Classed Pathways to Marriage: Hometown Ties, College Networks, and Life after Graduation

Abstract

Objective:

This article identifies mechanisms through which social class background shapes the marital outcomes of college-going white American women.

Background:

Scholars are interested in the relative influences of ascriptive and achieved characteristics on mate selection. Research indicates that social class background continues to influence the marriage patterns of college-educated Americans but does not identify the mechanisms through which this occurs.

Method:

The study analyzes six waves of longitudinal interviews with 45 women from differing social class backgrounds. The first interview was conducted at age 18, when women started college at a Midwestern public university. The final interview was collected at age 30, and was supplemented by a survey collecting the income, education, occupation, and debt of women and their spouses.

Results:

Women from privileged backgrounds were more likely to marry and married men who earned substantially more than the partners of less privileged women. Differences resulted from lifelong variation in social networks, originating in childhood. College did not interrupt long-standing exclusionary class networks. After graduation, social class background shaped where women moved, as well as with whom they worked and socialized.

Conclusion:

Higher education in the contemporary US may reinforce rather than interrupt class homogamy in marriage, even when students attend the same schools. The role of higher education in shaping classed social networks is in need of further study.

I know I’m gonna marry a guy that’s financially stable, just because that’s, like, my background.

--Tara, from a privileged family

Marriages tend to occur among those who are similar to each other (Schwartz, 2013). While ascribed factors such as class, religion, and nationality have long served as primary dimensions of marital sorting, achieved statuses such as education, earnings, and occupation have increased in salience. Scholars have grown interested in whether higher education reduces the influence of social background on marriage—particularly social class background (Blackwell, 1998; Kalmijn, 1991; Musick, Brand, & Davis, 2012; Nielsen & Svarer, 2009). If college and post-college life place students and degree holders in interaction, it is reasonable to expect a college degree to equalize marital prospects, as it does with earnings (Hout, 1988; Torche, 2011).

Yet recent scholarship finds that class background continues to affect marital prospects among Americans with a college degree (Musick et al. 2012). Precisely how class background shapes the marriage patterns of the college-educated has yet to be fully explained. Musick et al. (2012, pp. 56) suggest that “social and cultural factors” associated with class background “influence matching in the marriage market,” even among those who receive bachelor’s degrees. As their study involves quantitative analysis of population-level data, it is not designed to identify the mechanisms that generate the patterns that they describe. They hypothesize that “school and family environments” shape how “men and women assess potential mates” and note that “college-goers from less advantaged backgrounds” may end up “in a poor position to compete in the marriage market” but cannot observe how this occurs (pg. 54).

We explicate Musick et al.’s (2012) findings by documenting processes through which social class shapes the marital outcomes of white American women attending a flagship university in the Midwest. We draw on six waves of longitudinal interviews with a cohort of 45 women from differing class backgrounds, starting at age 18 and ending at 30. Our data identify the circumstances under which women met the individuals they came to marry, and how these relationships progressed to marriage. Among our sample, class privileged women were more likely to marry and to marry a high earner than less privileged women. The vast majority of marriages also occurred among those who shared class background and racial identity.

We explain these patterns by examining how dyad formation is shaped by social contexts key to marital union formation—childhood schools and neighborhoods, college social worlds, and post-college life. First, we find an enduring influence of class and race segregated hometown ties on marriage, as many women, particularly privileged women, married men they knew or could have known in childhood. Second, college did not disrupt exclusionary networks, as social life at the university women attended was highly segregated by class and race. Third, after graduation, women retained childhood and college ties. Class background shaped where they moved and with whom they worked and socialized. We also attend to exceptions (i.e., less privileged women married to high earners and privileged women married to low earners). Overall, we found continuity in women’s networks from childhood into adulthood.

How Marriages Form

Scholars have identified three complementary and interdependent mechanisms influencing dyad formation (Rivera, Soderstrom, & Uzzi, 2010). First, proximity is required for interactions leading to marriage to commence. As Blau and Schwartz (1997, p. 29, emphasis in the original) explain, “Rates of social association depend on opportunities for social contact.” Chance encounters and opportunities for interaction that can lead to relationship formation increase in proximity, as the probabilities of meeting rise (see Rivera et al., 2010; Small & Adler, 2019).

But proximity may not always be enough to generate ties. Existing social networks shape opportunities for interaction, information about potential partners, and the trust extended to a new tie (Rivera et al., 2010). People are likely to connect to those who can be vouched for by friends and family (Christakis & Fowler, 2009). Organizations such as schools and workplaces also shape how social networks connect people (Small, 2009). Spatial context influences the degree to which individuals from different backgrounds interact (Small & Adler, 2019). For instance, the internet enables people to reach outside existing networks; this may appeal to those disadvantaged in dating markets. In contrast, those who inhabit “environments full of eligible partners may not need to actively search for partners” (Rosenfeld & Thomas, 2012, p. 524).

The final mechanism of interest is preference for socially similar others. Referred to by Rivera et al. (2010, p. 94) as “assortative mechanisms,” the notion is that the formation, maintenance, and end of relationships depend on the “compatibility and complementarity of actors’ attributes.” Particularly in the case of romance, homophily on both ascribed and achieved characteristics is often sought (Blackwell & Lichter, 2004; McPherson, Smith-Lovin, & Cook 2001)—online as well as in real life (Hitsch, Hortaçsu, & Ariely, 2010; Robnett & Feliciano, 2011). Thus, while interracial marriage in the US has increased, it is still far lower than random mate selection would predict (Qian & Lichter, 2011). Proximity and networks—not just preferences—also drive homophily, as neighborhoods and networks tend to be populated with like others (Christakis & Fowler, 2009).

