writing letters of recommendations and refer to those resumes in the letters they write. There is reason to be concerned that, because of implicit biases, faculty are under-reading the resumes of African American students and thus under-describing black students’ abilities and promise. Second, admissions officials will read the resumes of prospective students to get a holistic picture of applicants and to assess students’ overall qualifications. Here, too, because implicit bias is so widespread, there is the potential for under-reading.

Of course, it’s hard to know the extent to which the resume bias problem we have described is a problem. We are aware of no empirical study quantifying the effect of resume bias on black college applicants. Our point is simply to suggest that there is reason to believe black applicants to colleges and universities are impacted by resume bias. It is another racial disadvantage black applicants across class likely experience that potentially renders their admissions files less competitive than those files would otherwise be.

C. Negative Institutional Culture

The final racial disadvantage we discuss that likely impacts the competitiveness of black students’ admissions files is negative institutional culture. To understand this phenomenon, it is helpful to note that targets of discrimination (for example, African Americans) focus not only on whether the people around them are explicitly or implicitly biased, but on whether the institutional cultures in which they are situated are explicitly or implicitly biased. In other words, socially marginalized groups focus on both discriminatory people and discriminatory institutions. They look for “cues” in the environment to ascertain whether that environment is one in which they are likely to succeed.¹²⁷ The following are examples of environmental cues to which black students are likely to attend:

Racial Demographics of Student Body. The lower the representation of black students within a particular institutional context, the greater the likelihood that those students will perceive that institution as unwelcoming and/or as one in which they are unlikely to ¹²⁸
Governing Ideology. The more an institution insists that race does not matter and encourages members within the institution to avoid discussing or engaging race, the more likely black students are to distrust the institution and question whether they belong.¹²⁹ In this respect, a school that promotes a formal commitment to colorblindness is going to feel less welcoming to black students than a school that encourages Black students may interpret a school’s commitment to colorblindness as a signal that black students should “tone down” the degree to which they are black and “act white.”¹³⁰ The more formally colorblind a school institutionally feels to black students, the stronger the pressure black students might experience to leave their “whole person” at home,¹³¹ compromise their sense of identity, and signal racial palatability.¹³²
Curricular Offerings and Content. There are at least two dimensions to how curricular offerings and content function as environmental First, institutions that offer very little in the way of courses on cultural and historical experiences outside of white European civilization are less likely to be ones in which black students feel comfortable or engaged.¹³³ We call this the “what teachers teach” dimension of the curriculum. But there is a “how teachers teach” problem as well. A teacher might teach the history of the U.S. Constitution without discussing race; engage in debates about whether black students have a right to attend, or are smart enough to be at, the very institution in which they are taking the course; explore “both sides” of the argument as to whether blacks are more criminally inclined than whites; and call on black students only when the topic of race comes up. We recognize that a teacher might offer pedagogical reasons for pursing any of the foregoing modes of engagement, including the rationale that they facilitate the robust exchange of ideas. Our point is to highlight a potential distributional effect of how teachers teach: The creation of an unwelcoming institutional environment for black students that diminishes the likelihood that black students will succeed.
Faculty and Administrative Leadership Demographics. Faculty and staff diversity, and not just student diversity, is an important environmental cue for black This might explain why, year after year, students across the country protest the lack of faculty and staff diversity.¹³⁴ Underlying this effort is not just a commitment to antidiscrimination and equality but a sense that the more racially diverse the faculty and administrative leadership of a school, the more African Americans are going to feel like they belong at that school. There are three more concrete ways to understand why black students might pay attention to faculty and staff diversity.
First, faculty and staff can serve as mentors—even for students whose areas of interest do not converge with the faculty or staff member’s. Second, black faculty and staff can serve as role models, signaling by their very presence that black students belongs at and can thrive in the institution. Third, black faculty and staff can function as bridge-builders to multiple parts of the university, including facilitating interactions between black students and nonblack faculty and staff in the students’ areas of interest. The foregoing three reasons explain why black students are likely to pay attention to faculty and staff diversity. That diversity is an environmental cue for the kind of faculty and staff relationships students think they will have, the level of difficulty they think they will experience navigating the institution, and the overall racial climate of the school.
Institutional Signs of Stigma and Exclusion. Given the history of racism in educational access in the United States, many colleges and universities continue to harbor enduring signs of stigma and exclusion, manifested, for instance, in the names of Buildings at both elite and nonelite colleges and universities are sometimes named after people who supported the most racially subordinating aspects of American history, for example, slavery. Yet, these are precisely the spaces in which black students are expected to learn and take exams. Over the past year, students across the country have been organizing to rename numerous buildings, explaining that the existing names stigmatize students of color and send a message that they do not belong.¹³⁵
A similar point can be made about college and high school mascots. For decades, Native Americans have been organizing around this issue. The problem here is not just one of cultural exploitation and appropriation,¹³⁶ it is also one of racial stigma and environmental marginalization. The existence of these sports mascots stigmatizes Native Americans as a group, depressing self-esteem, feelings of community worth, and perceived possibilities for future academic success.¹³⁷ This increased stigma renders the universities and high schools that the mascots purport to represent less welcoming and more hostile institutional spaces.¹³⁸
There are more subtle signs of stigma and exclusion in educational settings that bear emphasis as well. Consider, for example, the practice of displaying the portraits of the founding faculty members on the walls throughout the university. This ubiquitous and seemingly neutral practice exacts a cost on students who were excluded during the colleges’ founding. Inevitably, the portraits hang on the wall of the most prominent parts of the university, including classrooms. Inevitably, the figures these portraits represent are white and male. And, inevitably, some of people the images depict have meaningful ties to slavery or Jim Crow. Yet, black students must interact with those portraits—sometimes daily—and with the legacies of exclusion those portraits represent. Those interactions serve as an implicit reminder that the institution is not one in which black students have historically belonged.
Now consider a private high school at which the annual tradition during graduation is for the head of school to instruct parents or grandparents who are alumni of the school to stand up. After they do so, the principal invites everyone in attendance to recognize this group with a round of applause. Given, again, the history of race and educational access, that moment of recognition—which is not intended to be racially communicative—sends a clear signal about which groups, historically, have been insiders at the school and which have been outsiders. Every year the school sends the message.
Law Enforcement Police surveillance on high school and college campuses is another environmental cue on which black students focus. While the data on racial profiling on high school and college campuses is decidedly incomplete, black students across the country report being racially profiled by campus police.¹³⁹ The more black students on a particular college campus feel vulnerable to being racially profiled on a school’s campus, the more those students will feel like outsiders and potentially disengage from the school’s community.

Assuming that we are correct that the foregoing six items function as environmental cues for African Americans, one might reasonably ask whether evidence exists linking those items to the competitiveness of a black applicant’s admissions file. There is some, though not much, direct evidence on this point.

First, research shows that the more one feels like an outsider within a particular institutional setting, the less well one will perform. Consider, for example, a small group of black students taking an exam in a predominantly white environment. As we indicated earlier, the black students’ underrepresentation can operate as a negative environmental cue that compromises the students’ performance on the exam.¹⁴⁰ In a specific test of this reasoning, female students in one study who were exposed to a computer science classroom containing items typically associated with computer scientist “geeks” (such as a Star Trek poster, video games, and computer parts) reported less interest in computer science, manifested more self-consciousness about how they were perceived, and experienced less comfort fitting in with computer scientists compared to when they were exposed to a computer science classroom with neutral cues (such as a nature poster, books, and coffee mugs).¹⁴¹ These discrepancies persisted even when the female students thought they would be joining an all-female team of computer scientists.

Second, environmental cues can affect how white students view black students—and in ways that are relevant to admissions. Something as seemingly innocuous as the visual imagery of a school—literally, the posters on the wall—can affect how white students perceive and interact with black students. The types of posters and materials that predominantly white schools display to recognize Black History Month generally focus on cultural diversity, without explicit mentions of historical racism or slavery. Predominantly black schools, by contrast, generally emphasize black history in particular, explicitly referencing slavery, racism, and the Civil Rights Movement.¹⁴² The foregoing two clusters of images—the cultural diversity imagery from predominantly white schools and the black civil rights imagery from the predominantly black schools—create different environmental cues for white students. In an experimental setting with white participants who were completely blind to the origin of the materials, exposure to the predominantly white school materials diminished white students’ (1) perception of racism against blacks, (2) recognition of race-related barriers to opportunity, and (3) support for antidiscrimination remedial interventions.¹⁴³

The foregoing findings suggest that white students pay attention to environmental cues as well. The specific environmental cues of visual imagery at predominantly white schools make white students less willing to acknowledge the existence of racism and less motivated to mitigate the interpersonal and institutional effects of the phenomenon. White students’ disinvestment in antiracism along the preceding lines is not cost-free. Black students incur the expense. The lower white students’ level of regard for racism, the higher black students’ sense of alienation and exclusion and the less likely black students are to thrive in ways that enhance their college application profile.

