Consider two cases commonly said to the illustrate the conflicts past and future. (1) In New Mexico, the state human-rights commission ruled that a husband-wide photographer team violated state law barring discrimination on the basis of sexual orientation for refusing to photograph a same-sex commitment ceremony. The case may illustrate the overreach of some state antidiscrimination laws, though even this is unclear since the commission did not even consider the religious-exemption claim, the ruling is on appeal, and there's a good argument it violates the First Amendment and the state's own RFRA. But it does not exemplify the threat of gay marriage, since neither SSM nor civil unions are recognized in the state. (2) In New Jersey, a church refused to rent part of its beach-front property to a lesbian couple for their commitment ceremony. The basis for the action was not any marriage or civil union entered by the lesbians, but the church's own agreement to make the property available to the public in exchange for special tax treatment. I have discussed these and other cases here.

The absence of conflicts is suggestive, but not decisive, on the need for special religious protections in SSM bills. Perhaps it's too early and such cases are like a gathering storm. There are fifty states, with fifty sets of laws related to marriage, civil rights, and religious liberty. But if Doug Laycock is right, and I think he is, we should have expected conflicts to peak in a state in the immediate aftermath of SSM or civil-union recognition when emotions are highest and opposition is boldest, with a decline thereafter.

The really interesting question is why there have been so few conflicts. The main reason, I suspect, is common sense and forbearance on the part of both gay couples and those who object on religious grounds to gay marriage. Unless they have no other choices, few gay couples want to pay for marital goods or services from people who don't want to provide them. Few service providers object to gay marriage on religious grounds and, as Laycock suggests, fewer still believe their faith requires them to refuse goods or services (or housing) to gay couples. Plus, they want the business.

Another reason we've had few conflicts is that this unusually religious and pluralistic country already respects and protects religious beliefs and practices to an extent unseen anywhere else in the world. There are the federal and state constitutions protecting religious freedom. Just about every antidiscrimination law protecting gays has been the result of legislative compromise in which the scope of the law was limited, or exemptions were added, to minimize conflicts with religious objectors in the most likely contexts (like religious groups, and small businesses and landlords). Additionally, half of the states require by statute or judicial decision a compelling state interest, enforced by means narrowly drawn, for any state policy that burdens religion.

Obviously, religious individuals, businesses, and organizations sometimes lose religious-freedom claims. But they also win a lot of the time, most recently and prominently when the California Supreme Court left undisturbed a lower court decision allowing a religious school to exclude two students having a suspected lesbian relationship. The religious school was not a business, said the state courts, and thus not even subject to the state's civil-rights law.

That's the experiential backdrop for the legal arguments.

II. SSM and religious liberty: some specific legal arguments

In their latest response, Berg et al. suggest a couple of specific ways in which SSM increases the legal risk to religious liberty beyond the risk they would face under existing antidiscrimination laws alone.

The first specific concern is that SSM will weaken a possible defense for religious objectors: that when they discriminate against gay couples in providing goods or services they are not really discriminating against homosexual orientation but against all "extramarital conduct." The professors cite Christian Legal Society v. Walker, 453 F.3d 853 (7th Cir. 2006), for the fear that marriage-based distinctions will collapse for religious objectors in an SSM world. In that case, the Court granted a preliminary injunction against a law school that de-recognized a student group excluding from membership all those who have sex outside of traditional man-woman marriage (e.g., adultery, fornication, homosexual acts).

The panel held that this conduct-based exclusion likely didn't even violate the school's policy against sexual orientation discrimination. What difference would it have made if Illinois recognized gay marriage? The CLS's discrimination would still have been directed at certain sexual conduct (that outside of traditional marriage) rather than to sexual orientation, which is all the policy (and state antidiscrimination law) prohibits. Now maybe a court would hold that such a conduct-based exclusion is really tantamount to an orientation-based exclusion. Or it might hold that the religious objector does not consistently oppose all extramarital conduct and is using that as a pretext for discrimination against gays. But neither conclusion would turn on whether the state recognized SSM, much less on whether the excluded student was in a same-sex marriage.

Equally important here, the court held that the group was likely protected by constitutional expressive-association and free-speech principles. These are complicated claims. But whatever we think of them, what difference would it have made to the constitutional claims if Illinois recognized gay marriages? I see nothing in the opinion, or in the constitutional precedents upon which it relies, that suggests it would have made a difference.

The second specific concern is that, in the context of a state RFRA defense, SSM will strengthen the government's hand in claiming that it has a "compelling interest" to eliminate sexual-orientation discrimination sufficient to override a religious-liberty defense. Take the New Mexico photographer case. In the absence of SSM, the religious objector might argue that the state can hardly have a compelling interest in forcing her to take pictures of a gay commitment ceremony if the state does not itself recognize gay marriages. The state is being inconsistent, one might argue, and this very inconsistency undermines its compelling-interest claim. If the state recognizes SSM, it will have consistently advanced its antidiscrimination interest.

Maybe that's true, but I doubt SSM will make a difference. In strict scrutiny, courts generally defer to the state on what counts as a compelling interest and focus on whether the means used to achieve it are narrowly tailored. The state would have a good argument that it is has a compelling interest in eradicating sexual-orientation discrimination in whatever context that discrimination arises. I am unaware of state or federal cases holding that there is no compelling state interest in fighting anti-gay discrimination (the Boy Scouts v. Dale case does not so hold, in my view). I think the state can plausibly say it has a compelling interest in ending all private discrimination without recognizing SSM. There's a big difference between forbidding a single act of discrimination, a discrete matter implicating a narrow but compelling interest in eliminating private discrimination, and recognizing gay marriage, a comprehensive matter in which the state has all kinds of other competing interests to weigh. Recognizing gay marriage would impose many costs and obligations on the state that it does not undertake when it enforces its antidiscrimination law. Why does it have to do everything to maintain a compelling interest in doing one thing? A similar analysis would apply to the state's interest in the Catholic Charities case.

