Status: Past

College graduates are facing growing challenges in securing jobs that match their qualifications, yet these challenges may not be evenly distributed across social groups. This study examines how educational credentials translate into labor market outcomes, operationalized as education–occupation mismatch, and whether the burden falls disproportionately on certain groups. We distinguish between vertical mismatch (discrepancies between workers’ educational attainment and the level required for their occupations) and horizontal mismatch (between field of study and the type of education required for their occupations). Using national longitudinal data and multiple measures of mismatch, we find that education–occupation mismatch helps explain racial, gender, and nativity-based inequalities within the highly educated U.S. workforce. Racial and ethnic disparities are pronounced, with Black, Hispanic, and Asian graduates all facing disadvantages relative to their White peers in securing occupations commensurate with their credentials, though through distinct mechanisms. Gender differences are also notable: highly educated women are more likely to experience mismatch and remain mismatched for longer durations. Advanced degrees, STEM majors, and degrees from selective institutions reduce the likelihood of mismatch but do not eliminate these disparities. Finally, high-skilled immigrants, particularly those with foreign degrees, are especially vulnerable due to limited credential transferability, language barriers, and institutional constraints such as licensure requirements. These patterns point to the rise of a college-educated working class. Their broader societal implications will be discussed.

Using deidentified administrative records for non-international undergraduates graduating from 2014–2023 at a large, diverse public research university in the western United States, this study examines how family of origin structures college major choice and how incomplete measurement of family influences may bias estimates linking majors to early occupational status. Leveraging a comparative dyad design, we assess concordance in major selection and divergence in post-baccalaureate occupational standing and post-BA educational attainment across four groups that share progressively more family context: random cohort pairs, non-kin pairs matched on observed sociodemographic characteristics, sibling pairs, and twin pairs. We operationalize occupational standing with a Hauser–Warren–style Socioeconomic Index derived from linked employment records (Steppingblocks/Lightcast), with post-BA education outcomes drawn from National Student Clearinghouse data.

Selective colleges are sites of social closure. Scholarship examines how selective colleges evaluate applicants, but colleges do not passively accept applications. They recruit. If college recruiting territories are structured by shared infrastructures, then mechanisms of social closure operate upstream of admissions. This article examines how College Board Geomarkets and the Market Segment Model structure high school recruiting visits by selective colleges. Geomarkets carve metropolitan areas into smaller geographic units meant to define local recruiting territories. The Market Segment Model predicts how student demand varies by Geomarket as a function of social class. These market devices become embedded in organizational structures and in third-party products that produce college recruiting behavior. We analyze high school recruiting visits made in 2017 by 42 selective colleges to address three research questions. We find that visit patterns are consistent with the Market Segment Model, with visits concentrated in affluent Geomarkets (RQ1); Geomarkets explain substantial variation in which high schools receive visits (RQ2); and the benefits of being located in a popular Geomarket accrue disproportionately to affluent schools and to schools with moderate shares of Black, Hispanic, and Native students (RQ3). The market devices that college admissions offices incorporate constitute a shared infrastructure that performs social closure by matching a socioeconomic hierarchy of communities to a socioeconomic hierarchy of colleges prior to the evaluation of applications. More broadly, our findings suggest that market devices can facilitate infrastructural collusion, whereby shared reliance on third-party classificatory infrastructures produces coordinated behavior without explicit coordination.

Pathways look different at an R1 university than at a community college or a small liberal arts school, yet students across all three share a surprisingly similar challenge: not knowing what’s possible for them. This talk traces the founding and evolution of Stellic, a platform built not just for students to discover and navigate their career and academic pathways, but for the entire institution to move with them. Advisors gain real-time oversight into where students are headed. Administrators unlock a new layer of signal understanding, which informs pathway demand for curriculum decisions, course offerings, and resource planning. The founder will share a live platform demo, cross-campus impact stories, and the hard-won learnings from deploying across very different institutional contexts. What works at a research university often needs reinvention at a community college. What surprises us is how much the data converges. The talk closes with a founder’s honest look at where this is going: the bets we’re making, the futures we’re building toward, and the open questions we haven’t solved yet.