Equalization?

In this article we consider how proximity, networks, and preferences shape women’s access to higher-earning husbands as they move through different social contexts before, during, and after college. Prior scholarship offers predictions about the relative importance of these contexts for generating marital relationships, and the circumstances under which they influence the salience of class origin in union formation. Table 1 outlines competing views. Because men from privileged backgrounds are more likely than men from less privileged backgrounds to attend college and thus ultimately earn more (DiPrete & Buchmann, 2013; Hout, 2012), equalization in marital outcomes requires integration of those from different backgrounds in college and beyond. The disruption of early ties is associated with equalization of marital outcomes, while continuity of network ties across life stages works against equalization.

Table 1.

Contextual Influences on Marital Outcomes

	More Equalization (Disruption of Early Networks)	Less Equalization (Continuity of Ties)
Hometown Ties	Childhood social worlds are segregated by race and class, but delayed marriage, geographic mobility, decline of family control, and the rise of online dating disrupt these relationships.	Childhood social worlds are segregated by race and class. Despite social change, early class networks continue to play a powerful role in union formation, particularly for the privileged.
College Social Life	Universities offer efficient marriage markets and connect individuals across class background.	Networks retain their classed nature in college due to the residential and social segregation of college life by class and race.
Life After College	After college, those with degrees live and work in close proximity, integrating social networks, irrespective of class background.	After college, class background continues to shape where people live, work, and with whom they associate. Less privileged women are blocked from accessing higher-earning men.

Hometown ties.

Although historically people married very close to home, Rosenfeld (2009) demonstrated the declining influence of early social worlds on marriage; increasing temporal and geographic distance from families and neighborhoods of origin, often through higher education, produces marriages less homogamous on social background than in the past. Rosenfeld and Thomas (2012) also argue that the internet has displaced family, neighborhood, and friends as the primary venue for meeting partners—fundamentally disrupting hometown ties. However, as Bruch and Newman (2019) illustrate, geography still matters, as most people—even online—are looking to make a face-to-face match relatively close to home. People continue to rely on networks for introductions and information about a person’s background, as a “socially brokered introduction” reduces risk (Christakis & Fowler, 2009). This may be particularly true for elite class groups, which use social closure as a mechanism to reproduce advantage (Ostrander, 1984). When early social ties persist, we can expect to see less equalization in college marital outcomes.

College social life.

If postsecondary education is structured to encourage interaction among students from a wide array of backgrounds, equalization may occur. This possibility is limited by the sorting of individuals of different classes into different schools (Clotfelter, 2017). Within the same school, however, universities have been referred to as “very efficient marriage markets,” with a high “density of potential partners” and “lower search frictions” (Nielsen & Svarer, 2009). Many people marry someone from their alma mater, or a school of similar prestige (Arum, Roksa, & Budig, 2008; Ford, 2019), suggesting that college has the potential to disrupt hometown ties. But residential campuses are often segregated by class and race (Hamilton & Cheng, 2018; Karabel, 2005). Fraternities and sororities were founded to be exclusionary (Hechinger, 2017), in part to ensure that marriages occurred within class and race (Scott, 1965), and help to preserve segregated networks (Park & Kim, 2013). If students do not socialize across class, then college will not integrate.

Life after college.

If receipt of a bachelor’s degree equalizes not only earnings (Hout, 1988; Torche, 2011), but also patterns of geographic mobility and lifestyle, then we might expect college to equalize marital outcomes. With an average age of marriage in the late 20s (Payne 2012), college-educated Americans from a variety of backgrounds may meet after college in neighborhoods, workplaces, bars, restaurants, churches, gyms, or graduate school. A rise in online dating might also allow degree holders from a range of backgrounds to select on achieved, rather than ascribed, characteristics (Rosenfeld & Thomas, 2012). At the same time, some research suggests that a less privileged class background continues to disadvantage graduates in the labor market (Witteveen & Attewell, 2017). Research on college debt (Houle, 2014) and class differences in parental support during the transition to the labor force (Hamilton, 2016) suggests that those from privileged and less privileged backgrounds may not interact with the same people after college—even if they obtain a degree from the same university. If women from different class backgrounds occupy different social worlds after graduation, equalization in marital outcomes is unlikely to occur.

We have discussed proximity and networks—examining the conditions under which social contexts disrupt social ties. But what about women’s preferences? Heterosexual women, irrespective of class background, generally prefer to marry higher-earning men (referred to as “promising” or “marriageable” [see Killewald, 2016; Ludwig & Brüderl, 2018]). Principles of homophily suggest that women also prefer men from class backgrounds similar to their own. Realizing this preference is more possible for class privileged women, as they are likely to locate higher-earning men from within their social worlds. In contrast, less privileged women seeking higher-earning partners may have to marry men with whom they have less in common. Upwardly mobile less privileged women are more likely than privileged women to face a choice between cultural similarity and a higher-earning partner (Streib, 2015).

