RFRAs in Private Litigation: A Reply to Amar and Brownstein

By Michael Dorf

My latest Verdict column addresses an issue raised by Justice Scalia in last week's SSM oral argument. He asked whether recognition of a constitutional right to SSM would mean that states could no longer grant clergy who refuse to perform same-sex wedding ceremonies the power to peform any wedding ceremonies that have civil as well as religious consequences. The question has generally been treated as trivially easy: No one thinks that this will result. But as I explore in the column, the question is harder than commentators assume. If states themselves cannot deny people SSM, why can clergy deputized by the state do so? I ultimately conclude that the best explanation for why states may continue to give civil recognition to marriages performed by non-egalitarian clergy is that those clergy should not be understood as state actors when they perform religious ceremonies--even though the state gives civil effect to the ceremonies.

My Verdict column is partly styled as a response to a Verdict column last week by Professor Marci Hamilton, who--expressing the conventional wisdom--called Justice Scalia's question a "no-brainer." Although I ultimately agree with Professor Hamilton's bottom line, I try to show why Justice Scalia has raised a serious issue about the relationship between general obligations and religious liberty.

In this post, I want to address another question about that relationship, this one prompted by yet another Verdict column: A piece by Professors Vik Amar and Alan Brownstein that, in turn, responded to my earlier column about RFRAs in private litigation. If all of this sounds like a whole lot of inside baseball, I apologize. I do think there are important issues here--and while the differences among Professors Amar, Brownstein, Hamilton, and Dorf are relatively small and subtle, they may have important consequences down the line.

Let me briefly set the stage. In my April 8 column I began by acknowledging that there's nothing anomalous about a constitutional or statutory right applying in private litigation. I nonetheless offered two reasons for thinking that, absent express language addressing the private litigation question, a state or federal RFRA ought to be construed so as not to apply in private litigation. I'll recap those reasons in reverse order so as to focus attention on the main disagreement between Amar/Brownstein and me.

My second reason built on the proposition that courts and legislatures are properly reluctant to grant religious exceptions from general obligations where granting that exception imposes substantial burdens on third parties. I acknowledged that a third party could sometimes be subject to a substantial burden even in litigation between the religious claimant and the government but that a third party is just about always burdened by a religious exception if the issue arises in litigation between that "third" party and the religious claimant. Accordingly, I said that it may be simpler just to adopt a categorical rule that disallows RFRA claims in private litigation.

Professors Amar and Brownstein disagree with my proposed categorical rule. They would prefer a case-by-case balancing approach in which the burden on third parties is evaluated in both private litigation and litigation against the government. The disagreement here is of a familiar type. It invokes the tradeoff between rules and standards. My proposed rule--like just about all rules--is somewhat over- and under-inclusive relative to its background justification. The Amar/Brownstein standard is more finely calibrated but at the expense of uncertain application and higher administrative costs (because more people with ultimately losing claims would litigate). I don't think that it's obvious which approach is superior. I'm sure that all of us--Amar, Brownstein, and I--believe that in some circumstances a categorical rule is superior to a standard and in other circumstances vice-versa. My sense of the cases likely to arise leads me to think that the rule is preferable here, but absent substantial empirical evidence about the claims that actually arise, it's hard to know who has the better argument.

Accordingly, I'll leave our disagreement about my second rationale there and turn to my first rationale. I said in my original column that the application of a RFRA in private litigation is different from nearly every other constitutional right. (I say "constitutional" notwithstanding the fact that RFRAs are usually statutes, because a RFRA is meant to function similarly to a constitutional right.) How is RFRA different? The key is a point that American constitutional law is pervasively rule-dependent, as explained by Professor Matt Adler in a series of articles (in the Michigan Law Review, the Harvard Law Review, and Legal Theory). That is, constitutional rights are rights against rules, rather than rights to engage in conduct full stop. For example, Texas v. Johnson protects a right not to be prosecuted for burning a flag under a statute that forbids flag burning due to the message expressed by flag burning; it does not protect a right to burn a flag under a statute forbidding lighting a fire in public (so long as that statute is applied even-handedly so that the statutory rule is the actual rule).

Adler's normative view is controversial and one can quibble at the margins with his descriptive view. For instance, free speech doctrine at least purports to provide some protection for expression even as against content-neutral laws. But this sort of quibble at most requires a minor modification to Adler's description, wherein one says that constitutional law almost never provides protection for acts simpliciter. And even thus restated, Adler's analysis shows why RFRAs function differently from just about all other rights. RFRAs shield conduct; they don't attack rules.

Take an example I gave in my column. Suppose that a state law permits ex-wives but not ex-husbands to collect alimony. I acknowledged that this law should be subject to challenge in private litigation. The ex-husband seeking alimony from his ex-wife says that the law is constitutionally defective in drawing a sex line. Amar and Brownstein say the same thing about RFRAs. They say that whenever a private party raises a RFRA claim in private litigation, he or she also says the rule at issue is defective in failing to make an exception. But while this is a conceptually possible way of describing the situation, it's hardly the most natural way to describe it. It seems more natural to say that the rule is fine but that the RFRA claimant has a separate entitlement to be shielded from liability.

Consider a case of simple trespass. Suppose that Owen has posted on his property a sign saying "no trespassing." Whenever someone comes on Owen's property without his permission, Owen asks them to leave, and if they don't, Owen sues (for damages after they've left or for an injunction if they remain). Now suppose that Reggie comes on Owen's property. Owen asks him to leave. Reggie says he can't leave because a tree on Owen's property is holy in Reggie's religion, and Reggie needs to remain on Owen's property to worship at the tree for a week. Owen sues Reggie for trespass, and Reggie defends by invoking a state RFRA that is silent on whether it applies in private litigation.

If Amar and Brownstein are right, then we would say that the common-law tort of trespass is a RFRA-violative rule because it does not include an exception for religious trespasses. But it seems more natural to say that the common-law tort of trespass is fine; it's just that RFRA shields religious trespassers like Reggie, if the RFRA applies in private litigation.

Amar and Brownstein's description obscures how far-reaching are the consequences of applying RFRAs in private litigation, because it equates that application to examples like the alimony case, where the defect really is in the rule. But in the trespass case, we see that applying RFRA in private litigation essentially means waiving the state action requirement for RFRAs. It converts RFRA into a more powerful tool than the Constitution in a case like Shelley v. KraemerAllowing Reggie an exception from the common law of trespass means saying that the RFRA protects people's right to free exercise against their neighbors, not just against the state.

To be clear, a state could enact a RFRA as far-reaching as that. But the argument of my column was that absent strong evidence that it has, courts should presume that RFRAs do not so radically change our understanding of what a right to free exercise of religion means. Amar and Brownstein have not persuaded me otherwise.