TWO DECADES OF FAMILY CHANGE: THE SHIFTING ECONOMIC FOUNDATIONS OF MARRIAGE MEGAN M.SWEENEY University of California,Los Angeles Has the relationship between economic prospects and marriage formation in the United States changed in recent decades?To answer this question,a discrete-time event-history analysis was conducted using data from multiple cohorts of the Na- tional Longitudinal Surveys of Labor Market Experience.Among women,results indicate growth in the importance of earnings for marriage formation between the early baby-boom cohort (born between 1950 and 1954)and late baby-boom cohort (born between 1961 and 1965).Evidence of cohort change in the relationship be- tween men's economic prospects and marriage,however,is limited.Despite impor- tant racial differences in the economic and attitudinal context of marriage,key re- sults are generally similar for whites and for African Americans.Taken together, these findings imply that men and women are growing to resemble one another with respect to the relationship between economic prospects and marriage,although this convergence is driven primarily by changing patterns of marriage among women. These results are largely supportive of Oppenheimer's career-entry theory of mar- riage and suggest that Becker's specialization and trading model of marriage may be outdated. N THE DECADES since the mid-1960s, tion nearly doubled from 30 to 58 percent the United States has experienced great (U.S.Bureau of the Census 1994).The in- change in both the marriage and labor force creasing delay of marriage has been accom- participation rates of women.Between 1965 panied by rapid growth in the earnings of and 1993,the median age at first marriage married women but by only slow growth in rose almost four years,to 26.5 years for men the earnings of men(Cancian,Danziger,and and 24.5 years for women.During this pe- Gottschalk 1993).Social scientists fre- riod,married women's labor force participa- quently attribute declines in marriage to in- creases in women's economic independence Direct all correspondence to Megan M. resulting from these trends (e.g.,Becker Sweeney,Department of Sociology,UCLA,264 [1981]1992;Cherlin 1992;Preston and Haines Hall,Los Angeles CA 90095(msweeney Richards 1975;Waite and Spitze 1981). @soc.ucla.edu).A version of this paper was pre- sented at the meetings of the Population Asso- Commonly labeled the "economic indepen- ciation of America,Washington,D.C.,March dence hypothesis,"this argument assumes 1997.Support for this research was provided by that women with good prospects in the labor the National Institute of Child Health and Human market will be less likely to marry than will Development(HD07014)and by core support women with relatively poorer prospects.As from NICHD to the Center for Demography and women's rising incomes reduce economic Ecology,University of Wisconsin-Madison(P30 HD05876).I am grateful to Larry Bumpass, dependence on a spouse,many argue that the Maria Cancian,Steven Cook,Robert Hauser, incentive for women to marry has dimin- Robert Mare,Julie Phillips,and five anonymous ished. reviewers for their comments on previous ver- This perspective presumes a model of mar- sions of this paper. riage characterized by a high degree of spe- 132 AMERICAN SOCIOLOGICAL REVIEW,2002,VOL.67(FEBRUARY:132-147) This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWo DECADES OF FAMILY CHANGE: THE SHIFTING ECONOMIC FOUNDATIONS OF MARRIAGE MEGAN M. SWEENEY University of California, Los Angeles Has the relationship between economic prospects and marriage formation in the United States changed in recent decades? To answer this question, a discrete-time event-history analysis was conducted using data from multiple cohorts of the National Longitudinal Surveys of Labor Market Experience. Among women, results indicate growth in the importance of earnings for marriage formation between the early baby-boom cohort (born between 1950 and 1954) and late baby-boom cohort (born between 1961 and 1965). Evidence of cohort change in the relationship between men's economic prospects and marriage, however, is limited. Despite important racial differences in the economic and attitudinal context of marriage, key results are generally similar for whites and for African Americans. Taken together, these findings imply that men and women are growing to resemble one another with respect to the relationship between economic prospects and marriage, although this convergence is driven primarily by changing patterns of marriage among women. These results are largely supportive of Oppenheimer's career-entry theory of marriage and suggest that Becker's specialization and trading model of marriage may be outdated. IN THE DECADES since the mid-1960s, the United States has experienced great change in both the marriage and labor force participation rates of women. Between 1965 and 1993, the median age at first marriage rose almost four years, to 26.5 years for men and 24.5 years for women. During this period, married women's labor force participation nearly doubled from 30 to 58 percent (U.S. Bureau of the Census 1994). The increasing delay of marriage has been accompanied by rapid growth in the earnings of married women but by only slow growth in the earnings of men (Cancian, Danziger, and Gottschalk 1993). Social scientists frequently attribute declines in marriage to increases in women's economic independence resulting from these trends (e.g., Becker [1981] 1992; Cherlin 1992; Preston and Richards 1975; Waite and Spitze 1981). Commonly labeled the "economic independence hypothesis," this argument assumes that women with good prospects in the labor market will be less likely to marry than will women with relatively poorer prospects. As women's rising incomes reduce economic dependence on a spouse, many argue that the incentive for women to marry has diminished. This perspective presumes a model of marriage characterized by a high degree of speDirect all correspondence to Megan M. Sweeney, Department of Sociology, UCLA, 264 Haines Hall, Los Angeles CA 90095 (msweeney @soc.ucla.edu). A version of this paper was presented at the meetings of the Population Association of America, Washington, D.C., March 1997. Support for this research was provided by the National Institute of Child Health and Human Development (HD07014) and by core support from NICHD to the Center for Demography and Ecology, University of Wisconsin-Madison (P30 HD05876). I am grateful to Larry Bumpass, Maria Cancian, Steven Cook, Robert Hauser, Robert Mare, Julie Phillips, and five anonymous reviewers for their comments on previous versions of this paper. 132 AMERICAN SOCIOLOGICAL REVIEW, 2002, VOL. 67 (FEBRUARY: 132-147) This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 133 cialization in sex roles,with men expecting torical contexts.Indeed,trends that began to focus on market work after marriage and when the early baby-boomers were entering women expecting to concentrate their efforts adulthood in the late 1960s and 1970s- at home.Such a view suggests that good trends such as improved economic opportu- prospects in the labor market will increase nities for women and widespread movement the likelihood of marriage among men,who toward more egalitarian gender role ideals- will feel better prepared to marry,but reduce were more firmly established by the 1980s the likelihood of marriage among women, and early 1990s,when the late baby- who will see marriage as a relatively less at- boomers were moving into adulthood. tractive option than continuing to work out- Three questions about potential change in side the home.Yet models that were designed the process and context of marriage guide the to explain marriage before the revolutionary current research.First,as women are increas- economic changes of the 1960s and 1970s ingly expected to work outside the home over may be inappropriate for understanding mar- the course of their lives and as patterns of riage formation in more recent historical pe- consumption and gender role attitudes have riods.Building on the work of Oppenheimer changed,I ask whether women's economic (1988),I consider an alternative possibility: prospects(as indicated by their earnings, Rather than making marriage unnecessary or educational attainment,and employment sta- undesirable,changes in the labor market po- tus)have become more important for mar- sitions of both women and men-along with riage formation over time.Second,as women contemporaneous shifts in gender role atti- are increasingly able (and expected)to con- tudes and patterns of consumption-have al- tribute to the economic maintenance of their tered the nature of the marital bargain. families after marriage,I ask whether men's Through an investigation of the changing re- economic prospects have become somewhat lationship between economic prospects and less important for marriage formation over entry into first marriage,I examine the pos- time.Finally,given relatively large racial dif- sibility that the economic foundations of ferences in the economic and attitudinal con- marriage have shifted. text of marriage,I ask whether the nature of Although previous research has examined recent historical change in the relationship the association between marriage and eco- between economic prospects and marriage nomic prospects in particular historical peri- differs for blacks and for whites. ods,few studies have been explicitly de- signed to investigate historical change,thus THEORY,CONTEXT,AND limiting our understanding of the roots and PREVIOUS RESEARCH