Although our focus is on classed marital processes, we do not wish to naturalize processes producing racial marital homogamy. The women we studied, regardless of class background, occupied white worlds from childhood through adulthood. Even at college, racially segregated housing and social worlds made it possible for our white participants to form no ties with individuals of different racial backgrounds. This structured their marital outcomes. As Kao, Joyner, and Balistreri (2019) argue, youth who are part of diverse school communities are more likely to engage in interracial romantic relationships.

Data and Methods

We draw on a 12-year longitudinal study of a cohort of white women who began college in 2004 on the same residence hall floor at a large public university in the Midwest. Our book, Paying for the Party: How College Maintains Inequality (Armstrong & Hamilton, 2013), focuses on data from our year-long ethnography and the first five waves of interview data, beginning in 2004 and running annually until 2009, when most women graduated. A sixth wave of data, central to this article, was collected in 2016, when women were turning 30.

Large state flagships have historically catered to families from a wide variety of class backgrounds. Advantaged students of varied ability and ambition are channeled into schools like Midwest U—some lured by the promise of Greek life and a robust party scene. In contrast, it is often the most ambitious students from less privileged families that land at four-year public universities in their home state (see Zhou, 2019 on these selection effects). Like most flagship universities, Midwest U is also predominantly white. Given racial segregation on campus, the residence hall we studied was almost entirely white.

Women in our study were white and US born. They started college at 18 and without children. All but two identified as heterosexual. Existing research suggests that, on the whole, our sample includes women who should find it easier than others to meet and marry a “promising man.” Only women who participated in age 30 interviews are included. This corresponds to 85% (or 45 women) of the original 53 women living on the same-sex residence hall floor. All but three women were interviewed at least four times. The longitudinal qualitative interview design of this study uncovers processes producing marital patterns observed in quantitative scholarship. Although not the focus of this article, 59 interviews with the mothers and fathers of 41 women were conducted in 2008–2009, allowing for data triangulation (for more see Hamilton 2016).

Before the 2016 interview, women completed an online survey including questions about educational credentials, employment, and earnings; current relationship status; marital history; partner’s educational credentials, employment, and earnings; reproductive history; geographic and housing history; monthly mortgage payment; student loan, credit card, and other debt; and savings and preparation for retirement. Women then completed a semi-structured interview.

Age 30 interviews focused on women’s living situations, educational and career developments, romantic relationships, reproductive histories and desires, dependence on and relationships with parents, friendships, and overall financial security. During the interviews, we verified information from the survey that was confusing or incomplete. Interviews ranged from 30 minutes to 2 hours and were conducted over the phone. Women received a $100 Amazon gift card for their participation. In 2018, we also checked back with women to confirm developments that were ongoing in 2016 (e.g., engagements leading to marriage). For reference, the online Appendix lists all 45 women interviewed (by pseudonym), with details about their class origins and their educational attainment, occupations, and earnings at age 30.

As illustrated in Table 2, the class origins of women in this study were determined on the basis of parental education, occupation, income, and parental ability to fund a college education. Privileged women grew up in upper-class and upper-middle-class families. These families had substantial financial and educational resources and could offer support during and after college. Privileged families included two college-educated parents and typically at least one parent with an advanced degree. They were characterized by either CEO/CFO fathers paired with homemaker mothers, or two professionals (e.g., doctors, lawyers, professors, or accountants).

Table 2.

Women’s Class Backgrounds

Class	Mother Education	Father Education	Mother Occupation	Father Occupation	Income Category	Student Loans	Residency	Family Structure	N
Privileged
Upper	College degree	≥ College degree	Homemaker	CEO/CFO	> 525K	None	Out-of-state	Married	7 (16%)
Upper-middle	≥ College degree	≥ College degree	Professional	Professional	125K-525K	None	Mixed	Mostly married	19 (42%)
Less Privileged
Middle	College degree	College degree	Teacher, Management	Middle mgmt.	80K– 125K	Some	In-state	Mostly married	5 (11%)
Lower-middle	≤ Some college	Some college	Secretarial, Sales	Sales mgmt.	40K– 80K	Most	In-state	Mixed	9 (20%)
Working	≤ Some college	≤ Some college	Low-paying service work	Manual labor	< 40K	All	In-state	Mixed	5 (11%)
									45 (100%)

By contrast, less privileged women had limited financial and educational resources. We classified the small number of middle-class families as less privileged based in part on parental occupation; these parents worked in relatively lower paying, non-professional jobs such as food factory supervisor or tractor company distributor. Middle-class women had enough resources to participate in college social life, but—like other less privileged women—experienced a sense of exclusion and deprivation relative to privileged peers. Less privileged women, as a group, typically did not receive family support in the transition out of college.

Our analyses also required information about partner’s education and earnings, which was drawn from the 2016 survey. Partner’s class of origin (coded as either privileged or less privileged) was determined on the basis of several factors: 1) four-year college attendance; 2) debt upon completion of college; 3) parental education and occupation (when available); 4) location and wealth of hometown; 5) direct transfers of partner family wealth to the couple; and 6) women’s reports of men’s families’ affluence relative to their own.

The heart of the paper draws on longitudinal qualitative data to identify mechanisms producing class differences in marital outcomes. These analyses draw on 306 interviews, with particular attention to the wave 6 interviews. We constructed narrative accounts of how each woman met her partner, revisiting as many interviews as necessary. These narratives were analytical, as we read for the context and network ties that made the connection possible. We also focused on women’s evolving statements of what they did—and did not want—in a partner. We worked jointly on these narratives, sharing them back and forth. Early in the analysis, our approach was inductive. As we began to see patterns, we approached the data more deductively. At this point, we turned to Excel to systematically code how each woman met her partner (see the Appendix). This enabled us to verify that patterns were present across the data set. The Appendix also displays marital partners’ education, occupation, and income.

