Do white students enjoy an unfair advantage as compared to Asian Americans in admissions to certain universities? This Article explains the proper legal comparison under settled civil rights law for making this determination based on the number of white and Asian American applicants and admits for a given admissions cycle. This Article also raises questions regarding the accuracy of blaming affirmative action favoring African Americans—a Black bonus—for racial discrimination against Asian Americans—an Asian penalty. It does so by examining charges of Asian penalty made by amici in Fisher v. University of Texas at Austin and the similar charges made in the complaints in lawsuits challenging racial affirmative action by Harvard University and the University of North Carolina at Chapel Hill. The Article exposes two mathematical fallacies as well as the limitations of an oft-cited 2009 book that have been relied upon regularly to bolster the notion that admissions bonuses for African Americans and Latinos cause universities to operate racially discriminatory anti-Asian American admissions ceilings. The Article illuminates how efforts to link affirmative action and Asian penalty omit a key inquiry for identifying anti-Asian American discrimination—disparate impact analysis comparing Asian American selection rates to white selection rates.
I condemn the voices from my own community that are translating legitimate anger at ceilings on Asian admissions into unthinking opposition to affirmative-action floors needed to fight racism.
Determining whether white students enjoy an unfair advantage as compared to Asian Americans in admissions to a particular university is fairly straightforward. As a threshold matter, all that is required is a comparison of the white and Asian American rates of admission to the university, easily calculated if the number of white and Asian American applicants and admits is known for a given admissions cycle. Settled civil rights law—Title VI disparate impact law2—even offers a very basic mathematical rule for distinguishing between small racial variations in selection rates that suggest no racial unfairness and rate disparities between races that are sufficiently large to signal potentially illegal unfairness to the group selected at the lower rate.3 Yet, significantly, lawsuits purportedly focused on challenging racial discrimination against Asian Americans by Harvard University and the University of North Carolina at Chapel Hill (UNC)4 exclude any such comparison to instead blame affirmative action favoring African Americans—a Black bonus—for racial discrimination against Asian Americans—an Asian penalty. This Article examines charges of Asian penalty made by amici in Fisher v. University of Texas at Austin (Fisher I5 and Fisher II6) and the similar charges made in the Harvard and UNC cases.7 It illuminates how both sets of claims side step the more important comparison of white and Asian American selection rates, why doing so omits a key inquiry for identifying anti-Asian American discrimination, and exposes two mathematical fallacies underlying these claims as well as the limitations of an oft-cited 2009 book that has been relied upon regularly to bolster the incorrect notion that admissions bonuses for African Americans and Latinos cause universities to operate racially discriminatory anti-Asian American admissions ceilings.
Because whites most often make up the vast majority of applicants to elite universities, mathematical reality dictates that Asian American applicants denied admission to such institutions are almost always edged out by white applicants, not by an African American or Latino applicant who benefited from racial affirmative action.8 Nevertheless, critics of racial affirmative action, including a U.S. Supreme Court Justice,9 have promoted the fallacy that affirmative action for African American students causes universities to racially discriminate against Asian Americans,10 often invoking the specter of an “Asian penalty” and referencing a “boost” or “bonus” to African Americans in college admissions.11 Many commentators who contend ending affirmative action will end anti-Asian American discrimination in admissions fail to consider differences in the rates at which different racial groups are selected for admission yet rely commonly on two mathematical fallacies—the causation fallacy and the average-test-score-of-admitted-students fallacy—as well as one particular table12 in No Longer Separate, Not Yet Equal, a book published in 2009 by Thomas Espenshade and Alexandria Radford.13 Because the legal assault on racial affirmative action is increasingly framed as a path to ending anti-Asian American discrimination,14 I use this Article to explicate why the conflation of racial affirmative action with discriminatory ceilings on Asian American admissions is erroneous and undercuts proper legal analysis of the civil rights claims of rejected Asian American college applicants.
Part I of this Article explains in detail that the basic math of college admissions definitively refutes the purported causal link between Black bonus and Asian penalty—the causation fallacy. It presents old and new research findings on the causation fallacy to explicate why a university’s racial affirmative action policy, irrespective of its robustness, affects too few admission slots to be the mechanism by which a university imposes a racially discriminatory ceiling on Asian American admissions.
Part II debunks the average-test-score-of-admitted-students fallacy by making it clear that finding a mean (average) SAT15 score of all admitted African American students that is lower than that of all admitted Asian American students does not prove a particular university applied a different test score standard to the two racial groups. Part II does this by presenting research and offering examples explaining that it is not racial affirmative action that causes the average test score of all students admitted from one racial group to be lower than another.16
Part III presents details of the admissions policies of the University of Texas at Austin (UT Austin) deemed constitutional in Fisher II17 to interrogate whether those policies were in some way facially racially discriminatory against Asian American applicants or had a racially discriminatory effect on Asian Americans. Specifically, it compares the Asian American rate of admission for the race-conscious portion of the admission cycle challenged by Abigail Fisher in Fisher I and II to the African American, Latino, and white admissions rates, using mathematical and statistical tests for identifying racially disparate impact. This Part contradicts the claim, made by several Fisher I and II amici and Justice Alito in his Fisher II dissent, that the affirmative action policy of UT Austin negatively affected the admission of Asian Americans.18 Part III’s analysis reveals that: (1) no facial component of the Fisher admissions policy penalized Asian Americans on the basis of race; (2) white and Asian American applicants were admitted at over twice the rate of African American and Latino applicants under the race-conscious Fisher admissions process; and (3) the disparity between the higher white UT Austin admission rate and the Asian American admission rate was quite small, too small to suggest a racially discriminatory effect on Asian American applicants.
Part IV examines the complaints filed in the two federal lawsuits alleging that the 2014 admissions policies of Harvard and UNC violated the Equal Protection Clause and Title VI of the Civil Rights Act of 1964.19 It observes that, in lieu of allegations that Asian Americans were admitted at lower rates than white applicants, the complaints in the Harvard and UNC cases rely on a litany of anecdotes, historical accounts, and also invoke the causation fallacy and the average-test-score-of-admitted-students fallacy to suggest that a Black bonus in Harvard and UNC admissions somehow causes an Asian American penalty. Part IV also explains that the complaints filed against Harvard and UNC lack analysis of racially discriminatory effect—disparate impact analysis—and fail to even consider whether Harvard and UNC unjustifiably admitted Asian Americans at a lower rate due to a white selection advantage.
Lastly, Part V demonstrates that “using the treatment of Whites as a basis for comparison”20 is the proper mechanism for evaluating claims that a university operates a racially discriminatory anti-Asian American admissions ceiling if white students applied to and were admitted at a higher rate than Asian American applicants. This Part uses disparate impact analysis to compare Asian American and white admission rates from a dataset created by the authors of the Espenshade & Radford book).21 It finds that the dataset fits a white selection advantage fact pattern—a disparity between the white and Asian American selection rates large enough to constitute Title VI disparate impact. Part V thus supports the Article’s overarching conclusion that one of the most important inquiries in examining a potential civil rights claim by rejected Asian American college applicants is comparing Asian American and white rates of selection by race, not comparing percentages of admitted students, percentages of enrolled students, or the average SAT scores of all admitted students by racial group.
I. Causation Fallacy of Linking Affirmative Action and Anti-Asian American Ceilings
Opponents of racial affirmative action often introduce the causation fallacy22 to raise the specter that racial affirmative action confers unfair advantages on African American and Latino applicants—a Black bonus—that causes more qualified Asian American and white applicants to be disadvantaged in selective university admissions—an Asian penalty.23 The causation fallacy ignores the fact that white students are typically the overwhelming majority of applicants in the applicant pools to selective universities and the fact that African Americans and Latinos usually make up much smaller percentages of all applicants. In short, the causation fallacy obscures the reality that an Asian American admissions ceiling typically inures to the benefit of white applicants, not African Americans or Latinos.
Professor Mari Matsuda is among the legal scholars who have cautioned against confusing the effect of racial affirmative action policies with admissions policies that purposefully or effectively operate as limits on Asian American admissions.24 Although Matsuda does not use the term causation fallacy explicitly, her observations caution against falling prey to the fallacy. Matsuda explains this as the difference between ceilings, above which Asian American enrollment never rises, and floors, below which African American and Latino admissions do not fall.25 To make a similar distinction, Professor Jerry Kang introduces a term that distinguishes a policy having the purpose or effect of excluding Asian American applicants from affirmative action—“negative action.”26 Both scholars are clear that affirmative action—a policy designed with the purpose of including members of certain nonwhite racial groups, often African Americans and Latinos but also sometimes Asian Americans27—differs from racial discrimination against Asian American applicants on the basis of race. The work of Matsuda and Kang provides the conceptual framework for understanding why efforts to end racial affirmative action are unlikely to expose or end anti-Asian purposeful bias, anti-Asian effect discrimination, or policies and practices that place discriminatory ceilings on Asian American admissions.
Writing together, scholars Gabriel Chin, Sumi Cho, Jerry Kang, and Frank Wu add a third term—“neutral action”—to the lexicon distinguishing between affirmative action and negative action. Under neutral action, whites are not selected at significantly higher rates than Asian Americans.28 Chin, Cho, Kang, and Wu also articulate the view that neutral action for Asian Americans cannot be achieved simply by joining efforts to end pro-African American and pro-Latino racial affirmative action because it requires ending unfair white selection advantage.29 To make this point, they write:
What [Asian Pacific Americans] APAs must understand is that negative action against us does not result from affirmative action for other minorities. In fact, in cases of proven racial disparities between APA and White admission rates, the causes have been either stereotypical treatment of APA applicants or other preferences, such as that for alumni children, who tend to be predominantly White. Furthermore, eliminating affirmative action does not eliminate negative action. Regardless of whether a prestigious university practices affirmative action for other racial minorities, it may still enact informal measures to limit the number of APAs on campus.30
The point of this observation from Chin, Cho, Kang, and Wu is that ending affirmative action for African American and Latino applicants will not end white advantage. If affirmative action is eliminated, negative action could still advantage white applicants over Asian Americans in competing for the bulk of admissions slots, as well as in competition for the smaller number of previously affirmative action admission slots. This makes ending such affirmative action an inapt way to battle discrimination against Asian Americans.
Other legal scholars have analyzed and explained in explicit mathematical terms that racial affirmative action has a minuscule effect on white and Asian American college admissions. The seminal example of research on this subject was authored by now-California Supreme Court Justice Goodwin Liu. As a law professor, Justice Liu use the term “causation fallacy” to describe the false assertion that racial affirmative action was the reason that white applicants Jennifer Gratz and Allan Bakke were rejected by the University of Michigan and University of Davis Medical School, respectively.31 In The Causation Fallacy: Bakke and the Basic Arithmetic of Selective Admissions (Causation Fallacy), Justice Liu uses arithmetic to explain why it was erroneous for plaintiff Gratz to presume her rejection could be blamed on Michigan’s affirmative action policy.32 Justice Liu does this by calculating and comparing Gratz’s chances of admission with and without the existence of Michigan’s racial affirmative action policy.
Liu’s calculations reveal that Gratz would have been denied admission, irrespective of the University of Michigan’s racial affirmative action policy, because of the stiff competition she faced from 424 applicants “with grades and test scores comparable to hers.”33 According to his analysis, the University of Michigan’s low admission rate for students with Gratz’s level of admissions credentials was the cause of her rejection, not the fact that a racial affirmative action policy improved the chances of admission for some African American and Latino students. As Justice Liu notes, this is because there were too few African Americans in the applicant pool for African Americans’ improved admissions chances to have a large effect on Gratz’s admissions chances.34
Over a decade after Liu published his causation fallacy article, a group of UNC professors revisited Liu’s mathematical analysis. Published in 2016, the UNC professors’ update of Justice Liu’s 2002 article is titled Causation Fallacy 2.0: Revisiting the Myth and Math of Affirmative Action (Causation Fallacy 2.0).35 The Causation Fallacy 2.0 study demonstrates how little effect racial affirmative action has on white and Asian American admissions rates. It also demonstrates that all combined—non-affirmative action as well as affirmative action—admissions of African American and Latino students make up such a small proportion of admissions at many selective universities that the elimination of all African American and Latino applicants “from the respective admissions pools” would have little impact on the admissions odds of white and Asian American applicants to those schools.36
Specifically, the Causation Fallacy 2.0 study finds that the complete elimination of African American and Latino applicants from the Harvard admissions pool increases the admissions chances of white and Asian American applicants by only 1 percent.37 Similarly, under the same no-African-Americans-and-no-Latinos hypothetical at UNC, white and Asian American admissions were found to increase by only around 4 percent.38 The relatively small increases in overall rates of admission to Harvard and UNC if no African American nor Latino students were admitted at all shows the number of admissions slots impacted by affirmative action is typically too small to substantially decrease Asian American and white selection rates. It is also a first step in understanding that focusing on affirmative action is tangential to proving the existence of Asian penalty.39
II. The Average Test Score of Admitted Students Fallacy
The average-test-score-of-admitted-students fallacy is employed by critics of affirmative action who point to a difference in the group numerical average of the SAT test scores of all of the African American students admitted and Asian American (or white) students admitted as proof that a university is racially discriminatory in applying its SAT test score standard.40 Like the studies of the causation fallacy discussed in Part I, studies have debunked the average-test-score-of-admitted-students fallacy on two grounds. First, it is incorrect to view affirmative action as the cause of a numerical difference in the group average of the test scores of all admitted students from two different racial groups. Second, it is equally incorrect to treat such a difference as proof that a university has an explicit or implicit policy of applying a lower test score standard to the applicants belonging to the racial group with the lower group test score average.
Selective universities admit applicants based on a broad range of criteria unrelated to SAT test scores41 and, as a consequence of African Americans having, on average, lower SAT scores than Asian Americans, many African Americans admitted to selective universities who manage to edge out other applicants must do so based on non-SAT score attributes.42 As a mathematical matter, when the test score characteristics of the applicant pool vary substantially by race, a numerical difference in the group average SAT scores of admitted students will exist irrespective of whether a university considers race in admissions.43
Labor economists William Dickens and Thomas Kane have likewise explained why comparing the group average SAT scores of all members of a racial group admitted to a highly selective college or university does not prove that a university applies a racially differential test score standard nor does it prove that a university confers racial preferences on African Americans:44 When “the distribution [range] of standardized test scores” differs between any two groups, “the mean test score will be lower for people drawn from the second [lower test] distribution even though both must pass the same relatively high [test] standard.” 45 Thus, it is incorrect to infer Asian American applicants are required to meet a higher test standard even if the group average SAT score of all admitted Asian American students to a given university is higher than the SAT score of all African American admitted students. According to Dickens and Kane, this is both because “test scores (or what they represent) are only a small part of what is considered in most selection processes” and because “blacks’ test scores tend to be their weakest credential relative to whites.”46
A hypothetical illustrating the average-test-score-of-admitted-students fallacy is illustrative. Consider the existence of a numerical difference between the group average SAT math scores of all females and males admitted to a university’s engineering program. The fact that the admitted males’ group average SAT math score is higher than the admitted females’ group average SAT math score would not be proof that the engineering department’s admissions policy conferred gender preferences on female applicants or that the engineering department imposed a more stringent SAT math score requirement on male applicants. This is for two reasons. First, undergraduate engineering programs consider factors other than applicants’ math SAT scores in determining admission. Second, under this scenario, females, as a group, have, on average, lower SAT math scores than males, as a group—the distribution of SAT math scores differs between males and females.47
Therefore, if most female applicants to the university’s engineering program gain admission based on non-SAT math score qualification like a high grade in Advanced Placement Calculus, there would still be a numerical difference between the admitted females’ and males’ group average SAT math scores. If math SAT scores tend to be the weakest credential of female engineering school applicants relative to male engineering school applicants, a rejected male engineering school applicant who tried to rely on this numerical difference in the gender groups’ average SAT math scores as proof of discrimination against males would be incorrectly invoking the average-test-score-of-admitted-students fallacy. The gender difference in group average test scores would not be proof that the engineering program used a gender-differentiated SAT math score standard when it selected applicants.
