Intimate Relationship Development during the Transition to Adulthood: Differences by Social Class

Abstract

Compared to middle- and upper-class youth, lower-class youth have a higher prevalence of sexual activity and are more likely to cohabit or to marry early, but they are less likely to ever marry. Lower-class women have strong desires for marriage but difficulty in achieving common pre-requisites for marriage. Social class also shapes the relationships of special class-graded groups of youth such as sexual minorities, military service personnel, and prisoners. More research is needed on how the state and its laws and institutions constrain even the most intimate features of young lives.

Introduction

Romantic and sexual relationships first begin in adolescence and usually develop into more serious and committed relationships in early adulthood, often leading to cohabitation, joint parenthood, and marriage. On the heels of intense peer relationship development in early and middle adolescence (Brown, 1999), the late adolescent and early adult years are perhaps the period in the life course that is most occupied by social relationship development. From the early work of Dunphy (1963) to more recent studies, developmentalists have charted the normative pattern by which teens gain romantic experience first in mixed sex peer groups and later in dyadic relationships (Connolly & Goldberg, 1999; Brown, 1999; Furman & Wehner, 1994). Recent research ties these adolescent romantic experiences to cohabitation and marriage (Meier & Allen, 2007; Raley and others, forthcoming) and to relationship quality in young adulthood (Seiffge-Krenke, 2003), suggesting that adolescent romance is not trivial puppy love, but rather it carries developmental currency for the more serious relationships of adulthood (see too Collins, 2003).

Attachment theory posits that very early relationship experiences, especially the infant-mother relationship, are particularly important in facilitating successful relationships later in life (Bowlby, 1973). They are important because they provide cognitive representations of how relationships should operate (Furman & Wehner, 1994), and they allow individuals to build skills that can be invoked in later relationships (Collins & Sroufe, 1999). This suggests that parent-child relationships early in life should influence adolescent and young adult intimate relationships through the views they cast of how relationships operate. Recent theorizing suggests that important attachment relationships include those in adolescence (Allen & Land, 1999), suggesting that adolescent intimate relationships are not only the product of early life attachment relationships, but they may also be the attachment source generating models for future relationships. Adolescent romantic relationships may be a crucial source of ideas about how intimate partnerships work differently than other relationships, and help to build skills unique to romantic relationships (Giordano, 2003).

One way that adolescent romantic relationships are different than relationships with parents or peers is that they carry the unique potential for sexual activity. Indeed, sex is often part of romantic experience in adolescence as 63 percent of all teens have intercourse before they graduate from high school (Centers for Disease Control, 2005). Sex carries with it the potential to initiate family formation earlier than intended through pregnancy and childbirth. As young people stretch their dating lives well into their 20's or forgo marriage altogether, the risks of unintended and non-marital pregnancy rise. In the United States in 2001, 49 percent of all pregnancies and 30 percent of all births were unintended (Finer & Henshaw, 2006). Thus, romantic relationships and sexual experience during the transition to adulthood have the potential to significantly alter the course for future relationship formation.

Class Differences in Intimate Relationships

The course of romantic and sexual development from adolescence into adulthood is not universally experienced. In adolescence, class differences shape sexual experience more than relationship experience more generally. Tables 1 and 2 show the distributions of relationship and sexual experience by socioeconomic class using data from the National Longitudinal Study of Adolescent Health (Add Health; see Udry, 2003 for further information regarding Add Health). Add Health asked respondents, “In the last 18 months, have you had a special romantic relationship with anyone?” If the respondent answered “yes,” they are considered to have had a “romantic” relationship. If the respondent answered “no,” they were skipped to another section that asked whether they had ever done each of the following three things with a non-family member: held hands, kissed on the mouth, or told someone they liked or loved them. If they responded “yes” to all three of these items and they indicate all three happened with the same person, they are considered to have had a “liked” relationship. If the respondents report no romantic or liked relationships, but they had a sexual relationship in the past 18 months, they are considered to have had a “sex only” relationship.

Table 1.

