Debt, Cohabitation, and Marriage in Young Adulthood

Abstract

Despite growing evidence that debt influences pivotal life events in early and young adulthood, the role of debt in the familial lives of young adults has received relatively little attention. Using data from the NLSY 1997 cohort (N = 6,749) and a discrete-time competing risks hazard model framework, I test whether the transition to first union is influenced by a young adult’s credit card and education loan debt above and beyond traditional educational and labor market characteristics. I find that credit card debt is positively associated with cohabitation for men and women, and that women with education loan debt are more likely than women without such debt to delay marriage and transition into cohabitation. Single life may be difficult to afford, but marital life is unaffordable as well. Cohabitation presents an alternative to single life, but not necessarily a marital substitute for these young adults.

Introduction

Young adults increasingly delay marriage. American men’s median age at first marriage rose from 24.7 in 1980 to 28.2 in 2010. It increased even more among women, rising from age 22.0 to 26.1 (U.S. Census Bureau 2011), while the overall share of young adults married by age 30 declined (Cohn et al. 2011). Delays in labor market entrance and barriers to advancement, expensive housing, and increasing levels of debt are often attributed with contributing to marital delay (Furstenberg et al. 2004; Settersten and Ray 2010). At the same time, cohabitation rates have increased for young adults, with those in their late 20s having the highest cohabitation rates (Kennedy and Bumpass 2008). Numerous studies point to the importance of economic well-being and financial stability as predictors of marriage, particularly for men (Sassler and Goldscheider 2004; Sweeney 2002; Xie et al. 2003). Economic instability and a lack of economic resources matter less for cohabitation (Clarkberg 1999; Xie et al. 2003). Less attention has been paid to understanding whether the financial foundations required of cohabitation differ from those of marriage, how debt shapes union formation decisions, and whether these associations vary by sex.

Youth born in the early 1980s came of age during a period of expansive credit markets. Between 1992 and 2001, the percentage of young adults aged 25–34 who had credit card debt remained consistent, while the average debt holdings increased by 55 % (Draut and Silva 2004). It was also a period of increased college enrollment and dramatic changes in the financing of college, notably the decline in the purchasing power of federal grant aid and growth in the availability of student loan programs (Fitzpatrick and Turner 2007; Rothstein and Rouse 2011). As of 2003–2004, one-third of undergraduates borrowed federal loans up from a quarter of a decade before, and this was accompanied by a 26 % increase in the average loan amount (Wei and Berkner 2008). Understanding its role in the familial lives of young adults is important because of growing evidence that debt influences other pivotal life events in early and young adulthood, such as college completion (Dwyer et al. 2012), first career choice (Rothstein and Rouse 2011), and starting salary (Minicozzi 2005). Despite growing convergence in men’s and women’s roles, it remains unclear whether the effect of debt on union formation differs by gender. This omission is surprising, given that women pursue and obtain college degrees at higher rates than men (Buchmann and DiPrete 2006).

Building on recent work examining consumer debt and marriage in young adulthood (Chiteji 2007; Dew and Price 2011), this study tests whether individual debt has an independent and significant impact on transitions into marriage versus cohabitation, above and beyond traditional socioeconomic indicators. Two distinct types of credit obligations are analyzed. Credit card debt is the most common form of unsecured debt for young adults. Among those with education debt, however, student loans tend to account for the largest share of their debt portfolio. Data are from the 1997 National Longitudinal Study of Youth (NLSY97), a cohort of young adults born in the early 1980s. Discrete-time competing risk models test whether similar consumption, economic, and social factors predicting transitions into marriage also explain entrance into cohabiting unions and whether these relationships differ by sex.

Growth in Credit Card Debt and Education Loan Debt in Young Adulthood

Americans experienced 30 years of unprecedented availability and access to both unsecured and secured credit markets between the mid-1980s through the Great Recession (2007–2009) (Durkin 2000; Dynan and Kohn 2007; Lyons 2003). The rise in credit card debt levels was largely driven by financial deregulation changes in consumer credit (Watkins 2000) and technological changes that allowed companies’ to diversify risks across households and offer more attractive products (Johnson 2005; Watkins 2000). These policy changes and financial innovations increased the debt of existing customers and those able to gain access, like young adults (Dynan 2009; Weller 2010). Between 1992 and 2001, the average credit card debt of 18- to 24-year-olds increased 104 %, rising from $1,461 to $2,985, compared with an increase of 38 % for all households (Draut and Silva 2004). Jiang and Dunn (2013) calculated that the average credit card debt of young adults born between 1980 and 1984 exceeded that of their parent’s generation by $5,689, and of their grandparents’ generation by $8,156. Scholars point to the slowdown in real wage growth, which has not kept pace with the rate of inflation, as a cause for rising debt (Draut and Silva 2004). The inability to repay debt also increases the likelihood of rolling over debt with compounding interest into future periods.