meanings of contemporary marriage patterns (Modell 1999).I use a classic demographic Arguments suggesting that improvements in technique to study the process of social women's economic standing are responsible change:a comparison of the experiences of for recent declines in marriage derive theo- successive birth cohorts.Indeed,Ryder retical support from Becker's ([1981]1992) (1965)argues that the continued replace- “specialization and trading”model of mar- ment of one cohort by another greatly facili- riage.Borrowing ideas from the interna- tates transformations in societies,and fur- tional trade literature,Becker views single ther,"if change does occur,it differentiates men and women as trading partners who cohorts from one another,and the compari- choose to marry only when both partners be- son of their careers becomes a way to study lieve that they will be better off married than change"(p.844).To this end,the current single.All else held constant,the gains to analysis combines data from multiple marriage are greatest when men and women sources to compare the marriage formation specialize in the labor market and home,re- behaviors of the "early baby-boom"cohort spectively,and trade on their comparative (born between 1950 and 1954)and the "late advantages in these tasks.Becker argues that baby-boom"cohort (born between 1961 and "the gain from marriage is reduced...by 1965)in the United States.These cohorts higher earnings and labor force participation reached adulthood,and made decisions of married women,because the sexual divi- about marriage,in somewhat different his- sion of labor within households becomes This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 133 cialization in sex roles, with men expecting to focus on market work after marriage and women expecting to concentrate their efforts at home. Such a view suggests that good prospects in the labor market will increase the likelihood of marriage among men, who will feel better prepared to marry, but reduce the likelihood of marriage among women, who will see marriage as a relatively less attractive option than continuing to work outside the home. Yet models that were designed to explain marriage before the revolutionary economic changes of the 1960s and 1970s may be inappropriate for understanding marriage formation in more recent historical periods. Building on the work of Oppenheimer (1988), I consider an alternative possibility: Rather than making marriage unnecessary or undesirable, changes in the labor market positions of both women and men-along with contemporaneous shifts in gender role attitudes and patterns of consumption-have altered the nature of the marital bargain. Through an investigation of the changing relationship between economic prospects and entry into first marriage, I examine the possibility that the economic foundations of marriage have shifted. Although previous research has examined the association between marriage and economic prospects in particular historical periods, few studies have been explicitly designed to investigate historical change, thus limiting our understanding of the roots and meanings of contemporary marriage patterns (Modell 1999). I use a classic demographic technique to study the process of social change: a comparison of the experiences of successive birth cohorts. Indeed, Ryder (1965) argues that the continued replacement of one cohort by another greatly facilitates transformations in societies, and further, "if change does occur, it differentiates cohorts from one another, and the comparison of their careers becomes a way to study change" (p. 844). To this end, the current analysis combines data from multiple sources to compare the marriage formation behaviors of the "early baby-boom" cohort (born between 1950 and 1954) and the "late baby-boom" cohort (born between 1961 and 1965) in the United States. These cohorts reached adulthood, and made decisions about marriage, in somewhat different historical contexts. Indeed, trends that began when the early baby-boomers were entering adulthood in the late 1960s and 1970strends such as improved economic opportunities for women and widespread movement toward more egalitarian gender role idealswere more firmly established by the 1980s and early 1990s, when the late babyboomers were moving into adulthood. Three questions about potential change in the process and context of marriage guide the current research. First, as women are increasingly expected to work outside the home over the course of their lives and as patterns of consumption and gender role attitudes have changed, I ask whether women's economic prospects (as indicated by their earnings, educational attainment, and employment status) have become more important for marriage formation over time. Second, as women are increasingly able (and expected) to contribute to the economic maintenance of their families after marriage, I ask whether men's economic prospects have become somewhat less important for marriage formation over time. Finally, given relatively large racial differences in the economic and attitudinal context of marriage, I ask whether the nature of recent historical change in the relationship between economic prospects and marriage differs for blacks and for whites. THEORY, CONTEXT, AND PREVIOUS RESEARCH Arguments suggesting that improvements in women's economic standing are responsible for recent declines in marriage derive theoretical support from Becker's ([198111992) "specialization and trading" model of marriage. Borrowing ideas from the international trade literature, Becker views single men and women as trading partners who choose to marry only when both partners believe that they will be better off married than single. All else held constant, the gains to marriage are greatest when men and women specialize in the labor market and home, respectively, and trade on their comparative advantages in these tasks. Becker argues that "the gain from marriage is reduced ... by higher earnings and labor force participation of married women, because the sexual division of labor within households becomes This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
134 AMERICAN SOCIOLOGICAL REVIEW less advantageous"(p.55).Becker's theory in recent decades.Income growth since 1960 thus implies that having a good position in was greater for women than for men,and the the labor market will most likely increase proportion of women in the labor force has marriage among men but reduce marriage increased dramatically since 1960,particu- among women,again,all else held constant. larly among women who are white,married, Yet in her "career-entry"theory,Oppen- or who have young children.I Men,however, heimer(1988)suggests that changing condi- experienced some decline in labor supply tions in the labor market have fundamentally during this period (Wetzel 1995).Perhaps altered the nature of the marital bargain. not surprisingly,gender role attitudes in the Consistent with demography's historical em- United States also have changed since the phasis on the perceived economic feasibility 1960s,with an increasing proportion of the of marriage (e.g.,Dixon 1971;Easterlin population holding egalitarian sex role atti- 1980;Hajnal1965;Malthus[1798]1988), tudes (Barich and Bielby 1996;Thornton Oppenheimer argues that a certain standard 1989). of living must be obtained before marriage Changing consumption patterns also alter is considered affordable.In historical peri- the economic context of marriage.As ods when women are not expected to remain Bumpass (1990)stated in his presidential attached to the labor market throughout their address to the Population Association of lives,male labor market position is the key America,"[E]conomic need is a highly economic determinant of marriage.As amorphous concept,always seeming to out- women's patterns of labor force participation strip what we have"(p.489).Members of come to more closely resemble those of men. the baby-boom generation may expect a high however,Oppenheimer argues that the char- and rising standard of living based on expe- acteristics considered important in a spouse riences growing up in the relatively prosper- become more symmetrical for husbands and ous 1960s and early 1970s (Jones 1980).Re- wives.In particular,potential wives are in- cent declines in male earnings may increase creasingly evaluated on the basis of their the perceived necessity of a second income. own achieved socioeconomic status and fu- To the extent that owning a home symbol- ture labor market prospects,rather than on izes the middle-class lifestyle,rising hous- the basis of more traditional characteristics ing costs have further made this standard in- such as religion,family background,and creasingly difficult to achieve for a single- physical attractiveness.The "career-entry' earner family (Wetzel 1995).The economic perspective thus implies a positive effect of costs associated with raising children have women's good economic prospects on mar- also increased in recent decades (Casper riage,as well as growth over time in the im- 1995;England and Folbre 1999). portance of women's economic prospects for Taken together,these trends suggest marriage formation.As women's economic growth in the importance of wives'labor position improves,and as women can expect market position for marriage,but have am- to make increasingly large contributions to biguous implications for husbands'labor the economic maintenance of their families, market position.The combination of some we might further expect that male labor mar- decline in male economic standing with im- ket position would become somewhat less provements in standing among women sug- important for marriage formation.Indeed. gests that men's labor market position may Oppenheimer and Lew (1995)suggest that have become less important for marriage .the expectation of a regular work ca- formation.Yet changing patterns of con- reer may enable some women to 'afford'to sumption and the perceived economic re- marry a man who is unlikely to be a great quirements of supporting a family at an"ad- provider but who is desirable in other re- spects'”(p.109). I Although I consider entry into first marriage among single women,the changing economic THE CHANGING CONTEXT OF MARRIAGE roles of married men and women will affect both Much evidence supports the argument that what is valued in a partner and the level of eco- the economic context of marriage has shifted nomic achievement perceived to be necessary be- fore marriage. This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