Teen Relationship Experience by Family Income

	Type of Relationship
	Romantic	Liked	Sex Only	None
lowest quintile	52.94	9.23	9.3	28.53
quintile 2	51.41	7.65	7.07	33.88
quintile 3	54.74	6.03	7.64	31.6
quintile 4	54.84	4.48	7.22	33.46
highest quintile	58.26	3.07	6.59	32.07

^*data adjusted for complex sampling and weighted to population N=22,121,929

^*Authors' calculations from Add Health, Wave 1. When missing, family income is imputed from parental education, race/ethnicity, and family structure.

Table 2.

Teen Sexual Experience by Family Income

	Ever Had Sex
	female	male	total
lowest quintile	47.71	56.96	52.2
quintile 2	41.46	49.84	45.72
quintile 3	39.28	44.58	42.04
quintile 4	33.91	39.51	36.87
highest quintile	33.08	35.86	34.47

^*data adjusted for complex sampling and weighted to population N=21,837,633

^*Authors' calculations from Add Health, Wave 1. When missing, family income is imputed from parental education, race/ethnicity, and family structure.

There are only small differences in relationship experience across family income quintiles. Here we see that slightly over half of all adolescents report some romantic experience, with a slightly higher prevalence in the top income quintile. While the percentages reporting liked relationships are relatively small across all income groups, there is a noticeable trend with income: those in the lower income quintiles are more likely to report liked relationships than those in the higher income quintiles. Class differences are smaller in the percentages of those who had relationships that were sexual only. The only notable difference here is that those in the very bottom income quintile are slightly more likely than others to have sex-only relationships.

While class differences in relationship experience are minimal, there are rather striking differences in sexual experience. Sexual experience here is measured from several questions in the Add Health study that ask respondents whether they have had intercourse (one question is asked irrespective of relationship status, the other asks about intercourse in relationships). Here we use both measures to create an indicator for whether or not the respondent has ever had sex. Those on the bottom rung of the class ladder are substantially more likely to have had intercourse than those higher on the ladder, especially boys. The class gradient in sexual experience is clear: those in the highest income quintile are least experienced with successive increases in experience by income quintile; those in the lowest income quintile are most likely to be sexually experienced.

The most recent wave of Add Health data (collected in 2001) tracks respondents into their young adult years (ages 18–26). As shown in Table 3, results from these data suggest class-graded movement into early marriage and cohabitation. Young adults, particularly females, from lower-income families are more likely to make early transitions into both cohabitation and marriage. Approximately 28% of youth in the lowest income category have cohabited by the time they reach age 20 compared to only 15% of youth in the highest income category. Similar results hold for marriage. Youth in the lower income quintiles are more likely to marry early than youth from the highest income quintile. Given the correlation between early marriage and divorce (Lehrer, 2006), early-marrying youth may be setting the course for further economic setbacks stemming from divorce. Overall those from the higher income categories appear to be making family-related transitions later than youth from the lowest income quintiles.

Table 3.

Early Cohabitation and Marriage by Family Income

	Early Cohabitation (by age 20)			Early Marriage (by age 26)
	males	females	total	males	females	total
lowest quintile	22.71	32.73	27.7	16.99	25.92	21.44
quintile 2	20.09	33.96	27.08	19.82	27.96	23.94
quintile 3	17.71	27.78	22.47	15.52	25.01	20.01
quintile 4	12.01	21.44	16.56	13.39	23.22	18.15
highest quintile	10.8	18.99	14.89	11.47	15.45	13.47

^*data adjusted for complex sampling and weighted to population N=21,847,389

^*Authors' calculations from Add Health, Waves 1 and 3. When missing, family income is imputed from parental education, race/ethnicity, and family structure.

Add Health data captures early marriage but it does not tell us about the prevalence of marriage after age 26. Longitudinal data from the Youth Development Study shows a strong correlation between parents' income, education, and marital status and young adult transitions into marriage and parenthood (see Mortimer, 2003 for more on the YDS). These data suggest five patterns of joint movement into marriage and parenthood by ages 30–31: delayed family formation (no marriage or parenthood), marriage only (without parenthood), traditional married parents (marriage followed by parenthood), single parents (parenthood without marriage) and nontraditional married parents (parenthood followed by marriage). Youth with more highly educated parents and higher family incomes, those from two-parent families, and whites are more likely to follow the first three (more traditional) patterns than other youth (Eliason, Mortimer, Vuolo, & Tranby, 2007). That is, they are more likely than those from disadvantaged backgrounds to delay marriage, get married but delay parenthood, or marry and then have children.

