Weaving the Threads: The New Architecture of Federal Civil Rights Enforcement

The administration is narrowing disparate impact doctrine while building a far greater capacity to detect disparate outcomes. The machinery may prove more durable than the policy that built it.

In June, the Justice Department’s Office of Legal Counsel advised the Equal Employment Opportunity Commission that its longstanding approach to disparate impact liability under Title VII was constitutionally defective.¹ The Commission’s own chair had requested the opinion. OLC reasoned that liability resting on disproportionate effects alone can pressure employers toward precisely the race-conscious decision-making that Title VII and the Equal Protection Clause prohibit. The opinion does not make statistics irrelevant. It changes what they are for. A disparity may no longer supply the violation, but it can tell investigators where to look for one.

The Justice Department’s admissions work over the summer shows how much fits inside that distinction. In findings issued August 6 against a private law school, DOJ obtained applicant-level data, controlled for LSAT scores and undergraduate GPA, and reported that in the 2025 cycle a Black applicant had roughly three and a half times the probability of admission of an Asian applicant with similar academic credentials.² The Department captioned the relevant section “Statistical Evidence of Intentional Discrimination,” then paired the model with evidence concerning mission statements, application essays, reviewer instructions, internal tagging practices, and applicant characteristics associated with race. The theory was that a disparity surviving the controls called for an explanation, and that the explanation might be found in the design or administration of the ostensibly race-neutral process.

Findings against a public law school issued on September 9 follow the same architecture, under the same section caption.³ Controlling for LSAT scores and undergraduate GPA, DOJ compared admission outcomes by race and found highly statistically significant preferences in 2024 and 2025: Black applicants had 5.8 times the odds of admission of comparable White applicants in 2025 and 6.5 times in 2024, with Asian applicants disadvantaged to a similar degree. The accompanying evidence was of a different order. The letter opens its intent discussion with the dean’s published observation that institutions would need to find ways of achieving diversity that could not be documented as violating the Constitution. It then quotes video of the same dean describing an “unstated affirmative action” practice in faculty hiring, instructing colleagues that they could think it and vote it but should never articulate it, and adding that he would deny the conversation if deposed. An admissions dean’s email acknowledging the risk of becoming a litigation target, and choosing clarity about institutional values anyway, arrived with a Bates number attached. The review had been opened in March 2025 under a standing agreement between Justice and Education, expanded to the law school that April, and conducted cooperatively by both agencies.

The administration’s position is therefore more precise than a general retreat from statistical civil-rights enforcement. It rejects disparate outcomes as sufficient proof that a neutral policy is unlawful, while using those same outcomes to test whether the policy was neutral in the first place. That inquiry reaches well beyond an explicit racial checkbox. It reaches essays, interview protocols, socioeconomic indicators, first-generation status, disadvantage indices, recruitment pipelines, and scholarships. A narrower substantive doctrine can produce a wider investigation.

Finding the Targets

The proposed Admissions and Consumer Transparency Supplement to the Integrated Postsecondary Education Data System shows how quickly an investigative technique becomes regulatory infrastructure.⁴ ACTS was designed to collect unusually granular admissions and outcomes information from selective institutions, disaggregated by race and sex and cross-tabulated against test scores, GPA ranges, family income, parental education, admissions pathway, financial aid, and later academic outcomes. For the initial collection, the Department sought multiple years of historical data in order to establish a baseline extending to the period before Students for Fair Admissions, and said the resulting information could support risk-based enforcement practices.

For decades, civil rights enforcement generally began because someone complained, an event attracted attention, or the government otherwise acquired reason to examine a particular institution. A system designed like ACTS can begin before anyone has complained.

ACTS has been challenged in court, and preliminary relief has already narrowed its implementation.⁵ But focusing too heavily on the fate of the particular survey misses the more consequential development. The Department has articulated a regulatory objective: acquire enough institution-level data to identify unexplained disparities before a conventional investigation begins. ACTS is one mechanism for doing that. It is unlikely to be the last.

