Questions about harm related to religious protections are particularly weighty at this particular moment in history; religious exemptions have never been more controversial or hotly debated in legal scholarship. Particularly in light of Supreme Court cases like Hobby Lobby and Masterpiece Cakeshop, some scholars have advanced new theories that would place strict limits on the government’s ability to grant religious exemptions that result in harm to third parties. Iterations of this theory, referred to as the “third-party harm theory,” argue that “we should be disturbed by the claim that individual rights can be exercised in ways that harm others.”
In a forthcoming Indiana Law Journal article, “First Amendment ‘Harms,’” I provide a theoretical critique of the generic harm principle on which this theory relies and suggest more fruitful normative questions that we should be asking about harm.
Sufficient or Only Necessary?
Proponents of the third-party harm theory echo John Stuart Mill in claiming that the ability of individuals to exercise their religious liberty rights depends on whether such liberty causes “harm to others.” But third-party harm theorists take this harm principle a step further. Mill argued that harm was a “necessary” condition for restricting rights—government could not restrict individual liberty unless there was harm, but it need not do so. But third-party harm theorists argue that the mere presence of cognizable harm requires that government restrict religious rights. This significantly raises the stakes for determining what counts as cognizable “harm” under their theory.
Relying on a harm principle to justify government interference has a strong, intuitive appeal. At least superficially, it seems to be a theoretical shortcut to avoid other, difficult moral questions about which causes a government should or should not advance—questions on which there is little consensus in a pluralistic society. Pointing to harm seems like a neutral method for bypassing such moral conundrums. Unfortunately, determining what exactly we mean by “harm” unavoidably raises significant moral questions.
If the harm principle is broadened to include more expansive notions like dignitary harm, arguably every action can count as harm in some light. Thus, any action could justify, and indeed require, the government’s restriction of religious rights. However, if we do not want to adopt a purely subjective idea of harm that allows anything to count as harm that negatively impacts someone’s interest, then as Steven D. Smith has recognized, “harm” must become a term of art that pertains only to some sorts of interests and not others. But once a technical definition has been adopted, that definition must operate on top of a deep normative theory about which types of impact count as harm, and why. For example, if we use the “undue hardship” standard offered by some third-party harm theorists, it appears that “harm” is really being used as a technical term of art to mean the following: a negative impact that is, under Title VII case law, more than de minimis and that affects some subset of the population as a result of the exercise of religious rights. But the theory does not recognize as cognizable “harm” a negative impact that is experienced by the religious believer, or by third parties who benefit from the religious organization that is seeking a religious exemption. Nor does the theory offer a satisfactory normative explanation for why “harm” is not cognizable if it is experienced broadly by the population, or why “undue hardship” is the proper metric for what counts as harm.
Superficial Simplicity, but No Real Consensus
The normative appeal of the harm principle trades on its superficial simplicity. But once “harm” becomes a term of art, the normative justification for the theory becomes quite complex. And the plausibility of the harm principle trades on the assumption that there will be consensus about what constitutes harm. But there is no such consensus, only a plurality of views of what harm is.
Indeed, the lack of consensus on harm is highlighted by the fact that three different groups of third-party harm theorists define harm to mean three very different things. Some define it by a materiality standard as a burden that is relevant to decision-making; others define it by an undue hardship standard as a more than de minimis hardship that befalls subsets of the population; and others define “targeted” “material or dignitary harm” as a burden on those who “do not share the [religious] claimant’s belief.” None of these scholars provides clear normative justifications for why certain types of actions count as harm under their definition and others do not. What’s more, the “Do No Harm Act” defines harm completely differently, presenting a specific laundry list of actions that count as harmful, including religious objections to any sort of “health care item or service.”
Given this utter lack of consensus on what should count as harm, it is not surprising that scholars in other fields have observed how a heavy reliance on a generic harm principle almost always collapses in upon itself.
Not a Neutral Principle
This undertheorized reliance on a generic notion of harm fails to achieve its stated aims. If applied in an evenhanded way, the theory would remove religious exemptions for groups like religious minorities that third-party harm theorists generally acknowledge to be worthy of protection. These groups include Muslim prison inmates, Sikhs in the workplace, and Jews seeking kosher diets.
For example, consider the case of David Rasheed Ali, an observant Muslim and prison inmate who requested an exemption from his prison’s restrictive policies that prohibited him from wearing a kufi―a knit skullcap that his religious beliefs required him to wear. One might be tempted to conclude that wearing a kufi is both harmless and costless, making the decision to grant a religious exemption relatively straightforward. But even something as seemingly innocuous as a religious head covering contains a number of hidden potential costs and harms. These include: estimated hundreds of thousands of dollars spent in redistributing staff time and resources to implement a new policy; reduced resources for other inmates’ healthcare, activities, facilities, or food; a heightened physical risk for prison guards, who must enter an inmate’s “strike zone” in order to search such personal items; and an increased risk that deadly contraband might be hidden in headgear.