Results

Marital Circumstances at 30

By age 30, privileged women were slightly more likely to be married than less privileged women (20 out of 26, or 77% vs. 13 out of 19, or 68%). This finding is consistent with previous research (Musick et al., 2012; Smock & Schwartz, 2020). In Table 3, we examine the relationship between marital partner earnings and women’s class background. Privileged women were more frequently married to high earners. Only privileged women married someone who made more than $500,000. Eighty-five percent of married women from privileged households married someone earning at least $50,000. In contrast, only 38% of married women from less privileged families partnered with someone making at least $50,000.

Table 3.

Marital Partner Earnings by Women’s Social Class of Origin

Partner Education	Privileged Women	Less Privileged Women
Very High (500K+)	3 (15%)	----
High (100–250K)	6 (30%)	4 (31%)
Medium (50–99K)	8 (40%)	1 (7%)
Low (0–49K)	3 (15%)	8 (62%)
Total Married	20	13

There are also class differences in partner level of education (see Musick et al., 2012). As Table 4 indicates, all but one privileged woman married an individual with at least a four-year degree. By comparison, nearly a third of women from less privileged families married someone who did not have a four-year degree. Women’s access to college-educated partners is associated with their own level of education; three less privileged women did not obtain a BA, and two of these women partnered with men who did not have a BA. A greater percentage of women from privileged families married partners with advanced degrees (30% vs. 23%).

Table 4.

Educational Attainment of Marital Partner by Women’s Social Class of Origin

Partner Education	Privileged Women	Less Privileged Women
Less than Bachelor’s Degree	1 (5%)	4 (31%)
Bachelor’s Degree	13 (65%)	6 (46%)
Advanced Degree	6 (30%)	3 (23%)
Total Married	20	13

Table 5 examines the class background match between women and their partners. Three-fourths of privileged women and 69% of less privileged women married someone from a similar class background. Only five women from privileged families and four women from less privileged formed cross-class marriages. A high degree of class homogamy concentrates multigenerational economic resources (see Schwartz, 2013 for a review of the assortative mating literature). Our sample was also marked by homogamy along race and nationality; only one woman (Crystal) partnered with an individual who did not identify as white. All but four women partnered with individuals born in the United States.

Table 5.

Class Background of Marital Partner by Women’s Social Class of Origin

Partner Class Background	Privileged Women	Less Privileged Women
Privileged Partner	15 (75%)	4 (31%)
Less Privileged Partner	5 (25%)	9 (69%)
Total Married	20	13

Class and the Marital Benefits of College

As we detail below, class shaped marriage opportunities through the persistence of hometown ties, social closure in college, and access to different post-college marital markets. As women moved through the life course, these processes worked together to create divergent marital prospects for privileged versus less privileged women.

Hometown ties.

Almost a full third of the (married) women married someone they met—or reasonably could have met—in childhood or adolescence (see the Appendix). Privileged women were more likely than less privileged women to marry from within hometown networks (35% to 23%). The persistence of these early ties, despite college attendance, contributed to the persistence of class differences in marital outcomes.

As children, privileged women lived in pricey white neighborhoods and attended schools desired by other affluent parents. The boys in their social networks were also white and affluent. For instance, upper-class Melanie described attending summer camp and sharing high school friends with the upper-class man she ultimately married. Ben grew up in a Chicago suburb 10 minutes from her home. Virtually all of their privileged white peers attended a four-year university. Schools like Midwest U draw heavily from out-of-state communities like theirs, making it unsurprising that the two ended up at Midwest U. They even selected the same dormitory, known among privileged families as a “party dorm.” Here the pair pieced together their shared connections, leading to a romantic relationship that would, well after college, result in marriage.

Melanie did not rush into marriage. She, like other privileged women, employed a “watch and wait” strategy to ensure that their potential husbands would pan out as earners. After college, Melanie and Ben moved to Chicago and enjoyed big city social life together. As she noted in the wave 5 interview, her “dream [was] not to have to work,” so she was not ready to marry until Ben completed a law degree; with the degree in hand, it was clear that Ben stood to inherit his father’s debt collection company. As Melanie explained, she waited because she wanted to marry someone who was “definitely motivated… I don’t really like the real lazy type.” By wave 6, Ben had proven himself. They were married, Melanie became pregnant, and she was considering leaving her position as a teacher to be stay-at-home mother.

The web of connections between privileged women and their partners was striking, even when women did not meet directly through family or friends. Erica, who was on a steep ascent in her business career, met her future husband on an online dating site. James was also from an affluent family. She was drawn to him because of ties they shared. As she noted in her wave 6 interview: “You can see about how many mutual friends you have, and we had like over a hundred. He sends me a message, ‘Are you an Adams High School student?’ And I am like, ‘Yeah, who are you?’ He’s from my high school, from my neighborhood, one year younger than me…. We start talking,… dating, and pretty much immediately… hit it off.” Erica continued:

My grandfather used to own a boy’s camp, and my family’s house is on the same lake [as] the camp…. James went to that boys’ camp for five years. So, it’s like not only were we in the same high school and definitely crossed paths there, but we were probably at camp dances together…. His grandparents live in the same subdivision as my grandparents. His grandma went to elementary school with my papa. His great aunt and uncle went on vacation with my grandparents. I know his cousins…. It’s like a million different things.

