3B. Would Granting an Exemption from the Employer Mandate Violate the Establishment Clause?

(For an introduction to this series of posts, see here.)

In the preceding post, I’ve argued that Hobby Lobby might well prevail on the claim that the employer mandate substantially burdens its owners’ religious practice. This means that Hobby Lobby would be presumptively entitled to an exemption under RFRA, though that presumption could be rebutted if denying the exemption is the least restrictive means of serving a compelling government interest. (I’ll turn to that tomorrow.)

But for now, let me discuss the argument that granting such an exemption would violate the Establishment Clause, “[b]ecause exempting large, for-profit corporations from the contraception mandate would significantly burden female employees, along with all the wives and daughters covered by the policies of male employees.” (I quote here a Slate piece by Micah Schwartzman & Nelson Tebbe, though the argument was also made in much more detail in this law review article by Fred Gedicks & Rebecca Van Tassell.)

If the argument is right, then not only is Hobby Lobby (alongside similar employers) not entitled to a RFRA exemption — it couldn’t get an exemption even if Congress or the Administration were eager to create such an exemption. Thus, for purposes of this post I will assume that Hobby Lobby and similar employers are statutorily given this exemption, and I will ask whether that would be constitutional. I won’t ask whether RFRA calls for the exemption; that is a story for tomorrow.

1. To begin with, let’s be precise about the effect of an employer mandate exemption would have on employee: If the employees want certain implantation-preventing contraceptives, they would have to buy them with their own funds, rather than getting them for free through an employer-provided health insurance plan. They would thus be in essentially the same legal position as all employees are today, before the employer mandate kicks in, and as some employees will be in the future, if their employers are exempted from the mandate (for instance, because they have grandfathered plans). The employer isn’t forbidding its employees from using certain contraceptives. It’s just not paying for them.

The burden on employees would thus be a burden relative to what the ACA would provide in the absence of an exemption. The employees of a business that would get a religious exemption wouldn’t get a benefit (free implantation-preventing contraceptives) — apparently worth several hundred dollars per year — to which employees of most other businesses will be legally entitled. Keep this in mind, because it will be relevant when we compare it to what happens with other religious exemptions that are generally believed to be constitutional.

2. Religious exemptions often leave nonbeneficiaries worse off than they would have been but for the exemption. A few examples:

  1. Giving conscientious objectors an exemption from the draft means that others will be drafted in their place. This imposes a huge burden on those others, under the definition of “burden” used above: They have to give up years of their lives, risk death or serious injury, and quite likely be required to kill or injure the enemy. To be sure, the particular person being burdened may be hard to identify. But for every objector who is relieved from fighting (and apparently 170,000 c.o. deferments were given during the Vietnam War, for instance), more or less one other person will have to fill that spot.
  2. Giving the clergy exemptions from the duty to testify means that litigants who might want the clergy’s testimony — for instance, about what a penitent confessed to the clergy — will lose valuable evidence for the case, and may lose their case. This too imposes a huge burden on those litigants, under the definition of “burden” used above; they may lose millions of dollars to which they would be entitled if only the normal duty to testify were imposed on the clergy. (To be sure, if the privilege were struck down, some people wouldn’t talk to the clergy in the first place, but others still might, especially if they feel they need to unburden themselves, and even more especially if they feel that such confession is a religious necessity. So the privilege does on net cause some litigants to be worse off than if the privilege doesn’t exist.)
  3. Letting churches discriminate based on religion in hiring and firing — even as to clearly nonclergy employees, such as janitors or bus drivers — leaves employees and prospective employees of other religions (or of the same religion who lapse) worse off than if Title VII’s religious discrimination ban applied to the church, without the exemption. This too imposes a substantial burden on those employees, under the definition of “burden” used above; they may lose their jobs, or may feel substantial economic pressure to stay within the church, and hide any change of beliefs they may have had. Yet the Court upheld this exemption against such a challenge, in Corporation of Presiding Bishop v. Amos (1987).

Say, then, there are two people. Anita works for an employer who (by hypothesis) has been exempted from the usually applicable (with some secular exemptions) employer mandate as a result of a statutory religious objector exemption. As a result, she doesn’t get, say, $500/year worth of contraceptive benefits that she would have been legally entitled to but for the employer mandate.

Barbara is suing Don Defendant for $500,000. She has reason to think that Don has confessed to Carl Clergyman that Don is indeed liable, so she wants Carl to be ordered to testify about the communication. But Carl has been exempted from the usually applicable (with some secular exemptions) duty to testify as a result of a statutory clergy-congregant privilege. As a result, she doesn’t win the $500,000 that she would have been legally entitled to but for the clergy-congregant privilege.

