Symposium

Protecting free exercise under Smith and after Smith

This article is part of a symposium on the court’s decision in Fulton v. City of Philadelphia.

Thomas C. Berg is the James L. Oberstar professor of law and public policy at the University of St. Thomas (Minnesota). Douglas Laycock is the Robert E. Scott distinguished professor of law at the University of Virginia. 

Fulton v. Philadelphia is an important win for religious liberty. Philadelphia may not terminate its foster-care services contract with Catholic Social Services on the ground that CSS declines, because of its religious beliefs, to certify same-sex couples as foster parents. Teachings about sex and marriage are central to many religions; so are works of service. If religions lose the ability to serve because they act on their central teachings, the harm to free exercise is severe. The court prevented that here — and the result was unanimous.

Fulton applied the rule of Employment Division v. Smith: A law may burden religion if it is neutral and generally applicable, but if not, then the burden on religion must be justified by a compelling government interest. Fulton clarifies Smith in ways that strengthen protection.

The court made clear that general applicability is a separate requirement from neutrality; both must be satisfied. It held that a rule flunks general applicability when it gives officials discretion to grant exceptions, even if the officials never grant any: The discretion enables discrimination against religion. Nor can government discriminate just because it’s setting rules for its contractors rather than regulating the general public.

Fulton also makes clear that civil rights laws do not automatically and in every context serve a compelling government interest. Importantly, the liberals joined this holding.

Those points are significant. But the holding on general applicability turns on specific features of Philadelphia’s rules. Cities can rewrite their rules, eliminating discretionary exceptions, and perhaps satisfy general applicability.

The holding’s limits drew attack from Justice Samuel Alito, who (joined by Justices Clarence Thomas and Neil Gorsuch) argued that the court should overrule Smith and strictly scrutinize generally applicable laws. Justice Amy Coney Barrett, joined by Justice Brett Kavanaugh, wrote separately that “it is difficult to see why the Free Exercise Clause … offers nothing more than protection from discrimination.” So five justices said that Smith was mistaken, and there may be more.

Barrett and Kavanaugh followed Smith here because, they said, they’re uncertain what would replace it. They did not need to overrule it; the general-applicability ground was available. But some cases will rest primarily on challenging Smith, including a pending cert petition by a construction contractor who was denied a state license because he had religious objections to a requirement that he provide his Social Security number.

The court can overrule Smith before it resolves every follow-on issue. But we want to begin to address Barrett’s questions. We think the compelling-interest test should usually govern when a generally applicable law substantially burdens religion. That test, which applies to substantial burdens on several other fundamental rights, properly holds that only the prevention of significant harm can justify prohibiting religiously motivated conduct.

The compelling-interest test need not govern every situation. Laws that substantially interfere with religious organizations’ internal governance decisions, like their selection of leaders, are absolutely barred under the Hosanna-Tabor decision, which Barrett mentions. But we don’t think the test should be substantially weaker than “compelling interest.”

Barrett notes that the court has used a “more nuanced” approach than strict scrutiny when generally applicable laws affect speech or assembly. She may be referring to United States v. O’Brien, which applied intermediate scrutiny so weak that the court accepted a barely rational basis for punishing a protester who burned a draft card. But another expressive-conduct decision, Boy Scouts of America v. Dale, used strict scrutiny to hold that the Boy Scouts could not be forced to accept an openly gay scoutmaster. The court said that the nondiscrimination law in Dale “directly and immediately affects associational rights,” while the draft-card law “only incidentally affects the free speech rights of those who happen to use a violation of that law as a symbol of protest.”

The difference the court pointed to appears to be that a prohibition on symbolic conduct leaves open many other ways to express the same views. The court also allows content-neutral restrictions on speech if — but only if — they leave adequate alternative channels of communication. Prohibitions on religious practice are usually more like the law in Dale: They leave open no other way to follow the practice in question. If you face a substantial penalty for acting consistently with your religious tenet, it’s no answer to say you can still follow other tenets. If you are blocked from pursuing a form of religiously motivated service — like CSS placing foster children, or religious progressives giving food and water to undocumented migrants — it’s no answer to say you could do a different form of service. Religious practices are not fungible, and assessing whether they are close enough would involve courts in difficult religious judgments based on a mistaken premise.    

Barrett cited a pre-Smith decision, Gillette v. United States, that spoke of “substantial” rather than “compelling” government interests. And serious intermediate scrutiny would be far better than Smith’s total abdication of review. The danger is that intermediate scrutiny often declines into excessive deference, as in O’Brien.

The key point, as Fulton again emphasized, is that “[r]ather than rely on ‘broadly formulated interests,’ courts must ‘scrutinize[] the asserted harm of granting specific exemptions to partic­ular religious claimants.’” Exemptions are as-applied holdings; they allow the law to go forward in most cases while preserving religious freedom in particular applications. The interest underlying drug laws may be generally compelling, but not, the court found, as applied to limited use of a drug in worship services. The interest underlying nondiscrimination laws may be generally compelling, but less likely so when there are multiple alternatives to the objecting religious provider or when the alleged discrimination is inside the church itself.

This analytical structure led Congress to find, when it enacted the Religious Freedom Restoration Act, that the compelling-interest test “strikes sensible balances” between religious liberty and government interests. Multiple studies confirm that RFRA has produced far from absolute protection for religion. Religious exercise includes conduct, and government more often has compelling reasons to regulate conduct than to regulate speech. If applied in light of these considerations, compelling interest is a workable standard.

(Correction, June 21, 1:36 p.m.): This article has been updated to reflect that it was religious progressives more broadly, and not just Catholics, who were the subject of a ruling by a federal district court in Arizona regarding aid to undocumented migrants earlier this year.