Thus, even though Erica had not previously met James, when she found him on Tinder he was far from the “perfect stranger” Rosenfeld and Thomas (2012) envision people meeting online.

Growing up in privileged communities meant that women like Erica and Melanie had better odds of connecting with a partner who would move into a lucrative professional career. As noted above, Melanie’s husband Ben was positioned to take over his father’s successful debt collection company. James was a pharmacist. Privileged women, even privileged women who did not marry by age 30, often had “promising men” like Ben and James in their orbit.

Less privileged women, by contrast, did not have childhood ties with those likely to become high earners. The men in their rural white communities, with whom they shared geographic and cultural proximity, experienced barriers to college attendance. Thus, when less privileged women attended college, they typically left the men in their lives behind. At the start of college, they still envisioned marrying these men. Around half of less privileged women routinely drove home from college to see their boyfriends. As Stacey recollected of her hometown boyfriend during her second year of college, “He’ll be like, ‘I want to see you. Come home.’”

Hometown boyfriends were comfortable for less privileged women. As Megan described in her wave 3 interview, just before marrying her first husband: “We come from backgrounds that are similar. He’s from the country and has been on a farm for all his life, and that’s the way I am…. My family knew their family, and I know their family. I knew Bear real well. That’s his nickname, Bear.” Unfortunately, hometown boyfriends frequently did not share women’s ambitions. As Megan described a year later, shortly before her divorce, Bear had never thought about college for himself and was angered by her educational trajectory, “He wants me home. He wants me to be inferior to him…. He wants to have control over me and… to feel like he’s the dominant one…. The fact that I’m going to school and he knows I’m smart and he knows that I’m capable of doing anything that I want… it scares him.”

Megan’s upward mobility generated tension that contributed to the demise of this relationship. As her case illustrates, when less privileged women entered college, they expressed interest in a quicker path to marriage than their privileged peers (also see Hamilton and Armstrong 2009). They turned to culturally similar men—as would more privileged women well after graduation—but less privileged women were often forced to make a hard choice between men who shared their biographies and upward mobility, as men from home were not tracking alongside them educationally or professionally.

The jobs typically held by hometown men without college degrees in small Midwestern towns were lower paying and often involved manual labor. For instance, Olivia’s husband worked on the line at a meat-packing plant. Even when less privileged women’s partners brought in substantial salaries, these men were more likely to have financial challenges. Alyssa’s cohabiting partner Alex (also from a working-class background) was a medium earner, but he had previously declared bankruptcy (due, in part, to medical costs related to a chronic illness). He also paid child support for his daughter from a prior relationship. The couple did not share financial resources, despite having a child together. They decided not to marry as Alyssa did not want to inherit Alex’s financial baggage or have her earnings included in calculations for child support.

Given the economic limitations of hometown men, most less privileged women eventually ended these relationships as their own mobility prospects improved—pushing back their timeline for family formation and forcing a change in their romantic preferences. For instance, during her final year of college Alana decided that she and her hometown boyfriend of many years, whom she once envisioned marrying, needed to stop dating. The two cried together over the breakup, but “we were in two different places…. He doesn’t have the drive that I need someone to have.” Meeting men with this “drive,” however, would prove challenging.

Social closure in college.

Six women of the 33 married women (almost 20%) partnered with a college boyfriend (see the Appendix). Five of these marriages (Blair, Brenda, Lydia, Melanie, and Tara) were facilitated by Greek life and a sixth developed in the exclusive social world that formed around college athletes (Bailey dated and married a football player). All but one of the women involved (and all of the men) were from privileged backgrounds.

MU Greek life, populated almost exclusively by privileged students, served as the sort of “efficient marriage market” described above. For example, Lydia first met her husband when her sorority paired with his fraternity for a social event. Ethan was a good-looking informatics major, with lucrative career options. Tara connected with her future husband Chase, a business major, through Greek life in her first year in college. As she explained in the wave 1 interview, she first “met all the freshmen guys in [the fraternity] pledge class” and then “they introduced me—you know, just everyone introduced me” to Chase, who was a few years older.

At MU, like many other residential universities, Greek life dominated the heterosexual social scene. As first year students, women complained that fraternity parties were the only way to meet men (see Armstrong, Hamilton, & Sweeney, 2006). Fraternity parties facilitated casual sexual pairings—not relationships. Routine co-ed social activity, beyond attending parties, required sorority membership. For instance, in her second year of college, Nicole, who had opted out of rush her first year (only to join later), described what it was like to be non-Greek during a major campus social event: “It was all for the Greeks. Like if you weren’t Greek you really didn’t belong.” Social closure in Greek life was supported by residential segregation and was nearly complete. As Whitney explained during her last year of college, “It’s really weird to think, where do the thirty-some thousand students live? It baffles me just ‘cause I’m on [Greek row] and all I see is frats and sororities.”

Predominately white Greek organizations selected on class and racial privilege (see Armstrong & Hamilton, 2013; Hamilton & Cheng, 2018). Only four of the 19 women who joined sororities were less privileged—and all four were from middle class families, as opposed to lower-middle class or working-class families. The middle-class women who joined sororities never really felt like they belonged. As Emma, who was keenly attuned to the difference between her economic resources and those of her peers, observed of the recruitment process in wave 4, “They bring their Chanel purses or their Prada purses, or whatever they are wearing usually gives it away…. [The house members] can relate better to people with money because they have money. That’s how girls like that keep continuously being brought into [this sorority].”

