Last week, “Vandy Catholic”—a Catholic student group at Vanderbilt University—reluctantly decided to leave campus rather than affirm its compliance with the University’s new “nondiscrimination” policy, which requires religious student groups (but, interestingly, not all student groups) to open membership and leadership positions to “all comers,” without regard to religion. (Michael Stokes Paulsen helpfully evaluated the Vanderbilt policy for Public Discourse readers a few weeks ago.)

Like other controversies involving, for example, the Boy Scouts, or the Christian Legal Society, the goings-on at Vanderbilt reveal a troubling confusion about “discrimination,” a confusion that, as it spreads, will undermine religious freedom, institutional pluralism, and civil society. This confusion travels with a deeply illiberal failure to appreciate that the kind of liberal democracy we should embrace is not “total” or “comprehensive”; in Lawrence Alexander’s words, it is not “liberalism all the way down,” and it does not insist that the rules that govern in the political sphere and context—non-discrimination, neutrality, “all comers”, etc.—need to, or even should, govern in other spheres and contexts.

Like the Supreme Court’s decision in Christian Legal Society v. Martinez—which involved the policy not of a private institution such as Vanderbilt, but the government’s own institution, the Hastings College of the Law—the “Vandy Catholic” incident reflects an oversimplification of and confusion about the idea of discrimination: about why and when it is wrong, and about whether and when a liberal constitutional government has the power to prohibit or discourage it.

The ongoing “Law and Religion” story in American constitutional law, political thought, and public debate includes several related, but distinguishable, plot-lines. There is, for example, the elaboration of the idea—from James Madison’s famous Memorial and Remonstrance Against Religious Assessments in 1785 to contemporary school-voucher controversies—that our understanding of religious liberty and the First Amendment’s no-establishment rule allow but limit financial and other forms of cooperation between government and religious organizations. In addition, episodes involving exemptions for Quakers from military-service requirements, prosecutions (and persecutions) of Latter-Day Saints for polygamy, and the peyote-use rituals of the Native American Church have been highlights (for better or worse) in the tale of religious believers’ efforts to secure accommodations and exceptions from generally applicable laws. School-prayer policies, holiday displays,and Ten Commandments monuments have figured prominently in the continuing effort to work out the role of religious expression, symbols, and arguments in the public square.

Start your day with Public Discourse

Sign up and get our daily essays sent straight to your inbox.

The story’s different sub-plots surface and recede over time. One will enjoy front-page-headline status for a time and then yield to another, not so much because its conflicts are neatly resolved or its questions clearly answered—though they might seem to be, for a little while—but because another captures, or recaptures, our interest.

This appears to be happening now, as our focus and attention are increasingly drawn to the dramatic tension between the desire and efforts of governments to combat invidious and irrational discrimination, on the one hand, and the constitutional and other limits on these governments’ power and ability to do so, on the other. How should we think about this tension?

We believe that “discrimination” is wrong. And, because “discrimination” is wrong, we believe that governments such as ours—secular, liberal, constitutional governments—should take steps to prevent, discourage, and denounce it. We are right to believe these things. The proposition that it is not only true, but “self-evidently” true, that all human persons are “created equal” is foundational for us. The principle of equal citizenship holds near-universal appeal, even though we often disagree about that principle’s particular applications.

At the same time, it is not true that “discrimination” is always or necessarily wrong. Nor is it the case that governments always or necessarily should or may regulate or discourage it—say, through its expression and spending—even when it is wrong. “Discrimination,” after all, is just another word for decision-making, for choosing and acting in accord with or with reference to particular criteria. We do and should “discriminate”—we draw lines, identify limits, make judgments, act on the basis of preferences—all the time. As Alexander has put it, “All of us well-socialized Westerners know that discrimination against other human beings is wrong. Yet we also realize, if we think about it at all, that we discriminate against others routinely and inevitably.” The practice is ubiquitous and unremarkable: We don’t blame someone for drinking Brunello rather than Boone’s Farm or for preferring The French Laundry to Arby’s.

It is an obvious point, but still worth making: It is not “discrimination” that is wrong; instead, it is wrongful discrimination that is wrong. It is tempting and common, but also potentially misleading and distracting, to attach the rhetorically and morally powerful label of “discrimination” to decisions, conduct, and views whose wrongfulness has not (yet) been established. After all, there is no reason for governments such as ours to ban, regulate, or disapprove “discrimination” generally, as opposed to discrimination that has been shown, with reference to factors other than the mere use of decision-making criteria, to be wrong.

In addition, for good reason we do not believe that governments should or may prevent, correct, or even discourage every instance of wrongful discrimination. Some wrongs and bad conduct are beyond the authorized reach of government policy; some are too difficult or costly to identify, let alone regulate; others are, put simply, none of the government’s business. It is true, again, that we are committed to “equal justice under law.” This aspiration is engraved on the West Pediment of the Supreme Court itself. But, the idea of “a limited state in a free society” is no less well-pedigreed in our tradition. It is, in Michael McConnell’s words, among the “most fundamental features of liberal democratic order”and is the core of the constitutional enterprise. “Constitutionalism,” I have written elsewhere, is “the enterprise of protecting human freedom and promoting the common good by categorizing, separating, structuring, and limiting power in entrenched and enforceable ways,”and not all antidiscrimination efforts will fit well within this enterprise.