Similarly, the two reasons for a numerical difference in the group averages of the SAT scores of selected members of two racial groups is that universities do not select students solely based on SAT scores and because SAT scores are racially skewed as a general matter. First, correctly recognizing that SAT scores are not perfectly precise in predicting college and future life success, universities also consider factors such as high school grades, honors and awards, depth and breadth of extracurricular activities, written essays, and personal recommendations in selecting applicants for admission.48 Second, for reasons that range from the theoretical and measurement limitations of g-based standardized tests like the SAT49 to the fact that African Americans belong to the racial group that has been most severely harmed by American Jim Crow racism, housing segregation, public educational opportunity gaps, and disparities in economic opportunities,50 the group average of the SAT scores of all African American testtakers is typically lower than the group average of the SAT scores of testtakers of other racial groups.51
Just as a group average gender SAT math test score difference for engineering majors would not prove a female gender preference in admissions, the existence of a difference between the group average SAT scores of all admitted African Americans students and those of other racial groups is not proof that African American admitted students benefit from preferential treatment. In fact, the Dickens and Kane mathematical modeling study demonstrates that the numerical difference in the average test score of admitted students is a statistical artifact that exists even if admissions are completely blind of race.52 Dickens and Kane show that only one selection policy would eliminate all of the racial differences in group average test scores by race—the psychometrically unsound53 policy of admitting an entire freshman class of individuals who all have the exact same SAT score.54 Thus, the reality is that the existence of a difference in the group average of the SAT scores of admitted African American students and that of all admitted white and Asian American students is not reliable evidence—tells us little to nothing about—the role that either race or SAT scores play in a particular university’s selection process. In fact, a difference in a university’s African American and Asian American admitted student group average SAT score does not prove that the institution operates a racial affirmative action policy or considers race in any way at all.55
III. No Black Bonus and No Asian Penalty in Fisher v. Texas
In this Part, I present my original analysis56 showing that white and Asian American applicants fared very and essentially equally well in the UT Austin Fall 2008 admission cycle compared to African American and Latino applicants.57 Although the named plaintiff in Fisher v. University of Texas at Austin (Fisher I58 and Fisher II59), Abigail Fisher, contended her rejection from UT Austin violated her equal protection rights,60 my analysis of UT Austin admissions the year Fisher applied reveals that white students were selected under the race-conscious portion of admissions—the non-top 10 percent admissions process—at a higher rate than all other racial groups. Once white students whose high school grades ranked within the top 10 percent of their graduating class were admitted automatically, the remaining white applicants (who ranked below the top 10 percent) were admitted to the remaining limited slots by using an admissions policy that permitted but did not require the consideration of race. Wholly inconsistent with Fisher’s claim that her rejection constituted discrimination against her on the basis of race, whites were selected at significantly higher rates than Latino and African American students and at a very similar rate to Asian Americans under the policy Fisher challenged.61 This Part’s finding that non-top 10 percent Asian American and white students were admitted at more than double the rate of Latino and African American applicants62 severely undermines the contention by critics of affirmative action, like Justice Alito in his Fisher II dissent, who contend UT Austin’s racial affirmative action led to racial discrimination against Asian Americans.63
A. No Facial Racial Discrimination Against Asian Americans
Fisher unsuccessfully contended that UT Austin’s use of race failed to meet the strict scrutiny equal protection standard.64 Yet, as noted by Justice Kennedy in the Fisher II majority opinion, race was a limited consideration even in the race-conscious component of UT Austin admissions.65 In addition to all top 10 percent applicants being selected solely based on their class rank, even some non-top 10 percent applicants were selected based solely on high SAT scores and class rank according to a calculated academic index (AI) score.66 Race was not a factor in assigning AI scores because that score was an SAT score-driven numerical prediction of freshman grade point average (GPA).67 Likewise, it is undisputed that race was not a determinative factor in the rest of the race-conscious, non-top 10 percent component of UT Austin admissions.68
In fact, the only potentially race-conscious aspect of UT Austin admissions was the vertical placement of applicants on an admissions decision grid on which academic achievement (measured by the AI score) was plotted on the horizontal axis and a measure of personal achievement (non-SAT and non-GPA achievement) was plotted on the vertical axis.69 The part of the UT Austin selection process conceived as “holistic review” (because it considered personal achievement factors beyond students’ SAT scores and class rank) involved admissions readers assigning applicants a minimally race-conscious personal achievement index (PAI) score.70 On the specific facial role the race of an Asian American applicant played in admissions, only an applicant’s PAI score could be impacted by race. And, even then, that PAI score could have only a small and formula-limited effect on how high an applicant was plotted vertically on the UT Austin admission grid.71 The Fisher amici claiming affirmative action harmed Asian Americans neglected to explain how the PAI score operated and also neglected to explain the built-in limits on the role race played in its calculation—that admissions officers were not required to consider any applicant’s race. In addition, the amici failed to mention that race was the seventh of seven subfactors of six distinct factors permitted to be considered to assign a score that was itself a subfactor of the PAI score.72
The consideration of race as one of seven factors contributing to a student’s “special circumstances” was itself one of six personal achievement factors.73 In addition to race, there were six other non-racial special-circumstances factors, most of which were explicitly socioeconomic-focused.74 More specifically, an applicant’s PAI score was computed using a formula comprising two essay scores and a personal achievement (PA) score: “PAI = [(personal achievement score * 4)+(mean essay * 3)] / 7.”75 The PA score was assigned on a scale of 1–6 based on evaluation of six personal achievement factors.76 These six factors were: (1) “[l]eadership,” (2) “[e]xtracurricular [a]ctivities,” (3) “[a]wards/honors,” (4) “[w]ork experience,” (5) “[s]ervice to school or community,” and (6) “[s]pecial circumstances.”77 The sixth personal achievement subfactor—“[s]pecial circumstances”—itself had seven subfactors. Those were: (1) the socioeconomic status of the student’s family; (2) whether the student lived in a single-parent home; (3) the language spoken at the student’s home; (4) the student’s family responsibilities; (5) the socioeconomic status of the student’s high school; 6) how the student’s SAT or ACT score compared to the average score for the student’s high school; and (7) the student’s race.78
Because race was only one of the seven possible abovementioned subfactors of the PA score, itself a subfactor of the PAI score, the race consciousness of UT Austin’s policy was a restrained “factor of a factor of a factor.”79 Further limiting the role that race played in UT Austin admissions, the consideration of race was neither mandatory in determining any applicant’s PAI score, nor limited to any one particular racial group or groups.80 UT Austin applicants of any race, including white and Asian American applicants, could receive positive consideration under the racial “special circumstances” subfactor.81 Accordingly, any applicant, including any Asian American applicant, could be awarded a PA score of six, the highest possible score, with or without race being considered.82
Neither the Fisher amici alleging an Asian penalty nor Justice Alito’s dissent in Fisher II reconcile the factual reality of UT Austin’s very limited consideration of race83 and their unsubstantiated claims of anti-Asian American race discrimination by UT Austin.84 The role of race in UT Austin’s admissions policy is even more limited when all admissions—top 10 percent as well as non-top 10 percent admissions—are considered. In short, amici assertions and Justice Alito’s declaration in dissent that UT Austin used the facial consideration of race challenged by Abigail Fisher as the mechanism for racially discriminating against Asian Americans85 is empirically implausible in light of the fact that race was such a limited factor of a factor of a factor. The Fisher amicus briefs allege UT Austin’s use of racial affirmative action harmed Asian American applicants but rely almost exclusively on studies and analysis of other colleges and universities—they do not point to any UT Austin-specific admissions study of a policy or practice that facially or effectively racially discriminates against Asian Americans. Instead, the briefs recount prior historical instances of racial discrimination against Asian Americans by other institutions (not UT Austin), rely on the causation fallacy and average-test-score-of-admitted-students fallacy,86 and quote from the third chapter of the Espenshade & Radford book’s analysis of the 1997 admissions outcomes of other (non-UT Austin) universities.87 Given the Fisher amicus briefs lack support for their contention that a causal link exists between racial affirmative action employed by UT Austin and an Asian penalty,88 it is appropriate to compare the Asian American selection rate at UT Austin to the selection rate of other racial groups to determine whether the admissions policy at issue in the Fisher litigation imposed an Asian penalty.
B. Racial Comparison of Fisher Admissions Outcomes
This Subpart presents the type of analysis lacking from the Asian penalty claims discussed above—specific details regarding the facial role race played in the admissions policy in question, the overall numbers of applicants who applied and were admitted, disaggregated by race, and, finally and importantly, the legally recognized types of mathematical and statistical comparisons of racial group selection rates long utilized to detect disparate impact racial discrimination. Figure 1 below sets forth the outcomes of the Fall 2008 UT Austin admissions cycle that are crucial to evaluating the claim that UT Austin discriminated against Asian American applicants.
Figure 1 shows that, as required by the Texas Ten Percent Law,89 every Texas high school student ranking in the top 10 percent of his or her high school graduating class, based on GPA, was admitted, including all top 10 percent Asian American applicants.90 It also shows that, for each racial group, the number of top 10 percent applicants was exactly the same as the number of top 10 percent admits because applicants falling into this category were automatically admitted.91
Whereas all racial groups were admitted at the rate of 100 percent under the top 10 percent portion of UT Austin admissions, all racial groups were admitted at different rates under the race-conscious non-top 10 percent admissions, but not in a way that suggests an Asian penalty. Notably, whites fared better than all other races in the race-conscious portion of UT Austin’s selection process—the very process rejected white applicant Abigail Fisher unsuccessfully challenged as unconstitutional. The numerical reality of a white advantage in admissions is substantiated by the fact that the overall number of non-top 10 percent whites (2142) selected for admission was far greater than the number of non-top 10 percent admits for any other racial group. This number—2142 whites admitted under the race-conscious policy—is almost four times as many white students as Asian Americans students. Still, the number of non-top 10 percent Asian Americans admitted (565)—itself much smaller than the number of whites admitted—was far greater than the number of non-top 10 percent Black and Latino admits (146 and 403, respectively) combined.93
Irrespective of overall numbers of admits by race, knowing how Asian American rates of selection compared to the rates of selection of other racial groups is the analysis central to determining whether a particular selection policy has a racially discriminatory effect on Asian Americans. As a matter of basic arithmetic, the fact that substantially more white students applied to UT Austin than Asian American students means the Asian Americans who were rejected for admission were far more likely to have been displaced by a white applicant than an African American or Latino applicant. Comparing rates of admission allows a meaningful comparison of admissions outcomes when some racial groups constitute a numerical minority (typically true of African Americans and Latinos) and others constitute a numerical majority (typically true of white applicants).
C. Comparing Rates of Selection by Race
Whites were the overwhelming numerical majority and African Americans and Latinos were a numerical minority in the race-conscious non-top 10 percent UT Austin applicant pool. From that majority-white applicant pool, Figure 2 shows substantial disparities in rates of admission by race. The disparity it shows does not suggest a white or Asian penalty, it shows selection rates substantially favoring Asian Americans and whites.
For Fall 2008 UT Austin non-top 10 percent admissions, white and Asian American students were admitted at substantially higher rates than African American and Latino students. The selection rate for white applicants evaluated under the race-conscious non-top 10 percent policy was 22.3 percent—the highest admission rate of all racial groups.95 Non-top 10 percent Asian Americans were admitted at a similarly high rate of 21.7 percent,96 an only slightly lower (only 0.6 percent lower) rate than whites. By contrast, non-top 10 percent Latinos and African Americans were admitted at substantially lower rates than both whites and Asian Americans—10.4 percent for Latinos and 8.8 percent for African Americans.97 So, African American applicants—the racial group that opponents of affirmative action paint as displacing whites and Asian Americans due to “large racial preferences”98—were selected at the lowest rate of all racial groups during the admissions cycle at issue in the Fisher litigation. Both the white and Asian American admission rates were more than double the rates for Latino and African American students, statistically favoring white and Asian American applicants over African Americans and Latinos, not vice versa.99
My calculations and analysis in Table 1, below, confirm that the difference between the white and Asian American selection rates was neither statistically significant nor legally significant under Title VI disparate impact law.100 In Table 1, I apply the four-fifths (or 80 percent) rule to determine whether Asian Americans suffered “adverse impact” on the basis of race as defined by Title VI disparate impact regulations.101
Applying the four-fifths rule involves determining whether the ratio of the selection rate for the racial group alleging disparate impact—the group with the lower selection rate—is less than 80 percent of the selection rate for the racial group selected at the highest rate.102 If the ratio of the selection rate of Asian Americans compared to the rate of selecting whites is less than 80 percent, a rejected Asian American student would potentially have evidence of legally cognizable racial adverse impact, also called racially disparate impact, under established Title VI disparate impact law.103 If the ratio of the Asian American selection rate and the highest racial group selection rate—the white selection rate—is greater than 80 percent, however, Title VI disparate impact law considers the disparity too small to constitute legal evidence of disparate impact.104
Courts have also recognized the Pearson Chi Square statistical test (chi square test) as a method for assessing the likelihood that racial differences in selection rates happened at random or due to discrimination.105 If the likelihood is high that the difference in the admission rates of two racial groups resulted from chance, the difference is not deemed legally significant.106 If the odds are low, based on the chi square test—a finding suggesting the difference in two racial groups’ rates was not random, the difference is treated as legally significant racially discriminatory effect that must be justified.107 Table 1 reports both my original four-fifths and chi square analysis of the small disparity between UT Austin’s white and Asian American admission rates for the admissions cycle at issue in Fisher I and II.
Table 1. Lack of Adverse Impact Against Asian Americans in Admissions Policies Challenged in Fisher v. University of Texas108
|Fall 2008 UT Austin Non-Top 10% Selection Rates||White rate: 22.3%
Asian American rate: 21.7%
|Ratio of Higher and Lower Selection Rate||21.7
——— = .97
|Four-Fifths Rule Violated if Ratio
Less Than 0.80
|.97 > .80|
|Four-Fifths Rule Violated||No|
|What is p value for this Chi Square?||p=0.52|
|Is p Value Less Than 0.01?
(p value<0.01 is significant)
|No, 0.52 > 0.01
Disparity in rates is not
It is clear in Table 1 that the difference in UT Austin’s white and Asian American admissions rates are by no means large enough to be considered a violation of the four-fifths rule.109 The ratio of the selection rates for Asian Americans and whites is not less than 80 percent; it is significantly greater—0.97 (97 percent)—which means there is no racially disparate effect on Asian Americans under the four-fifths rule.110 This makes intuitive sense in light of the fact that the difference between the Asian American and white admission rates is less than one percentage point.
Table 1 also shows a chi square result of 0.41, with a p-value of 0.52—a result that is not statistically significant because the p-value is greater than 0.01. Both the four-fifths rule and chi square analysis in Table 1 show that the difference between the 21.7 percent Asian American and the 22.3 percent white selection rates is not statistical evidence of racially disparate impact against non-top 10 percent Asian American applicants to UT Austin.111
The lack of statistical evidence of adverse impact against Asian American applicants in Table 1 runs counter to claims that UT Austin’s race-conscious non-top 10 percent admissions policy racially discriminated against Asian Americans. Overall, the similar white and Asian American UT Austin selection rates (both close to 22 percent) and the lack of any component of UT Austin’s policy that facially discriminates against Asian Americans do not support the charge of Asian penalty in UT Austin admissions.112
IV. Fallacy Not Facts in SFFA Harvard and UNC Lawsuits
Edward Blum, the architect of several cases challenging traditional civil rights laws as racially discriminatory against whites113 (including Fisher v. University of Texas at Austin (Fisher I114 and Fisher II115)), is the president of an organization he founded called Students for Fair Admissions, Inc. (SFFA).116 SFFA is the named and organizational plaintiff in two lawsuits alleging that Harvard and UNC discriminate against white and Asian American students in violation of the Fourteenth Amendment Equal Protection Clause and Title VI of the Civil Rights Act of 1964.117 This Part’s examination of the complaints filed in the two SFFA cases reveals they do not contain the allegations most appropriate for an antidiscrimination lawsuit seeking to vindicate the civil rights of Asian Americans—allegations that Harvard and UNC operate admissions policies that either 1) facially discriminate against Asian Americans, 2) are facially nondiscriminatory but are executed in racially discriminatory manner against Asian Americans, or 3) have an unjustified racially discriminatory effect on Asian Americans based on comparisons of selection rates by race. Instead, the SFFA complaints rely heavily on two mathematical fallacies that feed falsities about affirmative action—the causation fallacy and the average-test-score-of-admitted-students fallacy.118 In so doing, they neglect the possibility that a white advantage in admissions could be the cause of an Asian penalty.
