Constitutional Rights and Their Relevance to Discrimination

Because in some discussions about exceptions, the status of constitutional rights and the duties of private citizens is a bit confusing, it helps first to clarify that. When the US Supreme Court held in 1973 in Roe v. Wade Footnote ² that women had a constitutional right to receive abortions, and when it held in Obergefell v. Hodges forty-two years later that same-sex couples had a constitutional right to marry,Footnote ³ those rulings directly controlled what legislators and government officials could do in various ways. They did not dictate how private individuals and organizations had to treat private citizens. People are free to make such choices on whatever grounds they choose in the absence of statutes or regulations that either demand the provision of services generally or forbid particular forms of discrimination. For most of our country's history, these statutes and regulations were much less extensive than they are now, entailing that both private individuals and organizations were much freer legally to rely on their own inclinations about how to treat others.Footnote ⁴ This freedom still exists about highly personal matters, such as whom to invite to one's home and whom to date. But for a great many interactions, various negative classifications are barred by antidiscrimination laws.

The Obergefell decision does not by itself preclude private discrimination against same-sex couples; a statute can explicitly, or implicitly, bar such negative treatment. A number of states now have such explicit laws, but many others and the federal government do not. The crucial issues about exemptions come up when such laws are being considered. As a matter of principle, should some exemptions be granted and, if so, how extensive should they be? A rather different concern is whether including exemptions will be needed to get an antidiscrimination law adopted. Interestingly, in Utah, the dominant Mormon Church, which does not itself accept same-sex marriage, did not oppose a law, given that it had significant exemptions.Footnote ⁵ Since the exemptions needed for passage can vary from state to state and are not really about what is right in principle, I shall essentially disregard that; but it can obviously matter for legislators who are deciding what to do in a climate of divided opinion about what is right. That climate is now pervasive in many parts of our country.

Lying in the background here is a constitutional issue I shall mention but not explore.Footnote ⁶ In the United States, if a state legislature chooses to grant an exemption that is much broader than those that exist with other antidiscrimination laws, or if it fails to pass any antidiscrimination statute to protect gay couples, either approach can arguably be seen as a denial of equal protection—treating these possessors of constitutional equality rights worse than others.

A different kind of constitutional question explored in depth by Steven D. Smith in The Rise and Decline of American Religious Freedom has relevance that is more indirect. In an extensive analysis, Smith explains various perspectives about religious liberty in the history of our country. He emphasizes that for most of that history nonestablishment concerned how the government reacted to private institutions and convictions, not actual religious references and practices within the government itself. Over the last decades, although the cases represent no clear doctrinal principles, the Supreme Court has curtailed what government institutions can do and say about religious premises. A central thesis of Smith's book, one certainly warranting reflection, is that this extension of nonestablishment to curb what can be done within the government has itself contributed to a kind of secular egalitarianism that sees religious views as basically misguided and not deserving of any accommodations from normal duties. This all contributes to an ideological split in the country and to strong resistance to any exemptions regarding same-sex marriage. In urging the particular cause-and-effect relationship between reduction in government expression of religion and a curtailment of religious exercise caused by negative views about religion, Smith is making a claim that, as he recognizes, is not widely shared. Although I believe his thesis is well developed and merits careful consideration, I myself am doubtful that the restrictions set on the government itself do, on balance, significantly impair free exercise. (Indeed, I explore numerous conflicts between free exercise and nonestablishment values in a forthcoming book on that subject.Footnote ⁷ )

Religion and Conscience

A significant question about many individual exemptions is whether they should be limited to genuine religious convictions or include other claims of conscience. Tied to this question is what actually counts as “religious” in this context. The exact boundaries of “conscience” are also far from precise. Mary Anne Waldron has written a book that explores the significance of conscience, how it is broader than specifically religious claims, and has, despite its inclusion in a Charter of Rights and Freedoms, received inadequate attention by Canadian courts and scholars (Waldron, 195–228). Waldron covers many specific issues that resemble those in the United States, among them Sunday Closing laws, the use of religious criteria for employment, abortion requirements, same-sex marriage, and the status of government employees. Waldron offers an interesting discussion of cases and controversies over assisted suicide, which may in the future become a more prominent issue in the United States.