Overall, family background and income appear to shape the nature and timing of adolescent and young adult intimate relationships. Adolescents from lower-income families are more likely to have sex and engage in sex-only relationships. Lower-income youth transition to cohabitation and marriage more rapidly than youth in higher income groups. Additionally, youth from higher-income backgrounds are more likely to follow normative patterns into marriage and parenthood. Finally, these differences may have long term effects in depressing the educational attainment and earnings of those from disadvantaged backgrounds.

Emerging Adulthood on the Margin

Jeffrey Arnett (2004) provides the useful frame of emerging adulthood to help us understand the elongated and increasingly heterogeneous experiences of the late teens through the third decade of life. He highlights five features of emerging adulthood (p. 8):

1. Identity exploration - a time of various possibilities, especially in love and work.

2. Instability - a time of frequent change in jobs, residence, romantic partners, etc.

3. Self-focus - a time with minimal responsibilities to others.

4. In transition - a time of feeling in-between, neither adolescent nor adult.

5. Possibilities - a time when hopes flourish and opportunities to transform abound.

All of these features apply to romantic and sexual relationships. Arnett's framework suggests that young people date a number of people to discover which features they like in partners and what sort of partner they will be (identity exploration, self-focus). This means a more frequent turn-over in partners than is common in later phases of the life course (instability). The framework also implies that during this period, young people are free from the constraints of, and obligations to parents and other family members, but they have not yet assumed the adult roles of committed partner or parent (self-focus, in transition). All of this transition leaves open a world of relationship (and other) possibilities. They may choose to date informally, cohabit, marry someone, or start a family sooner or later depending on what their exploration and self-focus yields.

Arnett recognizes that social class and family background shape the extent to which young people experience emerging adulthood (Arnett, 2004, p. 7). However, he highlights the exceptional cases of disadvantaged children who rise above their circumstances when they enter emerging adulthood (Arnett, 2004, Ch. 9). In fact, Arnett argues that the period of emerging adulthood (loosely defined as ages 18–30) is most important to those of meager resources or troubled family lives because it affords them the much-needed freedom to create a new future.

If this life stage is more critically important to improve the lives of those who have grown up with fewer resources, it is also more difficult for these young people to avail themselves of the exploration, self-focus, and possibilities that characterize emerging adulthood. Furstenberg (2006) makes the compelling case that social class influences development and that patterns set in place early in life are difficult to surmount. Thus, while emerging adulthood may be an age of possibilities, those possibilities are differentially constrained by social class.

One example of the ways in which social class shapes relationship development is evident in the ethnographic work of Edin and Kefalas (2005), who explore the question of why poor young women put motherhood before marriage. Rather than devaluing the institution of marriage, they find that poor women revere marriage. It is because they see marriage as such an important institution that they have numerous pre-requisites before they marry; common requirements are financial independence, a small house, enough money for a proper wedding, and stable employment for themselves and their intended spouse. Most would call these pre-requisites forward-thinking given the uncertainty that life can bring. For those in the middle-class, these seem like reasonable expectations; they may also be attainable for those in the working class with some hard work, persistence, and perhaps, luck. For the poor, however, these reasonable expectations are extremely difficult to meet given their current social and economic reality characterized by poor job (much less career) prospects and high levels of incarceration (Wilson, 1987). As several scholars have noted, marriage is becoming somewhat of a luxury good (Goldstein & Kenney, 2001; White & Rogers, 2000; Furstenberg, 2006) that is realistic only for those with a fair share of resources.

This explains why poor women have lower rates of marriage overall (even though they have higher rates of early marriage) than middle- or upper-class women; it does not explain why they have children at relatively young ages on average and outside of marriage. As Edin and Kefalas (2005) explain, most young disadvantaged women do not set out to have kids outside of marriage, but they also are not vigilant in trying to prevent pregnancy. Furthermore, many relationship partners express a desire to have a child with these young women, and this expressed desire is a symbol of the esteem in which they hold their partner (Anderson, 1999; Edin & Kefalas, 2005). This, combined with the young women's ambivalence about pregnancy and low opportunity costs, often leads to a pregnancy that is not entirely intended, but also not entirely unintended.

