We recently got married. Well, technically, we got married twice.
One fine day this spring, we put on nice clothes and publicly performed the rites and rituals recognized by our families and community as a wedding ceremony. As part of the day's events, we signed a Ketubah, the traditional Jewish wedding contract. Historically, the Ketubah included the groom's promise to provide "food, clothing, the necessities of life, and conjugal needs" for the bride, along with a statement of the dowry the bride brought to the marriage. Modern versions are often more egalitarian. Ours included a mutual promise to "work for one another," "live with one another," and "build together a household of integrity." Ketubot are typically beautifully calligraphed works of art, and we spent a lot of time choosing the right text and design for ours. It was witnessed by our rabbi and by two beloved friends. It hangs in our bedroom as a reminder of the commitment we have made to each other.
We also got married in the eyes of the law. Our state marriage license was printed at the city-county building on cheap paper after the clerk checked our IDs, filled our names into the anonymous blanks in the same text every other couple has to use, and gave us a pamphlet about syphilis. We had no say in the wording or the witnesses. We keep that license in the safe deposit box at the bank with our mortgage and the titles to our cars.
The contrast in the thinking behind our two marriage documents, and in how we have treated them now that we have them, captures the difference between thinking of marriage as a mutual contract and thinking of it as a license from the state. It's the difference between a relationship that requires consent and one that requires permission.
If you hang around with libertarians long enough, you'll almost certainly hear someone ask, "Why can't we just get the state out of the marriage business entirely?" Until two years ago, when Obergefell v. Hodges settled the question, you'd occasionally hear a certain stripe of libertarianish conservative call for privatizing marriage too, sometimes on principle and sometimes as a dodge around the question of whether the federal government should recognize gay unions. Sen. Rand Paul (R–Ky.), for instance, has long said, "I don't want my guns or my marriage registered in Washington." The Alabama state legislature has considered proposals that would more or less end the licensing of marriages in the state, presumably not because of a deep commitment to limited government.
As libertarians, we would prefer to deal with the government as little as we can, yet we still chose to involve the state in our marriage. The reasons we did so can shed light on the challenges involved in extracting the state from this institution, and also on why such a change might be worthwhile.
Cupid by Contract What does it mean in practice to say we want to get the state out of marriage?
One problem is that state marital provisions are one-size-fits-all, as with our fill-in-the-names-and-sign-here marriage license. Actual 21st century marriages are much more idiosyncratic—the wide range of pre-nuptial agreements demonstrates this, as does our personalized Ketubah. Many people might want the flexibility in marital arrangements that privatization allows. Writing in Slate in 1997, the Cato Institute's David Boaz imagined a kind of standard contract, much like a standardized will, that would work for many as-is but would also allow for more detailed arrangements. One can even imagine marriage contracts that are renewable at 5- or 10-year intervals, allowing couples to part ways amicably without many of the financial and emotional costs of divorce.
Economic arrangements could be varied along a number of dimensions, including considerations for children, agreements about how money will be spent, and worst-case-scenario planning for illness, death, or divorce. People could choose to have a religious wedding, with a contract designed by the institution in question, along with specifications for divorce and the rest. All of these kinds of contracts would be enforced through common-law mechanisms, with judges interpreting the texts and building up a body of legal precedent about how to resolve disputes. We could also imagine marriage certification services (Marital Underwriters Lab? BridalZoom.com?) who check over simple contracts for those who don't want to use a lawyer or a church.
Full privatization also implies that marital status must be irrelevant to the provision of government benefits. Otherwise, the state would still have an interest in how marriage is defined. Truly private marriage would allow for whatever variety of arrangements people desire.
Privatization of this sort might be especially attractive to those whose relationships currently do not have legal status. Plural marriages are the obvious example. Religious institutions might also support privatization, as it would enable them to have their own set of marital rules without fear that the state's requirements would force them to act against their beliefs. Tailored contracts, especially if such contracts became relatively standardized for particular kinds of partnerships, might be appealing to the very wealthy or to those with child custody complications. If such contracts included provisions for private arbitration (or even marital counseling), this might reduce some of the costs of divorce, making them attractive to all kinds of married people.
Marriage Without the State Historically, human beings have found a variety of ways to define and regularize marriage that don't require a license from the state. Such systems had the usual advantages of stateless orders: They offered more variety, they were more flexible, and they were more responsive to local needs. The problem was that these forms of marriage worked at a time when others in the community played a much larger role in determining not just what counted as a marriage but who could get married. Could the regularity and social enforcement required for functioning informal institutions exist in a 21st century context, when marriage is considered a much more private institution? It's an open question.