Waldron's book has three fundamental values for American readers.Footnote ⁸ The most obvious is that it provides a comparison of legal decisions and political disagreements in a country that resembles our own in some important respects. Second, it contains a careful analysis of discrete controversies and concepts like “religion” and “conscience” that extend well beyond the boundaries of Canada. Third, it provides accounts of general approaches that are definitely not limited to that country. According to Waldron, Canadian courts have avoided a “hierarchy of human rights,” instead favoring a balancing process when rights are in conflict. (To be clear, one can believe both in some degree of hierarchy and in balancing.) She concludes by writing that too much emphasis has been put on aims for uniformity and too little on accommodation and acceptance of difference. It is this latter approach that I believe must be given more attention in this country.

Some may believe that because religion is special and because the free exercise of religion is a fundamental right, a limit that excludes nonreligious conscience is always appropriate. Others think this is itself an inappropriate form of discrimination, one at odds with the principle that the government should not “establish” religion. They urge that the state should always be neutral. I think each of these positions is misguided. Sometimes religious claims have a strength that is hard to imagine for other assertions. Obvious examples are responses to prohibitions on the drinking of alcohol or ingestion of peyote and the killing of animals in certain ways. If wine and peyote are the center of Roman Catholic mass and Native American Church worship, respectively, we have no nonreligious analogue of similar strength. The same is true of the Orthodox Jewish belief that the religious requirement of kosher food demands that animals be killed a certain way. When it comes to being a pacifist, it is much easier to imagine a powerful conviction that is not religious in any typical sense.

The pacifist example, when looked at in light of US history and some Supreme Court decisions, helps to show just how complicated this issue about a limit to religious conscience can be. Early in the United States' history, a young man sometimes qualified for a draft exemption only if he was a member of a pacifist religious group, such as the Society of Friends. In 1940, Congress adopted a law that required individual religious convictions but did not demand group membership. In response to divided judicial interpretation, Congress in 1948 provided that “Religious training and belief in this connection mean an individual's belief in a Supreme Being involving duties superior to those arising from any human relation, but do not include essentially political, sociological, or philosophical views or a merely personal moral code.”Footnote ⁹ This language obviously was aimed to embrace a narrower concept of religion.

In two decisions, the Supreme Court essentially stretched the language beyond what it clearly meant.Footnote ¹⁰ In the second of the decisions, it was dealing with a claimant, Elliot Welsh, who had actually struck the word “religious” from his application, and referred to “reading in the fields of history and sociology.”Footnote ¹¹ The eight sitting Justices divided 4-4 over whether Welsh's views were covered by the statutory language. Four acknowledged that they were not.Footnote ¹² Four disagreed, stretching the language far beyond its actual meaning.Footnote ¹³ Justice Harlan took the statute for what it conveyed but concluded that since in this context drawing the line thus was unconstitutional, Welsh was entitled to the exemptionFootnote ¹⁴ The draft law and these cases tell us more generally how difficult it is to draw a clear and acceptable line between religious bases and other forms of individual conscience. However, it does not follow that that effort is never warranted.

Another important concern about a special place for religion is more general laws framed in terms of religion. We have two such crucial federal statutes, the Religious Freedom Restoration Act, or RFRA,Footnote ¹⁵ and the Religious Land Use and Institutionalized Persons Act.Footnote ¹⁶ RFRA, of course, has been the basis for the Supreme Court's protection of private enterprises regarding contraceptive insurance. Before looking at that, however, I must first note the source of that law and the relation between it and the coverage of nonreligious conscience.

Before 1990, the Supreme Court embraced a constitutional standard that supported claims of free exercise to deviate from various duties, and this standard was applied by lower courts in a variety of contexts. In Employment Division v. Smith,Footnote ¹⁷ the Supreme Court ruled that a religious person had no constitutional claim if a law was generally cast and not aimed at religious practices. In RFRA, Congress essentially reenacted what it took to be the prevailing free exercise standard prior to the Court's decision. We will look at the particular segments of RFRA in what follows, but the statute does raise the question whether singling out religion is appropriate. In terms of what Congress and many state legislatures have done, the Supreme Court's answer has been “yes,” that aspect of RFRA is not constitutional.Footnote ¹⁸ That conclusion alone does not answer what should be done about nonreligious claims if they are closely similar. For some circumstances, the granting of a RFRA exemption may actually require that a nonreligious one of the same kind be treated the same way. More broadly, RFRA and other similar statutes do not tell us what should be done about nonreligious claims in statutes dealing with narrower specific subjects, such as abortion and same-sex marriage.

