Expressive Discrimination: Universities' First Amendment Right to Affirmative Action — Part 3

Continuing my serial blogging on whether private universities can use a Boy Scouts expressive association theory to have race-based affirmative action.

Last week, I started blogging (and yesterday, I continued blogging) about my new article Expressive Discrimination: Universities' First Amendment Right to Affirmative Action, just published by the Florida Law Review. In these days of federal attacks on private DEI, maybe some private universities might find this useful as a strategy for fighting back against the Trump Administration!

Yesterday's post gave the main thesis, laying out under what conditions a Boy Scouts expressive association right might be viable under current caselaw. Today's post lays out some complications of the theory, answering four questions: (1) Does the race context make anything different? (No.) (2) Does the market context make anything different? (No.) (3) Does the Title VI conditional-spending context make anything different? (Yes, some.) and (4) Can public universities use this theory? (No.)

If you want to see the PDF with all the juicy footnotes, read the whole thing.

II. Some Complications

A. Is Race Different?

One might resist this whole argument by arguing that race discrimination is different from the sexual-orientation discrimination in Boy Scouts and 303 Creative.

But the racial angle shouldn't make any difference. The doctrine doesn't distinguish between race and other bases of discrimination. When one challenges governmental discrimination, Equal Protection doctrine draws strong distinctions—embodied in the applicable tiers of scrutiny—depending on whether the discrimination was based on race, sex, or something else. But that's not implicated here, because there's no state action in private universities' affirmative action programs. Here, the issue is compelled association, which (as discussed above) is like compelled speech.

Doctrinally, a person's liberty interest in saying something racist is the same as their interest in saying something anti-gay: the question is merely whether they're being made to say something they don't want to say. There's no doctrinal reason why 303 Creative would (or should) have come out differently if Lorie Smith had refused to make websites for interracial marriages.

And for associational freedom, an antidiscrimination law can burden a group's expression just as much when the group is being racially discriminatory as when it's being discriminatory against LGBT people. There's no doctrinal reason why Boy Scouts would (or should) have come out differently if it were a "Hitler Youth" organization that wanted to have white scoutmasters to teach white supremacy. (Or, in a more mainstream context, compare the racial theater and TV casting decisions discussed above with the gender-related decisions of someone making straight rom-coms or cisgender female beauty pageants.)

The cases cite each other without regard to the type of discrimination involved. Claybrooks v. ABC, recognizing a right to be racially discriminatory in TV show casting, relied heavily on Hurley—a case about an LGBT Irish-American group's participation in a parade—without considering whether race issues are different from LGBT issues. Similarly, Donaldson v. Farrakhan, finding a right to address a men-only audience, relied on Boy Scouts without discussing whether anti-LGBT discrimination is similar to sex discrimination (or even a type of sex discrimination, as later suggested by Bostock v. Clayton County).

Perhaps the difference isn't in the level of the burden on the speaker or association, but on whether the government would prevail under the relevant level of scrutiny. Can the government prevail more easily in a race case?

Recall that the appropriate level of scrutiny is somewhat unclear. If strict scrutiny is required, that raises two possibilities: (1) Would the government's interest in ending race discrimination be compelling, while the interest in ending sex or sexual-orientation discrimination isn't? (2) Does the race, sex, or sexual-orientation context affect whether narrow tailoring is present?

As to the first option, the Court has held that eradicating racial discrimination is compelling in cases such as Bob Jones University v. United States. But other interests have been held to be compelling too: Roberts talked about "Minnesota's compelling interest in eradicating discrimination against its female citizens." And the Roberts Court referred to discrimination more generally: A state's "commitment to eliminating discrimination and assuring its citizens equal access to publicly available goods and services . . . plainly serves compelling state interests of the highest order." In fact, "acts of invidious discrimination in the distribution of publicly available goods, services, and other advantages cause unique evils that government has a compelling interest to prevent."

For sexual-orientation discrimination, one could either rely on the broad Roberts formulation or think of sexual-orientation discrimination as a species of sex discrimination. The Tenth Circuit, in 303 Creative, stated (citing Roberts) that "Colorado has a compelling interest in protecting both the dignity interests of members of marginalized groups and their material interests in accessing the commercial marketplace."

