Thursday, June 04, 2020

Are Churches Like Restaurants? Like Political Rallies? If So, Why Isn't Peyote Like Wine?: The Comparator Problem In Religious Discrimination Cases

by Michael C. Dorf

In my essay on Tuesday, I criticized the dissenting Justices in South Bay United Pentecostal Church v. Newsom for seeming to misunderstand and/or mischaracterize California's rules governing in-person worship services, which, at the relevant time, were permitted if limited to 25% seating capacity and a cap of 100 people. As California argued and Chief Justice Roberts explained in his concurrence, comparable secular activities such as lectures, concerts, and live spectator sports were completely forbidden, so that far from discriminating against religion, California was giving religious worship services a special benefit (assuming that the opportunity to contract COVID-19 is a benefit, as all parties assumed). Justice Kavanaugh's dissent contrasted the treatment of worship services to other activities that were not subject to the 25% cap: "factories, offices, supermarkets, restaurants, retail stores, pharmacies, shopping malls, pet grooming shops, bookstores, florists, hair salons, and cannabis dispensaries." Yet, as the Chief Justice noted, these venues and activities are not really comparable in terms of health risk, because they do not occur in places "in which people neither congregate in large groups nor remain in close proximity for extended periods."

That's plainly right for most of the examples, but, a commenter on my earlier post observed, not necessarily all of them. Accordingly, another commenter concluded, California really was making a judgment not only about relative health risk but about the relative importance of religion and some other activities. And therefore, the commenters implied, my criticism of the Kavanaugh dissent for conflating a health risk judgment with an importance judgment missed the mark.

In today's follow-up essay, I'll say a few words about relative health risks--focusing on restaurants and political rallies--before turning to a question of constitutional doctrine that has implications and applications beyond the current context. Bottom Line: Chief Justice Roberts and I are right; Justice Kavanaugh and the commenters are wrong.

Much of the debate in the comments on my last essay focused on restaurants. As the case came before the SCOTUS, California had begun to permit in-person dining. The applicable rules include the admonition to "[a]djust maximum occupancy rules inside the establishment based on its size to limit the number of people inside" to allow for social distancing of at least six feet between customers at different tables. Kavanaugh and the commenter on my earlier essay are right that this provision--while aiming at the same goal--is not the hard cap applicable to worship services. Moreover, a now-famous study of a restaurant in China pretty strongly suggests that spacing in restaurants is likely to be inadequate to prevent COVID-19 spread, at least absent careful attention to ventilation issues. Still, I tend to agree with the commenter on my Tuesday essay who noted that even so, restaurants present a lower transmission risk than worship services, given the impact of singing at the latter.

That's not to say that I think California's rules governing restaurants are wise. In light of the mild climate, a more reasonable accommodation might be to continue to permit takeout and delivery, while limiting sit-down dining to outdoor seating areas. However, to say that California's rules governing restaurants are insufficiently protective of public health is not to say that they discriminate against religion. I'll return to that point after a digression about political rallies.

A commenter on my Tweet promoting Tuesday's essay observed what he thought was the irony that even as California is restricting worship services to 100, it is permitting much larger gatherings in the form of protests against racist policing. I too have been concerned that, much as I share the concerns of the peaceful protesters, mass protests will spread the virus. Many protesters wear masks, but some do not, and it is nearly impossible to maintain social distancing in a mass protest. Meanwhile, direct encounters with police and especially arrests create heightened risk of spread.

It is not clear to me, however, that California, its localities, or any of the other states or localities where mass protests have been occurring have any good public health options regarding the protests or that taking one approach to outdoor protests and a different approach to indoor sustained gatherings amounts to discrimination against religion. If California were allowing mass protests against racist policing but disallowing religious processions, that would be such discrimination. But that is pretty clearly not what California is doing. (See Postscript 1 below.)

That brings me to the broader theoretical point. On Tuesday I said that Justice Kavanaugh appeared to be conflating judgments about health risk with judgments about the importance of various activities. The comments on that essay lead me to think that he and perhaps some of the commenters were also conflating two other matters: the inquiry appropriate to determining whether discrimination against religion is justified under strict scrutiny; and the threshold question whether there is discrimination against religion in the first place.

Suppose that California or some other state limited worship services to 25% occupancy and a hard cap of 100, while permitting movie theaters to operate at 33% occupancy with a hard cap of 150. That would clearly be religious discrimination, because, on its face the rule distinguishes "worship" from other kinds of gatherings. The answer to the threshold question whether there is discrimination against religion is yes, and a reviewing court would move on to address whether that discrimination could satisfy strict scrutiny.