134 AMERICAN SOCIOLOGICAL REVIEW less advantageous" (p. 55). Becker's theory thus implies that having a good position in the labor market will most likely increase marriage among men but reduce marriage among women, again, all else held constant. Yet in her "career-entry" theory, Oppenheimer (1988) suggests that changing conditions in the labor market have fundamentally altered the nature of the marital bargain. Consistent with demography's historical emphasis on the perceived economic feasibility of marriage (e.g., Dixon 1971; Easterlin 1980; Hajnal 1965; Malthus [1798] 1988), Oppenheimer argues that a certain standard of living must be obtained before marriage is considered affordable. In historical periods when women are not expected to remain attached to the labor market throughout their lives, male labor market position is the key economic determinant of marriage. As women's patterns of labor force participation come to more closely resemble those of men, however, Oppenheimer argues that the characteristics considered important in a spouse become more symmetrical for husbands and wives. In particular, potential wives are increasingly evaluated on the basis of their own achieved socioeconomic status and future labor market prospects, rather than on the basis of more traditional characteristics such as religion, family background, and physical attractiveness. The "career-entry" perspective thus implies a positive effect of women's good economic prospects on marriage, as well as growth over time in the importance of women's economic prospects for marriage formation. As women's economic position improves, and as women can expect to make increasingly large contributions to the economic maintenance of their families, we might further expect that male labor market position would become somewhat less important for marriage formation. Indeed, Oppenheimer and Lew (1995) suggest that ".... the expectation of a regular work career may enable some women to 'afford' to marry a man who is unlikely to be a great provider but who is desirable in other respects" (p. 109). THE CHANGING CONTEXT OF MARRIAGE Much evidence supports the argument that the economic context of marriage has shifted in recent decades. Income growth since 1960 was greater for women than for men, and the proportion of women in the labor force has increased dramatically since 1960, particularly among women who are white, married, or who have young children. I Men, however, experienced some decline in labor supply during this period (Wetzel 1995). Perhaps not surprisingly, gender role attitudes in the United States also have changed since the 1960s, with an increasing proportion of the population holding egalitarian sex role attitudes (Barich and Bielby 1996; Thornton 1989). Changing consumption patterns also alter the economic context of marriage. As Bumpass (1990) stated in his presidential address to the Population Association of America, "[Elconomic need is a highly amorphous concept, always seeming to outstrip what we have" (p. 489). Members of the baby-boom generation may expect a high and rising standard of living based on experiences growing up in the relatively prosperous 1960s and early 1970s (Jones 1980). Recent declines in male earnings may increase the perceived necessity of a second income. To the extent that owning a home symbolizes the middle-class lifestyle, rising housing costs have further made this standard increasingly difficult to achieve for a singleearner family (Wetzel 1995). The economic costs associated with raising children have also increased in recent decades (Casper 1995; England and Folbre 1999). Taken together, these trends suggest growth in the importance of wives' labor market position for marriage, but have ambiguous implications for husbands' labor market position. The combination of some decline in male economic standing with improvements in standing among women suggests that men's labor market position may have become less important for marriage formation. Yet changing patterns of consumption and the perceived economic requirements of supporting a family at an "ad- ' Although I consider entry into first marriage among single women, the changing economic roles of married men and women will affect both what is valued in a partner and the level of economic achievement perceived to be necessary before marriage. This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 135 equate"level may offset any such change in studies have examined the relationship be- the significance of male economic prospects tween economic prospects and marriage for- for marriage. mation separately for black men and white It is also important to note,however,that men (Lloyd and South 1996;Oppenheimer the economic and attitudinal context of mar- et al.1997).Statistical tests generally were riage differs substantially by race.Black not performed to evaluate the significance of men have seen greater erosion in their own racial differences,however.Nor were racial labor market position in recent decades than differences in the nature of historical change have white men (U.S.Bureau of the Census explored. 1984,1991;Wilson 1987).Historically, Previous studies have drawn varying con- black women have been more likely to work clusions about the impact of labor market for pay than have white women,even when position on the marital behavior of women, factors such as education,family income, with notable differences across level of and number of children are controlled analysis.Studies taking an aggregate-level (Goldin 1990).Yet growth in income-par- approach-most often examining the pro- ticularly relative to that of same-race men- portion of married individuals in a particu- has been significantly greater for black lar local area-generally suggest that women than for white women (U.S.Bureau women's good economic prospects are asso- of the Census 1984,1991).In addition,ra- ciated with reduced marriage(Cready et al. cial differences are reported in attitudes to- 1997:Lichter,LeClere,and McLaughlin ward various aspects of family life (Carter 1991;McLanahan and Casper 1995;Preston 1993:South 1991).African Americans tend and Richards 1975;White 1981).Aggregate- to place greater emphasis on economic sta- level studies have been criticized,however, bility in marital decision-making than do because factors that produce aggregate-level whites,and black women are found to be variation in marriage prevalence may not re- less willing than white women to marry a late in the same way to marriage behavior man with fewer resources than themselves among individuals (Oppenheimer 1997). (Bulcroft and Bulcroft 1993).Although the Moreover,studies using cross-sectional data potentially offsetting nature of these factors can face difficulties identifying the correct leads to uncertain predictions about the ex- causal ordering of variables and tend to con- pected nature of racial differences in the trol only a very limited number of character- changing economic basis of marriage,it is istics relevant to marriage formation.This clearly important to consider variation in may lead to biased estimation of the coeffi- patterns of marriage by race. cients of interest.In contrast,individual- level studies of longitudinal data generally PREVIOUS RESEARCH AND THE report no relationship or a positive relation- ship between various indicators of women's PRESENT STUDY economic prospects and marriage formation Previous studies of the economic context of (Cherlin 1980:Goldscheider and Waite marriage have focused most often on three 1986;Lichter et al.1992;Oppenheimer and key indicators of position in the labor mar- Lew 1995:Thornton,Axinn,and Teachman ket:educational attainment,employment 1995:Waite and Spitze 1981).Although few status,and earnings.Regardless of time pe- studies directly test for significant racial dif- riod,data set,or analytical technique em- ferences in the effect of economic prospects ployed,these studies generally report posi- on marriage formation,and none explicitly tive effects of good economic prospects on marriage among men (e.g.,Cooney and riage,several studies instead report a negative Hogan 1991;Cready,Fossett,and Kiecolt effect of male education on marriage (e.g.,Lloyd 1991;Goldscheider and Waite 1986;Koball and South 1996:Mare and Winship 1991).Re- 1998;Oppenheimer,Kalmijn,and Lim sults from models with detailed categorical speci- fications of education suggest that these seem- 1997;Sassler and Schoen 1999).2 Several ingly contradictory results may be driven by the relatively high