Impairment of Conscience

What degree of impairment of one's convictions, religious or not, is needed to justify an exemption, and how is that administrable? This turns out to be a troublesome question to which Justice Alito's opinion in Burwell v. Hobby Lobby Footnote ¹⁹ provided a RFRA interpretation that is decidedly unsatisfactory.Footnote ²⁰

Let me make clear how, in contrast to the Court's first step, the law's coverage of individual claims may relate to claims on behalf of groups. The first requirement under RFRA is that the claimant must be a “person.” Given that the statute was designed to reintroduce the earlier constitutional standard, and that prior doctrine definitely did not make clear whether free exercise rights were possessed by for-profit businesses, the application of RFRA was hardly simple. The Court relied on an obscure provision in federal law that treats “person” very broadly “unless [the] context indicates otherwise,”Footnote ²¹ and it also gave some weight to the fact that no one doubted RFRA did include definite religious organizations such as churches and synagogues and religiously controlled charities. One could easily resist the Court's broad coverage in Hobby Lobby on the basis that for-profit businesses should not have the same status, partly because of their overarching objective to make money. Another problem is what to do about ordinary businesses with multiple stockholders who do not share the same convictions as the leaders. Justice Alito makes some effort to put these aside while covering closely held corporations; but what does that entail if 40 percent of the stocks are held by outsiders and business matters are controlled by a single family or unified group in charge? The Court's ruling that RFRA covers businesses is genuinely debatable, but it has no bearing in relation to individual claims of conscience. An individual is obviously a “person,” so plainly RFRA applies to her, whatever is done about companies.

RFRA's next requirement is that for someone to qualify for an exemption, performing a general duty must constitute a “substantial burden.” Here, the standard is set for whoever otherwise qualifies, and the Court's treatment in Hobby Lobby, if adhered to in the future will have a strong importance for individual claims of conscience relying on RFRA or a similar formulation, rather than a specific statutory right to forego certain behavior. Justice Alito's basic approach is that the burden depends fundamentally on the religious convictions of those making the claim, not on any outsider's assessment. This is an approach that makes sense as a matter of principle, but is deeply questionable as a general legal standard.

We can see how the principle is sound if we imagine two friends who are both subject to a military draft. One believes that any involvement in the military is deeply wrong, violating his core sense of how we should treat others. The second believes the nation should not have a draft and that it would be morally preferable to perform civilian service, but he does not conceive of submitting to the draft as violating any fundamental conviction. We could fairly say that submission would impose a “substantial burden” on the first man but not the second. Unfortunately, reference to one's personal sense does not show that the standard is generally viable for legal purposes.

Two key difficulties with actual application are whether people have nonconscience reasons to be excused from a general duty and whether the government is in a position to assess someone's asserted convictions. For example, there is a core problem with contraceptive insurance that can extend to some individual claims of exemption. The litigants in Hobby Lobby underwent considerable expenses to litigate, and they risked a negative decision; the leaders of the relevant companies definitely had strong convictions that their insurance should not cover some key contraceptives. But now their right is established. The alternatives used to provide this insurance for the workers of companies are not simple, and vary a bit depending on whether the companies use separate insurance companies or directly provide insurance for their own workers. In both instances, the end result is that the employers pay somewhat less for insurance if they are not charged for some contraceptive coverage. Given at least a modest financial incentive and official acceptance of any asserted claim, owners of closely held companies who have mild objections to the relevant contraceptives may well decide to claim that providing insurance constitutes a “substantial burden.” This may be especially true if the owners are connected to religious groups that regard the use of these contraceptives as immoral.

We can imagine similar candor problems for individual service to a same-sex married couple. Someone who has many acquaintances who are opposed to such marriages and who himself believes no constitutional right for it should exist may seek any exemption to essentially advocate a political position and perhaps obtain further approval of others that he serves. He may do so even if he does not believe that serving the couple in the relevant way would strongly violate his moral convictions. In terms of accepting claimants' assertions, it is very important whether we can easily conceive of claims that are less than sincere.

A central concern about administering a standard that rests on individual conditions is how far officials can delve into these. In contrast to draft exemptions, for a great many claims, officials need to simply accept what people assert. This is a very serious problem with making “substantial burden” totally subjective, and it is one about which the Court's approach to an organization's claims can bear on individual claims. The most obvious alternative approach to “substantial burden” is to ask how most people would regard the burden if they shared the core moral convictions that a claimant asserts. In a subsequent section of this essay I focus on degrees of involvement, which relate to what should count as a “substantial burden.”

