Ending the “Woke” Wars: A Federalism-Based Mechanism for Enforcing Civil-Rights Grant Conditions

abstract. President Trump’s executive orders attacking “illegal DEI” have provoked complaints about the executive orders’ vagueness and their burdens on free expression. But it is hardly unprecedented for presidents to use vaguely defined powers over federal revenue to impose controversial civil-rights mandates in ways that risk burdening grantees’ civil liberties. President Nixon’s “affirmative action” mandate imposed burdens on federal contractors analogous in their vagueness to Trump’s attack on “illegal DEI.” Likewise, “hostile educational environment” theories enforced against grantees pursuant to Title VI and IX have created risks of chilling freedom of expression similar to the risks posed by President Trump’s executive orders. In these respects, Trump’s “war on wokeness” makes use of the same “woke” tools that his followers had previously decried. Just because a practice has precedents, however, does not make it a good idea.

This Essay argues that this type of unilateral presidential power over civil-rights spending conditions inflames partisan polarization and unnecessarily crowds out pluralism about matters on which reasonable citizens can appropriately disagree. Helpfully, though, buried in Title VI and analogous statutes is a partial antidote: the often-ignored “pinpoint” provision limiting the sanctions that presidents can impose on grantees who resist their demands. By limiting penalties to the “program or part thereof” found to be in noncompliance, the pinpointing provisions restrain presidential power for the sake of a pluralism-promoting federalism. The Essay defends an interpretation of pinpointing that emphasizes proportionality and severability, thereby scaling federal leverage to actual federal investment, creating room for regional diversity on reasonably disputed interpretations of our national civil-rights commitments.

Introduction

Executive Orders 14,151 and 14,173—which target federally funded diversity, equity, and inclusion (DEI) programs—reveal a paradox in the Trump Administration’s so-called “War on Woke.”1 That “war” borrows heavily from the legal strategies that anti-“woke” warriors themselves associate with “woke” advocacy. “Wokeness” is admittedly a slippery concept. It evolved from its original meaning in 1930s Black vernacular to denote awareness of racial prejudice, to become a catch-all term for sensitivity to concerns of social justice by 2010, and to eventually be used in 2020 as a derisive epithet denoting a pretentiously exaggerated sensitivity to such concerns.2 Dismissing the term as mere rhetoric, however, ignores how both the popular press and public intellectuals use the term to describe the target of the Trump Administration’s campaign against the existing legal regime governing racial discrimination.3 Moreover, denunciations of “wokeness,” far from being empty rhetoric, have a fairly plain target: they repeatedly denounce theories alleging that ostensibly race-neutral social practices reflect “systemic” or “structural racism.”4 Making allowances for the inevitable fuzziness of political invectives addressed to a lay audience, those denunciations are obviously an attack on theories of systemic, structural, or institutional racism pressed by legal and social theorists from the late 1970s to the present. Those ideas urged a theory of racial discrimination imposing legal liability not only for intentional (sometimes denoted “attitudinal” or “personalistic”) racial discrimination, but also for institutions or structures that perpetuate racial inequalities or reflect implicit or unconscious racial stereotyping.5

So understood, the Trump Administration’s attacks on “wokeness” are not especially vague: they seek to roll back remedies for racially disparate impacts of formally race-neutral practices. Some of the rollback seeks to enforce the prohibition on the use of racial classifications announced by Students for Fair Admissions, Inc. v. President and Fellows of Harvard College.6 The Trump Administration’s hostility to DEI, however, goes beyond this limited goal of getting rid of express racial classifications to attack even race-neutral programs designed to address structural or systemic racism.7

Behind this ambition, however, lies an irony. Much of the Trump Administration’s campaign against “woke” theories itself rests on a “woke” theory—namely, a theory that DEI programs, even when race-neutral, nevertheless marginalize white people and, therefore, constitute a kind of institutional or systemic racism. The Trump Administration’s attack on wokeness is, in this sense, a kind of wokeness for white people.

In this Essay, I argue that the broad reach, vague terms, and extraordinary agency discretion conferred by Title VI of the Civil Rights Act of 19648 and analogous civil-rights statutes and policies facilitate the Trump Administration’s effort to flip “woke” antiracism on its head. Title VI and the three other statutes modeled after it all mandate that “no person in the United States shall on the basis of” some specified characteristic “be excluded from participation in, be denied the benefits of, or be subjected to discrimination under any program or activity receiving Federal financial assistance.”9 None of these requirements (which, for convenience, I will call “civil-rights spending conditions” or “CRSCs”) defines the concept of discrimination referenced by the “excluded/denied/subjected” phrase. Instead, Congress left it up to federal agencies to fill in the details. Federal agencies have done so through vague rules supplemented by somewhat less vague but formally nonbinding guidance. In response, grant recipients have developed internal bureaucracies to create compliance measures that federal agencies review through audits and investigations.10

This system of executive discretion through regulatory ambiguity has enabled the Trump Administration to shift policy radically, free from the restraints of clear rules or statutes. As I explain in more detail in Part I below, any attack on President Trump’s executive orders implementing Title VI must reckon with the exceptional ambiguity of the rules with which CRSCs have traditionally been implemented. Starting with President Kennedy’s mandate to federal contractors to institute “affirmative action” in E.O. 10,925, presidents have implemented their vision of civil rights by announcing undefined concepts that executive officials later cash out with more specific policies. “Illegal DEI,” a term used in the Trump Administration’s two executive orders discussed below,11 is certainly an undefined and exceptionally vague concept in 2025—but, as I explain below in Section I.A, no more so than “affirmative action” was between 1961 and 1969. It was only the Nixon Administration that finally filled out the idea of “affirmative action” with its revised “Philadelphia Plan” for minority hiring.12 Likewise, Trump’s executive orders are presently vague and await clarification through agency implementation. That implementation might possibly chill expression by recipients of federal grants who are fearful of running afoul of a ban on “illegal DEI.” But Kennedy’s executive orders encouraging “affirmative action” were later implemented by duties to take proactive steps to eliminate workplace or educational conditions contributing to racial or sex-based disparities. Those proactive steps included censorship of workplace or university speech deemed to create a “hostile environment” leading to such disparities.13 As I explain in Section I.B, however, well-established judicial doctrine has long tolerated bans on expression that create “hostile environments” deemed to discriminate on grounds forbidden by federal civil-rights statutes. Just as the vagueness of Trump’s recent executive orders mirrors that of longstanding antidiscrimination mandates, the burden on free expression imposed by Trump’s executive orders does not seem materially different from earlier burdens on expression at workplaces, college campuses, or K-12 schools.14

In sum, President Trump is using the “weapons of the woke”—that is, vague, difficult-to-implement antidiscrimination law—against the diversity, equity, and inclusion programs that wokeness created. His executive orders exploit longstanding ambiguity about permissible racial classifications and courts’ tolerance of the associated burdens on freedom of expression to destroy the prior legal regime by which Title VI was implemented.