I don’t know of any court that has taken the view that applying the clergy-congregant exemption from the duty to testify in Barbara’s case would violate the Establishment Clause violation, despite the likely burden this would impose on Barbara; nor do I think it likely that a court would so decide. But given that this is so, then how would the application of the hypothetical exemption from the employer mandate constitute an Establishment Clause violation in Anita’s case?

Now perhaps the burdens imposed by these exemptions are unfair. And quite likely these particular exemptions aren’t required by RFRA. (Recall that this isn’t the question we’re discussing in this post.) But it seems pretty well-settled that the exemptions are constitutionally permissible, notwithstanding these burdens.

Some distinctions could be drawn between these cases and the employer mandate, but I don’t generally find them particularly persuasive. Some have argued, for instance, that the draft exemption is inapposite because the particular person who had to serve in the objectors’ place couldn’t be easily identified. And I agree that the law does sometimes treat burdens on easily identifiable third parties differently from burdens on harder-to-identify third parties. But I doubt such a distinction is really mandated by the Establishment Clause. It can’t be that the government is free to impose such a massive burden — conscription for years, the possibility of having to kill, and the possibility of dying — on hard-to-identify but indubitably real third parties (tens of thousands of them), but that the Establishment Clause at the same time blocks the government from imposing a several-hundred-dollar-per-year burden on more readily identifiable employees.

Gedicks & Van Tassell argued that the draft exemption is different because it wouldn’t “be a factor in the decision of nonpacifists to comply with or evade the draft,” while the employer mandate exemption might be a factor in a person’s decision whether to work for Hobby Lobby and similar employers. But this too strikes me as constitutionally irrelevant, when the question is substantiality of burden. Perhaps nonpacifists had to comply with the draft in any event, because they’d go to prison otherwise — but I don’t see how that makes an Establishment Clause difference, if the essence of the harm is the burden of the exemption on nonbeneficiaries.

Amos involved nonprofits, and the concurrences stressed that point. But if denying employees a certain benefit is an Establishment-Clause-violating burden, then I don’t think that it becomes any less of a burden when the employees work for nonprofits (such as the Mormon Church gymnasium involved in Amos). And I’ve heard some defend the clergy-penitent privilege on the grounds that it exists alongside other secular privileges, such as the doctor-patient privilege, lawyer-client privilege, and so on. But the hypothetical religious exemption for the employer also exists alongside other secular exemptions, such as for grandfathered plans and for employers with fewer than 50 employees (who do have to provide the contraceptive coverage if they provide an insurance plan, but who aren’t obligated to provide such a plan). To be sure, the secular exemptions are different from the religious exemption. But the clergy-penitent privilege is likewise quite a bit broader than most other privileges, both because it lacks the exceptions that the other privileges have, and because it applies not just to government-licensed (and therefore usually quite expensive) professionals. For unmarried people, or people who don’t want to confide in a spouse, the clergy-penitent privilege is often the only one they can afford.

I should note that the conscientious objector cases did read “religion” very broadly, including deeply held conscientious beliefs — including ones unrelated to God or any higher power — held “with the strength of more traditional religious convictions.” It is possible that, absent such a saving construction, the conscientious objector privilege would have been struck down.

But only one of the eight Justices who participated in the case (Justice Harlan) concluded that a religious-objector-only exemption would have indeed violated the Establishment Clause. Even Justice Harlan solved the problem by slightly broadening the exemption to include such philosophical conscientious objectors rather than by saying that the burden imposed on third parties warranted striking down the exemption altogether. In Amos, the Court expressly rejected the view that the religion-only nature of the exemption made it unconstitutional. And the clergy-penitent privilege is likewise generally assumed to be constitutional, with no indication that it has to be broadened to include secular analogs of the clergy (and, as I noted, psychotherapists aren’t really secular analogs of the clergy).

But in any event, say that the loss of benefits caused by the employer mandate exemption is indeed unconstitutional because secular philosophical counterparts of the employer aren’t covered (think someone who is deeply opposed to any destruction of embryos for secular philosophical reasons) — the analog to Justice Harlan’s argument in Welsh. If so, then I would think that the way to avoid such unconstitutionality is, as in Welsh, to read “religion” as covering deeply held philosophical beliefs.

Indeed, the term “religion” in the other comprehensive federal accommodation scheme, the Title VII religious accommodation requirement, has generally been interpreted to cover such deeply held philosophical beliefs. (Title VII is narrower than RFRA in that it applies only to employment and only mandates reasonable accommodation, but it is also broader, in that it applies to private employers as well as to the government.) To be sure, the term “religion” for Free Exercise Clause purposes was read narrowly, to exclude nontheistic philosophical beliefs, in Wisconsin v. Yoder, but the term was read broadly, to include such beliefs, in Welsh. And if the broader reading is necessary to avoid unconstitutionality (which, again, is not clear), I would think that such a reading is preferable.