An intermediate position between a purely subjective approach and an inquiry about most people's reactions would be to require some external evidence that the leaders of a group do perceive such a burden. That could be shown by their making the claim itself, if it is hard to imagine anyone doing so without feeling a substantial burden. It could also be supported by membership in a religious body that definitely sees even minor, indirect involvement as definitely forbidden.

A recent book, The Crisis of Religious Liberty, edited by Stephen M. Krason, contains a number of essays that present Roman Catholic views on religious liberty and exemptions. Among these, Robert George, a leading natural law philosopher, argues that insights into what is good should, contrary to much contemporary liberal thought, take a priority over what is right; and that respecting human dignity includes a strong right of religious freedom. Gerard Bradley urges that religious liberty is a natural right now somewhat threatened. Kevin Schmiesing treats historical and present disagreements about whether the Catholic faith fits with American views of religious freedom; he uses issues about reproductive rights as an illustration of an uneasy relationship between the church and the state. Robert Destro concentrates on the Supreme Court's decisions concerning race and religious as confused and contradictory. Kenneth Grasso argues that the aim to achieve civic unity can threaten religious pluralism.

These essays and those by other authors in the book explore their subjects in considerable depth. They present not only a valuable source of how Roman Catholics may best see things but also how non-Catholics should take account of Catholic views. These insights are important in two ways. Much of what is addressed in the essays can have direct relevance for how one regards religious understanding and civil life. And given the substantial number of Roman Catholics in the United States, these essays can help readers understand the perspectives of serious Catholics and how they should be treated. To take a relevant example, one might be persuaded that if religious groups see behavior as fundamentally wrong, their members should have greater liberty to deny indirect services to those they see as morally “guilty” of certain legally protected exercises of rights than I have recommended in this essay and the two recent books.Footnote ²²

Harms Exemptions Cause to Others

A central objection to many exemptions, including ones from antidiscrimination laws, is that they cause harm to others. Indeed, some opposition to exemptions in regard to treatment of same-sex married couples is grounded in the idea that the harm to the couples is sufficient to make any exemptions inappropriate. That harm to others is relevant is obviously correct. If others will suffer significantly, that is definitely a reason to rule out an exemption or to limit its scope. On the other hand, the identification of a disadvantage of any sort does not necessarily rule out the desirability of an exemption. This is perhaps most obvious with pacifists. It has sometimes been urged that their exemption is unfair because some others will be drafted as a consequence of their being excused from service. There are two obvious counters to this concern: One is that if a pacifist will submit to criminal punishment rather than joining the military, he is effectively unavailable for service and another man will need to be drafted; the exemption avoids putting in jail people who do not by most criteria merit incarceration. The second concern involves genuine pacifists who may submit to military service in order to avoid jail time. What will they do when the time comes to battle the enemy? They may refuse to fire their weapons or otherwise follow orders because of their pacifist convictions. Such a consequence could be much worse for other soldiers than would the granting of an exemption to those who object to military service.

When it comes to antidiscrimination laws, exemptions can affect both the availability of services and the dignity of those who do not receive them. It is sometimes asserted that private individuals and organizations should not have a right to deny services. If one considers historical practices, this simple assumption faces two obstacles. The first is that before the enactment of antidiscrimination laws, private individuals and organizations were largely free to favor others on whatever grounds they chose. As a consequence, others often did suffer disadvantages. And when antidiscrimination laws were enacted, they nearly always contained some exemptions that could produce occasional disadvantages to those denied service or status by religious organizations. Within the last few years, the Supreme Court unanimously concluded that churches and other religious bodies, such as parochial schools, could fire “clerics”—understood broadly to include some teachers—on whatever grounds they chose.Footnote ²³ Further, the government could not even challenge such dismissals on the basis that the asserted grounds were insincere. This clearly works to the disadvantage, for example, of Roman Catholic women who would like to be priests and to “clerics” who correctly believe that their superiors have falsely stated a ground for dismissal. Although some disadvantage to others is not a simple, straightforward basis to deny an exemption, that concern should figure significantly in deciding whether one should be granted and what its scope should be.

Conscientious Objection to Performing Duties

In a liberal democracy, we do have different views about what it is morally right to do, but a basic premise is that we should generally not treat others negatively simply because of unalterable characteristics. It is, thus, more troublesome to allow negative treatment of most kinds toward others because of race, gender, or ageFootnote ²⁴ than based on acts in which they are now, or have been, engaged. In this respect, discrimination against gay individuals generally or same-sex married couples can formally be seen as concerning actions to which someone objects, but given the strong tendency of most people to seek sexual engagement with one or more of those to whom they are attracted, this comes very close to negative treatment based on a person's natural sexual inclinations (But cf. Krason, Crisis of Religious Liberty, 161).Footnote ²⁵

Assuming that a conviction that one should not perform an ordinary duty is based on what someone has done, is doing, or will do, rather than his or her unalterable characteristics such as race, gender, or age, does it matter how direct one's involvement with the actions will be? The obvious answer is “yes,” and I believe this is crucially important when one focuses on same-sex marriage.

