Under the Equal Protection Clause of the Fourteenth Amendment, "[n]o State shall … deny to any person within its jurisdiction the equal protection of the laws." The same equal protection obligation applies to the federal government through the Fifth Amendment. In general, the rule governs claims that the government is improperly treating individuals or groups differently. In most contexts, the government only has to show that distinctions it draws in law or policy are rational, rather than arbitrary. However, the Supreme Court has held that classifications based on race call for enhanced safeguards, known as "strict scrutiny," under the Equal Protection Clause. This In Focus outlines that analysis.
When Strict Scrutiny Applies
When a statute, regulation, or other government action distributes burdens or benefits based on race, ethnicity, or national origin, courts will impose a rigorous, "strict scrutiny" test to decide whether it violates constitutional equal protection principles. (While courts apply strict scrutiny in other contexts, including to decide whether content-based restrictions on speech comport with the First Amendment, this In Focus limits its discussion to racial classifications under the Equal Protection Clause.) To pass the strict scrutiny test, a law must be narrowly tailored to serve a compelling government interest.
The same test applies whether the racial classification aims to benefit or harm a racial group. Strict scrutiny also applies whether or not race is the only criteria used to classify. For example, if a grant program prioritizes three applicant groups: veterans, people with disabilities, and members of a minority racial group, the racial preference triggers strict scrutiny, even though it is not the only preference. Similarly, if race is a determinative factor in deciding who gets a benefit, the benefits program must pass strict scrutiny, even if nonracial factors also play a part. Benefits for federally recognized Indian tribes present a special case; such measures may hinge on a political status—tribal membership—rather than race.
Equal protection principles limit only intentional race-based actions. Classifications that have an unintentional effect on a racial group (sometimes called a disparate impact) are not subject to strict scrutiny. A government benefit offered based on income or home ownership, for example, would not face strict scrutiny even if members of a particular racial group less often qualify for the benefit. Race-based classifications affording no benefits or burdens will not be subject to strict scrutiny. Collecting racial demographic data, for example, rarely implicates equal protection.
When strict scrutiny applies, the government has the burden of proving both a compelling interest and narrow tailoring, and neither is easy to do.
A Compelling Government Interest and a Strong Basis in Evidence
Typically, racial classifications in legislation seek to remedy past discrimination. This goal can qualify as a compelling government interest. In practice, courts have viewed this interest as more compelling when the aim is correcting past government discrimination, or government participation in discrimination. The Supreme Court has stated that remedying general, "societal discrimination" is not a sufficiently compelling interest to satisfy strict scrutiny. Students for Fair Admissions, Inc. v. President and Fellows of Harvard College, 600 U.S. _ (2023). For several decades, the Court also recognized achieving a diverse student body as a compelling government interest, permitting some use of race in higher education admissions. In 2023, the Court held that two schools' race-based admissions programs relying on this interest violated equal protection, although the Court observed that schools may consider a student's individual life experience with race "be it through discrimination, inspiration, or otherwise." Id.
When the government aims to remedy discrimination, it must prove that there was in fact discrimination to establish a compelling government interest. In such cases, the Supreme Court has required that there be a strong basis in evidence—that is, an extensive and specific record in support. City of Richmond v. J.A. Croson Co., 488 U.S. 469 (1989). The evidence must align with the challenged remedial program. If a state government gives a construction contracting preference to Black and Asian-owned businesses, for example, the government must present evidence of discrimination against Black-owned construction businesses and Asian-owned construction businesses in that state's construction industry.
What types of evidence are courts looking for? When it comes to legislative action, courts have considered legislative findings, reports to Congress, testimony, and floor speeches—in short, the whole legislative record. In contrast, statements made after a law's enactment, and evidence not before legislators, cannot generally be used to show lawmakers' remedial aims.
Statistical evidence also may be important. Courts have sometimes cited agency data, congressional studies, or academic research included in the legislative record. Hearings and expert testimony might introduce this statistical evidence and illustrate its significance. Courts sometimes treat anecdotal evidence as relevant, particularly in showing that a statistical disparity is likely the result of discrimination rather than benign factors. However, courts have not viewed conclusory assertions by legislators as establishing a strong basis in evidence.
It is hard to say how much evidence is needed; courts evaluate each record in context. To justify a nationwide program, lawmakers need to show discrimination of national significance—rather than isolated instances—although state-by-state evidence is not required.