likelihood of marriage among 2 While most studies report an overall positive white men with less than 12 years of schooling effect of male educational attainment on mar- (Oppenheimer and Lewin 1999). This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 135 equate" level may offset any such change in the significance of male economic prospects for marriage. It is also important to note, however, that the economic and attitudinal context of marriage differs substantially by race. Black men have seen greater erosion in their own labor market position in recent decades than have white men (U.S. Bureau of the Census 1984, 1991; Wilson 1987). Historically, black women have been more likely to work for pay than have white women, even when factors such as education, family income, and number of children are controlled (Goldin 1990). Yet growth in income-particularly relative to that of same-race menhas been significantly greater for black women than for white women (U.S. Bureau of the Census 1984, 1991). In addition, racial differences are reported in attitudes toward various aspects of family life (Carter 1993; South 1991). African Americans tend to place greater emphasis on economic stability in marital decision-making than do whites, and black women are found to be less willing than white women to marry a man with fewer resources than themselves (Bulcroft and Bulcroft 1993). Although the potentially offsetting nature of these factors leads to uncertain predictions about the expected nature of racial differences in the changing economic basis of marriage, it is clearly important to consider variation in patterns of marriage by race. PREVIOUS RESEARCH AND THE PRESENT STUDY Previous studies of the economic context of marriage have focused most often on three key indicators of position in the labor market: educational attainment, employment status, and earnings. Regardless of time period, data set, or analytical technique employed, these studies generally report positive effects of good economic prospects on marriage among men (e.g., Cooney and Hogan 1991; Cready, Fossett, and Kiecolt 1991; Goldscheider and Waite 1986; Koball 1998; Oppenheimer, Kalmijn, and Lim 1997; Sassler and Schoen 1999).2 Several studies have examined the relationship between economic prospects and marriage formation separately for black men and white men (Lloyd and South 1996; Oppenheimer et al. 1997). Statistical tests generally were not performed to evaluate the significance of racial differences, however. Nor were racial differences in the nature of historical change explored. Previous studies have drawn varying conclusions about the impact of labor market position on the marital behavior of women, with notable differences across level of analysis. Studies taking an aggregate-level approach-most often examining the proportion of married individuals in a particular local area-generally suggest that women's good economic prospects are associated with reduced marriage (Cready et al. 1997; Lichter, LeClere, and McLaughlin 1991; McLanahan and Casper 1995; Preston and Richards 1975; White 1981). Aggregatelevel studies have been criticized, however, because factors that produce aggregate-level variation in marriage prevalence may not relate in the same way to marriage behavior among individuals (Oppenheimer 1997). Moreover, studies using cross-sectional data can face difficulties identifying the correct causal ordering of variables and tend to control only a very limited number of characteristics relevant to marriage formation. This may lead to biased estimation of the coefficients of interest. In contrast, individuallevel studies of longitudinal data generally report no relationship or a positive relationship between various indicators of women's economic prospects and marriage formation (Cherlin 1980; Goldscheider and Waite 1986; Lichter et al. 1992; Oppenheimer and Lew 1995; Thornton, Axinn, and Teachman 1995; Waite and Spitze 1981). Although few studies directly test for significant racial differences in the effect of economic prospects on marriage formation, and none explicitly 2 While most studies report an overall positive effect of male educational attainment on marriage, several studies instead report a negative effect of male education on marriage (e.g., Lloyd and South 1996; Mare and Winship 1991). Results from models with detailed categorical specifications of education suggest that these seemingly contradictory results may be driven by the relatively high likelihood of marriage among white men with less than 12 years of schooling (Oppenheimer and Lewin 1999). This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
136 AMERICAN SOCIOLOGICAL REVIEW examines racial differences in the nature of may differ by race.The direction of these historical change in the relationship between differences is uncertain,however,because of the two,previous research does suggest that offsetting effects in the economic and attitu- earnings may matter somewhat more for dinal contexts of marriage. marriage among black women than among white women (Oppenheimer and Lewin DATA 1999). Despite substantial reason to expect his- Data for the present study come from three torical change in the socioeconomic context sources:the Young Men (NLSM),Young of marriage,few recent empirical studies Women (NLSW),and Youth (NLSY)cohorts have directly investigated whether the nature of the National Longitudinal Surveys of La- and strength of the relationship between la- bor Market Experience.The Young Men's bor market position and marriage has shifted (NLSM)and Young Women's (NLSW) over time.Differences in sample definitions, samples are used to investigate marriage construction of key economic variables,and among members of the early baby-boom co- analytical approaches limit the usefulness of hort (born from 1950 to 1954),while data a meta-review of prior studies for under- from the Youth sample (NLSY)are used to standing change in marriage.Although two investigate marriage among members of the recent studies of cross-sectional variation in late baby-boom cohort (born from 1961 to marriage suggest a growing propensity to 1965).The early baby-boom cohort(NLSM marry among well-educated women relative and NLSW samples)largely includes mar- to their less educated peers (Goldstein and riage experiences from the late 1960s Kenney 2001;Qian and Preston 1993),no through the 1970s,while the late baby-boom research to date offers a formal test of co- cohort (NLSY sample)largely includes mar- hort change in the relationship of economic riage experiences during the 1980s and early prospects to marriage formation 1990s.The primary focus of these NLS co- I use multiple sources of longitudinal data horts was to examine the labor market expe- to systematically examine the nature and di- riences of young adults (Center for Human rection of cohort change in the relationship Resource Research 1997),making these data of economic prospects to the formation of ideal for my investigation.Further,all three first marriages.This improves the ability to survey groups include oversamples of examine the roots and meanings of contem- blacks,which facilitates the investigation of porary patterns of marriage and provides in- racial differences in the changing relation- sights into the likely direction of continued ship between economic prospects and mar- change in family life.I test several hypoth- riage. eses about how the relationship between The NLSY sample was interviewed annu- marriage and economic prospects may have ally from 1979 through 1994,the NLSM changed between the early and late baby- sample was interviewed annually from 1966 boom cohorts examined.First,I hypothesize to1971,and then in1973,1975,1976,1978, that the importance of women's economic 1980,and 1981.The NLSW sample was in- prospects (as measured by earnings,educa- terviewed annually from 1968 through 1973, tional attainment,and employment)for mar- and then in1975,1977,1978,1980,1982, riage formation will have grown between the and 1983.3 The 16 years of data analyzed in early and late baby-boom cohorts,with these this project for each cohort reflect first mar- effects becoming increasingly positive over riages formed by primary respondents be- time.The combination of improvement in tween the ages of 17 and 34.For the Young women's economic position with some de- Men's (NLSM)and Young Women's cline in men's economic position suggests (NLSW)samples,date of first marriage was that men are decreasingly expected to be the sole breadwinner for their families after mar- riage.I therefore expect that male economic 3 Although the Youth and Young Women's samples were also interviewed in subsequent prospects will have become less important years,these data are not used in the current over time for marriage.Finally,I expect that analysis.The Young Men's sample was not change in the economic basis of marriage reinterviewed after 1981. This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
