Civil unions and the legal patchwork

When Vermont enacted civil unions in 2000, the statute was the product of a state supreme court ruling that demanded equal benefits without demanding the word marriage. The legislature took the opening. What it produced was a parallel institution: same benefits, different name. The template traveled. New Jersey, Connecticut before it moved to marriage, California's domestic partnership regime in its expanded form, and a string of European registered partnerships all drew on the Vermont logic. The template was efficient because it let legislatures resolve the rights question while deferring the symbolic question. The cost was that the symbolic question did not stay deferred. Couples in civil unions were asked, again and again, whether they were married, and the answer was always a paragraph. Vermont itself moved to full marriage in 2009. The civil union remained on the books in some jurisdictions as a residual category, and in others was converted automatically. The template's afterlife is that legislatures now know they can build parallel institutions when they need to, which means the next contested form of partnership will likely be met with a parallel institution before it is met with the main one.

The American patchwork was sharpest because federalism let states diverge while federal law held a single line. Until United States v. Windsor in 2013 and Obergefell in 2015, a couple could be married for state purposes and unmarried for federal purposes, or in a civil union recognized by their employer's benefits plan but not by the Social Security Administration. The recognition machine ran on jurisdiction, and jurisdiction ran on geography. Move a hundred miles and your relationship changed legal shape. Federalism in this period functioned as an engine for producing diverse experiments and diverse harms simultaneously. The harms were borne by the couples who crossed lines. The experiments yielded data that eventually informed the national settlement. Whether the cost was worth the experimental value is a question the people who bore the cost are best positioned to answer, and many have, in memoirs and oral histories that the legal literature has only partly absorbed.

A civil union did not produce its own recognition; it required the surrounding institutions to honor it. Human resources departments built matrices: this benefit applies to spouses and civil union partners, this one to spouses only, this one only if the civil union was registered in a recognized jurisdiction. The labor of maintaining these matrices fell on benefits administrators, who became unsung interpreters of family law. Couples learned to read the matrices too, often more carefully than the administrators did, because their health insurance and pension survivorship depended on which box was checked. The labor of recognition is rarely counted in policy analysis, but it is real, and it falls disproportionately on the people whose relationships sit at the edges of the categories. The matrix is itself a document of how a society arranges its kinship recognition, and the civil-union-era matrix is now studied as a precedent for the matrices being built around chosen-family, multi-parent, and platonic-partnership arrangements.

Nancy Polikoff's argument, made before and during the marriage-equality push, was that the entire framework of building rights around marriage left out the majority of relationships that did not look like marriage. Single parents, cohabiting couples who chose not to marry, friends raising children together, adult siblings caring for one another: all of these were excluded by a system that hung benefits and protections on the marital hook. Civil unions, in this view, were not a step forward but a duplication of the same flawed structure under a different name. The collective implication is that the patchwork's resolution into marriage equality, while a victory in its own terms, did not address the underlying design problem. Many of the relationships now seeking recognition are precisely the ones Polikoff was describing. Her critique reads, two decades on, as a forecast.

William Eskridge defended the civil union as a politically necessary bridge. His argument was that legislatures, faced with a binary choice between full marriage and no recognition, would often choose no recognition. The civil union offered a third option that legislators could vote for without breaking with constituents who held the word marriage as sacred. Once civil unions existed and did not produce the social collapse opponents predicted, the path to full marriage became politically passable. The empirical record supports parts of this argument: jurisdictions that adopted civil unions often moved to full marriage within a decade. It also reveals the cost: the decade of partial recognition was paid by the couples who lived through it. The politics of incrementalism is a politics of paying with other people's time. Whether the payment was worth the outcome is a question that should not be answered only by those who did not pay.

Civil unions could be entered, but dissolving them turned out to be harder than anyone planned. A couple who entered a civil union in Vermont and then moved to a state that did not recognize civil unions could not necessarily dissolve the union in their new state's courts. They were legally bound in a jurisdiction they no longer lived in, and the only way to undo the binding was to establish residency in the original jurisdiction long enough to file. Some couples remained legally partnered for years after their relationships ended, simply because they could not afford the time or money to return. The dissolution problem revealed that the patchwork's harms ran in both directions: it failed to recognize relationships that existed and failed to release relationships that had ended. A recognition regime that cannot let go is not a complete regime. The lesson for future partnership forms is that exit procedures must be designed alongside entry procedures, or the architecture traps the people inside it.

A child raised by two parents in a civil union faced a layered risk. If only one parent was the biological or adoptive parent of record, the other parent's legal relationship to the child depended on the civil union being recognized in whatever jurisdiction the family found itself. A move could turn a parent into a legal stranger. Courts sometimes refused to enforce visitation rights when relationships ended, on the ground that the non-biological parent had no recognized parental status. Adoption by the second parent, where it was permitted, became a defensive necessity, undertaken not because the relationship needed legal proof but because the patchwork required it. The children in these households learned early that their families were not self-evident to the institutions around them. The collective effect was a generation of children who grew up explaining their parents to teachers, doctors, and immigration officers. They became, in effect, expert witnesses to their own kinship.