Narrowly Tailored Remedies
If a court finds a firm basis in evidence to support a compelling government interest in taking race-based action, it will look to see if the action is narrowly tailored to address that interest. On the whole, the court will consider whether the racial distinctions are necessary and whether they are over-inclusive or under-inclusive. The Supreme Court's plurality opinion in United States v. Paradise, 480 U.S. 149 (1987), identifies several relevant factors: "the necessity for the relief and the efficacy of alternative remedies; the flexibility and duration of the relief, including the availability of waiver provisions; the relationship of the numerical goals to the relevant labor market; and the impact of the relief on the rights of third parties."
First, to decide whether the race-based government action is necessary, a court may look for race-neutral options. Race-based government action might be more likely to survive strict scrutiny if race-neutral options have failed. A state employer might expand outreach and recruiting to minority applicants, for example, before implementing a race-based hiring goal. If these attempts do not sufficiently achieve the compelling interest, later race-based policies are easier to justify.
Next, courts may look to see if a racial classification is too broad. As part of the analysis, a court may consider the program's flexibility. Race-based quotas represent the least flexible options and are disfavored. Flexible measures include ways to opt into or out of the preference. For example, a preference for minority firms in construction subcontracting might exclude wealthier firms (which have made more progress in overcoming discriminatory barriers) to focus on smaller, under-resourced firms.
Adding waiver provisions may also help. If a government sets a minority hiring goal, for example, a waiver option might excuse noncompliance if hiring officials show good-faith efforts in outreach and recruiting, or show a lack of qualified minority candidates. How long the relief program lasts is also important. A preference with a sunset provision or a reauthorization requirement is more narrowly tailored than one of indefinite length, as it is less likely to last past the time when discrimination's effects have receded.
While courts disfavor quotas, they may allow targets or goals, provided they reflect relevant data. In evaluating a goal or preference, courts may consider the relevant market, or applicant pool. Courts may refer to the government's evidence supporting the compelling interest, reflecting demonstrated disparities in minority participation. A narrow plan for requiring prime contractors to hire minority subcontractors, for example, would tie hiring goals to availability of qualified minority subcontractors in each industry and region, rather than on statewide demographics.
Finally, a narrowly tailored remedy should minimize harm to third parties such as nonminority firms, applicants, or recipients. In competition for limited benefits, a court may find that racial preferences impose unjustified harms. In general, a racial classification is more problematic under this factor if it affects third parties' vested interests in some way, leaves third parties worse off than they were before (e.g., layoffs), or is unavailable to nonminorities.
Although there are several ways to tailor a remedy, it can be hard to predict judicial outcomes. Decisionmakers may use any or all of the Paradise factors. As a whole, the Paradise factors help courts assess whether a race-conscious remedy is under-inclusive or over-inclusive. Sometimes, however, courts consider this more directly, looking at whether race-based policies benefit those who have not suffered discrimination (i.e., are over-inclusive) or fail to benefit those who have (i.e., are under-inclusive). A narrowly tailored remedy avoids both.
Considerations for Congress
Applying precedent in this area can be difficult. Cases are few, as laws that use race are rare. Many of the Supreme Court's cases on the subject have produced splintered opinions with no clear majority rule. Historically, Supreme Court precedent on racial classifications comes almost entirely from three contexts—contracting, hiring, and higher-education admissions; and the Supreme Court has recently expressed disapproval of affirmative action in higher education. Extrapolating from the limited cases to assess the constitutional vulnerability of other kinds of race-based actions is difficult, given that equal protection analysis is context-specific. There are no bright-line rules.
Thus, if legislators undertake race-based actions, the more comprehensive the legislative record, the better. Remedying past discrimination is the most well-established government interest supporting race-based legislation, and a record supporting this interest generally includes detailed findings of discrimination and strong supporting evidence. Once a record is built, a more limited remedy more easily passes strict scrutiny. Sunset provisions, reauthorization requirements, race-neutral criteria, and waiver provisions can help. If legislation includes a numerical goal, it can be tailored to reflect available data. Finally, measures to minimize harms to third parties can reduce a statute's vulnerability to equal protection challenges under strict scrutiny.
As strict scrutiny is demanding, legislators may consider using nonracial classifications. These could be measures of health, education, income, access to resources (e.g., hospitals, transportation, or grocery stores), or proximity to hazards (e.g., pollution, underperforming schools, or high crime).