Postsecondary schooling would be unattainable for many without receiving some form of financial aid or grant assistance (Bound et al. 2007; Fitzpatrick and Turner 2007; Kane 1996). The majority of financial assistance programs are loan-based, despite various funding options available for low-income students (e.g., Pell grants, student loans) and tax incentives for those students coming from middle- to high-income households (e.g. tuition tax credits, 529 (college savings) plans). Loans have replaced much of the grant aid offered throughout the middle- to late-twentieth century (Fitzpatrick and Turner 2007). According to the National Center for Education Statistics (NCES 2009), 34 % of undergraduates held federal loans in 2007, compared with 27 % who received Pell grants ($2,600). The average college graduate left school with approximately $23,000 of college loan debt in 2008; in 1996, the average debt was $17,000 (Hinze-Pifer and Fry 2010). The replacement of grant assistance with student loan financial aid means that more young adults accrue a significant amount of debt, which can take years to pay down (King and Bannon 2002). Financial aid is oftentimes not enough to cover total college expenses, and a majority of college students have had to rely on credit cards to pay for additional costs and fees, such as room and board, books, and health insurance, which is usually required for matriculation (Lyons 2008; Draut and Silva 2004). As of 2008, only 2 % of undergraduates had no credit history, one-half held at least four credit cards (Sallie Mae 2009), and one in four students reported using credit cards to finance their education (Draut and Silva 2004).

Access to and use of credit cards is not limited to those attending postsecondary school, yet most empirical studies on debt behavior in young adulthood focus on graduates of four-year institutions. Fewer than 6 in 10 students who started a four-year degree in 2001 completed college in six years, and just 27.5 % of two-year program students completed their associate’s degree within three years (NCES 2012). Only 39.6 % of 18- to 24-year-olds were enrolled in degree-granting institutions in 2008 (NCES 2012), leaving a large proportion of the young adult population understudied.

The Role of Economic Resources in Cohabitation and Marriage in Young Adulthood

Early theories of marital formation argued that it was men’s economic position and labor market returns that determined their attractiveness in the marriage market, whereas those same attributes negatively predicted marriage for women (Becker 1981). Oppenheimer proposed that as women acquired skills similar to men’s, standards for a spouse would become more complementary, advantages in household production would be less valued, and socioeconomic achievements would be more desirable in the marriage market. Educational attainment and stable employment, for example, have been consistently positive predictors of marriage for men (Cooney and Hogan 1991; Goldstein and Kenney 2001; Oppenheimer et al. 1997; Sassler and Goldscheider 2004) and, more recently, for women (Qian and Preston 1993; Sassler and Schoen 1999; Sweeney 2002). Recent studies focusing on wealth (Schneider 2011) found that both financial assets and vehicle ownership increased the probability of a first marriage in a given year, especially for men.

Cohabitation, however, has become the modal relationship form for young adults (Amato et al. 2008; Sassler 2010), and scholars have increasingly focused their attention on the factors associated with first union transition: the entrance into either cohabitation or marriage. They have argued that the lack of enforceable trust that comes with a formal commitment keeps cohabitors from acquiring the social status and benefits of marriage (Cherlin 2004). This distinction influences the economic criteria that young adults assign to entering a cohabiting versus marital arrangement (Smock et al. 2005). Evidence suggests that the prerequisites for marriage and cohabitation differ. Cohabitors are disproportionately drawn from young adults with lower levels of educational attainment and poorer economic prospects; studies based on data from the 1970s, 1980s, and early 1990s found educational attainment to be either uncorrelated or negatively associated with transitions into cohabitation for women and men (Thornton et al. 1995). Despite school enrollment deterring both marriage and cohabitation (Sassler and Goldscheider 2004; Thornton et al. 1995), college graduates are less likely to cohabit (Kennedy and Bumpass 2008) and more likely to transition directly to marriage (McLanahan and Percheski 2008; Schwartz and Mare 2005). Furthermore, labor market earnings and earnings potential were found to be either not significant or positively associated with transitions into cohabitation for women and men (Clarkberg 1999; Xie et al. 2003). Low-skilled men and low-wage workers have experienced the largest declines in the likelihood of being married (Cherlin 2004). Qualitative studies have found that cohabitors view living together as more economical than maintaining two separate residences (Sassler 2004), and nearly one-third of cohabiting young adults mentioned that finances influenced their decision to cohabit (Sassler and Miller 2011; Taylor 2010). Such findings suggest that economic underpinnings are weaker for the formation of a cohabiting union compared with a marital union and will be associated not only with who enters cohabitation but also the timing.

Young adults across social classes express similar sentiments about what must be in place in order to be ready for marriage. These prerequisites include attaining stable work, economic security, some savings, and an effort to pay down outstanding debt (Cherlin 2009; Edin and Kefalas 2005; Manning et al. 2007; Smock et al. 2005). Single people with greater wealth have the resources that allow them to achieve marriageable status in a wealth-based and asset-valued marriage market (Schneider 2011). Debt accumulation has become a part of achieving financial independence and social mobility, but the acquisition of debt in young adulthood may have contributed to delays in marriage and increases in cohabitation for recent cohorts of young adults.