A. SFFA’s Attack on Racial Affirmative Action
Though critics of racial affirmative action have long invoked the prospect of anti-Asian American discrimination to argue that racial affirmative action is unfair and to suggest deficiencies in African Americans and Latinos as compared to Asian Americans,119 the Harvard and UNC lawsuits have been touted for their central focus on racial discrimination against Asian Americans.120 The two cases have garnered media attention as lawsuits alleging Asian American discrimination.121 Only the complaint filed in the Harvard case, however, opens with a description of the facts surrounding the rejection of an Asian American applicant.122 The UNC complaint describes the facts surrounding the rejection of a white applicant, not an Asian American applicant.123 Neither the Harvard nor the UNC complaints identify the rejected applicants they describe by name.124
After setting forth some of the admissions qualifications of an unnamed 2014 rejected Asian American applicant, the complaint in Students for Fair Admissions, Inc. v. President & Fellows of Harvard College125 charges Harvard with “intentionally discriminat[ing] against Asian-American applicants” and of operating an unconstitutional racial affirmative action policy.126 The SFFA complaint filed against UNC-Chapel Hill—the University of North Carolina’s most selective and flagship campus—describes the admissions qualifications of an unnamed white applicant who was not admitted to UNC’s 2014 entering class.127 The SFFA Harvard complaint alleges that Harvard’s consideration of race fails the Grutter v. Bollinger128 “narrow tailoring” equal protection requirement. It also seeks to overturn the holding in Grutter that increasing racial diversity among university students constitutes a constitutionally “compelling government interest” under the Equal Protection Clause.129
Like the SFFA complaint against Harvard, the UNC complaint accuses UNC of “intentionally and unconstitutionally discriminating against [members of the SFFA] on the basis of their race and ethnicity.”130 It also asserts: “Although UNC-Chapel Hill claims to use an applicant’s race and ethnicity only as one of many factors within its ‘holistic’ [admissions] system, statistical and other evidence establishes that race is a dominant factor in admissions decisions to the detriment of white and Asian-American applicants.”131 The complaint against UNC then goes on to allege, using text very similar to that in the Harvard complaint132, that UNC’s racial affirmative action policy is unconstitutional. It argues this on the grounds that the University’s consideration of race for the purpose of increasing diversity among its student population fails the Supreme Court’s narrow tailoring requirement as set forth in Grutter.133 In addition, like the Harvard case, the UNC case advocates the overruling of Grutter’s holding that universities having a compelling interest in considering race for the purpose of increasing diversity.134
B. SFFA’s Extraneous Facts
Significantly, the Harvard complaint neither includes nor mentions how the Asian American admission rate in 2014 compares to the white rate of admission in that year.135 Both SFFA complaints focus heavily on extraneous facts that are not legally cognizable evidence that Harvard and UNC discriminated against Asian Americans in the 2014 admissions cycle. For instance, instead of comparing selection rates by race, the Harvard complaint focuses on enrollment data in the form of an extended comparison of the percentage of Asian American students enrolled at Harvard and the percentage of Asian Americans enrolled at other universities and, even, several high schools.136 Without offering an explicit connection, the SFFA Harvard complaint implies that the percentage of Asian Americans enrolled at Harvard is unfairly low because of the percentage of Asian Americans enrolled at the particular selective universities and high schools mentioned in the complaint—UCLA, UC Berkeley, Hunter College High School in New York, and Thomas Jefferson High School for Science and Technology, “a magnet high school in Virginia.”
Similarly, a major portion of the Harvard complaint recounts historical accounts of Harvard’s 1920s and 1930s anti-Jewish discrimination, not how 2014 Asian American applicants were evaluated and selected by Harvard.137 Another significant portion of the complaint is comprised of anecdotal descriptions of the rejection of individual Asian American applicants and citations to newspaper accounts of the same during pre-2014 admissions cycles.138 Yet, because Harvard rejects so many extremely qualified students each year, such historical accounts and stories of well-qualified applicants being rejected are not, standing alone, compelling evidence that Harvard violated civil rights laws.139
At the same time, both the Harvard and UNC complaints are missing the key comparison needed to effectively demonstrate that the rejection of certain Asian American applicants was illegal140 because it violated either the Equal Protection Clause,141 Title VI,142 or Title VI disparate impact regulations.143 For example, despite alleging that Harvard violates Title VI by “intentionally discriminating against Asian Americans in admissions,” the complaint points to no evidence of documents or witness statements that suggests Harvard operated a covert, intentional policy of limiting Asian American enrollment in 2014.144 Instead of pointing to memos, emails, or statements that would be credible evidence that Harvard and UNC officials imposed an anti-Asian American admissions ceiling in 2014, the SFFA complaints focus on painting African Americans and Latinos, but not whites, as the students who are edging out Asian Americans.145
How SFFA describes the data in a table in the UNC complaint is an example of the failure to consider the role white advantage may play in Asian penalty. With no textual discussion that compares white and Asian American selection rates, the UNC complaint discusses a table—Table A, reporting 2006 academic selection rates by index score ranges and by race. By focusing only on differences between African American and Asian American selection rates, the UNC complaint misses an opportunity to evaluate whether a white selection advantage has caused an Asian penalty at UNC. In discussing the one table that does include the type of data useful in evaluating Asian penalty, albeit selection rate data from years prior to 2014, the UNC complaint notes only the higher selection rate for African American applicants that presumably result from UNC’s consideration of race for affirmative action purposes.146 Having fallen prey to the causation fallacy, the complaint’s discussion of Table A makes no mention of the white selection advantage that appears to exist due to the fact that whites are selected at higher rates than Asian American in the majority of index score ranges. As a consequence, the complaint invites the reader to infer that the higher selection rate for African Americans penalizes Asian Americans yet fails to note that selecting whites at a higher rate has a greater impact on Asian American odds of admission because the UNC applicant pool is likely comprised of far more white students than African American students.147
The only facially race-conscious admissions policy the Harvard complaint describes is Harvard’s overt and admitted policy goal of seeking to admit a critical mass of African American and Latino students and its tracking of projected enrollment by racial groups to accomplish this goal—a practice deemed constitutional under Grutter if exercised in a sufficiently restrained and “narrowly tailored” manner.148 Evidence that Harvard or UNC admissions officials were race conscious in implementing their racial affirmative action policies would only prove something the universities likely already openly admit and are willing to defend as constitutional—that they practice racial affirmative action.
Because the Supreme Court has held that the narrowly tailored consideration of race to achieve diversity is constitutional, as a matter of law,149 racial affirmative action is not per se racial discrimination against applicants who belong to racial groups not admitted through affirmative action. As observed by Professors Chin, Cho, Kang, and Wu, an Asian American applicant rejected by a university that practiced constitutional racial affirmative action in favor of African Americans and Latinos should be understood to have been subject to “neutral action.”150 Thus, evidence that Harvard and UNC operate a race-conscious racial affirmative action or even evidence that, in 2014, the institutions practiced racial affirmative action in an unconstitutional manner is different from evidence that proves the universities violate the civil rights of Asian American applicants.151 Comparing university selection rates by race is the way to prove that Harvard and UNC violated Title VI disparate impact law’s prohibition against recipients of federal funds operating facially neutral policies in a manner that has an unjustifiable adverse impact on Asian American applicants.152
The SFFA complaints fail to acknowledge the conceptual and practical reality that admissions practices with the purpose and effect of increasing the number of African American and Latino students admitted usually affect too few admission slots to be the policy that imposes an Asian penalty. It likewise fails to acknowledge that whites, the racial group making up the overwhelming majority of applicants and students admitted to most highly selective universities, are also the racial group that stands to benefit the most from the purposeful and effective implementation of an Asian American admissions ceiling. In other words, imposing an Asian American admissions ceiling would mean opening up more spots for whites.
If race discrimination against Asian Americans exists in Harvard or UNC admissions, it is more likely to be the product of bias in favor of whites. This is because the number of African American and Latino applicants, even under robust racial affirmative action policies, is too small to result in African Americans and Latinos edging out highly qualified Asian American applicants. A complaint vigorously targeting anti-Asian American ceilings should not be limited to arguing the implausible scenario that the admission of a comparatively small number and percentage of African American and Latino students causes the number of Asian Americans admitted to be unjustifiably and discriminatorily limited.153
V. Espenshade and Radford Endorse Racial
Affirmative Action Yet Feed the Fallacy
That it Causes Asian Penalty
Although it endorses racial affirmative action as the most viable policy for achieving racial diversity,154 No Longer Separate, Not Yet Equal (Espenshade & Radford book)155 has become a centerpiece of anti-affirmative-action advocacy that contends that affirmative action confers a Black bonus that is linked to an Asian penalty.156 This is likely because of (1) the book’s terminology—its use of terms like “[t]he black preference,”157 “bonus,”158 and “boost”159 as well as “[t]he Asian disadvantage”160 and “the black advantage”161—and (2) the book’s inclusion of a table, Table 3.5, it describes as reporting the effect of race in admissions converted into SAT points.162 Despite its causation-fallacy-reinforcing presentation, the Espenshade & Radford book is clear that its race effect findings should not be interpreted to mean admissions officials award African American students positive consideration because of their race, or that officials discriminate against Asian Americans because of their race.163
This Part illuminates how Espenshade and Radford’s interpretations of their own SAT-point equivalent findings about the role race plays in admissions differ vastly from the interpretations of their work by those who cite it regularly to criticize affirmative action and as evidence of an Asian penalty.164 In this Part, I also present my original analysis of the admissions data reported in the Espenshade & Radford book. Far from constituting smoking-gun evidence of Black bonus causing Asian penalty, the Espenshade & Radford book shows whites so greatly outnumber nonwhites in the public and private university applicant pools analyzed by Espenshade and Radford that the higher white admission rate impacts a far greater proportion of available admissions slots than racial affirmative action. This Part includes the same type of disparate impact analysis that I conducted in Part I.165 Yet, in contrast to the disparate impact analysis in Part I, which fails to show disparate impact against Asian Americans in Fall 2008 UT Austin admissions,166 this Part analyzes composite public and private university selection rates by race reported in Figure 3.3 of the Espenshade & Radford book and does find statistically significant disparate impact against Asian American applicants. Additionally, this Part shows that the book cited regularly for the proposition that African Americans enjoy a 310 SAT-point bonus and for the proposition that Asian Americans suffer a 140 SAT-point penalty reports a disparity in selection rate data between white and Asian American applicants sufficiently large to constitute prima facie evidence of a violation of Title VI disparate impact law stemming from white advantage as opposed to Black bonus.167
A. Espenshade and Radford Dataset and Black Bonus and Asian Penalty Findings
Of its ten chapters, the third chapter of the Espenshade & Radford book has garnered the most attention from opponents of racial affirmative action. This chapter includes logistic regression estimates of odds of admission by race under six hypothetical admissions models built by the authors using preexisting databases and responses to mailed surveys of students who attended eight of ten universities for which data was available to the authors168 for admissions in 1983, 1993, and 1997.169 One table in particular, Table 3.5, and a textual description of Table 3.5 have been enlisted repeatedly to argue that affirmative action results in racial discrimination against Asian Americans.170 That portion of the book reads as follows:
The second column of Table 3.5 indicates the size of admission preferences at private [surveyed] institutions. Once again, black applicants receive the largest admission bonus—equivalent to 310 SAT points. A black candidate with an SAT score of 1250 could be expected to have the same chance of being admitted as a white student whose SAT score is 1560, all other things equal in model 5. . . . On the other hand, an Asian candidate with a 1250 SAT score would be just as likely to be admitted at a private [surveyed] institution as a white student with an SAT score of 1110 . . . .171
Amici opposing UT Austin’s affirmative action policy in Fisher v. University of Texas at Austin172 and in the Harvard complaint rely on the SAT-point equivalents from Table 3.5173 Not mentioned, however, is the fact that Espenshade and Radford use a conversion methodology in creating Table 3.5, which they acknowledge inflates racial differences. The authors explicitly state that their approach “will have the effect of magnifying the size of the racial (or social class) preference or disadvantage as measured by their ACT-or SAT-point equivalents.”174 Those invoking Table 3.5 as evidence of Black bonus causing Asian penalty also fail to note that Espenshade and Radford explicitly caution that the SAT-point equivalents they report “will also be sensitive to the point in time they are measured” and that “[t]here is some evidence to suggest that the strength of the admission preference for underrepresented minority [African American and Latino] students has been declining.”175
It is also significant, but rarely mentioned when the Espenshade & Radford book is invoked to support the contention that universities confer large bonuses on African American applicants, that the authors themselves warn against interpreting the SAT-point differentials in Table 3.5 as proof that either African American applicants receive large SAT point bonuses or that Asian Americans suffer SAT point penalties based on race. They include this warning because the regression models underlying Table 3.5 use a “NSCE data set [that] does not include all of the information that admissions officers presumably evaluate when making admissions decisions.”176 Espenshade and Radford explain that because key information about applicants’ academic and non-academic credentials was not available to them (not part of the dataset they used), Table 3.5 should not be interpreted to mean African American applicants have a better chance of admission “just because” they are African American and that additional analysis is necessary to determine whether a particular selective university discriminates against Asian Americans.177
The authors’ exact words in explaining that Table 3.5 does not prove universities give bonus points to African American and Latino applicants because of their race are: “[I]t would be a mistake to interpret the data in Table 3.5 as meaning that elite college admission officers are necessarily giving extra weight to black and Hispanic candidates just because they belong to underrepresented minority groups.”178 The Espenshade & Radford book authors likewise caution that Table 3.5 is not dispositive evidence that selective universities racially discriminate against Asian Americans for the same reason—incomplete data—with these words:
“A similar reasoning pertains to the so-called Asian disadvantage in admissions. . . . It is likely that incorporating in our models an even fuller range of academic performance measures as well as these other nonacademic factors would cast the effect of coming from an Asian background in a different light.”179
With these comments, the Espenshade & Radford book authors are telling readers that the 310 SAT-point racial effect reported in Table 3.5 is not, in their view, caused by admission officers applying a different SAT test score standard for African American applicants. In fact, the authors say they believe the size of the racial effects they have found “would be diminished” if their analysis “were able to include these other considerations in [their] models”180 and observe explicitly that their findings are “not able to settle the question of whether Asian applicants experience discrimination in elite college admissions.”181
B. Using the Espenshade and Radford Dataset to Examine White Selection Advantage
The assertion that Black bonus from racial affirmative action causes significant decreases in Asian American admission rates is a fallacy because it ignores the mathematical reality that the number of African American students applying to historically majority-white selective universities is typically too small relative to other racial groups, including Asian Americans, to have this impact. It is, however, quite possible that a university could employ admissions practices that disadvantage Asian American applicants but inure to the benefit of the comparatively large number of white applicants who apply to those same selective universities. If a university’s admissions outcomes are found to result in a legally cognizable Asian penalty, as a matter of basic math, it is extremely unlikely to be caused by a Black bonus.
Although Chapter 3 of the Espenshade & Radford book is inconclusive on the question of Asian penalty,182 it does include selection data that I analyze in this Subpart. Relying on the racial group percentage data reported in Figure 3.3 of the Espenshade & Radford book along with the public and private university total applicant numbers reported in Table 3.2 of the book, I calculate the number of public and private university applicants and admitted students by race and present these findings in Tables 2 and 3 as well as Figures 3-6. In Table 4, I present my disparate impact analysis comparing the white and Asian American selections rates reported by Espenshade and Radford. In contrast to the finding above of no racially disparate impact in Table 1 due to the very similar white and Asian American selection rates at UT Austin,183 Table 4 below finds legally significant disparities in Asian American and white selection rates in the Espenshade and Radford composite dataset.184 However, it should be noted that because the admissions data underlying the tables and figures in this Subpart is composite data from many different universities and colleges over various time periods, Table 4 is illustrative of disparate impact methodology but not proof of Asian penalty at any single college or university within the dataset.
Figures 3 and 4 of this article are pie charts that illustrate my finding that white applicants are the racial group making up the greatest percentage (highest proportion) of both the selective public and private university applicant pools reported in the Espenshade and Radford dataset—85.6 percent and 55.7 percent of total applicants, respectively.185 Figures 3 and 4 also show that Asian Americans make up a relatively small portion of the public university applicant pools—5.3 percent of total applicants—yet a larger portion of the private university applicant pools—30.9 percent of total applicants.186 African Americans make up a small proportion of both the public and private university applicant pools—7 percent and 6.4 percent of total applicants, respectively.187 Likewise, Latino applicants are the smallest percentage of public university applicants in the Espenshade and Radford dataset—making up 2.1 percent of total public university applicants—and 6.9 percent of the total private university applicants.188
Figure 3. Proportion Reported by Espenshade & Radford Dataset and Calculated Public Universities Applicant Numbers by Race189
Figure 4. Proportion Reported by Espenshade & Radford Dataset and Calculated Private Universities Applicant Numbers by Race190
When whites make up the vast majority of applicants, their admission at a higher rate than Asian Americans becomes critical in evaluating Asian penalty. To make this clearer, below I present admissions selection rate data along with data showing how dramatically whites outnumber other racial groups in the Espenshade and Radford dataset. Tables 2 and 3 as well as Figures 5 and 6 show that the number of white applicants in both the public and private university admissions pools is vastly larger than the number of African American applicants, that whites are admitted at a higher rate than Asian Americans, and that whites make up the numerical majority of all selected students. Table 2 and Figure 5 show there are over ten times as many white pubic university applicants (23,258) as African American applicants (1901). White applicants are chosen for 13,071 of the 14,971 admission slots at public universities in the dataset—87.3 percent. Table 3 and Figure 6 show close to ten times as many white private university applicants (28,992) as African American applicants (3331). White applicants are chosen for 7451 of the 12,388 slots at private universities in the Espenshade and Radford dataset—60.1 percent.
Table 2. Calculated Public University Applicants and Admitted Students by Race Based on Espenshade & Radford Reported
* Higher admission rates for numerically underrepresented racial groups are legally justified if satisfy standards set forth by the Supreme in Court in Grutter v. Bollinger.
† Higher admission rate for racial group that has been historically the majority would need to be legally justified in order to comply with Title VI disparate impact law.