There are two important aspects of this question. The more straightforward is what it is right for the law to provide. The other is the extent to which the Supreme Court's ruling about closely held businesses in Hobby Lobby bears on individual claims of conscience. On the latter aspect, Justice Alito's approach to “substantial burden” in RFRA is highly relevant. If that burden depends simply on what is subjectively perceived by those required to perform a duty, rather than some objective standard, then the degree of directness is not by itself legally relevant to whether one has a right to avoid performance. A person who feels she should not provide even the most indirect service, would have that right. If this approach applies to the owners of businesses, it also does so for claims of individual conscience. I have already indicated that to make “substantial burden” completely subjective is quite troubling, especially when people may have financial or political reasons to refuse performance, even though carrying out their duty would not deeply trouble their conscience.

Before focusing on what are sensible lines, I shall say a little about the provision of employee health insurance and how that relates to the paying of taxes. When free exercise constitutional protections were still fairly broad, an Amish employer sought to be excused from paying Social Security taxes, given that his employees were Amish and the religious group did not believe in accepting Social Security. The Supreme Court rejected the claim.Footnote ²⁶ How different is providing health insurance? The closely held company that does so does not directly choose or encourage the purchase of contraceptives and has no involvement in their actual use. If one sees the insurance situation as different from taxes because the nonpayment of taxes by some increases the burden on other taxpayers, in fact the requirement in Hobby Lobby that insurance companies themselves bear the costs actually entails a similar outcome in many circumstances. The reason is that if a business directly provides insurance for its workers, a separate insurance company must be drawn in to cover contraceptives, and that company will effectively be reimbursed by having to pay less to the government.Footnote ²⁷ Although it is highly unlikely to happen, one might well say that the logic of Hobby Lobby concerning burden and indirectness supports the notion that individuals should be able to rely on RFRA to refuse to pay some or all taxes, if contributing to certain government expenditures violates their conscience.

I now turn to more specific issues about indirectness, focusing mainly on abortions and same-sex marriage. This discussion does actually bear on how a generally cast statute like RFRA should best be interpreted, since the capacity of legislators to focus on specific problems may suggest less broad interpretation of general laws. After the Supreme Court discerned a constitutional right to abortion in Roe v. Wade, Congress adopted statutory law that provided that federal funding could not be used to require that medical facilities and individuals “perform or assist” abortions.Footnote ²⁸ The law went further and required that the owners of the facilities themselves could not insist that their attitudes about abortions be imposed on their workers. Hospitals performing abortions could not insist that all doctors and nurses do so; hospitals opposed to abortions could not penalize doctors and nurses who participated in them outside their facilities.

Abortions are clearly a subject for which exemptions from standard duties are appropriate. Although many of those with competing views about the morality of abortion believe otherwise, I do not think that ordinary reason can tell when life that warrants protection begins. There are powerful reasons, including the rights of pregnant women and the ineffectiveness of enforcing a criminal law forbidding abortions, to have a legal right to obtain abortions. However, it remains true that the fact that people believe that abortion is a form of unwarranted killing provides a strong reason for not insisting that these people actually perform them. Occasionally, it is asserted that when people undertake jobs, they should simply perform all their duties, but that hardly is persuasive for doctors and nurses who engaged in their occupations when abortion was still criminal in virtually all states or did so later when exemptions were well established.

The federal statute and similar state laws raise a number of questions. These include whether a line should be drawn at religion, whether a qualification to exemptions is needed, and whether “perform or assist” is a desirable formulation. As with pacifism, we can imagine nonreligious convictions that abortions should not be undertaken. Some people with these views may have been persuaded by the kinds of ostensibly nonreligious arguments made by natural law scholars; others may have clung to this feature of a religious connection that existed early in their lives but from which they have now departed; still others may rest on a kind of personal intuition generated by seeing electronic photographs of babies-to-be within the wombs of women. Even if the great majority of objections to abortions will be religiously grounded, it does not make sense to limit an exemption in that way.

