Since the 1970s, federal civil rights statutes—and many state analogues—have imposed liability for ‘disparate impact’: when facially neutral policies disproportionately harm protected groups. Liability can attach even without evidence of conscious discriminatory intent, though employers are not liable where a policy’s adverse effect is justified by business necessity.
On April 23, 2025, President Donald Trump issued Executive Order 14281, titled Restoring Equality of Opportunity and Meritocracy, seeking “to eliminate the use of disparate-impact [discrimination] liability in all contexts to the maximum degree possible.” This executive order—along with its accompanying fact sheet and the rescission of 2022 guidance—marks the latest chapter in a longstanding debate over the meaning of a fundamental term in antidiscrimination statutes and regulations: “equal opportunity.”
But prohibiting disparate-impact discrimination is “equal opportunity” according to federal legislation and decades of judicial precedent, neither of which an Executive Order can override. EO 14281 may remove disparate-impact compliance as a condition of federal funding and contracting, but unless Congress or the courts say otherwise, employers remain bound by the longstanding disparate impact framework embedded in federal law.
Background: What is Disparate Impact Liability?
Disparate impact liability is a well-established legal doctrine in federal legislation and case law that unlawful discrimination can occur without conscious intent, prohibiting employment and other policies that are facially neutral but disproportionately harm individuals in protected classes—such as by race, sex, age, or disability—without a sufficient business justification. In contrast to the “disparate treatment” theory (which prohibits similarly situated individuals differently based on a protected characteristic like race or gender), plaintiffs need not show proof of discriminatory intent to establish liability. The Supreme Court, Congress, and the executive branch have long recognized that intent-based standards alone are inadequate to combat discrimination rooted in unconscious bias, structural disadvantage, and proxy variables (e.g., a facially neutral screening factor that tracks closely with race or gender).
Disparate impact was first recognized by the U.S. Supreme Court in Griggs v. Duke Power Co., 401 U.S. 424 (1971), where the Court struck down a high school diploma and aptitude test requirement that disproportionately excluded Black applicants and could not be justified as job-related. With the Civil Rights Act of 1991, Congress codified this standard into Title VII at 42 U.S.C. § 2000e–2(k), requiring employers to prove that any policy or practice with a disparate impact is “job-related and consistent with business necessity.”
While the precise contours of disparate-impact liability vary depending on the context, it is a critical aspect of antidiscrimination laws governing housing (under the Fair Housing Act), education, credit, federal contracting, and increasingly, in the use of algorithmic decision-making. Some state civil rights laws governing employment and housing include disparate-impact provisions, such as New York’s Human Rights Law, Illinois’s Human Rights Act, and California’s Fair Employment and Housing Act. At the federal level, a private right of action exists in the workplace context (under Title VII) and in the fair-housing context (as upheld in Texas Department of Housing and Comunity Affairs v. Inclusive Communities Project, 576 U.S. 519 (2015)), but private disparate impact suits under Title VI are no longer cognizable under Alexander v. Sandoval, 532 U.S. 275 (2001).
Notwithstanding EO 14281’s characterization that “disparate-impact liability all but requires individuals and businesses to consider race and engage in racial balancing to avoid potentially crippling legal liability,” the law does not require employers to engage in intentional discrimination to avoid disparate impact. To the contrary, the Supreme Court made clear in Ricci v. DeStefano, 557 U.S. 557 (2009) that Title VII prohibits race-based employment decisions unless the employer has a “strong basis in evidence” that it would otherwise be subject to disparate impact liability. The Court held that an employer’s decision to discard firefighter promotion exam results—out of concern that the results would have a disparate impact on Black candidates—constituted intentional discrimination against the successful white and Hispanic candidates. This case reinforced that disparate impact concerns must be grounded in rigorous evidence, and that employers cannot use demographic outcomes as a blanket justification to disregard individual test results or qualifications.