Table 3. Calculated Private University Applicants and Admitted Students by Race Based on Espenshade & Radford Reported Selection Rates192
* Higher admission rates for numerically underrepresented racial groups are legally justified if satisfy standards set forth by the Supreme in Court in Grutter v. Bollinger.
† Higher admission rate for racial group that has been historically the majority would need to be legally justified in order to comply with Title VI disparate impact law.
Even at the highest of the admission rates, Table 2 and Figure 5 show African Americans fill only 1112 of the 14,971 public admission slots and Table 3 and Figure 6 show African Americans fill only 1033 of the 12,388 private admission slots, 7.4 percent and 8.3 percent, respectively. This means that even when a university’s racial affirmative action policy is sufficiently robust to admit African Americans (or Latinos) at higher rates than whites and Asian Americans (such as the 58.5 percent African American rate for public universities and the 31.0 percent African American rate for private universities),193 the higher African American selection rate has little effect on the admission chances of white and Asian American applicants because of the comparatively small number of African American (or Latino) applicants and admits.
The proportions of Asian American applicants and the Asian American rates of selection vary greatly in the Espenshade and Radford dataset between the private university and public university admissions pools. Specifically, Table 2 and Figure 5 show Asian Americans comprise only 1440 of 27,171—8.3 percent of—students in the public university applicant pools—and are selected for 575 of the 14,971 public university slots,194 whereas whites, again make up over 80 percent of that pool and 87.3 percent of those admitted students.195 Table 3 and Figure 6 show Asian Americans make up 16,084 of 52,051 of the applicants in the private university applicant pools—23.9 percent—and are selected for 2959 of 12,388 private university admissions slots, with whites comprising over 60 percent of that pool and 60.1 percent of those admitted students.
Figure 5. Calculated Public University Applicants & Admits Compared by Race Based on Espenshade & Radford Dataset
Figure 6. Calculated Private University Applicants & Admits Compared by Race Based on Espenshade & Radford Dataset
Thus, when whites make up the vast majority of a university’s applicant pool, it is a mathematical reality that the white selection rate impacts a far greater number of admissions slots than the African American and Latino selection rates. Nevertheless, claims of Asian penalty by Justice Alito in Fisher II,196 some Fisher amici,197 and the SFFA Harvard and UNC complaints198 fail to frame discrimination against Asian Americans as potentially stemming from the accused university conferring an unfair advantage on white applicants over Asian Americans. This ignores the reality that, absent evidence of purposeful racial discrimination against Asian Americans, the strongest legal basis for asserting Asian penalty under settled law is to allege that facially race-neutral admissions practices have the effect of unjustifiably selecting Asian Americans at a lower rate than other races.
If, as is the case in the dataset from the Espenshade & Radford book, the selection rate for Asian Americans is lower than the selection rate for African Americans and Latinos, critics of racial affirmative action are correct to note that a university has a legal obligation to justify the disparity. The key observation missing from the SFFA lawsuits and other charges of Asian penalty focused on critiquing racial affirmative action is that, as matter of law, selecting African Americans and Latinos at higher rates than whites and Asian Americans can be legally justified under both the Equal Protection Clause and Title VI if the manner in which race is considered in admissions is narrowly tailored to achieve the university’s constitutionally recognized, compelling interest in admitting a racially diverse group of students.199 Specifically, a university should be able to argue successfully that a higher admission rate for a racial group included in the institution’s affirmative action policy is legally justified because the racial affirmative action policy itself is a legitimate “nondiscriminatory rationale” for the higher African American and Latino admission rates.200
Affirmative action, however, cannot be relied upon to justify a white selection advantage in admissions. As explained above, Title VI disparate impact law requires a university that receives federal funding to justify admitting members of different races at significantly different rates.201 The existence of narrowly tailored affirmative action policy would not suffice to justify a white selection advantage—a university selecting white students, members of a racial group making up the vast majority of applicants and admitted students, for admission at higher rates than Asian Americans. For this reason, comparing the rate at which a university selects white applicants to the Asian American selection rate should be a central aspect of analyzing Asian penalty.
Table 4 does just that using the Espenshade and Radford dataset. It compares the white and Asian American selection rates to public and private universities because (1) the higher rates for African Americans and Latinos impact only a very small portion of the available admissions slots and (2) whites are the racial group both admitted at the highest rate and not included in university racial affirmative action programs because they fill the overwhelming majority of admissions slots. While there is no finding of disparate impact in Part I’s analysis comparing the almost identical white and Asian American admission rates for the Fisher admissions cycle,202 Table 4 compares the white and Asian American admission rates reported in Figure 3.3 of the Espenshade & Radford book and finds the disparity between white and Asian American admissions rates is statistically significant for both the public and private university datasets.
Table 4. Adverse Impact Analysis of Asian American Admit Rates Reported in Table 3.3. of Espenshade & Radford203
|Public Universities||Private Universities|
|Selection Rates||Asian American rate: 39.9%
White rate: 56.2%
|Asian American rate: 18.4%
White rate: 25.7%
|Higher and Lower Selection Rate Ratio||39.9
——— = .71
——— = .72
|Four-Fifths Rule Violated if Ratio Less Than 0.80||.71 < .80||.72 < .80|
|Four-Fifths Rule Violated||Yes||Yes|
|What is p Value for this Chi Square?||p<0.00||p<0.00|
|Is p Value Less Than 0.01?
(p value<0.01 is significant)
|Yes, disparity in rates is statistically significant||Yes, disparity in rates is statistically significant|
Table 4 reveals that the disparity in the white and Asian American admission rates in the Espenshade and Radford dataset is legally significant under the same four-fifths rule and chi square analysis used in Table 1 of Part I.204 Both the public and private university Asian American selection rates are significantly lower than the respective white rates under the 80 percent rule and chi square analysis. The 39.9 percent Asian American composite public university admission rate is less than four-fifths rule of the 56.2 percent white admission rate. The same is true when the lower 18.4 percent composite private university Asian American admission rate is compared to the higher 25.7 percent white rate. In addition, chi square analysis of the differences between Asian American and white admission rates results in chi square values of 145.18 and 310.65, respectively and very low p-values, far less than 0.01. These very low p-values of less than 0.00 mean there is an extremely low likelihood that the white–Asian American selection rate disparities in Table 4 happened by chance.
In fact, both the chi square analysis and four-fifths rule analysis support the conclusion that the racial differences in the Espenshade and Radford composite admission rates in Table 4 are not the result of chance. Accordingly, if Asian American and white admission rates at a particular real-world university differed to the same extent as those in the dataset analyzed in Table 4, Title VI disparate impact law would require that the institution, if a recipient of Title VI federal funds, justify the use of the selection criteria that led to the legally significant racial disparity in selection rates.205 In short, Table 4 is the type of analysis missing from the SFFA lawsuits against Harvard and UNC. If Harvard and UNC selected Asian American applicants at substantially lower rates than white applicants, similar to the data in Table 4, it would be strong evidence to support a claim of discrimination against Asian American applicants,206 far stronger than linking Black bonus and Asian penalty.
Failing to acknowledge that anti-Asian American admissions policies are factually distinct from affirmative action policies ignores the possibility that white advantage—bias in favor of white applicants—could be impacting a far greater number of admissions slots than racial affirmative action for African Americans and Latinos. Presuming, without statistical proof, that a Black bonus penalizes Asian Americans operates to obscure the importance of comparing white and Asian American selection rates to smoke out racial discrimination against Asian Americans. There is a major distinction between attacking racial affirmative action—policies to include groups who ultimately only fill a numerical minority of admissions slots—and combatting racially exclusionary anti-Asian American discrimination.207 Investigating the existence and legality of a white admissions advantage—a higher selection rate for white applicants than Asian American applicants—is critical to interrogating Asian penalty.
Despite their intuitive appeal, claims of discrimination against Asian Americans based on the causation and average-test-score-of-admitted-students fallacies are fundamentally misleading. The mathematical reality is that a higher admission rate for the group that makes up the vast majority of all applicants—typically whites—affects the admission chances of Asian American applicants far more than racial affirmative action for racial groups that constitute a small numerical minority of applicants and admitted students. Likewise, the existence of a numerical difference in the group test score average of all admitted African American and all admitted Asian American students is not proof of the extent to which race plays a factor in a university’s admissions process.
Determinations of whether Asian Americans are the victims of a racially discriminatory admissions ceiling at a particular university should be based on analysis of (1) the facial details of the university’s admissions policy, (2) evidence as to whether universities are purposefully racially discriminatory in applying policies that are facially race-neutral as to Asian Americans, and, the primary focus of the Article, (3) comparison of the Asian American selection rate to the selection rate of whites—a racial group not covered by racial affirmative action because it constitutes a majority of applicants and admits to selective universities. Reliance on anecdotal accounts of highly qualified Asian applicants being rejected, racial differences and variation in overall enrollment numbers, and sensational invocations of the causation and average-test-score-of-admitted-students fallacies may be effective in stirring opposition to racial affirmative action but it does little to combat the likely cause of Asian penalty—white selection advantage. Consequently, examining whether a white advantage exists in admissions should be moved to the forefront of Asian penalty litigation if the goal of such efforts is truly to identify and vindicate anti-Asian American race discrimination.
. Mari J. Matsuda, Where is Your Body? And Other Essays on Race, Gender, and the Law 153–54 (1996).
. 42 U.S.C. § 2000d (2012); 34 C.F.R. § 100.3(b) (2016) (setting forth U.S. Department of Education regulations promulgated to enforce Title VI that prohibit recipients of federal funds from engaging in policies with unjustified racially disparate impact).
. See infra note 101 and accompanying text (explaining “four-fifths” rule for identifying a legally significant racial disparity in selection rates).
. See Complaint, Students for Fair Admissions, Inc. v. President & Fellows of Harvard Coll., 308 F.R.D. 39 (D. Mass. 2015) (No. 14-cv-14176), 2014 WL 6241935 [hereinafter Harvard Complaint]; Complaint, Students for Fair Admissions, Inc. v. Univ. of N.C., No. 1:14-cv-00954 (M.D.N.C. Nov. 17, 2014), 2014 WL 6386755 [hereinafter UNC Complaint].
. Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (Fisher I).
. Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198 (2016) (Fisher II).
. See generally Harvard Complaint, supra note 4 (alleging discrimination against Asian Americans by Harvard); UNC Complaint, supra note 4 (alleging discrimination against Asian Americans and whites by UNC).
. As William Kidder has noted in describing the case of the rejection of an Asian American applicant with “a strong transcript and say a 1510 or 1430 or 1360 on the SAT, it is exceedingly more likely that the student admitted instead was a White applicant with slightly lower academic credentials, not a Black or Latino applicant given an affirmative action plus factor.” William C. Kidder, Negative Action Versus Affirmative Action: Asian Pacific Americans are Still Caught in the Crossfire, 11 Mich. J. Race & L. 605, 615–16 (2006).
. In Fisher II, Justice Samuel Alito’s dissenting opinion focused extensively on the unproven and empirically unsupported claim that Asian Americans who applied to the University of Texas at Austin (UT Austin) suffered racial discrimination in admissions that stemmed from racial affirmative action). See Fisher II, 136 S. Ct. at 2216 (Alito, J., dissenting).
. See, e.g., id. at 2216, 2227 n.4 (Alito, J., dissenting) (describing UT Austin’s admissions policy as discriminating against Asian American students and invoking the causation fallacy by suggesting that increasing the number of African American and Latinos “by giving them an admissions boost” necessarily discriminates against Asian American applicants under “the laws of mathematics”); Ward Connerly, Creating Equal: My Fight Against Race Preferences 176 (2000) (author recounting his own assertion that under racial affirmative action “the son of a black four-star general would receive a preference over the daughter of an Asian dishwasher”); Stephan Thernstrom & Abigail Thernstrom, Reflections on The Shape of the River, 46 UCLA L. Rev. 1583, 1629 (1999) (reviewing William G. Bowen & Derek Bok, The Shape of the River: Long Term Consequences of Considering Race in College and University Admissions (1998)) (asserting that Asian Americans derive the greatest benefit when affirmative action is eliminated); Stuart Taylor, Symposium: Extrapolating from Fisher—Racial Preferences Forever, SCOTUSblog (June 23, 2016, 4:42 PM), http://www.scotusblog.com/2016/06/symposium-extrapolating-from-fisher-racial-preferences-forever [https://perma.cc/V96Z-JPE9] (describing UT’s Austin admissions policy as “discriminat[ing] flagrantly against Asian-Americans”).
. See, e.g., Richard Sander & Stuart Taylor, Jr., Mismatch: How Affirmative Action Hurts Students It’s Intended to Help, and Why Universities Won’t Admit It (2012). Professor Richard Sanders and journalist Stuart Taylor assert that “the ‘Asian penalty’ . . . has become a term of art among experts.” Id. at 276. They characterize affirmative action as the “boosting [of] blacks and Hispanics up to more elite institutions.” Id. at 277. They also rely on the book No Longer Separate, Not Yet Equal by Thomas Espenshade and Alexandria Radford, reporting that “black applicants received ‘an admissions bonus . . . equivalent to 310 SAT points’ relative to whites and more relative to Asians.” Id. at 18 (quoting Thomas J. Espenshade & Alexandria Walton Radford, No Longer Separate, Not Yet Equal 93, 137 (2009))
. Espenshade & Radford, supra note 11, at 92 tbl.3.5. This specific table presents results of the authors’ regression analysis of Fall 1997 admissions data provided by private universities and public universities reported as “SAT-point equivalents” in a table that reports African American applicants receive an advantage equivalent to 310 SAT points and Asian Americans are penalized the equivalent of 140 SAT points. Id. at 92–93, 92 tbl.3.5, 411 n.1.
. For an example of critics of affirmative action relying on Espenshade and Radford’s book, see Sander & Taylor, supra note 11, at 18, presenting the SAT-point equivalents in Espenshade & Radford, supra note 11, at 92 tbl.3.5.
. See SCOTUSblog on Camera: Edward Blum (Complete), SCOTUSblog, http://www.scotusblog.com/media/scotusblog-on-camera-edward-blum-complete [https://perma.cc/R6K6-2N9Y] (presenting a video of Edward Blum describing his plan to sue Harvard and UNC).
. This Article uses “SAT” to refer to the standardized college admissions test administered by the College Board. See Peter Applebome, Insisting It’s Nothing, Creator Says SAT, Not S.A.T., N.Y. Times, Apr. 2, 1997, at B6. Prior to 1994, “S.A.T.” was an abbreviation for the test’s former title—the Scholastic Aptitude Test. See id. Today, however, SAT is not an abbreviation because its letters do not stand for anything. Id.
. It also explains that if, on average, members of one racial group have higher scores than another racial group—as is the case for Asian Americans and African Americans, the average scores of all African Americans admitted will be lower than the average of all Asian Americans admitted because universities admit a significant portion of students based on non-SAT score criteria. In other words, while many of the students selected for admission to selective universities have perfect, or nearly perfect SAT scores, some students, including many African American students, are admitted to selective universities because they demonstrate excellence as measured by non-SAT-score-related admissions criterion. Cf. Kimberly West-Faulcon, More Intelligent Design: Testing Measures of Merit, 13 U. Pa. J. Const. L. 1235, 1266–68 (2011) (arguing that selecting students based on rank order SAT scores would constitute a deviation from merit because SAT scores explain only 13 percent of the variation in testtakers’ first-year college grades and, thus, to be based on true merit, selection must also consider other factors responsible for the 87 percent of variation of early college grades not explained by SAT scores).
. Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198 (2016) (Fisher II).
. See, e.g., id. at 2216, 2227 n.4 (Alito, J., dissenting) (arguing that UT Austin’s consideration of race in admissions should be deemed unconstitutional because UT Austin failed to explain “why the underrepresentation of Asian-American students in many classes justifies its plan, which discriminates against those students” (emphasis omitted)); Brief of Amicus Curiae the Asian American Legal Foundation & the Asian American Coalition for Education (Representing 117 Affiliated Asian American Organizations) in Support of Petitioner, Fisher II, 136 S. Ct. 2198 (No. 14-981), 2015 WL 5345842 [hereinafter Fisher II Brief of Asian American Legal Foundation]; Brief of Amicus Curiae California Ass’n of Scholars in Support of Petitioner at 29, Fisher II, 136 S. Ct. 2198 (No. 14-981), 2015 WL 5345837 [hereinafter Fisher II Brief of California Ass’n of Scholars] (arguing UT Austin’s race conscious policy is unconstitutional by describing a study of “elite private colleges” as finding that “Asian applicants with perfect SAT scores of 1600 had the same changes of being accepted as white applicants with 1460s and African-American applicants with 1150s”); Brief Amicus Curiae of Pacific Legal Foundation, Center for Equal Opportunity, American Civil Rights Institute, Project 21, National Ass’n of Scholars, Individual Rights Foundation, & Reason Foundation in Support of Petitioner at 32, 33, Fisher II, 136 S. Ct. 2198 (No. 14-981), 2015 WL 5345847 [hereinafter Fisher II Brief of Pacific Legal Foundation] (expressing the view that consideration of race in admissions results is discrimination against Asian Americans because “universities across the country continue to use race as a predominant factor in admissions” in a manner that supports the conclusion that “Asians are frequently discriminated against at least as much, and sometimes even more than whites”).