The answer to that question is not immediately apparent. Perhaps the state would say that while moviegoers sit in their seats mostly quietly, people at worship services routinely sing, which creates a greater risk of transmission. If so, the court would next inquire whether the state applies the same rule applicable to worship services to other gatherings where people sing, such as secular choir rehearsals. If it doesn't, then that under-inclusiveness would mean that the law fails strict scrutiny.

Note, however, that the careful examination of how the state treats religious and comparable non-religious activities only occurs after the determination that the state has engaged in religious discrimination. By contrast, Justice Kavanaugh and (perhaps one or more of) the commenters on my last essay seemed to be applying strict scrutiny to determine whether there was religious discrimination in the first place. But that's not how the doctrine works.

Employment Division v. Smith itself is illustrative. Oregon forbade everyone from using peyote. In dissent, Justice Blackmun noted that "respondents' use of peyote seems closely analogous to the sacramental use of wine by the Roman Catholic Church," and even Justice Scalia's majority opinion acknowledged that the "exercise of religion" may involve "sacramental use of . . . wine." Oregon forbade peyote but not wine. Under the logic of Justice Kavanaugh and the commenters on my last essay, that should have counted as at least prima facie discrimination against practitioners of faiths that use peyote instead of wine in their rituals, calling for an evaluation of the relative health (or other) dangers of peyote and wine. Yet it did not.

To be sure, in free exercise cases as in equal protection cases, sometimes a law that is neutral on its face might have been adopted for an impermissibly discriminatory purpose, but to establish illicit purpose requires more than showing that the lines the state drew are not perfectly logical and disproportionately burden religion; it requires a showing of anti-discriminatory animus.

And there was nothing like such a showing in South Bay United. On the contrary, the only evidence of  intent on this point shows that the state intended to exempt worship services from the stricter rules that would apply were they secular gatherings, as the face of the order revealed. Thus, under the constitutional doctrine as it stands, comparisons to restaurants and rallies are wholly inapposite.

But wait. Wasn't the Supreme Court in Smith somewhat obtuse to treat Oregon's law as neutral? Shouldn't the state's different treatment of peyote and wine count as discrimination based on religion (in favor of Catholics and Jews, who use wine sacramentally, and against practitioners of Native religion, who use peyote sacramentally)? Perhaps so, but if so, that would require a rather dramatic modification to Smith.

As I explained in Tuesday's essay, there is growing dissatisfaction on the political right with Smith, but none of the dissenters in South Bay United so much as hinted that they were voting as they did because they thought Smith was no longer controlling. Nor did they hint that they want to dramatically narrow what counts as a religion-neutral law under Smith. Instead, they purported to apply existing law.

They did so in error. For purposes of making a threshold determination whether there is religious discrimination at all, restaurants and outdoor political rallies are considerably less comparable to worship services than wine is to peyote. Justice Kavanaugh's examples were inapt.

Postscript 1: On the Volokh Conspiracy, Prof. David Bernstein calls into question policies in NYC and NJ allowing mass protests against racist policing but not allowing mass religious gatherings for funerals. He implies but does not specifically say that such policies are impermissibly content-based in violation of free speech and, insofar as they treat religious gatherings less favorably than secular ones, a form of religious discrimination. Although I object to Prof. Bernstein's casual repetition of conspiracy theories about the motives of "the left" propounded by his right-wing "social media friends," he has a point. Government's response to the public health dangers posed by mass gatherings should not be based on the views expressed. (Bernstein also criticizes individuals who condemned protests against shelter-in-place orders but have condoned the current protests. However, that criticism does not directly implicate any constitutional issues, because government officials did not restrict the anti-lockdown protests.) 

Postscript 2: After I drafted but before I published the foregoing essay, I received an enlightening email from Professor Alan Brownstein providing some additional rationales for the California policy as it came before the SCOTUS in South Bay United. I quote his analysis below (with his permission) because I think it makes very clear that California is not discriminating (in a pejorative sense) based on religion. Indeed, Prof Brownstein's insights would go a long way towards justifying the distinctions California has drawn, even if they were subject to heightened scrutiny. He writes:

To help explain the distinctions California is drawing with its regulations, it might be useful to give examples of gatherings that would be prohibited or restricted even if they took place in an open business.

So, for example, book stores may be open, but authors reading from their work to an audience and signing book copies in the book store would be a prohibited gathering. Similarly, restaurants could not host a wedding reception or bar mitzvah celebration that extended beyond guests dining at spread out tables. Gatherings are problematic wherever they occur.

I also tried to come up with my own list of reasons why gatherings for worship at a church, synagogue, or mosque create particular risks of contagion or enforcement problems for state authorities. (I admit it is heavily influenced by my experience with Jewish services -- which I currently attend via Zoom.)

Communal touching or handling of objects (e.g. prayer or hymn books, tallit or yarmulkes, Torah scrolls.)

Touching of persons.

Singing (in a choir or by congregants).

Communal sharing of food or wine.

