The United States Supreme Court
The University of Michigan Admissions Lawsuits

Gratz v. Bollinger and Grutter v. Bollinger

Supporting the University, the American Psychological Association as amicus curiae (pp 4-23): Racial diversity in higher education helps to combat unconscious racial prejudice and stereotypes that lead to discrimination. U-M's Gurin Report provides solid evidence of the educational benefits of racial diversity. (The complete brief is online here.)

Comments by Curtis Crawford,* indented in italics.


A. Discrimination And Prejudice Persist
In Significant And Demonstrable Ways,
Even In Many Who Believe That They
Are Not Prejudiced And Do Not Discriminate

1. Discrimination. *Although the Nation has made remarkable strides in eliminating overt discrimination against racial and ethnic minorities, numerous studies over the past twenty years demonstrate that discrimination persists in almost all aspects of American life.* Many of these studies are summarized in Faye J. Crosby, Affirmative [-5-] Action Is Dead; Long Live Affirmative Action (Yale Univ. Press) (forthcoming) at 225-228 ("Crosby"). 3 In addition, the brief amicus curiae in this case of the American Sociological Association et. al., to which we respectfully refer the Court, reviews at some length the large body of research that demonstrates that racial and ethnic minorities continue to have unequal access to jobs, housing, and education.


* Amicus strongly believes that "discrimination against racial and ethnic minorities" is wrong. Does it equally condemn discrimination in their favor?

2. Prejudice. Whereas the term "discrimination" describes unequal treatment, "prejudice" has to do with thoughts and feelings. See Crosby, at 20. Research "shows a dramatic decrease in the proportion of White Americans who express overtly hostile feelings or overtly derogating thoughts about people of color." Crosby, at 229. 4 That has led some scholars to discount racism as an explanation for the serious inequalities that beset racial and ethnic minorities. 5 But in fact a series of recent studies show that prejudice is tenacious and pervasive even among those who [-6-] maintain explicit attitudes of equality. Thus, many people who firmly believe that they have open and favorable attitudes about people of various races and ethnicities will demonstrate that they implicitly (unconsciously) harbor a variety of racial and ethnic prejudices that can translate into subtle discriminatory behaviors.

 One might assume that racial and ethnic prejudices are widespread in all racial and ethnic groups. Does amicus share this view, or believe instead that racial prejudice is wholly or primarily a white failing?

These unconscious prejudices have been described as "aversive racism" Ė so-called because those who harbor prejudice would find it aversive to recognize their own racism. "* * * [R]acial bias is expressed in indirect ways that do not threaten the aversive racistís non-prejudiced self-image. * * * [D]iscrimination occurs when [that] bias is not obvious or can be rationalized on the basis of some factor other than race." John F. Dovidio & Samuel L. Gaertner et al., Aversive Racism and Selection Decisions: 1989 and 1999, 11 Psychol. Sci. 315 (2000).

Aversive racism has been demonstrated by (among others) a series of carefully designed experiments among young, liberal Whites. In 1989, the researchers asked White college students to evaluate candidates for a position of resident counselor. One third of the candidates were highly qualified, one third were distinctly unqualified, and the middle third were strong candidates but not unambiguously so. The students regularly endorsed the candidates who were clearly well qualified, whether the candidates were White or Black.6 Similarly, the students rejected the unqualified candidates equally without regard to race. But when the candidates were of intermediate quality, White students recommended hiring the White candidate significantly more often than the Black candidate and felt more strongly about [-7-] their recommendations for the Whites than for the Blacks. Ten years later the researchers repeated the experiment with a new crop of students and got exactly the same results. During those ten years, the students showed a noticeable decrease in the direct verbal expression of racial prejudice Ė but their actual behavior remained unchanged. 7 One reason for the existence and persistence of prejudice is straightforward. We need categories to absorb and understand our world. Prejudice and stereotypes are the natural by-products of ordinary perception, categorization, learning, memory, and judgment. There is, accordingly, widespread agreement among social scientists that social categorization processes Ė including racial and ethnic stereotyping Ė are virtually automatic, operating at perhaps a more basic level and often independent of conscious attitudes, beliefs, and perceptions. 8

 The explanation of "prejudice and stereotypes" in this paragraph suggests that they are characteristic faults of homo sapiens, rather than of particular races or ethnicities.