Women in class and race exclusive Greek organizations met and dated men in similarly exclusive organizations, or men that could be vouched for by their Greek friends. These were the primary ties privileged women formed during college that led to marriage. Less privileged women—even middle-class women (with the exception of Blair, whom we return to later)—did not have enough access to Greek life, or fit in without enough ease, to form lasting romantic relationships with men in fraternities.

When less privileged women (especially working-class and lower-middle-class women) were actively searching for a partner at the university, they were rarely in the same social spaces as college men from privileged backgrounds—unless in a service capacity. Without discretionary income or time, these women were the ones slinging the drinks and serving the food. As Heather reflected during her wave 3 interview, after transferring out of Midwest U, “I missed out on a lot of crap because I had to work. Like, [on the week of the biggest social event of the year] I didn’t go out once… I worked at [a chain restaurant as a waitress] and I made good money, but…I had to work really late at night and on the weekends I had to work 12 to 13 hour shifts.”

Instead of dating college peers, less privileged women frequently dated their co-workers. For instance, Amanda dated several fellow employees of the local big box store where she worked. Alan, whom she dated on and off for several years, was seven years older than her and had not graduated from college. His earning prospects were limited. Amanda’s mother encouraged her to meet a fellow college student. As Amanda reported in wave 2, “She’s like, ‘You’re in college, and there’s twenty thousand guys on campus. I’m sure you will be able to find somebody.’ And I’m like, well, I don’t know anybody right now.” Amanda’s entire social network was composed of co-workers. She did not form a single new tie in college.

Our longitudinal analysis allowed us to see that all but two less privileged women (i.e., Amanda and Alyssa) eventually left behind the men that they met in service positions as they pursued better career opportunities. For example, Valerie, whose work study position was in dining, entered a long-term cohabiting relationship with the dining hall chef that continued after she graduated. He was 13 years older than her and supported a son on his limited salary. The relationship ended when Valerie left him, her job, and the town itself in search of mobility.

We combed through the transcripts and identified fewer than five cases of lower-middle-class or working-class women dating fellow MU students—even for short periods—as the barriers were substantial. None of these connections led to marriage. The MU men that less privileged women dated came from less privileged families. For instance, after breaking up with her hometown tie, Alana met Michael in the Outdoor Adventure Club. The couple moved out west after college in pursuit of seasonal work as ski instructors. Michael, however, had accrued a lot of debt during college. As a result, in the wave 5 interview Alana reported that he was “not ready to get married right now, and he has a lot of financial issues that need to be cleared up.” A year later we learned that his financial stressors and high-risk employment as a forest firefighter led the couple to break up.

Post-college networks.

After college, privileged women were geographically mobile. They followed hometown ties and college sweethearts to urban centers and met additional men with strong career trajectories. Less privileged women typically returned home to small towns where it was more difficult to locate high-earning men. Thus, although around 40% of both privileged and less privileged women married someone they connected to after college through friends or work (see the Appendix), class background continued to shape whom women met.

Privileged women received substantial financial backing from their parents in the transition out of college and throughout their twenties (see “The Bank of Mum and Dad” in Friedman and Laurison, 2019). Supported by their parents, they typically made seamless transitions to thriving urban locales upon graduation. Networks of privileged students moved to Chicago, New York, Los Angeles, Dallas, Atlanta, and other similar cities. Women lived with their college friends in hip neighborhoods where local bars hosted events for graduates from MU and similar schools. In new cities, these women interacted with high-earning college-educated men, including those they already knew. It was in this context that hometown and college connections, such as Ben and Melanie’s, moved toward marriage.

Post-college social networks could even bring together MU graduates. For instance, as Lisa described in her wave 6 interview, she met her husband Daniel for the first time at a big MU social event years after graduation. They connected at a fraternity house; Daniel had lived there when he was a student, and Lisa’s younger brother was a member. When they began talking, they realized the overlap in their college experiences. Daniel had lived in Lisa’s first year dormitory. They “had a ton of mutual friends” in college. Lisa noted that they must have crossed paths many times. Their eventual pairing is a testament to how a shared alma mater shapes marital patterns—even though most marriages only occur many years after graduation.

Privileged women were introduced to high-earning men through friends, work, or both. As Sophie detailed in the wave 6 interview, her husband attended a similarly prestigious state flagship and was a member of a fraternity where Sophie “knew a lot of the same people.” They were introduced when “Clark was in med school with a few guys that I’d grown up with. Just being in [my large urban city], I’d gotten close again with those guys.” Clark was a pediatric anesthesiologist and Sophie a neo-natal nurse, so they also shared professional interests.

As these examples suggest, privileged women’s post-college dating markets reflected lifelong access to others from affluent families, whose odds of making it to college and enjoying career success were already amplified. Many of these women took for granted that they would always be surrounded by “promising men.” As Tara remarked in her wave 5 interview, “I feel like social classes fit together. The people I run around with come from money, and they’re all on the right path to finding good careers… I don’t feel like I would have to go and search for [them]. It’s what I hang out with or what I am attracted to, but it is important to me.”