How does this picture of non-marriage but early childbearing among the poor square with the idea of emerging adulthood? The late teens and twenties are a time of great instability for poor youth, but their instability is of a different character than that of their middle-class counterparts. Whereas Arnett discusses frequent residential moves into and out of the parental home, apartments with friends, and dormitories as “the best illustration of the instability of emerging adulthood” (Arnett, 2004, p. 11), instability for poor youth is instead characterized by bouts of unemployment, incarceration, pregnancy, and child birth. This instability is arguably more consequential than the residential hops and varying degrees of independence that middle-class youth experience. Poor youth often live with extended family, with or without their romantic partners and children, to pool resources while they transition to adulthood and sometimes long after that (Goldscheider & Goldscheider, 1999).

Are the late teen and early adult years a time for exploration in romantic relationships and self-focus for poor youth? Indeed, those from disadvantaged backgrounds cohabit at a higher rate than those with more family resources. While their cohabitation behavior may seem congruent with relationship exploration as described by Arnett (2004), the motivation for cohabitation may be quite different across the class spectrum. Arnett suggests cohabitation is a way for emerging adults to experiment with partners and test out marriage-like relationships. However, among lower-class individuals, cohabiting unions may be less about exploration and more about practical matters like saving money (Smock & Manning, 1997). Living together without marriage is financially prudent in several ways: individuals can take advantage of economies of scale by combining expenses, cohabitation is a way to establish a partnership without the cost of even a modest wedding, and finally, cohabitation requires a weaker commitment than marriage to fulfill long-term economic responsibilities. Therefore, lower-class young adults cohabit at higher rates than others, but the degree to which their cohabiting unions are initiated as a means of partner exploration is questionable.

In addition to the material pre-requisites for marriage, several young women interviewed by Edin and Kefalas (2005) expressed other varieties of exploration and self-focus as reasons for delaying marriage, although it is unclear if they actually participate in these activities. For example, when asked why one should delay marriage, one woman responded, “Get out there, see the world. Travel, you know. Know what you want out of life before you say `I do.'” (p. 125). While this sentiment is consistent with the exploration and self-focus features of emerging adulthood, there is no indication that this, or any other woman in their study, actually traveled or saw more of the world than her immediate surroundings. Moreover, as Edin and Kefalas (2005) note, “The obvious irony is that…these young women have already taken on what most Americans consider the most significant adult social role of their lifetimes, that of a parent” (p. 124). Having a child is likely to at least partially foreclose the exploration and self-focus opportunities enjoyed by others in emerging adulthood.

Is emerging adulthood an age of romantic possibilities for poor youth? As suggested by Arnett, it is a time when young people have a chance to break free from their up-bringing, and this chance may be particularly valuable for those from disadvantaged backgrounds. Relationship development for middle-and upper-class youth often occurs in highly age-segregated and class-homogeneous settings (for example college campuses) that are conducive to meeting eligible others and exploring relationship possibilities. However, youth who move directly from high school to the work world may spend more time in age-heterogeneous settings (for example job settings) and have fewer opportunities to meet desirable, eligible partners. By choosing a partner from a different socio-economic background young adults may indeed break free from their disadvantaged roots. Unfortunately it is more the exception than the rule. Recent research indicates that since the early 1970s there has been a steady decline in marriages across class lines (Schwartz & Mare, 2005). While marrying up the social ladder may be attractive in terms of increasing one's socio-economic status, it may also bring difficulties within the relationship. Class differences can breed power inequities, especially when women start at a lower class position, as is more common in cross-class partnerships (Tichenor, 2005; Blumberg & Coleman, 1989). Thus, Arnett's notion of an age of possibilities as it relates to relationships is increasingly improbable for poor youth; and even if accomplished, it may have undesirable side effects.