Social scientists have devised a number of tests for measuring the strength of unconscious stereotyping in different individuals. One, widely used, is the Implicit [-8-] Association Test (IAT). 9 It is a reaction time experiment, measuring the speed with which two concepts can be associated. Not surprisingly, people quickly complete the task when "insects" and "bad" are paired together, and respond to the pairing of "insects" and "good" much more slowly. Because reaction times are mapped in milliseconds, the associative process is automatic and therefore not subject to the kind of conscious control that otherwise makes it difficult to know whether experiment participants are producing results according to consciously desired outcomes.

The results of the research on associative processes conclusively demonstrate that unconscious stereotyping and prejudice, including race stereotyping, is widespread. 10 To take but one example, test respondents consistently make more ready associations between, on the one hand, faces of African Americans and words having negative concepts (e.g., bomb, devil, awful) and, on the other, faces of Whites and positive concepts (e.g. peace, joy, love). Hundreds of thousands of individuals have taken the test producing similar results. [-9-]

3. Unconscious Stereotyping And Biased Behavior. Although implicit prejudices operate in the unconscious mind, they can have significant real-world effects. Recent empirical research indicates that automatic prejudice plays an important role in producing discriminatory behavior and judgments 11 and that measures of implicit prejudice are significant predictors of the level of discriminatory behaviors and judgments.12 For example, people high in implicit prejudice are unfriendly towards African American interaction partners,13 and form negative and stereotypical impressions of minorities.14

 In the Brief thus far, all the research examples of derogatory racial prejudice have been of attitudes unfavorable to blacks or to other people of color. This suggests that amicus may view racial prejudice as primarily a white failing.

[-10-] Research shows that implicit prejudice and stereotypes become exaggerated for token minorities in groups. Psychologists have identified a number of negative processes associated with token representation, one of which is increased visibility and the likelihood that tokens are treated in terms of their group identity rather than personal qualities.15 As a consequence, people with token status are frequently "encapsulated" in roles stereotypic of their group in employment settings.16 Moreover, people who feel that they are tokens of their group experience negative feelings of [-11-] distinctiveness,17 associated with sensitivity to being stereotyped, and a loss of self-confidence,18 which can interfere with cognitive performance.19

Feelings of distinctiveness can also invoke "stereotype threat" whereby students behave in ways that conform to the negative cultural stereotype of their group, even when it has an adverse effect on cognitive performance.20

These findings demonstrate that implicit prejudices can be *a major factor in limiting both the opportunities of members of minority groups* and the effectiveness of culturally diverse work groups.21

 * Why would "implicit prejudices" operate to limit "the opportunities" only "of minority groups"? Does amicus believe that whites are not also the victims of implicit prejudices and of the discrimination that flows from them?

[-12-] B. Diversity In Higher Education Reduces Unconscious Stereotyping
And Promotes Productive Intergroup Relations

Automatic biases cannot be eliminated merely by conscious efforts to develop racially tolerant attitudes. Rather, there is "abundant evidence that stereotypes that operate unconsciously defend their territory fiercely, influencing social interactions even when perceivers are consciously vigilant and motivated to defeat them."22

But it does not follow that unconscious biases cannot be combated and defeated. Rather, one promising strategy for attacking unconscious biases is to "create the social conditions that allow new associations and new learning about social groups that blur the bright line that demarcates [-13-] social groups."23 *After all, if stereotypes represent our classification strategy for making sense of a complex world, then they are necessarily shaped by the world we are classifying.24 If we meaningfully change that world Ė "the larger societal context that controls these phenomena"25 Ė we can thereby change individualsí classification strategies and thus reduce both unconscious bias and its behavioral effects.* In keeping with these premises, considerable research has been undertaken recently into the malleability of unconscious stereotyping. It proves, as Justices of the Bakke Court understood and as our intuitive instincts might predict, that positive exposure to members of other racial groups helps reduce such biases.26 That research is consistent with, [-14-] and reinforces, the long-standing finding that *positive contact between people of different races and ethnicities "typically reduces intergroup prejudice."27* For example, **in one recent study White students who interacted with a Black experimenter showed significantly reduced automatic race prejudice, compared to interaction with a White experimenter.28**


* True, although note that the "contact between people of different races and ethnicities" needs to be "positive." The Brief's theory of the causes and cure of racial prejudice, if extended to nonwhites as well as whites, seems eminently sensible. Moreover, less racial prejudice should mean less racial discrimination, an indispensable national goal.

** Still another example from the Brief in which those who manifest racial prejudice are white. Do psychologists not investigate the formation and extent of racial and ethnic prejudice in nonwhites?