3. So religious accommodations do often leave nonbeneficiaries worse off, as a result of giving the benefits to the beneficiaries. And, as I noted, they can leave nonbeneficiaries worse off to a considerably greater degree than requiring them to pay several hundred dollars per year out of pocket.

But there are indeed some precedents that suggest that excessive burdens on nonbeneficiaries do indeed pose Establishment Clause problems. The one precedent that squarely stands for this proposition is Estate of Thornton v. Caldor, Inc. (1985).

In Thornton, a Connecticut statute legally required private employers to give any employee a day off on the employee’s Sabbath: “No person who states that a particular day of the week is observed as his Sabbath may be required by his employer to work on such day. An employee’s refusal to work on his Sabbath shall not constitute grounds for his dismissal.” And the Court held that this was unconstitutional.

“Under the Religion Clauses,” the Court reasoned, “government … must take pains not to compel people to act in the name of any religion.” “[T]he Connecticut statute imposes on employers and employees an absolute duty to conform their business practices to the particular religious practices of the employee by enforcing observance of the Sabbath the employee unilaterally designates.” “[T]here is no exception when honoring the dictates of Sabbath observers would cause the employer substantial economic burdens or when the employer’s compliance would require the imposition of significant burdens on other employees required to work in place of the Sabbath observers.” “This unyielding weighting in favor of Sabbath observers over all other interests contravenes a fundamental principle of the Religion Clauses, so well articulated by Judge Learned Hand: ‘The First Amendment … gives no one the right to insist that in pursuit of their own interests others must conform their conduct to his own religious necessities.’”

But in Amos, two years later, the Court made clear that Thornton was limited. Recall that there, too, the statute had an “unyielding weighting in favor of” churches that wanted to hire only coreligionists, with no exceptions when the firing of someone who had strayed from the faith would impose an especially grave burden on him. But the Court said Thornton was inapplicable:

This is a very different case than Estate of Thornton v. Caldor, Inc.…. In effect, Connecticut had given the force of law to the employee’s designation of a Sabbath day and required accommodation by the employer regardless of the burden which that constituted for the employer or other employees. In the present cases, appellee Mayson was not legally obligated to take the steps necessary to qualify for a temple recommend, and his discharge was not required by statute.

As I read it, this limits Thornton to situations where the government imposes a highly burdensome legal obligation on private parties, such as employers, to take steps to accommodate someone. It doesn’t apply in situations where, as in Amos, the government is lifting a legal obligation from private parties — even when this lifting of the legal obligation means that other private parties will be worse off than if the legal obligation remained in place.

The one other Supreme Court case potentially supporting the “employer mandate exemption would violate the Establishment Clause” argument is Cutter v. Wilkinson, which did cite Thornton in discussing the government’s lifting legal obligations from private parties. Cutter did say that “tak[ing] adequate account of the burdens a requested accommodation may impose on nonbeneficiaries” while “alleviat[ing] exceptional government-created burdens on private religious exercise” (and not discriminating among denominations) was sufficient to make a law constitutional, and did say that “[o]ur decisions indicate that an accommodation must be measured so that it does not override other significant interests”; both times the Court cited Thornton.

But the Court didn’t explain what should be done in cases like Amos, where the accommodation didn’t take account of the burdens imposed on nonbeneficiaries (such as the fired employee). Moreover, Cutter involved RLUIPA, a federal law that, in relevant part, mandated religious exemptions in state prisons (and similar institutions), and the opinion closed by saying, “Should inmate requests for religious accommodations become excessive, impose unjustified burdens on other institutionalized persons, or jeopardize the effective functioning of an institution, the facility would be free to resist the imposition. In that event, adjudication in as-applied challenges would be in order.” Inmate requests that impose unjustified burdens on other inmates might (like the law in Thornton but unlike an employer mandate exemption or the law in Amos) involve legally coercive constraints imposed on other inmates. And inmate requests that jeopardize the effective functioning of a state prison would involve federal constraints on a state government, which again are different from the federal government simply lifting burdens on religion imposed by its own law.

In any event, that’s the somewhat uncertain state of the precedent. I think that on balance the argument that any employer mandate exemption would be constitutionally forbidden is unpersuasive — given the other exemptions I mentioned above — but I agree that Thornton and especially Cutter leave some doubt on the matter.

Tomorrow, I will turn to what I think is the much thornier question, which is whether an employer mandate exemption is required by RFRA. And there the arguments about the loss of benefits to third parties become, I think, much more serious.