The case law thus resists any distinction between eradicating different forms of discrimination in terms of the strength of the governmental interest. Any antidiscriminatory governmental purpose is likely to be held compelling.

And as to narrow tailoring, it seems that the degree of tailoring of any antidiscrimination law depends on how broadly or narrowly it sweeps, not on the basis of the discrimination that it prohibits. One could conclude that, given that eradicating the dignitary harms of discrimination is a compelling interest, the most narrowly tailored means of achieving that goal is a complete prohibition. Or one could conclude (as did the Tenth Circuit) that the governmental goal is ensuring access to services. If that's the case, it depends on how unique someone's services are; as long as one can get the same product elsewhere, one can afford to make exceptions, but as to unique products, a complete prohibition is the most narrowly tailored way of achieving the goal. These theories may be plausible, but they don't distinguish race from sexual orientation.

Or perhaps—sticking with the "equal access" goal—the degree of tailoring has something to do with the scope of the statute's exceptions. A complete prohibition might be minimally tailored, a prohibition with some targeted exceptions more tailored, and so on. Here, one might ask to what extent the availability of exceptions prevents the government from achieving its goal. The background societal level of discrimination against a particular group might thus be relevant in assessing whether an antidiscrimination law is really the government's least restrictive alternative; so would the degree of competitiveness or "monopoly power" in a particular industry.

Perhaps—one might argue, under this view—a racial-discrimination prohibition must be super-strict, because only an absolute law can succeed in combating racial discrimination. Perhaps a prohibition against sexual-orientation discrimination can afford to not be so absolute, because anti-LGBT discrimination isn't as widespread and thus LGBT people are more likely to find someone willing to deal with them. (Even if one custom website designer won't deal with them, another custom website designer will; the uniqueness of that particular website designer's product wouldn't be relevant.)

But one problem with that theory is that the cases don't ask that question. Another problem is that even if the relative degrees of animosity toward different groups were relevant, it's not clear which way that cuts in today's society. What's more widespread today, anti-LGBT animus or racial animus? Surely interracial couples can more easily find a wedding-website designer than can same-sex couples. And if that's the case, a principle that allows for anti-LGBT expressive exceptions should apply at least as strongly in the racial case.

B. The Market Context and the Level of Scrutiny

1. Maybe This Doesn't Apply in the Market Context?

Last year, one could have objected that Boy Scouts has never been extended to the commercial context. The Boy Scouts is a nonprofit, volunteer organization. Hurley is about a parade—likewise a nonprofit, volunteer organization. Avoiding compelled speech is all well and good, and so is expressive association, but should antidiscrimination law give way even when commercial relations are involved?

Perhaps one could have argued that universities are still covered by Boy Scouts because they're generally nonprofits. But perhaps not: there's nothing magical about nonprofits (the Jaycees was also a nonprofit), there's no sharp distinction between how nonprofit and for-profit firms act, and there's no strong difference between how nonprofits and for-profits are treated in fields such as antitrust. Thus, despite universities' expressive nature, one might think the commercial, rather than volunteer, aspect is primary. Student admissions and faculty hiring are about commercial relations—for faculty, this is about one's very livelihood, and for students, it's about a multi-year contract, perhaps hundreds of thousands of dollars of debt, and preparation for a career.

In arguing that commercial relations are different, one could have looked to various sources, though they wouldn't all have had the same implications for universities.

First, one could have looked at Runyon v. McCrary, which denied a private school's right to discriminate against black students. Some might find this the closest case on point to university affirmative action, since it involved a school (though, admittedly, a for-profit school) trying to choose its students based on racial criteria. But, as noted above, Runyon didn't squarely present the issue.

Second, one could have looked at Justice Powell's concurrence in Runyon. Justice Powell thought that (as a statutory matter) § 1981 applied because the school there was "operated strictly on a commercial basis" and was "part of a commercial relationship offered generally or widely." He distinguished that school from "[a] small kindergarten or music class, operated on the basis of personal invitations extended to a limited number of preidentified students." Universities would certainly fall on the commercial side of Justice Powell's line, though Justice Powell wasn't making any constitutional point.