. 42 U.S.C. § 2000d (2012).
. Jerry Kang, Negative Action Against Asian Americans: The Internal Instability of Dworkin’s Defense of Affirmative Action, 31 Harv. C. R.-C. L. L. Rev. 1, 3 (1996).
. Espenshade & Radford, supra note 11.
. The causation fallacy in selective higher education admissions is the incorrect notion that admitting African American and Latino students at higher rates under a racial affirmative action policy causes the admission chances of white and Asian American students to be substantially lower than they would otherwise be in the absence of racial affirmative action. Goodwin Liu, The Causation Fallacy: Bakke and the Basic Arithmetic of Selective Admissions, 100 Mich. L. Rev. 1045, 1046 (2002).
. See, e.g., Sander & Taylor, supra note 11, at 12 (“[I]ncreasingly, racial admissions preferences are used to advantage Hispanics, biracial Americans, and black foreign nationals and to disadvantage Asian Americans.”).
. See Matsuda, supra note 1, at 153–54.
. See id.
. See generally Kang, supra note 20 (explaining “negative action” against Asian Americans as in force if a university denies admission to an Asian American applicant who would have been admitted had that applicant been white).
. This is an appropriate distinction because universities have the legal authority to adopt and have adopted racial affirmative action policies designed to increase the admission rate for Asian Americans. See, e.g., Students for Fair Admissions, Inc. v. President & Fellows of Harvard Coll., 308 F.R.D. 39, 49 (D. Mass.), aff’d, 807 F.3d 472 (1st Cir. 2015) (noting that Harvard had such a policy) Complaint at 2–3, Rios v. Regents of the Univ. of Cal., No. 99-0525 (N.D. Cal. Feb. 2, 1999) (noting that UC Berkeley had an affirmative action policy for Filipino applicants prior to passage of the state’s anti-affirmative action law, California Proposition 209).
. Gabriel J. Chin et al., Beyond Self-Interest: Asian Pacific Americans Toward a Community of Justice, A Policy Analysis of Affirmative Action, 4 UCLA Asian Pac. Am. L.J. 129, 159 (1996) (“First, [Asian Pacific Americans] APAs could be included in race-based affirmative action. Second, APAs could be excluded from affirmative action and treated indistinguishably from Whites who are similarly ineligible. Third, APAs could be capped by an admissions ceiling such that they are denied admission in order to admit more Whites (not other racial minorities). These three regimes may be called affirmative action, neutral action, and negative action.”).
. Id. at 159–60.
. Id. (footnote omitted).
. See Liu, supra note 22, at 1046.
. Id. at 1073. Of those 424 candidates with similar grades and test scores, 378 were white and forty-six were nonwhite. Id. Gratz applied to the University of Michigan with a grade point average (GPA) of 3.8 and an ACT test score of 25 (out of 36), see Petition for Writ of Certiorari at 4, Gratz v. Bollinger, 539 U.S. 244 (2002) (No. 02-516), 2002 WL 32101145, at *4, and Justice Liu showed that her admissions chance was reduced, at most, by 7 percent—from 39 percent to 32 percent. Liu, supra note 22, at 1073 (explaining that admitting African Americans and Latinos with “grades and test scores comparable” to those of Jennifer Gratz at a rate of 100 percent only decreased Gratz’s admissions chances by 7 percent because of how greatly the number of white applicants exceeded the number of minority applicants). Had there been no racial affirmative action policy, there was still a 61 percent likelihood that the university would have rejected Gratz. Id. at 1074 (“[T]he smallness of the pool of minority applicants and the relevance of nonobjective criteria in selecting among large numbers of white applicants conspire to limit the effect on white applicants of substantial preferences for minority applicants.”). “[G]iven the predominance of [white and Asian American] applicants with lesser qualifications, it is more probable that Gratz was displaced by a [white or Asian American] applicant than an [African American or Latino] applicant.” Defendant-Intervenors’ Response to Defendants’ Motion for Summary Judgment as to Plaintiffs’ Claim for Declaratory and Injunctive Relief, Gratz v. Bollinger, 353 F. Supp. 2d 929 (E.D. Mich. 2005) (No. 97-75231), 2000 WL 35504966.
. Liu, supra note 22, at 1074.
. See generally Sherick Hughes et al., Causation Fallacy 2.0: Revisiting the Myth and Math of Affirmative Action, 30 Educ. Pol’y 63 (2016) (considering the effect of eliminating all African American and Latino applicants from selective admissions).
. Id. at 80, 81 fig.1 (reporting admit rates and percent change in admit rates for white and Asian American applicants when African American and Latino applicants are removed from Harvard, University of Michigan, UNC, and UT Austin applicant pools).
. Id. (from 5.84 percent to 6.84 percent).
. Id. (from 27.59 percent to 31.68 percent).
. The number of white and Asian American applicants to certain selective universities far outstrips the number of Latino and African American applicants because a very large proportion of all Asian American students attend “a relatively small number of elite and highly visible schools.” Nat’l Comm’n on Asian Am. & Pac. Islander Research in Educ., Asian Americans and Pacific Islanders: Facts, Not Fiction, Setting The Record Straight 6 (2008), https://secure-media.collegeboard.org/digitalServices/pdf/professionals/asian-americans-and-pacific-islanders-facts-not-fiction.pdf [http://perma.cc/49Z3-6EUV]. Because Asian American applications are “concentrated in a small number of schools,” the percentage of Asian Americans in the applicant pool for the 5 percent of all Title VI institutions where they are concentrated is relatively high compared to other nonwhite racial groups. See id. at 6, 8 (explaining that the concentration of Asian Americans in a small percentage of American universities in just a few states, including California, New York, and Texas, gives the false impression that Asian Americans are “taking over” American colleges and universities when many Asian American students attend colleges that are “nonselective or minimally selective”).
. Several were briefs submitted in Fisher I and Fisher II that rely on this fallacy. See e.g., Brief Amici Curiae For Richard Sanders And Stuart Taylor, Jr. in Support of Neither Party at 3, 15, Fisher v. Univ. of Tex. at Austin (Fisher I), 132 S. Ct. 1536 (2012) (No. 11-345), 2012 WL 1950266, at *3, *15 [hereinafter Fisher I Sander & Taylor Brief] (describing differences in African American as compared to white and Asian American mean SAT scores at UT Austin as “staggering” and implying, without asserting, that the lower average SAT scores of African American were definitive proof that UT Austin conferred “very large preferences” and “very large preferences [on] blacks”).
. Examples of non-SAT-score-related admissions criteria relied upon by selective universities include a high GPA, letters of recommendation, an SAT score that is significantly higher than the average SAT score of students attending the applicant’s high school, an SAT score higher than the average SAT score of students from the same socioeconomic background, the quality of an applicant’s written essays, an applicant’s work experience, leadership experience, and community service. Moreover, selective universities do not admit students based solely on SAT-score rank order because institutions correctly understand that the SAT college admissions test is a useful, but far from perfect, predictor of admissions-related merit. See West-Faulcon, supra note 16, at 1266–68.
. This is not to say Asian American applicants lack stellar non-SAT score admissions credentials. It just means that fewer African Americans who win in the selective admissions process do so based on having perfect or close to perfect SAT scores. See William T. Dickens & Thomas J. Kane, Racial Test Score Differences as Evidence of Reverse Discrimination: Less Than Meets the Eye, 38 Indus. Rel. 331, 356 (1999)
. See id.
. Id. at 357 (“[O]ne cannot simply compare the test scores of blacks and whites on the same job or at the same school to determine whether the process of choosing people for that job or school is race-blind.”).
. Id. at 334–35.
. Id. at 338.
. See Muriel Niederle & Lise Vesterlund, Explaining the Gender Gap in Math Test Scores: The Role of Competition, J. Econ. Perspectives, Spring 2010, at 129, 129 (noting that a “gender gap has been documented for a series of math tests including the AP calculus test, the mathematics SAT, and the quantitative portion of the Graduate Record Exam (GRE),” despite the fact that “[t]he number of math and science courses taken by female high school students has increased”). Scholars have also noted that a study found “a two to one male-female ratio” among “students who scored 800 on the math SAT in 2007.” Id. (citing Glenn Ellison & Ashley Swanson, The Gender Gap in Secondary School Mathematics at High Achievement Levels: Evidence From the American Mathematics Competitions, J. Econ. Perspectives, Spring 2010, at 109).
. Reasons for the former and the latter are explained, at least in part, by the psychometric characteristics of the conventional college admissions tests like the SAT. As to the former, universities do not rely solely on SAT scores to make admissions decisions because the test is not a perfect predictor of future academic or real-world success. With respect to the latter, social science research that has emerged over the last several decades supports the theory that part of the racial skew in scores on currently-used college admissions tests is attributable to theoretical inadequacies in the theories of intelligence underlying such tests. See, e.g., West-Faulcon, supra note 16, at 1275–77 (explaining various critiques of the theory of intelligence as the general factor (“g”) and the increased predictive power of mental tests designed to measure modern non-g-based theories of intelligence with smaller and, in the case of some intelligence tests, no racial differences in scores).
. Far from perfect in its prediction, the 13 percent power of predicting first-year college GPA based on SAT score and the 23 percent predictive power of using SAT score combined with high school GPA to predict the same early college outcomes are helpful but not complete information for universities to use in assessing the academic merit of applicants. See id. at 1264–68.
. The tremendous historical and current racial differences between African Americans and Latinos as compared to whites and (some) Asians, in income, wealth and educational opportunity, all of which have been and continue to be exacerbated by racially segregated housing and exclusion of nonwhites from racially and income integrated neighborhoods with higher quality public school systems, are well-known and well-documented. See Gary Orfield & Jongyeon Ee, Civil Rights Project, Segregating California’s Future: Inequality and Its Alternative 60 Years After Brown v. Board of Education 40 (2014), http://civilrightsproject.ucla.edu/research/k-12-education/integration-and-diversity/segregating-california2019s-future-inequality-and-its-alternative-60-years-after-brown-v.-board-of-education/orfield-ee-segregating-california-future-brown-at.pdf [https://perma.cc/RW5V-EH8N] (“There is a clear pattern of intense double segregation by race and poverty for black and Latino children in California’s metro areas . . . . [T]he default for a white or Asian family is a middle-class school, while the default for a Latino or black family is a school of concentrated poverty.”). This is not to say that there are not significant intragroup differences in addition. See, e.g., Frank Shyong, The Term ‘Asian’ May Be Overly Broad but California is Stuck With It, L.A. Times, (Oct. 21, 2015 2:00 AM), http://www.latimes.com/local/california/la-me-asian-veto-20151021-story.html [https://perma.cc/97YF-XHTX] (describing tremendous variation in socioeconomic status and educational attainment among Asian racial subgroups).
. See West-Faulcon, supra note 16, at 1270 (describing racial skew in SAT scores).
. Dickens & Kane, supra note 42, at 337–38. Further demonstrating that there is no relationship between the existence of numerical difference in racial group mean SAT scores and racial preferences, mathematical models found some racial difference in the overall average of all selected students when racial groups’ means were compared “no matter where the [test score cut-off] range is positioned.”
. See, e.g., West-Faulcon, supra note 16, at 1275–77 (describing measurement limitations of standardized test like the SAT).
. According to the mathematical modeling, the only way to admit a class of Asian American and African American students with the same average SAT score for both racial groups, would be to require every single applicant to the college or university in question to have “the same test score.” Dickens & Kane, supra note 42, at 338.
. Id. at 337–38.
. See generally Kimberly West-Faulcon, Forsaking Claims of Merit: The Advance of Race-Blindness Entitlement in Fisher v. Texas, in 29 Civil Rights Litigation and Attorney Fees Annual Handbook 335, 347, 352 (Steven Saltzman & Cheryl I. Harris eds., 2013) (citing Univ. of Tex. at Austin Office of Admissions, Implementation and Results of the Texas Automatic Admissions Law (HB 588) at the University of Texas at Austin: Demographic Analysis of Entering Freshman Fall 2008 (2008), https://utexas.app.box.com/s/jie61tqplwb45kexwoefowwgz99wsqro/1/7736553677/23476775929/1 [hereinafter UT Freshmen Fall 2008 Report]) (analyzing UT Austin admissions data reported in the UT Freshman Fall 2008 Report, supra, and reporting the number of applicants and admits within the racial categories of white, Asian American, African American, and Latino for 2008 in Figure 1 and the number of automatic admits and total admits within these racial categories for 2008 in Figure 3).
. The overall admission rate (under the combined top 10 percent process and the non-top 10 percent process) for Asian American applicants to UT Austin in 2008 was 53.2 percent, whereas the white admission rate was 46.9 percent. In contrast, the African American and Latino overall admission rates were 32.5 and 21.7 percent, respectively. I have calculated these overall admission rates using the Fall 2008 Asian American and white applicant and admission numbers set forth in Figure 1. See infra Figure 1.
The total number of Asian American applicants was 4344 (1744 top 10 percent + 2600 non-top 10 percent) and the total number of Asian American admits was 2309 (1744 top 10 percent + 565 non-top 10 percent), making the overall admit rate 53.2 percent (2309/4344=0.5316). The total number of white applicants was 14,038 (4440 top 10 percent + 9598 non-top 10 percent) and the total number of white admits was 6582 (4440 top 10 percent + 2142 non-top 10 percent), making the overall admit rate 46.89 percent (6582/14,038=0.4689).
The total number of African American applicants was 2234 (582 top 10 percent + 1652 non-top 10 percent) and the total number of African American admits was 728 (582 top 10 percent + 146 non-top 10 percent), making the overall admit rate 32.6 percent (728/2234=0.3259).
The total number of Latino applicants was 12,081 (2218 top 10 percent + 3863 non-top 10 percent) and the total number of Latino admits was 2621 (2218 top 10 percent + 403 non-top 10 percent), making the overall admit rate 21.7 percent (2621/12,081=0.217).
. Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (Fisher I).
. Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198 (2016) (Fisher II).
. Brief for Petitioner at 2, Fisher I, 133 S. Ct. 2411 (2013) (No. 11-345), 2015 WL 5261568, at *2.
. See infra Figure 2 (showing UT Austin admission rates for white and Asian American non-top 10 percent Fall 2008 applicants was 22.3 percent and 21.7 percent, respectively, while the Latino and African American rates were 10.4 percent and 8.8 percent, respectively).
. See infra Figure 2.
. See Fisher II, 136 S. Ct. at 2215–16, 2227 n.4 (Alito, J., dissenting).
. Brief for Petitioner, supra note 60, at 26.
. See Fisher II, 136 S. Ct. at 2207.
. See Plaintiff’s Statement of Facts in Support of Motion for Partial Summary Judgment at 19, Fisher v. Univ. of Tex. at Austin, 645 F.Supp.2d 587 (W.D. Tex. 2009) (No. 1:08-CV-00263-SS) [hereinafter Plaintiff’s Statement of Facts] (“Some number of non-Top Ten Percent Law applicants may be automatically admitted to a particular academic program based solely on their high AI [academic index] score. UT Austin does not appear to maintain admission figures for the number of applicants admitted based on high AI scores.” (citation omitted)).
. See UT Freshmen Fall 2008 Report, supra note 56, at 2.
. Fisher acknowledged that “a number of the 216 non-Top 10% minority enrollees would have been admitted without regard to their race.” Brief for Petitioner, supra note 60, at 9, (stating that some non-top 10 percent “[minority enrollees] were admitted based solely on high AI scores” and that “[m]any others would have been admitted under an AI-PAI system unaffected by race”).
. See Brief for Respondents at 26, Fisher v. Univ. of Tex. at Austin (Fisher I), 133 S. Ct. 2411 (2013) (No. 11-345), 2012 WL 3245488, at *26.
. See Memorandum in Support of Defendant’s Cross-Motion for Summary Judgment and in Opposition to Plaintiff’s Motion for Partial Summary Judgment at 3–5, Fisher v. Univ. of Tex. at Austin, 645 F. Supp. 2d 587 (W.D. Tex. 2009) (No. 08-00263-SS); see also UT Freshmen Fall 2008 Report, supra note 56, at 2–3. The academic index score could range from 0 to 4.1. UT Freshmen Fall 2008 Report, supra note 56, at 2.
. See UT Freshmen Fall 2008 Report, supra note 56, at 2–4.
. For an example of this neglect, see Brief for the Asian American Legal Foundation and the Judicial Education Project as Amici Curiae in Support of Petitioner at 6, 9–12, Fisher I, 133 S. Ct. 2411(No. 11-345), 2012 WL 1961250, at *6, *9–12 [hereinafter Fisher I Brief of Asian American Legal Foundation], which asserted that “[e]very slot allocated to someone who would not have been admitted but for his race is a slot denied to someone else who would have been admitted but for their race” and suggested that because African American and Latino students are admitted with lower SAT scores than Asian American students, racial affirmative action harms Asian American applicants to UT Austin.