Men’s economic attributes have long been considered primary in the marriage market, with women’s economic factors having weak or insignificant effects on marital transitions. In recent years, however, women have surpassed men in academic achievement and college completion (Buchmann and DiPrete 2006), and women’s earnings have become increasingly important as a predictor of marriage among recent cohorts (Sassler and Schoen 1999; Sweeney 2002). This suggests that men and women are beginning to resemble one another in the relationship between economic prospects and marriage. Whether debt will have similar effects for men and women remains an open question.

How Does Debt Operate in the Relationship Market for Young Adults?

Cohabitation is an economically attractive living arrangement because couples benefit from cost sharing and economies of scale, without the social and financial expectations of marriage. The informal versus formal distinction between cohabitation and marriage may deter cohabitors from investing in relationship-specific capital. Research on intrahousehold resource allocation finds that married couples generally pool income and manage resources jointly, but cohabitors are more likely to maintain independent money-management systems (Brines and Joyner 1999; Treas 1993; Winkler 1997). For example, cohabitors are more likely to retain separate bank accounts, which negatively influences relationship quality and relationship commitment (Addo and Sassler 2010). At the same time, living together in an informal union can allow a couple to sort out or improve their financial situation (Dew and Price 2011) while allowing each partner to become familiar with the other’s financial status.

These behavioral differences suggest how debt would impact the decision to enter into marriage versus cohabitation.¹ In a marital union, debt can be considered an individual financial burden brought into the union, which removes financial resources from the joint household. In a cohabiting union, debt remains the responsibility of the debtor, decreasing only one partner’s resources, assuming that cohabitors maintain separate financial systems. If young adults prefer to be financially established prior to marriage (Cherlin 2004), cohabitation will be more likely if debt is high, and marriage will be more likely if debt is low or nonexistent. The directionality of the association remains unchanged even when assuming debt values are revealed: marriage will be more likely when an individual has found a partner willing to assume his/her current debt. The formation of a union occurs in the presence of nonzero debt if there has been a consensus to share assets for marriage or not share assets for cohabitation. This leads to the following hypotheses:

Although this analysis does not use an exchange model explicitly, the relationship market chosen may reveal the preference of both the respondent and the partner. If a young adult chooses marriage, s/he reveals a preferred union as well as the ranking of relationship choices to be married over cohabitating and remaining single. In the current analysis, cohabitation and marriage are modeled as competing risks. A person not only chooses to enter a union but also jointly decides the type of union entered: cohabitation or marriage, as opposed to continued singlehood. Modeling the choices as separate binary outcomes might misrepresent the relationship given that the three states are correlated, interdependent events. The decision to transition into a coresidential relationship is not necessarily sequential, with the decision to form a union followed by the selection between cohabitation or marriage (Manning and Smock 2005). The three choices are separate and distinct but not substitutable events.

Variation in Union Formation by Type of Debt

There is reason to believe the type of debt held by young adults matters for their attractiveness in the relationship market. Although both credit card debt and education loans could be considered investment debts given that most young adults lack the income to acquire many of the goods they need, the structural dynamics of the two types of debt vary. Compared with average education loan debt, average rates of credit card indebtedness are low (Chiteji 2007). Outstanding credit card debt, however, often carries large penalties in the form of high interest rates (Baek and Hong 2004), and the accumulation of credit card debt is associated with negative financial practices (e.g., overspending) and poor fiscal management skills (Drentea 2000). High interest rates and short repayment periods also increase the incentives to pay down credit card debt faster. Conversely having education loan debt is more normative. The principal amounts tend to be larger and most borrowers can often choose their repayment periods, a factor contributing to a longer payoff time horizon (Avery and Turner 2012). Education loan characteristics can also vary significantly by type. Private loans tend to have higher interest rates than federal loans. With need-based subsidized loans, the federal government pays interest while enrolled, and borrowers repay their loans after college at a subsidized rate. Students can also apply for unsubsidized federal loans that allow them to borrow independent of financial need; they are, however, responsible for repaying the interest and principal amounts. At any rate, loan repayment is deferred until after college or a grace period upon withdrawal (Avery and Turner 2012). Federal and local policies can also influence individual behavior toward debt (Poterba 2001). As of this writing, interest payments on some qualified education loans are tax-deductible. And unlike credit card debt and other unsecured debts, education loans are extremely difficult to discharge through consumer bankruptcy, although there are multiple systems in place for borrowers to defer repayment (Avery and Turner 2012).