Part II argues that extreme oscillations in the interpretation of CRSCs with each change of presidential administration are costly. I examine how disputes about the prevalence of and proper remedies for racial discrimination divide Americans along partisan lines. These disputes contribute to partisan polarization which paralyzes the governmental process and even promotes political violence. When a President enforces a partisan interpretation of CRSCs—an interpretation rejected by their political opponents—it raises the stakes of presidential elections in ways that exacerbate such polarization.

Part III argues that those costs could be mitigated by federalism-based limits on presidential power. The Part begins by arguing that a narrow construction of Title VI can ameliorate partisan divisions by limiting the power of federal agencies to coerce grantees into acceding to views about civil rights that grantees reject and even deplore. Next, it explains that Title VI has built within itself mechanisms for such a federalist solution to partisan polarization in its so-called “pinpoint provision.” The pinpoint provision requires that the termination of funding for grantees who violate CRSCs “shall be limited in its effect to the particular program, or part thereof, in which such noncompliance has been so found.”15 Under the provision, presidents can still enforce their interpretation of CRSCs, but grantees have greater power to reject such interpretations because they have less money at stake. According to the reading of the pinpoint provision urged here, federal agencies could still threaten to strip grantees of federal grants when those recipients reject the federal executive’s reading of Title VI. Those threats, however, would be narrowly confined to subnational activities specifically receiving federal funding (for instance, university labs), as opposed to other programs controlled by the grantee but financed through own-source revenue (for instance, university athletics). That limit radically diminishes the power of the federal government to bend subnational governments to the President’s will. So understood, “pinpointing” limits the federal power to terminate grants by the amount of federal revenue actually at stake in a disagreement over Title VI’s meaning.

Limits on grant termination would not mean that CRSCs would go unenforced. Private causes of action could still be used to enforce the statute itself, while the Department of Justice (DOJ) could seek injunctions to enforce formal legislative rules implementing Title VI. Pinpointing, however, would limit the potent threat of grant termination to nudge recalcitrant subnational governments into complying with interpretations of Title VI that have neither been read into Title VI by the courts nor vetted through the notice-and-comment rulemaking process by agencies.

I. ambiguity and free expression in civil-rights spending conditions

President Trump’s executive orders’ vague terms and expression-chilling effects fit comfortably within well-established readings of both Title VI and the U.S. Constitution. This Part turns to the history of presidential implementation of civil rights to show just that.

E.O. 14,151 denounces “diversity, equity, and inclusion” programs as “forced illegal and immoral discrimination programs” and calls for the “termina[tion]” of “equity-related grants.”16 Section 3(b)(iv)(B) of E.O. 14,173 further requires federal grantees to “certify that [they do] not operate any programs promoting DEI that violate any applicable Federal anti-discrimination laws.”17 In National Association of Higher Education Diversity Officers v. Trump,18 Judge Abelson of the District of Maryland preliminarily enjoined the executive orders on the grounds that: (1) E.O. 14,151’s “termination provision” was unconstitutionally void for vagueness; and (2) E.O. 14,173’s “certification provision” burdened grantees’ freedom of expression in violation of the First Amendment. As shown below, however, both of these holdings run up against longstanding traditions of the executive branch enforcing mandates to achieve racial equality with vaguely worded policies and courts permitting burdens on freedom of expression incidental to that enforcement.

A. The Tradition of Ambiguity in Executive Orders Implementing Racial Equality

Judge Abelson held that E.O. 14,151’s threat to terminate grants related to DEI is unconstitutionally vague, violating grantees’ Fifth Amendment due-process right to notice of the content of legal requirements.19 The court stated that the termination provision “leaves the private sector at a loss for whether the administration will deem a particular policy, program, discussion, announcement, etc. to be among the ‘preferences, mandates, policies, programs, and activities’ the administration now deems ‘illegal.’”20 In this Section, I argue that presidents have long implemented racial equality through vaguely worded executive orders, and that Abelson is wrong to see President Trump as an exception to that tradition.

The vagueness of which Judge Abelson complains has been baked into CRSCs since 1961, when President Kennedy issued E.O. 10,925 calling for federal contractors to take “affirmative action” to ensure nondiscrimination in employment.21 E.O. 10,925 did not define what affirmative action meant. Although E.O. 11,246, President Johnson’s 1965 sequel to E.O. 10,925, was more detailed in its specification of contractors’ reporting duties, Johnson’s order did not define affirmative action beyond noting that it “shall include” a wide variety of employment practices ranging from “employment, upgrading, demotion, or transfer; recruitment or recruitment advertising” to “selection for training, including apprenticeship.”22

In short, presidents, since the onset of the civil-rights revolution, have employed vague antidiscrimination mandates in executive orders. Clarification about what “affirmative action” actually meant, therefore, required a long, painstaking process of bureaucratic implementation. The evolution of the so-called “Philadelphia Plan” illustrates both the vagueness of the initial concept of affirmative action and how subsequent political conflict—not a clear presidential definition—gradually defined it.

In an effort to increase Black employment following riots in the summer of 1967, midlevel federal bureaucrats in Philadelphia devised an “operational plan” to require contractors bidding on federal construction projects to declare goals for hiring Black workers.23 Those hiring goals, however, did not include firm numbers of necessary hires: only after the contract was awarded, contractors would work out the actual numbers through regular meetings with federal compliance officers.24 Moreover, the Plan was kept deliberately vague with respect to the percentages of Black hires needed to satisfy E.O. 11,246 out of fear that any specific percentage would be seen as an illegal quota in violation of Section 703(j) of Title VII’s prohibition on “preferential hiring” to correct “imbalances” in the workforce.25 Rather than define hiring percentages, the first version of the Philadelphia Plan left it to federal contractors to propose hiring goals.26 That very ambiguity would be the cause of the Plan’s initial (albeit temporary) defeat. The Comptroller General (the officer in charge of the General Accounting Office, an office internal to Congress advising it on budgetary and revenue matters) opined that, because the Plan did not specify a particular hiring quota, it violated federal procurement rules by imposing a post-bid requirement on low bidders.27 Following the Comptroller General’s opinion, Johnson allowed the Plan to die.28

In an unexpected “Nixon-to-China”-style maneuver, however, the incoming Nixon Administration revived the Philadelphia Plan in the spring of 1969.29 President Nixon may have been motivated primarily by his Machiavellian political desire to pit two Democratic constituencies—the civil-rights establishment and trade unions—against each other.30 His revival of the Plan, however, became the seed for eventually transforming “affirmative action” from an empty catchphrase into a viable legal concept.