In terms of a possible qualification to a right not to perform abortions, the answer is “yes.” If someone's service is desperately needed for a woman to receive an abortion, he should not be able to refuse. Virtually all state laws, unlike the federal statute, contain such qualifications, and courts are inclined to read them in even if they are not explicit. I shall not explore just how these qualifications may best be cast.

“Perform or assist” is a sensible formulation. When thinking of how to draw lines of legal coverage, we need to consider what makes sense in principle and what gives adequate notice and is administrable. Not infrequently the sensible legal boundary needs to be more straightforward than what would be ideal if perfect application were feasible. For abortions, “assist” is the crucial border. It is not totally precise, but if we think of what goes on in a hospital, most degrees of involvement are plainly on one side or the other. If a nurse hands instruments to a doctor who is performing an abortion, she is definitely “assisting.” The officer who allows patients into a hospital, the janitor who cleans rooms, and the nurse who simply attends to all patients who are staying in the hospital, are not assisting. Although we can imagine borderline situations, such as a nurse giving special concentrated care to a woman who is to receive or has just received an abortion, “assist” is clear enough to be an appropriate, workable standard.

What makes sense in terms of claims for exemptions regarding same-sex marriage? I believe that in the case of same-sex marriage, a crucial distinction exists between actual participation and subsequent services other than adoption. All of us engage with others with whom we have core differences in some moral convictions. We rarely, if ever, refuse involvement in unrelated matters simply because of those differences. But if we are asked to help perform an act that we believe is deeply wrong, we will say “no,” at least if we have the courage to do so.

If we reflect on what providers of services, individuals and businesses, have done about most matters, we can reach a similar conclusion. Many religious individuals and organizations believe that abortions are wrong, that people should not be engaged in sexual relations outside of marriage, and that civil divorces are not by themselves sufficient to end a marriage. Yet few, if any restaurants and hotels are denying service to unmarried couples, or people who have remarried after civil divorces, or women known to have had abortions. Of course, much of this information is not obvious, but if this really mattered deeply we might expect at least some inquiries of those who request services. In fact, services are rarely denied to individuals who have been known to have committed serious crimes, including abuse of family members. In brief, the provision of ordinary services does not typically seem to entail any serious involvement in the practices of those who receive the service.

If one believes same-sex marriage is against God's will, actual participation in the ceremony itself is different. We certainly do not expect churches and clerics opposed to such marriage to have to perform the ceremony, and it is highly likely that the right to decline will be seen as constitutionally protected by the Free Exercise Clause. Exactly what counts as genuine participation in the ceremony of the marriage is arguable, but to take two actual cases, I believe being the main photographer, taking hundreds of photographs, many of which will last a lifetime, should count as participation,Footnote ²⁹ but simply baking a cake for a wedding celebration should not.Footnote ³⁰ A cake baker should not get any exemption unless he must convey a specific message to which he strongly objects.

When it comes to ordinary public services, it becomes very hard to distinguish negative treatment of gay married couples from negative treatment of all those with evident homosexual inclinations. Strong evidence that this distinction is less than crucial is that virtually no one providing services tries to assure that heterosexual couples are properly married and have been so after being genuinely single or going through appropriate terminations of their prior marriages. Interestingly, in this respect Pope Francis said on June 16, 2016, that many couples now were not properly married because the partners did not have the right sense of commitment to a permanent union. Even if they could figure out whether that was true about a particular couple, would we expect Roman Catholic individuals and closely held companies run by Catholics to refuse services? The answer is “no.” Of course, one thing that is special about gay couples is that if they display physical affection, they are likely to be married or unmarried romantic couples. If they seek an apartment to live together and use the same last name, they are likely to be married, although they could simply be friendly brothers or sisters. For gay couples, little or no inquiry is usually needed to figure out one's disinclination to equal treatment if that extends to all homosexual involvement or all same-sex marriages or both. As noted, it is usually much harder in the case of same-sex marriage to distinguish negative treatment based on fundamental characteristics from a negative response to forms of behavior.

