Opinion of the Court
example, courts can ask whether temporary racial segregation of inmates will prevent harm to those in the prison. See Johnson, 543 U. S., at 512–513. When it comes to workplace discrimination, courts can ask whether a race-based benefit makes members of the discriminated class “whole for [the] injuries [they] suffered.” Franks v. Bowman Transp. Co., 424 U. S. 747, 763 (1976) (internal quotation marks omitted). And in school segregation cases, courts can determine whether any race-based remedial action produces a distribution of students “compar[able] to what it would have been in the absence of such constitutional violations.” Dayton Bd. of Ed. v. Brinkman, 433 U. S. 406, 420 (1977).
Nothing like that is possible when it comes to evaluating the interests respondents assert here. Unlike discerning whether a prisoner will be injured or whether an employee should receive backpay, the question whether a particular mix of minority students produces “engaged and productive citizens,” sufficiently “enhance[s] appreciation, respect, and empathy,” or effectively “train[s] future leaders” is standardless. 567 F. Supp. 3d, at 656; 980 F. 3d, at 173–174. The interests that respondents seek, though plainly worthy, are inescapably imponderable.
Second, respondents’ admissions programs fail to articulate a meaningful connection between the means they employ and the goals they pursue. To achieve the educational benefits of diversity, UNC works to avoid the underrepresentation of minority groups, 567 F. Supp. 3d, at 591–592, and n. 7, while Harvard likewise “guard[s] against inadvertent drop-offs in representation” of certain minority groups from year to year, Brief for Respondent in No. 20–1199, at 16. To accomplish both of those goals, in turn, the universities measure the racial composition of their classes using the following categories: (1) Asian; (2) Native Hawaiian or Pacific Islander; (3) Hispanic; (4) White; (5) African-American; and (6) Native American. See, e.g., 397