Another way environmental cues impact white students to the detriment of black students takes us back to our discussion about racial demographics. The more black students are underrepresented numerically in classrooms, the less likely their peers are to treat them as leaders, even when they are well-liked (which is unusual, because perception of leadership skills and popularity typically overlap).¹⁴⁴ Perceptions of black students both as leaders and as likeable figures improve as their numerical representation increases.¹⁴⁵ In contrast, perceptions of white students do not depend on their numerical representation; whether white students are the numerical minority or majority, they tend to be rated highly by their peers with respect to both likeability and leadership.¹⁴⁶ These findings directly bear on admissions. Admissions officials consider demonstrated leadership activity a plus factor in a student’s admissions file. Because black students at predominantly white schools have lower likeability and leadership ratings than white students, white students are less likely to elect these students to precisely the kind of leadership positions (in, for example, student government) that admissions officials look for when they read applications.

* * *

This section began with a twofold question: Is there empirical evidence to support our contention that African Americans across class are vulnerable to the racial disadvantages we discussed in Part I, and is there evidence linking those disadvantages to critical aspects of an admissions file? The answer, we have suggested, is “yes.” To summarize: Empirical evidence reveals that (1) environmental cues impact how black students experience their institutional environments in ways that can compromise their academic performance and access to extracurricular activities and leadership roles; (2) stereotype threat negatively impacts black students’ performance on standardized test, grades, and overall intellectual engagement inside and outside the classroom; and (3) explicit and implicit biases potentially negatively impact everything from the way teachers grade and evaluate the work of black students, to whether and how they mentor and discipline black students, to whether teachers actively engage black students in class and/or encourage them to attend office hours, to the tone and substance of the letters of recommendations teachers write on behalf of black students, to whether teachers are responsive to black students’ questions or concerns, to whether they facilitate internships or educational-enhancing job opportunities for black students.

For reasons of analytical clarity, we discussed the foregoing racial disadvantages seriatim. In fact, however, they mutually reinforce and amplify each other,¹⁴⁷ creating an admissions scale that looks something like Figure 4:

Figure 4

As we explained earlier, we cannot articulate precisely how much the scales lean in the direction Figure 4 depicts. And, note that, for the most part, we have focused on how whites are advantaged by black disadvantages. We had said very little about white racial advantages per se.¹⁴⁸ A significant dimension of our model is that black disadvantage appears as a thumb on the scale for whites. We are largely agnostic on whether one should frame the imbalance Figure 4 represents in terms of white advantages, black disadvantages, or both. Our broader project is to suggest that affirmative action does not enter an admissions world in which blacks and whites are competing on a level playing field. To mix metaphors, the scales are already tipped in favor of whites. That is precisely the imbalance affirmative action attempts to counteract.

Conclusion

Our goal in this Essay was to challenge the dominant framing of affirmative action as a racial preference that privileges blacks who are not disadvantaged. We did so by highlighting the racial disadvantages African Americans across class experience up to and in the context of admissions. We suggested that these disadvantages can diminish the overall competitiveness of an applicant’s admissions file. Because affirmative action attempts to offset this negative effect, judges, policymakers, and the public at large should view the policy as a mechanism to level the playing field, not a racial preference.

We conclude by suggesting that part of the reason it is difficult for people to see affirmative action as a mechanism that levels the playing field for all African Americans is because the debate about affirmative action overstates the middle-class status of blacks. Recall that one of the images of African Americans that opponents of affirmative action paint is of a person who is not racially disadvantaged but economically privileged. The representation of African Americans as class-privileged completely elides that black middle-class families are frequently far from economically secure—and generally not nearly as economically stable as average white middle-class families.

A recent report by Demos and the Institute for Assets and Social Policy at Brandeis University highlights a significant “racial wealth gap” between whites and blacks.¹⁴⁹ Across multiple dimensions along which families might accumulate wealth—homeownership/economic return on homeownership, educational attainment/economic return on educational attainment, income level/economic return on income, and inheritance—whites do better than African Americans.¹⁵⁰

Moreover, the foregoing racial wealth gaps between whites and blacks transcend class. For example, “for every $1 in wealth that accrues to median Black households associated with a higher income, median white households accrue $4.06.”¹⁵¹ Similarly, “for every $1 in wealth that accrues to Black families associated with a college degree at the median, White families accrue $11.49.”¹⁵²

As with any study, the one from which we derive the preceding data is undoubtedly vulnerable to criticism. But the broad findings of the research—that there is a meaningful wealth gap between whites and blacks that persists across class—is largely (if not entirely) uncontroversial. All of this is to say, the black middle class and the white middle class are in very different economic positions. The former is less economically secure than the latter.¹⁵³

Yet opponents of affirmative action continue to paint the black middle class as economically privileged. Some go so far as to treat black middle-class applicants to colleges and universities as if they were the children of President Obama. Consider, for example, what Richard Kahlenberg has said about the black middle class: “But as President Obama has acknowledged, daughters like his—who have grown up with privilege—don’t need preferential consideration when applying to college. Yet it is privileged minority students who are the center of Fisher vs. University of Texas II . . . .”¹⁵⁴

Kahlenberg’s argument here is just one example in which critics of affirmative action sometimes employ “Sasha and Malia Obama as the poster children for privileged blacks who do not deserve to, but who are most likely to, benefit from current race-conscious affirmative action.”¹⁵⁵ President Obama himself has been asked whether his children should benefit from affirmative action,¹⁵⁶ as if that question is somehow relevant to whether the average black middle-class child should benefit from the policy.

It is hard to imagine a political debate about affirmative action, or educational policy more generally, in which the children of a white president are invoked to raise questions about the fairness of admissions for the rest of the country’s youth. People would be quick to note that, with respect to privilege, those children are in an exceptional position vis-à-vis other children—and certainly few, if anyone, would employ a white president’s children to stand in for white middle-class applicants to colleges and universities.

Obama’s children, on the other hand, do stand in for black middle-class college applicants. Notwithstanding that, in terms of privilege, the President’s daughters are obviously unrepresentative of African Americans as a whole, opponents of affirmative action continue to ask rhetorically whether Sasha and Malia should be beneficiaries of affirmative action. The answer in the negative extends beyond Obama’s children. It applies to black middle-class students writ large. This is precisely why liberals and progressives must more forthrightly defend affirmative action for African Americans who are middle-class. Doing so both exposes the simplistic and inaccurate way in which opponents of affirmative action describe middle-class African Americans (for example, as the children of Obama) and provides an opportunity to explain why it is a mistake to conceptualize affirmative action as a racial preference.

Our hope is by providing a specific indication of the ways in which racial disadvantage transcends class in the context of admissions—even for middle-class black students—we have disrupted the foregoing dominant and misleading representations of race, class, and racism: the framing of black middle-class students as economically privileged, not racially disadvantaged, and the framing of affirmative action as reverse discrimination, not racial remediation. These representations (in which liberals have acquiesced) have been normatively and doctrinally consequential. They have made affirmative action less constitutionally secure and more difficult normatively to defend—and they have made white applicants to colleges and universities the underserving victims of racial discrimination and black applicants the underserving beneficiaries of racial preferences.

[1]. See generally Luke Charles Harris & Uma Narayan, Affirmative Action and the Myth of Preferential Treatment: A Transformative Critique of the Terms of the Affirmative Action Debate, 11 Harv. BlackLetter L.J. 1 (1994). We focus on African Americans in this Article for the same reasons Harris and Narayan do: “[T]he hostility to, and the ambivalence about, affirmative action policies is most powerfully articulated in the context of discussions about race-based policies as they pertain to African Americans. We think that it is easier to stereotype these policies when Blacks are viewed as their principal beneficiaries . . . .” Id. at 3.

[2]. See Maurice R. Dyson, De Facto Segregation & Group Blindness: Proposals for Narrow Tailoring Under a New Viable State Interest in Pics v. Seattle School District, 77 UMKC L. Rev. 697, 737 (2009); see also Michelle Alexander, The New Jim Crow: Mass Incarceration in the Age of Colorblindness 221–61 (2012); Sheryll Cashin, Place, Not Race: A New Vision of Opportunity in America (2014); Lisa J. Holmes, Comment, After Grutter: Ensuring Diversity in K-12 Schools, 52 UCLA L. Rev. 563, 592–94 (2004); Sharon Hsin-Yi Lee, Comment, Justifying Affirmative Action in K-12 Private Schools, 23 Harv. BlackLetter L.J. 107, 125–26 (2007).

