The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Do the Ethnic Categories Used by Universities for "Diversity" Purposes Make Sense?

No, argues an amicus brief filed on my behalf in the pending Supreme Court affirmative action litigation.


Some highlights of the brief, authored by Cory Liu:

Harvard and UNC's race-conscious admissions policies divide applicants into the following categories for purposes of determining eligibility for race-based
advantages in the admissions process: (1) Asian; (2) Native Hawaiian or Pacific Islander; (3) Hispanic; (4) White; (5) African American; and (6) Native American.

As Professor David Bernstein has shown, these racial and ethnic categories were created in the mid-1970s by federal bureaucrats whose only goal was to unify the racial and ethnic categories federal agencies used for recordkeeping. David E. Bernstein, The Modern American Law of Race, 94 S. CAL. L. REV. 171, 197–200 (2021); see also DAVID E. BERNSTEIN, CLASSIFIED: THE UNTOLD STORY OF RACIAL CLASSIFICATION IN AMERICA (forthcoming 2022). The categories came about in a haphazard manner without any input from anthropologists, sociologists, ethnologists, or other experts.

The bureaucrats who created the categories expressly warned that they "should not be interpreted as being scientific or anthropological in nature, nor should they be viewed as determinants for eligibility for participation in any Federal program." 43 Fed. Reg. 19,260, 19,269 (May 4, 1978).

There was never even a hint in the development of the categories that they were established for achieving educationally beneficial diversity in higher education. See Hugh Davis Graham, The Origins of Official Minority Designation, in THE NEW RACE QUESTION: HOW THE CENSUS COUNTS MULTIRACIAL INDIVIDUALS 289 (Joel Perlmann & Mary C. Waters eds., 2002)….

Harvard and UNC's racial and ethnic categories match the categories adopted by federal agencies, including the Department of Education. Whatever value the categories may have in allowing for consistency in data collection, they lump together members of very diverse groups into arbitrary categories. As Michael Omi and Howard Winant, two of the leading sociologists of race in the United States, point out: "These racial categories are rife with inconsistencies and lack parallel construction. Only one category is specifically racial, only one is cultural, and only one relies on a notion of affiliation or community recognition." MICHAEL OMI & HOWARD WINANT, RACIAL FORMATION IN THE UNITED STATES 122 (3d ed. 2015); see also PETER H. SCHUCK, DIVERSITY IN AMERICA: KEEPING GOVERNMENT AT A SAFE DISTANCE 164 (2003) (describing the racial categories as "almost comically arbitrary").

Harvard and UNC cannot explain why they use these particular racial and ethnic categories in their admissions policies….

Given the unduly broad nature of the "Asian" category, it is no surprise that only a minority of people assigned to that category identify as "Asian" or "Asian American." See JANELLE WONG ET AL., ASIAN AMERICAN POLITICAL PARTICIPATION: EMERGING CONSTITUENTS AND THEIR POLITICAL IDENTITIES 162 (2011) (finding that less than 40% of Indian, Chinese, and Filipino respondents identified as "Asian" or "Asian-American," even as a secondary identity)….

The question of who counts as "Hispanic" has continually befuddled federal and state authorities. See, e.g., Marinelli Constr. Corp. v. State, 613 N.Y.S.2d 1000, 1002 (N.Y. App. Div. 1994) (denying Hispanic status to a person of Italian-Argentine descent); Major Concrete Constr., Inc. v. Erie Cty., 521 N.Y.S.2d 959, 960 (NY. App. Div. 1987) (denying Hispanic status to a person with one Mexican grandparent); In re Rothschild-Lynn Legal & Fin. Servs., SBA No. MSBE94-10-13-46, 1995 WL 542398, at *3–4 (Apr. 12, 1995) (granting Hispanic status to a Sephardic Jew whose ancestors had fled Spain centuries earlier); In re DCS Elecs., Inc., SBA No. MSBE-91-10-4-26, 1992 WL 558961, at *4 (May 8, 1992) (recounting agency's conclusion that someone with "blond hair and light skin" was not Hispanic); In re Kist Corp., 99 F.C.C.2d 201, 216–17, 248 (1983) (granting partial minority credit for Hispanic status to a person with one Cuban grandparent); In re Storer Broad. Co., 87 F.C.C.2d 190, 191–93 (1981) (accepting Sephardic Jewish heritage as evidence of Hispanic status); In re Lone Cypress Radio Assocs., Inc., 7 FCC Rcd. 4403, 1992 WL 690184, at *5 (1992) (concluding that while being one-fourth Hispanic is enough to classify someone as Hispanic, being one-eighth Hispanic is not); Participation by Disadvantaged Business Enterprise in Department of Transportation Programs, 62 Fed. Reg. 29,548, 29,550 (May 30, 1997) (reaffirming Department of Transportation decision to classify "persons of European Spanish and Portuguese origin" as Hispanic, even though the latter group is not of Spanish origin or culture)….

There is a tremendous amount of ethnic, cultural, linguistic, and religious diversity within the category of people that Harvard and UNC classify as white…. The category includes, among others, Welsh, Norwegians, Greeks, Moroccans, Chaldeans, Afghans, Iranians, and North African Berbers. To place people descended from all these groups into one category is inconsistent with the goal of achieving genuine educational diversity….

Neither Harvard nor UNC has explained why a white Catholic of Spanish descent, classified as Hispanic, gets an admissions preference for contributing to educational diversity, but a dark-skinned Muslim of Arab descent, an Egyptian Copt, a Hungarian Roma, a Bosnian refugee, a Scandinavian Laplander, a Siberian Tatar, or a Bobover Hasid—all classified as "white"—do not. Similarly, it is hard to see how diversity is better accomplished by admitting an additional "Hispanic" student of Mexican ancestry over an equally or better qualified student whose parents immigrated from Turkmenistan, who would be the only Turkman in the entire student body, because the Turkman is arbitrarily classified as "white."…

A descendant of American slaves who grew up in a working-class, majority-black neighborhood in Milwaukee does not contribute to diversity in the same way as a child of an African diplomat, nor as a black-identified applicant with multiracial ancestry who grew up in an overwhelmingly white small town in Montana…. Yet they all fall into the same diversity category at Harvard and UNC….

Similarly, the experiences of a Navajo Indian who grew up on the tribe's reservation in Arizona are quite different from those of a person with one-sixty-fourth Cherokee ancestry and a European surname whose appearance and life are indistinguishable from his "white" neighbors' except that he has inherited tribal membership….

The racial and ethnic categories that Harvard, UNC, and universities across the country use in their admissions policies were created by executive-branch bureaucrats who specifically warned that they were not scientific or anthropological in nature and should not be used to determine eligibility for benefits in race conscious policies. The categories are imprecise, over and underinclusive, and are not narrowly tailored to achieve educationally beneficial diversity.

Note that the brief does not address the issue of whether a different version of affirmative action for "diversity" purposes could past constitutional muster,  nor whether affirmative action programs based on values other than "diversity" may be constitutional. The brief only addresses the notion that basing affirmative action justified by educational diversity on categories created for entirely different purposes and not meant to reflect any sociological or anthropological reality cannot pass strict scrutiny.