On the other hand, failing to grant an exemption causes spiritual and dignitary harm to Ali, who must violate his conscience. And numerous studies suggest that providing religious protections for inmates decreases prison violence and results in significant, rehabilitative positive externalities, not just for other inmates and security staff, but for society at large. In light of these competing and varied externalities, simply to rely on a supposedly neutral harm principle does not resolve the quandary.
Or consider the case of Kawal Tagore, an IRS employee. Ms. Tagore brought an action against the IRS for refusing to allow her to wear her kirpan, a dull blade that is one of five articles of faith that Sikh adherents are required to wear. Because the case involved the federal government acting as an employer, both the Religious Freedom Restoration Act (RFRA) and Title VII’s de minimis standard applied. Under the undue burden standard—the standard some third-party harm theorists advocate as the proper approach—the court simply accepted the IRS’s word for it that to require that security officials ensure that the employee’s blade was dull enough was more than a de minimis harm to the employer. But under RFRA’s rigorous strict scrutiny review, the analysis was much different. The same court noted that the government had been inconsistent in its enforcement of its policy. For instance, it let other weapons into the building for lawful purposes, and it had let Sikh individuals wear their kirpans in buildings like the White House. Thus, the court determined that RFRA’s “fact-sensitive inquiry” required reversal of the lower court. It issued a remand for further inquiry into less restrictive alternatives to accommodate this employee’s religious beliefs. But Ms. Tagore’s religious beliefs would not have been protected under the standard that third-party harm theorists advocate.
Finally, consider the case dealing with a request by Jewish prison inmates in Florida for a kosher dietary accommodation. Some third-party harm theorists say that such a religious accommodation is “normatively unproblematic” so long as the government is taxing just the public—as opposed to nonreligious inmates—for this kosher diet. But this distinction proves both illusory and unworkable as a protection of religious rights in the real world. For example, in one Eleventh Circuit case where an inmate in Florida requested a kosher dietary exception, the state officials responded that the costs of providing such an accommodation would result in less funding for “roofs for prisons, mental health and medical care for inmates, and salaries for security staff,” and that reduced resources could even compromise the “security and safety of the institutions.” The Department also argued that the increased costs of the diet would result in a “hiring freeze,” as well as a situation where the Department “would have to eliminate 246 staff positions in its already minimally staffed facilities.” All of these harms, if real, would likely constitute more than de minimis negative externalities for identifiable third parties. The government was likely exaggerating to some extent the consequences from increased costs in both of these cases, but the officials were correct to note that to spend more funding on kosher meals effectively operates as a tax on other inmates by leaving fewer resources for their needs.
If a religious exemption could be defeated any time the government could argue that the exemption effectively taxed some segment of the population disproportionately, few if any religious exemptions could survive. In all the illustrations above, the exemptions for religious minorities—which third-party harm theorists view as normatively desirable—involved cognizable harm. Thus these exemptions likely could not have survived an evenhanded application of the third-party harm theory.
Weighing Competing Harms
Given the normative and descriptive shortcomings of the third-party harm theory, it is not surprising that courts are not, in fact, treating the presence of generic harm alone as categorically requiring the government to restrict religious rights. Instead, at times courts recognize and allow significant amounts of harm to third parties in order to protect every type of First Amendment right—not just religious rights. While the presence of harm may be a necessary condition to justify the government’s restriction of First Amendment rights, the harm must also have certain characteristics.
Based on these characteristics, courts weigh a variety of competing harms, classifying them in three specific categories: (1) prohibited harms (which are categorically impermissible); (2) probative harms (which can be balanced against one another); and (3) inadmissible harms (which are given no weight, regardless of how severely or disproportionately the third parties experience them).
This descriptive framework has important normative implications. A clear understanding of the role that harm plays in courts’ treatment of various First Amendment rights highlights the fact that the protection of any right inherently concerns harms competing from either side of the ledger. The real question with which courts are often grappling is what the proper balance of “harm” ought to be. This framework does not treat the presence of any harm as a sufficient condition for restricting any right. Rather, it treats harm as part of an equation: it must be weighed in a consistent way for the government to determine the most socially beneficial outcome. In addition, it is worth noting that the issue of how to define harm under the third-party harm theory need not arise in this context—or certainly not with the same acuteness. Whichever definition a court may employ for harm that affects third parties, one must grapple with how an action will reciprocally harm groups in a similar way on the other side of the ledger. Such a reciprocal approach mitigates the risks involved with technical definitions of harm, which can be used to justify special pleading to prohibit some sorts of religious beliefs that may be socially disfavored.
Which brings us back to the “Do No Harm Act.” It would gut the Religious Freedom Restoration Act by removing religious liberty protections that “harm” others. That would be a mistake. The protection of any First Amendment right inherently involves balancing competing harms on both sides. Moving beyond a false dichotomy of “harm” versus “no harm” allows one to ask much more fruitful normative questions about harm, including whether there is a justifiable trade-off between the specific harm and the social goods that it provides; whether institutions can be modified to mitigate avoidable harm; and whether disproportionate harms can be distributed in more just ways. Asking these questions will not immediately resolve all the conflicts that arise in the religious context, but it can help point us to the results that will most likely reduce overall social harm by protecting both the interests of third parties and important religious rights.