[3]. Though Richard Sander is not the first person to articulate the theory of mismatch, his scholarship in the area has arguably garnered the most traction. See generally Richard H. Sander, A Systematic Analysis of Affirmative Action in American Law Schools, 57 Stan. L. Rev. 367 (2004); Richard Sander & Stuart Taylor, Jr., Mismatch: How Affirmative Action Hurts Students It’s Intended to Help, and Why Universities Won’t Admit It (2012). There is a fairly broad literature criticizing his methodology. See generally Ian Ayres & Richard Brooks, Does Affirmative Action Reduce the Number of Black Lawyers?, 57 Stan. L. Rev. 1807 (2005); David L. Chambers et al., The Real Impact of Eliminating Affirmative Action in American Law Schools: An Empirical Critique of Richard Sander’s Study, 57 Stan. L. Rev. 1855 (2005).

[4]. 771 F.3d 274 (5th Cir. 2014), cert. granted, 135 S. Ct. 2888 (2015).

[5]. See Brief Amicus Curiae of Gail Heriot and Peter N. Kirsanow, Members of the United States Commission on Civil Rights in Support of Petitioner, Fisher v. Univ. of Tex. at Austin, No. 14-981, (U.S. Sept. 9, 2015); Brief Amicus Curiae of Pacific Legal Foundation et al. in Support of Petitioner, Fisher v. Univ. of Tex. at Austin, No. 14-981 (U.S. Sept. 10, 2015).

[6]. See, e.g., Cheryl I. Harris, Fisher’s Foibles: From Race and Class to Class Not Race, 64 UCLA L. Rev. Disc. (forthcoming 2016)

[7]. Cf. Khalil Gibrand Muhammad, The Condemnation of Blackness: Race, Crime, and the Making of Modern Urban America (2010) (discussing the ways in which crime is written into race).

[8]. Id. at 272 (“Nothing in the world is easier in the United States than to accuse a black man of a crime.” (quoting Du Bois, Negro Editors on Communism, Crisis (1932))).

[9]. We are not suggesting that proponents of mismatch theory are necessarily racist. Our claim is that the theory helps to legitimize and further entrench a pernicious racial stereotype about African Americans. For a review of the literature criticizing mismatch theory, see Richard Lempert’s essay in this symposium. Richard Lempert, Mismatch Theory in an Empirical Light, 64 UCLA L. Rev. Disc. (forthcoming 2016).

[10]. See Tung Yin, A Carbolic Smoke Ball for the Nineties: Class-Based Affirmative Action, 31 Loy. L.A. L. Rev. 213, 225 (1997).

[11]. One might also ask why opponents of affirmative action frame the remedial issue as though it involved a choice between racial remediation or class remediation. On this point, see Cheryl Harris’s contribution to the symposium. Harris, supra note 6. Sometime opponents of race-based affirmative action also argue that class-based affirmative action is an effective mechanism to achieve racial diversity. Bill Kidder explains why focusing on class alone will fail to racially diversify colleges and universities. William C. Kidder, How Workable Are Class-Based and Race-Neutral Alternatives at Leading American Universities?, 64 UCLA L. Rev. Disc. (forthcoming 2016); see also Liliana M. Garces, Lessons From Social Science for Kennedy’s Doctrinal Inquiry in Fisher v. University of Texas II, 64 UCLA L. Rev. Disc. 18 (2016) (highlighting the diversity costs of eliminating affirmative action). One might also ask why opponents of affirmative action rarely criticize legacy admissions. Generally, students who are legacy admits are decidedly privileged and take advantage of having a generational relationship to elite colleges and universities. And what about class-privileged whites more generally? Why don’t opponents of affirmative action target them? After all, their private school trajectories are often gateways to America’s most prestigious universities.

[12]. Harris & Narayan, supra note 1, at 7.

[13]. Whites, on the other hand, are presumptively neither. The most salient image of whites in the affirmative action debate is as innocent victims of reverse discrimination.

[14]. Of course, African Americans of all class backgrounds are included in the diversity rationale for affirmative action. Our point is that proponents of affirmative action rarely expressly name and defend African Americans who are not class disadvantaged as beneficiaries of the policy.

[15]. See Kermit Roosevelt III, The Ironies of Affirmative Action, 17 U. Pa. J. Const. L. 729, 750 n.65 (2015).

[16]. See generally Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265 (1978) (Powell, J., concurring) (suggesting that diversity is a compelling state interest for affirmative action because diversity contributes to the “robust exchange of ideas”).

[17]. David S. Schwartz, The Case of the Vanishing Protected Class: Reflections on Reverse Discrimination, Affirmative Action, and Racial Balancing, 2000 Wis. L. Rev. 657 (2000). But see Devon W. Carbado & Cheryl I. Harris, The New Racial Preferences, 96 Calif. L. Rev. 1139 (2008) (explaining how the formal commitment to colorblindness in the context of admissions produces racial preferences).

[18]. Though we do not develop the intraracial diversity argument here, we think it is important to make clear that it is a significant benefit of affirmative action programs. See generally Devon W. Carbado, Intraracial Diversity, 60 UCLA L. Rev. 1130 (2013); Vinay Harpalani, Diversity Within Racial Groups and the Constitutionality of Race-Conscious Admissions, 15 U. Pa. J. Const. L. 463 (2012); Elise Boddie, Commentary on Fisher: The Importance of Diversity Within Diversity, SCOTUSblog (Oct. 11, 2012, 10:50 AM), http://www.scotusblog.com/2012/10/commentary-on-fisher-the-importance-of-diversity-within-diversity/ [https://perma.cc/2GXA-E9F3].

[19]. See infra Part I.

[20]. See Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 314 (1978) (plurality opinion) (“[T]he interest of diversity is compelling in the context of a university's admissions program . . . .”); see also Grutter v. Bollinger, 539 U.S. 306, 329 (2003) (“Our conclusion that the Law School has a compelling interest in a diverse student body is informed by our view that attaining a diverse student body is at the heart of the Law School's proper institutional mission . . . .”).

[21]. See infra Part I.

[22]. It bears mentioning that Justice Powell explicitly rejects social discrimination as a justification for affirmative action. Bakke, 438 U.S. at 310 (“[T]he purpose of helping certain groups whom the faculty of the Davis Medical School perceived as victims of ‘societal discrimination’ does not justify a classification that imposes disadvantages upon persons like respondent, who bear no responsibility for whatever harm the beneficiaries of the special admissions program are thought to have suffered.”).

[23]. See generally Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013).

[24]. Transcript of Oral Argument at 43–44, Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (No. 11-345).

[25]. Id. at 43. The caveat is that if a university admits that it engaged in intentional discrimination in the past against a particular group, that institution may use that “identified discrimination” as a remedial justification for affirmative action. See Bakke, 438 U.S. at 309, 343.

[26]. See generally William Julius Wilson, The Truly Disadvantaged: The Inner City, The Underclass, and Public Policy (2d ed. 2012).

[27]. Transcript of Oral Argument at 60, Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (No. 11-345).

[28]. See generally Neil Gotanda, A Critique of “Our Constitution is Color-Blind”, 44 Stan. L. Rev. 1 (1991) (linking the Supreme Court’s colorblind approach to race to biological notions of race); Carbado & Harris, supra note 17 (same); Ian F. Haney López, The Social Construction of Race: Some Observations on Illusion, Fabrication, and Choice, 29 Harv. C.R.-C.L. L. Rev. 1 (1994).

[29]. See Brief Amicus Curiae of Kimberly West-Faulcon in Support of Respondents, Fisher v. Univ. of Tex. at Austin, No. 14-981 (U.S. Nov. 2, 2015) (challenging the idea of a “black bonus”).

[30]. See Lilly v. City of Beckley, 615 F. Supp. 137, 140 (S.D. W. Va. 1985).

[31]. See Richard D. Kahlenberg, The Remedy: Class, Race, and Affirmative Action xxiii–xxiv (1996); see also Valerie Strauss, On Charles Murray, the Black Lawyer’s Son, the White Plumber’s Son and College Admissions, Wash. Post (Mar. 8, 2012), https://www.washingtonpost.com/blogs/answer-sheet/post/on-charles-murray-the-black-lawyers-son-the-white-plumbers-son-and-college-admissions/2012/03/08/gIQAN
odczR_blog.html [https://perma.cc/DM2Z-8JVD].

