Women’s underrepresentation in male-dominated fields of study is often attributed to a gender gap in success expectations. Prior research shows that female students report lower confidence in their academic abilities and future success than men, not only in male-dominated but also in female-dominated and gender-integrated fields. These lower success expectations can negatively affect women’s academic paths, making it important to understand the factors that shape them. Although existing studies highlight the positive role of social experiences in shaping women’s success expectations, particularly in male-dominated fields, women are also more likely than men to experience discrimination, which may undermine these benefits. However, the role of discrimination experiences in shaping success expectations in higher education remains underexplored. This study examines how female students’ discrimination experiences contribute to the gender gap in students’ success expectations and investigates whether the relationship between discrimination experiences and success expectations among female students varies across fields of study with varying gender ratios. Using data from 9,512 students collected as part of the Student Survey in Germany (SiD), we analyze female students’ discrimination experiences and their associations with success expectations in male-dominated, female-dominated, and gender-integrated fields of study. The results show that women, particularly in male-dominated fields, report higher levels of discrimination than men. These experiences contribute to the overall gender gap in success expectations. However, detailed analyses reveal that only relational discrimination experiences, such as social exclusion or unfair treatment, are significantly associated with female students’ success expectations, revealing a seeming paradox: this association emerges primarily in female-dominated rather than male-dominated fields, even though female students experience more discrimination in the latter. This suggests that the consequences of discrimination for success expectations are context-dependent and may reflect different coping strategies across disciplines. Overall, our findings highlight the importance of distinguishing between types of discrimination and contexts when examining gender disparities in success expectations. Furthermore, they underscore the need for targeted interventions addressing relational discrimination to support women’s academic confidence in higher education.
Gender disparities in academia persist worldwide, yet empirical evidence remains limited on where, at which stages, and to what extent women exit academic careers. Using a uniquely comprehensive, decade-long national dataset from Japan (2014–2023), this study provides system-level evidence of gendered attrition across six disciplinary fields and four academic ranks. By distinguishing among the potential pool, applicant pool, and hiring outcomes, the analysis separates application-stage processes from hiring-stage outcomes—an analytical distinction rarely possible in other national contexts due to data limitations. The findings indicate that the application stage constitutes the primary bottleneck in women’s academic careers. Across fields, women’s underrepresentation appears to be driven less by disadvantage at the point of selection than by the fact that many qualified women do not apply for academic positions. Once applications are submitted, hiring outcomes are generally neutral or, in some cases, favorable to women, with lecturer-level recruitment showing evidence of proactive intervention. Rank-level analyses further indicate that women’s attrition is pronounced at career entry and continues at mid- and senior-career stages in most fields. Field-level analyses highlight substantial heterogeneity in how and when attrition occurs. By separating application-stage attrition from hiring outcomes and examining patterns across ranks and disciplines, this study refines existing accounts of gender inequality in academia. The results highlight the importance of organizational and structural factors shaping application behavior and cumulative disadvantage, and underscore the policy relevance of collecting detailed applicant-flow data. Overall, the findings suggest that stage-specific and discipline-sensitive interventions are essential for addressing persistent gendered disparities in academia.
Adopting an “equal but different” perspective, this study investigates long-term changes in gendered patterns of university enrolment and field-of-study choice in Italy, focusing on the care-technical divide among cohorts of upper secondary school graduates from 1995 to 2022. Leveraging a unique longitudinal harmonization of survey data from the National Institute of Statistics with administrative educational databases, the analysis accounts for gender differences in prior secondary school track choice and academic performance. Results show that the female advantage in Higher Education (HE) participation largely disappears once differences in prior schooling are considered, suggesting that the gender reversal in enrolments reflects gendered educational pathways rather than a net advantage for women. However, a modest and persistent female advantage emerges in the second half of the 2000s, after the incorporation of traditionally feminized fields such as teaching and nursing into HE. Most importantly, gender disparities in field-of-study choice, particularly along the care-technical divide, have remained strikingly stable across cohorts, despite profound changes in overall participation levels. These findings highlight the resilience of horizontal gender segregation in HE and call into question the extent to which the expansion of women’s access to university has translated into more gender-equal educational outcomes.