It is interesting, in this respect, that the sense of the wrongfulness of homosexual actions can exist for some who both have that inclination and belong to a religious denomination that teaches that it is deeply wrong. After the Orlando killing of forty-nine people in the Pulse Nightclub, which primarily served gay people, it turned out that the killer had visited there with some frequency, and the question arose whether he himself had homosexual inclinations. On June 17, 2016, news host Rachel Maddow interviewed Sohail Ahmed, a British citizen who acknowledged that as a young Muslim accepting an extreme Islamic view, he felt his own homosexual tendencies were deeply wrong and needed to be countered.Footnote ³¹ He seriously considered violent terrorist acts against gays as a way to overcome his sexual attraction by favoring what he believed God would want him to do. A more traditional illustration would be Roman Catholics who have had homosexual inclinations and, believing they were seriously wrong, have decided to live without sexual involvement. Some with that perspective have become priests and nuns. These examples help to make clear that in our world we have no complete or simple division between religious believers and those sexually attracted to others of the same gender. Some of the latter are believers who are deeply troubled by their inclinations; and many religious persons, thankfully an increasing number, believe that, despite certain passages in the Bible, either nothing is negative about same-sex relationships or that even if those relationships are somehow less desirable, people involved in them should be treated with acceptance and respect.

The reality is that virtually all individuals and businesses provide services to almost anyone who has engaged in acts that they think are wrongful, or have failed to do what they believe is morally needed, such as a proper marriage. This fact about our social life strongly suggests that claims of broad privileges to treat same-sex couples unfavorably are largely a consequence of outright negative views about anyone with homosexual inclinations, or a form of political opposition to same-sex marriage, or both. To be clear, individuals who believe such marriage is basically wrong may often find it hard to single out what their conscience directly tells them not to do from these other influences toward negative treatment. But I believe we need to come back to the general provisions of services to people who may or may not have behaved in accord with what the providers think is right. This reality strongly supports not granting a broad exemption from equal treatment and not perceiving such treatment as a “substantial burden” under a statute like RFRA.

Should an exemption here be cast in terms of religion? Although increasingly the objections to same-sex marriage are ones of religion and not the culture generally, when it comes to the limited range of actual participation in the ceremony, extending the privilege to any claim of conscience makes sense. Matters become more complicated if one thinks of broader exemptions. Some readers will almost certainly believe that what I have suggested is intrinsically too narrow, and some legislators will conclude that in their state, or for the federal government, a broader exemption is needed to achieve political passage of an antidiscrimination law. My sense is that if a broader exemption is politically necessary, when it comes to ordinary services after marriage, a limit to religious bases is appropriate because it is hard, in the case of same-sex marriage, to imagine powerful nonreligious claims of conscience to providing those services, and broadening this aspect of an exemption will contribute to political divisiveness.

One form of involvement is itself neither direct participation in the ceremony nor an ordinary unrelated service after the marriage takes place. That is what adoption agencies and those who work within them do, and what individual parents may ask for when they put their children up for adoption. I will not go into detail here, but we do need to recognize that adoption is special in two ways. Providing children is a much more direct involvement in a marriage than are ordinary services. And, perhaps even more important, the strong reasons for a legal right to same-sex marriage do not themselves establish that these marriages are relatively indistinguishable from ordinary marriages insofar as children are concerned. Perhaps it is helpful for children to have at least one operating parent who is of the same gender, and it may usually be desirable to have close exposure to adults of both genders. These are definitely not solid reasons to rule out same-sex couples as adoptive parents, since many children are now well raised by such couples and many other children are raised by single parents or by male-female couples who are in frequent conflict. But it would probably be going too far to insist that for adoption the genders of a couple must be treated as completely irrelevant. To draw a comparison, people generally cannot treat others worse because of their age or economic status; but an adoption agency can definitely prefer a couple in their thirties with adequate income over a poor couple in their sixties.Footnote ³² In short, this is one particular subsequent involvement with a marriage that is fairly direct and for which significant reasons exist neither to allow a complete bar on service nor to provide absolutely equal treatment for every possible adoption.

Private Citizens and Public Officials

Apart from whether a person is directly involved or not, what position does he occupy, as between the public and private spheres? For exemptions, the key difference is between government officials and private citizens. Let me start with how the position of an actor relates to constitutional rulings by the Supreme Court. In a prior section of this essay I explain that such rulings do not bear directly on what most private citizens may or may not do. Of course, a decision that same-sex couples have a right to marry does directly concern what those couples may do; but it does not dictate how other private citizens and enterprises must treat those couples. That depends on an antidiscrimination law or some other form of statutory requirement.

Matters are different for government bureaus and officials. Basically they must give effect to a constitutional right. It is partly as a consequence of this reality that when Kim Davis instructed her office not to perform any marriages for same-sex couples, that was both a radically implausible claim of any right to decline, but also an unpersuasive starting point for any private claims for exemptions. Does it follow that no exemptions or other excuses from performing duties are warranted for government officials? Some people believe so, but I think that absolute approach is not sensitive to all that properly counts.