For most of human history, for most people, marriage was a way to ensure that they and the communities they inhabited were able to survive economically. When most humans lived at the margins, ensuring that there was sufficient labor to work the fields or run the family business was essential. Marriage provided an economic partnership for organizing production, and the ability to raise children together meant that those children could contribute to that production process.
Because of the community's interest in ensuring good marital matches, what counted as being married was decided in decentralized ways by particular communities, especially among the poor. Marriage was much more a matter of custom than of formal rules. Permissions came not from a magistrate, but from neighbors and kin. Arranged marriages can be seen as an extremely strong form of permission, where parents or others possess an exclusive right to contract a marriage for their heirs, even when it overrides the desires of young people. More commonly, that right to grant permission was used to veto a marriage rather than to force one.
But in the West at least, marriage has been a matter of consent between partners for a very long time. As far back as the Magna Carta we find specific language protecting widows—a particularly vulnerable category—from forced marriages. But consensual marriage still required permission from parents, relatives, and community. What's more, consent is not the same thing as love. Consensual marriages based on economic and political concerns were the standard up until the last 200 years or so.
Until then, customary practices predominated, especially in remote areas where the reach of the state was limited. The practice of "jumping the broomstick" as a public indication of intent to marry was common among slaves in the American south, but also had a long history in other communities as well. Some communities recognized jumping backward over a broomstick as an intent to divorce. Medieval Ireland's rules for marriage were part of a legal system that operated outside of a formal state. There were 10 different forms of Irish marriage, most of which depended on what each partner brought to the union in terms of property. Property mattered, as it did elsewhere, because it defined the rights the person had with respect to divorce as well as the rules governing inheritance. There were also a series of fines for a variety of classes of illegal marriages, including marriage by stealth, abduction, or rape. Scotland's famously relaxed marriage laws allowed for a much lower age of consent than in England, and permitted nearly any adult to perform a marriage between two consenting individuals. Dramatic scenes of elopements to Gretna Green thus became a feature of the English novel and of English life.
There is no specific moment when marriage became defined by the state rather than private institutions. In part, this is because church and state were so intertwined for so long.
Going to the Chapel Today we think of churches as being part of civil society, and those who propose the privatization of marriage are generally insistent that houses of worship should be able to determine their own rules for the marriages they will chose to sanctify. If houses of worship do not wish to marry same-sex or interfaith couples, that's up to them.
But through much of history, the line between church and state that allows for such nuance was nearly nonexistent—in part because the church frequently acted like a state. The church's rules covering marriage enjoyed the force of law, so much of what the church did was equivalent to state involvement. And the church's numerous rules about who could marry, divorce, remarry, and adopt often were designed to work to the church's material benefit.
For example, for a long time one could not marry a relative closer than a seventh cousin. There is no biological justification for such a rule, and you can imagine how difficult it was to provide proof that a couple wasn't breaking it. So the church sold indulgences to waive the rule, which was a convenient source of revenue. Similarly, the church's centurieslong prohibition on adoption was a way to ensure that childless couples did not have heirs and would be more likely to will their property to the church. Rules against remarriage for widows also made it more likely that property would go to the church while creating a class of women who could become nuns. This is likewise why the church prohibited the Jewish practice of levirate marriage, where a widow would be expected to marry her dead husband's brother to keep her property in his family.
It's tempting to simply propose that governments accept as valid any marriages performed by other private institutions, but this only shifts the battle one level higher. Which institutions would count? Are evangelicals going to stand for plural marriages in the Church of Satan? Will progressives accept arranged marriages between much older men and very young girls?
In the 12th century, the Catholic Church attempted to require that marriages be solemnized in a church, but that was unenforceable in a world where tradition saw, in the words of historian Stephanie Coontz, "mutual intent or the blessing of a parent sufficient" for that purpose. What constituted mutual intent varied across time and communities, early on requiring mutual promises followed by sexual intercourse, but later requiring only the exchange of vows. By 1215, the Church required three things for a marriage to be valid: (1) a bride with a dowry; (2) announcements beforehand of the intent to marry; and (3) marriage in a church. Even here, however, the state played no official role, and other religious groups had their own traditions for what constituted a valid marriage.
The Protestant Reformation meant that states were increasingly closely identified with the particular sacramental practices that were legal within their borders, so the distance between church and state narrowed. No longer competing institutions, the church and state were often nearly synonymous. Enlightenment thinkers' insistence on creating separation between church and state pulled the two apart again, but the more secular world of the Enlightenment meant that the balance of power had shifted. Modernity meant that, in the West, the state took primacy over the church in many matters that had previously been primarily theological—including marriage.
Marriage, American-Style In the United States, family law, including marriage law, has long been the purview of individual states. But in 18th and early 19th century America, those laws were difficult to enforce in the face of communities where established customs defined marital status and where clergy were often rare visitors.