[32]. We do not mean that critics of affirmative action are, across the board, unconcerned about working-class and poor whites. Our point is that concerns about poor whites are often rehearsed in the context of (but not outside of) debates about affirmative action and other forms of racial remediation. See Richard D. Kahlenberg, Class-Based Affirmative Action, 84 Calif. L. Rev. 1037, 1038, 1061 (1996) (discussing the benefit of socioeconomic affirmative action over “toxic . . . biological preference[s]” like race or gender; “[w]hereas a racial preference will unfairly benefit Bill Cosby’s offspring over the son of a white sanitation worker, class preferences help those who need it”). The majority of Kahlenberg’s work that discusses white poverty only does so in criticism of race-based affirmative action policies. Work that explores the ways poor whites are ignored often does not come from vocal critics of affirmative action, and is still written in contrast to ameliorative programs that seek to alleviate poverty for people of color. See Leonard Pitts Jr., White Poverty Exists, Ignored, Miami Herald (Oct. 5, 2014, 8:49 AM), http://www.miamiherald.com/opinion/opn-columns-blogs/leonard-pitts-jr/article2518087.html [https://perma.cc/J9TP-5VYJ]; see also Pablo Eisenberg, Poverty Among Whites Demands Philanthropy’s Attention, Chron. Philanthropy (Sept. 3, 2015), https://philanthropy.com/article/Opinion-Poverty-Among-Whites/232799 [https://perma.cc/T3SE-LBWA].

[33]. Importantly, African American liberals, and not just conservative whites, sometimes articulate this view. See Cashin, supra note 2; see also Mark Finkelstein, Henry Louis Gates: End Affirmative Action for Affluent African-Americans, MRC NewsBusters (Oct. 22, 2013, 9:18 AM), http://newsbusters.org/blogs/mark-finkelstein/2013/10/22/henry-louis-gates-end-affirmative-action-affluent-african-american [https://perma.cc/Y6VG-8ZSC] (“I think that for a lot of reasons, political and also practical and economic, we should think about affirmative action for the poor. And I grew up in West Virginia with poor white people. They need affirmative action as much as my people do. And I think it would be a savvy thing to reconsider. Also I'm upper middle class. My daughters were born at the Yale-New Haven hospital. They have a privileged life. Do they really need to benefit from affirmative action? Affirmative action was a class escalator when I went to Yale, and I think it still should be. So I want to get more more [sic] poor black people into the middle class and I want to get more poor white people in the middle class as well.” (quoting Henry Louis Gates on Morning Joe (MSNBC television broadcast Oct. 22, 2013))).

[34]. See generally Stephan Thernstrom & Abigail Thernstrom, America in Black and White: One Nation, Indivisible (1997).

[35].. See Jason Irizarry, Cultural Deficit Model, Education.com (Dec. 23, 2009), http://www.education.com/reference/article/cultural-deficit-model/ [https://perma.cc/T3JK-FGD8].

[36]. See Katherine Y. Barnes, Is Affirmative Action Responsible for the Achievement Gap Between Black and White Law Students?, 101 Nw. U. L. Rev. 1759, 1763 (2007).

[37].. See Albert G. Mosley & Nicholas Capaldi, Affirmative Action: Social Justice or Unfair Preference? 122 (1996).

[38]. See Devon Carbado & Mitu Gulati, Acting White?: Rethinking Race in “Post Racial” America 1–2 (2013).

[39]. See Dinesh D’Souza, The End of Racism: Principles for a Multiracial Society xxv–xxvi (1995).

[40]. See D’Souza, supra note 39, at xxv–xxvi.

[41]. Even President Obama encouraged black students not to associate studying with “acting white.” See Ivory A. Toldson, The ‘Acting White Theory’ Doesn’t Add Up, Root (Jan. 30, 2013) http://www.theroot.com/articles/politics/2013/01/acting_white_theory_black_academic
_achievement_based_on_other_factors.html [https://perma.cc/PZ8F-J5QX].

[42]. See generally Robert S. Chang, Toward an Asian American Legal Scholarship: Critical Race Theory, Post-Structuralism, and Narrative Space, 81 Calif. L. Rev. 1241 (1993) (defining as dominant culture’s perception that minority group achieves a higher degree of socioeconomic success, allowing it to justify ignoring the unique discrimination faced by that group).

[43]. There is literature contesting the very idea of merit, particularly the association between merit and standardized tests. See Lani Gunier, Reframing the Affirmative Action Debate, 86 Ky. L.J. 505, 511–13 (1998) (providing examples of how the “SAT, the LSAT, and the GRE” are “coachable tests” and revealing that “[e]xamples like these show clearly that we’re using certain aptitude tests to credentialize a social oligarchy and we’re mistakenly calling it merit”); Harris & Narayan, Affirmative Action and the Myth of Preferential Treatment, supra note 1 (arguing that the “transition to greater inclusiveness has provoked, among other things, some rethinking about the traditional criteria of ‘merit’ for admission to and promotion within various American institutions, and some reexamination of assessment procedures once thought to be unquestionably ‘neutral’”); Daria Roithmayr, Deconstructing the Distinction Between Bias and Merit, 10 La Raza L.J. 363, 366–67 (1998) (observing that merit has been conceptualized as “ahistorical, objective measures of ability,” and arguing that this understanding of merit obscures that “what constitutes ability itself is subjective and constructed under particular historical circumstances by particular social groups;”). See generally Lani Guinier, The Tyranny of Meritocracy: Democratizing Higher Education in America (2015) (challenging the ways in which standardized testing stands in for merit); Phoebe A. Haddon & Deborah W. Post, Misuse and Abuse of the LSAT: Making the Case for Alternative Evaluative Efforts and a Redefinition of Merit, 80 St. John’s L. Rev. 41 (2006) (focusing specifically on the LSAT); Charles R. Lawrence III, Two Views of the River: A Critique of the Liberal Defense of Affirmative Action, 101 Colum. L. Rev. 928 (2001).

[44]. See Erika Frankenberg et al., The Civil Rights Project, Harvard Univ., A Multiracial Society With Segregated Schools: Are We Losing the Dream? (2003); see also John Kucsera & Gary Orfield, The Civil Rights Project, UCLA., New York State’s Extreme School Segregation: Inequality, Inaction and a Damaged Future (2014).

[45]. See Cheryl I. Harris, Whiteness as Property, 106 Harv. L. Rev. 1707, 1709 (1993); see also Ian Haney López, White by Law: The Legal Construction of Race (10th ed. 2006).

[46]. See Camille Gear Rich, Marginal Whiteness, 98 Calif. L. Rev. 1497 (2010).

[47]. See generally Frances E. Aboud, The Formation of In-Group Favoritism and Out-Group Prejudice in Young Children: Are They Distinct Attitudes?, 39 Developmental Psychol. 48 (2003); Marilynn B. Brewer, The Psychology of Prejudice: Ingroup Love or Outgroup Hate?, 55 J. Soc. Issues 429 (1999); Miles Hewstone et al., Intergroup Bias, 53 Ann. Rev. Psychol. 575 (2002).

[48]. See Marilynn B. Brewer & Donald T. Campbell, Ethnocentrism and Intergroup Attitudes: East African Evidence 85 (1976); Brewer, supra note 47, at 432; Marilynn B. Brewer, In-Group Bias in the Minimal Intergroup Situation: A Cognitive-Motivational Analysis, 86 Psychol. Bull. 307 (1979); Anthony G. Greenwald & Thomas F. Pettigrew, With Malice Toward None and Charity for Some: Ingroup Favoritism Enables Discrimination, 69 Am. Psychologist 669 (2014).

[49]. See Samuel Gaertner & Leonard Bickman, Effects of Race on the Elicitation of Helping Behavior: The Wrong Number Technique, 20 J. Personality & Soc. Psychol. 218 (1971); Donald A. Saucier et al., Differences in Helping Whites and Blacks: A Meta-Analysis, 9 Personality & Soc. Psychol. Rev. 2 (2005).

[50]. See Ian Ayres et al., To Insure Prejudice: Racial Disparities in Taxicab Tipping, 114 Yale L.J. 1613 (2005); Michael Lynn et al., Consumer Racial Discrimination in Tipping: A Replication and Extension, 38 J. Applied Soc. Psychol. 1045 (2008).

[51]. See Lauren A. Rivera, Pedigree: How Elite Students Get Elite Jobs (2015); Margery Austin Turner et al., Discrimination in Metropolitan Housing Markets: National Results from Phase I HDS (2002); Marc Bendick Jr. et al., Employment Discrimination in Upscale Restaurants: Evidence From Matched Pair Testing, 47 Soc. Sci. J. 802 (2010).

