So long as the state defines, confers, and uses marital status as a vehicle for protecting and supporting associations of intimate caregivers, also known as families, it must offer the status to same-sex intimate caregiving couples. A commitment to equal treatment before the law demands it. Period.Footnote ¹

That the state defines, confers, and relies upon marital status in this manner is, however, deeply problematic. This arrangement—the establishment of marriage—conflicts with postmodern, liberal democratic feminist commitments to freedom, equality, and meeting the demands of care (See Cott Reference Cott2000; Metz Reference Metz2010). It distracts and detracts from what really ought to be the focus of public debate and policy: how to meet the demands of care without threatening freedom and equality.

The “demands of care” refers to the fact that no human being is wholly independent. Rather, as theorists of care have long reminded us, human beings are deeply interdependent. We cannot survive, much less thrive, without lifelong care, that is, physical, emotional, and social attention by others (See e.g., Engster Reference Engster2007; Fineman Reference Fineman1995, Reference Fineman2004; Kittay Reference Kittay1999; McClain Reference McClain2006; Minow and Shanley Reference Minow and Shanley1996; Tronto Reference Tronto1993). And giving care, especially the uniquely valuable variety of intimate care, is risky. By intimate care I mean care that is unremunerated, typically exchanged in relationships involving utter or provisional dependency/interdependency over extended periods of time and in a shared domicile. Intimate care is unmonitored and noncontractual, and rarely involves strictly reciprocal exchange of diverse and often incommensurable “goods” (e.g., psychological, social, spiritual, physical, financial, and material goods) (Kittay Reference Kittay1999; Tronto Reference Tronto1993). In the United States, most intimate care, though not all, takes place in families, broadly defined.Footnote ²

Crucially, although intimate care is uniquely consonant with human attachment, particularity, diversity and freedom, it is also rife with risk (Elshtain Reference Elshtain1982; Fineman Reference Fineman1995; Kittay Reference Kittay1999; Tronto Reference Tronto1993). Of particular relevance to the present discussion is the risk to the caregiver. Because this labor is (even ideally) unpaid, unmonitored, and not strictly reciprocal, caring for vulnerable people generates its own physical and material vulnerability, what M. A. Fineman (Reference Fineman1995) calls “derivative dependency.”Footnote ³ In instrumental terms: by using their fixed resources on others without guarantee of timely and relevant return, caregivers reduce their ability to protect themselves against the normal vagaries of existence, thus increasing their vulnerability to those vagaries. Such is the risk of caregiving. Of course, there are numerous tangible and intangible benefits that typically flow from caregiving.Footnote ⁴ To call caregiving risky is not to deny this. It is simply to highlight the fact that all of these potential benefits notwithstanding, when one pours one's limited resources into other human beings, one needs to rely on others to make up the difference. Such is life. And thankfully so, for as philosophers as old as Plato have argued, only gods and monsters exist without care of other human beings.

That relationships of intimate care between unequals (for example, parent and child) involve risk for both the dependent, who is vulnerable to noncare and abuse, and the caregiver, who gives without promise of like return, may seem obvious. But caregiving is also risky among equals. Although return of some sort is typically expected, the unmonitored, unpredictable, and often incommensurable nature of caregiving means that even between able-bodied, able-minded adults, caregiving involves serious material and physical risk. Because the risk involved in intimate care generates systematic vulnerabilities and, therefore, disincentive, any society that wishes this work to be done well and its benefits and burdens distributed justly (within and among families) must offer a degree of insurance against these risks. So, the fact that we all need it and that giving care, itself, generates further need for care, coupled with a commitment to fairness, points to the need for society-wide, coordinated action to ensure against the risks of giving the essential good of care. Roughly, this is what I mean when I refer to the demands of care.

Historically, societies have provided this insurance in various ways. Some inculcate the value of obligation of children to care for aging parents. The risks of giving care are reduced by the socially enforced tradition that children will return the care when parents need it. More familiar to citizens of postindustrial, capitalist democracies are laws that favor parents over unrelated adults or marital partners in key decision-making scenarios (e.g., life and death) and marital property regimes that value nonmonetary contributions. Both legal regimes are based on the assumption that certain individuals (e.g., parents or spouses) have contributed to the survival and flourishing of others (e.g., children or spouses) and on the normative view that these contributions ought to be recognized and valued at crucial junctures. In the first instance, care entails special decision-making privileges. In the second, unpaid labor is valued in material terms. In both cases, the recognition of the contributions of care decreases the chance that the giving of care will only reduce the provider's stores of resources.