Ricci and its extensive progeny demonstrate that the law draws a distinction between acknowledging systemic bias and engaging in identity-based decision-making without clear legal justification. In short, (1) employers can be liable for discrimination when facially neutral practices harm protected groups without business necessity; (2) efforts to remedy potential disparate impact must be carefully designed and evidence-based to avoid crossing into intentional discrimination, the latter of which is the “core prohibition of Title VII.” 557 U.S. at 577.
EO 14281: Sweeping, Largely Symbolic, Agency Directives
EO 14281–bundled with a related Fact Sheet and a Notice of Rescission of its 2022 Report on Lawful Uses of Race or Sex in Federal Contracting Programs –aims to eliminate race- and gender-conscious tools that have been long used as a means to achieve equal opportunity, notwithstanding the order’s perplexing framing of that concept. After this executive order, disparate-impact discrimination is no longer a bar to receiving federal contracts or grants, but practically speaking this is merely a footnote to January’s anti-DEI executive orders. The thrust of EO 14281 is in its policy pronouncements, consisting of ideological rhetoric embracing “a colorblind society,” followed by instructions to the Attorney General and agencies to take “appropriate action…consistent with the policy of this order.”
In its purpose section, the EO accuses disparate-impact laws of threatening the “bedrock principle of the United States … that all citizens are treated equally under the law.” Its title calls for “restoring equality of opportunity,” a deflection from the very reason disparate-laws came to be: “equality of opportunity” did not exist. At our country’s founding, women were barred from voting, Black individuals were enslaved, and–even long after these particular injustices were remedied–state and private actors treated these and other marginalized groups unequally and unfavorably with the law’s full backing. Until the Civil Rights movement, marginalized groups suffering discrimination had no meaningful legal recourse to vindicate the dream of “equal opportunity” against deeply embedded hatred and prejudice. The order invokes a historical ideal of ‘equal opportunity’ without reference to the remedial role civil rights law has historically played in addressing systemic inequality.
This EO seeks to build on January’s executive orders, which deployed similar “equal opportunity” rhetoric to imply that “DEI is discrimination,” again without acknowledging past discrimination. The implication that DEI is inherently discriminatory appears to overlook the fact that many DEI programs are implemented in response to existing legal frameworks, including disparate impact standards. With EO 14281, the administration may seek to address that apparent inconsistency, stating that “[d]isparate-impact liability is wholly inconsistent with the Constitution.” (EO 14281, Section 1.) A direct pronouncement of unconstitutionality is unusual in an executive order, and lacks any operative legal effect–this framing raises separation-of-powers concerns and appears inconsistent with settled precedent.
In its effort to dismantle disparate impact as a basis for discrimination liability, this executive order includes the following agency directives:
- Section 4 states that “all agencies shall deprioritize enforcement of all statutes and regulations to the extent they include disparate-impact liability.” Indeed, “deprioritiz[ing]” resource allocation is the most an order like this has the authority to do; the agencies are bound to enforce federal statutes, including disparate-impact provisions.
- The order revokes the Department of Justice’s authority to enforce disparate-impact discrimination under Title VI (section 3) and directs the Attorney General to initiate the repeal or amendment of implementing regulations related to disparate-impact liability (section 5). Simultaneously with the EO, the DOJ rescinded its 2022 guidance on lawful uses of race and sex in federal contracts. Note: These provisions apply only to discrimination in programs and activities receiving federal financial assistance under Title VI, and not to workplace discrimination under Title VII, which is codified by statute.
- The EO also sets its sights on state and local disparate-impact laws, requiring the AG to report on state laws with disparate-impact provisions, and identify “appropriate measures to address any constitutional or other legal infirmities” (section 5) and to “determine whether Federal authorities preempt State laws… that impose disparate-impact liability… or whether such laws… have constitutional infirmities that warrant Federal action” (section 7).