[52]. Harris & Narayan, supra note 1, at 25.

[53]. See Kimberly West-Faulcon, Forsaking Claims of Merit: The Advance of Race Blindness Entitlement in Fisher v. Texas, in 29 Nat’l Lawyers Guild, Civil Rights Litigation and Attorney Fees Annual Handbook 335 (Steven Saltzman & Cheryl I. Harris eds., 2013) (analyzing the percentage representation of African Americans in University of Texas’s 2008 admissions pool); Kimberly West-Faulcon, Obscuring “Asian Penalty” With Illusions of “Black Bonus”, 64 UCLA L. Rev. Disc. (forthcoming 2016). West-Faulcon’s work builds on an earlier effort by Goodwin Liu that demonstrated the diminished effect affirmative action has on white applicants. See Goodwin Liu, The Causation Fallacy: Bakke and the Basic Arithmetic of Selective Admissions, 100 Mich. L. Rev. 1045 (2002).

[54]. See generally Sherick Hughes et al, Causation Fallacy 2.0: Revisiting the Myth and Math of Affirmative Action, 30 Educ. Pol’y 63 (2016).

[55]. Id.

[56]. We should be clear to note that we do not present Figure 7 as an empirical claim but as heuristic device to demonstrate the minimal (potentially at most 1 percent) impact of affirmative action on whites. Note, for example, that the schematic does not reference Asian Americans who have a very robust presence in the admissions pools of elite colleges and universities. Indeed, some opponents of affirmative action argue that there is a disconnect between the qualification of Asian Americans and their percentage representation in admissions pools and the rate at which elite colleges and universities admit them. The claim, more specifically, is that affirmative action functions as a “penalty” for Asian Americans. See West-Faulcon, supra note 53 (discussing this issue). It is beyond the scope of our Article to engage this Asian penalty claim. We deploy Figure 6 to discuss the impact of affirmative action on whites.

[57]. See Carbado & Gulati, supra note 38.

[58]. See generally, e.g., Claude M. Steele, A Threat in the Air: How Stereotypes Shape Intellectual Identity and Performance, 52 Am. Psychologist 613 (1997); Claude M. Steele & Joshua Aronson, Stereotype Threat and the Intellectual Test Performance of African Americans, 69 J. Personality & Soc. Psychol. 797 (1995). For a recent review, see generally Steven J. Spencer et al., Stereotype Threat, 67 Ann. Rev. Psychol. 415 (2016).

[59]. For other discussions of how stereotype threat is implicated in debates about affirmative action, see generally Rachel D. Godsil, Why Race Matters in Physics Class, 64 UCLA L. Rev. Disc. 40 (2016); Jerry Kang & Mahzarin Banaji, Fear Measures: A Behavorial Realist Revision of Affirmative Action, 94 Cal. L. Rev 1063 (2006).

[60]. See, e.g., David M. Amodio & Patricia G. Devine, Stereotyping and Evaluation in Implicit Race Bias: Evidence for Independent Constructs and Unique Effects on Behavior, 91 J. Personality & Soc. Psychol. 654, 654–55 (2006) (demonstrating that people hold significant levels of implicit stereotypes of Blacks as unintelligent, athletic, and rhythmic); Lei Guo & Summer Harlow, User-Generated Racism: An Analysis of Stereotypes of African Americans, Latinos, and Asians in YouTube Videos, 25 Howard J. Comm. 291, 291–92 (2014) (showing that 61% of most-viewed YouTube videos referencing Blacks, Latinos, or Asians contained stereotypes about Blacks, and that 21% of these videos included stereotypes about Blacks as unintelligent or uneducated); Paul Verhaeghen et al., Prime and Prejudice: Co-Occurrence in the Culture as a Source of Automatic Stereotype Priming, 50 Brit. J. Soc. Psychol. 501, 506 (2011) (examining the strength of association between various identity groups and stereotypic traits and finding strong associations between “Black” and “ignorant” and “lazy,” among others).

[61]. See Jenessa R. Shapiro & Steven L. Neuberg, From Stereotype Threat to Stereotype Threats: Implications of a Multi-Threat Framework for Causes, Moderators, Mediators, Consequences, and Interventions, 11 Personality & Soc. Psychol. Rev. 107, 112–13 (2007).

[62]. Such stress can be observed even at the physiological level. See generally Jim Blascovich et al., African Americans and High Blood Pressure: The Role of Stereotype Threat, 12 Psychol. Sci. 225 (2001).

[63]. See, e.g., Sapna Cheryan et al., Ambient Belonging: How Stereotypical Cues Impact Gender Participation in Computer Science, 97 J. Personality & Soc. Psychol. 1045 (2009); Jennifer A. Mangels et al., Emotion Blocks the Path to Learning Under Stereotype Threat, 7 Soc. Cognitiive & Affective Neuroscience 230 (2012); Robert J. Rydell et al., Stereotype Threat Prevents Perceptual Learning, 107 Proc. Nat’l Acad. Sci. 14042 (2010); Valerie Jones Taylor & Gregory M. Walton, Stereotype Threat Undermines Academic Learning, 37 Personality & Soc. Psychol. Bull. 1055 (2011).

[64]. Sian L. Beilock et al., More on the Fragility of Performance: Choking Under Pressure in Mathematical Problem Solving, 133 J. Experimental Psychol. 584 (2004) (discussing how a high pressure to achieve on a math exam can lead individuals to “choke”—perform poorly compared to actual ability—because, under pressure, they anxiously over-monitor each component of their behavior and thoughts, which interferes with automated processes and dramatically increases working memory load, harming performance); Sian L. Beilock et al., On the Causal Mechanisms of Stereotype Threat: Can Skills That Don’t Rely Heavily on Working Memory Still Be Threatened?, 32 Personality & Soc. Psychol. Bull. 1059 (2006) (following their 2004 study, cited immediately above, Beilock and colleagues ran two experiments on the nature of performance impairment under stereotype threat, and found that stereotype threat resulted in the same increase in over-monitoring as choking under pressure); Sian L. Beilock et al., Stereotype Threat and Working Memory: Mechanisms, Alleviation, and Spillover, 136 J. Experimental Psychol. 256 (2007) (elaborating on the mechanism outlined in the two previously cited studies by showing that stereotype threat negatively affects performance on tasks relying on working memory resources due to its taxation of cognitive resources); Jean-Claude Croizet et al., Stereotype Threat Undermines Intellectual Performance by Triggering a Disruptive Mental Load, 30 Personality & Soc. Psychol. Bull. 721 (2004) (finding that the activation of stereotypes about intellectual inferiority heightened cognitive workload, as measured by heart rate variability, and disrupted performance on a test); Michael Johns et al., Stereotype Threat and Executive Resource Depletion: Examining the Influence of Emotion Regulation, 137 J. Experimental Psychol. 691 (2008) (expanding on the previously cited research by finding that the increased cognitive load was specifically driven by attempts to regulate emotions, such as anxiety, depleting executive resources, and suppressing performance); Toni Schmader & Michael Johns, Converging Evidence That Stereotype Threat Reduces Working Memory Capacity, 85 J. Personality & Soc. Psychol. 440 (2003) (conducting one of the earliest studies to show that the negative effect of stereotype threat on exam performance was mediated by increased cognitive load that reduced students’ working memory capacity).

[65]. For meta-analyses on this subject, see Hannah-Hanh D. Nguyen & Ann Marie Ryan, Does Stereotype Threat Affect Test Performance of Minorities and Women? A Meta-Analysis of Experimental Evidence, 93 J. Applied Psychol. 1314 (2008); Gregory M. Walton & Steven J. Spencer, Latent Ability Grades and Test Scores Systematically Underestimate the Intellectual Ability of Negatively Stereotyped Students, 20 Psychol. Sci. 1132 (2009).

[66]. See generally, for example, the aforementioned meta-analysis by Nguyen & Ryan, supra note 65, which identified 151 empirical reports on the effects of stereotype threat.

[67]. Claude M. Steele & Joshua Aronson, Stereotype Threat and the Intellectual Test Performance of African Americans, 69 J. Personality & Soc. Psychol. 797 (1995).

[68]. Sean F. Reardon et al., Effects of the California High School Exit Exam on Student Persistence, Achievement, and Graduation (Stanford Univ. Inst. for Research on Educ. Policy & Practice, Working Paper No. 2009-12. 2009). http://web.stanford.edu/
group/cepa/workingpapers/WORKING_PAPER_2009_12.pdf

[69]. See, e.g., Geoffrey L. Cohen et al., Recursive Processes in Self-Affirmation: Intervening to Close the Minority Achievement Gap, 324 Sci. 400 (2009); Geoffrey L. Cohen et al., Reducing the Racial Achievement Gap: A Social Psychological Intervention, 313 Sci. 1307 (2006).