. UT Freshmen Fall 2008 Report, supra note 70, at 2.
. See id. at 3.
. Id. at 2; see also, Affidavit of Kedra B. Ishop at 1–2, Fisher v. Univ. of Tex. At Austin, 645 F. Supp. 2d 587 (W.D. Tex. 2009) (No. 08-00263-SS), [hereinafter Ishop Affidavit], reprinted in Plaintiff’s Statement of Facts, supra note 66, at 221–22.
. UT Freshmen Fall 2008 Report, supra note 56, at 2 (listing the seven subfactors for the special circumstances personal achievement factor).
. Fisher v. Univ. of Tex. at Austin (Fisher II), 136 S. Ct. 2198, 2207 (2016) (“[T]here is no dispute that race is but a ‘factor of a factor of a factor’ in the holistic-review calculus.” (quoting Fisher, 645 F. Supp. 2d at 608)).
. See Oral Deposition of Kedra Ishop at 57, Fisher, 645 F. Supp. 2d 587 (No. 08-00263-SS), reprinted in Plaintiff’s Statement of Facts, supra note 66, at 78–105 (including deposition testimony from UT Austin Associate Director of Admissions Kedra Ishop stating: “[R]ace, within the context of the rest of the application, can be beneficial to any applicant, to Whites as well as minority applicants.”); see also Fisher II, 136 S. Ct. at 2207 (“[T]he consideration of race, within the full context of the entire application, may be beneficial to any UT Austin applicant—including whites and Asian-Americans.”).
. See Ishop Affidavit, supra note 77, at 2 (“No numerical factor is assigned to any component of the [Personal Achievement Index]. And race may be beneficial for minorities or non-minorities alike, depending on the factors of a particular situation.”).
. See id.
. See Ishop Affidavit, supra note 77, at 4 (explaining that “92% of the Texan spaces were awarded to Top 10% applicants” the year Abigail Fisher applied to UT Austin such that only “[a]pproximately 841 spaces remained” for non-top 10 percent applicants like Fisher).
. See, e.g., Fisher II Brief of the Asian American Legal Foundation, supra note 18; see also Fisher II Brief of California Ass’n of Scholars at 29, supra note 18 (relying on a study of “elite private colleges” to support its allegation that UT Austin racially discriminates against Asian Americans in the application of its SAT test score standard without acknowledging that UT Austin is not a private college and was not part of the study), and Fisher II Brief of Pacific Legal Foundation at 32, supra note 18 (referencing “studies” and “anecdotal reports of racial discrimination in universities nationwide” to support claim that “Asians are frequently discriminated against” even though none of the studies and anecdotal reports include references to or data from UT Austin admissions); see also, e.g., Fisher II, 136 S. Ct. at 2216, 2227 n.4 (Alito, J., dissenting) (asserting that UT Austin’s consideration of race to provide “a boost to African-Americans and Hispanics inevitably harms” Asian Americans because it “decreas[es] their odds of admission” without mention of fact that selecting whites at a higher rate than Asian Americans has a far greater negative impact on Asian American admissions because the number of white applicants to UT Austin is substantially larger than the number of Asian American applicants and because the number of African American and Latino applicants is so much smaller than either the number of Asian American or white UT Austin applicants that a higher selection rate for those groups has very minimal impact on the odds of an Asian American student being admitted) .
. See sources cited supra note 84.
. See supra Parts I, II for detailed explanations of these fallacies.
. Espenshade & Radford, supra note 11. The briefs neither acknowledge that the data relied upon in Espenshade and Radford’s book is not UT Austin admissions data, nor do they note that the dataset in Espenshade & Radford’s book relies on admissions cycles from several decades ago. See, e.g., Fisher II Brief of California Ass’n of Scholars, supra note 18, (discussing various universities, including UCLA School of Law, University of Delaware, the Wright State University School of Medicine, University of South Alabama College of Medicine, George Mason University School of Law, Harvard, Yale, and Princeton); Fisher II Brief of Pacific Legal Foundation, supra note 18 (describing Asian American admissions at University of Wisconsin but not UT Austin). See infra Part V for details on the data sources within Espenshade & Radford’s book.
. See, e.g., Fisher II Brief of Asian American Legal Foundation, supra note 18, at 4 (characterizing Asian Americans as the group “most harm[ed]” by UT Austin’s affirmative action policy but never presenting details of UT Austin’s admissions policy). The Asian American Legal Foundation brief in Fisher II relies heavily on the average-test-score-of-admitted-students fallacy and describes various news articles and opinion pieces asserting the existence of racial discrimination against Asian Americans by colleges other than UT Austin. Id. at 25–27. It also describes the admissions controversy surrounding consideration of race in admission to a selective San Francisco high school and Harvard University’s discrimination against Jewish students in the 1920s and 1930s. Id. at 19–20.
. H.B. 588, 1997 Leg., 75th Reg. Sess. (Tex. 1997).
. See infra Figure 1.
. The data presented in Figure 1, originally printed in West-Faulcon, supra note 56, at 352, is based upon my analysis of publicly available UT Austin admissions data. See UT Freshman Fall 2008 Report, supra note 56, at 6, 8.
. See supra Figure 1.
. The data presented in Figure 2 originally printed in West-Faulcon, supra note 56, at 349, is based upon my analysis of UT Austin admissions data. See UT Freshman Fall 2008 Report, supra note 70, at 6, 8.
. See supra Figure 2.
. See supra Figure 2.
. See supra Figure 2.
. Sander & Taylor, supra note 11, at 6 (describing that due to “large racial preferences, . . . many whites and Asians . . . get passed over [in place of] the many blacks and Hispanics who receive [these] preferences”)
. See supra Figure 2.
. A violation of Title VI regulations does not require proof of purposeful discrimination. In Alexander v. Sandoval, 532 U.S. 275 (2001), the Supreme Court ruled that no private right of action exists to enforce Title VI disparate impact regulations. Id. at 292–93. Whether rejected minority applicants may bring Title VI effect-discrimination cases under 42 U.S.C. § 1983—an approach endorsed by Justice Stevens in his dissent in Sandoval, 532 U.S. at 299–301 (Stevens, J., dissenting)—has not been decided definitively by the Court. Compare id., with Johnson v. City of Detroit, 446 F.3d 614, 629 (6th Cir. 2006) (holding a federal regulation alone cannot create a right enforceable through § 1983), and Save Our Valley v. Sound Transit, 335 F.3d 932, 939 (9th Cir. 2003) (holding “regulations cannot independently create rights enforceable through § 1983”).
Even if private enforcement of Title VI regulations is precluded, however, individuals are still permitted to file complaints with the U.S. Department of Education Office of Civil Rights (OCR) alleging that an institution’s admissions policies have a Title VI discriminatory effect on the basis of race. See Brence D. Pernell, Note, Aligning “Educational Necessity” with Title VI: An Enhanced Regulatory Role for Executive Agencies in Title VI Disparate Impact Enforcement, 90 N.Y.U. L. Rev. 1369, 1380 (2015) (“[P]rivate parties may file disparate impact complaints with the [Department of Education], which has the power to investigate, review, and revoke federal funds pursuant to Title VI.”). Title VI itself provides: “No person in the United States shall, on the ground of race, color, or national origin, be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.” 42 U.S.C. § 2000d (2012). Liability under Title VI is identical to the federal Equal Protection Clause’s jurisprudence in the requirement that plaintiffs prove a discriminatory purpose. See Washington v. Davis, 426 U.S. 229, 240 (1976) (discussing cases holding that challengers have to show purposeful discrimination). Title VI implementing regulations set forth the “specific discriminatory actions” that recipients of federal funds must avoid. 34 C.F.R. § 100.3(b) (2016) (emphasis omitted).
. See 29 C.F.R. § 1607.4(D) (2016) (explaining the “four-fifths rule” for adverse impact). The four-fifths rule is explained as follows:
A selection rate for any race, sex, or ethnic group which is less than four-fifths (4/5) (or eighty percent) of the rate for the group with the highest rate will generally be regarded by the Federal enforcement agencies as evidence of adverse impact, while a greater than four-fifths rate will generally not be regarded by Federal enforcement agencies as evidence of adverse impact.
. Id.; cf. Kimberly West-Faulcon, The River Runs Dry: When Title VI Trumps State Anti-Affirmative Action Laws, 157 U. Pa. L. Rev. 1075, 1129–30 (2009) (applying four-fifths analysis to admissions cycles at University of California at Los Angeles (UCLA) and the University of California at Berkeley (UC Berkeley)).
. 29 C.F.R. § 1607.4(D) (explaining that a selection rate of less than 80 percent of the racial group with the highest selection rate will be treated as evidence of adverse impact under Title VI implementing regulations).
. Chi square analysis is an example of a statistical test plaintiffs typically rely upon in litigation to prove disparate impact. See, e.g., Powers v. Alabama Dep’t of Educ., 854 F.2d 1285, 1298 n.21 (11th Cir. 1988) (“Several courts have approved chi-square analysis as an alternative to standard deviation analysis . . . .”); Castaneda v. Partida, 430 U.S. 482, 496 n.17 (1977) (observing that when expected results and actual results differ by more than two or three standard deviations, the statistical disparity in selection rates constitutes evidence of a Title VI disparate impact). Courts have interpreted Title VI to require that the racial disparity exists not just within the overall applicant pool but also within the subpopulation of qualified applicants. In other words, there is proof of Title VI discriminatory effect if, considering only qualified applicants, the racial disparities are large enough to be statistically significant. See, e.g., Hazelwood Sch. Dist. v. United States, 433 U.S. 299, 307–08 (1977) (“Where gross statistical disparities can be shown, they alone may . . . constitute prima facie proof of a pattern or practice of discrimination.”).
. Castaneda, 430 U.S. at 496 n.17.
. See supra Figure 2 (showing my calculations of white and Asian American admit rates for UT Fall 2008).
. The Asian American admission rate of 21.7 percent was close to 100 percent of, not less than 80 percent of, the white admission rate of 22.3 percent. Dividing the 21.7 percent rate by the 22.3 percent rate equals 0.97, a ratio greater than 0.80 (80 percent). See supra Table 1.
. See supra Table 1.
. It is beyond the scope of this Article, but theoretically possible, to argue that Asian American applicants are so superior to applicants of all other racial groups that it violates the Equal Protection Clause and Title VI disparate impact law to only admit them at rates equal to white applicants. If a rejected Asian American plaintiff made such a claim, evidence that Asian Americans were admitted at equal or even dramatically higher rates than whites would not be dispositive. The U.S. Supreme Court requires that equal protection plaintiffs prove both racially discriminatory effect, see, e.g., Palmer v. Thompson, 403 U.S. 217, 225 (1971), and racially discriminatory purpose, see, e.g., Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 272 (1979); Washington v. Davis, 426 U.S. 229, 241 (1976).
It is possible that a rejected Asian American plaintiff might prove racially discriminatory effect against Asian American applicants based on a comparison of a subset of all applicants. For instance, if such a plaintiff could demonstrate that within the pool of highly qualified applicants, Asian Americans were disproportionately rejected as compared to white applicants, then the racial disparity of admission rates could constitute evidence that UT Austin’s admissions practices had a racially discriminatory effect on Asian American applicants. None of the amicus briefs alleging that racial affirmative action policies harm Asian Americans, however, suggest that the universities harbor a racially discriminatory purpose, in the sense that they consider race for the purpose of excluding Asian Americans. For example, the California Association of Scholars asserted: “The CAS is not arguing that college administrators bear ill-will toward Asians (or towards Scots-Irish, Cajuns, Hmong, or any other underrepresented group that is ignored by fashionable diversity policies).” Fisher II Brief of California Ass’n of Scholars, supra note 18, at 30. If the reviewing court interprets the purpose of the Equal Protection Clause and Title VI as dismantling racial hierarchies and protecting racial groups from exclusion, however, such a claim would also be weaker based on the analysis I set forth in this Article.
. I presume allegations of Asian penalty, if framed as legal claims, would be assertions that a university has violated either: 1) the Equal Protection Clause, as in Palmer v. Thompson, 403 U.S. 217, 225 (1971), which held that racially exclusionary effects were necessary to prove an equal protection violation; 2) Title VI, 42 U.S.C. § 2000d (2012); or 3) Title VI disparate impact regulations, 34 C.F.R. § 100.3(b) (2016).
. Edward Blum, a former Texas stockbroker, is an opponent of race-conscious policies, ranging from the Voting Rights Act of 1965 to university affirmative action policies, on a mission to put an end to “race-based laws and policies.” Joan Biskupic, Special Report: Behind U.S. Race Cases, a Little Known Recruiter, Reuters (Dec. 4, 2012, 1:50 PM) http://www.reuters.com/article/us-usa-court-casemaker-idUSBRE8B30V220121204 [https://perma.cc/F2R6-EV96]; Stephanie Mencimer, Here’s the Next Sleeper Challenge to Affirmative Action: The Guy Who Engineered the Fisher Case in the Supreme Court Isn’t Done Yet, Mother Jones (Jul. 19, 2016, 6:00 AM) http:// http://www.motherjones.com/politics/2016/07/abigail-fisher-going-stay-mad [https://perma.cc/TN4K-6AJX]. Indeed, Blum procured both the financial funding and the named plaintiff for the Fisher case. See Biskupic, supra. In Blum’s pursuit of a viable plaintiff to challenge the University of Texas’s affirmative action: “He set up a Web address, utnotfair.org, which asked spurned University of Texas students to contact Blum and relate their experiences. He gave speeches to the Young Conservatives of Texas and similar groups, and hounded everyone he knew in the state.” Id.
. Fisher v. Univ. of Tex. at Austin, 133 S. Ct. 2411 (2013) (Fisher I).
. Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198 (2016) (Fisher II).
. Harvard Complaint, supra note 4, at 8; see Mencimer, supra note 113 (describing Students for Fair Admissions as “a group founded and run by” Blum).
. The University of North Carolina at Chapel Hill (UNC) lawsuit includes both equal protection and Title VI claims while the Harvard lawsuit only has Title VI claims because unlike UNC, a public university, Harvard is a private university. See UNC Complaint, supra note 4, at 56–63; see also Harvard Complaint, supra note 4, at 100–15. As noted in the Harvard complaint, the need for proving discriminatory intent to prove a violation under the Title VI statute has been deemed identical to the equal protection standard. Harvard Complaint, supra note 4, at 102; see also Guardians Ass’n v. Civil Serv. Comm’n, 463 U.S. 582, 680 n.1 (1983) (Powell, J., concurring in judgment) (plurality opinion) (finding intentional discrimination necessary to violate Title VI. Proving a violation of Title VI regulations does not require the same proof of intent. Guardians, 463 U.S. at 584 (White, J.) (plurality opinion) (Justice White “conclud[ing], as [did] four other justices, in separate opinions, that the Court of Appeals erred in requiring proof of discriminatory intent” in a fractured opinion in which five justices concluded the lower court erred in requiring proof of intent and three justices explicitly found a disparate impact cause of action under Title VI regulations); see also Victor Suthammanont, Rebalancing the Scales: Restoring the Availability of Disparate Impact Causes of Action in Title VI Cases, 54 N.Y. L. Sch. L. Rev. 27, 32 n.40 (2009) (“In sum, the five justices would have found a disparate-impact cause of action under some section of Title VI, but disagreed as to the nature of the action and the remedies available.”).
. See supra Parts I and II for explanations of both of these fallacies.
. See, e.g., Stephan Thernstrom & Abigail Thernstrom, America in Black and White: One nation, Indivisible 400–01 (1997) (equating SAT math scores with admissions-related merit and asserting that African Americans are so dramatically underqualified to attend universities like California Institute of Technology and Massachusetts Institute of Technology that “Cal Tech might have needed to give some preference to black applicants to get even as many as 1.1 percent” to critique racial affirmative action as conferring large racial preferences that negatively affect Asian Americans). But cf. West-Faulcon, supra note 16, at 1266–68 (explaining the limitations in the predictive power of the SAT and the facts that the SAT leaves roughly 87 percent of the variation in first-year college grades unexplained and that the SAT has far less capacity to predict latter grades in college, college graduation, or success beyond college). However, critics including Stephan and Abigail Thernstrom have also compared “Asian students” to “black students” and described findings that Asian American students spend twice as much time doing homework than the average student, cut classes less often, pay more attention to the teacher, and are more eager to get good grades in contrast to “Blacks.” Thernstrom & Thernstrom, supra, at 383 Morevoer, they contend that African American students cut classes more, pay less attention, do less homework, and view doing well in school as “betraying one’s race,” but also believe that social success requires embracing “the values celebrated in rap music.” Id. Similarly, some have asserted that African American students suffer from a “cultural barrier to achievement” that needs to be overcome. Id.; cf. Matsuda, supra note 1, at 153–54 (writing on the use of Asian Americans as a wedge on the issue of racial affirmative action); Chin et al., supra note 28, at 161. (“We call upon conservatives to cease using APAs as their ‘racial mascot’ to arrogate moral authority in furtherance of regressive policies.”).