As these examples remind us, across time and place, including in twenty-first-century United States, the demands have often been met through marriage. Or, more accurately, most societies rely upon some variety of a formal, social institution that enacts a thick, comprehensive account of the relationships among members of intimate caregiving units—and those units' relationship to the society—to meet a privileged set of these demands. Definitional difficulties aside, these institutions get called “marriage” by most commentators, academic and other.

As I have argued elsewhere, in a diverse, liberal democratic polity, the use of marriage to meet the demands of care is deeply problematic.Footnote ⁵ To start with, in the American context at least, the features that make marriage different from civil union are the very features that led the Framers of the federal and many state constitutions to insist that church and state be separated by nonestablishment clauses. As an institution, and especially in terms of what it “asks” and assumes of the public authority that defines and controls it, marriage is more like a bar mitzvah than like a registered domestic partnership or a business partnership.

What makes marriage different from civil union is that it elicits ethical recognition—the moral approval and the complex normative account of the relationship it names.Footnote ⁶ This is precisely the reason that the battles over marriage are so fierce. Marriage is a lot more than the civil institution created by the state. The “m word” matters because it both calls upon and depends upon vast (but varied) social, extralegal meanings. Think of what many understand to be happening when a couple stands in front of its community—whether a collection of hippies in a northern California redwood grove or a traditional congregation in a Catholic church—and accepts the marital status bestowed by that community. The community's conferral and the couple's acceptance of marital status are understood as acts intended to alter the way each sees and relates to the other. The special value of marriage is more than any civic meaning it conveys. For this reason, Justice J. Sosman of the Massachusetts Supreme Judicial Court can joke about “the ‘m’ word” and replacing marriage with civil union without worrying that she will be taken seriously.Footnote ⁷ Even a registered domestic partnership status that gives all but the marital label to its bearers is not marriage.Footnote ⁸ And we all know it.

Note what this special quality of marriage assumes about the relation between the community that confers the label and the recipients of it. The conferring authority must represent and defend a comprehensive story about the relationships (between the individuals in the couple, and the couple and the community) it labels and represents. The public performance of these stories transforms by enacting, celebrating, and calling for public defense of the shared understandings they express. Think of bar mitzvah or baptismal status: They are conferred with the aim of altering beliefs, to integrate individuals into the norms and beliefs that define the community. By calling on the ethical authority of a community of shared worldviews, these statuses alter self-understandings. If I am right that marriage is fundamentally different from civil union because of its power to transform beliefs and, especially, self-understanding—a popular and plausible claim—then when the state acts as a controlling public authority in defining and limiting access to the status, it assumes the role of ethical authority. The state's ability to do so in turn rests on the existence of a community of comprehensive belief.

We can now see why it is especially appropriate to speak of marriage as “established.” Just as the “establishment of religion” refers to the state's involvement in defining, inculcating, and reproducing a particular religious worldview and institution, so the “establishment of marriage” highlights the state's integral role in reproducing and relying on belief in a particular institutional form of intimate life and its tie to the community. The comparison to religion highlights the second fact not appreciated by pro-establishment advocates, even progressives: When it fills the role of primary controller of marriage, the state fails on two counts (beyond the equality issue discussed previously). It fails liberty. The role of the public authority in marriage is not merely incidental. It is integral to the potential of marriage. As the primary public authority in marriage, the state engages in a role, the basic purpose of which is to alter the self-understandings of those who assume, and those who honor, the label. In so acting, the state intrudes on the realm of belief, and thus it violates one of the most basic liberal mechanisms for protecting liberty. For good reason, the role of the ethical authority is far more familiar to us from the realm of religion than that of politics. By design, the liberal state is more distant, neutral, and uninvolved with the beliefs of its citizens than it would seem to be in the picture just sketched. The familiar idea behind the limited state is that freedom consists, in large part, in individuals being free from interference to live according to their own design. Marriage, however, asks for and depends on much more from its regulating authority than what the liberal state can or should try to give.