Narrowing the Doctrine, Widening the Investigation

That objective is increasingly pursued across agency lines. During a public webinar in late August, HUD’s Assistant Secretary for Fair Housing and Equal Opportunity described regular coordination among senior civil rights officials across the administration, including a monthly gathering of officials from multiple agencies.⁶ Other speakers returned repeatedly to the theme that the administration’s civil rights priorities were being pursued through coordinated federal action rather than isolated agency programs. That official came to HUD from the Department of Education, where he had served as Acting Assistant Secretary for Civil Rights, which illustrates the same point.

The August 28 medical-school reviews provide a public example of the same approach. HHS described the reviews expressly as a joint effort with DOJ and the Department of Education, extending beyond admissions to scholarships, fellowships, residency programs, contracting, program participation, and resource allocation.⁷ Five weeks earlier, the same agency had finalized a rule eliminating disparate impact liability from its Title VI regulations. It then opened investigations focused on intentional discrimination and the possible use of facially neutral criteria as proxies for protected characteristics. The doctrinal aperture narrowed while the investigative one opened.

The range of legal authorities now in use extends that pattern considerably. In August, a professional services firm agreed to pay the United States $21.5 million to resolve allegations involving race- and sex-conscious employment practices.⁸ The matter was not principally resolved as a conventional employment case. DOJ alleged that the firm had falsely certified compliance with nondiscrimination requirements contained in its federal contracts, converting the underlying civil rights allegations into False Claims Act exposure. A private relator had filed the action and received $4.3 million of the recovery. Other agencies have reached similar conduct through the statutes they administer. EEOC has proceeded under Title VII against a research university, alleging retaliation against an employee who objected to race-based separation during mandatory DEI training. HUD has used the Fair Housing Act on race-focused university housing, HHS pairs Section 1557 with Title VI in medical education, and Justice has opened compliance reviews into scholarships at one public university and alleged race discrimination at another. Volume 2 of this series described how a Section 117 foreign-funding review generated separate civil rights scrutiny at one research university, which is the same phenomenon seen from another angle.

Covering One Another’s Gaps

Treasury and the IRS have now proposed another doorway. Their September 4 proposal would provide that a private school engaging in racial discrimination in the administration of admissions, scholarships, athletics, or other educational programs does not qualify for federal tax exemption under Section 501(c)(3).⁹ The underlying principle is old, traceable to a 1971 revenue ruling the Supreme Court sustained in Bob Jones. Its breadth and placement are what matter. The proposal operates through the tax code rather than a particular federal funding relationship, and it applies to private institutions while excluding governmental ones.

A public flagship may face constitutional theories unavailable against a private college. A private institution may face tax-exemption consequences that do not always apply to the public one (although they may apply to the public school’s affiliated foundations). A federal contractor may face False Claims Act exposure. A medical school may face Section 1557. An employer may face Title VII. A recipient of federal financial assistance remains subject to Title VI.

Different federal agencies are covering one another’s gaps.

The treatment of facially neutral criteria follows from the same logic. In the admissions findings, the government has pointed to first-generation status, Pell eligibility, socioeconomic indicators, and similar factors as possible evidence that institutions were attempting to reproduce racial outcomes after Students for Fair Admissions. The public law school’s findings sharpen the point, because DOJ relied on leadership statements about pursuing racial diversity through race-neutral means as evidence of intent. None of that makes first-generation status, income, geography, or disadvantage inherently suspect. It makes purpose, design, implementation, and documentation consequential. A criterion adopted because an institution wants to reach economically disadvantaged students presents one question. The same criterion selected because decision-makers expect it to recreate a preferred racial composition presents another. The policy can look identical on its face. The legal difference may appear only in the emails, presentations, committee discussions, and analyses explaining why it was adopted.

Private actors add another layer. The False Claims Act action described above was initiated by an advocacy organization through the qui tam mechanism, with a direct financial interest in the outcome. Separate organizations have filed complaints concerning university housing, scholarships, admissions, and other programs. Enforcement capacity is accordingly no longer limited to the priorities and personnel of the federal agencies themselves. Private organizations can identify institutional practices, compare them across campuses, select the legal vehicle most likely to attract federal interest, and put a matter before the government. Under the FCA, they may also be compensated if the theory succeeds.

