Americans often talk as if marriage were a private, personal relationship. But when two people live together for their own strictly private reasons, and carve out their own, strictly private bargain about the relationship, we call that relationship not marriage but "cohabitation." In America, it is now more popular than ever. More men and women are moving in together, sharing an apartment and a bed, without getting married first. The latest Census Bureau figures show four million couples living together outside of marriage (not counting gay couples), eight times as many as in 1970. And many more people have cohabited than are currently doing so; recent figures show that almost two-thirds of young adult men and women chose to cohabit first rather than marry directly.

Most cohabitations are quite short-lived; they typically last for about a year or a little more and then are transformed into marriages or dissolve. Although many observers expected the United States to follow the path blazed by the Nordic countries toward a future of informal but stable relationships, this has not happened. We see no sign that cohabitation is becoming a long-term alternative to marriage in the U.S. It has remained a stage in the courtship process or a temporary expediency, but not typically a stable social arrangement. Thus, by resembling marriage in some ways and differing from it in others, cohabitation brings some but not all of the costs and benefits of marriage.

The Cohabitation Deal versus the Marriage Bargain

Cohabitation is a tentative, non-legal coresidential union. It does not require or imply a lifetime commitment to stay together. Even if one partner expects the relationship to be permanent, the other partner often does not. Cohabiting unions break up at a much higher rate than marriages. Cohabitors have no responsibility for financial support of their partner and most do not pool financial resources. Cohabitors are more likely than married couples to both value separate leisure activities and to keep their social lives independent. Although most cohabitors expect their relationship to be sexually exclusive, in fact they are much less likely than husbands and wives to be monogamous.

A substantial proportion of cohabiting couples have definite plans to marry, and these couples tend to behave like already-married couples. Others have no plans to marry and these tentative and uncommitted relationships are bound together by the "cohabitation deal" rather than the "marriage bargain." In fact, couples may choose cohabitation precisely because it carries no formal constraints or responsibilities.

But the deal has costs. The tentative, impermanent, and socially unsupported nature of cohabitation impedes the ability of this type of partnership to deliver many of the benefits of marriage, as does the relatively separate lives typically pursued by cohabiting partners. The uncertainty about the stability and longevity of the relationship makes both investment in the relationship and specialization with this partner much riskier than in marriage. Couples who expect to stay together for the very long run can develop some skills and let others atrophy because they can count on their spouse (or partner) to fill in where they are weak. This specialization means that couples working together in a long-term partnership will produce more than the same people would working alone. But cohabitation reduces the benefits and increases the costs of specializing-it is much safer to just do everything for yourself since you don't know whether the partner you are living with now will be around next year. So cohabiting couples typically produce less than married couples.

The temporary and informal nature of cohabitation also makes it more difficult and riskier for extended family to invest in and support the relationship. Parents, siblings, friends of the partners are less likely to get to know a cohabiting partner than a spouse and, more important, less likely to incorporate a person who remains outside "the family" into its activities, ceremonies, and financial dealings. Parents of one member of a cohabiting couple are ill-advised to invest in the partner emotionally or financially until they see if the relationship will be long term. They are also ill-advised to become attached to children of their child's cohabiting partner because their "grandparent" relationship with that child will dissolve if the cohabitation splits up. Marriage and plans to marry make that long-term commitment explicit and reduce the risk to families of incorporating the son- or daughter-in-law and stepchildren.

The separateness of cohabitors' lives also reduces their usefulness as a source of support during difficult times. Julie Brines and Kara Joyner, writing in the American Sociological Review, argue that cohabitors tend to expect each person to be responsible for supporting him or herself, and failure to do so threatens the relationship. The lack of sharing typical of cohabitors disadvantages the women and their children in these families relative to the men, because women typically earn less than men and this is especially true for mothers.

Another drawback of cohabitation is that it seems to distance people from some important social institutions, especially organized religion. Most formal religions disapprove of and discourage cohabitation, making membership in religious communities awkward for unmarried couples. The result is that individuals who enter a cohabitation often reduce their involvement in religious activities. In contrast, people who get married and those who become parents generally become more active. Finally, while young men and women who define themselves as "religious" are less likely to cohabit, those who do cohabit subsequently become less religious.