Indeed, a thorough review of the field suggests that "face to face interaction . . . is importantly related to reduced prejudice."29 The beneficial effects of such contact with other groups emerge across many prejudice measures, outgroup targets, and contact settings.

These recent findings give even more credence to the testimony of leading educators and psychologists advanced by the University of Michigan in the District Courts below, which demonstrated that attending college in a diverse setting reduces stereotyping and results in "more diverse friends, neighbors, and work associates." Expert Reports, 115. Moreover, the research we have described helps to explain why simply taking diversity workshops without minorities participating or reading great works by minority leaders (which are alternative approaches recommended by [-15-] opponents of race conscious admissions policies) will be ineffective in removing prejudice and stereotypes and allowing the kind of deep and complex multi-cultural thinking that the University wishes to foster.

In sum, extensive recent psychological research demonstrates that unconscious racial and ethnic stereotyping and prejudice are pervasive and persistent in our society. Other research shows, however, that these implicit attitudes and responses can be ameliorated when students from diverse racial and ethnic backgrounds live and work with each other intensively, both in and out of the classroom. We recognize that "societal discrimination" has been ruled an inadequate basis for raced-based classifications designed to provide "raced-based relief." City of Richmond v. J.A. Croson Co., 488 U.S. 469 (1989). But we believe that, putting aside the consideration of remedies for past wrongs, diversity may constitute a compelling governmental interest because it is critical to promoting future harmonious and productive relations among different racial and ethnic groups as they must increasingly interact in a shrinking global community. Accordingly, *the research results we have described above strongly support the conclusion of the Sixth Circuit below that the government has a compelling interest in creating a racially diverse student body in higher education.*


* Not quite. The "conclusion of the Sixth Circuit below" did not contain the word, racially. As announced by Chief Judge Martin, it read: "Because this court is bound by Justice Powell's Bakke opinion, we find that the Law School has a compelling state interest in achieving a diverse student body." (Opinion, at end of Part A, Sec. 1, Grutter v. Bollinger, 2002 FED App. 0170P (6th Cir.)(online here.) Powell's famous opinion in Bakke v. Board of Regents had held diversity of thought and viewpoint, not of race, as a compelling educational interest.

However, it has been much disputed whether Powell's view was ever adopted by the Court, and whether it should now be ratified, revised or rejected. In Gratz and in Grutter, the Court could hold that the creation of racial diversity in higher education is indeed a compelling governmental interest, justifying racial preference in admissions. Do "the research results" described by amicus "strongly support" such a holding?

Not at all. From the fact that positive interracial activity can reduce racial prejudice and discrimination, it does not follow logically or morally that government or its agencies should practice racial discrimination in order to make such activity more available. The colleges and universities, and the state and federal governments, are powerful teachers. When they practice racial discrimination, they teach that it is right, and always there are people ready to learn and apply this lesson.

In the 1950s and '60s, most Northern and Western states, and then the federal government, adopted civil-rights laws barring discrimination based on race or ethnicity in education, employment, public accommodations, governmental services, housing, etc. Typical was the language of Title VI, ß 601, of the Civil Rights Act of 1964, concerning educational programs such as those at U-M: "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." These laws clarified and taught an American ideal implied in the Declaration of Independence and the 14th and 15th Amendments to the U.S. Constitution. Their adoption reflected and reinforced the willingness of the majority in the nation's most numerous and powerful racial and ethnic groups to limit radically the areas of permissible racial discrimination.

In the late 1960s, Michigan, like many Northern and Western schools, enrolled substantial numbers of black students via affirmative-action programs that conferred a strong preference based on race. These policies were also effective teachers, and student activists were quick to learn. According to the Ann Arbor News, "In 1970, the newly formed Black Action Movement demanded that blacks account for at least 10 percent of the undergraduate population. BAM backed its demands with picketing, demonstrations and eventually a strike that shut down much of the university. In the spring of 1970, BAM presented a list of demands to the administration: That blacks be 10 percent of undergraduate students, that the number of black faculty be increased, that a tutoring program for minority students be expanded, that an activity center for black students be created along with an academic center for African-American studies." (Ann Arbor News, 3/23/03)

Notice that all five demands imply racial preference and/or racial segregation. Throughout the country, whether initiated by administrators or faculty, or forced by students, racial preference in admissions has spread to racial discrimination in: faculty hiring, student financial and tutorial assistance, student organizations, living quarters, courses taught, readings assigned, and departments established.