Even though outstanding credit card debt may be a signal of financial independence and access to financial resources, it can also be a marker of current and future instability. Significant credit card debt may indicate financial irresponsibility, making someone unattractive in the marriage market but not in the cohabitation market, where financial requirements are lower because of the decreased likelihood of income (or debt) pooling. Young adults holding nonzero credit card debt may fare better in the cohabitation market, where entry costs are considered lower than marriage (Sassler 2010). Young adults may choose to cohabit instead of marry as a means to cost-share. Therefore, credit card debt reduces the relative price of cohabitation by increasing the price of marriage. The ability to take on credit card debt can also help defray moving costs and pay rent, increasing the attractiveness of cohabitation relative to continued singlehood, and leading to the following hypothesis:

Education loans are considered an investment on an appreciating asset (education), which represents future earnings potential and economic stability. Youth holding nonzero education debt are potentially attractive partners in the marriage market given their expected future earnings potential; however, they are also more likely to delay marriage, prioritizing career and financial stability over marriage (Fry 2010). Additionally, the structure of postsecondary enrollment (e.g., dormitory living, delayed employment, and extended training) may act as an indirect deterrent to union formation in early and young adulthood, prolonging the search process, leading to my third hypothesis:

Evidence suggests both male and female economic resources are important for marital formation (Qian and Preston 1993; Sweeney 2002), and that a woman’s economic resources matter for both the likelihood and the timing of transition (Oppenheimer 1997). Women with greater economic resources, advanced educational attainment, and labor market rewards could subsidize their spousal search, prolonging it in order to find a better match. Women have also outpaced men in college attendance and completion (Buchmann and Diprete 2006), leading to the following hypothesis regarding gender and union transitions:

Method

Data

Data are from the NLSY97 (Bureau of Labor Statistics 2009), an annual study following a nationally representative sample of 12- to 16-year-olds living in the United States as of December 31, 1996. The NLSY97 includes extensive information about the youth labor market and educational experiences, as well as their familial and relationship backgrounds. The survey also ascertains information on wages, income, and educational debt at every survey year. After reaching their 20th and 25th birthdays, respondents were asked to complete an assets module containing questions about all financial and nonfinancial asset holdings, asset values, and outstanding debts. This study starts in the first survey wave after the respondent completed the age-20 assets module and continues through the 2009 survey year. The panel nature of the data allows one to follow the youth up to eight years after the age-20 assessment.

Two sample restrictions were imposed on the data. First, any youth who transitioned to a first cohabitation or first marriage prior to the age-20 asset module was excluded from the analysis, removing 1,095 women and 572 men.² Second, youth without a complete union history and who missed two consecutive interviews during the study period and experienced a union transition were removed, eliminating an additional 548 young adults. The final sample is more proportionally male; is more likely to come from two-parent households with parents who have, on average, one full year more of schooling than the omitted sample; and has wealthier parents. Compared with the omitted group, the final sample contains fewer Hispanics but more blacks, which is consistent with previous research indicating ethnoracial differences in the timing to first coresidential union (Addo 2012).³ Finally, although the dropped sample reported higher rates of full-time employment, they were less likely, on average, to have completed college, to hold advanced degrees, or to be currently enrolled in a postsecondary program.

Multiple imputation using the chained equations method in STATA is applied to maintain maximum sample size for those missing information on independent variables. The variables, described in detail in the next section, include residence in rural region in childhood (15.3 % missing person-years), grew up with both parents in household (10.7 %), paternal education (19.3 %), current region of residence (3.6 %), and total value of all assets (9.0 %). The estimation method works well with categorical and binary variables (White et al. 2011). The final analytic sample follows 3,025 women and 3,744 men who contributed 14,681 and 19,373 person-years, respectively, to the analysis.

Cohabitation and Marriage

The main dependent variables are union transitions. Young adults can transition from a single state into first cohabitation or first marriage. Cohabitation is defined in the NLSY97 as a sexual relationship in which a respondent resides with a person of the opposite sex for a minimum of one month. In each survey round, respondents are asked their current marital status as well as month and year of first cohabitation and first marriage.

Debt Measures

The variable capturing credit card debt is coded based on responses to the following question: “Do you have any other debts that you currently owe money on that we have not already talked about? (Examples include store bills, credit cards, loans obtained through a bank or credit union, margin loans through a stockbroker, and other installment loans. Include credit cards only if the respondent carries a balance.)” Two questions related to government and private educational loans were asked every survey year (by semester) for youth currently enrolled in any type of postsecondary or advanced degree program after high school: “Other than assistance you received from relatives and friends, how much did you borrow in government subsidized loans or other types of loans while you attended this school/institution?,” and “How much is still owed on (this/these) loan(s)?” The variable is created by using a summated yearly figure of all outstanding government and private loans taken out by the respondent for educational study. The median value is assigned to youth who entered in a range (i.e., $0–$1000 was assigned a value of $500). The continuous debt measures were logged, lagged by one period, and included along with an indicator variable equal to 1 if the respondent had no debt, credit card debt, or education loan debt in the respective models. This is done to distinguish both qualitatively and quantitatively between those with no debt and those with some nonzero amount (see Sweeney 2002).⁴

Education, Labor Market, and Financial Characteristics

Current educational attainment is categorized into less than a high school diploma, high school diploma, some college, and bachelor’s degree or more. Enrollment status captures whether the respondent was enrolled in a degree program in any month during the calendar year. The variable is disaggregated into the unenrolled and those attending two- and four-year programs; those enrolled in K–12 are grouped with the unenrolled, but professional degree or postsecondary enrollees are included with those attending four-year programs because of small cell size. Including those with less than a high school diploma and the unenrolled population along with the college-goers and the graduates is important given that they are also accessing credit markets and making decisions related to relationship formation.