The key to the transformation was in who set the hiring goals. The revived Philadelphia Plan assigned to the federal government, rather than to federal contractors, the task of defining hiring goals for Black workers.31 “Affirmative action,” so defined, thus involved race-conscious state action. Southern Democrats and conservative Republicans in Congress quickly united to denounce this version of “affirmative action” as a racial “quota” that violated Section 703(j) of Title VII.32 In response, the Nixon Administration developed a conceptual distinction between “goals” and “quotas.”33 The distinction was rooted in the goals’ character as (1) a percentage range rather than a precise number and (2) a rebuttable presumption that contractors could overcome by a showing of “good faith” effort.34 In addition, the Nixon Administration, analogizing its actions to the U.S. Supreme Court’s desegregation of public schools and Congress’s elimination of literacy tests for voting, emphasized that its hiring goals were a properly race-conscious response to the building trade unions’ long history of racial discrimination against Black workers.35 In the ensuing showdown with Congress, the Nixon Administration ultimately prevailed in defeating efforts to repeal the Philadelphia Plan, aided politically by labor protests over the summer of 1969 that turned violent in Chicago and Pittsburgh.36

The Philadelphia Plan’s legislative victory and President Nixon’s subsequent invocation of the distinction between “goals” and “quotas” created our modern concept of affirmative action, almost a decade after President Kennedy had used the phrase in E.O. 10,925. Even after this victory, however, the quotas versus goals distinction that undergirded the defense of the Philadelphia Plan remained an embattled and uncertain legal category. It is not at all clear that the defenders of the distinction in the Nixon Administration took it seriously beyond using it as a convenient weapon against labor-union racism.37 Moreover, the exact content of the distinction was never certain. Why, after all, was a percentage range less of a quota than a fixed number, when both were obviously numerical?38 Justifications for the “goal”/“quota” distinction varied. The Nixon Administration emphasized that goals created only a rebuttable presumption that a contractor could overcome by showing their good-faith, even if unsuccessful, efforts to meet the hiring goal.39 The Supreme Court emphasized, by contrast, the remedial aspect of affirmative action in United Steelworkers of America v. Weber40 and Fullilove v. Klutznick,41 with Justice Powell’s decisive opinion in Regents of the University of California v. Bakke adopting the “goal”/“quota” distinction as the basis for striking down the University of California’s admissions quota.42 Whatever their varying justifications, however, the executive and judicial branches nevertheless converged on a concept of “affirmative action” almost two decades after Kennedy proposed the idea in an executive order, but that concept still today remains disputed in its scope and justification.43

Judge Abelson’s holding that President Trump’s executive orders are unconstitutionally vague, therefore, runs up against a longstanding tradition of executive orders containing vague CRSCs that are made specific only through later executive implementation. Abelson is correct that the executive orders do not define what the Trump Administration considers “illegal DEI discrimination and preferences,” “[p]romoting diversity,” “illegal DEI and DEIA policies,” or what types of “DEI programs or principles” the Administration considers unlawful and is seeking to prevent.44 The same complaint, however, could be leveled against the term “affirmative action” in E.O. 10,925 and E.O. 11,246, which left contractors in the dark about what sort of “action” they should take, and what exactly such action should “affirm.” Would it have sufficed simply to try to revise collective-bargaining agreements with craft unions that had excluded Black workers from apprenticeship programs? Was it enough to urge those unions to accept Black trainees, or did the contractors also have to hire nonunion workers in defiance of those collective-bargaining agreements? Were race-based hiring goals acceptable, given the prohibition on racial preferences in Section 703(j) of Title VII? As a preeminent historian of the Philadelphia Plan notes, federal contractors “worried that the goals were too vague and would have preferred outright quotas.”45 Likewise, the “good-faith” defense to failure to achieve a hiring “goal” was “deliberately left undefined so as to give contractors that exercised affirmative action as much leeway as possible should their efforts fail to produce results.”46

President Kennedy’s and Johnson’s executive orders, and President Nixon’s later implementation of both, said nothing whatsoever about the specific version of affirmative action that was required. These questions were answered only gradually and incompletely by executive actions following the executive orders.

The references to “illegal DEI” in President Trump’s E.O. 14,151 and E.O. 14,173 are no more obscure than the references to “affirmative action” in E.O. 10,925 and E.O. 11,246. In ordinary usage, “diversity, equity, and inclusion” refers to policies that promote the representation of underrepresented racial minorities. Trump’s executive orders do not specify how such policies might violate the law, but guidance from Trump Administration officers indicates that race-conscious DEI measures might conflict with the prohibition on racial classification announced in Students for Fair Admissions, Inc. v. President and Fellows of Harvard College.47 How exactly such a conflict might arise is not explained in detail by any of this guidance. Numerous lawsuits challenging various public and private efforts to promote racial diversity and inclusion, however, provide at least as much clarity about the scope of “illegal DEI” as Presidents Kennedy and Johnson’s executive orders provided about the scope of required “affirmative action.”48

In sum, it seems a bit late in the day to complain about conceptual vagueness in executive orders implementing Title VI. Indeed, those who are opposed to race-conscious remedies for racial inequality like affirmative action might reasonably complain of a double standard that privileges the vague concept of “affirmative action” over the equally vague idea of “illegal DEI.” For the same reason, even as President Trump wages his war on wokeness, he is drawing from the same toolbox that gave rise to the policies he denounces.

B. The Tradition of Courts Tolerating Burdens on Free Expression from CRSCs

Complaints that President Trump's executive orders chill speech seem just as foreclosed by longstanding precedent as complaints about the orders’ vagueness. The illegality of Trump’s executive orders, according to Judge Abelson, resulted not only from their vagueness but also from the fact that they could chill freedom of expression in violation of the First Amendment. The danger to free speech, Abelson explained, arose from the “certification provision” of Section 3(b)(iv)(B) of E.O. 14,173, which requires each grantee “to certify that it does not operate any programs promoting DEI that violate any applicable Federal anti-discrimination law.”49 Abelson deemed that provision to be an unconstitutionally coercive demand that grant recipients self-censor their pro-DEI advocacy.50

Judge Abelson’s concern about unconstitutional chilling of expression runs up against a line of precedents peculiar to antidiscrimination law that are unusually tolerant of “incidental” speech restrictions aimed at discriminatory conduct. This tolerance of content-based speech restrictions is most closely associated with the doctrine of Meritor Savings Bank v. Vinson, where the Court held that sexual harassment in the workplace, if sufficiently severe or pervasive, can create a hostile or abusive work environment constituting sex-based discrimination in violation of Title VII.51 As Professor Eugene Volokh argued four years after Meritor Savings Bank was decided, this prohibition on workplace harassment often imposes civil liability on employers for permitting what was “core protected speech.”52 Such speech includes, for example, political advocacy, religious proselytizing, and other communication about public affairs that in other contexts would be protected by the First Amendment.53 Moreover, as Professor Richard Fallon observed, liability for harassing speech “runs afoul of what the Supreme Court has often trumpeted as perhaps the central tenet of the First Amendment”—content neutrality—because liability is predicated on the communicative impact of the speech.54