[70]. See, e.g., Gregory M. Walton & Steven J. Spencer, Latent Ability: Grades and Test Scores Systematically Underestimate the Intellectual Ability of Negatively Stereotyped Students, 20 Psychol. Sci. 1132 (2009) (finding that stereotype threat accounted for a score reduction of just under one fifth of a standard deviation (approximately 0.18) for negatively stereotyped students. Given that one standard deviation of the past year’s SAT was 351 points across all three sections of the exam, losing 0.18 of a standard deviation to stereotype threat would be the equivalent of a sixty-three point loss).

[71]. This is almost certainly a conservative estimate because this research took place in the students’ classrooms, an uncontrolled field setting where no experimental exercise can eliminate stereotype threat or its effects completely.

[72]. Walton & Spencer, supra note 70.

[73]. See e.g., Marie Mazerolle et al., Stereotype Threat Strengthens Automatic Recall and Undermines Controlled Processes in Older Adults, 23 Psychol. Sci. 723 (2012).

[74]. See e.g., Beilock et al,, supra note 64, at 257; Michael Johns et al., Knowing Is Half the Battle: Teaching Stereotype Threat as a Means of Improving Women’s Math Performance, 16 Psychol. Sci. 175 (2005); Nguyen & Ryan, supra note 65; Toni Schmader, Gender Identification Moderates Stereotype Threat Effects on Women’s Math Performance, 38 J. Experimental Soc. Psychol. 194 (2002).

[75]. See Nai Chi Jonathan Yeung & Courtney von Hippel, Stereotype Threat Increases the Likelihood That Female Drivers in a Simulator Run Over Jaywalkers, 40 Accident Analysis & Prevention 667 (2008).

[76]. Jeff Stone et al., Stereotype Threat Effects on Black and White Athletic Performance, 77 J. Personality & Soc. Psychol. 1213, 1216–20 (1999).

[77]. Joshua Aronson et al., When White Men Can’t Do Math: Necessary and Sufficient Factors in Stereotype Threat, 35 J. Experimental Soc. Psychol. 29, 32–35 (1999).

[78]. Cheryl R. Kaiser & Nao Hagiwara, Gender Identification Moderates Social Identity Threat Effects on Working Memory, 35 Psychol. Women Q. 243 (2011).

[79]. See generally Cheryan et al., supra note 63; Michael Inzlicht & Talia Ben-Zeev, A Threatening Intellectual Environment: Why Females Are Susceptible to Experiencing Problem-Solving Deficits in the Presence of Males, 11 Psychol. Sci. 365 (2000); Charles G. Lord & Delia S. Saenz, Memory Deficits and Memory Surfeits: Differential Cognitive Consequences of Tokenism for Tokens and Observers, 49 J. Personality & Soc. Psychol. 918 (1985); Mary C. Murphy et al., Signaling Threat: How Situational Cues Affect Women in Math, Science, and Engineering Settings, 18 Psychol. Sci. 879 (2007); Valerie Purdie-Vaughns et al., Social Identity Contingencies: How Diversity Cues Signal Threat or Safety for African Americans in Mainstream Institutions, 94 J. Personality & Soc. Psychol. 615 (2008).

[80]. See Inzlicht & Ben-Zeev, supra note 79; Denise Sekaquaptewa & Mischa Thompson, Solo Status, Stereotype Threat, and Performance Expectancies: Their Effects on Women’s Performance, 39 J. Experimental Soc. Psychol. 68 (2003); Denise Sekaquaptewa & Mischa Thompson, The Differential Effects of Solo Status on Members of High-and Low-Status Groups, 28 Personality & Soc. Psychol. Bull. 694, 701–04 (2002), [hereinafter Sekaquaptewa & Thompson, 2002]; Mischa Thompson & Denise Sekaquaptewa, When Being Different Is Detrimental: Solo Status and the Performance of Women and Racial Minorities, 2 Analyses Soc. Issues & Pub. Pol’y 183 (2002).

[81]. See Markus Appel & Nicole Kronberger, Stereotypes and the Achievement Gap: Stereotype Threat Prior to Test Taking, 24 Educ. Psychol. Rev. 609, 624–25 (2012); Markus Appel et al., Stereotype Threat Impairs Ability Building: Effects on Test Preparation Among Women in Science and Technology, 41 Eur. J. Soc. Psychol. 904 (2011); Mangels et al., supra note 63; Rydell et al., supra note 63; Taylor & Walton, supra note 63.

[82]. Stereotype threat can have further downstream effects as well, leading to the avoidance of academically oriented careers or even dropping out of school altogether. See Claude M. Steele, Race and the Schooling of Black Americans, Atlantic (Apr. 1992), http://www.theatlantic.com/magazine/archive/1992/04/race-and-the-schooling-of-black-americans/306073/ [https://perma.cc/CM7D-3NUK].

[83]. Appel & Kronberger, supra note 81; Taylor & Walton, supra note 63.

[84]. See generally Daniel Katz & Kenneth W. Braly, Racial Prejudice and Racial Stereotypes, 30 J. Abnormal & Soc. Psychol. 175 (1935); Daniel Katz & Kenneth Braly, Racial Stereotypes of One Hundred College Students, 28 J. Abnormal & Soc. Psychol. 280 (1933).

[85]. Gordon W. Allport, Attitudes, in A Handbook of Social Psychology 798, 798–844 (Carl Murchison ed., 1935); Anthony G. Greenwald & Mahzarin R. Banaji, Implicit Social Cognition: Attitudes, Self-Esteem, and Stereotypes, 102 Psychol. Rev. 4 (1995).

[86]. Associated Press, AP Poll: U.S. Majority Have Prejudice Against Blacks, USA Today (Oct. 27, 2012, 8:37 AM), http://www.usatoday.com/story/news/politics/2012/10/
27/poll-black-prejudice-america/1662067/ [https://perma.cc/SGP3-J8VV].

[87]. See Michael Tesler & David O. Sears, Obama’s Race: The 2008 Election and the Dream of a Post-Racial America 18 (Benjamin I. Page et al. eds., 2010). For additional conceptualizations of this more recent form of racism, see also Donald R. Kinder & Lynn M. Sanders, Divided by Color: Racial Politics and Democratic Ideals (1996); Lawrence Bobo et al., Laissez-Faire Racism: The Crystallization of a Kinder, Gentler, Antiblack Ideology, in Racial Attitudes in the 1990s: Continuity and Change 15, 15–42 (Steven A. Tuch & Jack K. Martin eds., 1997); Samuel L. Gaertner & John F. Dovidio, The Aversive Form of Racism, in Prejudice, Discrimination, and Racism 61, 61–89 (John F. Dovidio & Samuel L. Gaertner eds., 1986); John B. McConahay, Modern Racism, Ambivalence, and the Modern Racism Scale, in Prejudice, Discrimination, and Racism 91, 91–125 (John F. Dovidio & Samuel L. Gaertner eds., 1986).

[88]. See Donald R. Kinder & Tali Mendelberg, Individualism Reconsidered: Principles and Prejudice in Contemporary American Opinion, in Racialized Politics: The Debate About Racism in America 44, 44–74 (David O. Sears et al. eds., 2000).

[89]. See P.J. Henry & David O. Sears, The Symbolic Racism 2000 Scale, 23 Pol. Psychol. 253, 254 (2002).

[90]. For example, an individual with symbolically racist beliefs might agree with statements such as, “It’s really a matter of some people not trying hard enough; if blacks would only try harder they could be just as well off as whites,” and disagree with statements such as, “Generations of slavery and discrimination have created conditions that make it difficult for blacks to work their way out of the lower class.” See Henry & Sears, supra note 89, at 279.

[91]. For example, arguments against forms of racial remediation employing symbolic racism may include “Welfare cheats could find work if they tried,” “Blacks should not be given a status they have not earned,” and “Whites have worked hard for their neighborhoods and for their neighborhood schools,” among others. See Donald R. Kinder & David O. Sears, Prejudice and Politics: Symbolic Racism Versus Racial Threats to the Good Life, 40 J. Personality & Soc. Psychol. 414, 416 (1981).

[92]. See Kinder & Mendelberg, supra note 88, at 58–59, 61, 73.