136 AMERICAN SOCIOLOGICAL REVIEW examines racial differences in the nature of historical change in the relationship between the two, previous research does suggest that earnings may matter somewhat more for marriage among black women than among white women (Oppenheimer and Lewin 1999). Despite substantial reason to expect historical change in the socioeconomic context of marriage, few recent empirical studies have directly investigated whether the nature and strength of the relationship between labor market position and marriage has shifted over time. Differences in sample definitions, construction of key economic variables, and analytical approaches limit the usefulness of a meta-review of prior studies for understanding change in marriage. Although two recent studies of cross-sectional variation in marriage suggest a growing propensity to marry among well-educated women relative to their less educated peers (Goldstein and Kenney 2001; Qian and Preston 1993), no research to date offers a formal test of cohort change in the relationship of economic prospects to marriage formation. I use multiple sources of longitudinal data to systematically examine the nature and direction of cohort change in the relationship of economic prospects to the formation of first marriages. This improves the ability to examine the roots and meanings of contemporary patterns of marriage and provides insights into the likely direction of continued change in family life. I test several hypotheses about how the relationship between marriage and economic prospects may have changed between the early and late babyboom cohorts examined. First, I hypothesize that the importance of women's economic prospects (as measured by earnings, educational attainment, and employment) for marriage formation will have grown between the early and late baby-boom cohorts, with these effects becoming increasingly positive over time. The combination of improvement in women's economic position with some decline in men's economic position suggests that men are decreasingly expected to be the sole breadwinner for their families after marriage. I therefore expect that male economic prospects will have become less important over time for marriage. Finally, I expect that change in the economic basis of marriage may differ by race. The direction of these differences is uncertain, however, because of offsetting effects in the economic and attitudinal contexts of marriage. DATA Data for the present study come from three sources: the Young Men (NLSM), Young Women (NLSW), and Youth (NLSY) cohorts of the National Longitudinal Surveys of Labor Market Experience. The Young Men's (NLSM) and Young Women's (NLSW) samples are used to investigate marriage among members of the early baby-boom cohort (born from 1950 to 1954), while data from the Youth sample (NLSY) are used to investigate marriage among members of the late baby-boom cohort (born from 1961 to 1965). The early baby-boom cohort (NLSM and NLSW samples) largely includes marriage experiences from the late 1960s through the 1970s, while the late baby-boom cohort (NLSY sample) largely includes marriage experiences during the 1980s and early 1990s. The primary focus of these NLS cohorts was to examine the labor market experiences of young adults (Center for Human Resource Research 1997), making these data ideal for my investigation. Further, all three survey groups include oversamples of blacks, which facilitates the investigation of racial differences in the changing relationship between economic prospects and marriage. The NLSY sample was interviewed annually from 1979 through 1994, the NLSM sample was interviewed annually from 1966 to 1971, and then in 1973, 1975, 1976, 1978, 1980, and 1981. The NLSW sample was interviewed annually from 1968 through 1973, and then in 1975, 1977, 1978, 1980, 1982, and 1983.3 The 16 years of data analyzed in this project for each cohort reflect first marriages formed by primary respondents between the ages of 17 and 34. For the Young Men's (NLSM) and Young Women's (NLSW) samples, date of first marriage was 3 Although the Youth and Young Women's samples were also interviewed in subsequent years, these data are not used in the current analysis. The Young Men's sample was not reinterviewed after 1981. This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 137 not asked in every interview year.Thus the 1980s that are available from other surveys analytical sample for the early cohort is lim- were primarily obtained retrospectively ited to men who responded in 1976,and to (Smock 2000),although little is known about women who responded in either 1978 or the quality of retrospective reporting on co- 1983,when more complete marital histories habitation.The lack of information on co- were collected.I retain the cross-sectional habitation is a drawback of using the NLS sample and the oversample of blacks in all data,but the generally brief duration of co- surveys.Because of data limitations,the habitation for most couples should lead to samples are restricted to blacks and whites, relatively small differences in estimations of with Hispanics coded by race rather than the timing of first marriages versus first ethnicity.4 To minimize problems from left- unions of any type.Indeed,fully half of co- censoring of important economic covariates habitations end within 16 months,and only associated with marriage formation,I further one in ten couples is still cohabiting (and un- limit the analysis to individuals under age 19 married)after five years (Bumpass and at first interview.The analytical samples in- Sweet 1989).It is important to keep in mind, clude a total of 3,631 women and men in the however,that the present study sheds light early cohort and 4,920 women and men in on just one piece of the broader union forma- the late cohort.All descriptive statistics are tion process-entry into a legal marriage. weighted to adjust for design issues,such as the oversampling of blacks,and for differ- VARIABLES AND METHODS ential nonresponse to the interviews(Center for Human Resource Research 1997). I use logistic regression analysis to estimate It is desirable to have information col- discrete-time hazard models of the effects of lected at as many time points as possible for economic prospects on entry into first mar- an event-history analysis of marriage.While riage.This approach permits the estimation the Youth sample (NLSY)was interviewed of effects of fixed and time-varying annually from 1979 through 1994,the Young covariates on entry into marriage,and avoids Men's (NLSM)and Young Women's the assumption of proportional hazards (NLSW)samples were interviewed some- (Allison 1995).The dependent variable in what less regularly.Wherever possible,the the analysis is a dichotomous indicator of annual measures are constructed from retro- whether a marriage occurred in the interval spective questions about activities during the between two given years,with time-varying noninterview year.This approximation is independent variables fixed at the beginning generally feasible for measures of school en- of the interval.Data are organized into per- rollment,for example.In other cases,where son-year records,with one record for each these retrospective questions were not asked, annual interval in which respondents were at missing data are generally imputed from the risk of first marriage,including intervals in previous interview year.5 A variable indicat- which a first marriage occurred.For this ing a noninterview year for the early baby- analysis,the risk of marriage is assumed to boom cohort (NLSM and NLSW)is in- begin at age 17.Because the focus of this cluded in all models. analysis is on the comparison of coefficients Unfortunately,it is not possible to exam- between cohort models,the models them- ine cohabitation using the NLS data because, selves are kept relatively simple,with few like most surveys conducted prior to the mid- complex variable or model specifications. 1980s,the relevant questions were not asked Table 1 shows the mean values of the in- consistently.Data on cohabitation prior to the dependent variables,separately by race and sex.This analysis uses three well-estab- lished measures of standing in the labor mar- 4 It is not possible to identify Hispanic eth- ket to reflect economic prospects:earnings, nicity in the NLSM sample,and ethnicity is not educational attainment,and employment sta- ascertained until the 1993 interview in the NLSW sample. tus.Both a continuous measure of logged 5A variety of approaches to dealing with miss- earnings (wage and business income)and a ing data were considered,with substantive results dummy variable for zero earnings in the robust to choice of method prior year are constructed,based on the as- This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 137 not asked in every interview year. Thus the analytical sample for the early cohort is limited to men who responded in 1976, and to women who responded in either 1978 or 1983, when more complete marital histories were collected. I retain the cross-sectional sample and the oversample of blacks in all surveys. Because of data limitations, the samples are restricted to blacks and whites, with Hispanics coded by race rather than ethnicity.4 To minimize problems from leftcensoring of important economic covariates associated with marriage formation, I further limit the analysis to individuals under age 19 at first interview. The analytical samples include a total of 3,631 women and men in the early cohort and 4,920 women and men in the late cohort. All descriptive statistics are weighted to adjust for design issues, such as the oversampling of blacks, and for differential nonresponse to the interviews (Center for Human Resource Research 1997). It is desirable to have information collected at as many time points as possible for an event-history analysis of marriage. While the Youth sample (NLSY) was interviewed annually from 1979 through 1994, the Young Men's (NLSM) and Young Women's (NLSW) samples were interviewed somewhat less regularly. Wherever possible, the annual measures are constructed from