. See Blum, supra note 14. Like Blum, other modern critics of racial affirmative action, including Richard Kahlenberg, invoke Asian Americans this way by basing their critiques of affirmative action on its supposedly negative effect on Asian Americans. See, e.g., Sam Sanders, New Affirmative Action Cases Say Policies Hurt Asian-Americans, NPR (Nov. 20, 2014, 6:28 PM), http://www.npr.org/sections/codeswitch/2014/11/20/365547463/new-affirmative-action-cases-say-policies-hurt-asian-americans (quoting Richard Kahlenberg criticizing affirmative action as hurting Asian Americans). Kahlenberg has been described as a “paid consultant for the plaintiffs” in the Students for Fair Admissions, Inc. (SFFA) Harvard and UNC lawsuits. Id.
. See, e.g., Asian-Americans: The Model Minority Is Losing Patience, Economist (Oct. 3, 2015), http://www.economist.com/news/briefing/21669595-asian-americans-are-united-states-most-successful-minority-they-are-complaining-ever [https://perma.cc/EZ87-TKLZ]. (erroneously describing the SFFA Harvard and UNC lawsuits as “brought by the group of Asian students against Harvard and the University of North Carolina”). Rather, the race of each rejected applicant within the plaintiff organization is unstated except the specific descriptions of one Asian student in the Harvard complaint and one white student in the UNC complaint. See Harvard Complaint, supra note 4, at 8–9; UNC Complaint, supra note 4, at 8–10.
. Harvard Complaint, supra note 4, at 8–9.
. UNC Complaint, supra note 4, at 8–9 (asserting that the organization plaintiff, the Students for Fair Admissions, has at least one member who applied for and was denied admission to UNC for 2014 admission noting the “[a]pplicant is white”).
. See Harvard Complaint, supra note 4; UNC Complaint, supra note 4. Whereas lawsuits challenging affirmative action policies, historically, have had specifically identified, named lead plaintiffs, including Allan Bakke in the Bakke case, Jennifer Gratz and Barbara Grutter in the Gratz and Grutter cases, and, more recently, Abigail Fisher in the Fisher cases, no student is identified as a named plaintiff in either the Harvard or UNC cases; the only individual identified by name in the Harvard and UNC complaints is Edward Blum, Harvard Complaint, supra note 4, at 8; UNC Complaint, supra note 4, at 8, who is white and the self-described architect of the Harvard and UNC cases. See Blum, supra note 14. Blum also arranged the Fisher cases, see supra note 113, and the case that prompted the Supreme Court to overturn a key provision of the Voting Rights Act of 1965, see Biskupic, supra note 113, at 3 (describing Blum’s role in conceiving and orchestrating the filing and litigation of Shelby County v. Holder, 133 S. Ct. 2612 (2013)).
. 308 F.R.D. 39 (D. Mass.), aff’d, 807 F.3d 472 (1st Cir. 2015).
. Harvard Complaint, supra note 4, at 1, 43.
. UNC Complaint, supra note 4, at 8–9.
. 539 U.S. 306 (2003).
. Harvard Complaint, supra note 4, at 108, 115–16 (“‘Diversity’ is not an interest that could ever justify the use of racial preferences under the Fourteenth Amendment and Title VI. Even if there were a compelling government interest in ‘diversity’ in the abstract, however, the use of racial preferences in the educational setting nevertheless should be forbidden . . . .”). See Grutter, 539 U.S. at 323, for the Court’s discussion of the compelling interest of racial diversity in universities originally set forth in the opinion of Justice Powell in Regents of the Univ. of Cal. v. Bakke, 438 U.S. 265, 311–13 (1978) (reasoning that the First Amendment affords universities a unique academic freedom to create a diverse student body so long as race is one of multiple components of “diversity”).
. UNC Complaint, supra note 4, at 2.
. Id. at 17, 56 (“Only using race or ethnicity as a dominant factor in admissions decisions could account for the decision to admit certain African-American and Hispanic applicants and deny admission to certain white and Asian-Americans applicants.”).
. See Harvard Complaint, supra note 4, at 5, 111 (“Only using race or ethnicity as a dominant factor in admissions decisions could, for example, account for the remarkably low admission rate for high-achieving Asian-American applicants.”).
. UNC Complaint, supra note 4, at 55.
. Id. at 56. Many who continue to oppose racial affirmative action policies in litigation seek to challenge the notion that explicitly considering race accomplishes compelling diversity interests. See, e.g., Complaint at 8, Grutter v. Bollinger, 137 F. Supp. 2d 874 (E.D. Mich. 2001) (No. 97-75928), 1997 WL 34642450 (“[The university] had no compelling interest to justify their use of race in the admissions process, and were not motivated by . . . an interest in educational diversity . . . .”); see also, e.g., Complaint at 6, Gratz v. Bollinger, 353 F. Supp. 2d 929 (E.D. Mich. 2005) (No. 97-75231) (“[The university] had no compelling interest to justify their use of race in the admissions process, and were not motivated by . . . an interest in educational diversity . . . .”); Motion For Preliminary Injunction at 10 n.2, Fisher v. Univ. of Tex. at Austin, 645 F. Supp. 2d. 587 (W.D. Tex. 2009) (No. 1:08-cv-00263-SS), 2008 WL 7318505 (“Plaintiffs do not challenge UT Austin’s goal of ‘student body diversity’ only to the extent that it is this interest, and not some other, that serves as the university’s justification for re-introducing race into the undergraduate admissions process.”)
. See Harvard Complaint, supra note 4. Evidence of substantial differences in white and Asian American admission rates to Harvard and UNC in 2014 may very well surface during the course of litigation. My goal in this Article is to delineate the salience of comparing white and Asian American rates of selection under the admissions policy rejected Asian Americans challenge as imposing an Asian penalty. There are, in fact, a few references to rates of admission in the Harvard SFFA complaint, such as one reference to findings of lower Asian American admission rates from1979 through 1988, see id. at 37, a discussion of a working paper the complaint describes as finding that Asian Americans were being admitted at lower rates to “several Ivy League colleges” between 1994 and 2012, id. at 45–46, and a quote of a statement from a Harvard dean saying Asian American and white admission rates were lower than other groups in 1995. However, none of these sources include 2014 Harvard or UNC admission rate data. Moreover, they seem focused on comparing Asian American selection rates to African American and Latino rates of selection, not comparing Asian American and white rates. This emphasis may stem from the fact that SFFA seeks to convince the Supreme Court to overturn its endorsement of narrowly tailored consideration of race in admissions in the Grutter and Fisher II cases. See id. at 113, 2 (arguing that narrowly tailored affirmative action is not constitutionally permissible, unless used as a “last resort,” as it creates racial penalties for certain groups) (emphasis added) (quoting City of Richmond v. J.A. Croson Co., 488 U.S. 469, 518 (1989) (Kennedy, J., concurring in part and concurring in the judgment)). The UNC complaint includes a table showing 2006 UNC admit rates by race but otherwise fails to make any mention of how the white selection rate compares to the Asian American selection rate. See UNC Complaint, supra note 4, at 18–19, 18 tbl.A (presenting a table titled “Academic Index for All Admitted Students (2006)” and subsequently discussing racial group differences without noting differences between Asian and white applicants). Similarly, the complaint includes a table with average SAT scores of admitted students in 2012 disaggregated by racial groups; again the complaint fails to compare white and Asian American SAT score averages of admitted students. See id. at 19 (describing Asian American and white admitted students as “non-preferred students,” choosing to average the two groups’ high school GPAs and SAT score averages together, and proceeding to compare that average to the average of admitted underrepresented minorities).
. See Harvard Complaint, supra note 4, at 54–55, 54 tbl.B (comparing the percentage of enrolled students who are Asian American at Harvard to the percentage of enrolled Asian students at California Institute of Technology from 1991–2013 and also comparing Asian American enrollment at Harvard to higher percentages of Asian American enrollees at select universities and high schools). The complaint employs this comparison to illustrate that higher percentages of Asian American enrollees exist at schools with race-neutral admissions policies. Id. at 53. The complaint claims: “Harvard’s remarkably stable admissions and enrollment figures [of racial groups] over time are the deliberate result of systemwide intentional discrimination designed to achieve a predetermined racial balance of its student body.” Id. at 5.
. Id. at 11–43; see also Jerome Karabel, The Chosen: The Hidden History of Admission and Exclusion at Harvard, Yale, and Princeton 91 (2005).
. Instead of statistical evidence of racially disparate impact, the complaint sets forth numerous comments by a variety of institutions and individuals relying on personal experiences. This includes statements attributed to sometimes unidentified Harvard officials, high school directors of admission, and to admissions officers from universities other than Harvard. Harvard Complaint, supra note 4 at 56. The complaint also presents accounts from a Wall Street Journal article by Daniel Golden, which recounts experiences of Asian American students with high academic qualifications not admitted to Harvard and other Ivy League universities. Id. at 61–64, 65 (describing in detail high school students of Chinese and Korean descent who were rejected by Harvard as well as statements by a Chinese American doctoral student who was accepted by Princeton, noting that “it is a fact that Asians need higher academic achievements than their peers to get admitted to the same school”). The Harvard complaint also uses excerpts from materials produced by commercial test preparation and college consulting companies, which assert that Asian American students should take steps to differentiate themselves from other Asians and avoid fitting into Asian racial stereotypes. Id. at 57–60.
. In fact, Harvard has a history of limiting admissions of other nonwhite racial groups, including African Americans, see Karabel, supra note 137, at 23 (describing Harvard in the twentieth century as being “[a]lmost exclusively white [but] in some years . . . enroll[ing] a handful of blacks”). Likewise, Harvard rejects many thousands of African American applicants each year. For the Fall 2014 admissions cycle that is the subject of the SFFA lawsuits, Harvard rejected over 3000 African American students and UNC rejected over 2000 African American students. See Black First-Year Students at the Nation’s Leading Research Universities, J. Blacks Higher Educ. (Dec. 31, 2014), https://www.jbhe.com/2014/12/black-first-year-students-at-the-nations-leading-research-universities [https://perma.cc/55KP-AHAB] (reporting 3468 African American applicants but only 236 African American admissions to Harvard and 3274 African American applicants but only 880 African American admissions to UNC for Fall 2014).
. In this Article, I presume that the racial penalty on Asian American applicants claim is based on the view that the institutions consider the race of Asian American applicants in a manner that violates either the Fourteenth Amendment Equal Protection Clause, Title VI of the Civil Rights Act of 1964, or federal disparate impact regulations implementing Title VI.
. The legal standard for proving a violation of the Equal Protection Clause requires that equal protection plaintiffs prove both racially discriminatory effect, see, e.g., Palmer v. Thompson, 403 U.S. 217, 225 (1971), and racially discriminatory purpose, see, e.g., Pers. Adm’r of Mass. v. Feeney, 442 U.S. 256, 272 (1979); Washington v. Davis, 426 U.S. 229, 241 (1976) (imposing the purpose and effect requirements on equal protection plaintiffs).
. The legal standard for proving a violation under the Title VI statute is the same as the equal protection standard. See Guardians Ass’n v. Civil Serv. Comm’n, 463 U.S. 582, 584 (1983).
. Proving a violation of Title VI regulations does not require the same proof of intent. See id. Likewise, a civil rights plaintiff alleging a Title VI disparate impact violation is required to prove: 1) a federally funded university admitted members of its racial group at a rate that is lower than the admission rate of the racial group admitted at the highest rate; and 2) that the disparity between the difference in rates meets the judicially determined threshold that triggers the legal requirement that the university prove the admissions criteria causing the racial disparity in rates is an “educational necessity.” See, e.g., Sharif ex rel. Salahuddin v. N. Y. State Educ. Dep’t, 709 F. Supp. 345, 361 (S.D.N.Y. 1989) (explaining Title VI disparate impact legal standard). I filed such a claim on behalf of rejected Filipino (as well as African American and Latino) applicants who applied to UC Berkeley for admission in Fall 1998. See Complaint, supra note 27, at 1–3 (alleging rejection of Filipino American, African American, and Latino applicants by UC Berkeley violated the Equal Protection Clause, Title VI law for disparate treatment, and Title VI regulations for unjustified disparate impact); cf. Charles R. Lawrence III, Two Views of the River: A Critique of the Liberal Defense of Affirmative Action, 101 Colum. L. Rev. 928, 946–47 (2001) (observing that the Rios/Castaneda lawsuit “directly turns the upside down logic of ‘reverse discrimination’ right side up” and posits “a different view of what constitutes equality, a different remedy requested, and, ultimately, a different conception of justice”); Evelyn Nieves, Civil Rights Groups Suing Berkeley Over Admissions Policy, N.Y. Times, Feb. 3, 1999, at A9 (“Five civil rights organizations have sued the University of California at Berkeley on behalf of more than 750 black, Hispanic, and Filipino-American students for what they say are discriminatory policies . . . .”).
. See Harvard Complaint, supra note 4, at 44.
. See, e.g., UNC Complaint, supra note 4, at 18 & tbl.A (arguing that the data in Table A proves that race is a dispositive fact essentially guaranteeing admission of African American applicants); Harvard Complaint, supra note 4, at 71 (describing and comparing changes in percentages of African American and Asian American enrollment at Harvard between 2002 and 2012).
. Id. at 18 tbl.A. (showing very high selection rates of African American applicants with the highest possible UNC academic index score—between 3.0 and 3.299, at a higher rate than Asian Americans). Universities may have ample justification for admitting African American students who fall in this category at substantially higher rates than similarly situated Asian American and white applicants because the number of such African American students is comparatively small and because the “yield rate”—the percentage of the very high-scoring African American students who decide to enroll subsequent to their admission—is often significantly lower than similarly situated Asian American and white applicants. Cf. Liu, supra note 22, at 1076 (describing findings by Bowen and Bok that for selective universities the yield “tends to be lower for highly qualified black candidates than for comparable white candidates because the black candidates are likely to be admitted by more schools” (quoting Bowen & Bok, supra note 10, at 33–34)).
. UNC Complaint, supra note 4, at 18 tbl.A. Table A also shows that white UNC applicants with academic index scores ranging from 2.4 to 3.099—a span of seven index score ranges—were selected at higher rates than Asian Americans applicants with the same index scores. Id. The complaint did not include comparable data for the 2014 admissions cycle, the cycle challenged in the litigation. See id.
. See Grutter v. Bollinger, 539 U.S. 306, 311 (2003). The Supreme Court in Grutter upheld a public university’s admissions policy under strict scrutiny, which sought “to achieve that diversity which has the potential to enrich everyone’s education” “[b]y enrolling a ‘critical mass’ of underrepresented minority students” in order to “ensure their ability to make unique contributions to the character of the [educational institution].” See id. at 306 (original alterations omitted). The Court has described this limited use of race. See id. at 334 (“Universities can . . . consider race or ethnicity more flexibly as a ‘plus’ factor in the context of individualized consideration of each and every applicant.”); see also Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198 (2016) (Fisher II) (upholding the race-conscious component of UT Austin’s undergraduate admissions policy because the university proved it had a compelling interest in racial diversity and that the manner in which it considered race was sufficiently “narrowly tailored” to satisfy the legal standard of strict scrutiny).
. See Grutter, 539 U.S. at 327 (“When race-based action is necessary to further a compelling governmental interest, such action does not violate the constitutional guarantee of equal protection so long as the narrow-tailoring requirement is also satisfied.”).
. Chin et al., supra note 28, at 159 (explaining “neutral action” as Asian Americans being excluded from a university’s racial affirmative action policy but otherwise being treated indistinguishably from Whites in admission).
. Scholars have used the term “negative action” to describe what I describe as “white advantage.” Such discrimination has long been held to violate the Fourteenth Amendment Equal Protection Clause. See id; see also Yick Wo v. Hopkins, 118 U.S. 356, 373–74 (1886) (holding that applying a rule or requirement differently on the basis of race violates the Equal Protection Clause stating: “Though the law itself be fair on its face and impartial in appearance, [it is unconstitutional] if it is applied and administered by public authority with an evil eye and an unequal hand [on the basis of race.]”).
. Disparate impact analysis identifying and comparing Asian American rates of admission to white rates of admission is the type of statistical evidence that can be relied upon to prove a violation of Title VI disparate impact law. See Guardians Ass’n v. Civil Serv. Comm’n, 463 U.S. 582, 590 (1983). Likewise, a civil rights plaintiff alleging a Title VI disparate impact violation is required to prove a federally funded university employs a facially neutral admissions practice that “has a disproportionate effect” that negatively impacts Asian American applicants. See Sharif ex rel. Salahuddin v. N. Y. State Educ. Dep’t, 709 F. Supp. 345, 361 (S.D.N.Y. 1989) (explaining the Title VI disparate impact legal standard requires proof of a statistical disparity in selection rates that is lacking “educational necessity”).
