854 Journal of Marriage and Family movement emerged among gay men and lesbi- tection.In the 1950s,half of all unmarried preg- ans in the United States,who sought the right to nant women in the United States married before marry with increasing success.Clearly,mar- the birth of their child.whereas in the 1990s. riage remains important to many people in the only one fourth married (U.S.Census Bureau, United States.Consequently,I think the inter- 1999).Finally,evolutionary theory cannot esting question is not why so few people are explain the persistence of the formal wedding marrying,but rather,why so many people are style in which people are still marrying (see marrying,or planning to marry,or hoping to below).Studies of preindustrial societies have marry,when cohabitation and single parent- found that although many have elaborate ceremo- hood are widely acceptable options.(This ques- nies,others have little or no ceremony (Ember. tion may be less relevant in Canada and the Ember,Peregrine,2002;Stephens,1963). many European nations where the estimated The mid-20th century specialization model of proportions of who will ever marry are lower.) economist Gary Becker (1965,1981)also seems less relevant than when it was intro- The Gains to Marriage duced.Becker assumed that women were rela- tively more productive at home than men,and The dominant theoretical perspectives on mar- that men were relatively more productive (i.e., riage in the 20th century do not provide much they could earn higher wages)in the labor mar- guidance on the question of why marriage re- ket.He argued that women and men could mains so popular.The structural functionalists increase their utility by exchanging,through in social anthropology and sociology in the marriage,women's home work for men's labor early-to mid-20th century emphasized the role market work.The specialization model would of marriage in ensuring that a child would have predict that in the present era,women with less a link to the status of a man,a right to his pro- labor market potential would be more likely to tection,and a claim to inherit his property marry because they would gain the most econom- (Mair,1971).But as the law began to recognize ically from finding a husband.But several studies the rights of children born outside marriage,and show that in recent decades,women in the United as mothers acquired resources by working in States and Canada with less education (and there- the paid work force,these reasons for marriage fore less labor market potential)are less likely to become less important. marry (Lichter,McLaughlin,Kephart,Landry, Nor is evolutionary theory very helpful. 1992;Oppenheimer,Blossfeld,Wackerow, Although there may be important evolutionary 1995;Qian Preston,1993;Sweeney,2002; influences on family behavior.it is unlikely that Turcotte Goldscheider,1998).This finding humans have developed an innate preference suggests that the specialization model may no for marriage as we know it.The classical longer hold.Moreover,the specialization model account of our evolutionary heritage is that was developed before cohabitation was wide- women,whose reproductive capacity is limited spread,and offers no explanation for why cou- by pregnancy and lactation (which delays the ples would marry rather than cohabit. return of ovulation),seek stable pair bonds with From a rational choice perspective,then, men,whereas men seek to maximize their fertil- what benefits might contemporary marriage ity by impregnating many women.Rather than offer that would lead cohabiting couples to being "natural,"marriage-centered kinship was marry rather than cohabit?I suggest that the described in much early-and mid-20th century major benefit is what we might call enforceable anthropological writing as the social invention trust (Cherlin,2000;Portes Sensenbrenner. that solved the problem of the sexually wander- 1993).Marriage still requires a public commit- ing male (Tiger Fox,1971).Moreover,when ment to a long-term,possibly lifelong relation- dependable male providers are not available, ship.This commitment is usually expressed in women may prefer a reproductive strategy of front of relatives,friends.and religious con- relying on a network of female kin and more gregants.Cohabitation,in contrast,requires than one man (Hrdy,1999).In addition,mar- only a private commitment,which is easier to riages are increasingly being formed well after break.Therefore,marriage,more so than cohab- a child is born,yet evolutionary theory suggests itation,lowers the risk that one's partner will that the impetus to marry should be greatest renege on agreements that have been made.In when newbom children need support and pro- the language of economic theory,marriagemovement emerged among gay men and lesbi￾ans