retrospective questions about activities during the noninterview year. This approximation is generally feasible for measures of school enrollment, for example. In other cases, where these retrospective questions were not asked, missing data are generally imputed from the previous interview year.5 A variable indicating a noninterview year for the early babyboom cohort (NLSM and NLSW) is included in all models. Unfortunately, it is not possible to examine cohabitation using the NLS data because, like most surveys conducted prior to the mid- 1980s, the relevant questions were not asked consistently. Data on cohabitation prior to the 1980s that are available from other surveys were primarily obtained retrospectively (Smock 2000), although little is known about the quality of retrospective reporting on cohabitation. The lack of information on cohabitation is a drawback of using the NLS data, but the generally brief duration of cohabitation for most couples should lead to relatively small differences in estimations of the timing of first marriages versus first unions of any type. Indeed, fully half of cohabitations end within 16 months, and only one in ten couples is still cohabiting (and unmarried) after five years (Bumpass and Sweet 1989). It is important to keep in mind, however, that the present study sheds light on just one piece of the broader union formation process-entry into a legal marriage. VARIABLES AND METHODS I use logistic regression analysis to estimate discrete-time hazard models of the effects of economic prospects on entry into first marriage. This approach permits the estimation of effects of fixed and time-varying covariates on entry into marriage, and avoids the assumption of proportional hazards (Allison 1995). The dependent variable in the analysis is a dichotomous indicator of whether a marriage occurred in the interval between two given years, with time-varying independent variables fixed at the beginning of the interval. Data are organized into person-year records, with one record for each annual interval in which respondents were at risk of first marriage, including intervals in which a first marriage occurred. For this analysis, the risk of marriage is assumed to begin at age 17. Because the focus of this analysis is on the comparison of coefficients between cohort models, the models themselves are kept relatively simple, with few complex variable or model specifications. Table 1 shows the mean values of the independent variables, separately by race and sex. This analysis uses three well-established measures of standing in the labor market to reflect economic prospects: earnings, educational attainment, and employment status. Both a continuous measure of logged earnings (wage and business income) and a dummy variable for zero earnings in the prior year are constructed, based on the asI It is not possible to identify Hispanic ethnicity in the NLSM sample, and ethnicity is not ascertained until the 1993 interview in the NLSW sample. 5 A variety of approaches to dealing with missing data were considered, with substantive results robust to choice of method. This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
138 AMERICAN SOCIOLOGICAL REVIEW Table 1.Sample Means by Race and Sex for Variables Used in Analysis of Transition to First Marriage:Men and Women,Age 22,from the National Longitudinal Surveys of Young Women(NLSW),Young Men (NLSM),and Youth (NLSY) White Women Black Women White Men Black Men Early Late Early Late Early Late Early Late Independent Variable Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Earnings (log) .74 .75 .71 .72 .80 .79 .79 .75 No income .16 .11 .26 28 .03 .05 .06 .11 Educational attainment: Less than 12 years .10 .08 32 .17 .10 .13 .28 .22 12 years .48 .45 .43 38 .45 ,49 13 to 15 years .29 28 .19 .28 .17 .22 16 or more years .12 16 03 05 ,12 .12 .03 .04 Currently enrolled in ,13 19 2 .2 .11 .13 school Currently employed .65 .70 72 .69 .65 .55 Service in military .12 07 .13 .10 Family Background Variablesa Two-parent family at age 14 87 .80 .60 .86 .79 .63 .50 Family head's job is 29 .29 04 7 26 .28 04 .06 professional or managerial Mother's education: Less than 12 years .31 .25 .66 .49 .36 22 68 .45 12 years .45 .48 .14 31 .43 .51 .11 33 13 or more years .20 .24 .05 .12 .16 .22 04 .14 Residence Lives in South .27 .29 .55 .57 .24 .27 .51 .49 Lives in SMSA .70 .69 .73 .73 .61 .63 51 .66 Number of cases 1,203 1.474 517 877 1,440 1,608 471961 Note:Sample means are weighted a Measured at the time of the first interview. sumption that having some earnings is quali- from time spent in school,I constructed an tatively different from having no earnings. additional measure indicating current school Consistent with other national data,the enrollment.A dummy variable indicates em- NLS cohorts show some growth over time in ployment at the time of interview.For men, women's earnings and some decline in men's an additional measure indicated whether re- earnings,particularly among black men. spondents were on active duty in the military Consistent with national trends in education, (these men are coded 0 on the employment some decline over time is seen in the com- variable),as military service has been found pleted education of young white men,likely to impact marital transitions in prior re- reflecting the opportunity to postpone service search.Given the timing of the Vietnam War, in the Vietnam War by extending one's time it is not surprising that substantially more in school (Bernhardt et al.1999:Mare 1995) early-cohort than late-cohort men were en- and the related higher rates of military par- listed in the military at age 22. ticipation among less educated men.To dis- As prior research has found that family tinguish effects of accumulated education background characteristics are related to This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
138 AMERICAN SOCIOLOGICAL REVIEW Table 1. Sample Means by Race and Sex for Variables Used in Analysis of Transition to First Marriage: Men and Women, Age 22, from the National Longitudinal Surveys of Young Women (NLSW), Young Men (NLSM), and Youth (NLSY) White Women Black Women White Men Black Men Early Late Early Late Early Late Early Late Independent Variable Cohort Cohort Cohort Cohort Cohort Cohort Cohort Cohort Earnings (log) .74 .75 .71 .72 .80 .79 .79 .75 No income .16 .11 .26 .28 .03 .05 .06 .11 Educational attainment: Less than 12 years .10 .08 .32 .17 .10 .13 .28 .22 12 years .48 .45 .43 .47 .38 .45 .42 .49 13 to 15 years .29 .28 .19 .30 .31 .28 .17 .22 16 or more years .12 .16 .03 .05 .12 .12 .03 .04 Currently enrolled in .13 .19 .12 .16 .20 .21 .11 .13 school Currently employed .65 .70 .52 .51 .72 .69 .65 .55 Service in military .12 .07 .13 .10 Family Background Variables a Two-parent family at age 14 .87 .80 .60 .45 .86 .79 .63 .50 Family head's job is .29 .29 .04 .07 .26 .28 .04 .06 professional or managerial Mother's education: Less than 12 years .31 .25 .66 .49 .36 .22 .68 .45 12 years .45 .48 .14 .31 .43 .51 .11 .33 13 or more years .20 .24 .05 .12 .16 .22 .04 .14 Residence Lives in South .27 .29 .55 .57 .24 .27 .51 .49 Lives in SMSA .70 .69 .73 .73 .61 .63 .51 .66 Number of cases 1,203 1,474 517 877 1,440 1,608 471 961 Note: Sample means are weighted. a Measured at the time of the first interview. sumption that having some earnings is qualitatively different from having no earnings. Consistent with other national data, the NLS cohorts show some growth over time in women's earnings and some decline in men's earnings, particularly among black men. Consistent with national trends in education, some decline over time is seen in the completed education of young white men, likely reflecting the opportunity to postpone service in the Vietnam War by extending one's time in school (Bernhardt et al. 1999; Mare 1995) and the related higher rates of military participation among less educated men. To distinguish effects of accumulated education from time spent in school, I constructed an additional measure indicating current school enrollment. A dummy variable indicates employment at the time of interview. For men, an additional measure indicated whether respondents were on active duty in the military (these men are coded 0 on the employment variable), as military service has been found to impact marital transitions in prior research. Given the timing of the Vietnam War, it is not surprising that substantially more early-cohort than late-cohort men were enlisted in the military at age 22. As prior research has found that family background characteristics are related to This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 139 marriage formation (e.g.,Michael and Tuma role of student with the responsibilities as- 1985),several family background variables sociated with marriage and a family are included in the analysis,all measured at (Thornton et al.1995).No statistically sig- the time of first interview.These include nificant effects are found for family back- dummy variables indicating whether the re- ground variables in the early cohort,but a spondent was living in a two-parent family positive effect on marriage is associated at age 14,whether the respondent's father with living in the South or outside of an (or head of household)was employed in a SMSA. managerial or professional occupation,as The focus of the analysis lies in identify- well as a categorical measure of the ing how these effects have changed over respondent's mother's educational attain- time.Compare the first column in Table 2 ment.Consistent with national trends,data