Third, one could have looked at Justice O'Connor's concurrence in Roberts. Justice O'Connor argued that expressive-association rights should belong to expressive associations, not commercial ones. Expressive associations, she wrote, should get to define their membership. "[T]he formation of an expressive association is the creation of a voice, and the selection of members is the definition of that voice"—and "state regulation of its membership will necessarily affect, change, dilute, or silence one collective voice that would otherwise be heard." But freedom of commercial association, she wrote, is only minimally protected.

Justice O'Connor's view would probably cut in favor of the university's First Amendment claims: universities are predominantly engaged in expression, and—especially if they have a strong race-related mission statement and choose their members to inculcate their values—state regulation of their membership could dilute their message. Though, it's hard to say for sure, because a lot of money is changing hands and the relations are contractual. Green v. Miss USA illustrates the range of possible disagreement here: is a for-profit beauty pageant an expressive association that promotes certain ideals of beauty and "female role models," or is it a glorified "multi-level marketing business"?

Fourth, one could have looked at the expressive-boycott case law. The basic rule is that boycotting activity isn't protected by the First Amendment, whereas the expression surrounding boycotts is. But for a while, some language in FTC v. Superior Court Trial Lawyers Ass'n provided some support for an economic/noneconomic distinction. There, the Court ruled against striking court-appointed defense attorneys for indigent defendants, and it distinguished its holding from its pro-boycotters holding in NAACP v. Claiborne Hardware Co. In the Court's view, what made the NAACP boycott different was that it was for "[e]quality and freedom"; the lawyers' strike/boycott, by contrast, was an economic boycott whose "immediate objective was to increase the price that [the strikers] would be paid for their services," no matter their "altruistic . . . motives." Regulation of such "economic boycotts" was well within "the government's 'power to regulate . . . economic activity.'"

Perhaps "the better reading" of these cases "is that many but not all elements of political boycotts are expressive" and that the economic/noneconomic distinction doesn't really play a role. Still, even if it's a misguided dictum, the Superior Court Trial Lawyers language is there; what do we do with it? Universities fit uneasily into this dichotomy because their pursuit of affirmative action is grounded in ideas of equality and freedom, even though that plays out through economic activity. (In any event, Rumsfeld v. FAIR ultimately made clear that boycotting activity is unprotected—not because of any economic/noneconomic distinction but because boycotting simply isn't expressive activity.)

2. The Possible Irrelevance of Markets

So if we look at the older materials, we're left with some uncertainty about whether the Boy Scouts theory applies in a market context at all and, if it doesn't, what that means for universities' hiring and admissions decisions. But today, after 303 Creative, any economic/noneconomic distinction is harder to maintain.

First, as a matter of first principles, compelled-speech and expressive-association claims are no less relevant in economic contexts than in noneconomic ones. If the NAACP wanted to be run by a black person, the American Nazi Party by a white person, or the National Association of Women by a woman, it's an equal burden on those groups' expression, whether those directors were paid or volunteer. If—as a writer or director—I want to have race-conscious casting in Othello or Hamilton, it seems like an equal imposition on my artistic vision whether or not the actors are employees. This is so even though we're talking about someone's livelihood.

Second, various cases applying this framework over the years have come out in favor of the First Amendment claimant even though economic transactions were involved. In Claybrooks, the district court upheld the right of The Bachelor and The Bachelorette to discriminate based on race in casting, even though participants were paid a stipend. In Miss USA, the Ninth Circuit upheld the right of a beauty pageant to limit itself to cisgender women, even though winners received valuable prizes. Hurley, too, noted that protections against compelled speech apply generally—they're not just for the press but are also "enjoyed by business corporations generally and by ordinary people engaged in unsophisticated expression as well as by professional publishers."