There is no easy way to explain why liability for creating a hostile environment under antidiscrimination law does not violate the First Amendment’s protection of expression, and yet courts have repeatedly imposed such liability without any such explanation. The closest the Supreme Court has ever come to defending such liability from First Amendment concerns occurred in R.A.V. v. City of St. Paul.55 The R.A.V. Court reasoned that Title VII’s prohibition on workplace discrimination is a generally applicable law “directed not against speech but against conduct,” permitting harassing speech to “be swept up incidentally within the reach of a statute directed at conduct rather than speech.”56

As several legal scholars and Justice White’s R.A.V. concurrence have noted, the R.A.V. majority’s legal reasoning leaves much to be desired.57 It is certainly correct that, as far back as United States v. O’Brien,58 the Court has upheld incidental restrictions on speech imposed by generally applicable laws aimed at conduct other than speech.59 The R.A.V. majority’s justification for speech restrictions imposed by hostile-environment liability, however, does not easily fit into First Amendment doctrine. The First Amendment does not ordinarily allow generally applicable laws to be enforced against speech for the purpose of preventing harms caused by that speech’s persuasive or emotional force on listeners, or its “communicative impact.”60 Enforcing generally applicable laws to curb expression’s communicative impact on listeners is, in Professor Volokh’s phrase, “content-based as applied,” and content-based speech restrictions are ordinarily subject to strict scrutiny.61 It is one thing to prohibit vandals from spray-painting slogans on a governmental building to preserve the building’s tidy aesthetic appearance. It is quite another to prohibit vandals from spray-painting slogans on a governmental building because the government seeks to prevent onlookers from being persuaded or offended by what those slogans declare.62

Sexual-harassment liability seems more akin to the latter than the former. Regardless of the speech neutrality of Title VII in general, liability for creating a “hostile environment” through sexist or racist speech plainly hinges on the communicative impact of such speech on listeners aggrieved by hostility.63 One cannot be driven out of a classroom or workplace by pervasive or severe hostility unless one understands and is offended by the hostile message.

Despite the force of the argument for subjecting at least some “hostile-environment” claims to First Amendment strict scrutiny, lower courts, following R.A.V.’s guidance, have generally upheld such statutory claims against First Amendment defenses.64 The reasoning underlying Title VII workplace harassment, moreover, has migrated from the workplace to programs and activities covered by Title VI and Title IX, barring recipients of federal grants from creating or permitting a racially or sexually hostile educational environment.65 Occasionally, lower courts have cited free-speech concerns when limiting employers’ liability by requiring a lot of evidence of pervasiveness for hostility to rise to a Title VII violation.66 Despite these occasional judicial reservations, however, there is a well-established line of lower-court precedents repeatedly applying antidiscrimination laws to impose liability on various forms of workplace expression, ranging from background music with misogynistic lyrics to religious proselytizing by a supervisor.67

In denouncing DEI policies as illegal, President Trump’s executive orders implicitly embrace this “hostile environment” line of precedent. For instance, E.O. 14,173 declares that “[h]ardworking Americans who deserve a shot at the American Dream should not be stigmatized, demeaned, or shut out of opportunities because of their race or sex.”68 Stigmatizing or demeaning persons because of race or sex is precisely the mechanism by which a hostile environment in the workplace or classroom constitutes illegal discrimination. President Trump’s executive orders do not explain how DEI programs stigmatize anyone. The February 2025 guidance from the Department of Education’s Office for Civil Rights (OCR), however, fills in the gaps, asserting that DEI programs “teach students that certain racial groups bear unique moral burdens that others do not” and thereby “stigmatize students who belong to particular racial groups based on crude racial stereotypes.”69 These general statements echo lawsuits brought to challenge diversity or sensitivity training as inflicting a hostile workplace environment on white plaintiffs.70

In denouncing DEI programs as “stigmatiz[ing]” and “demean[ing],” E.O. 14,173 implicitly invokes a tradition of “hostile environment” doctrine that is resistant to Judge Abelson’s First Amendment reasoning. Abelson is surely correct that the First Amendment ordinarily would protect the speech involved in DEI programs.71 But if such speech imposes a hostile work environment on white employees, then—as discussed earlier in this Section—a longstanding line of cases deems limits on that speech to be a merely “incidental” burden on expression in the service of a “generally applicable” prohibition on conduct.

That line of cases also undermines Judge Abelson’s reliance on Agency for International Development v. Alliance for Open Society, International.72 In Alliance for Open Society, the Court held that a federal statute requiring grantees to certify that they opposed prostitution violated the grantees’ First Amendment rights to freedom of speech.73 Judge Abelson invoked Alliance for Open Society’s principle that the federal government may not “leverag[e] its funding to restrict federal contractors and grantees from otherwise exercising their First Amendment rights outside the scope of the federal funding.”74 According to Abelson, Section 3(b)(iv)(B) of E.O. 14,173 violated this principle by requiring grantees and contractors to certify that they did “not operate any programs promoting DEI that violate any applicable Federal anti-discrimination laws.”75 The violation allegedly results from the fact that the required certification applied to all of the grant recipients’ programs, however funded, rather than just their programs that received federal funding.

The analogy to the grant restriction at issue in Alliance for Open Society breaks down, however, because E.O. 14,173 requires certification that the grant recipient did not “violate any applicable Federal anti-discrimination laws.”76 As explained above, such violations have long been deemed to be “conduct” unprotected by the First Amendment, even when the conduct partly consists of speech. Certifying that one is not engaging in illegal conduct, therefore, bears no analogy to the coerced certification at issue in Alliance for Open Society, which required grant recipients to adopt a policy opposing prostitution. In the simple dichotomy of R.A.V.—and of the lower courts that have followed it—refusing to oppose prostitution is not illegal conduct but protected speech. By contrast, violating antidiscrimination laws is not protected speech but prohibited conduct.

The burdens on free expression imposed by the Trump Administration’s executive orders are, in sum, hardly unprecedented. Nor has the Trump Administration departed from past administrations’ practice by seeking to leverage federal grants to control grantees’ programs that are funded without federal grant revenue. Judge Abelson’s objection to such “leveraging” federal grants reveals the extent to which he has missed the peculiar legal traditions governing Title VI and related antidiscrimination laws. Since Congress enacted the Civil Rights Restoration Act in 1987 to overrule the Court’s decision in Grove City College v. Bell,77 Title VI and Title IX conditions have applied to “program[s] or activit[ies]” that do not directly receive any federal funding.78 (That is precisely why public colleges’ and universities’ athletic programs are covered by Title IX despite receiving no direct federal funding: they are nevertheless part of a federally funded “agency” or “department”). In short, the CRSCs imposed by Title VI and Title IX routinely violate the antileveraging principle that Abelson invokes to enjoin President Trump’s executive orders. If Abelson is right that such leveraging violates the First Amendment, then a long line of precedents upholding liability for an entire institution based on one part of that institution’s receiving federal funds would have to be reversed.