Long term continuing stationary presence of individuals in proximity to each other, even with appropriate social distancing, in enclosed environment.

The size of the same group in a single room.

Two more general enforcement concerns might be:

The fact that worship services of certain faiths may be more problematic from a public health perspective than the services of other faiths because the worship services of those faiths create more of a risk of contagion. The state wants to avoid allowing some faiths to hold services while prohibiting the services of other faiths. Or interfering to a far greater extent in the services of some faiths than others.

The state's interest in avoiding the intrusive monitoring of religious services that might be required to guarantee that public health requirements are being adequately satisfied. I think the state is and should be much less concerned about having a public officer visit and monitor the operation of a restaurant than having state agents monitor a worship service and call into question religious practices that arguably risk contagion.


Samuel Rickless said...

I wonder whether Kavanaugh et al. would have been on stronger ground if they hadn't based their argument on Equal Protection but had instead based it on (Substantive) Due Process. The assumption would be that freedom of religious exercise is a fundamental right (whether the criterion of fundamentality is history and tradition or the centrality/importance of the relevant activity to people's lives). Once a fundamental right is infringed or abridged, as it is in the case of government prohibition of religious gatherings (above 25% occupancy), this immediately triggers strict scrutiny. There is therefore no need to enquire into whether there has been discrimination on grounds of religion. Perhaps, if one is maximally charitable, this is where Kavanaugh et al. were really coming from.

But this wouldn't help the dissenters anyway. In the application of strict scrutiny to this case, the question is not whether the end is compelling (the end is clearly to save lives, and this is a compelling end if anything is) but whether the means are narrowly tailored to achieving the end. The relevant factors here are:

1. Necessity
2. Availability of neutral alternatives
3. Flexibility and duration
4. Impact on rights of third parties
5. Level of stereotyping

The 25% occupancy rule seems necessary to significantly lower the risk of viral transmission (according to our best evidence), the neutral alternatives would either significantly increase the risk (raising the cap to 50% occupancy, eg) or burden the right further (by banning religious gatherings outright), it is clear that the measures being taken, though inflexible, are limited in duration, there is no impact on the rights of third parties, and there is no stereotyping associated with the rule. The means, then, do seem narrowly tailored to achieving the end of saving a significant number of lives.

Note that I am intentionally bringing the Court's analysis of narrow tailoring in Equal Protection cases to bear on what I have suggested the dissenters analyze as a Due Process matter. Maybe this is inappropriate. But, in that case, we need more guidance from the Court on how to treat discrimination with respect to the burdening of a fundamental right.

Suppose, to take a fanciful example, that the freedom to protest peacefully is a fundamental right, and that members of certain religious denominations protest by forming human towers (in the way that gymnasts and cheerleaders do). Suppose, further, that government imposes a 12 foot limit on the height of human towers at protests, on grounds of public safety (people could fall off the towers and injure others). The question, for me, is whether the Court should analyze this simply by looking at the question of whether the height restriction is necessary to achieving the end of protecting the public, or whether the Court should analyze this by looking at whether the restriction is narrowly tailored to achieving that end. I would think that the Court should do the latter.

Michael C. Dorf said...

Sam: Kavanaugh et al don't use SDP because (a) they think they hate it (because it gives rise to Roe); and (b) what you're describing as SDP in this context would mean overruling Smith. As I said in my original post, they probably want to overrule Smith, but they don't think they can do so w/o a case fully presenting the question.

Joe said...

A basic thing that stood out for me was Roberts' statement on the high test required for the in effect emergency relief requested & that it was not met.

I think the restaurant rules/guidelines very well might be too weak but they still are different from churches & how weak they are is far from clear to me. Normal litigation and continual changing rules as the situation requires (as Roberts referenced) to me at this point was the response justified. Roberts to me is still correct. The details in the essay here help show that; after all, again, there is a high bar.

I'd add that I don't think the "what about the protests" (see also, Balkanization Blog post) adds as much as some might think. A host of the Gay USA program (Chris Cooper of the birder/white woman with a dog racist incident sometimes guest hosts), e.g., complained that police limited a NYC protest she was involved in given social distancing rules. It was not a funeral gathering and small.

Yes, the mass protests in response to George Floyd's murder are being allowed. But, I'm unsure if "allowed" is exactly apt since there is -- as compared to any number of gatherings -- just so much you can do with such a public expression of anger and protest. The city did on Monday start a nightly curfew to protect public order, one of limited success. Various attempts to restrain the mass action here led to problems.

It very well is concerning so many people are together like this and some have expressed fears that it will lead to new outbreaks. But, it is unclear to me how useful it is to compare it to day by day policies.

Salemicus said...