- Section 6 asks the EEOC chair and the Attorney General to review pending positions, investigations, consent decrees, and injunctions relying on a disparate-impact liability theory and “take appropriate action…consistent with the policy of this order” (e.g., presumably to drop disparate-impact prosecutions, or possibly amend consent decrees).
The extent to which these directives will result in concrete EEOC enforcement shifts remains to be seen. Administrative regulations cannot override judicial precedent or statutory text. And while the EEOC will likely rework some of its web content (it seems to have removed a technical assistance document called “Assessment of Software, Algorithms, and Artificial Intelligence for Disparate Impact,” for example), sources like the EEOC’s most recent annual report and its “Systemic Enforcement” page do not emphasize disparate impact as a standalone enforcement priority. This is unsurprising; under Ricci, disparate impact is already regarded as secondary to disparate-treatment as a basis for discrimination liability, and many actions that allege disparate impact also allege disparate treatment based on the same facts, evidence, and statistical analysis. If the facts supporting a disparate-impact claim also support a disparate-treatment claim, it would not warrant “deprioritizing” under this order.
The EO’s other general calls for “appropriate action” following reports on preemption and unconstitutionality likewise may not amount to much. Executive-branch agencies have carefully circumscribed authority, cannot declare laws unconstitutional, and lack statutory grounds for federal preemption. Aggressive invocation of preemption authority and agency interpretation sits uneasily with the administration’s public embrace of federalism and deregulation–and the law does not support it. Federal preemption requires a statutory basis, which it would be difficult for the AG to find given that federal law and regulation has for decades endorsed and enforced disparate-impact theory. As the Supreme Court stated in Wyeth v. Levine, 555 U.S. 555 (2009): “the historic police powers of the States were not to be superseded by the Federal Act unless that was the clear and manifest purpose of Congress.” And to the extent that federal agencies were entitled to deference on interpreting the statutes they enforce, the recent Supreme Court decision in Loper Bright Enterprises v. Raimondo curtailed that substantially, with Chief Justice Roberts holding in a majority opinion that “Courts must exercise their independent judgment in deciding whether an agency has acted within its statutory authority.”
Provisions of other recent executive orders have been subject to emergency challenge, e.g., as infringing on the First Amendment rights of private parties, exceeding executive branch authority, and/or violating separation of powers. Some such challenges have already succeeded—for example, on Friday, law firm Perkins Coie obtained a permanent injunction preventing the Department of Justice from enforcing an executive order blocking its attorneys from government access. Even where courts have declined to order interim relief on facial challenges to an executive order’s text (as the District of Columbia federal court did Friday), they are poised to block any attempts on the administration’s part to enforce the orders unconstitutionally. EO 14281 may face similar legal challenges, to the order itself or to actions the EEOC or other agencies take in line with it (such as dropping lawsuits).
No Legal Impact on Private Lawsuits under Title VII—including Class Actions
The EO may shift agency priorities or reframe internal enforcement policies. But it does not bind courts and cannot extinguish legal rights that Congress has created; courts are required to apply the law. Even if federal agencies curtail their use of disparate impact in enforcement, private plaintiffs remain fully entitled to pursue such claims. So long as they meet procedural prerequisites—such as filing a charge with the EEOC—they can still bring cases under Title VII and other statutes. Courts adjudicating those cases are bound to follow the law, notwithstanding executive branch policy; Executive Orders have no legal effect when they contradict congressional statutes.
Inasmuch as it directs the EEOC to deprioritize disparate impact, it might even reduce barriers to private plaintiff suits. If the EEOC decides not to spend time enforcing disparate-impact statutes, it may issue a “notice of right to sue” more quickly to private plaintiffs who wish to bring such claims. As disparate-impact cases often involve a particular policy or practice, they tend to be well suited for class and collective relief, further underscoring the importance for employers and any businesses involved in employment decisions of ensuring that their algorithms and practices are in compliance with disparate-impact laws.