in the United States, who sought the right to marry with increasing success. Clearly, mar￾riage remains important to many people in the United States. Consequently, I think the inter￾esting question is not why so few people are marrying, but rather, why so many people are marrying, or planning to marry, or hoping to marry, when cohabitation and single parent￾hood are widely acceptable options. (This ques￾tion may be less relevant in Canada and the many European nations where the estimated proportions of who will ever marry are lower.) The Gains to Marriage The dominant theoretical perspectives on mar￾riage in the 20th century do not provide much guidance on the question of why marriage re￾mains so popular. The structural functionalists in social anthropology and sociology in the early- to mid-20th century emphasized the role of marriage in ensuring that a child would have a link to the status of a man, a right to his pro￾tection, and a claim to inherit his property (Mair, 1971). But as the law began to recognize the rights of children born outside marriage, and as mothers acquired resources by working in the paid work force, these reasons for marriage become less important. Nor is evolutionary theory very helpful. Although there may be important evolutionary influences on family behavior, it is unlikely that humans have developed an innate preference for marriage as we know it. The classical account of our evolutionary heritage is that women, whose reproductive capacity is limited by pregnancy and lactation (which delays the return of ovulation), seek stable pair bonds with men, whereas men seek to maximize their fertil￾ity by impregnating many women. Rather than being ‘‘natural,’’ marriage-centered kinship was described in much early- and mid-20th century anthropological writing as the social invention that solved the problem of the sexually wander￾ing male (Tiger & Fox, 1971). Moreover, when dependable male providers are not available, women may prefer a reproductive strategy of relying on a network of female kin and more than one man (Hrdy, 1999). In addition, mar￾riages are increasingly being formed well after a child is born, yet evolutionary theory suggests that the impetus to marry should be greatest when newborn children need support and pro￾tection. In the 1950s, half of all unmarried preg￾nant women in the United States married before the birth of their child, whereas in the 1990s, only one fourth married (U.S. Census Bureau, 1999). Finally, evolutionary theory cannot explain the persistence of the formal wedding style in which people are still marrying (see below). Studies of preindustrial societies have found that although many have elaborate ceremo￾nies, others have little or no ceremony (Ember, Ember, & Peregrine, 2002; Stephens, 1963). The mid-20th century specialization model of economist Gary Becker (1965, 1981) also seems less relevant than when it was intro￾duced. Becker assumed that women were rela￾tively more productive at home than men, and that men were relatively more productive (i.e., they could earn higher wages) in the labor mar￾ket. He argued that women and men could increase their utility by exchanging, through marriage, women’s home work for men’s labor market work. The specialization model would predict that in the present era, women with less labor market potential would be more likely to marry because they would gain the most econom￾ically from finding a husband. But several studies show that in recent decades, women in the United States and Canada with less education (and there￾fore less labor market potential) are less likely to marry (Lichter, McLaughlin, Kephart, & Landry, 1992; Oppenheimer, Blossfeld, & Wackerow, 1995; Qian & Preston, 1993; Sweeney, 2002; Turcotte & Goldscheider, 1998). This finding suggests that the specialization model may no longer hold. Moreover, the specialization model was developed before cohabitation was wide￾spread, and offers no explanation for why cou￾ples would marry rather than cohabit. From a rational choice perspective, then, what benefits might contemporary marriage offer that would lead cohabiting couples to marry rather than cohabit? I suggest that the major benefit is what we might call enforceable trust (Cherlin, 2000; Portes & Sensenbrenner, 1993). Marriage still requires a public commit￾ment to a long-term, possibly lifelong relation￾ship. This commitment is usually expressed in front of relatives, friends, and religious con￾gregants. Cohabitation, in contrast, requires only a private commitment, which is easier to break. Therefore, marriage, more so than cohab￾itation, lowers the risk that one’s partner will renege on agreements that have been made. In the language of economic theory, marriage 854 Journal of Marriage and Family