This is a lot of racial discrimination. Amicus offers no evidence that the decrease in racial discrimination from greater student racial diversity is greater than, equal to, or anywhere near, the increase of discrimination that results from race-conscious admissions.


The District Court in Gratz concluded, based upon the record before it, that "a racially and ethnically diverse student body produces significant educational benefits such [-16-] that diversity, in the context of higher education, constitutes a compelling governmental interest under strict scrutiny." Gratz v. Bollinger, 122 F. Supp. 2d 811, 824 (E.D. Mich. 2000), cert. granted, 123 S. Ct. 602 (2002). In reaching that conclusion, the court relied in part on research undertaken by Dr. Patricia Y. Gurin, professor of psychology and women's studies at the University of Michigan.30 The court quoted with approval the reportís conclusion that "students learn better in a diverse educational environment, and they are better prepared to become active participants in our pluralistic, democratic society once they leave such a setting." Id. at 822.

Dr. Gurinís work analyzed three databases. The first contained initial, 4-year, and 9-year national data involving 184 colleges and universities from the Cooperative Institutional Research Program conducted by UCLAís Higher Education Research Institute. The other two provided a more in-depth look at diversity experiences at the University of Michigan. One was the Michigan Student Survey, conducted prior to these lawsuits and involving initial and 4 year data from the 1990 entering undergraduate class. The other contained pre-participation, class completion, and 4 year data from a study of those in the 1990 entering class who took an introductory course in the Intergroup Relations, Conflict and Community Program at the University. By examining these three databases, Dr. Gurin was able to take both a "macro" and "micro" look at [-17-] how diversity experiences in higher education are related to educational and citizenship outcomes.31

Dr. Gurinís analysis of the three databases found consistent, statistically-significant results. In particular, she found that opportunities for enrollment in ethnic studies courses and classes dealing with issues of race and ethnicity, and for informal interactions with racial and ethnic minorities, including both friendships and more general interactions, were directly linked to such favorable outcomes as less automatic, improved active thinking; more active engagement in the learning process; increased ability to understand the perspectives of others; and greater motivation and engagement in society. Gurin Report at 146.

Amici supporting petitioners have attacked Dr. Gurinís research with more vigor than accurate scientific criticism. But as we now show, the critiques themselves depend on numerous flawed assumptions and methods that render them unpersuasive.

First, *the Gurin Report has been wrongly criticized for not comparing the levels of diversity in student bodies at various universities ("structural diversity") directly with educational outcomes.* Brief for Amicus Curiae National Association of Scholars ("NS") in Support of Petitioners in Gratz v. Bollinger, No. 02-516, at 7. But Dr. Gurinís research was not focused on structural diversity by itself. What her research shows is that structural diversity is one [-18-] factor making possible campus diversity experiences that are associated with learning outcomes.32 A campus could be full of minority students yet still have a segregated environment without meaningful interactions between different racial and ethnic groups. Dr. Gurin finds that interactions between groups are needed to produce educational benefits, and posits that a university must foster such interactions.33 **In Dr. Gurinís words, "Structural diversity is essential but, by itself, usually not sufficient to produce substantial benefits; in addition to being together on the same campus, students from diverse backgrounds must also learn about each other in the courses they take and in informal interaction outside of the classroom."** Id. at 6. 34


* If the facts had shown a significant correlation between the amount of student-body racial diversity at various campuses and educational benefits to students, would Dr. Gurin have included that in her report? Absolutely! It would have been important evidence for her thesis. Her problem was that the facts in UCLA's Cooperative Institutional Research Program, a data base she used for other purposes, showed no such correlation.

** Gurin is right that student-body racial diversity, though essential for the educational benefits it provides, will not provide them if not put to good use. The same is true of a college library. However, both are factors whose educational contribution can be compared through statistical analysis. Controlling for other factors, one can determine whether there is a correlation between the quality of the library (or student-body racial diversity) and the quality of student learning.

Second, National Association of Scholars attacks the Gurin Report on the ground that, for the vast majority of her measures, she reported that even extremely large increases in her diversity measures account for less than 1 percent of student learning outcomes. NS Gratz Br. at 18. Labeling these measures as showing either no effects or "only a miniscule effect on students," NS argues that the Gurin report [-19-] is irrelevant to the case. Id. at 18-19. But the small effects reported by Dr. Gurin are due, in part, to her separate consideration of each individual experience or outcome in order to show "a remarkable consistency of results." Gurin Supp. Rep. at 8. The effect size for any single measure will often be small because a measure of an individual experience or outcome will typically have low reliability; and reliability will set limits on effect size. However, replicating findings across several measures suggests a consistent and stable set of results unlikely to be due to chance.