Class-Graded Special Populations

When studying class differences in adolescent and young adult relationships, we often use school-based or large-scale survey data. Doing so leaves us lacking in our understanding of several groups who are typically understudied in surveys or interviews but who are likely to have unique relationship experiences: the incarcerated population, those serving in the military, and sexual minority youth (gay, lesbian, and bisexual). These three groups make up a considerable proportion of the young adult population today. While the full class spectrum is present in each of these populations, membership in them is highly class-graded. The prison population is overwhelmingly comprised of lower-class youth (Pettit & Western, 2004); the military draws predominantly from young adults of working-class backgrounds (Halbfinger & Holmes, 2003); and sexual minorities are disproportionately middle-class (Black and others, 2000).

All three of these populations face unique circumstances that shape their romantic development. Future researchers would be wise to focus on their relationship experiences. Such research would allow us to better understand the relationship development of these under-studied yet numerically significant sub-populations and should reveal how class-graded institutions shape young lives. Below we detail what is known about romantic development in each of these populations and offer some ideas for next steps in research efforts to complete the picture.

Sexual Minorities

Relatively little is known about the relationship experiences of sexual minority youth, and what we do know is often dependent on the definitions used in data collection (see Black and others, 2000 for a review of prominent large-scale surveys). Using data compiled from multiple studies, Savin-Williams and Ream (2007) discuss the prevalence rates of sexual/romantic attraction, behavior, and identity among youth. Females report higher levels of homosexuality on all three measures, with up to 13% reporting same-sex attraction, 11% reporting same-sex behavior, and 8% homosexual identity; corresponding rates for males are 5%, 5%, and 3% (see too Diamond, 2003).

In addition to descriptive data, Diamond and Dubé (2002) find that sexual minority youth are less likely than heterosexual youth to report any romantic relationships, but we know little beyond this. Research on gay, lesbian, and bisexual adults sheds some light on the experiences of sexual minority youth. The same-sex relationships of adults are on average less homogamous along dimensions of social class, race/ethnicity, and age than heterosexual partners (Andersson and others, 2006; Jepsen & Jepsen, 2002; Rosenfeld & Kim, 2005). Recent work by Meier, Hull, and Ortyl (2007) suggests that this may be because sexual minority youth have more accepting relationship views. Using Add Health data, the authors find that sexual minority youth rate racial homogamy and having enough money as less important in their relationships than heterosexual youth.

While little is known about the social class of sexual minorities' families of origin, research by Black and others (2000) suggests that on average lesbians and gay men attain higher education levels than their heterosexual counterparts. However, these higher education levels do not directly translate into higher incomes; partnered gay men earn less than other men with similar education levels while lesbian women (of any partnership status) generally earn more than other females. Thus, we expect the relationship experiences of sexual minority youth to differ from heterosexual youth given their extended time in higher education, more accepting attitudes toward relationship partners, and the shifting cultural, social, and legal landscape for same-sex partnerships.

Incarcerated Young Adults

Incarceration disproportionately affects young adults, males, and minorities from poor (especially poor black) neighborhoods (Messner and others, 2001). In addition, a criminal record is one element that distinguishes many of those in what William Julius Wilson (1987) termed the ghetto underclass, signaling the strong association between incarceration and social class. While 18- to 29-year-olds make up 16.5% percent of the U.S. population, they comprise 35 percent of the incarcerated population (U. S. Census, 2000). Thus, a large proportion of incarcerated individuals are removed from civilian life while in their prime relationship development years.

Arguably, intimate relationships are more important for individuals with delinquent or criminal backgrounds than they are for others. Several scholars have shown that increased social bonding, largely through commitments to mainstream institutions like marriage and family, is an important “turning point” in the life course of delinquent youth (Uggen & Massoglia, 2003; Cernkovich & Giordano, 2001). Yet, our knowledge about the relationships of incarcerated young adults while in prison is quite limited. Young adult prison inmates are unlikely to be married, but most have children (Uggen & Massoglia, 2003, Table 15-1). Recent ethnographic research on female partners of male prisoners suggests that many men maintain relationships with an intimate partner while in prison. While most women view imprisonment as an unwanted interruption in their relationships, for some women incarceration serves to hold together a relationship that was otherwise on shaky ground (Comfort, 2007).