The “no exceptions” approach for a government worker rests on the premise that if one's conscience is sharply opposed to performing any ordinary official duties, he should either never take the job in the first place or give it up when the conflict reveals itself. That disregards or oversimplifies two important factors about many government duties. The first is that a person may take a job with no expectation that it will cover certain matters. A woman who joined a government agency dealing with marriage and family issues two decades ago would not have expected that same-sex couples would obtain a right to marry. To say that that woman should now either perform these regardless of what her conscience tells her or give up her job is harsh if the couples who seek marriages can do so without inconvenience or embarrassment because others in the office will perform them. The second relevant consideration is that many government offices perform multiple tasks. If a worker objects in conscience to performing only one, that may have little or no effect on his overall contribution or the effectiveness of the staff as a whole. One example I offer here is personal, although it fell short of an absolute refusal on the basis of conscience. When I was a deputy solicitor general dealing with most criminal cases, my ordinary responsibilities would have included reviewing and editing a government brief that defended intensive surveillance of organizations opposed to the Vietnam War. When I told Solicitor General Erwin Griswold that I believed the government's position was deeply unjust and I would prefer not to work on the brief, he shifted the task to someone else. My not working on this one brief was a very small segment of my overall responsibilities. A more general and presently common example is government lawyers in states with capital punishment who are not required to work on those cases if they have strong objections to the death penalty. When one thinks about this kind of example, one can see that it is not always best to simply require government employees to perform every duty that falls within their responsibilities.

Three clarifications are important here. The first is obvious. These possible concessions to individual conscience are radically different from making a whole office unavailable to perform its duties. If a right to marriage has been clearly established, and you control the office in your jurisdiction that performs marriages and issues marriage licenses, the office itself does need to perform those responsibilities for all couples with the right to marry. The fact that the objection in conscience of Kentucky county clerk Kim Davis led to nonissuance by her entire office was a highly unconvincing kind of claim for exemption and perhaps could better be seen as a strong objection to what the Supreme Court had done.

The second clarification concerns what is plain when someone takes a job. If offices themselves must perform a duty that is central to their responsibilities, any argument for a special concession is much less compelling for someone who takes a job fully aware of its main components.

The third clarification involves the way in which someone is excused from duties. There can be a formal exemption or, as in my personal example, an exercise of discretion by one's supervisor. There can also be an announced policy by a supervisor—say, not to insist that lawyers work on capital punishment cases—that provides advance notice but falls short of a definite legal right. When it comes to government duties, these less formal bases for concessions to conscience may often be more appropriate than formal exemptions, although some antidiscrimination statutes have included government officials in the exemptions they provide.

The Government's Interest in Requiring Performance

A crucial question for any possible exemption is how strong the government's interest is in not affording it. This is both a matter of legislative judgment and interpretation of many statutes that, like RFRA, are generally cast. When we consider what exemptions are wise, restricting those to fairly direct involvement is not only responsive to the power of an objection in conscience, it serves the government interest in avoiding broader discrimination that impairs the availability of services and the dignity of those who suffer negative treatment. For any specific proposed exemption, legislators need to consider the balance of respecting claims of conscience and how our society may benefit from enforcing a general requirement without exceptions.

In regard to RFRA, I shall simply note the sense with which the law was passed and the Supreme Court's treatment in Hobby Lobby. That treatment, insofar as it is clear, obviously covers individual claims, as well as those by organizations. If a “person” makes a RFRA claim and suffers a “substantial burden,” he (or it) will lose only if the government has a “compelling interest” that cannot be served by a “less restrictive means.” That reads like a very strict standard, and when that formulation has been applied by the Supreme Court in racial discrimination and core free speech cases, it has been rigorous. However, when the Court used this language in respect to free exercise claims to depart from ordinary duties, it was much less demanding about what the government had to show. And both it and many lower courts sustained applications for which the government need might well be seen as less than overwhelming.Footnote ³³ Given that RFRA was enacted to reintroduce the earlier constitutional standard, this language should be perceived essentially in the same way. The majority opinion in Hobby Lobby does not rest on the lack of compelling interest, but does assert that other companies providing the insurance is a less restrictive means. Since, as noted, the ultimate cost for the shift will sometimes fall on the government, the circumstances are not so easy to distinguish from the Amish tax case. Insofar as Hobby Lobby does require a somewhat more powerful government showing that an exemption would be inappropriate, it can benefit claims by individuals as well as closely held corporations.