As with President Trump’s vague language of “illegal DEI,” the Administration’s certification provision in E.O. 14,173 and similar attempts to leverage funding to suppress DEI use the weapons of the woke—that is, Title VI and Title IX hostile-environment liability that undermines conventional First Amendment defenses—to make war on wokeness. Ironically, the Republican Party took precisely Judge Abelson’s stance regarding these weapons several years earlier, dedicating their 2020 Convention to attack the “cancel culture” that such liability is alleged to encourage.79 By contrast, the Trump Administration instead seeks to expand such liability and the “cancel culture” they allegedly promote to include new litigants (white persons conceived as an aggrieved minority group) who want to “cancel” a new kind of “hostile environment” (DEI programs that allegedly stigmatize white workers or students).80 The merits of any such expansion of liability would turn on the particular DEI programs targeted by specific agency actions implementing the executive order. Until such implementation takes place, however, it is difficult to defend Abelson’s position that the executive orders were invalid on their face. Those orders are simply too similar to the past implementation of Title VI to be facially invalid.

Critics of the Trump Administration’s attack on “illegal DEI” might argue that the attacks are distinguishable from earlier implementations of “affirmative action on grounds of sincerity of motivation. Such critics might concede that the meaning of and justification for “affirmative action” were never very clear but nevertheless assert that earlier affirmative obligations to eliminate racial disparities produced by facially race-neutral policies was reasonably suggested by statutory language and affirmed by judges and politicians from both ends of the political spectrum. By contrast, such critics may dismiss the Trump Administration’s attack on “illegal DEI” as a merely cynical political exploitation of racial grievances without a substantial theory of illegality.

Such an intent-based distinction between the two rival readings of CRSCs, however, might be too charitable to the Johnson and Nixon Administrations and too dismissive of the Trump Administration’s case against DEI programs. The proponents of “affirmative action” had unsavory motivations as well: As noted above, President Nixon in particular hoped to divide labor unions from civil-rights advocates with his affirmative-action proposal. More generally, some of DEI’s critics charge theorists who push theories of institutional or systemic racism with dogmatic yet incoherent and oppressively divisive accusations of racism.81 As for the legal merits, the question of whether liability based on racial disparities shades into illegal racial classifications has been an unsettled question for decades.82 Indeed, fighting the Trump Administration’s executive orders on grounds of vagueness or censorship—losing grounds when one looks to history and doctrine—will risk provoking plausible accusations that the bench and academia are applying a partisan double standard.

As explained in the next Part, the more promising legal theory for attacking the Trump Administration’s executive orders focuses less on their familiar vagueness and burdens on free expression and more on their unprecedented reliance on total funding cutoffs regardless of the scale of the alleged transgression against Title VI and analogous statutes.

II. how abrupt funding cutoffs exacerbate partisan polarization

While President Trump’s executive orders may not be facially invalid, the Trump Administration’s unprecedented use of CRSCs and their fundamentally vague enforcement scheme raise novel worries requiring judicial intervention. Those worries appear in the Administration’s now-notorious penchant for cutting off all federal funds from recalcitrant grantees while ignoring the statutory procedures and limits governing grant termination.83 As explained below, this practice of peremptory and total grant termination, unprecedented in the history of Title VI, dramatically raises the stakes of partisan disagreements over the meaning of racial equality. Presidents’ forcing of their political party’s interpretation of CRSCs on the entire nation through such funding cutoffs exacerbates our current crisis of crippling partisan polarization.

The ambiguity of antidiscrimination law, described in Part I above, reflects the political conflicts buried in the vague terms of CRSCs. In the case of “affirmative action,” the unresolved tension lies between citizens’ desire to end racial inequality and a deep discomfort with the racial classification of persons.84 Public opinion is conflicted over the relative priority of race neutrality and racial equality, with surveys indicating that the public, by large majorities, disapproves of race-conscious admissions for higher education and hiring yet simultaneously approves of promoting racial diversity in education and employment.85 In the case of “hostile environment” discrimination, this unresolved tension leads to an ambiguous line between preventing discriminatory abuse in the workplace, classroom, or other settings where CRSCs apply and preserving ordinary civic freedoms. As explained in Part I, the text of neither Title VI nor Title IX resolves these tensions. The executive and judicial branches have instead finessed these tensions with vague rules and nonbinding guidance. That legal ambiguity has spurred regulated organizations to hire specialists who manage compliance with those vaguely defined requirements.86 By promulgating and monitoring the enforcement of “best practices,” these specialists act as brokers between the agencies and the organizations that agencies regulate, enlarging the power of the latter without fundamentally disrupting the governmental needs of the former.87

President Trump’s executive orders purport to cut through this ambiguous modus vivendi by threatening grantees with a loss of all federal funding for failure to abide by its new interpretation of CRSCs. That interpretation not only requires rigid color blindness in a maximal—arguably overbroad—reading of Students for Fair Admissions but also even suggests that purportedly race-neutral means for promoting diversity (for example, eliminating reliance on SAT test scores for university admissions) are forbidden.88

This enforcement of Title VI is novel and aggressive not only in its ends but also in its means. The Trump Administration has asserted extraordinary unilateral power over federal revenue—what Professors Matthew B. Lawrence, Eloise Pasachoff, and Zachary S. Price felicitously term “appropriations presidentialism”89—by threatening funding cutoffs that ignore hearings and statutory limits.90 Because of the pervasive ambiguity of CRSCs described in Part I, however, courts cannot easily constrain these presidential maneuvers by showing a conflict between the Trump Administration’s interpretation and some plain text or precedent defining forbidden discrimination. For instance, President Trump has threatened to deprive universities of billions of dollars in federal grants based on findings of occasional campus antisemitism.91 There is no agency rule or judicial precedent, however, defining how many discriminatory incidents it takes to create a hostile environment sufficient to violate Title VI. Instead, agency guidance contains the hedging language typical of the pervasively vague agency oversight described in Part I, in which even a single incident usually does not, but sometimes might, qualify as a Title VI violation.92

Total cutoffs of all federal grants are an unprecedented mechanism for enforcing CRSCs. Since the 1970s, agencies have rarely used funding cutoffs to resolve the tensions underlying CRSCs. It is a familiar point that federal agencies rarely terminate federal grants of any sort, for reasons ranging from reluctance to harm grant beneficiaries to the political influence of subnational politicians.93 Agencies have long regarded funding cutoffs as a blunt instrument.94 In any case, whatever the willingness of agencies to cut off funds, federal agencies have not found it necessary to do so in order to enforce CRSCs. In an interview with the author, veteran administrators from the Department of Education’s OCR observed that during their several decades at the agency, the agency was routinely able to obtain compliance through persuasion and negotiation—albeit possibly aided by the educational institution’s awareness of grant termination as an enforcement mechanism.95 A survey of decisions from the Department of Education’s Civil Rights Reviewing Authority shows only five funding-termination cases during the 1990s and none between 1998 and 2017.96

Why have agencies been so reluctant to use funding cutoffs to enforce civil-rights mandates prior to the Trump Administration? One reason might be that they are an unusually divisive methods for settling disputes about unusually divisive values.

