Deinstitutionalization of American Marriage 849 may even assist in the risky business of predict- The Growth of Cohabitation ing the future of marriage.To some extent,sim- ilar changes in marriage have occurred in the In the 1970s,neither I nor most other American United States,Canada,and much of Europe,but researchers foresaw the greatly increased role of the American situation may be distinctive.Con- cohabitation in the adult life course.We thought sequently,although I include information about that,except among the poor,cohabitation would Canadian and European families,I focus mainly remain a short-term arrangement among child- on the United States. less young adults who would quickly break up or marry.But it has become a more prevalent and more complex phenomenon.For example, THE DEINSTITUTIONALIZATION cohabitation has created an additional layer of OF MARRIAGE complexity in stepfamilies.When I wrote my Even as I was writing my 1978 article,the article,nearly all stepfamilies were formed by changing division of labor in the home and the the remarriage of one or both spouses.Now, increase in childbearing outside marriage were about one fourth of all stepfamilies in the undermining the institutionalized basis of mar- United States,and one half of all stepfamilies in riage.The distinct roles of homemaker and Canada,are formed by cohabitation rather than breadwinner were fading as more married marriage (Bumpass,Raley,Sweet,1995;Sta- women entered the paid labor force.Looking tistics Canada,2002).It is not uncommon,espe- into the future,I thought that perhaps an equita- cially among the low-income population,for ble division of household labor might become a woman to have a child outside marriage,end institutionalized.But what happened instead her relationship with that partner,and then was the "stalled revolution,"in Hochschild's begin cohabiting with a different partner.This (1989)well-known phrase.Men do somewhat new union is equivalent in structure to a step- more home work than they used to do,but there family but does not involve marriage.Some- is wide variation,and each couple must work times the couple later marries,and if neither has out their own arrangement without clear guide- been married before.their union creates a first lines.In addition,when I wrote the article,I out marriage with stepchildren.As a result,we now of 6 births in the United States occurred outside see an increasing number of stepfamilies that do marriage,already a much higher ratio than at not involve marriage,and an increasing number midcentury (U.S.National Center for Health of first marriages that involve stepfamilies. Statistics,1982).Today,the comparable figure More generally,cohabitation is becoming is 1 out of 3 (U.S.National Center for Health accepted as an alternative to marriage.British Statistics,2003).The percentage is similar in demographer Kathleen Kiernan (2002)writes Canada (Statistics Canada.2003)and in the that the acceptance of cohabitation is occurring United Kingdom and Ireland (Kiernan,2002). in stages in European nations,with some na- In the Nordic countries of Denmark,Iceland, tions further along than others.In stage one, Norway,and Sweden,the figure ranges from cohabitation is a fringe or avant garde phenome- about 45%to about 65%(Kiernan).Marriage is non;in stage two,it is accepted as a testing no longer the nearly universal setting for child- ground for marriage;in stage three,it becomes bearing that it was a half century ago. acceptable as an alternative to marriage;and in Both of these developments-the changing stage four,it becomes indistinguishable from division of labor in the home and the increase in marriage.Sweden and Denmark,she argues, childbearing outside marriage-were well under have made the transition to stage four:in con- way when I wrote my 1978 article,as was trast,Mediterranean countries such as Spain, a steep rise in divorce.Here I discuss two more Italy,and Greece remain in stage one.In the recent changes in family life,both of which early 2000s,the United States appeared to be in have contributed to the deinstitutionalization of transition from stage two to stage three (Smock marriage after the 1970s:the growth of cohabi- Gupta,2002).A number of indicators sug- tation,which began in the 1970s but was not gested that the connection between cohabitation fully appreciated until it accelerated in the and marriage was weakening.The proportion of 1980s and 1990s,and same-sex marriage, cohabiting unions that end in marriage within which emerged as an issue in the 1990s and has 3 years dropped from 60%in the 1970s to come to the fore in the current decade. about 33%in the 1990s (Smock Gupta).may even assist in the risky business of