While Dr. Gurinís approach is an acceptable research strategy, psychologists often aggregate individual measures into indices when multiple measures examining a common theme are studied (e.g., combining five measures dealing with citizenship behaviors into an overall index of citizenship). Such indices or aggregations are commonly used because they increase reliability,35 thus minimize the impact of measurement error or "noise" for any one measure. When Dr. Gurinís individual measures are aggregated, subsequent findings using these indices show stronger relationships than are observed with the individual measures, thus refuting the claim that the effects in her research are trivial. These indices provide further evidence of the statistically significant relationship between both classroom/informal interactions and learning/democracy outcomes. Gurin Supp. Rep. at Table 1.

The NS critique also fails to recognize that although effect size is widely used in social and behavioral research as a measure of the strength of a relationship, "a large effect size is not the only way to demonstrate that an effect is [-20-] important."36 The key point is that the importance of the findings depends upon the context in which they are made. Small effect sizes may have enormous impact over time.37 For example, today, physicians routinely advise men over a certain age to take an aspirin each day to prevent heart attacks. In the original study that led to this advice, the effect size between taking an aspirin and having a heart attack was .0011 -- which resulted in almost twice as many heart attacks in the placebo group as in the aspirin group.38 The effect was considered so important that the study was discontinued early because it was determined that aspirin could no longer ethically be withheld from the placebo-control group.39

[-21-] Moreover, in higher education, involving some 15 million college students plus additional graduate students every year, even small benefits can yield substantial rewards.

Third, National Association of Scholars erroneously accuses Dr. Gurin of having replaced the accepted standard for categorizing a finding as statistically significant (5%) with one that allowed her to exaggerate the importance and size of her results. NS Gratz Br. at 19-20. Dr. Gurin did use the 5% level of significance for her analyses of the large samples of White students. But due to the small numbers of students in the minority samples, she chose a different level (10%). Gurin Report at 115. That decision reflects a methodologically sound balancing of concerns for "Type I" error (finding significance when the result is in actuality by chance) and "Type II" error (finding no significance when in fact there is a significant relationship). Dr. Gurinís decision to use a less strict level of significance for her small sample was an eminently reasonable research decision. See, e.g., Jacob Cohen, Statistical Power Analysis for the Behavioral Sciences (1988).

Fourth, the National Association of Scholars asserts that Dr. Gurinís findings are flawed because they rely upon self-reports instead of more objective outcome measurements. NS Gratz Br. at 16. *But in fact use of self-report measures is common practice in social science research that involves large data bases.* Moreover, a focus on what National Association of Scholars terms "traditional measurements of academic outcome," such as "grades, graduation rates, admissions to graduate school, and performance on seven standardized tests," NS Gratz Br. at [-22-] 13, takes an unrealistically narrow view of academic achievement and fails to capture a broad range of outcomes widely valued by educational institutions, such as, for example, continuous learning, intellectual interest and curiosity, multicultural tolerance and appreciation, interpersonal skills, and social responsibility, citizenship, and involvement.40

 * Does amicus mean to suggest that colleges should substitute self-reports of academic ability and achievement for grades and test scores as the bases of admission, evaluation and graduation? If objective measures are important in determining whether students are learning what colleges have to teach, why are they not important also in determining whether students are learning from campus racial diversity?

Fifth, the National Association of Scholars faults Dr. Gurinís research on the ground that various classroom diversity experiences she finds relevant to her conclusions (such as enrollment and participation in ethnic studies courses and attending racial/cultural awareness workshops) could be equally beneficial to White students without the "presence on campus of any students of another race." NS Gratz Br. at 10. But it should be obvious that, in the real world, discussing racial and ethnic issues in all-White settings is bound to have a different impact on the students than if racial and ethnic minorities were also present and participating. Indeed, even National Association of Scholars concedes that "informal interaction" such as having close friends of another race, interracial socializing, and sharing personal feelings and problems with members of other races, "presuppose[s] the presence of some students of another race on campus." NS Gratz Br. at 10.