Whether the state has advanced its compelling interest through narrowly tailored means should be the focus. Since many other photographers (and adoption agencies) are available to gay couples, the burden on particular religious objectors does not seem necessary to achieve the state's interest. But the means analysis does not hinge on whether the gay couple can marry under state law.

Let me be clear: I think both the New Mexico photographer and Catholic Charities should be exempt from compliance with state antidiscrimination law in the contexts in which these cases arose. The first result can and should be accomplished through existing limits on antidiscrimination law and religious-freedom principles. The second may require an exemption specifically crafted for religious organizations providing special-needs adoption services. But neither case has anything to do, as I see it, with gay marriage.

I'm hesitant to say that there will never be a case in which SSM will have made the difference between winning and losing for a religious-liberty claimant. Never is a long time, and never say never when it comes to courts. But at this point I do not think concerns about the direct legal effects of SSM should play much, if any, role in the debate over SSM.

The most that can be said is that SSM will help reinforce an idea that is already well underway in the law and in our culture: homosexuality is a benign variation of human sexuality. Every legal change in the status and protection of homosexuals over that past 50 years, from the elimination of sodomy laws to the creation of gay student groups, has both reflected and reinforced this trend. Each could have been, and was, opposed on the ground that it would promote a world view religious traditionalists oppose. Each made it marginally harder for religious traditionalists to teach their children that homosexuality is wrong, since the law no longer fully backed that teaching. Each increased by some degree the possibility that the cultural/legal environment would become more hostile for religious traditionalists. I don't doubt that marriage is another step in this direction, more or less significant than some of the others in terms of its impact on a traditionalist world view. (Probably less significant than legalizing homosexual sodomy but more significant than, say, allowing homosexuals to have security clearances.)

Even the process of sensitizing us to discrimination against a group like gays, however, occurs against the backdrop of a very deep and still popular commitment to protecting religious freedom. It's no accident that Congress overwhelmingly passed RFRA or that many states have passed their own versions. After almost 150 years of expanding civil-rights laws in scope and breadth, of hate crimes laws, of equality for women and blacks, of major changes in marriage, and of liberalization of attitudes toward gays, we're still the most religiously tolerant, diverse, and observant Western country.

III. How should religious freedom be protected in SSM bills?

As I read their draft exemption and their reply to my questions about it, the professors' proposal leaves in place whatever antidiscrimination requirements are already embodied in state law. It simply negates any antidiscrimination obligation they might have arising from the provision of marriage-related services, etc. It does not eliminate their obligations under general state antidiscrimination law that arise independent of any marriage. If that's right, none of the cases they have cited, and none that I have reviewed, would come out differently under their proposal. The New Mexico case, the New Jersey pavilion, and even the Catholic Charities case would have been resolved for (or against) the religious objectors under existing doctrine since none of them turn on the marital status of the gay couples involved. I think any potential overreach of antidiscrimination law can best be addressed in the context of those laws by, e.g., providing a religious exemption for small businesses and landlords, religious nonprofit businesses, and personal and non-essential service providers (like wedding photographers, florists, and others).

The substantive case for crafting special religious protection in state legislation authorizing SSM is not very strong, for reasons I've given above. The political case for adding special religious protection seems much stronger, since doing so may allow legislators to alleviate reasonable fears and reduce the opportunity for demagoguery against gay marriage -- all while protecting gay families in the law. While I don't think the special protections drafted by the professors are necessary to secure religious liberty, I'm also at a loss to see how they hurt much, as long as two things are kept in mind.

First, any special religious exemption should provide (as the professors suggest) a "hardship exception" for gay couples who cannot readily and easily find the goods (like flowers) or services (like a caterer) elsewhere. The draft of the proposal sent to the Connecticut legislature two weeks ago provides no such hardship exception. It's easy to imagine that a gay couple in an urban area can find an adoption service or a photographer. It's harder to imagine that in the middle of Montana. It's also hard to justify allowing, say, a Catholic hospital to bar a person from his dying spouse's room on the grounds that the hospital does not recognize their marriage as valid. All that may be needed to deal with these hardship cases is an explicit provision in the marriage context applying the compelling-interest test common in state statute and decisional law. I'd be happy to work with the professors on drafting something like that to deal with such cases.

Second, while I would like to see greater protection for individuals whose religious beliefs are burdened by antidiscrimination law, I am wary of introducing that idea into the provision of services to the public by employees of the state. Because of their position as representatives of the state, because they administer laws that benefit the public, and because they are paid by public tax money, these employees are treated differently than we treat private employees or members of the public themselves (see, e.g., free-speech rights). The only cases in which the existence of SSM or civil unions actually have imposed obligations on religious dissenters are ones in which marriage-license bureau clerks have refused to do their jobs. Even these have been few. If we could be sure that an exemption covering state employees would be limited to purely ministerial acts (like stamping a marriage application) for which there is another person readily available (not 100 miles away at the next courthouse), there would be no practical harm in it. But as I read the professors' draft exemption, there is no qualitative limit on the claims that could be made by state employees, including judges who refuse to preside over SSM divorces. The state administers so many benefits under marriage law that an exemption for government employees from doing their jobs has potentially unlimited applications.