Third, there's now a decades-old tradition of cases upholding First Amendment rights even in the economic/commercial/corporate context—for instance, in campaign finance, in commercial speech, in other deregulatory contexts, and in the general tendency to treat corporations as rightsholders. This civil rights expansion, recognizing that people don't check their free-speech rights at the door when they do business or use the corporate form, has progressed unabated—even in the face of persistent complaints about First Amendment Lochnerism and "weaponizing" the First Amendment. The Court has written that "[i]t is too late to suggest 'that the dependence of a communication on the expenditure of money itself operates to introduce a nonspeech element or to reduce the exacting scrutiny required by the First Amendment'"—and that "too late" was over forty-five years ago.

Finally, we now have 303 Creative, where Lorie Smith operated through a for-profit corporation. What difference did this make? None at all:

Of course, as the State emphasizes, Ms. Smith offers her speech for pay and does so through 303 Creative LLC, a company in which she is "the sole member-owner." But none of that makes a difference. Does anyone think a speechwriter loses his First Amendment right to choose for whom he works if he accepts money in return? Or that a visual artist who accepts commissions from the public does the same? Many of the world's great works of literature and art were created with an expectation of compensation. Nor, this Court has held, do speakers shed their First Amendment protections by employing the corporate form to disseminate their speech. This fact underlies our cases involving everything from movie producers to book publishers to newspapers.

Justice Sotomayor, in her dissent, argued that public accommodations laws reflect a "simple, but powerful, social contract" where one agrees to serve the public at large by choosing to sell to the public at large. She characterized the Hurley and Boy Scouts regime as applying "to private, nonprofit expressive associations" and not to "clearly commercial entities" such as 303 Creative LLC, and she argued that "any burden on the company's expression is incidental to the State's content-neutral regulation of commercial conduct." The significance of 303 Creative lies in how clearly and forthrightly the majority rejected the relevance of the public accommodations context and the commercial/noncommercial distinction championed by Justice Sotomayor.

303 Creative strongly undermines any contrary hints that one might glean from Runyon or Roberts. It's true that "the First Amendment does not generally protect liberty of contract." But when we have a regulation that would violate the First Amendment when applied to speech done for free, 303 Creative strongly supports the argument that the regulation also violates the First Amendment when applied to speech done for money.

3. Strict Scrutiny Analysis

There still remains the possibility that a ban on discrimination in an economic context would satisfy the relevant level of scrutiny, while a ban on discrimination in a volunteer context wouldn't. As above, note that there's some uncertainty about what the level of scrutiny truly is. Perhaps it's strict scrutiny, perhaps it's more; but the best evidence as to how that scrutiny would come out is 303 Creative itself, which treated the commercial context as irrelevant. This makes it unlikely that eradicating discrimination in the economic context would be any more likely to survive than eradicating discrimination in the volunteer context.

C. Title VI and Conditional Federal Spending

Educational contracting—for faculty or students—is governed by regulatory statutes such as § 1981 or Title VII. Suppose the Boy Scouts theory overcomes these statutes because forcing universities to accept members they don't want would interfere with their message. Universities would still have to contend with non-regulatory statutes like Title VI, under which they could lose their federal funding. (Harvard was, in fact, sued under Title VI, which has, like § 1981, been interpreted to be coextensive with Equal Protection.)

Under current law, universities will not be entitled to an exemption from Title VI, even if the courts accept the Boy Scouts theory—because the government isn't forcing you to do anything when it induces behavior by offering conditional funding. This will be a significant issue for most universities, since most universities receive federal funding, even if just in the form of federal financial assistance for their students.

Originally, the threat of losing funding was relatively mild. As a statutory matter, the Supreme Court held in Grove City College v. Bell that sex discrimination would only put a limited amount of university funding at risk under Title IX of the Education Amendments of 1972—the "program or activity" implicated by student financial aid would be only the student financial aid office. But Congress overruled this interpretation in the Civil Rights Restoration Act of 1987 (CRRA)—not only with respect to Title IX but also with respect to a number of other antidiscrimination laws, including Title VI—by specifying that "program or activity" means "all the operations" of a university. So now all of a university's funding is jeopardized by any discrimination in any unit. And few universities would agree to adopt affirmative action at the cost of losing all their funding.