This paper examines the relationship between gender differences in nonpecuniary workplace preferences and human capital investments. Using survey data on traditional bachelor’s degree-seeking first-time undergraduates, I document trends in the relative importance of five workplace preferences (balancing work and family, balancing work and leisure time, being an expert, helping others, and making decisions) to salary. I find that workplace preferences became less important relative to salary as students approached the workforce. Second, I find that workplace preferences, particularly preferences for helping others as a part of one’s job, were positively associated with female-dominated majors and occupations (e.g., education and health care fields) and negatively associated with male-dominated majors and occupations (e.g., business and STEM fields). Third, I find that preferences for helping others explain a significant portion (12
Academic mobility is a central issue in the study of academic professions, shaping career trajectories, institutional dynamics, and the distribution of knowledge production. While international research has explored its determinants and implications, evidence from Global South contexts remains scarce. Brazil, with one of the largest higher education systems in the world, provides a valuable case to examine how institutional, sociodemographic, and policy factors shape academic mobility. This study investigates the employment trajectories of professors affiliated with Brazilian higher education institutions between 2003 and 2021, based on microdata from the Annual Social Information Report (RAIS). The analysis accounts for gender, age, degree, institutional type, distance, and remuneration. Nearly half of faculty (46.9
Universities are increasingly expected to translate research into economic and social value, yet the policy designs through which they govern technology transfer vary across higher education systems. This study compares 695 currently operative institutional documents on university technology transfer from the United States, the United Kingdom, Australia, and China. Using text mining, rule-based coding, and manual content analysis, it examines five governance dimensions: transfer modalities, funding mechanisms, ownership and IP governance, organizational arrangements, and revenue distribution. The findings show convergence in the centrality of IP protection and researcher incentives, but also complicate simple state-versus-market typologies. U.S. universities combine market coordination with public accountability and professionalized intermediation; U.K. universities link commercialization to impact-oriented accountability infrastructures; Australia relies on procedural, partnership-based, TTO-centered coordination; and China combines state steering with high inventor rewards and selective market-style mechanisms. Across systems, revenue distribution is not the dominant governance signal; policy attention concentrates more strongly on IP governance, transfer pathways, and organizational coordination. The study offers a configurational account of technology transfer governance, showing how universities adapt to innovation pressures through nationally differentiated combinations of state, market, and academic logics.
This study examines university dropout by focusing not on students’ intentions but on those who have effectively discontinued their studies. Data were collected using an ad hoc questionnaire administered to a sample of 2183 former undergraduates from four Spanish public universities who dropped out between 2010 and 2020. The instrument combined sociodemographic and academic information with self-reported accounts of satisfaction, engagement, reported withdrawal-related circumstances, and decision-making processes. Principal component analysis reduced the variables to eleven dimensions, which were subsequently used in a two-step cluster analysis to identify distinct patterns within students’ retrospective accounts. Three interpretive profiles emerged. The first comprises disengaged and misaligned students, characterized by low satisfaction and engagement and by retrospective attributions related to institutional or program-level shortcomings. The second includes low-commitment and externally influenced students, who show uniformly low levels of engagement and whose withdrawal narratives appear weakly grounded in personal or academic motivations. The third profile encompasses constrained but committed students, who report high levels of satisfaction and engagement but retrospectively associate their withdrawal with external constraints such as economic pressures, work–study incompatibility, or family responsibilities. These findings illustrate how former students retrospectively interpret and organize their dropout experiences and underscore the importance of considering students’ own accounts of their academic trajectories. More specifically, the results suggest that a substantial proportion of withdrawal narratives are structured around constraints that participants perceive as being beyond their individual control, pointing to the potential relevance of equity-oriented support policies and institutional flexibility measures. Overall, the study contributes to a more differentiated and context-sensitive understanding of how former students reconstruct and make sense of dropout processes in higher education.
Compared with other high-income countries, the ratio of women among all researchers is particularly low in Japan. In academic STEM fields, especially physics and engineering, this figure is even worse. Against this backdrop, Japan’s top-ranked national universities are implementing direct affirmative hiring (DAHW) for women researchers, with the aim of mitigating gender imbalance in STEM fields. However, this rapidly emerging hiring policy is caught between two incompatible institutional logics: diversity promotion and meritocracy. Drawing on semi-structured interviews and abductive thematic analysis, this research examines the reasoning STEM faculties deploy in justifying their support of or opposition to DAHW, and the potential consequences of DAHW for women. Overall, the analysis reveals stronger support for meritocracy than for diversity promotion, and that the policy increases negative effects, such as low self-esteem and discriminatory experiences, for women. Based on the analysis, we recommend that academic communities put in more effort not only to increase the pool of female researchers but also to actively seek out potential job applicants.