Young adult workers confronting marginalization and growing precarity use cultural repertoires to interpret their situations and take action. Yet many of these cultural tools can seem contradictory, and exposure to contradictory tools can create confusing signals about how to pursue advancement. Data from life story interviews with low-income, Black-identifying youth from New Orleans, Louisiana, instead demonstrate how marginalized young adults synthesize heterogeneous cultural schemas, and how resulting cultural aggregates relate to mobility. Abductively building on theory linking cultural ends and means, evidence shows how actors combine common cultural schemas regarding self, opportunity structure, and social action. These produce shared cultural models, or ideal-typical mobility ideologies oriented around themes of inner transformation, outer mobilization, and social emancipation. Despite persistent precarity, mobility ideologies perpetuate social myths of agency among respondents by aligning self and action in three corresponding narrative identities: metaphorical gardeners, climbers, and seekers, respectively. Mobility ideologies interact and align with both actor resources and social expectations to varying degrees. Counterintuitively, cultural mismatch or incongruence between actor mobility ideologies and ecological expectations correlates with advancement when it helps respondents work against reproductive forces like stereotyping and labor market discrimination. Findings highlight how combinatory and interactive cultural processes—as opposed to cultural repertoires alone—relate to inequality.

Although general education requirements comprise one-third of U.S. four-year degree requirements, we know little about their role in shaping student course-taking pathways. Drawing on administrative data from Texas linked with general education requirements at 32 four-year institutions, I study the role of general education in U.S. higher education. I show that general education requirements dramatically shape student course-taking, particularly in the first two years of a student’s academic career. Using a difference-in-differences design, I estimate the causal effect of adding a general education designation to a pre-existing course: adding a general education designation increases course enrollments by 43 percent and departmental enrollments by 49 percent. These enrollment increases translate into increased instructional staffing and course offerings within affected departments. I argue that changing general education designations offers state policymakers a seemingly neutral mechanism that reshapes student pathways while reallocating resources and instructors according to their ideological priorities.

College majors are important stratification mechanisms, shaping occupational trajectories, postcollegiate earnings, and patterns of inequality by gender, race, and class. Yet the processes through which students and fields of study come together have received limited theoretical and empirical attention. We advance a theory of major selection as a form of organizational differentiation that both builds on and mediates parallel processes earlier in the life course. Our longitudinal conceptualization links the development of academic subjectivities in K-12 education with postsecondary organizational conditions. Compared with their prior schooling, college students exercise greater agency over their course-taking, face greater ambiguity about the purposes of school, and confront complex curricular choice architectures. These conditions are sustained by an organized anarchy in the provision of courses and majors that simultaneously enables and constrains selection. Our theory recognizes the cumulative and contingent character of major differentiation and illuminates enduring problems of stalled college progress.

AI-supported digital tools offer great promise for improving college students’ academic engagement and progress, yet little research has focused on whether and how the character of tool implementation might influence tool uptake in specific campus contexts. Our study documents variation in implementation of the same digital tool across four public university campuses in a large state system. Scaled computational observation across campuses reveals substantial variation in magnitude and style of tool use. Site visits and qualitative interviews with university staff reveal parallel variation in how the tool is embedded in technical infrastructures and inter-office divisions of labor and how data are used to inform tool use. Findings suggest that the uptake of digital tools, and consequently their sustained adoption, is partly dependent on how tools are embedded in socio-technical systems that intertwine technologies, organizational conditions, and interpersonal relationships in context- specific ways.

This presentation examines two projects looking at the decision to change majors at a large West Coast research university. In the first, we examined individual differences in changes of major relating to the gender composition of each major. While many students changed their majors, only a minority changed to one of a different gender composition. Students who persisted in or changed to a female-dominated major were more likely to cite prosocial career goals as a motivating factor in their studies. In the second project, we tracked the timing and frequency of changes in university major, finding that first-generation students make later and more frequent major changes. Relatedly, first-generation students are also more likely to enter university with an undeclared major and to wait longer to declare a major. Furthermore, first-generation students are less likely to switch to a similar major (e.g., biology to human biology) compared to other students. This has notable costs in terms of time-to-graduation and extra credits earned.

Collaborators: Nayssan Safavian (University of California, Irvine), Yannan Gao (University of Tubingen), Ann-Lena Dicke (University of California, Irvine), Sirui Wan (University of Wisconsin-Madison), Jutta Heckhausen (University of California, Irvine), Jacquelynn S. Eccles (University of California, Irvine), Richard Arum (University of California, Irvine).