First, the underlying values at stake are themselves particularly controversial. Disagreements about the prevalence of and appropriate remedies for racial discrimination sharply divide Republicans and Democrats and have dramatic effects on political behavior like voting.97 Moreover, disagreements over these questions are unusually fraught with emotion.98 Presidents elected by narrow majorities, therefore, enforce partisan interpretations of CRSCs only by overriding the passionately felt views of a large minority of Americans.

Moreover, total funding cutoffs are an unusually divisive method for enforcing CRSCs. While at least one scholar has advocated for the increased use of funding cutoffs,99 the tactic is a notably aggressive mechanism for resolving these polarizing disagreements. Funding cutoffs give preeminent power to a partisan executive rather than the less partisan judicial branch. Unlike private causes of action, agencies can enforce not only Title VI itself but also the agency’s interpretation of the statute embodied in rules and guidance.100 And unlike an action by the Department of Justice seeking an injunction to enforce agency rules in federal court, the agency can impose a funding cutoff through an administrative adjudication conducted by personnel internal to the agency.101 The aid recipient can appeal that cutoff decision to a federal court, but, like most agency adjudications, the agency’s decision is typically subject only to the usual “arbitrary and capricious” review in which the agency benefits from a presumption of regularity and factual support.102

Total funding cutoffs to enforce CRSCs, therefore, give presidents the power to force cities and states to adopt the view of civil rights endorsed only by a narrowly partisan majority of citizens with no plain justification in statutory text and no prescreening by the judiciary. On issues like racial equality, where rival views are matters of deep emotional conviction, such a power seems calculated to exacerbate partisan polarization. Because the relevant statutory terms are ambiguous, the centralized imposition of the divisive interpretation gratuitously denies equal concern and respect to rival and equally reasonable viewpoints.103 Such a power also raises the stakes of presidential elections by assigning to presidents the job of making radical transformations in civil rights based on a few percentage points of votes in battleground states. Intensifying the stakes of the presidency is unwise in a system already afflicted with political emotions so intense that they lead to political violence.104 Because this polarization is made possible by the longstanding vagueness of civil-rights mandates and permissiveness of courts with respect to incidental First Amendment violations, a new approach to CRSC enforcement is necessary.

III. title vi’s pinpoint provision as federalist compromise on crscs

There is an alternative to making the President the one-size-fits-all arbiter of civil rights: Courts could limit the President’s power to cutoff federal funds unilaterally when states and their subdivisions challenge presidential interpretations of CRSCs not plainly specified by statutory text. As explained below, such a limit on the President’s “cutoff power” has several virtues. First, it exploits an attractive virtue of federalism by broadening subnational power to adopt different policies on divisive issues. Second, such a statutory limit actually already exists in the text of Title VI and other statutes modeled on Title VI. The so-called “pinpoint” provision of Title VI limits the scope of any funding cutoff, thereby curtailing presidential power to impose novel readings of CRSCs that have not been tested in court.

A. The Federal Virtues of Decentralizing Polarizing Disputes About CRSCs

First, consider the federalism virtues of limiting presidential power over CRSCs. Allowing subnational governments to have more leeway to resolve the statutory ambiguities described in Part I would give each side, red and blue, some share of power, thereby advancing a vision of federalism as value pluralism.105 Different regions, states, and cities have different predominant attitudes towards race. Why not let each go its way on matters where federal statutes have nothing clear to say and any presidential view will likely be a partial and partisan posture that nearly half the nation will reject?

The obvious answer is that protection of racial equality has been a prerogative of the national government since the ratification of the Fourteenth Amendment. Diversity on matters of less urgency might be tolerable in a federal regime. As far back as James Madison’s Federalist No. 10, however, it has been conventional wisdom that subnational majorities are not likely to protect discrete and insular minorities from unified, racially or culturally homogenous majorities.106 That states and cities cannot be given a free hand to define for themselves the appropriate level of racial equality is suggested not just by political theory but also by statutory language. Title VI and other civil-rights statutes imposing CRSCs plainly require some sort of uniform baseline of racial equality in all federally financed programs and activities, although its language does not specify what that level ought to be.

Any federalist limit on presidents’ power to interpret Title VI, therefore, must ensure that grantees who fail to adhere to some minimum level of uniform national enforcement of racial equality in federally financed programs suffer some meaningful financial consequence. At the same time, the ambiguity of the statutory command outlined in Part I suggests that it should be geared towards constraining the means by which presidents can carry out their interpretation of the CRSCs rather than the substantive definition of racial discrimination that Title VI and analogous statutes do not provide.

Fortunately, Title VI’s so-called “pinpointing” provision contains the seeds of just such a means-based federalist compromise. That provision provides that any termination of funds “shall be limited to the particular political entity, or part thereof, or other recipient as to whom such a finding [of noncompliance with Title VI] has been made and, shall be limited in its effect to the particular program, or part thereof, in which such noncompliance has been so found . . . .”107 At a minimum, the plain text of this provision prohibits an agency from terminating all funds to a recipient organization merely because a “program” or “part” of that organization did not comply with Title VI. The “effect” of any loss of federal funds must instead somehow be limited (or “pinpointed”) to the noncomplying “program or part” of a recipient’s various programs and activities (hence, the slang term “pinpointing” provision).

By limiting the effect of a funding cutoff to the extent of noncompliance, Title VI’s pinpointing provision could theoretically limit the power of the President to impose controversial readings of Title VI on recipients without requiring any judicial evaluation of the substantive merits of those readings. In particular, enforcement of the pinpoint provision might render illegal many of the Trump Administration’s funding cutoffs without forcing courts to issue any opinion on difficult First Amendment or Due Process questions. Instead, recipients would sacrifice some funding proportional to the extent of their disagreement with the President, with small violations leading to proportionally small sacrifices. Small-scale violations—say, an administrator’s inadequate sanctions for students’ isolated anti-Semitic remarks within, for example, the School of Music—would be remedied by loss of only that School’s federal funding, leaving the rest of the University’s funding intact. Such a provision would function something like Guido Calabresi and A. Douglas Melamed’s liability rules: it would provide recipients with the option to buy their way out of executive demands at a price scaled to the burden imposed on the federal interest by noncompliance with the President’s reading of Title VI.108

B. The Pinpointing Provision’s Limit on Funding Cutoffs

Is such a reading of Title VI consistent with the history of how the pinpointing provision has been construed? That history is limited. Because until now the federal government has rarely tried to terminate funding to enforce Title VI, there are only a handful of judicial or agency interpretations of the pinpoint provision. Nevertheless, the sparse precedents that exist indicate that the federalism-promoting reading of the pinpointing provision described above is consistent with statutory text and purpose.