Labor market controls include a measure of the youth’s logged predicted annual earnings, lagged one year. This was estimated from the young adults’ hourly wage earnings if they worked full-time year-round, using all available waves of the young adult pre- and post-transition, and was estimated separately by gender (Haurin et al. 1997; Whittington and Peters 1996). Measures of current employment status include indicators for full-time work, having worked 30 or more weeks, and at least 30 hours per week in the previous year. All education and labor market explanatory variables are time-varying.

The total value of all financial and nonfinancial assets at the start of the study period, except the value of primary residence, is included as a proxy of wealth (Schneider 2011). Also included is an indicator for bank account ownership, which captures respondents’ connectedness to formal bank institutions or reveals economic disadvantage (Garasky et al. 2008). A dummy variable equal to 1 indicates those who are “unbanked” (lacking a checking or savings account). Young adult households younger than age 24 have the highest rates of unbanked persons, with percentages declining with age (FDIC 2012).

Additional Controls

Factors expected to impact union formation and timing and considered exogenous to the youth’s relationship type and timing decision are also included. Time-invariant controls for family background are the mother’s and father’s educational attainment as of 1997, whether the youth resided in a rural area at age 12, a variable equal to 1 if the youth lived with both biological parents from birth through age 14, and an indicator equal to 1 if the parental respondent reported negative net wealth in the 1997 survey. Given racial and ethnic differences in young adult cohabitation and marital rates (Addo 2012; Amato et al. 2008), the sample is categorized into four ethnoracial categories: non-Hispanic white (reference group), non-Hispanic black, Hispanic, and mixed race. In addition, all models control for whether the youth currently resides in a rural area, as well as her/his birth year, age, and age squared.

Analysis Plan

To estimate the role of early debt holdings while controlling for the other covariates on transitioning to cohabitation and marriage in early adulthood, I generate hazard function estimates using maximum likelihood (Allison 1984). This modeling technique allows for the inclusion of both time-varying and invariant regressors in the estimation. Respondents are followed for every year they are at risk of transitioning from single status into a union type. For the competing risks (hazard) models, when the decision to cohabit or marry is jointly determined, multinomial logistic regressions are estimated. Given that the outcome can be one of two events—cohabitation or marriage—the hazard rates estimated here represent the conditional probability that a youth will transition out of singlehood into a coresidential union given the other event has not occurred.

Standard errors are clustered at the individual level using the Huber/White procedure, which assumes that observations are independent across and not within respondents. The final data set is arranged in a person-year format, with each young adult contributing an observation for every survey year they remain single from age 20 until they transition to their first union. All observations after transitioning are censored. This is important both to avoid reverse causation, given that prior union history can influence current debt levels, and to permit modeling the importance of financial health in the relationship market during this transitional phase in the life course. All tables list the relative risk ratios, the antilog of the estimated coefficients. A likelihood ratio test comparing a pooled model of both gender and distinct models rejected the null hypothesis at p < .000; therefore, all models were run separately for women and men.

Results

Descriptive Statistics

Figure 1 plots the unconditional hazard rates of transitioning to cohabitation and marriage by gender over the study period. At every age, both men and women have a greater hazard of cohabiting than marrying. Women transition to cohabitation at earlier ages than men and at greater rates across the study period. The hazard of a first union increases with age for both women and men. By contrast, the hazard rates for marriage are low and exhibit a slow and steady increase, peaking at age 29 (the oldest age by the end of the study period) for both women and men. The majority of the sample remained single over the study period: 52 % of women and 62 % of men. Women were nearly twice as likely to transition to cohabitation (31 %) as to directly marry (16.9 %), compared with 24.9 % of men who cohabited and only 13 % who married. These transition rates are in line with current research showing cohabitation as the modal pathway to coresidential relationships in young adulthood (Sassler 2010).

Table 1 compares the rates of indebtedness and average debt for young women and men by first union status. More than 34 % of the young women held credit card debt (averaging $2,582), compared with 29 % of the men (averaging $3,057). Although women were more likely than men to hold any debt, a closer look reveals that such differences are concentrated among women who remained single and those who cohabited; there are no significant differences in the proportion of women and men who held debt and married.