with the second column,which shows a par- from the National Longitudinal Surveys in- allel model for white women in the late dicate decline over time in the proportion of baby-boom cohort.Although the pattern of young people growing up in two-parent effects in the two cohorts of white women is families and some increase in the level of generally similar,the effect of women's mothers'educational attainment.Finally,as earnings on marriage formation increases various local-area characteristics have been over time as predicted.The level of women's found to be related to marriage in prior re- earnings was not significantly related to search (e.g.,Lichter et al.1992),dummy marriage in the early baby-boom cohort of variables are included in the models for cur- white women,yet significantly stronger ef- rent residence in the South and in a Standard fects are observed among white women in Metropolitan Statistical Area (SMSA).Mean the later cohort.Indeed,each unit increase values are assigned in the case of missing in logged earnings is associated with a 21- data on earnings,and separate indicators of percent increase in the odds of marriage for missing data on this and other variables are the later cohort of women,while having no included in all models,as appropriate (see earnings is associated with a 32-percent re- note 5).With the exception of family back- duction in the odds of marriage. ground variables,all explanatory measures To determine if change in marriage has are time-varying proceeded in a similar fashion by race,the last two columns of Table 2 estimate a par- RESULTS allel set of models for black women.Despite important racial differences in the economic The first stage of the analysis investigates and attitudinal contexts of marriage,patterns change over time in the relationship be- of change in marriage are similar by race.As tween economic prospects and marriage for- observed for white women,only the effect mation among women.Beginning with mar- of earnings has changed significantly over riage among white women of the early time,with growth in the importance of earn- baby-boom cohort,shown in the first col- ings for marriage formation observed among umn of Table 2,results indicate that al- black women.Yet several significant racial though high earnings are associated with an differences are noted in the general pattern increase in the odds of marriage,this effect of women's marriage.For example,having is not statistically significant.With respect accumulated fewer than 12 years of school- to education,having at least 16 years of ing (relative to having exactly 12 years)is schooling is associated with 52 percent associated with increased marriage among greater odds (exp[.42])of marriage than white women in both cohorts,but has no sig- having 12 years of schooling.Consistent nificant effects on marriage among black with previous research,current school en- women.6 In explaining a similar finding in rollment is significantly associated with de- lay in marriage among white women of the 6 Although the size of the effect of having 16 early baby-boom cohort-student status is or more years of education is smaller for the late associated with a 58-percent reduction in the cohort of black women than for the early cohort, odds of marriage.This is not surprising the difference between these coefficients is not given the potential incompatibility of the statistically different from zero. This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 139 marriage formation (e.g., Michael and Tuma 1985), several family background variables are included in the analysis, all measured at the time of first interview. These include dummy variables indicating whether the respondent was living in a two-parent family at age 14, whether the respondent's father (or head of household) was employed in a managerial or professional occupation, as well as a categorical measure of the respondent's mother's educational attainment. Consistent with national trends, data from the National Longitudinal Surveys indicate decline over time in the proportion of young people growing up in two-parent families and some increase in the level of mothers' educational attainment. Finally, as various local-area characteristics have been found to be related to marriage in prior research (e.g., Lichter et al. 1992), dummy variables are included in the models for current residence in the South and in a Standard Metropolitan Statistical Area (SMSA). Mean values are assigned in the case of missing data on earnings, and separate indicators of missing data on this and other variables are included in all models, as appropriate (see note 5). With the exception of family background variables, all explanatory measures are time-varying. RESULTS The first stage of the analysis investigates change over time in the relationship between economic prospects and marriage formation among women. Beginning with marriage among white women of the early baby-boom cohort, shown in the first column of Table 2, results indicate that although high earnings are associated with an increase in the odds of marriage, this effect is not statistically significant. With respect to education, having at least 16 years of schooling is associated with 52 percent greater odds (exp[.42]) of marriage than having 12 years of schooling. Consistent with previous research, current school enrollment is significantly associated with delay in marriage among white women of the early baby-boom cohort-student status is associated with a 58-percent reduction in the odds of marriage. This is not surprising given the potential incompatibility of the role of student with the responsibilities associated with marriage and a family (Thornton et al. 1995). No statistically significant effects are found for family background variables in the early cohort, but a positive effect on marriage is associated with living in the South or outside of an SMSA. The focus of the analysis lies in identifying how these effects have changed over time. Compare the first column in Table 2 with the second column, which shows a parallel model for white women in the late baby-boom cohort. Although the pattern of effects in the two cohorts of white women is generally similar, the effect of women's earnings on marriage formation increases over time as predicted. The level of women's earnings was not significantly related to marriage in the early baby-boom cohort of white women, yet significantly stronger effects are observed among white women in the later cohort. Indeed, each unit increase in logged earnings is associated with a 21- percent increase in the odds of marriage for the later cohort of women, while having no earnings is associated with a 32-percent reduction in the odds of marriage. To determine if change in marriage has proceeded in a similar fashion by race, the last two columns of Table 2 estimate a parallel set of models for black women. Despite important racial differences in the economic and attitudinal contexts of marriage, patterns of change in marriage are similar by race. As observed for white women, only the effect of earnings has changed significantly over time, with growth in the importance of earnings for marriage formation observed among black women. Yet several significant racial differences are noted in the general pattern of women's marriage. For example, having accumulated fewer than 12 years of schooling (relative to having exactly 12 years) is associated with increased marriage among white women in both cohorts, but has no significant effects on marriage among black women.6 In explaining a similar finding in 6 Although the size of the effect of having 16 or more years of education is smaller for the late cohort of black women than for the early cohort, the difference between these coefficients is not statistically different from zero. This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
140 AMERICAN SOCIOLOGICAL REVIEW Table 2.Coefficients for Women from the Logistic Regression of Transition to First Marriage on Selected Independent Variables:National Longitudinal Surveys of Youth(NLSY)and Young Women (NLSW) White Women Black Women Early Cohort Late Cohort Early Cohort Late Cohort Independent Variable Coef.S.E. Coef.S.E. Coef.S.E. Coef.S.E. Earnings(log) .07 (.04) .19*t (.04) .01 (.06) .22+(.05) No earnings -.25 (.13) -.38 (.13) -.19 (.17) -.59 (.15) Educational Attainment Less than 12 years .24 (.12) 36 (.11) -.01 (.15) -.04 (.15) 13 to 15 years .41* (.11) .23 (.10) -.27+ (.23) 03 (.13) 16 or more years 42 (.16) ,41 (,11) .85 (.26) 37 (.19) Currently enrolled in school -.87 (.10) -.71° (.09) -.25# (.16) -09t (.14) Currently employed 03 (.09) -.01 (.09) .14 (.16) .08 (.13) Family Background Variables Two-parent family at -.16 (.11) 02 (.08) 22* (.14) .14 (.10) age 14 Family head's job is -11 (.09) -.06 (.08) -09 (.36) .07 (.21) professional/managerial Mother's education: 12 years -.06 (.09) .00 (.08) -.12 (.19) -.06 (.11) 13 or more years -.21 (.12) -21 (.10) -.25 (.31) -31 (.18) Residence Lives in South .27* (.08) 41 (.07) .07 (.14) .35 (.11) Lives in SMSA -.26 (.08) -.22 (.08) -.03 (.15) -.15 (.12) Constant -3.87* (.80) -4.14(.51) -3.71*(.89) 4.66 (.58) Log-likelihood -2,673 -3.446 -1.049 -1,720 Number of person years 6,794 10,185 4,122 8.230 Nore:Standard errors are in parentheses.For mother's education the omitted category is"less than high school";for respondent's education it is"12 years."Models also contain controls for age and indicators for missing data. Significantly different from the early cohort coefficent at p<.05(two-tailed tests) Significantly different from the white(within-cohort)coefficient at p<.05(two-tailed tests) "p<.05 (two-tailed tests) their own research,Oppenheimer and Lewin the early cohort,although this effect itself is (1999)suggest that the least educated white not statistically significant even among women may prefer to marry at an earlier age, black women. but are hesitant to do so given their poor In the next stage of the analysis,I investi- economic position.School enrollment also gate whether economic prospects have be- has significantly stronger marriage-delaying come less important over time for marriage effects for white women than for black formation among men.The first column of women in both cohorts.Finally,living in a Table 3 displays results for white men in the two-parent family at age 14 has a stronger early baby-boom cohort.Early cohort white positive effect on marriage formation among men's earnings display the expected positive black women than among white women in relationship to marriage formation,with This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