predict￾ing the future of marriage. To some extent, sim￾ilar changes in marriage have occurred in the United States, Canada, and much of Europe, but the American situation may be distinctive. Con￾sequently, although I include information about Canadian and European families, I focus mainly on the United States. THE DEINSTITUTIONALIZATION OF MARRIAGE Even as I was writing my 1978 article, the changing division of labor in the home and the increase in childbearing outside marriage were undermining the institutionalized basis of mar￾riage. The distinct roles of homemaker and breadwinner were fading as more married women entered the paid labor force. Looking into the future, I thought that perhaps an equita￾ble division of household labor might become institutionalized. But what happened instead was the ‘‘stalled revolution,’’ in Hochschild’s (1989) well-known phrase. Men do somewhat more home work than they used to do, but there is wide variation, and each couple must work out their own arrangement without clear guide￾lines. In addition, when I wrote the article, 1 out of 6 births in the United States occurred outside marriage, already a much higher ratio than at midcentury (U.S. National Center for Health Statistics, 1982). Today, the comparable figure is 1 out of 3 (U.S. National Center for Health Statistics, 2003). The percentage is similar in Canada (Statistics Canada, 2003) and in the United Kingdom and Ireland (Kiernan, 2002). In the Nordic countries of Denmark, Iceland, Norway, and Sweden, the figure ranges from about 45% to about 65% (Kiernan). Marriage is no longer the nearly universal setting for child￾bearing that it was a half century ago. Both of these developments—the changing division of labor in the home and the increase in childbearing outside marriage—were well under way when I wrote my 1978 article, as was a steep rise in divorce. Here I discuss two more recent changes in family life, both of which have contributed to the deinstitutionalization of marriage after the 1970s: the growth of cohabi￾tation, which began in the 1970s but was not fully appreciated until it accelerated in the 1980s and 1990s, and same-sex marriage, which emerged as an issue in the 1990s and has come to the fore in the current decade. The Growth of Cohabitation In the 1970s, neither I nor most other American researchers foresaw the greatly increased role of cohabitation in the adult life course. We thought that, except among the poor, cohabitation would remain a short-term arrangement among child￾less young adults who would quickly break up or marry. But it has become a more prevalent and more complex phenomenon. For example, cohabitation has created an additional layer of complexity in stepfamilies. When I wrote my article, nearly all stepfamilies were formed by the remarriage of one or both spouses. Now, about one fourth of all stepfamilies in the United States, and one half of all stepfamilies in Canada, are formed by cohabitation rather than marriage (Bumpass, Raley, & Sweet, 1995; Sta￾tistics Canada, 2002). It is not uncommon, espe￾cially among the low-income population, for a woman to have a child outside marriage, end her relationship with that partner, and then begin cohabiting with a different partner. This new union is equivalent in structure to a step￾family but does not involve marriage. Some￾times the couple later marries, and if neither has been married before, their union creates a first marriage with stepchildren. As a result, we now see an increasing number of stepfamilies that do not involve marriage, and an increasing number of first marriages that involve stepfamilies. More generally, cohabitation is becoming accepted as an alternative to marriage. British demographer Kathleen Kiernan (2002) writes that the acceptance of cohabitation is occurring in stages in European nations, with some na￾tions further along than others. In stage one, cohabitation is a fringe or avant garde phenome￾non; in stage two, it is accepted as a testing ground for marriage; in stage three, it becomes acceptable as an alternative to marriage; and in stage four, it becomes indistinguishable from marriage. Sweden and Denmark, she argues, have made the transition to stage four; in con￾trast, Mediterranean countries such as Spain, Italy, and Greece remain in stage one. In the early 2000s, the United States appeared to be in transition from stage two to stage three (Smock & Gupta, 2002). A number of indicators sug￾gested that the connection between cohabitation and marriage was weakening. The proportion of cohabiting unions that end in marriage within 3 years dropped from 60% in the 1970s to about 33% in the 1990s (Smock & Gupta), Deinstitutionalization of American Marriage 849