In sum, the methodology of the Gurin Report is consistent with scientific standards for research. Moreover, [-23-] Dr. Gurinís research confirms numerous studies brought to the Courtís attention by other amici and respondents which demonstrate that diversity experiences in college and graduate school are associated with a variety of important educational objectives, such as the ability to think in "deeper, more complex ways" (Gurin Report at 100); to develop "more positive views of conflict, and the perception that diversity is not inevitably divisive in our society" (id. at 127); and to become more motivated and competent to live and work in a complex and diverse society" (id. at 101, 127).


3. We are prepared to lodge a copy of the manuscript with the Clerk of the Court.
[return to text]

4. In 1948, for example, 63% of White respondents objected to having a Black neighbor, and 41% objected to having Mexican or Mexican American neighbors. By 1987, the objectors had dropped to 13% and 9%. John F. Dovidio et al., Cognitive and Motivational Bases of Bias: Implications of Aversive Racism for Attitudes toward Hispanics, in Hispanics In The Workplace 75, 90-91 (Stephen B. Knouse et al. eds., 1992). Support for integrated schooling grew from 32% in 1942 to 90% in 1983. Howard Schuman et al., Racial Attitudes in America: Trends and Interpretations (1988). [return to text]

5. See, e.g., Stephen Thernstrom & Abigail Thernstrom, America In Black and White: One Nation, Indivisible (1997). [return to text]

6. Throughout this brief "Black" and "African-American" are used interchangeably. [return to text]

7. John F. Dovidio & Samuel L. Gaertner et al., Aversive Racism and Selection Decisions: 1989 and 1999, 11 Psychol. Sci. 315 (2000). Dovidio and Gaertner are not the only experimental social psychologists to document the persistence of racial preferences. See, e.g., Monica Biernat & Theresa K. Biernat, Categorization and Stereotyping: Effects of Group Context on Memory and Social Judgment, 29 J. Experimental Soc. Psychol. 166 (1993). [return to text]

8. See, e.g., Steven J. Spencer et. al., Automatic Activation of Stereotypes: The Role of Self-Image Threat, 24 Personality & Soc. Psychol. Bull. 1139 (Nov. 1998); J. Blascovich et al., Racism and Racial Categorization, 72 J. Personality & Soc. Psychol. 1364 (June 1997). [return to text]

9. Anthony G. Greenwald et. al., Measuring Individual Differences in Implicit Cognition: The Implicit Association Test, 74 J. Personality & Soc. Psychol. 1464 (1998). [return to text]

10. See, e.g., Patricia G. Devine, Stereotypes and Prejudice: Their Automatic and Controlled Components, 56 J. Personality & Soc. Psychol. 5 (Jan. 1989); Steven J. Spencer et. al., Automatic Activation of Stereotypes: The Role of Self-Image Threat, 24 Personality & Soc. Psychol. Bull. 1139 (Nov. 1998). See also, John F. Dovidio et al., On the Nature of Prejudice: Automatic and Controlled Processes, 33 J. Experimental Soc. Psychol. 510 (1997). [return to text]

11. Russell H. Fazio & Michael A. Olson, Implicit Measures in Social Cognition Research: Their Meaning and Use, 54 Ann. Rev. Psychol. 297, 306, 310 (2003). [return to text]

12. See, e.g., Allen R. McConnell & Jill M. Liebold, Relations Among the Implicit Association Test, Discriminatory Behavior, and Explicit Measures of Racial Attitudes, 37 J. Experimental Soc. Psychol. 435 (2001); William von Hippel et al., The Linguistic Intergroup Bias as an Implicit Indicator of Prejudice, 33 J. Experimental Soc. Psychol. 490 (1997); Denise Sekaquaptewa et al., Stereotypic Explanatory Bias: Implicit Stereotyping as a Predictor of Discrimination, 39 J. Experimental Soc. Psychol. 75 (2003). [return to text]

13. John F. Dovidio et al., Implicit and Explicit Prejudice and Interracial Interactions, 82 J. Personality & Soc. Psychol. 62 (Jan. 2002); John F. Dovidio et al., On the Nature of Prejudice: Automatic and Controlled Processes, 33 J. Experimental Soc. Psychol. 510 (1997); Russell H. Fazio et al., Variability in Automatic Activation as an Unobtrusive Measure of Racial Attitudes: A Bona Fide Pipeline?, 69 J. Personality & Soc. Psychol. 1013 (Dec. 1995); Allen R. McConnell & Jill M. Liebold, Relations among the Implicit Association Test, Discriminatory Behavior, and Explicit Measures of Racial Attitudes, 37 J. Experimental Soc. Psychol. 435 (2001). [return to text]

