The present doctrinal development begins with Church of Lukumi Babalu Aye v. Hialeah (1993), in which the Court invalidated a ban on animal sacrifice. The Court had previously held that there was no right to religious exemptions from neutral laws. But this law wasn’t neutral. It targeted an unpopular religion of Caribbean immigrants. The laws, the Court concluded, were “drafted with care to forbid few killings but those occasioned by animal sacrifice.” The state said that it had a legitimate interest in preventing cruelty to animals. The Court retorted that the city “devalues religious reasons for killing by judging them to be of lesser import than nonreligious reasons.”
In Hialeah, animals could be killed, sometimes painfully, for all sorts of nonreligious reasons. Live rabbits were used to train grayhounds. The city’s laws, the Court said, “fail to prohibit nonreligious conduct that endangers these interests in a similar or greater degree than Santeria sacrifice does.” That showed an impermissible purpose. “The record in this case compels the conclusion that suppression of the central element of the Santeria worship service was the object of [Hialeah’s] ordinances.”
The claim in Lukumi built on language in Smith distinguishing earlier cases in which it had overturned state refusals of unemployment benefit to claimants who refused to work for religious reasons. Where there is “a system of ‘individualized governmental assessment of the reasons for the relevant conduct,'” the Smith court had declared, government “‘may not refuse to extend that system to cases of ‘religious hardship’ without compelling reason.” When Justice Scalia wrote that, he probably did not intend to lay down a sweeping new rule. Michael McConnell is right that this language had “one function only: to enable the Court to reach the conclusion it desired in Smith without openly overruling any prior decisions.”
Lukumi is the source of the underinclusiveness rule that the Court adopted in Tandon. But the Tandon rule does not readily follow from Lukumi, which did not specify the standard it was applying. As James Oleske has shown in considerable detail, “the Court viewed Lukumi as an extreme case and deliberately left unclear the appropriate methodology for deciding closer cases.” Lukumi does not adopt any form of MFN. It is a precursor of MFN, not an example of it.
The Lukumi Court was divided on evidentiary questions, but it agreed that hostility toward some unpopular religion was the trigger for strict scrutiny. Justice Kennedy, writing for the majority, concluded that the record of its enactment “discloses animosity to Santeria adherents and their religious practices.”
As the Court later explained, the question under Lukumi was whether a law “had the object of stifling or punishing free exercise.” A single secular exemption could hardly suffice to prove that religion was singled out. The question of what to do if presented with such an exemption was not before the Court.
Impermissible purpose was indispensable to Lukumi‘s analysis. Its most fundamental difference from Tandon is that no such purpose was found in the California Covid regulation. That difference is clear in the remedies that the court provided. In Tandon, the Court required a religious exemption to the secular ones. In LukumiOleske observes, “the remedy in the case went far beyond the granting of a religious exemption to ensure parity. Instead, the. Lukumi Court completely voided Hialeah’s ordinances after determining that they were ‘designed to persecute or oppress a religion or its practices.'” That is appropriate when the state has acted with an improper purpose, such as racism or the endorsement of a religion.
So how did the Lukumi rule change?
The most basic form of MFN, using the existence of a single exception to trigger strict scrutiny, was first adopted by Justice Alito when he was a Third Circuit judge, in Fraternal Order of Police v. Newark (1999). Call this MFN-1.
In Newark, the city did not permit police officers to grow beards. However, it is exempted those with pseudo folliculitis barbae (PFB), a medical condition that makes shaving impossible. Two Sunni Muslim officers, whose religion requires men to grow beards, claimed that they were being unconstitutionally discriminated against. Alito agreed: a law discriminates when it “creates a categorical exemption for individuals with a secular objection but not for individuals with a religious objection.” The police department impermissibly “has made a value judgment that secular (ie, medical) motivations for wearing a beard are important enough to overcome its general interest in uniformity but that religious motivations are not.”
There are crucial differences from Lukumi. The Hialeah ordinances were “drafted with care to forbid few killings but those occasioned by religious sacrifice.” The claims in Newark did not allege that the no-beards rule itself was adopted for the purpose of harming Muslims. The “hostility” was manifested only in the failure to make an exception for them. The remedy was to create that exception, rather than to invalidate the underlying rule.
The test is still discriminatory purpose, but in Newark the failure to accommodate religion, while any Trade secular interest is accommodated, is taken as conclusive evidence of that forbidden purpose. No further evidence is necessary, even though there is no evidence of intention to harm, nor any plausible way to construe the statutory scheme as having the object of harming minorities.
The best justification for the result in Newark would be that, comparator or not, the state interest was not especially pressing and would not have been appreciably jeopardized by religious accommodation. The medical exemption would then be understood as evidence supporting a finding of selective sympathy and indifference—which would be discriminatory. That of course would mean overruling Smith and directly balancing the interests. More generally, MFN-1 can be useful, but that usefulness is evidentiary, helping to determine whether religion has been devalued, and its usefulness is intertwined with the larger question of Smith.
Later forms of MFN are not so easily defended.