Women and men were considerably more likely to hold credit card debt than education debt. Cohabiting women and men held more credit card debt than did single women. Married men were also more likely than single men to have credit card debt, and in larger amounts. Men who remained single had greater amounts of credit card debt than did women who remained single—the one category where men’s debt levels exceeded women’s. Single women and men are more likely than their cohabiting and married counterparts to have education debt, and both single and cohabiting women are more likely to have education debt than their male counterparts. Of note, however, is that the amount of education debt does not differ significantly by sex.

Close to 7 % of the women in the sample reported holding both credit card and education loan debt, compared with only 3 % of the men. A little more than one-half of the women in the sample reported not having either type of debt, compared with more than 60 % of the men. Women held debt at higher rates in every union category. These results support Chiteji’s (2007) findings that the majority of young adults do not have outstanding credit card debt, with high debt loads concentrated among a minority of young adults. Descriptive statistics for all independent variables used in the analysis are provided in Table 4 in the appendix.

Competing Risks Models: The Decision to Cohabit or Directly Marry Versus Remain Single

The multinomial logistic regression models for women are presented in Table 2, and Table 3 shows the results for men. Model A uses all explanatory variables, including the educational, labor market, and financial measures.⁵ Model B adds in the total debt measure (combined credit card and education loan debt measure), and Model C enters in the debt measures separately. Introducing debt into the model as an additional explanatory variable, along with the youth’s educational attainment and labor market characteristics, tests whether debt is acting as a mediator or operates independently from the other economic resources previously used as predictors of relationship formation. Although debt value can independently signal one’s financial state, it could also work in tandem with other financial and economic measures to provide an overall assessment of financial health. If debt is sending its own independent signal on the relationship market about the respondent, there should be no significant change in the magnitude of the estimates on the other economic resource measures. Asterisks indicate significant differences relative to remaining single, and underlined risk ratios represent significant differences between cohabitation and marriage.

The multivariate competing risks results for young women are presented in Table 2. In Model A, the estimates indicate that educational attainment is positively associated with a first union transition. Single women with less than a high school diploma have decreased risk of directly marrying, whereas having some postsecondary schooling increases the risk of marriage relative to women with only a high school diploma. The least-educated women are also at greater risk of cohabiting than marrying, as indicated by the underlined ratios. There is no association between transitioning to first cohabitation and educational attainment. These results are consistent with the education results reported for women using other data sources (Sassler and Schoen 1999; Sweeney 2002). Women with more education, even those in recent cohorts, are more likely to marry directly.

Current enrollment in any type of postsecondary degree program deters cohabitation, and four-year college enrollment decreases the odds of marriage. The results indicate that school enrollment is perceived as incompatible with early union formation. On the other hand, women who reported holding full-time jobs have an increased probability of cohabitation and direct marriage. Having positive assets aids in the transitions to cohabitation and marriage. Among this recent cohort of young women, positive economic attributes are associated with transitioning out of singlehood into a first coresidential union.

The addition of the total debt measure in Model B has an additive impact on the other socioeconomic attributes in the relationship market; total debt is also an independent and significant predictor of union formation. The relative risk ratio of a one-unit increase in logged total debt is 1.035 for transitioning to cohabitation relative to remaining single and 0.985 for direct marriage relative to cohabiting. In other words, young women with nonzero debt have an increased risk of cohabitation relative to remaining single and relative to directly marrying (Hypothesis 1). The financial status indicators show that being unbanked increases a young woman’s odds of cohabitation relative to marriage in any given year. Net financial assets are positively and significantly related to transitions into a first cohabitation or direct marriage.

The final results, from Model C, assess whether the type of debt held matters for the union decision choice. The competing risks models reveal that relative to remaining single, cohabitation is the relationship choice for women with credit card debt (Hypothesis 2). A one-unit increase in logged credit card debt is associated with a risk ratio of 1.057 for transitioning to cohabitation relative to remaining single. Young women with education loan debt exhibit a 0.928 decreased risk of directly marrying relative to remaining single. Partially consistent with Hypothesis 3, women with education loan debt are less likely to transition to marriage but are also more likely to transition to cohabitation.

The regression results for young men, whose economic attributes have historically mattered more for marital formation (Sassler and Goldscheider 2004), are provided in Table 3. In Model A, the competing risks model indicates that men with less than a high school diploma have lower risks of transitioning into cohabitation and marriage relative to remaining single, and increased chances of cohabitation over direct marriage. High school dropouts are about one-half as likely to marry in any given year compared with men with high school diplomas. Men with bachelor’s degrees or more are more likely to directly marry than to cohabit, relative to men with a high school diploma.

Compared with the unenrolled, being currently enrolled in a two- or four-year degree program significantly deters cohabitation, with two-year enrollees slightly more likely to transition than four-year attendees; men enrolled in four-year degree programs also have decreased odds of directly marrying, consistent with previous studies (Axinn and Thornton 1992; Sassler and Goldscheider 2004). Men enrolled in four-year degree programs are also more likely to transition to marriage than to cohabitation, as indicated by the underlines in Table 3. Although school enrollment tends to deter cohabiting relationships, advanced degrees increase the probability of a transition, particularly into marriage.