In contrast, the marital markets of less privileged women were comparatively thin. Most moved back home to small towns. This often meant partnering with non-college educated men who were low earners and had financial baggage—or opting out of relationships altogether. For example, when Emma moved back home, she despaired of her dating prospects. As Emma noted in wave 6 of the men in her post-industrial Midwest hometown, “The selection is slim pickings, if you ask me…. They’re just not people that I would date.” Emma remained unmarried because she could not find a partner that she believed would contribute to, rather than threaten, her tenuous economic position.

A few less privileged women found stable earners in their communities. However, these women located college-educated men who had arrived from outside of town, but with whom they shared cultural and geographic roots. For instance, Karen married a fellow teacher in her rural Midwestern town. He had grown up in similar area about an hour south. After two difficult relationships with hometown men—including her relationship with Bear—Megan connected with Bobby in her graduate counseling program. Bobby shared Megan’s love for the country, but also her ambitions. Yet, while men like Bobby were the most “promising” in the area, they were still relatively low earners and not always possible to find.

Exceptions.

Examining privileged women partnering with low earners and less privileged women partnering with high earners sheds further light on how unequal class marital outcomes developed in our sample. These are the exceptions that prove the rule—that is, they are cases in which proximity, networks, and preferences operated to produce a less common outcome. Collectively, exceptions reveal the centrality of heterosexual privilege and highlight that the everyday flow of interactions from childhood, through college, and beyond more often than not channel people into class homogamous marriages.

Three privileged women married low earners. Two of these women, Brianna and Leah, married women. Lesbian identity reduces the likelihood of having a higher earning partner, as women generally earn less than men. Although individual lesbians may earn more than straight women, the household income of lesbian couples is lower on average than that of heterosexual married couples (for a review see Ahmed, Andersson, & Hammarstedt, 2011).

Nicole, the sole privileged heterosexual woman who married a low earner, illustrates the consequences of not having enough resources to track alongside “promising men.” Her upper-middle-class parents exhausted their resources paying for her out-of-state degree and could not underwrite a life in New York City. Nicole was unable to capitalize on ties to advantaged peers formed through Greek life. She lived at home, outside the city, where her social networks consisted of high school friends with limited economic prospects. As Nicole described in the wave 5 interview, “None of the boys have steady jobs.” By wave 6, she had married a man with a lackluster academic record whom she knew from a previous job as a summer camp counselor. Nicole had imagined that Zac would “go do something with his dad” (a successful entrepreneur), but Zac floundered, working retail until starting as a teaching assistant making $40,000. Nicole was disappointed, stating, “We’re definitely not as financially comfortable as we’d like to be.”

Four less privileged women (Blair, Carrie, Crystal, and Valerie) married high earners. All four, with varying degrees of intentionality, successfully disrupted their class networks. As these cases suggest, less privileged women met “promising men” only if they reconfigured their social networks—often a daunting task. Above we noted that Blair was one of the few less privileged women to join a sorority, which led her to meet and marry a fraternity man from a background more privileged than her own. Beauty and interactional skill enabled her to access more privileged marital networks. There were costs to her efforts, though, as she was miserable in college, continuously aware that she was different from her privileged peers. As she noted of her sorority sisters in wave 5, “All their dads were doctors and lawyers, and I just didn’t fit in.”

Carrie tried online dating. It worked. As she explained in wave 6, “I met accountants, pharmacists. I was like, this is awesome. They were nice, normal guys.” She met Thomas, a lawyer who had graduated from Midwest U, a year before she did. Unlike Lisa and Daniel, who shared network ties at MU, Carrie and Thomas’ social worlds in college had not overlapped. The shared degree created a common experience, which helped them to connect online without prior ties. Yet few women leveraged the capacity of online dating to extend networks. Only six participants married someone they dated online, and most—particularly the privileged—used online dating to verify that the man was already in-network.

Geographic moves were another way to disrupt networks. Some less privileged women moved to cities, where they could make new ties. Valerie and Crystal both did this. Valerie’s move—alluded to above—was particularly intentional. For years after graduation, she worked in student affairs at MU earning $44,000; during this time, she cohabited with the dining hall chef, mentioned earlier. She did not see him as viable long-term partner, nor did she see other dating possibilities in her small town. As Valerie noted in wave 6, “I turn 30 next year and I don’t want to be waiting around here, especially if I’m gonna be single. [This town] is not a good place to do that if you’re not 22 or younger.” In a check-in, we confirmed that she moved to Chicago to improve both her career and dating options. She rapidly met and married a man working in IT—one making more than three times that of the college-town chef.

The marriages of Blair, Carrie, Crystal, and Valerie raise the question: Why didn’t other less privileged women do what they did? Others tried and were not successful. Alana also disrupted her geographic networks by moving west with her college boyfriend. Yet Michael, and all the men she dated after him, were as economically precarious as she was. Whitney moved to Chicago, where she barely had enough money to live. As she noted in wave 5, “It’s still paycheck to paycheck.” She eventually moved home, without meeting a promising man.

In addition to being risky, disrupting networks incurred emotional costs. Moving, leaving behind family and friends, and trying to fit into new cultural worlds are hard, as Blair’s story illustrates. While some women were willing to incur these costs, other less privileged women sought economic security without having to rupture ties. Class privilege ensures that affluent women do not have to rewire networks to marry high-earning men.