Cohabitation has become an increasingly important-but poorly delineated-context for child rearing. One quarter of current stepfamilies involve cohabiting couples, and a significant proportion of "single-parent" families are actually two-parent cohabiting families. The parenting role of a cohabiting partner toward the child(ren) of the other person is extremely vaguely defined. The non-parent partner-the man in the substantial majority of cases-has no explicit legal, financial, supervisory, or custodial rights or responsibilities regarding the child of his partner. This ambiguity and lack of enforceable claims by either cohabiting partner or child makes investment in the relationship dangerous for both parties and makes "Mom's boyfriend" a weak and shifting base from which to discipline and guide a child.

What Cohabitation "Produces"

As the previous section showed, marriage fosters certain behavioral changes-by both the couple and those around them-that cohabitation simply doesn't encourage: each partner can specialize; in-laws can get involved; children and their parent's spouse can invest in a mutual relationship; and so on. What, though, are the empirical results of these behavioral changes, and of the many other ways in which the two options differ?

Before seeking to answer this question, it must first be acknowledged that cohabiting couples, especially those with no plans to marry, tend to differ from married couples even before the cohabitation begins. Living with someone rather than marrying attracts people less committed to marriage, and less likely to be successful at it. Thus, selection of people with less to offer a partner and less to gain from marriage accounts for some of the poorer outcomes of cohabitors. But, as we shall see, at least some of the evidence suggests that cohabiting itself also contributes to those outcomes.

Domestic Violence. A recent Census Bureau report speculated that perhaps so many children were being born to unmarried mothers because women were avoiding marriage out of fear of domestic violence and child abuse. Is this a reasonable fear? My own analysis of data from the 1987/88 National Survey of Families and Households shows that married people are about half as likely as cohabiting couples to say that arguments between them and their partner had become physical in the previous year (eight percent of married women compared to 16 percent of cohabiting women). When it comes to "hitting, shoving, and throwing things," cohabiting couples are more than three times more likely than the married to say things get that far out of hand. One reason cohabitors are more violent is that they are, on average, younger and less educated. But even after controlling for education, race, age, and gender, people who live together are 1.8 times more likely to report violent arguments than married people.

It matters a great deal, however, whether cohabiting couples have definite plans to marry. Engaged cohabitors are no more likely to report violence than married couples, but cohabitors with no plans to marry are twice as likely to report couple violence as either married or engaged couples. Women in uncommitted cohabiting relationships seem to be especially at risk of violence directed toward them. The well-being of married and engaged cohabiting couples is substantially higher on this dimension than uncommitted cohabiting couples. Some researchers suggest that commitment to the relationship and to the partner reduces violence. These differentials seem to support that view.

Sex. Sex appears to be a key part of the cohabiting "deal." According to the 1992 National Health and Social Life Survey, cohabiting men and women make love on average between seven and seven and a half times a month, or about one extra sex act a month than married people. But cohabiting men and women are less likely than those who are married to be monogamous, although virtually all say that they expect their relationship to be sexually exclusive. Renata Forste and Koray Tanfer find in the National Survey of Women that four percent of married women had a secondary sex partner compared to 20 percent of cohabiting women and 18 percent of dating women. Women's behavior changed dramatically when they married, with a huge decline in the chances of having a secondary sex partner. Forste and Tanfer conclude that marriage itself increases sexual exclusivity; cohabitation is no better than "dating" on this dimension.

Housework. Women who are living with men tend to do more housework than women living alone or with other women. A recent study by Scott South and Glenna Spitze shows that once they take into account the presence of children and others and characteristics of the partners, married women spend 14 hours more than married men do. Women who are cohabiting spend about ten hours more on housework than cohabiting men. On this dimension, then, cohabitation would seem to be a better deal for women than marriage. Some economists, however, would argue that husbands compensate their wives for their time in work for the family by sharing their income with them, while cohabiting women generally don't share their partner's earnings, so they may be doing extra housework without extra pay.

Wealth. Married couples link their fates-including their finances. This is a more attractive proposition if one's intended has a decent income and few debts. But if not, living together is a way to avoid taking on the debts-current or future-of the partner. It also allows couples to avoid the "marriage penalty" in tax code-an issue for two-worker couples with fairly equal incomes (but couples with unequal earnings could see tax benefits if they marry and share income). Since the income of one's spouse (but not one's cohabiting partner) is counted in determining eligibility for benefits under government programs like Food Stamps and the Earned Income Tax Credit, the implicit tax on marriage in these programs can be very high, as Eugene Steuerle, writing in this publication, has pointed out.