[93]. With respect to social policy support, several authors have found that symbolic racism is predictive of affirmative action attitudes. See generally Michael Hughes, Symbolic Racism, Old-Fashioned Racism, and Whites’ Opposition to Affirmative Action, in Racial Attitudes in the 1990s: Continuity & Change 45, 45–75 (Steven A. Tuch & Jack K. Martin eds., 1997); Cardell K. Jacobson, Resistance to Affirmative Action: Self-Interest or Racism?, 29 J. Conflict Resol. 306 (1985); David O. Sears et al., Is It Really Racism?: The Origins of White Americans’ Opposition to Race-Targeted Policies, 61 Pub. Opinion Q. 16 (1997). Researchers have also found that symbolic racism is predictive of behavior, specifically voting for Proposition 209, an anti-affirmative action ballot measure. See Jacqueline N. Sawires & M. Jean Peacock, Symbolic Racism and Voting Behavior on Proposition 209, 30 J. Applied Soc. Psychol. 2092 (2000). Rabinowitz and colleagues likewise found that symbolic racism predicted whites’ opposition to policies designed to benefit blacks, even controlling for factors such as conservative ideology and attitudes about the role of government; in contrast, it did not predict whites’ attitudes towards social programs over and above other political leanings when the policy did not specifically target blacks. See Joshua L. Rabinowitz et al., Why Do White Americans Oppose Race-Targeted Policies? Clarifying the Impact of Symbolic Racism, 30 Pol. Psychol. 805 (2009). Interestingly, symbolic racism also predicted support for crime policy, even in the absence of any explicit reference to race. Green and colleagues looked at the effect of symbolic racism on two types of crime policies: punitive policies, in which the goal was to discourage crime by making it too costly, and preventive policies, in which the goal was to prevent crime by changing the structural conditions believed to produce it, including poverty and unemployment. The researchers hypothesized that, because symbolic racism reflects the belief that blacks do not play by the “rules” of American society, endorsement of symbolic racism would be associated with higher support for a punitive response to crime (for example, a response that implies blacks have bad morals and must be coerced to behave using punishment) and lower support for a preventive response (which reflects an appreciation for the structural factors that contribute to crimes committed by blacks). The data confirmed these expectations. See, e.g., Eva G. T. Green et al., Symbolic Racism and Whites’ Attitudes Towards Punitive and Preventive Crime Policies, 30 L. & Hum. Behav. 435 (2006).

[94]. In examinations of vote preference, Donald Kinder and David Sears (1981) found that symbolic racism predicted voting for the white, over the black, mayoral candidate in two Los Angeles elections, a finding that was later replicated in an additional sample. See generally Kinder & Sears, supra note 91; McConahay, supra note 87. Jonathan Knuckey and Byron Orey found that symbolic racism predicted voting for a Louisiana gubernatorial candidate in a race against a black opponent, and Susan Howell found that it predicted support for David Duke, a former grand wizard of the Ku Klux Klan who ran for office in Louisiana as well. Jonathan Knuckey & Byron D’Andra Orey, Symbolic Racism in the 1995 Louisiana Gubernatorial Election, 81 Soc. Sci. Q. 1027 (2000); Susan E. Howell, Racism, Cynicism, Economics, and David Duke, 22 Am. Pol. Q. 190 (1994).

[95]. See Henry & Sears, supra note 89.

[96]. Id. at 267.

[97]. See id. at 278.

[98]. Tyrone A. Forman et al., Race, Place, and Discrimination, in Perspectives on Social Problems: Public Harassment 231 (Carol Brooks Gardner ed., 1997) (noting: (1) “[T]he nature and extent of discrimination among Blacks and Whites remains unclear”; (2) “[p]rior research provides few estimates of whites’ experience of discrimination; (3) “there is considerable amount of empirical evidence on African Americans’ experience of discrimination” and (4) the discussion of the three categories does not discuss any contrast between the self-reporting of blacks or whites).

[99]. See generally Ronald C. Kessler et al., The Prevalence, Distribution, and Mental Health Correlates of Perceived Discrimination in the United States, 40 J. Health & Soc. Behav. 208 (1999).

[100]. Quaylan Allen, “They Think Minority Means Lesser Than”: Black Middle-Class Sons and Fathers Resisting Microaggressions in the School, 48 Urban Educ, 171 (2013).

[101]. Id. at 186–87.

[102]. Id. at 181.

[103]. See generally, e.g., Anthony D. Ong et al., Racial Discrimination and the Stress Process, 96 J. Personality & Soc. Psychol. 1259 (2009).

[104]. Keeanga-Yamahtta Taylor, Black Lives Matter on Campus Too, Al Jazeera Am. (Nov. 29, 2015), http://america.aljazeera.com/opinions/2015/11/black-lives-matter-on-campus-too.html.

[105]. Matt Ferner & Nick Wing, It Shouldn’t Be Hard to Accept That Racism Is a Problem at Mizzou, Huffington Post (Nov. 13, 2015, 11:30 AM), http://www.huffington
post.com/entry/mizzou-racist-incidents_us_5644f33ce4b045bf3dee583d.

[106]. Racism on Campus: Stories From New York Times Readers, N.Y. Times (Nov. 17, 2015), http://www.nytimes.com/2015/11/18/us/racism-on-campus-stories-from-new-york-times-readers.html [https://perma.cc/2358-RKYE].

[107]. In Subpart 4, we revisit explicit racial bias in the context of highlighting how black students also contend with negative institutional culture.

[108]. See Ivan E. Bodensteiner, The Implications of Psychological Research Related to Unconscious Discrimination and Implicit Bias in Proving Intentional Discrimination, 73 Mo. L. Rev. 83, 102 (2008).

[109]. See Icek Ajzen & Nicole Gilbert Cote, Attitudes and the Prediction of Behavior, in Attitudes and Attitude Change 289, 294 (William D. Crano & Radmila Prislin eds., 2008).

[110]. William A. Cunningham et al., Implicit Attitude Measures: Consistency, Stability, and Convergent Validity, 12 Psychol. Sci. 163 (2001).

[111]. See Brian A. Nosek et al., Understanding and Using the Implicit Association Test: II. Method Variables and Construct Validity, 31 Personality & Soc. Psychol. Bull. 166 (2005); see also N. Sriram & Anthony G. Greenwald, The Brief Implicit Association Test, 56 Experimental Psychol. 283 (2009) (providing information on the brief Implicit Association Test (IAT), which is a newer and commonly used form of the standard IAT).

[112]. See generally Irene V. Blair et al., Implicit Attitudes, in 1 APA Handbook of Personality and Social Psychology: Attitudes and Social Cognition 665 (Mario Mikulincer et al. eds., 2015).

[113]. Andrew Scott Baron & Mahzarin R. Banaji, The Development of Implicit Attitudes Evidence of Race Evaluations From Ages 6 and 10 and Adulthood, 17 Psychol. Sci. 53 (2006).

[114]. Godsil, supra note 59, at 58.

[115]. See generally Linda van den Bergh et al., The Implicit Prejudiced Attitudes of Teachers Relations to Teacher Expectations and the Ethnic Achievement Gap, 47 Am. Educ. Res. J. 497(2010).

[116]. Id., at 514–15, 518–19.

[117]. Id.

[118]. See Sian L. Beilock et al., Female Teachers’ Math Anxiety Affects Girls’ Math Achievement, 107 Proc. Nat’l Acad. Sci. 1860 (2010).

[119]. Arin N. Reeves, Nextions Yellow Paper Series, Written in Black & White: Exploring Confirmation Bias in Racialized Perceptions of Writing Skills (2014).

[120]. See generally Frances Trix & Carolyn Psenka, Exploring the Color of Glass: Letters of Recommendation for Female and Male Medical Faculty, 14 Discourse & Soc’y 191 (2003).

[121]. Id. at 214–15.

[122]. See generally Rhea E. Steinpreis et al., The Impact of Gender on the Review of the Curricula Vitaeof Job Applicants and Tenure Candidates: A National Empirical Study, 41 Sex Roles 509 (1999).

[123]. See generally Kimberle Crenshaw, Demarginalizing the Intersection of Race and Sex: A Black Feminist Critique of Antidiscrimination Doctrine, Feminist Theory and Antiracist Politics, 1989 U. Chi. Legal F. 139 (asserting that black women’s experiences are too often marginalized in discussions about gender inequality and arguing for an intersectional approach to antidiscrimination law and politics).

[124]. Marianne Bertrand & Sendhil Mullainathan, Are Emily and Greg More Employable Than Lakisha and Jamal? A Field Experiment on Labor Market Discrimination (Nat’l Bureau of Econ. Research, Working Paper No. 9873, 2003), http://www.nber.org/
papers/w9873.

[125]. Id. at 10.

[126]. Id.