14. See Robert C. Dunton & Russell H. Fazio, An Individual Difference Measure of Motivation to Control Prejudiced Reactions, 23 Personality Soc. Psychol. Bull. 316 (1997); Florak et al., When Do Associations Matter? The Use of Automatic Associations Toward Ethnic Groups in Person Judgments, 37 J. Experimental Soc. Psychol. 518 (2001); William von Hippel et al., The Linguistic Intergroup Bias As an Implicit Indicator of Prejudice, 33 J. Experimental Soc. Psychol. 490 (1997); Russell H. Fazio & Laura E. Hilden, Emotional Reactions to a Seemingly Prejudiced Response: The Role of Automatically Activated Racial Attitudes and Motivation to Control Prejudiced Reactions, 27 Personality and Soc. Psychol. Bull. 538 (May 2001). [return to text]

15. Shelly E. Taylor et al., Categorical and Contextual Bases of Person Memory and Stereotyping, 36 J. Personality and Soc. Psychol. 778 (July 1978). [return to text]

16. E.g., women in nurturing roles; see J. Yoder, Looking Beyond Numbers: The Effects of Gender Status, Job Prestige, and Occupational Gender-Typing on Tokenism Processes, 57 Soc. Psychol. Q. 150 (1994). [return to text]

17.Yolanda F. Niemann & John F. Dovidio, Tenure, Race/Ethnicity and Attitudes Toward Affirmative Action: A Matter of Self-Interest?, 41 Sociological Perspectives 783 (1998). [return to text]

18. Laurie L. Cohen and Janet K. Swim, The Differential Impact of Gender Ratios on Women and Men: Tokenism, Self-Confidence, and Expectations, 21 Personality & Soc. Psychol. Bull. 876 (Sept. 1995). [return to text]

19. Charles G. Lord & Delia S. Saenz, Memory Deficits and Memory Surfeits: Differential Cognitive Consequences of Tokenism for Tokens and Observers, 49 J. Personality & Soc. Psychol. 918 (Oct. 1985). [return to text]

20. See generally Claude M. Steele, A Threat in the Air: How Stereotypes Shape Intellectual Identity and performance, 52 Am. Psychol. 613 (June 1997). [return to text]

21. Decades of discrimination have left ethnic minority students disproportionately condemned to be educated in under-funded or impoverished schools where they have less opportunity than students at affluent schools to demonstrate that they are capable of performing well in challenging courses. Jonathan Kozol, Savage Inequities: Children in Americaís Schools (1991). "[B]ecause racism targets and injures individuals based on their group membership, its antidote must also target the group if its individual members are to overcome barriers to equal opportunity." Michele A. Wittig, Taking Affirmative Action in Education and Employment, 52 J. Soc. Issues 145, 157 (1996). [return to text]

22. Mahzarin R. Banaji & Anthony G. Greenwald, Implicit Gender Stereotyping in Judgments of Fame, 68 J. Personality & Soc. Psychol. 181, 197 (1995); Irene V. Blair, & Mahzarin R. Banaji, Automatic and Controlled Processes in Stereotype Priming, 70 J. Personality & Soc. Psychol. 1142 (1996); Anthony G. Greenwald & Mahzarin R. Banaji, Implicit Social Cognition: Attitudes, Self-Esteem, and Stereotypes, 102 Psychol. Review 4 (1995). [return to text]

23. Mahzarin R. Banaji et al., The Social Unconscious, in Blackwell Handbook of Social Psychology (A. Tesser & N. Schwartz eds., 2001). [return to text]

24. Shelly E. Taylor et al., Categorical and Contextual Bases of Person Memory and Stereotyping, 36 J. Personality and Soc. Psychol. 778 (July 1978); Charles G. Lord & Delia S. Saenz, Memory Deficits and Memory Surfeits: Differential Cognitive Consequences of Tokenism for Tokens and Observers, 49 J. Personality & Soc. Psychol. 918 (Oct. 1985). Indeed, there is "clear evidence that automatic responses to category cues depend on the surrounding context." Irene V. Blair, The Malleability of Automatic Stereotypes and Prejudice, 6 Personality & Soc. Psychol. Rev. 242 (2002). [return to text]

25. Susan T. Fiske, Stereotyping, Prejudice, and Discrimination, in 2 The Handbook of Soc. Psychol. (D. Gilbert et. al. eds., 1998). [return to text]