The most influential is Board of Public Instruction v. Finch.109 In Finch, the Fifth Circuit held that in order to terminate three separate federal grants to a school district that had failed to desegregate its teachers and students adequately, the Department of Health, Education, and Welfare (HEW) had to make specific findings that illegal racial segregation affected each of those grants.110 The Finch court observed that the federal grant for supplementary educational centers possibly “would have been used for a facility entirely separate from the rest of the school system.”111 Likewise, it was possible that “the grant for adult educational classes supported a program that was administered in an entirely desegregated manner even if the elementary and high school classes were not.”112 Finch, therefore, understood the provision’s language of “particular program” to refer to the particular federal statute that authorized the grant.113 A school district’s racially discriminatory use of funds authorized by one federal statute did not necessarily imply that the same recipient spent funds derived from a different statutory “program” in a racially discriminatory manner. Finch understood the pinpointing provision, therefore, as a way of insulating “innocent” expenditures of federal funds from blameworthy ones. “[T]here will also be cases from time to time,” Finch stated, “where a particular program . . . is effectively insulated from otherwise unlawful activities.”114 Given such “insulation,” Finch asserted that “Congress did not intend that such a program suffer for the sins of others. HEW was denied the right to condemn programs by association . . . . In this way the Act is shielded from a vindictive application.”115

Finch is worth a detailed exposition because its interpretation of Title VI’s pinpointing provision has been widely adopted as the authoritative reading.116 In particular, Congress itself seems to have acquiesced in Finch’s interpretation of Title VI when it enacted the Civil Rights Restoration Act to overrule the Court’s decision in Grove City College.117 In Grove City College, the Court held that a private college’s enrollment of students who received federal financial aid brought only those parts of the institution that received such aid under the coverage of Title IX.118 Because only the college’s financial-aid office directly received the funds, only that office was obliged to certify that it complied with Title IX. In response, Congress enacted the Civil Rights Restoration Act to amend the definition of “program” in Title VI, Title IX, Section 504 of the Rehabilitation Act, and the Age Discrimination in Employment Act.119 Under this new definition, a “program” receiving federal assistance means, among other things, “all of the operations of—a department, agency, special purpose district, or other instrumentality of a State or of a local government” or “a college, university, or other postsecondary institution, or a public system of higher education.”120 If a university receives a grant for its labs from the National Institutes of Health (NIH), for instance, then “all of [its] operations,” including, for instance, its athletic program, are covered by Title IX. That Title IX applies to all of the university’s operations, however, does not say anything about the sanction that the university should suffer for violating Title IX: coverage and sanction are simply two distinct legal questions. The legislative history of the Act indicates that Congress did not intend to disturb the limits imposed on sanctions by the pinpoint provision as understood in Finch when it enacted the Civil Rights Restoration Act to overrule Grove City College’s limit on coverage.121

The Finch test seems to provide a normatively attractive way to limit presidential power over fund cutoffs without staking out any position on the substantive question of which forms of discrimination are forbidden by Title VI. If a university provides an inadequate response to anti-Semitic actions by undergraduate students, for instance, then this failure would likely constitute national-origin discrimination under Title VI.122 That Title VI violation, however, would likely not allow the NIH’s Office of Civil Rights to terminate federal funds provided by the NIH for the university’s labs, unless the undergraduates at fault were somehow involved with those labs as, for instance, employees. The NIH, after all, provides funding for medical research pursuant to the Public Health Service Act,123 and this statute—the relevant “program,” under Finch’s understanding of the term—does not provide funds for undergraduate education. In Finch’s phrase, if the labs are “effectively insulated from otherwise unlawful activities” of the undergraduate athletic program, then “Congress did not intend that such a program suffer for the sins of others.”124

C. Refining the Finch Test to Promote Federalism

Finch’s requirement that federal programs be assessed statute by statute could substantially limit the Trump Administration’s power to strip away federal funds from an institution. But two ambiguities in the “pinpointing” idea dramatically affect how much it constrains the power to cut off funds.

First, how carefully must a recipient work to ensure that noncompliance in one part of its operations does not “infect” other parts? Suppose, for instance, that a single professor teaching a single undergraduate course makes anti-Semitic remarks sufficiently severe or pervasive to create a hostile educational environment in that course. Suppose also that the university receives federal aid for its undergraduate courses in the form of Pell Grants for low-income students enrolled in the undergraduate college.125 If the university fails to discipline the offending professor in a manner deemed suitable by the Department of Education’s OCR, then could OCR seek to terminate all Pell Grants for all undergraduates enrolled at the university? Suppose that no students receiving Pell Grants were actually enrolled in the course affected by a hostile educational environment: Could the university argue that its Pell Grant-funded “program, or part thereof” was, therefore, in full compliance with Title IX? Or could OCR respond that, because federally assisted undergraduates were eligible to enroll in the offending course, that course “infected” every other curricular decision that the students made?

Nothing in Finch answers this question about the degree to which fund recipients must quarantine noncompliant programs from compliant ones. Instead, Finch offers only the cryptic observation that “[i]f funds provided by the [federal] grant . . . support a program which is infected by a discriminatory environment, then termination of such funds is proper.”126

Second, nothing in Finch explains whether there is a requirement of proportionality between funding sanctions and noncompliance with CRSCs. Suppose that one student who received a Pell Grant was enrolled in a course affected by a hostile educational environment in violation of Title VI. Does this leakage of federal funds derived from a single federal “program” into noncompliant activities justify suspension of the entire Pell Grant program at the university? Finch does not say. Or should the loss of Pell Grant funds be reduced only by the proportion that was expended in a noncompliant manner? If the latter, then how should the burden of the fund cutoff be distributed? Should the hapless student who enrolled in the offending class lose his or her entire Pell Grant? Should the university be obliged to use its own revenues to continue the student’s financial aid previously supplied by the federal government? Or should every student at the university who received a Pell Grant lose some much smaller pro rata share of their federal financial assistance? Again, nothing in the pinpointing provision or Finch’s interpretation of it addresses the question of proportionality.

Moreover, because fund cutoffs have been so rare, precedents from either courts or agencies provide little guidance. Sometimes the Department of Education has deemed federal funds to be a “program, or part thereof, in which such noncompliance has been . . . found” merely when those funds financed programs that generated benefits for students in noncompliant programs. In deciding whether to cutoff grants from a school district, for instance, the Department of Education in Capistrano United School District held that the school district’s Chapter 2 funds, money earmarked for instructional materials and counseling, were part of a noncompliant program because they generally benefitted the district’s entire educational program, including the specific “Class and/or Co-curricular activity” at issue in the case.127 The fund cutoff, however, seemed only distantly connected to the particular violation of Title IX, which consisted of the district’s retaliation against a single teacher for alleging discrimination because of her sex.128 By contrast, in Maine v. United States Department of Agriculture,129 Judge Woodcock held that Maine’s alleged violation of Title IX in allowing transgender women and girls to play on women’s and girls’ athletic teams did not permit the Department of Agriculture to terminate Maine’s funding for federal food assistance provided pursuant to the Child Nutrition Program.130 Woodcock reasoned that the pinpointing provision “must mean that the repercussions of a particular program’s noncompliance should be experienced by that same program,” and the school lunch program was not the same as the athletic program.131 By the logic of Capistrano, however, one could conclude that federally funded school lunch program was part of a noncompliant program to the extent that school lunches benefited the entire school, including the allegedly noncompliant athletic teams.