The breadth of the campaign changes how litigation should be understood. Higher education has followed the administration’s record in court closely, and understandably so. Some initiatives have been enjoined, some agency theories rejected, and others will be tested. But similar institutional practices have been reached through enough different statutes, agencies, and forms of leverage that judicial wins and losses provide only a partial measure of what is happening. If one data-collection mechanism is blocked, investigators may seek the information directly. If one statutory theory fails, another agency may possess a different authority reaching the same conduct. If a regulation is set aside, an older instrument may remain available. If government enforcement proves cumbersome, a private relator or complainant may proceed independently. No individual theory needs to prevail everywhere. Enough of them need to remain viable. The government is increasingly pursuing objectives through cases.

What Proves Durable

There is another reason institutions should not build their response around the politics of the present administration: the architecture has no fixed political direction. In August, the ranking member of the House Judiciary Committee demanded records concerning whistleblower allegations involving federal antisemitism investigations of universities.¹⁰ Former federal attorneys and other employees assigned to the Task Force to Combat Anti-Semitism alleged that certain investigations had predetermined outcomes and that career staff were pressured to support conclusions they believed the evidence did not justify. Those allegations are disputed.

The mechanics are not. A university can find itself answering fundamentally different questions about the same conduct over time. One part of the federal government may ask why the institution did not act more aggressively. Another may later ask whether it was coerced into acting at all. Documents created for one investigation can become exhibits in congressional oversight. Settlement discussions can be revisited. Agency communications can become relevant to inspector-general reviews or subsequent litigation. Internal assessments prepared to answer one administration can acquire a second life under another. A whole-of-government enforcement model produces a whole-of-government paper trail, and that record has no preferred political direction.

Much of the current civil rights agenda may change after the administration that created it. Executive orders can be rescinded, regulations rewritten, OLC opinions withdrawn, enforcement priorities reversed, and legal theories rejected by courts. But policy durability and institutional durability aren’t the same thing. What may prove harder to unwind are the capabilities being built around the policy: the datasets, the statistical methods sharpened across successive investigations, the standing agreements and relationships among civil rights offices, the habit of interagency referral, the experience applying old statutes in new settings, the private incentive structure, and the operating assumption that information already in the government’s possession may be interrogated for purposes unrelated to the program that collected it.

A future administration could therefore restore aggressive disparate impact enforcement while inheriting a considerably stronger apparatus for identifying disparate outcomes than existed before this administration set out to dismantle disparate impact liability. It could reject the current administration’s substantive priorities and keep its data, its methods, its interagency relationships, and its broader conception of where a civil rights violation might be found. The targets could change. The equipment would not have to.

Predicting which party controls Washington in the future isn’t a strategy. Leadership should know what information about the institution already sits elsewhere in the federal government, what disparities appear when its data are examined across programs, why the practices producing those outcomes were adopted, and what the institutional record says about that purpose. It should understand which representations to federal agencies depend on civil rights compliance, and which other agencies could become interested in the same facts. And before treating the defeat of one theory as the end of a matter, counsel should ask what other authority could reach substantially the same conduct.

A whole-of-government approach needs a whole-of-university response.

AJM Institutional Counsel and Strategy advises universities, colleges, and health systems on governance structures and decision-making frameworks for federal matters that reach across agencies, statutes, and administrations.

 

NOTES

1. Office of Legal Counsel, U.S. Department of Justice, “Constitutionality of Disparate-Impact Liability Under Title VII,” 50 Op. O.L.C. __ (June 9, 2026) (opinion of Assistant Attorney General T. Elliot Gaiser and Deputy Assistant Attorney General Joshua J. Craddock, issued at the request of EEOC Chair Andrea R. Lucas). https://www.justice.gov/olc/media/1444871/dl

2. Letter from Assistant Attorney General Harmeet K. Dhillon to Marc P. Berger, Latham & Watkins LLP, re: Duke University School of Law — Title VI Findings, August 6, 2026. https://www.justice.gov/crt/media/1456491/dl