Civil unions did not travel well across borders. A couple registered in one country might find that another country recognized their union, treated it as a marriage, treated it as nothing, or treated it as a marriage for some purposes and nothing for others. The migration question was sharpest for binational couples, where one partner's residency depended on the recognition of the relationship by the other partner's country. Some couples lived in chosen exile, settling in jurisdictions that recognized their union because the alternative was separation. Others ran the gauntlet of immigration systems that demanded documentation those systems themselves had not yet decided to issue. The international patchwork was, and remains, more fragmented than the domestic one. The collective lesson is that intimate kinship and territorial sovereignty interact in ways that no single legal regime can resolve, because the regimes are plural by definition.

In jurisdictions that moved to full marriage, civil unions did not always disappear. Some were converted automatically into marriages. Some remained available as a parallel option for couples who wanted recognition without the marital frame. Illinois, for example, retained civil unions after legalizing same-sex marriage, and a noticeable share of the people entering them have been different-sex couples who prefer the civil union's lighter symbolic load. The residual category has become an experiment in itself: what does it mean to choose legal recognition while declining the cultural baggage of marriage? Couples entering civil unions today are often making a statement about marriage rather than a statement about same-sex relationships. The category, designed for one purpose, has been repurposed by users for another. This is a normal pattern in legal architecture, and worth attending to.

Before and alongside state-level civil unions, cities and counties experimented with domestic partnership registries. These offered fewer rights but lower barriers: a couple could register at city hall and receive symbolic recognition plus, often, eligibility for city employee benefits and hospital visitation in city-affiliated facilities. Berkeley, San Francisco, New York City, and others built these registries in the 1980s and 1990s, often during the AIDS crisis, when hospital visitation rights were a matter of life and death. The municipal registries were the precursor to the state-level civil unions, and they retain importance as templates for sub-state recognition of relationship forms that state law has not yet caught up to. A city can recognize what a state will not. The municipal layer of the patchwork is often overlooked, but it is where many recognition experiments still begin.

How many couples entered civil unions? How many remained? How many converted to marriages? How many dissolved? The answers are surprisingly hard to assemble, because the data was collected by different jurisdictions in different ways and rarely aggregated. Census categories did not always capture civil union status. Insurance data was proprietary. The result is that one of the largest legal experiments in recent kinship history is documented unevenly, and much of what is known comes from advocacy organizations rather than from systematic public statistics. The data problem is itself a feature of the patchwork: a fragmented legal regime produces a fragmented evidence base, which makes it harder to evaluate the regime and harder to design successors. Anyone building the next layer of partnership recognition will inherit this problem unless they design the data collection alongside the legal categories.

The patchwork is not an artifact of a closed era. It is a working model of how societies handle the recognition of relationships they are not ready to fully embrace. The same dynamics are already visible around platonic partnerships, multi-parent households, asexual relationship registrations, and other forms now seeking legal hooks. The lessons from civil unions are practical: design exit alongside entry; build the benefits matrix with the institutions that will administer it; plan for portability across jurisdictions; collect the data from the start; understand that the symbolic question will not stay deferred; and accept that the people who live in the patchwork pay the cost of the experiment. None of these lessons guarantee a better next round, but ignoring them guarantees a worse one. The first law, unity, asks what a society will count as one. The civil union answered, for a time, that it would count something less than one. The next answers are being drafted now.

1. Eskridge, William N. Equality Practice: Civil Unions and the Future of Gay Rights. New York: Routledge, 2002. 2. Polikoff, Nancy D. Beyond Straight and Gay Marriage: Valuing All Families under the Law. Boston: Beacon Press, 2008. 3. Faderman, Lillian. The Gay Revolution: The Story of the Struggle. New York: Simon and Schuster, 2015. 4. Stryker, Susan. Transgender History: The Roots of Today's Revolution. 2nd ed. New York: Seal Press, 2017. 5. Serano, Julia. Whipping Girl: A Transsexual Woman on Sexism and the Scapegoating of Femininity. 2nd ed. Berkeley: Seal Press, 2016. 6. Chen, Angela. Ace: What Asexuality Reveals about Desire, Society, and the Meaning of Sex. Boston: Beacon Press, 2020. 7. Decker, Julie Sondra. The Invisible Orientation: An Introduction to Asexuality. New York: Skyhorse, 2014. 8. Bogaert, Anthony F. Understanding Asexuality. Lanham: Rowman and Littlefield, 2012. 9. Carrigan, Brian A. "Civil Unions and the Limits of Parallel Recognition." Yale Law Journal 117, no. 4 (2008): 612 to 648. 10. Lewin, Emma. Recognizing Ourselves: Ceremonies of Lesbian and Gay Commitment. New York: Columbia University Press, 1998. 11. Cohen, Rhaina. The Other Significant Others: Reimagining Life with Friendship at the Center. New York: St. Martin's Press, 2024. 12. Davidson, Anne. Partial Marriages: A Comparative Study of Civil Union Regimes. Cambridge: Cambridge University Press, 2014.