Finally, the high-earning men that less privileged women married were not the same as the high-earning men many privileged women married. Crystal’s husband was from a hard-working immigrant family, Valerie’s husband was not American, Carrie’s husband was upwardly mobile from a less privileged background, and Blair’s husband’s family turned out to be much less affluent than Blair had assumed. Less privileged women never gained access to men with the levels of privilege held by Ben and James—the husbands of Melanie and Erica.

Discussion

We uncovered mechanisms driving class disparities in marriage outcomes among the college-educated that are not visible in existing survey data (e.g., see Musick et al. 2012). As we detail above, women’s childhood schools and neighborhoods were segregated by class and race. Class differences in networks persisted through college, as social life at the university was also segregated. After graduation, the influence of differing social networks, class disparities in the ability to move to urban areas, and class variation in whom women encountered socially and at work combined to provide women from privileged and less privileged families access to divergent marital markets. Even online dating was used by privileged women to search existing networks, rather than to extend or disrupt their social worlds. Thus, as a result of class network continuity across the life course, we observed less class equalization in access to high-earning spouses than would otherwise be expected.

We also explored exceptions—privileged women married to low earners and less privileged women married to high earners. These cases revealed that heterosexual privilege is key to partnering with high earners and highlighted the depth of resources necessary to remain in-network with advantaged men likely to become high earners. We learned that for less privileged women to marry high earners a difficult process of network disruption and reconstruction was necessary. It was here that online dating facilitated matches that would not otherwise occur.

Overall, the social worlds of class privileged women might be envisioned as a river, where they swim alongside privileged boys who become privileged men, as the river moves from childhood, through college, and into pricey urban neighborhoods, and, in some cases, shared professional worlds. Less privileged women swim in a different river, alongside boys who face structural barriers to achieving adult economic security. As our exceptions reveal, some upwardly mobile women switch rivers. But this is no small feat—and attending college does not, in and of itself, lead this switch to occur.

Our findings suggest that in their enthusiasm to emphasize the historical decline in the role of early ties in union formation, Rosenfeld (2009) and Christakis and Fowler (2009) may have overstated the case, at least for class privileged individuals. The evidence that “the locations and circumstances under which people meet partners have been changing over the past century” (Christakis and Fowler 2009, p. 67) is persuasive. But a decline over time in the role of early ties does not erase the importance of these ties. Closely guarded networks continue to be a defining feature of upper-class life (see Ostrander, 1984).

That Musick et al. (2012) find similar patterns in representative data suggests that it is not only this university that maintained classed networks. College is often thought to be an equalizing force. However, research has long documented that college frequently operates as a class preserving institution (Karabel, 2005), in which historically white Greek organizations play a key role in preserving class and race boundaries (Hamilton & Cheng, 2018; Park & Kim, 2013).

Just as research has documented organizational variation in the ability of universities to provide intergenerational income mobility (Chetty et al., 2017), we predict similar variation in the extent to which universities disrupt students’ classed social networks. College is one of the few points in life where it is realistic or possible for large number of people to switch rivers—that is, to reconstruct their whole networks. Whether or not college succeeds in doing so is highly consequential, and not only for marriage.

What might increase the possibilities of network disruption? First, privileged and less privileged students need to attend the same schools. In our current postsecondary system, large flagship universities are one of the few types of four-year schools that bring together substantial numbers of youth from a wide array of backgrounds. As our study suggests, however, this is not enough. Universities also need to create conditions in which students can interact positively across class. For example, schools should consider “difference-education interventions” that build understanding of how students from dissimilar class backgrounds might experience college (see Stephen, Hamedani, & Destin, 2014). Universities could also engineer living arrangements to be socioeconomically diverse and support these units with programming.

Schools that allow—or even encourage—class-segregated social, athletic, and academic activities will predictably generate class-segregated romance. As we noted, historically white fraternities and sororities are typically exclusionary along multiple dimensions. But other organizations may be class exclusionary as well. Rivera (2015) points out that country club sports like rowing and lacrosse are so strongly associated with affluence that they are used by elite employers as a proxy for class background. We encourage universities to take steps to eliminate or reconfigure organizations and activities that are class (and often racially) exclusive. However, increasing reliance on tuition and donations limits organizational willingness to disrupt activities valued by wealthier families (Armstrong and Hamilton, 2013).

The adult lives of the women that we studied, privileged and not, continued to be segregated not only by class, but also by race. They lived in predominately white neighborhoods and sent their children to daycare and schools that included few racially marginalized families. These decisions actively work against racial equality (Lewis & Diamond, 2015). Research on colorblind racism highlights the importance of recognizing what may be seemingly nonracial choices in reinforcing the post-Jim Crow racial order in the US (Bonilla-Silva, 2015).

Although we focus on the experiences of white women, there is emerging work on the romantic experiences of highly educated Black women (e.g., see Clarke, 2011; Ford, 2018). As women often marry within race, the structural racism encountered by Black men makes it difficult for Black women to locate “promising men.” We need more research on how individuals in a variety of intersectional locations navigate marriage (see Lichter, Price, & Swigert et al., 2020).

This article contributes to a broader discussion about the ways in which a four-year college does, or does not, equalize. Scholars tend to focus narrowly on earnings as evidence of the potential for class equalization offered by college. However, marital prospects—along with other factors, such as student debt, ongoing parental support, and geographic mobility—are key to women’s life chances. Scholars should continue to develop greater understanding of how advantage or disadvantage across multiple domains are involved in reproducing inequality.