Historian Nancy Cott reports that despite marriage laws, "informal marriage was common and validated among white settlers from the colonial period on." Couples who met community standards more or less married themselves, demonstrating the continued importance of consent rather than the law. Cott adds that "cohabitation and reciprocal economic contributions" also mattered for indicating that a couple was married, but that "consent was the first essential." Before the state's reach became sufficiently great, communities were much more concerned about whether pairs were functioning as married couples were expected to function than whether they had followed the formal rules of the state or even the church.
Even in these situations where the state was mostly absent, marriage was not a matter of "anything goes." The standards for determining a valid marriage were self-policed by the communities in question. Where law and customary practice conflicted, even to questions of tolerating divorce or sex outside of marriage, customary practice that permitted such things under the right conditions often won the day. Some of the communal norms about defining marital status were strong enough that they became codified by judges in case law. "Informal marriage was valid unless it was specifically prohibited," writes Cott. In cases where judicial intervention was needed to determine if a couple was married, judges generally deferred to community norms about what counted as married, operating on the assumption that couples who were cohabitating and otherwise acting like a married couple were properly considered one, absent clear evidence to the contrary.
It's tempting to think that a 21st century version of privatized marriage could simply recreate this world. However, the increased heterogeneity of human beings and their romantic relationships has made a single, common norm of marriage unsustainable, requiring that society move from informal, communal norms to some more explicit and formalized contractual relationship.
To this day, states and comparable jurisdictions have wide latitude in defining the rules for marriage and divorce. Until 2008, for example, the District of Columbia required a blood test for syphilis in order to get a marriage license, while neighboring Virginia and Maryland did not. Alimony and child support rules differ widely from state to state as well. In a fashion not unlike those Gretna Green elopements, it was common for couples to cross state lines to marry or divorce in states whose laws were more amenable to their individual circumstances. Several dozen movies between 1910 and 1947 focused on the infamous ease and convenience of the divorce laws in Reno, Nevada. And one of the reasons the economist F.A. Hayek took a job in Arkansas when he first moved to the U.S. in the 1950s was to take advantage of the state's liberal divorce rules.
Marrying Like a State We got married for a complex but fairly ordinary list of reasons.
We wanted to spend more time enjoying life with the person we loved best. Steve wanted someone to edit his prose. Sarah wanted someone to organize her chaos. We wanted to raise our two sets of kids together, to work together, to be there for the other in case of emergencies, and to ensure that we had a legally acknowledged relationship for financial and medical reasons. We also wanted to model a committed and loving relationship for our children by formalizing our promises to one another. We wanted to combine our households and our lives for fun, for pleasure, for efficiency, and for the sake of our budgets.
Given all these self-interested reasons that we, and other people, get married, why should the state determine who counts as wed? Even if we agree that marriage is good for the people getting married and that maintaining the institution is good for society because it contributes significantly to better child raising, why is it licensed and controlled by external authorities?
For an answer, look to the self-interest of state actors. The power to define the terms of marriage is the power to raise revenue, incentivize behavior that benefits the organization, and determine who is eligible to receive the benefits the organization provides. In Seeing Like a State, his study of the state's power to organize and manipulate its citizens, James Scott argues that governments and similar institutions need to impose categories and rigid, artificial organizational schemes on people in order to accomplish the institutions' various goals. Once the state imposes those categories, they become part of how we think about what those categories are organizing.
The state's increased role in defining what counts as marriage, even as the state has reduced restrictions on who one can marry, has happened in parallel with the growth of state involvement in many other aspects of people's lives. The state's interest in defining and approving of marriages is entangled with the role marital status plays in a host of government programs. When privatizers call to remove the state from marriage, it sounds as if there's only one plug to be pulled; in fact, there are thousands. As long as those programs exist, and as long as they depend to some degree on a clear definition of marital status, the state is unlikely to get out of the business of defining marriage.
This becomes the dilemma. The battles over marriage, including future debates over plural marriage, can indeed be defused if they are de-politicized. But as long as marriage matters for so much else, it cannot be de-politicized.
It's tempting to simply propose that governments accept as valid any marriages performed by other private institutions, but this only shifts the battle one level higher. Which institutions would count? Are evangelicals going to stand for plural marriages or weddings in the Church of Satan? Will progressives accept arranged marriages between much older men and very young girls? When governments need to know marital status, marriage cannot be de-politicized.