The good news for universities, though, is that while the government does have some leeway to induce behavior through funding, the unconstitutional conditions doctrine prevents the government from withdrawing a university's entire funding based on discrimination by just one unit. Thus, universities should have the ability to make an informed choice to adopt affirmative action—and thus lose federal funding—in certain units; perhaps, for instance, a university might choose to adopt affirmative action in a particular unit that doesn't receive a lot of federal funding and can therefore afford to lose it.

1. The Unconstitutional Conditions Doctrine

People constantly waive their rights as a condition of receiving government benefits. For instance, we all have a First Amendment right not to talk about constitutional law, but constitutional law professors waive that right if they want to make money as constitutional law professors at public universities—and the same is broadly true for one's on-the-job speech whenever one accepts a government job.

The case law is clear that there's nothing inherently invalid about such inducements, even when what's being induced is something the government couldn't mandate directly. "[A] legislature's decision not to subsidize the exercise of a fundamental right does not infringe the right." After all, one is always "free to decline the federal funds." In the case of expressive association, "That the Constitution may compel toleration of private discrimination in some circumstances does not mean that it requires state support for such discrimination." "Congress is free to attach reasonable and unambiguous conditions to federal financial assistance that educational institutions are not obligated to accept." Thus, in various post–Boy Scouts cases, courts have allowed governments to refuse to subsidize the Boy Scouts unless it stopped discriminating.

The same is true throughout constitutional law. Indeed, there's a certain logic behind one's ability to bargain away one's rights. If constitutional rights represent areas of personal autonomy, the right whether to exercise a right and the right not to exercise a right are as fundamental as the right itself—and if one can choose not to exercise one's right, it makes sense that one should be able to agree not to exercise that right for whatever reason one likes, including in exchange for something of value.

And yet, constitutional law imposes limits on the government's ability to make such deals. Despite Justice Oliver Wendell Holmes Jr.'s quip that a policeman "may have a constitutional right to talk politics, but he has no constitutional right to be a policeman," modern constitutional law doesn't adopt the extreme laissez-faire attitude that the government can offer whatever deals it pleases. And perhaps for good reason: the government has such overwhelming control over access to important discretionary benefits that it could induce whatever behavior it likes if given the chance—achieving indirectly what it couldn't achieve directly through regulation.

Thus, various cases have asserted that "a government may not require an individual to relinquish rights guaranteed him by the First Amendment as a condition of public employment," and that "the government may not deny a benefit to a person on a basis that infringes his constitutionally protected . . . freedom of speech even if he has no entitlement to that benefit." Of course, the idea that "the government has no duty to subsidize" and the idea that "the government can't require you to give up any rights as conditions of benefits," in their strong forms, are irreconcilable, so these statements shouldn't necessarily be taken at face value; the point here is just to show that there are some constitutional restrictions on the government's ability to condition funds.

This "unconstitutional conditions doctrine" is context-dependent, and in fact there's no single coherent doctrine that applies throughout all areas of constitutional law. But one persistent theme is that the conditions imposed need to be related, or germane, to the purpose of the government program. For government employees, the content of one's on-the-job speech is usually maximally relevant to the government's purpose, while off-duty speech is usually much less so. When public-school students are involved, we weigh the student's interest in speaking against the school's interest in suppressing the speech—but if the speech would have no effect on school operations, presumably that balancing would always cut in favor of the student because restricting the student's speech would be unlikely to be germane to any valid school interest.

2. The Significant Independent Grants Principle

Germaneness isn't a problem here, since there's a connection between federal educational funding and Title VI compliance: if you're going to take federal money to educate people, the federal government can demand that the money be spent evenhandedly, without racial discrimination. (The same goes for Title IX and other conditional-funding statutes.)

But the unconstitutional conditions doctrine is subject to an important constraint: the "significant independent grants" principle. This principle has been important in federalism cases involving the Spending Power, where the Court has stated that while Congress can withdraw program funds if the states don't agree to administer its program, it can't put "significant independent grants" at risk. A similar principle applies in First Amendment cases.