140 AMERICAN SOCIOLOGICAL REVIEW Table 2. Coefficients for Women from the Logistic Regression of Transition to First Marriage on Selected Independent Variables: National Longitudinal Surveys of Youth (NLSY) and Young Women (NLSW) White Women Black Women Early Cohort Late Cohort Early Cohort Late Cohort Independent Variable Coef. S.E. Coef. S.E. Coef. S.E. Coef. S.E. Earnings (log) .07 (.04) 19*t (.04) .01 (.06) .22*t (.05) No earnings -.25 (.13) -.38* (.13) -.19 (.17) -.59* (.15) Educational Attainment Less than 12 years .24* (.12) .36* (.11) -.01 (.15) -.04t (.15) 13 to 15 years .41* (.11) .23* (.10) -.27t (.23) .03 (.13) 16 or more years .42* (.16) .41* (.11) .85* (.26) .37* (.19) Currently enrolled in school -.87* (.10) -.71* (.09) -.25t (.16) -.09t (.14) Currently employed .03 (.09) -.01 (.09) .14 (.16) .08 (.13) Family Background Variables Two-parent family at -.16 (.11) .02 (.08) .22t (.14) .14 (.10) age 14 Family head's job is -.11 (.09) -.06 (.08) -.09 (.36) .07 (.21) professional/managerial Mother's education: 12 years -.06 (.09) .00 (.08) -.12 (.19) -.06 (.11) 13 or more years -.21 (.12) -.2 * (.10) -.25 (.31) -.31 (.18) Residence Lives in South .27* (.08) .41* (.07) .07 (.14) .35* (.11) Lives in SMSA -.26* (.08) -.22* (.08) -.03 (.15) -.15 (.12) Constant -3.87* (.80) -4.14* (.51) -3.71* (.89) -4.66* (.58) Log-likelihood -2,673 -3,446 -1,049 -1,720 Number of person years 6,794 10,185 4,122 8,230 Note: Standard errors are in parentheses. For mother's education the omitted category is "less than high school"; for respondent's education it is "12 years." Models also contain controls for age and indicators for missing data. Significantly different from the early cohort coefficent at p < .05 (two-tailed tests) Significantly different from the white (within-cohort) coefficient at p < .05 (two-tailed tests) *p < .05 (two-tailed tests) their own research, Oppenheimer and Lewin (1999) suggest that the least educated white women may prefer to marry at an earlier age, but are hesitant to do so given their poor economic position. School enrollment also has significantly stronger marriage-delaying effects for white women than for black women in both cohorts. Finally, living in a two-parent family at age 14 has a stronger positive effect on marriage formation among black women than among white women in the early cohort, although this effect itself is not statistically significant even among black women. In the next stage of the analysis, I investigate whether economic prospects have become less important over time for marriage formation among men. The first column of Table 3 displays results for white men in the early baby-boom cohort. Early cohort white men's earnings display the expected positive relationship to marriage formation, with This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 141 Table 3.Coefficients for Men from the Logistic Regression of Transition to First Marriage on Selected Independent Variables:National Longitudinal Surveys of Youth(NLSY)and Young Men (NLSM) White Men Black Men Early Cohort Late Cohort Early Cohort Late Cohort Independent Variable Coef.S.E. Coef.S.E. Coef.S.E. Coef.S.E. Earnings(log) .27 (.04) .26 (.04) 29 (.08) .26 (.06) No earnings -61* (.21) -74 (.20) -.85 (.35) -.41 (.21) Educational Attainment Less than 12 years .05 (.11) 28* (.10) -.27 (.15) -20￥ (.15) 13 to 15 years -.03 (.09) .26*+(.10) .17 (.20) .20 (14) 16 or more years -.03 (.12) .42+(.11) .37 (.34) .74 (.19) Currently enrolled in school -21" (.09) -.68*+(.11) -.40° (.19) -.40 (.19) Currently employed 44* (.10) .14t (.10) .14 (.19) .43 (.15) Service in military 39 (.42) .63 (.22) -1.06 (.63) .85+(32) Family Background Variables Two-parent family at -05 (.10) .08 (.08) .12 (.14) -04 (.10) age 14 Family head's job is -13 (.08) .04 (.08) -.02 (.46) -.33 (.24) professional/managerial Mother's education: 12 years -.06 (.07) -.13 (.08) -.08 (.20) 00 (.12) 13 or more years -25 (.11) -.31 (.11) -.73 (.43) .13t (.17) Residence Lives in South .18 (.08) 33 (.07) 39 (.17) .22 (.11) Lives in SMSA -.28* (.08) -.36 (.08) -,17 (.15) .08t (.14) Constant -4.50 (.55) -4.74 (.47) 6.18 (.12) -6.07 (.69) Log-likelihood -3.413 -3.483 -1,023 -1.619 Number of person years 10.057 13.653 3,717 9,780 Note:Standard errors are in parentheses.For mother's education the omitted category is "less than high school";for respondent's education it is"12 years."Models also contain controls for age and indicators for missing data. t Significantly different from Early Cohort coefficent at p<05(two-tailed test). Significantly different from white(within-cohort)coefficient at p<.05(two-tailed test). 'p<.05(two-tailed test) each additional unit of logged income asso- tistically significant effect of military ser- ciated with a 31-percent increase in the odds vice is found on marriage among early co- of marriage;having no income is associated hort white men,while having a college-edu- with 46 percent lower odds of marriage.Al- cated mother is associated with delayed mar- though no significant effects are found for riage.Finally,similar to the findings for early cohort white men's accumulated edu- women,residence in the South and living cation,school enrollment is associated with outside of an SMSA are associated with rela- a 19-percent decrease in the odds of men's tively earlier transitions into first marriage marriage,and employment is associated with among white men in the early baby-boom 55 percent greater odds of marriage.No sta- cohort. This content downloaded from 128.103.149.52 on Mon,12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions
TWO DECADES OF FAMILY CHANGE 141 Table 3. Coefficients for Men from the Logistic Regression of Transition to First Marriage on Selected Independent Variables: National Longitudinal Surveys of Youth (NLSY) and Young Men (NLSM) White Men Black Men Early Cohort Late Cohort Early Cohort Late Cohort Independent Variable Coef. S.E. Coef. S.E. Coef. S.E. Coef. S.E. Earnings (log) .27* (.04) .26* (.04) .29* (.08) .26* (.06) No earnings -.61 * (.21) -.74* (.20) -.85* (.35) -.41 (.21) Educational Attainment Less than 12 years .05 (.11) .28* (.10) -.27 (.15) -.20t (.15) 13 to 15 years -.03 (.09) .26*t (.10) .17 (.20) .20 (.14) 16 or more years -.03 (.12) .42*t (.11) .37 (.34) .74* (.19) Currently enrolled in school -.21* (.09) -.68*t (.11) -.40* (.19) _.40* (.19) Currently employed .44* (.10) .14t (.10) .14 (.19) .43* (.15) Service in military .39 (.42) .63* (.22) -1.06 (.63) 85*t (.32) Family Background Variables Two-parent family at -.05 (.10) .08 (.08) .12 (.14) -.04 (.10) age 14 Family head's job is -.13 (.08) .04 (.08) -.02 (.46) -.33 (.24) professional/managerial Mother's education: 12 years -.06 (.07) -.13 (.08) -.08 (.20) .00 (.12) 13 or more years -.25* (.11) -.31* (.11) -.73 (.43) .13t (.17) Residence Lives in South .18* (.08) .33* (.07) .39* (.17) .22* (.11) Lives in SMSA -.28* (.08) -.36* (.08) -.17 (.15) .08t (.14) Constant -4.50* (.55) -4.74* (.47) -6.18* (.12) -6.07* (.69) Log-likelihood -3,413 -3,483 -1,023 -1,619 Number of person years 10,057 13,653 3,717 9,780 Note: Standard errors are in parentheses. For mother's education the omitted category is "less than high school"; for respondent's education it is "12 years." Models also contain controls for age and indicators for missing data. t Significantly different from Early Cohort coefficent at p < .05 (two-tailed test). t Significantly different from white (within-cohort) coefficient at p < .05 (two-tailed test). *p < .05 (two-tailed test) each additional unit of logged income associated with a 31-percent increase in the odds of marriage; having no income is associated with 46 percent lower odds of marriage. Although no significant effects are found for early cohort white men's accumulated education, school enrollment is associated with a 19-percent decrease in the odds of men's marriage, and employment is associated with 55 percent greater odds of marriage. No statistically significant effect of military service is found on marriage among early cohort white men, while having a college-educated mother is associated with delayed marriage. Finally, similar to the findings for women, residence in the South and living outside of an SMSA are associated with relatively earlier transitions into first marriage among white men in the early baby-boom cohort. This content downloaded from 128.103.149.52 on Mon, 12 Aug 2013 19:18:11 PM All use subject to JSTOR Terms and Conditions