What’s more, it is not only that over-enthusiastic or insufficiently deliberate campaigns against “discrimination,” in the name of “equality,” can conflict with, or even undermine, the fundamental and core idea of liberal, constitutional, and therefore limited government. There is also the need for an appropriate—not a paralyzing, but an appropriate—humility about our ability to identify confidently and to operationalize through law and policy the content of our ideal of social and political equality. One does not have to insist that the “idea of equality” is—in Peter Westen’s words—entirely “empty” to admit that it, and the nondiscrimination norm, are hotly and reasonably contested, and more easily admired than understood.

And so, to start again: When we say that “discrimination” is wrong, what we actually mean is that wrongful discrimination is wrong, and when we affirm that governments should oppose it, we mean that governments should oppose it when it makes sense, all things considered, and when it is within their constitutionally and morally limited powers, to do so. To label a decision or an action “discrimination” is simply to note that one factor or another was taken into account in the course of a decision; it is to invite, but not to answer, the questions whether that decision or action was wrong, and whether the public authority may or should forbid or discourage it.

Answering these questions requires careful consideration of many factors and variables: Who is the decision-maker? Who are the affected parties? What is the criterion-for-decision? How will the decision, and others like it, affect our ability to respect and vindicate other goods? How costly would it be to regulate or to try to prohibit such decisions? Is the social meaning of the particular decision in question such that it “belies the principle that people are of equal ultimate worth,” or is it something else? And, is the decision one that a “limited state in a free society” has the authority to supervise?

In other words, and as usual, context matters. It is not enough merely to report the occurrence of “discrimination,” or to invoke the ideal of equality, or to declare particular decisions “invidious” or “odious,” or certain criteria “suspect.” These terms communicate something important and troubling about certain instances of “discrimination,” but it is crucial to remember that they add something to what they modify.

So, what about “religion”? “Discrimination” that is motivated by religion, or that is engaged in by religious believers and institution, or that uses religion as a criterion, is sometimes wrong. The government should not do it, and it should—sometimes—discourage or regulate it. That said, “discrimination” on religious and other grounds is sometimes a dimension of religious liberty that governments may, and perhaps even must, accommodate or allow. Sometimes, a government such as ours will, may, and should regulate discrimination that targets religious status, or is motivated by religious belief, or is engaged in by religious actors. Sometimes, on the other hand, a government such as ours may, will, and should protect, or at least leave alone, such discrimination. Sometimes, it is wrong—wrong in a way that implicates the concerns of a liberal, constitutional government—for religious communities and actors to “discriminate.” But, sometimes, it isn’t. In the former type of case, such a government will want to respond and—so long as it is within its authority and it is not too costly, all things considered, to do so—probably should; in the latter type, however, such a government should be unbothered.

To say this is not to be blasé about the ideal of equal citizenship, or about the threats that both state and nonstate discrimination pose to it. We are deeply committed to the antidiscrimination norm and we believe that the coercive, expressive, and pedagogical functions of law should be deployed against wrongful discrimination. We are also, however, irrevocably (though it sometimes seems less deeply) committed to limited government and religious liberty. These commitments, it sometimes seems, are in tension, even in conflict, especially under conditions of social pluralism. We cannot avoid trade-offs, compromises, sacrifices, and prioritizing.

The rhetorical, moral, and legal power of the antidiscrimination norm can sometimes distort or distract our thinking about how we do and should protect religious freedom through law. The near-universal, if sometimes unreflective, conviction that “discrimination” is wrong means that assertions of religious freedom are sometimes heard as requests that political authorities tolerate a wrong—i.e., “discrimination”—which they would otherwise prohibit, penalize, or discourage. Such requests then raise the question of whether it is “worth it” for the authorities to do so—that is, whether doing so would complicate too much the government’s own projects, or conflict too glaringly with its values. And so, when they are granted, accommodations are regarded all around as concessions. Sometimes, to be sure, we do and probably should think about honoring rights in terms of protecting, or simply tolerating, a liberty to do even the wrong thing (so long as the wrong thing is not too wrong). Our free-speech decisions and doctrines provide many examples, including the Supreme Court’s recent rulings protecting depictions of animal cruelty, hateful funeral protests, and over-the-top-violent video games.

We should not forget, though, that one dimension of the freedom of religion is, sometimes, precisely the freedom to “discriminate,” and that this freedom should be protected not simply because such discrimination is an all-things-considered tolerable wrong—sometimes it is, sometimes it isn’t—but because it is inextricably tied to something good—a human right—and is, sometimes, beyond political authorities’ legitimate reach.