Even though the EEOC seems to have removed its 2023 tactical guidance on artificial intelligence algorithms and disparate impact and may not initiate future enforcement actions on that sole basis, private plaintiffs may well continue to bring individual and representative challenges based on AI bias–not only against employers but vendors whose programs influence hiring results. In one currently-pending case, Mobley v. Workday, Inc., Case No. 3:23-cv-00770-RFL (N.D. Cal.), a plaintiff brought disparate-treatment and disparate-impact claims against a software vendor whose AI hiring tool allegedly screened out older, disabled, and Black candidates at statistically significant rates. He seeks to pursue his case on behalf of himself and others similarly situated, on a class and collective basis. On the defendant’s motion to dismiss, the court dismissed the disparate-treatment claim based on lack of evidence of discriminatory intent, but green-lighted the disparate-impact claim, finding that the plaintiff plausibly alleged a facially neutral process with a disproportionate impact—and that Workday could be treated as an agent of its employer clients under Title VII.
Section 7(b): Skills-Based Hiring: Next Steps for a Diverse, Equitable Workplace?
Among the order’s otherwise laser-focused attack on race- and gender-conscious initiatives is the following clause:
The Attorney General and the Chair of the Equal Employment Opportunity Commission shall jointly formulate and issue guidance or technical assistance to employers regarding appropriate methods to promote equal access to employment regardless of whether an applicant has a college education, where appropriate.
Section 7(b). This provision seems to reference a common subject of DEI advocacy and employment lawyer callouts: many job postings require a four-year degree even where it is unnecessary to perform the job. Educational criteria often screen out qualified job applicants from marginalized groups disproportionately, because those groups are also disproportionately excluded from access to college education. We can expect that access to diminish even further with this administration’s executive actions; unnecessary educational requirements are a common “facially neutral” practice that can give rise to a workplace disparate-impact claim, perhaps why the administration included this clause in this order.
Amidst 2025’s executive actions and the Supreme Court’s 2023 decision restricting race-conscious admissions in higher education, the burden to advance equal opportunity falls more than ever to employers. As this EO acknowledges, unnecessarily excluding candidates from certain educational backgrounds prevents equal workplace opportunity. Even before the EEOC and the AG promulgate the referenced guidance, employers can reevaluate their hiring criteria with this in mind.
Conclusion and Employer Takeaways
This president’s calls to disregard disparate-impact court decisions and statutes are best understood in the present historical context: as a reaction to 2020’s public outcry against the persistent structural racism that deprives racial minorities of equal rights in all settings, from workplaces to law enforcement. This backlash began in earnest in late 2020 (e.g., with Executive Order 13950), well before this year’s executive orders. Particularly in an administration whose public statements favor deregulation and states’ rights, such heavy-handed “facially neutral” rhetoric reveals a degree of irony.
It bears restate that the direct legal effect of this particular executive order is limited, particularly on private-sector employers. The primary impact of EO 14281 may well be in how its policy agenda–amplified and often sensationalized–influences those whom it threatens but does not bind: private actors, state and local governments, and other branches of the federal government. Consistent with the January orders, EO 14281 quotes selectively from foundational texts and ignores civil rights history, leveraging laws enacted to remedy centuries of legally-sanctioned discrimination to prioritize the rights of groups that historically benefited from it. These moves maximize political dramatic effect, hoping for a self-fulfilling prophecy when market forces and other branches of government acquiesce preemptively.
In this complex and sensitive climate, employers are uniquely poised to understand and advance equal opportunity, through actions that may include:
- Removing unnecessary screening criteria for job applications (apropos of the EO Section 7(b))
- Carefully considering the use of AI recruiting algorithms, and ensure that such algorithms are audited for bias;
- Embedding authentic values into the company’s culture; overhauling or eliminating initiatives that operated as merely performative exercises;
- Careful and considered use of data, in compliance with privacy laws and policies and with respect for data subjects.