Grove City College and Rumsfeld v. FAIR noted that conditions need to be "reasonable," but they didn't go into further detail. Other cases have been more specific, though. Eugene Volokh summarizes the doctrine: "While the government may generally place conditions on the use of benefits that it provides, it generally may not control the use of the recipient's other assets as a condition of providing the benefit. We might call this the No Governmental Restrictions on Use of Private Funds Principle." For instance:

• In Regan v. Taxation with Representation, the Court held that the government could deny § 501(c)(3) tax deductions to organizations that engaged in lobbying. Nothing prevented those groups from creating a separate lobbying affiliate under § 501(c)(4).
• In FCC v. League of Women Voters of California, the Court struck down a prohibition on editorializing by stations that received federal funds. The problem here was that stations weren't even allowed to editorialize with their own funds; that problem would be fixed if the stations could establish separate affiliate organizations.
• In Rust v. Sullivan, the Court upheld the prohibition on granting federal funds under Title X of the Public Health Service Act to "programs where abortion is a method of family " The organizations that received funding were allowed to advocate abortion in separate activities that maintained "objective integrity and independence" from Title X projects.
• In Agency for International Development v. Alliance for Open Society International, the Court struck down a funding condition providing that anti-HIV funding couldn't go to organizations unless they had "a policy explicitly opposing prostitution and sex trafficking." It wasn't just that organizations couldn't use those funds "to promote or advocate the legalization or practice of prostitution or sex trafficking"—the statute even controlled what recipient organizations did with their own money, separate from that funding program. The Court noted that a condition can't "seek to leverage funding to regulate speech outside the contours of the program itself," endorsing the "dual structure" theory of Taxation with Representation and the segregated project theory of Rust v. Sullivan.

This distinction doesn't help an organization if it wants to choose its top-level administrators in a discriminatory way, because that choice of leaders necessarily affects the entire institution. So if a university wants to keep federal funding, it should choose its high-level decisionmakers nondiscriminatorily—here, that means the president, provost, and all other university-wide officials.

But under this distinction, a university would be able to set up separate departments or units—for instance, a law school—keeping the units where it intends to keep federal funding tightly segregated from the ones where it has no federal funding (or where it's ready to give up funding). The university should set up appropriate firewalls. If the funding is received in the form of aid to particular students, that might mean those students shouldn't be allowed to enroll in the discriminating departments. Within individual units, the university should be able to discriminate under the Boy Scouts theory. If the federal government insists on regulating the discriminating law school just because the nondiscriminating medical school gets federal funding, then it will be violating the No Governmental Restrictions on Use of Private Funds Principle from League of Women Voters and Alliance for Open Society.

D. Can Public Universities Use This Theory?

Students for Fair Admissions treated public and private universities the same, as far as their ability to discriminate based on race is concerned. However, the First Amendment expressive-discrimination workaround described here is available only to private universities. The reason is simple: because of the state-action doctrine, private universities have virtually no constitutional constraints. Therefore, the only constraints that prevent private universities from having affirmative action programs are statutory—in the federal context, Title VI and § 1981. And the Boy Scouts theory relies on the idea that a constitutional right such as the expressive-association right trumps any mere statutory right.

Making a similar argument for public universities would be trickier. First, as governmental organizations, public universities are often thought of as lacking First Amendment rights, which are generally thought to apply only to private parties (whether individuals or collective entities such as corporations). Second, even if public universities have First Amendment rights, they're still state actors, so they're also bound to respect the Equal Protection rights of the people it deals with. We would then have to resolve the conflict between two competing constitutional rights—which would be trickier.

This Section explores potential ways for public universities to get around the problem. First, perhaps public universities (and other government bodies) have First Amendment rights of their own. Second, perhaps universities—public or private—are "First Amendment organizations" that deserve particular deference in their academic decisionmaking. Third, perhaps public universities, because of their noncoercive nature, either shouldn't be considered state actors at all or should be subject to a reduced version of constitutional constraints.

But all these theories have little to no support in current doctrine. So this Section is more of a thought experiment.