3. Letter from Assistant Attorney General Harmeet K. Dhillon to Tania Faransso, Wilmer Cutler Pickering Hale and Dorr LLP, re: University of California, Berkeley, School of Law — Title VI Findings, September 9, 2026. https://www.justice.gov/crt/media/1460556/dl

4. “Agency Information Collection Activities; Comment Request; Integrated Postsecondary Education Data System (IPEDS) 2024-25 Through 2026-27,” 90 Fed. Reg. 39385 (Aug. 15, 2025). https://www.federalregister.gov/documents/2025/08/15/2025-15536/agency-information-collection-activities-comment-request-integrated-postsecondary-education-data

5. Commonwealth of Massachusetts v. U.S. Department of Education, No. 1:26-cv-11229 (D. Mass.), preliminary injunction entered April 3, 2026 and extended April 24, 2026.

6. The Federalist Society, Civil Rights in the Second Trump Administration: A Whole-of-Government Shift (webinar, Aug. 27, 2026) (remarks of Hon. Craig Trainor, Assistant Secretary for Fair Housing and Equal Opportunity, U.S. Dep’t of Housing and Urban Development; Hon. Brittany Bull Panuccio, Commissioner, U.S. Equal Employment Opportunity Commission; and Hon. Devon Westhill, Assistant Secretary for Civil Rights, U.S. Dep’t of Agriculture; Kenneth L. Marcus, moderator). Video and audio available at https://fedsoc.org/events/civil-rights-in-the-second-trump-administration-a-whole-of-government-shift

7. Press Release, “HHS’ Office for Civil Rights Launches Compliance Reviews of Five Medical Schools for Potential Title VI and Section 1557 Violations,” U.S. Department of Health and Human Services, August 28, 2026. https://www.hhs.gov/press-room/hhs-office-civil-rights-launches-compliance-reviews-five-medical-schools-potential-title-vi-section-1557-violations.html. The Title VI rule eliminating disparate impact liability was announced July 23, 2026. https://www.hhs.gov/press-room/hhs-restores-title-vis-promise-equal-protection.html

8. Press Release, “Deloitte Agrees to Pay $21.5M to Resolve Alleged Employment Discrimination Violations,” U.S. Department of Justice, August 25, 2026. https://www.justice.gov/opa/pr/deloitte-agrees-pay-215m-resolve-alleged-employment-discrimination-violations. The qui tam action is captioned United States ex rel. American Alliance for Equal Rights v. Deloitte LLP, No. 4:25-cv-00458; the relator received $4.3 million, and Deloitte separately agreed to pay $2.4 million to Indiana and Florida. The settlement was the second resolution under the Civil Rights Fraud Initiative, following a $17 million settlement with IBM in April 2026. On the other matters described in this paragraph, see Press Release, “EEOC Sues Washington University for Retaliation,” U.S. Equal Employment Opportunity Commission, August 12, 2026, https://www.eeoc.gov/newsroom/eeoc-sues-washington-university-retaliation (EEOC v. The Washington University, No. 4:26-cv-01289 (E.D. Mo.)); Press Release, “Justice Department Investigates William & Mary’s Scholarships & Student Benefits For Unlawful Race-Based Criteria,” U.S. Department of Justice, August 17, 2026, https://www.justice.gov/opa/pr/justice-department-investigates-william-marys-scholarships-student-benefits-unlawful-race; and Press Release, “Justice Department Investigates Alleged Race Discrimination at the City University of New York,” U.S. Department of Justice, June 9, 2026, https://www.justice.gov/opa/pr/us-justice-department-investigates-alleged-race-discriminat

9. “Racial Nondiscrimination in Private Schools,” REG-119986-25, RIN 1545-BS05 (proposed Sept. 4, 2026). https://www.federalregister.gov/d/2026-18127

10. Letter from Ranking Member Jamie Raskin, House Committee on the Judiciary, to Assistant Attorney General Harmeet K. Dhillon, August 18, 2026. https://democrats-judiciary.house.gov/sites/evo-subsites/democrats-judiciary.house.gov/files/evo-media-document/2026-08-18-raskin-to-dhillon-doj-re-title-vi-investigation.pdf