I think you are still missing the force of Kavanaugh's argument. This case would be comparable to Smith if California had placed restrictions on choral singing (say). That would be a facially neutral rule (but arguably impermissible due to animus or if Smith was wrongly decided, etc). Instead, California has singled out religious services as a type. A silent church service is treated differently to a silent office meeting of equal duration. That's prima facie discrimination.

Would you make the "typical members of class A have quality X, so it's OK to attribute that quality to all of them" argument in other discrimination contexts? Suppose California allows ethnicity A to start restaurants without a permit, but requires a permit from ethnicity B, and argues that ethnicity B's restaurants are typically so noisy and polluting that they are comparable to factories of ethnicity A (which do require a permit). Even if the differential pollution claim is true, is that a winning argument?

If Oregon had banned the use of peyote solely in religious ceremonies, but allowed it in all other contexts, Smith would have come out the other way.

Joe said...

CA did not "single out" religious services.

Kavanaugh asserted that the differences weren't relevant. That is the true debate here and the discussion argues the Roberts side, basically.

Secular religious ceremony-like meetings were treated in a comparable way.


Reference was made to alcohol/peyote and this came up in the oral argument in the second case particularly as well. There is some real argument that especially for Native Americans, alcohol has been more dangerous than a limited use drug. OTOH, there was some argument made that it was a hallucinogenic drug with certain potent effects like other restricted drugs & in fact some (all failed) have made religious arguments for other drugs like marijuana. The dissent in Smith also dealt with that while others might have a more open-ended libertarian argument.

Asher Steinberg said...

I didn't expect to find this so unpersuasive. A couple things:

A peyote/wine distinction (which wasn't at issue or discussed in Smith -- Blackmun only says in a footnote that an exception for sacramental peyote could be granted because one was granted for sacramental wine during Prohibition, not that Oregon was discriminating by treating peyote differently from wine today) was facially neutral as between religions, because peyote and wine are not only or even mostly used for sacramental purposes. This set of policies, on its face, discriminates against religious services, not some larger class of things that happens to include religious services; it has religious-service-specific rules. The question then is, notwithstanding that it is a facially religion-based rule that imposes different occupancy requirements if you're a church than if you're a restaurant than if you're a grocery store, might the set of policies nevertheless be deemed neutral because of some set of relevant characteristics connecting the things treated like churches and the things treated like restaurants. Really, given what you say about churches and movie theaters, i.e., that an occupancy distinction between them would trigger strict scrutiny, I don't know why you don't concede strict scrutiny applies here. You can't really think it's safer for hundreds of people to sit 6 feet apart in a restaurant for a couple hours while loudly conversing and eating their food than it is for hundreds of people to sit six feet apart in a movie theater for a couple hours, whispering occasionally and nibbling on their popcorn.

Singing, on its own, is a risible basis on which to say restaurants and churches are different enough to call differential treatment of the two neutral. People quite often speak more loudly (and more continuously -- there is no part of the dinner where someone sermonizes for half an hour from a great distance while everyone else is quiet) in many restaurants, and therefore expectorate more, than they sing in church. And many of the things Brownstein says about religious services are also true of restaurants.

In the counterfactual world you've constructed where (a) Smith actually engaged with a wine/peyote distinction, and (b) that distinction isn't facially neutral in a way that a distinction between churches and restaurants undeniably is not, which is an extremely counterfactual world indeed, restaurants and churches would still be more comparable than peyote is to wine. Peyote's a hallucinogen; wine is not.

David Bernstein said...

Mike, I'm puzzled why you assert that I was spreading "right-wing" conspiracy theories about "the left." I didn't mention right nor left, just social media friends complaining about elites imposing their values on the public, and this was about motive, not conspiracy. At least in my case, those who were doing the complaining were a mixture of progressives, conservatives, and libertarians, not "the right," and my most vociferous pro-lockdown social media friends, otoh, were libertarians, not liberals.

Joe said...
This comment has been removed by the author.
tjchiang said...

Two points:

1. You are shifting the argument quite a bit. Your prior post was accusing Kavanaugh of being "almost incomprehensibly incompetent" for thinking that California was making judgments about what is and is not essential. I think your argument is now a doctrinal formality that, yes, honestly speaking California probably is making judgments about what is and is not essential, but we don't scrutinize that deep on the first step question about whether a law is discriminatory against religion.

2. I'll concede the doctrinal point that a faithful read of Smith probably does make it difficult to show that a facially neutral law is motivated by animus, and that "animus" is the standard. But I think the SCOTUS conservatives would find impermissible motive not just with affirmative animus but also a condescending disdain towards religion, and this is not nearly such a revolution of doctrine as you suggest. Also, what one thinks about a lawmaker's motives depend strongly on one's priors. You and I might read California's categories as just inevitable and benign imperfections in drawing category boundaries; people who don't trust the California government are much more likely to see a conspiracy on the same facts.