1. Governments as First Amendment Rightsholders

Whether governments have free-speech rights is an unresolved question of First Amendment law. Most courts—usually in dicta—have "reflexively rejected the notion, relying on the assumption that the First Amendment can only restrict, not protect, state actors." As Justice Potter Stewart wrote in his concurrence in CBS v. Democratic National Committee, "The First Amendment protects the press from governmental interference; it confers no analogous protection on the Government." For the contrary view, one could point to some California cases; the Bakke/Grutter suggestion of deference to university judgment; and Justice Stevens's dissent in United States v. American Library Ass'n, where he argued that federal conditions on municipal library funding violate the unconstitutional conditions doctrine. The Supreme Court has noted but not decided the question. So far, we don't have anything authoritative at the level of federal law.

Paul Secunda suggests that assuming that the law school members of the Forum for Academic & Institutional Rights—including the public law schools—are expressive associations may have inadvertently opened the door to public-university employment-discrimination rights. But the Rumsfeld v. FAIR analysis is hardly a holding that public universities are expressive associations, and Secunda himself believes that the Supreme Court would reject this suggestion if push came to shove.

Most recently, in Walker v. Texas Division, Sons of Confederate Veterans, Inc., the specialty-license-plates case, the Court tantalizingly suggested the existence of state-government First Amendment rights: "[J]ust as Texas cannot require [Sons of Confederate Veterans (SCV)] to convey the State's ideological message, SCV cannot force Texas to include a Confederate battle flag on its specialty license plates." But, again, this was just dictum.

That public universities may have First Amendment rights is an intriguing theory, and beyond the scope of this Article. But even if we granted public universities First Amendment rights, we would still have to resolve more issues.

First, a state university with First Amendment rights would be able to assert the expressive-association theory against federal antidiscrimination laws such as Title VI or § 1981. But the state university might not be able to assert its rights against state law—presumably the state can control its own organizations. (It might be different if the state constitution grants the university system some independence from the rest of state government.) Nor could a federal university (such as West Point or the University of the District of Columbia) assert its rights against federal law.

Second, public universities, as state actors, are bound by Equal Protection, including the holding in Students for Fair Admissions. So, any expressive-association right would have to be balanced against the Equal Protection rights of whoever is harmed by an affirmative action policy. That would be tougher than in the private case, where a constitutional right is up against a mere statute.

And in a matchup between the public university's expressive-association right and the people's Equal Protection rights, the former might not do very well. Recall that Grutter, though it didn't explicitly examine the university's First Amendment right, did (following Justice Powell's Bakke concurrence) connect universities' interests with "the expansive freedoms of speech and thought associated with the university environment." Despite that, Grutter still insisted on strict scrutiny. (And the same was true in some post-Grutter cases, even though they deferred to universities' judgment.)

Who knows, maybe it would have been different if the Court had entertained the idea of universities' First Amendment rights rather than mere First Amendment–inspired interests. But based on hints from Grutter, if public universities had any First Amendment–inspired scope of action with respect to affirmative action, it was limited.

Justice Anthony Kennedy, in his dissent, similarly broadly endorsed Justice Powell's First Amendment rationale, but insisted on "rigorous judicial review" and "strict review"—which, in his view, prohibited the affirmative action program. Justice Clarence Thomas's separate opinion denied that "the Law School's assessment of the benefits of racial discrimination and devotion to the admissions status quo [were] entitled to any sort of deference, grounded in the First Amendment or anywhere else." In his view, there was "no basis for a right of public universities to do what would otherwise violate the Equal Protection Clause."

The combined views of Justices Kennedy and Thomas now seem to have carried the day: the only place the First Amendment shows up in Students for Fair Admissions is in Justice Sotomayor's dissent.

2. Universities as "First Amendment Institutions"

Perhaps it's not enough to vaguely gesture, Justice Powell–style, to the importance of universities in free expression. Perhaps one should take seriously that "[e]ducation is different" and build on the tradition of judicial deference to universities' decisions in the First Amendment context (e.g., in Grutter) and elsewhere.

Paul Horwitz argues that one should explicitly recognize universities as part of a unique set of institutions (including "the press, religious associations, libraries, and others") that "play a fundamental role in our system of free speech"; in his view, they are "First Amendment institutions" and should be treated differently as a doctrinal matter. Such institutions should be "granted significant presumptive authority to act, and courts [should] defer substantially to actions taken by those institutions within [their] sphere of autonomy."