Both labor market characteristics—full-time employment and earnings—are positively related with transitioning into cohabitating unions. The risks of cohabitation and direct marriage relative to remaining single as well as of direct marriage relative to cohabitation are positively associated with the value of total assets, which is consistent with the results found for women. Unbanked young men, however, have greater odds of cohabitation over the other two relationship states, whereas being unbanked increased women’s odds of cohabitation only relative to marriage.

The addition of the combined debt measure to the male competing risk models (Model B) does not significantly alter the relationships of the other measures of economic stability and earnings potential, with one exception: men with at least a bachelor’s degree now have a decreased risk of cohabiting relative to remaining single, although they remain more likely to directly marry than cohabit. Contrary to Hypothesis 1, having no debt is negatively associated with transitioning from single into a union for young men. Men’s economic attributes more strongly predict first cohabitation, whereas the same economic characteristics for women are stronger predictors of direct marriage. Not only are women’s economic attributes a significant predictor of their first coresidential union, but they also appear to matter more for marital formation than do men’s economic attributes (Hypothesis 4).

For men, the estimates in Model C indicate a positive correlation related to transitioning into cohabitation for those with nonzero credit card debt (Hypothesis 2). Men who reported zero credit card debt have decreased odds of transitioning into either cohabitation or direct marriage. Education loan debt does not increase men’s risks of transitioning into cohabitation relative to marriage and continued singlehood, nor does it decrease their risks of marriage as it did for the female sample. The lack of debt—specifically, no credit debt—matters more for union transitions. Contrary to Hypothesis 3, only credit card debt increases transitions into cohabitation relative to remaining single.

In an additional sensitivity analysis (available upon request), the multinomial regression models were run on young adults with at least some college (sample size of 1,305 women and 1,025 men) and only college graduates (1,061 women and 800 men). The association between the debt measures and union transitions did not change substantially in magnitude or significance for young women. The negative relationship between enrollment in two-year degree progression and cohabitation and four-year degree programs were no longer significant relative to remaining single, nor was the positive relationship between transitioning into cohabitation with total assets. For the young men, the relationship between total debt and direct marriage relative to remaining single was both negative and significant; however, credit card and education loan debt did not differ.

It is possible that the relationships between education debt and educational attainment could be driving these associations, especially for those who have not completed their studies. Models interacting education debt with education were tested, given that many young adults enter and leave postsecondary schooling without obtaining a degree. Interaction results were significant only among male college graduates, who had an increased risk of cohabitation relative to remaining single regardless of whether they carried education loan debt.

All models were then run on a pooled data set combining women and men and added gender and nonzero debt interactions to address whether gender differences in the relationship between debt amount and debt types are significant. The interaction of being female and positive total debt is significant at p < .10 for transitions into cohabitation relative to remaining single and marrying, providing additional support for Hypothesis 4. When delineated by debt type, gender did not have an added impact on positive credit debt for women. Education loans, however, were more highly associated with transitions into cohabitation relative to singlehood and relative to marriage for women than for men (p < .10). These results highlight that the economic attributes of young women—from their educational attainment to their labor market and financial characteristics—are associated with their first union choice, particularly regarding transitions into marriage, with debt playing an independent role on women’s transitions during young adulthood. Debt increases cohabitation and deters marriage, suggesting that the financial underpinnings related to cohabitation and marriage differ. The findings also imply the type of union entered is sensitive to debt type.

Discussion

This study suggests that credit card and education loan debt, an increasingly dominant feature in many young adults’ debt portfolios (Houle 2014), is an important factor in union formation decisions during this stage in the life course. Educational attainment and labor market characteristics still matter for relationship formation, but so does financial status—and, specifically, debt, at least for this cohort of young adults. Although wealth and marriage remain significantly associated among young adults (e.g., Schneider 2011), my findings extend previous work by revealing that debt matters as well, for both the cohabitation and the marriage market. Debt is positively associated with transitioning out of singlehood, but the absence of debt is not, suggesting that single life in young adulthood may be difficult to afford. Married life, however, is unaffordable as well. Cohabitation presents an alternative to single life but is not necessarily a substitute for these adults.