Finally, incarceration sends social ripples through the home communities of imprisoned men. Estimates suggest that incarceration has increased by more than 600% in the past three decades (Bureau of Justice Statistics, 2002). This dramatic increase, coupled with its disproportionate toll on poor black men, has lead some to suggest that incarceration has created a severe imbalance in the marriage market for black young adults, and it may be a primary factor responsible for their disproportionately low marriage rates (for example Lichter and others, 1992).

Future research should examine how incarceration shapes not only marriage prospects, but relationship experiences more generally. Since most who are incarcerated will re-enter society after they serve time, it would be interesting to assess how ex-convicts' experiences in prison shape their relationship skills and potential to meet partners. For example, research on how the stigma associated with a criminal record shapes labor market prospects (Pager, 2003) could be extended to examine how it shapes relationship development processes as well.

Soldiers

The issue of gender, race, and class diversity in the U.S. military are timely and “hot-button” issues,¹ though these issues were first considered with the demise of the draft and the subsequent all-volunteer force (Lundquist & Smith, 2005). The Department of Defense has tracked changes in the demographic profile and population representation of all branches of the military for the last 30 years. The military is a relatively young workforce; 87% of all new recruits were 18–24 years old (Maxfield, 2007). According to their reports, all branches of the military have seen growth in the percent of females (15% of all active duty soldiers) and minorities. And, while a Department of Defense press release in 2005 suggests that the army is “strongly middle class” (Garamone, 2005), a 2003 New York Times article titled “Military Mirrors Working-Class America” suggests otherwise, and focuses on the draw of the military for those lacking the economic resources necessary to attend college (Halbfinger & Holmes, 2003). Demographic evidence partially supports the claim that the military is increasingly working-class. In 2005, only 87% of enlisted Army soldiers had high school degrees compared to 96% in 1995.

Further research on military demographics has come out of the Military Family Research Institute at Purdue University. Overall, 53% of military service members are married–only 7% are married to others in the military. Active duty men are more likely than active duty women to be married, and 43% have children (Military Family Research Institute, 2004). Though research has focused on the effects of military service on marriage, much of this research is drawn from out-dated survey data (Lundquist & Smith, 2005; Lundquist, 2004; Teachman, 2007). Military service-members are more likely to marry and they marry earlier than their civilian counterparts, with particularly strong effects for blacks (Lundquist, 2004; Teachman, 2007).

Though we can infer from demographic data what some of the associations between military service and marriage may be, we know relatively little about the dating patterns of service-members. Active-duty military personnel experience a constrained environment for relationship development given the extensive amount of time spent on military bases, frequent moves between bases, infrequent leaves, predominance of males, and the relative rarity of dual-military marriages. How do soldiers engage in courtship? Where are soldiers meeting their spouses? Furthermore, how are gay/lesbian soldiers faring under the current “Don't Ask, Don't Tell” policy? Future research should explore these issues as military service is an increasingly common experience for working class men and women, and it undoubtedly shapes relationship development in the critical young adult years.

Conclusion

In adolescence, youth experience fairly similar rates of romantic involvements across the social class spectrum. However, adolescent sexual activity is clearly class-graded. There is evidence that class barriers profoundly shape young adult relationships. Lower class young adults are more likely to cohabit, but less likely to ever marry; they have strong desire for marriage, but difficulty in achieving the pre-requisites for marriage. These differences disadvantage the lower-class and make it less likely that they will experience the emerging adulthood Arnett describes as fully as their middle- and upper-class counterparts. Even as emerging adulthood is arguably a more important life stage for the disadvantaged if their potentials are to be actualized, its distinguishing experiences are less accessible to them.

Researchers have taken advantage of recent large, rich nationally representative studies like Add Health to document adolescent and young adult romantic and sexual experience. In addition, attachment theories have been extended to consider adolescent attachment relationships, and Arnett has theorized about a new, longer and less standardized transition to adulthood. Thus, both theoretical and empirical developments have enriched our understanding of intimate relationships during the transition to adulthood. Yet, several numerically significant sub-populations have unique circumstances that may impinge upon relationship development. These groups are often outside the scope of the scholarly lens. Undoubtedly, this is largely due to the difficulties in studying these populations. Yet, if we are to understand how class shapes relationships during the transition to adulthood, we must consider how society's highly class-graded institutions, like the military and the criminal justice system, constrain and shape even the most intimate features of young lives.