26. Irene V. Blair, The Malleability of Automatic Stereotypes and Prejudice, 6 Personality & Soc. Psychol. Rev. 242 (2002); see Regents of the University of California v. Bakke, 438 U.S. 265, 314 (1978) (Opinion of Powell, J.) [return to text]

27. Thomas Pettigrew & Linda R. Tropp, Does Intergroup Contact Reduce Prejudice? Recent Meta-Analytic Findings, in Reducing Prejudice and Discrimination 93 (Stuart Oskamp ed., 2000). [return to text]

28. Brian S. Lowery et. al., Social Influence Effects On Automatic Racial Prejudice, 81 J. Personality & Soc. Psychol. 842 (Nov. 2001) [return to text]

29. Pettigrew & Linda R. Tropp, Does Intergroup Contact Reduce Prejudice? Recent Meta-Analytic Findings, in Reducing Prejudice and Discrimination 93, 109 (Stuart Oskamp ed., 2000). [return to text]

30. Expert Report of Patricia Y. Gurin, Gratz v. Bollinger, No. 97- 75321 (E.D. Mich.) & Grutter v. Bollinger, No. 97-75928 (E.D. Mich.), in The Compelling Need for Diversity in Higher Education 99, 100 (1999) ("Gurin Report"). [return to text]

31. In all three sets of analyses, to the extent possible, Dr. Gurin statistically accounted for or "controlled" for other variables which were "known or expected to contribute" to the outcomes in order to account for these variableís possible effects so as not to overestimate the effects. Gurin Rpt. App C. at 1. [return to text]

32. Evidence for the Educational Benefits of Diversity in Higher Education: Response to the Wood & Sherman Critique by the National Association of Scholars of the Expert Witness Report of Patricia Gurin in Gratz v. Bollinger and Grutter v. Bollinger, by Patricia Gurin, May 30, 2001, at http://www.umich.edu/~urel /admissions/research/ ("Gurin Supp. Report"). [return to text]

33. For further insight into why this may be so, see the discussion of the malleability of implicit stereotypes at 12-13 supra. [return to text]

34. See also, Patricia Gurin et. al., Diversity and Higher Education: Theory and Impact on Educational Outcomes, 72 Harv. Educ. Rev. 330, 333 (2002) ("[S]tructural diversity is a necessary but insufficient condition for maximal educational benefits * * *."). [return to text]

35. Seymour Epstein, Aggregation and Beyond: Some Basic Issues on the Prediction of Behavior, 51 J. Personality 360 (1983). [return to text]

36. Deborah A. Prentice & Dale T. Miller, When Small Effects Are Impressive, 112 Psychol. Bull. 160 (July 1992). [return to text]

37. See, e.g., Douglas G. Mook, In Defense of External Invalidity, 38 Am. Psychol. 379 (Apr. 1983); Robert P. Abelson, A Variance Explanation Paradox: When a Little is a Lot, 97 Psychol. Bull. 129 (Jan. 1985); Robert Rosenthal & Donald B. Rubin, A Note on Percent Variance Explained as A Measure of the Importance of Effects, 9 J. Applied Soc. Psychol. 395 (1979). [return to text]

38. Robert Rosenthal, How Are We Doing in Soft Psychology?, 45 Am. Psychol. 775 (June 1990). [return to text]

39. Steering Committee of the Physicianís Health Study Research Group, Preliminary Report: Findings From The Aspirin Component Of The Ongoing Physicians' Health Study (Special Report), 318 New Eng. J. Med. 262 (Jan. 1988). See also, Deborah M. Barnes, Promising Results Halt Trials of Anti-AIDS Drug AZT, 234 Science 15 (Oct. 1986). (The clinical trials of the drug AZT, now widely prescribed for people with AIDS, were cancelled because it was felt that the benefits of AZT (an effect size of .073) were too great to deny it to the patients in the control condition); Robert Rosenthal, Methodology, in Advanced Soc. Psychol. 17, 35 (Abraham Tesser ed., 1995) (The effect size in the original trials of Cyclosporine, a drug that reduces the likelihood of organ rejection, was .0225). [return to text]

40. Such outcome criteria can be identified through a web search of colleges and universities; from university resources; and from related research. See Paul R. Sackett et al., High-Stakes Testing in Employment, Credentialing, and Higher Education: Prospects in a Post-Affirmative-Action World, 56 Am. Psychol. 302 (Apr. 2001).[return to text]