Gender differences suggest the economic burden debt presents for a marital union is treated differently within men’s and women’s relationship markets. Total debt, credit card debt, and education loan debt all increased the odds of cohabitation for women, consistent with Hypotheses 1, 2, and 3. Yet, only positive credit card debt was associated with transitioning into cohabitation from singlehood for young adult men. The analyses indicate the existence of an economic threshold for cohabitation, one that may differ for young men and women. Women with debt may be considered unattractive in the marriage market but still have some resources (e.g., employment, education, and access to credit) that can positively contribute to a nonmarital coresidential household. The findings for women also support the qualitative research showing that debt is a barrier to marriage but not to cohabitation (Sassler and Miller 2011; Smock et al. 2005). Young men who lack observable indicators of financial sophistication (e.g., bank accounts, and nonzero credit card debt) may be considered suitable relationship partners but may be relegated to the cohabitation market for their first coresidential union. Men with no credit card debt, however, do not enter either the cohabitation market or the marriage market. This may potentially reflect lack of access to credit, which is a sign of financial immaturity. Men with debt are more able to transition to coresidential unions than women, indicating that men continue to dictate the terms of union formation within this recent cohort of young adults (Sassler and Miller 2011). Men appear willing to accept a working partner but not one who is a potential financial burden on the household. These findings might explain the growing socioeconomic divide between those who marry and those who chose to delay or opt not to marry, as well as the rising median age at first marriage for women and men.

My findings also indicate that women’s economic attributes are increasingly important for marriage formation, extending prior research based on earlier samples of young women (Oppenheimer 1997; Sassler and Schoen 1999; Sweeney 2002). Early union transition type and timing of marriage remain associated with measures of educational attainment and positive indicators of current financial health and future economic stability. Current trends in the marriage market reflect labor market fluctuations, which have seen the rewards to highly skilled men increase disproportionately in size relative to low-skill wages (Madrick and Papanikolaou 2010). Dwyer et al. (2013) found the relationship between debt and enrollment to be gendered, with the odds of dropping out of college more sensitive to education loan debt for young adult men than they are for young women. My results indicate that education loan debt is also a contributing factor to continued social stratification and economic mobility within society. Although greater educational attainment is associated with transitioning to marriage for women, women are more likely to pay a penalty for their education loan debt than men. The accrual of debt from pursuing greater educational attainment may have unintended consequences for women, with cohabitation being the more attractive coresidential option earlier in the life course (Hypothesis 4).

This study is not without limitations. First, given that the data are right-truncated, this is not a study of nonmarriage but rather a study of delayed marriage. Second, debt is a stock quantity, meaning that it is a point-in-time measure. It is difficult to ascertain from the survey how long it took the young adults to accumulate debt and how long it will take them to pay it off. Third, aside from the unbanked proxy variable, testing for actual credit access and whether a youth was credit-constrained was not available because it was not explicitly assessed until later interviews. The results presented reflect actions related to credit use. Fourth, this study analyzes how outstanding debt and debt type potentially alter behavior for young adults navigating in the cohabitation and marriage markets—not whether young adults with access and outstanding debt are inherently different from those with debt. By distinguishing those with nonzero debt from those with no debt, it is possible that at least two distinct socioeconomic and demographic groups are combined into the no-debt category: the credit-constrained and the wealthy. Descriptive statistics indicate that young adults sampled fall predominantly into the lower- and middle-income class and that this study has captured the behavior of a policy-relevant and growing demographic: young adults saddled with credit card and education loan debt. Another potential concern is that differential access to education loans may influence enrollment; however, several studies have found access to financial aid for postsecondary schooling is not a constraint for enrollment decisions (Carneiro and Heckman 2002; Stinebrickner and Stinebrickner 2008), although recent research has suggested that household credit constraints may negatively impact children’s college enrollment decisions (see Lovenheim 2011). And finally, although no strong causal claims are made, my results do provide strong evidence that debt plays a nontrivial role in the relationship decisions of young adults coming of age in the twenty-first century.

Marriage as an institution has undergone significant changes during recent decades (Cherlin 2005). Contemporary young adults now emphasize individual financial and personal responsibility as a necessary precursor for marriage, with a shift away from shared obligations and asset accumulation starting early in one’s adult life. Marriage has come to represent the finish line rather than a starting point of young adult life (Cherlin 2004), and it is no longer something that the majority of young adults enter into in early adulthood. The relationships among economic resources, union formation, and marital timing remain an area of demographic interest. Future studies should focus on the role of debt on family-building behaviors, such as fertility decisions, relationship quality, and remaining married or partnered. Ongoing compositional changes in the labor market, education market, and the financial landscape are salient and undoubtedly impact the American household and family. This article examines one specific structural factor that emerged in the last three decades within the lives of America’s young adults: indebtedness. The evidence suggests that debt, operating through the differences in the debt structure of credit card and education loan debt, may have an independent influence on the first coresidential union choice, above and beyond the traditional educational and labor market characteristics of “good fortune” presumed necessary for marital formation (Sassler and Goldscheider 2004; Schneider 2011). These results pinpoint the need to revisit the relationship between economic resources and early union formation, particularly during a period when cohabitation has become more common and rising inequality has contributed to a new marital divide in America.

Acknowledgments

The author would like to acknowledge Sharon Sassler, Daniel T. Lichter, and Francine D. Blau for their helpful comments on early versions. A preliminary version of this article was presented at the 2012 annual meeting of the Population Association of America in San Francisco, CA.

Appendix

Notes

References