The American government's burgeoning role in marriage has always been driven by a succession of social issues where control over defining marriage was a trump card. In the 19th century, the most obvious examples were slavery and race. Roughly half of the original 13 colonies prohibited interracial marriage. By the end of the Civil War, the vast majority of states did so. Far more states criminalized interracial marriage than interracial sex, and interracial marriages were one clear exception to the law's deference to community-based informal norms. Even if interracial couples met all of the local expectations, they did not get the exemption from the law that white couples usually did. Maintaining this racial inequality even as social norms worked to change it required the full force of the law. Over the course of the 20th century, some states rescinded their laws and many others enforced them only selectively until the Loving v. Virginia decision overturned them all in 1967.
In the 20th century, as always, the definition of marriage was closely tied to economics, but with the new twist that the state was providing many more incomes. "Marriage bar" policies of both governments and private employers early in the century prohibited hiring married women, making marital status a way to control women's behavior and enforce a particular vision of married life. A number of New Deal programs (including Social Security) used marital status as a way of determining benefits, so individual states had to clarify their marriage procedures and ensure that they were followed. A federal policy during the Depression that no family could hold more than one government job made marital status more of a concern to Washington.
The history of the U.S. tax code has been deeply entangled with marital status. In his history of the tax code's effects on women, Taxing Women, tax law scholar Edward McCaffery explains how various changes in rates and filing status have reflected the attempts by politicians and others to privilege the "traditional" male-headed single-earner household. Joint filing under the current tax code creates a "secondary earner bias" that causes the spouse whose labor force participation is more marginal to have his or her first dollar taxed at the same rate as the last dollar earned by the primary earner.
As women are far more likely to be the secondary earner, this feature continues to discourage married women from working even as it encourages marriage among people with higher earning capacity. For high-earning couples with relatively equal incomes, getting married means a higher tax bill than staying single, while high-earning couples with disparate incomes see lower taxes if they marry. These features create incentives, at least on the margin, for a particular type of marriage.
On the spending side, marriage matters for eligibility and benefit levels for a large number of government programs. In the mid-1990s, the General Accounting Office reported 1,049 federal government laws that recognize marital status. To the extent that marriage reduces benefit levels and increases the tax burden for poorer couples, the combination of the two discourages marriage among the poor.
The structure of the Earned Income Tax Credit (EITC) means that for many poor couples, marriage means a reduction in benefits. Unmarried parents can both take advantage of the family EITC, and the one without the resident child can also get the childless worker EITC. Married parents, meanwhile, are not eligible for all three credits. The power of the state to define who is and is not married has implications not just for the financial situation of potential partners, but for the viability of the institution of marriage and the corresponding social benefits it brings.
The fight over same-sex marriage illustrates many of the same themes as the shift of heterosexual marriage. The broader cultural move of marriage from permission to consent, and from economic to companionate unions, gave same sex-couples even more grounds to wonder why their partnerships remained unsanctioned. But the consent of marital partners still required exogenous permission from the state, so the same-sex marriage question was inevitably a political question.
As the political battle grew, libertarians were quick to suggest that such battles were unavoidable as a result of the state's role in defining marriage. A number of libertarians argued for eliminating the state's role in marriage entirely as a way to cut the Gordian knot.
Recognizing that such a radical solution was a non-starter politically, other libertarians—including Reason and the Libertarian Party—backed state-sanctioned same-sex marriage as a second-best option as early as the 1970s. Given the long-standing classical liberal commitment to equality before the law, this position was a legitimate one. The state's involvement in marriage was hardly likely to disappear in the near future. We should expect the debate over plural marriage to raise all of these issues again.
Getting the State Out of Everything If the problem with getting the state out of marriage is that marriage matters for so many other things that the state does, why not focus on the state doing fewer of those things? Reducing the state's role in other areas is already a good idea, but all of those policies can also be seen as intermediary steps, part of eliminating the state's role in marriage.
For example, an anti-poverty agenda that includes things like ending occupational licensure, reducing or eliminating the minimum wage, removing restrictive zoning laws that raise housing prices and discourage people from starting small businesses, and introducing meaningful competition into K–12 education could reduce the need for government welfare programs. Such programs rely on marital status to categorize benefit recipients even as program incentives often work to undermine marriage and its social benefits. Transforming anti-poverty policies reduces the political importance of marital status at the same time it strengthens marriage as a social institution.
One could make a similar argument about policies from taxation to health care. Reducing and flattening tax rates are already part of most libertarian thinking about taxes, but they've rarely been seen as a tactic to reduce the state's role in marriage by reducing the scope of government policies that rely on marital status. Similarly, freeing up health care markets from government regulation and ending the tax-favored treatment of employer-provided health insurance are good ideas in themselves, and such changes will also have the side benefit of reducing the importance of marital status for public policy.
Getting the state out of marriage requires that we get the state out of a whole number of other things first. Then the only marriage contract we'll need will be the one we made between ourselves, and not the one between ourselves and the state.
This article originally appeared in print under the headline "Getting the State Out of Marriage".