I want to talk to you about the Free Exercise Clause of the Constitution.  This post is long.

My view of the Free Exercise Clause is one part of a larger approach to constitutional adjudication involving the religion clauses.  For those who have been thirsting feverishly to know more about that approach, fear not: soon enough, I will flood the zone.  Suffice it for now to say that one of the most serious criticisms of my approach is that it is insufficiently predictable.  It is sometimes said, not without reason, that my approach is not rule-like enough, and that it is therefore damaging to rule of law values.

These are fair criticisms, and I do my best to address them.  I do this in part by taking a close look at the way in which a selection of district and intermediate appellate courts have applied that putatively most rule-like of all religion clause rules: neutral laws of general application do not violate the Free Exercise Clause.

What I find is: that rule is not nearly as inviolable as many who invoke it believe.  In fact, knowing when that rule will apply actually depends on having the sense of a host of context-dependent and issue-specific factors.  The trouble, as I have explained before, is the issue of general applicability.  Employment Division v. Smith carved out the unemployment compensation cases from its holding.  But, per this amicus brief, it is more accurate to think about this carve-out not as an “exception” but as a corollary to the rule itself, which creates a kind of graduated spectrum of general applicability. Laws which are not “generally applicable” are lifted out of the Smith “rule” and receive judicial balancing.  How do we know when a law is not “generally applicable”?

It falls to courts to determine what “generally applicable” means along the spectrum.  It cannot mean that the law has no exceptions, period; that would destroy the rule.  And yet “generally applicable” must mean something.  What it means is the subject of judicial interpretation–for now, very much in the common law style.  And that means that the Smith rule is much less predictable than its supporters suppose: “If the vice of pluralistic approaches is that they are predictable only to those who know how they will be applied, that is no less true of monistic approaches.”  Chapter 8, The Tragedy of Religious Freedom.  That is not enough, by itself, to convince you to adopt my approach.  For that, you need to buy the book!

Here is a brand new HHS Mandate case to show the predictable unpredictability of Smith, Geneva College et al. v. Sebelius, decided Wednesday by the U.S. District Court for the Western District of Pennsylvania.  The case is somewhat unusual inasmuch as the plaintiffs are both nonprofits and for-profits.  The nonprofits’ case was dismissed on standing grounds (only the Eastern District of New York, to my knowledge, has not followed this route for nonprofits).  As to the for-profits, after discussing the issue of a corporation’s exercise of religion and the RFRA claim, the court rested its decision to deny the motion to dismiss with respect to plaintiffs’ free exercise decision on an analysis of the issue of general applicability.  Here’s a substantial chunk of the decision, beginning around page 46:

There is little doubt that the mandate’s requirements are facially neutral in the sense that they are directed toward benefiting the public health, and are not explicitly targeted at any particular religious conduct. The court’s analysis, however, must extend beyond the face of the regulations in question. The Court of Appeals for the Third Circuit has acknowledged that

the Free Exercise Clause’s mandate of neutrality toward religion prohibits government from ‘deciding that secular motivations are more important than religious motivations.’ . . . Accordingly, in situations where government officials exercise discretion in applying a facially neutral law, so that whether they enforce the law depends on their evaluation of the reasons underlying a violator’s conduct, they contravene the neutrality requirement if they exempt some secularly motivated conduct but not comparable religiously motivated conduct.

Tenafly Eruv Ass’n, Inc. v. Borough of Tenafly, 309 F.3d 144, 165-66 (3d Cir. 2002). The process of implementing the objected to requirements has been replete with examples of the government impermissibly exercising its discretion by exempting vast numbers of entities while refusing to extend the religious employer exemption to include entities like SHLC.

The primary example of the “categorical exemption” rejected in Fraternal Order of Police in the present case is the grandfathering provision in the ACA, which exempts as many as 191 million entities from the mandate’s requirements. The grandfathering exemption impacts secular employers to “at least the same degree”—and likely far more—than religious objections from entities like SHLC. Blackhawk, 381 F.3d at 209. The fact that the government saw fit to exempt so many entities and individuals from the mandate’s requirements renders their claim of general applicability dubious, at best. Elsewhere in their briefing, defendants respond that the number of grandfathered plans will continue to decrease as time goes on. Even if this comes to fruition (which is not a certainty), the secular exemption for employers with fewer than fifty full-time employees that choose not to provide any insurance coverage remains. 26 U.S.C. § 4980H(c)(2)(A). Taken together, these categorical exemptions for secular entities and individuals raise a concern that the mandate’s requirements are not generally applicable.

In addition to the secular exemptions, the government continues to engage in an impermissible “religious gerrymander” by extending exemptions to an increasing number of religiously-affiliated entities. Although the court of appeals in Blackhawk and Fraternal Order of Police was not faced with the situation where, as here, some religious conduct is exempted, the fact that defendants continue to carve out exemptions, see generally 78 FED. REG. 8,456, while subjecting SHLC and other similarly-situated close corporate entities to the mandate’s requirements, raises a suggestion of “discriminatory intent” against close corporate entities seeking to advance the religious beliefs of their owners. Fraternal Order of Police 170 F.3d at 362. On the present record, this court finds that the Hepler plaintiffs raised plausible claims that the sheer number of exemptions—both secular and religious—to the mandate’s requirements burdened their free exercise rights to an extent sufficient to trigger strict scrutiny. The court already analyzed the mandate’s requirements under the compelling government interest test in the RFRA context and found that they do not survive strict scrutiny; therefore, for the same reasons, the First Amendment claim is sufficient, and the motion to dismiss this claim must be denied.

Let’s set to the side, for the moment, the issue of the proper interpretation of “general applicability.”  This court interpreted in a certain way; other courts, as I show, interpret it differently.

The problem with the “general applicability” issue isn’t that one court may decide a case in a way you might like, and another court may decide a different case in a way you might not.  The real burn of it is that the very unpredictability that Smith aimed to eliminate has seeped right back in.  No matter how rule-like Smith tried to be, it could not squeeze out of constitutional adjudication what is and must be true about it (at least as to issues like these).  And if the response is that we can solve all of this by clarifying Smith and making it even more rule-like, my reply is: the more you squeeze, the more slips through.