[127]. Ian F. Haney López, Institutional Racism: Judicial Conduct and a New Theory of Racial Discrimination, 109 Yale L.J. 1717, 1794 (2000).

[128]. Susan R. Rankin & Robert Dean Reason, Differing Perceptions: How Students of Color and White Students Perceive Campus Climate for Underrepresented Groups, 46 J. C. Student Dev. 43, 52–55 (2005); Thomas F. Pettigrew & Joanne Martin, Shaping the Organizational Context for Black American Inclusion, 43 J. Soc. Issues 41 (1987); Linda Serra Hagedorn et al., An Investigation of Critical Mass: The Role of Latino Representation in the Success of Urban Community College Students, 48 Res. Higher Educ. 73 (2006); Valerie Purdie-Vaughns et al., Social Identity Contingencies: How Diversity Cues Signal Threat or Safety for African Americans in Mainstream Institutions, 94 J. Personality & Soc. Psychol. 615 (2008).

[129]. Purdie-Vaughns et al., supra note 128, at 615–630 (2008).

[130]. See generally Carbado & Gulati, supra note 38.

[131]. See Elise C. Boddie, The Indignities of Color Blindness, 64 UCLA L. Rev. Disc. 64 (2016).

[132]. Id. at 67 (discussing “racial palatability” and “racial salience”).

[133]. This problem has engendered debates about race and cannons on knowledge and whether institutions should have diversity requirements. See Thomas F. Nelson Laird & Mark E. Engberg, Establishing Differences Between Diversity Requirements and Other Courses With Varying Degrees of Diversity Inclusivity, Penn. St. Educ. Equity Workshop (2011), http://equity.psu.edu/workshop/assets/pdf/fall11/divreq.pdf [https://perma.cc/QT24-GFRD].

[134]. See, e.g., Madeleine Pauker, UC Berkeley’s Persistent Lack of Faculty Diversity Prompts Efforts to Address Issue, Daily Californian (July 12, 2015), http://www.dailycal.org/2015/07/12/ uc-berkeleys-persistent-lack-of-faculty-diversity-prompts-efforts-to-address-issue/ [https:// perma.cc/9744-D4TU]; Jan Ransom, Brandeis Students Occupy Building to Protest Lack of Racial Diversity, Boston Globe (Nov. 23, 2015), https://www.bostonglobe.com/metro/ 2015/11/22/brandeis-students-occupy-building-protest-lack-racial-diversity/I1yUrfTbEpF dyW2ilajvJM/story.html [https://perma.cc/GZN7-X7NK].

[135]. See, e.g., Jess Bidgood, Amherst College Drops ‘Lord Jeff’ as Mascot, New York Times (Jan. 26, 2016), http://www.nytimes.com/2016/01/27/us/amherst-college-drops-lord-jeff-as-mascot.html [https://perma.cc/3TSE-GPJE]; Blake Neff, Princeton Students Take Over President’s Office, Demand Erasure of Woodrow Wilson, Daily Caller (Nov. 18, 2015), http://dailycaller.com/2015/11/18/princeton-students-take-over-presidents-office-demand-erasure-of-woodrow-wilson/ [https://perma.cc/5PRH-QJQT]; Tanaz Ahmed, After Charleston Shooting, Yale Students Petition to Rename Calhoun College, USA TODAY College (July 9, 2015), http://college.usatoday.com/2015/07/09/after-charleston-shooting-yale-students-petition-to-rename-calhoun-college/ [https://perma.cc/SK34-CZ9Q].

[136]. See, e.g., andré douglas pond cummings & Seth E. Harper, Wide Right: Why the NCAA’s Policy on the American Indian Mascot Issue Misses the Mark, 9 U. Md. L.J. Race, Religion, Gender & Class 135 (2009).

[137]. Stephanie A. Fryberg et al., Of Warrior Chiefs and Indian Princesses: The Psychological Consequences of American Indian Mascots, 30 Basic & Applied Soc. Psychol. 208 (2008).

[138]. See generally Team Spirits: The Native American Mascots Controversy (C. Richard King & Charles Fruehling Springwood eds., 2001).

[139]. See, e.g., Kira Brekke, Black Professor Speaks Out About Being Racially Profiled Near Campus, Huffington post: Black Voices (Dec. 10, 2015, 5:22 PM), http://www.huffington post.com/entry/steve-locke-racially-profiled_us_56688025e4b009377b236c54 [http://perma.cc/ 32T9-YJHH]; Scott Jaschick, Yale Police Aim Gun at Columnist’s Son, Turning Spotlight on Racial Profiling on Campus, PBS Newshour (Jan. 26, 2015, 1:22 PM), http://www.pbs. org/newshour/rundown/yale-police-point-gun-columnists-son-bring-spotlight-back-racial-profiling-campus/ [https://perma.cc/37G3-QDFN]; Shereen Marisol Meraji, USC Students Allege Racial Profiling by LAPD, NPR Code Switch (May 8, 2013, 11:53 AM), http://www.npr.org/sections/codeswitch/2013/05/09/182175917/L-A-s-Police-Department-Faces-Allegations-Of-Racism [https://perma.cc/G3KL-8ZXW].

[140]. See Sekaquaptewa & Thompson, 2002, supra note 80.

[141]. See Cheryan et al., supra note 63.

[142]. Phia Shante Salter, Representations of Black History as Intentional Worlds of Oppression and Liberation 15 (Sept. 23, 2010) (unpublished Ph.D. dissertation, University of Kansas), https://kuscholarworks.ku.edu/handle/1808/7721 [https://perma.cc/J22M-JCKS].

[143]. Id.

[144]. John D. Coie et al., Peer Group Behavior and Social Status,in Peer Rejection in Childhood, 17 (1990), https://books-google-com.ezproxy.cul.columbia.edu/books/about/ Peer_Rejection_in_Childhood.html?id=OC84AAAAIAAJ.

[145]. Melissa Faye Jackson et al., Classroom Contextual Effects of Race on Children’s Peer Nominations, 77 Child Dev. 1325 (2006)

[146]. Id.

[147]. Barbara Reskin, The Race Discrimination System, 38 Ann. Rev. Sociology 17 (2012).

[148]. This is consistent with the literature on white privilege, which frames white advantage as the collection or privileges white people have precisely because they are not black. The classic articulation is by Peggy Macintosh. See generally Peggy Macintosh, White Privilege and Male Privilege: A Personal Account of Coming to See Correspondences Through Work in Women’s Studies (Wellesley Coll. Ctr. for Research on Women, Working Paper No. 189, 1988).

[149]. See generally Laura Sullivan et al., The Racial Wealth Gap: Why Policy Matters (2015); see also Melvin L. Oliver & Thomas M. Shapiro, Black Wealth/White Wealth: A New Perspective on Racial Inequality (2d ed. 2006).

[150]. Sullivan et al., supra note 149, at 6.

[151]. Id. at 3.

[152]. Id. at 20.

[153]. See generally Dalton Conley, Being Black, Living in the Red: Race, Wealth, and Social Policy in America (2010); Michael Fletcher, A Shattered Foundation, Wash. Post (Jan. 24, 2015), http://www.washingtonpost.com/sf/investigative/2015/ 01/24/the-american-dream-shatters-in-prince-georges-county/ [https://perma.cc/25WR-ZY3B] (exploring African American families in Prince George’s County who built up a middle-class community, but without the economic security of their neighbors in white suburbs, are losing their wealth); Tami Luhby, Housing Crisis Hits Blacks Hardest, CNN (Oct. 19, 2010, 8:15 AM), http://www.cnn.com/2010/LIVING/10/19/inam.housing.fore closure.money/ [https://perma.cc/9WN3-KMCX] (arguing that the black middle class was built through homeownership, and that the foreclosure crisis decimated the rate of black homeownership).

[154]. Richard D. Kahlenberg, Texas’ College Admissions Policies Give the Well-To-Do a Leg Up, L.A. Times (Dec. 8, 2015, 5:00 AM), http://www.latimes.com/opinion/op-ed/le-oe-1208-kahlenberg-college-affirmative-action-supreme-court-20151208-story.html [https://perma.cc/ KS65-XKKS].

[155]. Cheryl I. Harris, Fisher’s Foibles: From Race and Class to Class Not Race, 64 UCLA L. Rev. Disc. (forthcoming 2016) (manuscript at 8–9) (on file with author).

[156]. See Eugene Robinson, A Question of Race vs. Class, Wash. Post (May 15, 2007), http://www.washingtonpost.com/wp-dyn/content/article/2007/05/14/AR2007051401233.html [https://perma.cc/CUN5-VH3D].