In this role, the state also fails marriage and, by extension, all families who depend on its special powers to bolster their existence. The liberal state is ill suited to serve as an ethical authority; so when it functions as the public authority of greatest significance vis-à-vis marriage, the state actually threatens marriage by undermining the force of its constitutive sway. Again, the example of bar mitzvah status is helpful. We would doubt both the effectiveness of the status and the justice of the action were the state to start defining, setting standards for, and regulating access to bar mitzvah status. To effect its psychosocial transformation, the status must come from a community of shared religious belief and from Jewish religious leaders, as representatives of the community of belief that gives the status meaning. Were the state to try to fill this role, it would rightly be seen as inappropriate—a violation of even the least restrictive variety of state neutrality—and, crucially, ineffective.

If marriage requires constitutive recognition from an ethical authority, yet the liberal democratic state cannot (and should not try to) effectively provide such recognition, what is to be done? Must we choose between a healthy marital institution and liberal commitments and institutional arrangements? No. The suggestion implicit in the comparison to religion —that there may be good reason for the state to withdraw from its pivotal role in controlling marital status—is compelling. The state should get out of the business of marriage. The state should withdraw from its current role in defining and controlling marriage; it should not confer marital status or use “marriage” as a category for dispersing benefits or assigning legal obligations. Marital status should be left to voluntary associations, religious, ethical, and cultural entities that wield ethical authority more effectively and justly than does the state.

None of this is to say that the state should withdraw from its role in supporting intimate care. On the contrary: government has a crucial role to play in this arena. As I have argued, we all need care. Moreover, giving care itself generates further need for care, together with a commitment to fairness. The demands of care thus point to the need for society-wide, coordinated action to ensure against the risks of giving the essential good of care. To the extent that the state represents and has the power to affect polity-wide concerns, as is the case in our postindustrial, liberal democratic polity, the demand for societal efforts implicates government. So, in our case, the demands of care require us to consider how the state can and should be involved in recognizing and regulating intimate caregiving associations of all types.

All of this means that care, not marriage, should be the focus of our postmodern, liberal democratic feminist scholarly and political energies. Even more to the point, we need to turn our attention to the question of how the state can and should be involved in protecting and supporting intimate care without threatening freedom or equality. This is no simple matter. Consider some of the challenges.

First, what should we count as intimate care, worthy of social support? The reciprocal, material, social, and psychological give-and-take of two able-bodied, able-minded, sexually intimate adults—with or without children? What about a pair of elderly, same-sex, nonsexually intimate friends? Should they be able to avail themselves of a preset status that provides tax and “kinship” recognition benefits equivalent to, say, those of a unit of parents and children? Should the state offer recognition and support to polygamous units of intimate caregiving? What about a unit containing a nonbiologically related teenage girl and an adult man?

Second, for those caregiving units the state does protect and support, we must consider how their freedom is guarded – but not too much. As any mother on public assistance can tell us, certain types of state support come with severe limits on choices. And on the flip side, as feminists have long noted, too much privacy has a way of protecting little tyrants within families.

Third, there are very serious questions about how to support citizens who do not find themselves in intimate caregiving relationships—by choice or not. Since we all need care, these might be the most vulnerable among us. The argument that their freedom is violated when they are taxed to support the caregiving and care-receiving of others deserves serious attention, if only because it is so widespread.Footnote ⁹

Fourth, one of the great benefits of removing the public discussion about intimate care from behind the veil of marriage is that the real costs, benefits, and distribution of intimate care will be exposed. In this light—and of crucial significance—the means by which men and women of the middle and upper classes and the global North, in general, manage to participate in the well-paid labor force and have families will become clearer, and it will become an unavoidable fact and part of public discussion. The costs and injustices of the second shift and the global care chains will, and should, take center stage in the discussions about family policy. These are just a few of the crucial questions that deserve attention.

So, yes, we should support gay marriage, and certainly must oppose anti–same–sex marriage initiatives in the meantime. But we should not do so uncritically or solely. We cannot let this issue distract us from the demands of care and the challenges they present in a diverse polity such as ours.