In figuring out how to define pinpointing, therefore, courts write on a slate that is, if not blank, at least covered with little more than indecipherable scribble. That lack of clarity provides an opportunity for a court to construe Title VI’s pinpointing provision to avoid what Finch calls “vindictive application” of CRSCs. Two principles could provide such a constraint. First, pinpointing should require a principle of proportionality where possible. Second, pinpointing should require a presumption of severability.

“Proportionality where possible” would require agencies to scale funding cutoffs to the extent of noncompliance whenever the federal funding program allows funds to be broken into discrete units that can be identified with a noncomplying “part” of the program. Individualized student aid provides a good example of such possible proportionality. If the funding in question takes the form of aid to individual students, then the proportion of federally assisted students who are enrolled in a noncompliant program should be the measure of any cutoff of such federal financial assistance. In the hypothetical of the single student who both receives a Pell Grant and is enrolled in a course affected by a hostile educational environment, the reduction of financial aid to the noncompliant university should be no greater than the grant received by that student. A federal subsidy for the construction of a building in which such a course was taught, by contrast, cannot be broken into discrete compliant and noncompliant parts, so the university would have to repay the entire amount of the subsidy.

The “presumption of severability” would require that, where a federal program funds an activity (for instance, reduced-cost school lunches), the agency must overcome a presumption that the activity does not generate spillover benefits for noncompliant parts of the recipient’s activities (for instance, its athletic program). Judge Woodcock seems to have implicitly relied on this presumption in the Maine case. Such a presumption might be rebutted by specific evidence that the federal program funded the recipients’ noncompliant activities. For instance, the Trump Administration might be able to show that the federal school-lunch program somehow contributed to a school’s noncompliant athletics, because the school’s teams’ viability somehow depended on the availability of federally subsidized meals.

Unlike some statutes, Title VI and other statutes modeled on its language do not expressly codify these two presumptions of proportionality or severability.132 These principles nevertheless help make sense of Finch’s idea that CRSCs be “shielded from a vindictive application” in which small, episodic violations are leveraged into presidential control over an institution’s entire educational program. If “schools and programs are not condemned en masse or in gross, with the good and the bad condemned together,” as Finch urges, then it seems nonsensical to terminate billions of dollars because a few thousand, at most, were affected by a noncompliant part of a large institution. Likewise, if federal programs fund activities (say, eating lunch) different from the activities in which violations have been found (say, playing on a sports team), then the burden on the agency is properly to show that the activities are causally connected—in Finch’s language, that one “infects” the other. If Congress ratified Finch as part of the pinpointing provision, then the two presumptions offered here are sound readings of that provision.

Beyond capturing the meaning of Finch, however, the two presumptions offered above advance a brand of federalism defended above in Section III.A—a kind of federalism that reduces the stakes of acquiring or losing national power by reducing the influence of national politicians to the scale of their actual financial investment.133 Limiting funding cutoffs does not eliminate federal oversight over federally assisted activities, but it constrains unilateral presidential power in an era when presidents and their partisan opponents bitterly disagree with each other over the meaning of CRSCs. Where presidents are determined to force through their vision of civil rights by bringing actions seeking injunctions against grantees, they should embody that vision in the form of a binding rule of law—that is, a legislative rule rather than a mere guidance—enforceable by the Department of Justice. If they do not want to invest that political effort in pushing through a rule and instead proceed by guidance like OCR’s ubiquitous “Dear Colleague” letters, however, then their power to press the often eccentric and casually adopted nonbinding interpretations contained in such guidance through funding cutoffs should be significantly limited in terms of the money at risk.

As noted above, such limits transform presidential interpretations of CRSCs into liability rules in Calabresi and Melamed’s sense of the term: they preserve the power of recipients of federal aid to buy their way out of ideologically charged visions of civil rights with which they disagree.134 To the extent that federalism desirably preserves value pluralism, this limit on fund cutoffs is consistent with at least one attractive vision of federalism.135

“Federalism” is a notoriously slippery term, so it is useful to distinguish between two distinct ways in which pinpointing promotes “federalism.” First and most obviously, protecting subnational governments from unilateral presidential power to terminate federal grants confers autonomy on such governments to decide how best to carry out nondiscrimination mandates regarding their own operations. Second, to the extent that such limits liberate private grantees from presidential control, these limits give subnational governments correspondingly more space in which to regulate those private organizations. Presidentially imposed CRSCs can preempt subnational governments’ rival views about how best to promote civil rights in private organizations.136 The Trump Administration, for instance, has taken the extreme view that Title VI forbids even such race-neutral methods of pursuing racial diversity. 137 That interpretation of Title VI could impede a state legislature’s ability to require colleges to promote racial diversity by getting rid of policies with racially disparate impacts, because the state policy could cause private organizations to forfeit federal revenue.138 By contrast, if the pinpointing provision places meaningful limits on presidential power to terminate grants, then private organizations would be at correspondingly less risk of losing federal revenue because they complied with state laws that presidential interpretation of CRSCs would forbid.

Conclusion

The ambiguity of CRSCs has long allowed shifting and vaguely defined presidential interpretations, whether advancing “affirmative action” or attacking “illegal DEI.” Courts have in turn tolerated these broad federal mandates notwithstanding potential First Amendment concerns. Major swings in CRSC interpretation when enforced through total grant terminations inflame partisan polarization by imposing one faction’s contested vision of civil rights nationwide.

Title VI, however, carries within its text an obscure but potent limit on presidential imposition of factitious norms: Title VI’s “pinpoint” provision, which limits funding cutoffs to the specific federally financed program in violation, could counter unilateral presidentialism with pluralism. Construed according to principles of proportionality and severability, this provision would let states and cities reject controversial federal interpretations at a manageable financial cost. This approach would not gut civil-rights enforcement—private suits and DOJ injunctions remain—but it would prevent presidents from leveraging vague CRSCs to force partisan change on unwilling jurisdictions. By scaling federal leverage to the actual funds at stake, such a federalist reading could reduce the stakes of national elections, temper ideological conflict, and preserve space for regional diversity in areas where Congress has left key terms undefined, thereby turning Title VI from a tool of partisan escalation into a mechanism for managing deep national disagreement.

Roderick M. Hills, Jr. is the William T. Comfort III Professor of Law, New York University Law School. The author is deeply thankful to Howard Kallem and Sue Bowers, both of whom are both former officers in the Office of Civil Rights (OCR) in the U.S. Department of Education for instructing him on the practice of OCR regarding pinpointing and providing comments on earlier drafts and to Richard Pildes for his comments. None of these commenters are, of course, responsible for the author’s errors. The author is also grateful to Kevin Pollack and Sofia Costatino for superb research assistance.