Universities are well-suited for such treatment because they're easy to recognize, are environments for pursuing truth and finding meaning, operate within highly structured "disciplines," comply with "scholarly standards," and are characterized by "myriad bureaucratic rules" and other "structural mechanisms." (This recognition needn't be pro-free-speech: Horwitz argues that courts should even uphold a university's speech restrictions if they're grounded in the university's values related to "the nature of desirable speech on campus" or the university's "academic mission.") In the strongest form of "First Amendment institution" university deference (which Horwitz doesn't favor and which certainly isn't the law), universities could be given "near-absolute discretion" in areas including "the composition of the student body based explicitly on considerations of race or gender."

This "universities as First Amendment institutions" view would be a departure from current doctrine; the current view is what Horwitz calls the acontextual or "institutionally agnostic" view, which embodies, at most, what he calls "weak-form treatment of universities as First Amendment institutions." Moreover, if a university significantly abandons academic-freedom and free-speech norms to fit in the Boy Scouts theory, the "First Amendment institutions" theory might not even apply on its own terms.

3. Noncoercive Competitors as Not State Actors

Finally, one might question whether public universities should be considered state actors at all.

Usually, private parties aren't considered state actors and don't need to observe constitutional rights. For instance, private schools aren't state actors—even if their money largely comes from public contracts. But private parties can be considered governmental if they have clear indicia of publicness, such as fulfilling a traditionally exclusive public function or somehow being entwined with the public sector. Thus, private prisons are always state actors—at least with respect to their prisoners—because they perform a coercive function that's unavailable to ordinary private people.

On the public-sector side, though, the rule is much more automatic: public employees and formally public entities (i.e., ones that are part of the traditional apparatus of state government) are virtually always considered state actors. One exception to this principle is public defenders, because their relationship to their clients is similar to any other lawyer–client relationship, because they need to be independent, and because they're in a role that's adverse to the government. But that's a narrow exception: otherwise, public employees are classified as state actors with no further analysis.

Do we need to do this? If we can follow a functional analysis, whereby nominally private parties can sometimes be considered state actors if they act in a way that's somehow governmental, why not follow the same analysis in reverse and sometimes (beyond the narrow context of public defenders) consider nominally public actors as though they were private? Perhaps "the automatic treatment of all actions of the government as 'state action'—or at least as all equally state action—should be qualified in favor of a more thoroughgoing functionalism."

If we did this, a good candidate for non-state-actor status would be public universities. They're non-coercive (nobody attends or works at a university unless they want to), their relationship with students or faculty is regulated by contract, and public universities compete with (and can sometimes be hard to distinguish from) private universities. Moreover—perhaps like public defenders—academic-freedom and free-speech norms suggest that professors and students are expected to be independent from (and sometimes adverse to) the government. And people at universities often have their own ethical norms, from generalized academic-freedom norms to field-specific codes such as legal or medical ethics. So perhaps public universities should be considered just like private universities—and thus have First Amendment rights, including expressive-association rights.

This slightly resembles Richard Epstein's argument in favor of a "flexible standard that measures the constitutionality" of public-university affirmative action, when done on those universities' own initiative, "by comparing it with the practices done by private, competitive institutions on a voluntary basis." Public universities' "resemblances to private universities are far greater than those to public police forces"; public universities should therefore "be able to imitate the set of practices undertaken by private universities."

Epstein isn't denying that universities are state actors, though; under his view, public universities would merely be subject to a sort of reasonableness review where the standard is the prevailing practice in comparable private-sector institutions: "[W]e take as the measure of the reasonableness of the state action the question of whether it follows the voluntary patterns and choices made by the private institutions with which it competes."

Epstein's argument—a sort of "constitutional variation of the business judgment rule"—is one way to allow public-university affirmative action; the no-state-action rule is another. Both are premised on the idea that public universities are essentially voluntary organizations that compete with and resemble private universities. However, neither of these positions has much support in current doctrine. I'm sympathetic to the no-state-action position (and thus the expressive-association rights of public universities)—and, more generally, I would welcome a functionalist rethinking of state action. But that rethinking would be a fundamental change in state-action doctrine.

[Note: this serial-blogging will continue in later posts.]