Selection of those with few resources into cohabitation-and/or the negative effects of the cohabitation bargain-combine to leave couples who are living together with relatively little money. LingXin Hao, writing in Social Forces, shows that among all families with children, cohabiting couples have the lowest average level of wealth, comparable to families headed by a single mother. Intact two-parent families and stepfamilies have the highest level of wealth, followed at a distance by families headed by a single father. Unlike single-parent families, cohabiting couples have two potential earners, so their very low levels of wealth are less expected. But expected or not, they are a cause for concern, especially for the children living in these families.

Emotional Well-Being. Marriage is, by design and agreement, for the long run. Married people, thus, see their relationship as much more stable than cohabiting couples do. And for any couple, thinking that the relationship is likely to break up has a dampening effect on the spirits. The result: cohabitors show lower psychological well-being than similar married people. Specifically, cohabitors report being more depressed and less satisfied with life than do married people. And according to sociologist Susan Brown, worrying that one's relationship will break up is especially distressing for cohabiting women with children, who show quite high levels of depression as a result.

Perhaps, however, cohabiting people are more depressed because depressed and dissatisfied people have trouble getting married. Not so, says Brown. She found that cohabitors' higher levels of depression are not explained by their scores before the start of the union. Rather it is a person's perception of the chances that the relationship will break up that seems to be the chief culprit in his or her poor emotional well-being.

Divorce. People often believe that living together in a "trial marriage" will tell potential partners something about what marriage would be like. The information gained could help couples make good choices and avoid bad ones; cohabiting before marriage could lead to better marriages later. Evidence from the National Survey of Families and Households shows how widespread this belief is. Most cohabitors say that making sure that they are compatible before marriage is an important reason that they wanted to live together.

But a large body of recent evidence now shows quite consistently that people who cohabit and then marry are much more likely to divorce than people who marry without living together. An initial conclusion might be that cohabitation changes people's attitudes in ways that make them less committed to the institution of marriage. However, research conducted by Lee Lillard, Michael Brien, and myself shows that people who cohabit have other characteristics that both lead them to cohabit in the first place and make them poor marriage material. Thus, in the case of divorce, selection would seem to account for the differences between marriage and cohabitation.

Recovering the Wheat without the Chaff

The cumulative evidence clearly suggests that compared to marriage, uncommitted cohabitation-cohabitation by couples who are not engaged-is an inferior social arrangement. Couples who live together with no definite plans to marry are making a different bargain than married couples or engaged cohabitors. The bargain is very much not marriage, and is "marriage-like" only in that couples share an active sex life and a house or apartment. Cohabiting men tend to be quite uncommitted to the relationship; cohabiting women with children tend to be quite uncertain about its future. Levels of domestic violence are much higher in these couples than in either married or engaged cohabiting couples. Children in families headed by an unmarried couple do much worse than children in families with married parents. Uncommitted cohabitation delivers relatively few benefits to men, women, or children. This social arrangement also probably benefits communities less than marriage.

Clearly, the men and women who choose uncommitted cohabitation do not have the same characteristics as those who marry without first living together or who live together while planning their wedding. This selection into cohabitation of people less likely to build a successful marriage seems to account for their higher chances of divorce should they ultimately marry. But cohabitation itself seems to cause attitudes to change in ways inimical to long-term commitment, to damage emotional well-being, and to distance people from religious institutions and from their families. There is also some evidence that cohabitation is less beneficial for children than marriage is. And there is some suggestion that marriage-but not uncommitted cohabitation-reduces domestic violence.

If cohabitation is inferior to marriage, then we as a society would benefit from more of the latter and less of the former. Encouraging marriage over cohabitation involves undoing a whole series of legal and social changes that have undercut the privileged status of marriage. This doesn't mean we should encourage a return to the old model of marriage and the family. But in the justified effort to overcome the sexism and inflexibility of the "1950s" marriage, we may have been too willing to throw the baby out with the bathwater. If we wish to retrieve for more people the benefits that marriage delivers that cohabitation does not, it is important to begin the process of re-privileging marriage now.

The Communitarian Network
2130 H Street, NW, Suite 703
Washington, DC 20052
202.994.6118
comnet@gwu.edu