. My point is that accusing Harvard of racial balancing is a promising means of convincing the federal court to strike down the institution’s racial affirmative action policy but, if Harvard admissions officials have an either conscious or unconscious enrollment limit they are inclined to impose on Asian American enrollment to preserve Harvard’s predominantly white character, eliminating racial affirmative action will neither expose or rectify that type of anti-Asian bias in admissions. See Matsuda, supra note 1, at 153–54.
. Espenshade & Radford, supra note 11, at 376 (“[W]e have looked for but have not found any feasible policy alternative to the current practice of race-sensitive admission that has the capacity to generate the same minority student representation on campus.”).
. Id. A shorter 2005 study by Professor Thomas Espenshade and Chang Chung, which contends that Asian Americans would be the “biggest winners” if universities eliminated racial affirmative action, Thomas J. Espenshade & Chang Y. Chung, The Opportunity Cost of Admission Preferences at Elite Universities, 86 Soc. Sci. Q. 293, 298 (2005), has been critiqued by William Kidder. Kidder, supra note 8, at 614. Similar to the focus of this Article, Kidder explains that the analysis by Espenshade and Chung fails to distinguish between “negative action” against Asian Pacific Americans, which is caused by unfair white advantage, and the effects on Asian Pacific Americans that result from racial affirmative action to increase inclusion of African Americans and Latinos. Id.
. See, e.g., Fisher II Brief of Asian American Legal Foundation, supra note 18, at 26–27 (“[A]n Asian American student has to score 140 points higher than a White student, 270 points higher than a Hispanic student, and 450 points higher than a Black student on the SAT to be on equal footing” (quoting Coal. of Asian-American Ass’ns, Complaint Against Harvard University and the President and Fellows of Harvard College for Discriminating Against Asian-American Applicants in the College Admissions Process 13 (2015), http://asianamericanforeducation.org/wp-content/uploads/2015/09/Aisan-Complaint-Harvard-Document-20150515-Final.pdf [https://perma.cc/3BTA-D8MB]) (citing Espenshade & Radford, supra note 11, at 92 tbl.3.5)); Fisher II Brief of California Ass’n of Scholars, supra note 18, at 29 (citing Espenshade & Radford, supra note 11); see also, e.g., Harvard Complaint, supra note 4, at 44–45 (citing Espenshade & Radford, supra note 11, at 92 tbl.3.5); Brief Amicus Curiae of Jonathan Zell in Support of Petitioner at 15, Fisher v. Univ. of Tex. at Austin, 136 S. Ct. 2198 (2016) (No. 14-981), 2015 WL 5244357, at *15 [hereinafter Fisher II Brief of Jonathan Zell] (citing Espenshade & Radford, supra note 11, at 92 tbl.3.5); Brief of Richard D. Kahlenberg as Amicus Curiae in Support of Neither Party at 6, Fisher II, 136 S. Ct. 2198 (No. 14-981), 2015 WL 5345843, at *6[hereinafter Fisher II Brief of Richard Kahlenberg] (citing Espenshade & Radford, supra note 11, at 92 tbl.3.5); Sander & Taylor, supra note 11, at 18 (citing Espenshade & Radford, supra note 11, at 92 tbl.3.5).
. Espenshade & Radford, supra note 11, at 93.
. Id. at 127. The summary of Chapter 3 of the Espenshade & Radford book makes unqualified assertions about African American and Latino applicants having “an admissions advantage” and “a boost” of a certain number of ACT and SAT test score points. Id.
. Id. at 93.
. Id. at 92 tbl.3.5. The authors assert:
An easier way to interpret the findings associated with race and social class in Table 3.4 is to convert them into more familiar terms. . . . Table 3.5 transforms the strength of admission preferences for members of different racial and socioeconomic groups in models 2 and 5 into their ACT-and SAT-point equivalents, using ACT points at public institutions and SAT point totals at private schools.
Id. at 91–92.
. Id. at 94. (“[I]t would be a mistake to interpret the data in Table 3.5 as meaning that elite college admission officers are necessarily giving extra weight to black and Hispanic candidates just because they belong to underrepresented minority groups.”)
. The Espenshade & Radford book is about far more than the role of race in college admissions. The lengthy, 409-page book considers the role that race and class play in college applications, admissions, enrollment, how students and families pay for college, college academic performance, the degree to which college students experience interaction with students of other races in curricular and extracurricular college activities, the rationales for and efficacy of racial affirmative action policies, and whether race-blind class-based affirmative action can produce the same level of racial diversity as racial affirmative action used in concert with socioeconomic affirmative action. See Espenshade & Radford, supra note 11, at 10–13. Notably, the Espenshade & Radford book includes findings contrary to the claim that racial affirmative action should be ended because class-based affirmative action makes it unnecessary. Id. at 355–61 (describing and then strongly rejecting “the eloquent case” for race-blind, class-based affirmative action). Richard Kahlenberg espouses this alternative affirmative action policy. See generally Fisher II Brief of Kahlenberg, supra note 156; Richard D. Kahlenberg, The Remedy: Class, Race, and Affirmative Action (1st ed. 1996). Espenshade and Radford describe their conclusion in the following manner: “In this exhaustive examination of a wide variety of potential admissions policies, we have looked for but have not found any feasible policy alternative to the current practice of race-sensitive admission that has the capacity to generate the same minority student representation on campus.” Espenshade & Radford, supra note 11, at 376.
. See supra Table 1.
. In Part III, I found no statistical evidence to support racial discrimination against Asian Americans by UT Austin. See supra Part I.
. This is not actual admission data from any one particular real-world university and thus cannot be relied upon in an actual lawsuit because the Espenshade and Radford admission rate data is a compilation of many decades-old admissions datasets from eight different universities. See infra note 169 (explaining that the Espenshade & Radford book’s data is from the 1983, 1993, and 1997 admissions cycles at eight different universities).
. The Fisher amici fail to note that Harvard University is not one of the colleges and universities that provided data for the book or that, although UNC is among the potential list of thirty-five institutions, it is unknown whether UNC is actually one of the colleges or universities in the NSCE database used in the book’s analysis. Espenshade & Radford, supra note 11, at 411 n.1. Similarly, neither the UT Austin nor Harvard were among the schools invited to volunteer data for the Espenshade book. Id. Yet, neither the anti-affirmative-action Fisher amici nor the SFFA Harvard complaint acknowledge this fact. See Fisher II Brief of Asian American Legal Foundation, supra note 18; Fisher II Brief of California Ass’n of Scholars, supra note 18, at 29; Fisher II Brief of Jonathan Zell, supra note 156, at 15; Fisher II Brief of Richard Kahlenberg, supra note 156, at 6, 7, 26; see also Harvard Complaint, supra note 4, at 44 (describing Espenshade & Radford’s data as originating from “a group of three elite public and four elite private colleges” without clarifying Harvard was not included).
. Espenshade & Radford, supra note 11, at 10. Espenshade and Radford note that data from two of the ten universities was not included in their analysis because the institutions were two historically black colleges and universities (HBCUs). Id. at 413–14. There was also data supplied by the anonymous participating universities. Id. The most recent of the admissions cycles analyzed in the Espenshade & Radford book took place almost two decades ago in 1997, and the oldest admissions data dates back to 1983. See id. at 10. This is significant in evaluating the value of the Espenshade & Radford book’s findings in Table 3.5 in lawsuits challenging Harvard and UNC 2014 admissions. Since these long ago admissions cycles, there has been substantial change in selective university admissions, based, in part, on the Supreme Court’s 2003 ruling in Gratz v. Bollinger, which cautions that holistic admissions policies are more likely to survive strict scrutiny analysis than numerically-driven admissions policies. Gratz v. Bollinger, 539 U.S. 244, 270–71 (2003).
. The Fisher amicus brief submitted to the Supreme Court by the Asian American Legal Foundation (AALF) and a group called the Asian American Coalition for Education (AACE) describes a complaint letter the latter group filed with the OCR as highlighting “compelling evidence that Harvard and other elite colleges discriminate against undergraduate Asian American applicants to maintain their informal quotas.” Fisher II Brief of Asian American Legal Foundation, supra note 18, at 25. The AALF brief says that the Espenshade & Radford book found: “Asian American applicants have 67% lower odds of admission than white applicants with comparable tests scores.” Id. at 26 (citing Coal. of Asian-American Ass’ns, supra note 156, at 12–13, and Espenshade & Radford, supra note 11). The complaint also notes Espenshade and Radford found that “when applying to top private universities an Asian-American student has to score 140 points higher than a White student, 270 points higher than a Hispanic student and 450 points higher than a Black student on the SAT to be on equal footing.” Id. (quoting Coal. of Asian-American Ass’ns, supra note 156, at 13) (citing Espenshade & Radford, supra note 11, at 94).
. Espenshade & Radford, supra note 11, at 93.
. 136 S. Ct. 2198 (2016) (Fisher II). The amici include, for example, Fisher II Brief of Asian American Legal Foundation, supra note 18 at 26; Fisher II Brief of California Ass’n of Scholars, supra note 18, at 29; Fisher II Brief of Jonathan Zell, supra note 156, at 15; and Fisher II Brief of Richard Kahlenberg, supra note 156, at 6 n.12. Other critics of affirmative action have cited prior research by Espenshade. See, e.g., Fisher I Sander & Taylor Brief, supra note 40, at 3 (“[I]n fact the racial preferences used by the University of Texas, and those used by most flagship state universities, elite colleges, and graduate professional schools are very large indeed” without acknowledging that that the study did not analyze any data from UT Austin (citing Thomas J. Espenshade et al., Admissions Preferences for Minority Students, Athletes, and Legacies at Elite Universities, 85 Soc. Sci. Q. 1422 (2004)).
. Harvard Complaint, supra note 4, at 44. (asserting that publicly available data on Harvard’s admissions constitutes “statistical evidence [which] establishes that Harvard is intentionally discriminating against Asian Americans by making it far more difficult for Asian Americans than for any other racial and ethnic group of students to gain admission to Harvard”).
. Espenshade & Radford, supra note 11, at 92 n.29 (emphasis added) (explaining that the process of first converting the odds ratios in Table 3.4 “back into logistic regression coefficients” and then interpolating the coefficients on race and class into the ACT scale for public institutions and the SAT scale for private institutions magnified the size of the racial effects the authors purport to have found).
. Id. at 92–93 n. 29.
. Id. at 94 n.32.
. Id. at 94.
. Id. at 94–95.
. Id. at 94 (“[I]f we were able to include these other considerations in our models, we believe the effect of being black or Hispanic per se would be diminished.”). In a footnote on the same page, Espenshade and Radford explain that the database they use for their regression analysis “does not include all of the information that admission officers presumably evaluate when making admissions decisions. Omitted, for example, are letters of recommendation, students’ personal statements, lists of extracurricular activities and other talents, and work experience.” Id. at 94 n.32.
. Id. at 95.
. See id.
. See supra Table 1.
. See infra Table 4.
. Espenshade & Radford, supra note 11, at 73, fig. 3.3.
. Admission rates in Figure 3 are taken from Table 3.3 of the Espenshade & Radford book. Id. at 80 tbl.3.3. The total numbers of applicants to public and private institutions are taken from Table 3.2. Id. at 71 tbl.3.2. I calculated the number of white, Asian American, Latino, and African American applicants to public and private institutions by multiplying total number of public university applicants reported in Table 3.2 of the Espenshade & Radford book—27,171—by the proportion reported in the book’s Figure 3.3. See id. at 73 fig.3.3. For example, the number of white applicants—23,258—reported in Figure 3 was calculated by multiplying .856 (85.6 percent) by 27,171 (the total number of public university applicants).
. Admission rates in Figure 4 are taken from Table 3.3 of the Espenshade & Radford book. Id. at 80 tbl.3.3. Total numbers of applicants to public and private institutions are taken from Table 3.2. Id. at 71 tbl.3.2. I calculated the number of white, Asian American, Latino, and African American applicants to public and private institutions by multiplying total number of private university applicants reported in Table 3.2 of the Espenshade & Radford book—52,051—by the proportion reported in the book’s Figure 3.3. See id. at 73 fig.3.3. For example, the number of white applicants—28,992—reported in Figure 4 of this Article was calculated by multiplying .557 (55.7 percent) by 52,051 (the total number of private university applicants).
. Admission rates in Table 2 are taken from Table 3.3 of the Espenshade & Radford book. Espenshade & Radford, supra note 11, at 80 tbl.3.3. Total numbers of applicants to public and private institutions are taken from Table 3.2. Id. at 71 tbl.3.2. The numbers of white, Asian American, Latino, and African American applicants to public and private institutions were calculated based on percentages in Figure 3.3 and numbers of admitted white, Asian American, Latino, and African American students were calculated by applying admission rates in Table 2 to admit numbers calculated. See id. at 80 tbl.3.3.
. Admission rates in Table 2 are taken from Table 3.3 of the Espenshade & Radford book. Id. at 80 tbl.3.3. Total numbers of applicants to public and private institutions are taken from Table 3.2. Id. at 71 tbl.3.2. The numbers of white, Asian American, Latino, and African American applicants to public and private institutions were calculated based on percentages in Figure 3.3 and numbers of admitted white, Asian American, Latino, and African American students were calculated by applying admission rates in Table 2 to admit numbers calculated. See id. at 80 tbl.3.3.
. The higher admission rate for these racial groups would be legally justified, because the racial affirmative action policy is a legitimate nondiscriminatory rationale for the higher African American and Latino admission rates. See Johnson v. Transp. Agency, 480 U.S. 616, 626 (1987).
. See infra Tables 2 and 3; Figures 5 and 6.
. See infra Tables 2 and 3; Figures 5 and 6.
. See Fisher v. Univ. of Tex. at Austin (Fisher II), 136 S. Ct. 2198, 2216, 2227 n.4 (2016) (Alito, J., dissenting).
. See, e.g., Fisher II Brief of the Asian American Legal Foundation, supra note 18;Fisher II Brief of California Ass’n of Scholars, supra note 18; Fisher II Brief of Pacific Legal Foundation, supra note 18.
. Harvard Complaint, supra note 4, at 54, 67 (comparing admissions statistics, including the percentage of enrolled and admitted students who are Asian American for various admissions cycles, that do not constitute evidence of discriminatory effect on Asian American applicants—a racial disparity in rates of selection—in 2014 admissions); UNC Complaint, supra note 4, at 18-19 (doing the same because providing only comparison of selection rates is for 2006 UNC admissions).
. See Grutter v. Bollinger, 539 U.S. 306, 343 (2003).
. See Johnson v. Transportation Agency, 480 U.S. 616 (1987), for comparable analysis under Title VII employment discrimination law. In Johnson, the Supreme Court held that a legally valid gender-based affirmative action policy is a “nondiscriminatory rationale” for considering race in hiring under Title VII law. Id. at 626 (explaining that the existence of an affirmative action plan provides “a nondiscriminatory rationale” for the decision not to hire a rejected applicant).
. See supra Part III for a discussion of Title VI disparate impact law.
. See supra Table 1.
. Table 4 was created from admission rate data taken from the Espenshade & Radford’s book. See Espenshade & Radford, supra note 11, at 128.
. See supra Table 1 (explaining four-fifths rule and chi square anlaysis).
. The basic principle underlying the Title VI theory of effect discrimination is that individuals of all races should enjoy equal access to federally funded institutions and services. U.S. Comm’n on Civil Rights, Civil Rights Under Federal Programs: An Analysis of Title VI of the Civil Rights Act of 1964 4–5 (1968), http://www.law.umaryland.edu/marshall/usccr/documents/cr1101968.pdf [https://perma.cc/FK6U-RES5]. U.S. Department of Education regulations promulgated to enforce Title VI of the Civil Rights Act of 1964 state that a recipient of federal funds may not “utilize criteria or methods of administration which have the effect of subjecting individuals to discrimination because of their race, color, or national origin.” 45 C.F.R. § 80.3 (2016). If the university cannot justify the use of a selection criterion that has a racially discriminatory impact—a criterion that causes admission rates to diverge from racial parity—the U.S. Department of Justice may find that institution to be in violation of Title VI disparate impact regulations.
. See West-Faulcon, supra note 102, at 1097–98 (explaining that a legally significant racial disparity in admission rates to a particular university constitutes prima facie evidence of Title VI racially disparate impact against the racial group selected at the lower rate). Title VI disparate impact regulations require a university to justify the disparity and, even if the university can defend it, rejected Asian American applicants could prove a Title VI disparate impact violation if they can present a less discriminatory admissions practice as an alternative. Id.
. Considering race for the purpose of inclusion is factually distinct from considering race for the purpose of exclusion. See Justice Breyer’s dissent in Parents Involved for a similar observation:
The upshot is that the cases to which the plurality refers, though all applying strict scrutiny, do not treat exclusive and inclusive uses the same. Rather, they apply the strict scrutiny test in a manner that is ‘fatal in fact’ only to racial classifications that harmfully exclude; they apply the test in a manner that is not fatal to racial classifications that seek to include.
Parents Involved in Cmty. Sch. v. Seattle Sch. Dist. No. 1, 551 U.S. 701, 833 (2007) (Breyer, J., dissenting).