The Volokh Conspiracy

Mostly law professors | Sometimes contrarian | Often libertarian | Always independent

Religion and the Law

A Framework for Analyzing a Church Autonomy Defense

Some doctrinal tools to appropriately limit church autonomy.


This continues a blog series about my new paper, "The Limits of Church Autonomy." You can find Post 1 here, Post 2 here, and Post 3 here.

Both church autonomy and accountability are important principles. The problem in the current law is that the courts are unclear on how to avoid allowing one to swallow up the other. This section proposes some analytical revisions that would help to clean up what has become a doctrinal mess. In terms of doctrine, the shifts suggested here are modest and can be done without any change by the Supreme Court.

Religious conduct

The starting point is that church autonomy protects religious conduct. There has been confusion about whether conduct has to be based on religious belief, or whether reasons are needed at all for church conduct to receive church autonomy protection. Particularly in the employment cases, courts (including the Supreme Court) have frequently said that there need be no showing that particular conduct is religiously motivated. But in other cases, courts have said that only when the challenged action is religious is it protected.

The best way to reconcile the cases is to note that church autonomy protects religious conduct. But this is not the same as saying that everything happens within a church is religious, nor that there must be an articulated religious reason for every protected act. Instead:

The easiest case is when the conduct at issue is religious is done directly pursuant to religious doctrine or beliefs.

Beyond this, a set of classic, recognizable internal governance actions by religious institutions should be considered presumptively religious, without requiring the institution to provide a religious reason for each action. (This is the best way to rationalize the holding in ministerial exception cases that the church need not provide reasons for its employment decision to receive protection.) This would cover the selection, supervision, and retention of ministers; matters of membership; and matters of church discipline.

Rather than an open-ended protection of church government, the forms of church governance here need to be carefully limited. The presumption, historically, was that two forms of church discipline were immunized from judicial scrutiny: verbal corrections and excommunication. By verbal, I mean correction by words, whether written or spoken. By excommunication, I mean the removal of the person from membership in the religious institution and the exclusion of that person from ritual or sacramental observances.

In sum: conduct taken as part of church governance can be religious, even if it does not always have specific religious reasons for each individual act.


After considering whether conduct is "religious"—including presumptively religious matters of internal church governance—we turn to consent. Where consent comes in is to solve the problem of when to treat particular conduct as religious that does not easily fit the categories that have become classic instances of church governance in the case law.

Suppose, for instance, that Church X has an unusual form of discipline: a ceremonial slap on the wrist. This doesn't fall into the narrow and cautious category of verbal rebuke that the cases (construed conservatively) have recognized. Arguably this is because physical slap is not a standard part of church practice. But if it were, there would have been as many battery suits as there have been defamation suits arising out of church discipline.

The solution on this point is consent. A religious institution can have the benefits of immunity when there is a clear consent to the later-challenged conduct. This is likely to be most important with unusual forms of discipline. A member could consent to a physical slap as a form of discipline, for instance. If consent was not coerced and was informed, this will count as a consent defense under tort law.

Here's an example of how this could come into play: a mandatory reporting statute requiring church leadership to report child abuse. Imagine that a religious group objects to this as an infringement on its ability to self-govern. But there is no existing established tradition that reporting crime infringes on the prerogatives of the religious institution's self-governance. To the contrary, the history suggests that this is what one would have expected religious entities to do, building on the English and common law history taken into the American context. Meanwhile, the people being protected by the mandatory reporting laws—children, the abused—either can't or won't consent to not report. This just isn't going to be covered by church autonomy.

Consent also helps to solve the problem of privileging a particular set of religious practices. One possible objection to the approach outlined so far is that it privileges traditional forms of religion. Christian churches have had by far the most litigation in America. So of course, what is "recognizable" as church governance in the caselaw is going to be disproportionately shaped by this Christian tradition. If we basically freeze identifiable church governance activities with the easily-recognizable functions in the case law, that is effectively freezing a set of practices that are most identified with church practices: selection of ministers, discipline in the form of rebuke and disfellowshipping. On what basis could an unusual practice make its way into the category of church governance? Again, this is where consent could come into the picture.

The approach outlined so far will cover most cases that actually arise. It will also take the odds and ends of doctrinal tools already in use by the courts and organize them into a logically-structured analysis, rather than leaving them to be drawn upon in a grab-bag manner, with unpredictable and inconsistent results.

A backstop

Yet all that said, one might still worry about a more drastic reductio of church autonomy: a religious institution that actually has criminal conduct as part of its religious precepts. Why doesn't religious autonomy protect human sacrifice or ritual sex with underage victims? For this, we might formulate a backstop principle, such as (at minimum) that church autonomy does not protect the direct cause of physical harm. This catches the egregious but exceptional case that slips through the other filters: if there is a religiously motivated act that causes physical harm, the perpetrator should still be subject to accountability before the civil authorities. One could base this on the history I sketched in an earlier post, and on early American history: by the time of the Constitution's framing, it was well understand that whatever religious freedom meant, it didn't protect "licentiousness" or acts violating the "public peace."

The backstop is not perfect (and in the paper I reflect on some of the shortcomings). But it's a starting point, and one that is entirely consistent with existing caselaw.

Lower courts can draw on this structure of analysis (and lots of helpful citations are in the paper) to (hopefully) decide these cases in ways that are clearer and avoid some of the analytical missteps that I described in an earlier post.