Branch campuses have become a global phenomenon in higher education. This study investigates whether the establishment of elite cross-city branch campuses enhances or crowds out the scientific research performance of incumbent local universities in China. Leveraging the staggered entry of 44 elite branch campuses across 26 cities as a natural experiment, we employ a staggered difference-in-differences (DID) and a triple-difference (DDD) design using a comprehensive panel dataset from 1994 to 2020. Our results show that the establishment of cross-city branch campuses significantly increases scientific research output of local universities, with a 12.9
Using the High School Longitudinal Study of 2009 (HSLS:09) baseline, transcript records, and the Postsecondary Education Administrative Records (PEAR; degree completions through 2021), we examine whether taking AP Calculus, Biology, Chemistry, Physics, or Statistics predicts earning a postsecondary credential in four STEM fields: health professions; biology, life science, and agriculture; physical science and engineering; and mathematics and computer science. To address selection bias, we estimate average treatment effect on the treated models using propensity score weighting, assess balance, and fit weighted logistic regressions, with subgroup analyses by sex, race/ethnicity, and their intersection. Results indicate strong course–field alignment: AP Biology is associated with substantially higher likelihood of earning a biology/life science/agriculture degree and AP Calculus and Physics are positively associated with earning a physical science/engineering degree. AP Statistics shows no consistent relationship with STEM degree attainment. We also find heterogeneous relationships by sex and race/ethnicity, including negative associations with health professions for some AP math/science courses.
High-impact practices (HIPs)—structured, effortful learning experiences such as internships, undergraduate research, service learning, study abroad, learning communities, and capstone projects—are widely promoted as mechanisms to strengthen student engagement and academic achievement. However, in stratified higher education systems, participation in HIPs may yield unequal academic returns. This study examines these dynamics in Colombia by identifying latent participation pathways among 1689 final-year undergraduate students enrolled in accredited private universities. Using latent class analysis (LCA), four participation configurations were identified—comprehensive, academic-focused, practical, and minimal—with systematic variation by gender and academic discipline. Higher cumulative grade point averages (GPAs) were observed only among students in the comprehensive pathway, indicating that academic value depends on the coherent alignment of multiple academically integrative practices rather than on isolated participation. No significant GPA advantage emerged for partial or practice-oriented configurations. These findings indicate that institutional and disciplinary ecologies condition whether experiential learning translates into academic achievement, challenging universalist assumptions about HIP effectiveness. The study reframes HIP participation as a configuration-dependent opportunity structure and highlights the limits of access-based evaluation approaches when HIPs are scaled within unequal higher education systems.
A growing number of public universities in the US have adopted a decentralized budget model known as responsibility center management (RCM). Compared with the traditional centralized budget model, RCM gives greater budgetary responsibility and decision-making to individual academic units, meaning that units take on greater roles in generating revenue and managing costs. Leveraging the staggered adoption of RCM across public universities, this study finds that RCM is positively associated with the production of master’s degrees, but negatively associated with the production of doctoral degrees. Field-level differences are found, with the positive effect on master’s degrees most evident for STEM, a high-demand field. Transition to a decentralized budget model appears to affect how academic units approach degree production, potentially leading to a greater focus on revenue generation, with the possibility of reinforcing resource disparities across academic fields.
This study examines how participation in dual enrollment (DE), including variation in course subject and instructional setting, is associated with students’ postsecondary access, initial enrollment level, and early credential attainment during the initial postsecondary years. Using nationally representative data from the High School Longitudinal Study of 2009 (HSLS:09) and grounded in the academic momentum framework and informed by Social Cognitive Career Theory as a secondary interpretive perspective, the analysis investigates how different DE experiences relate to students’ postsecondary trajectories. Logistic and multinomial regression models, combined with propensity score matching and multi-treatment weighting, were used to compare outcomes among students who completed different types of DE courses. Findings show that DE participation is associated with a higher likelihood of postsecondary enrollment and early credential attainment within the observation window, with variation across course subjects and instructional settings. Mathematics DE was associated with comparatively more favorable early credential completion patterns in within-DE comparisons, although subject differences were modest overall. In addition, DE offered on college campuses or online was associated with higher odds of early bachelor’s attainment within approximately 3 years after high school, whereas high school-based DE showed weaker patterns. These findings highlight the importance of attending to both the subject and setting of dual enrollment when designing policies intended to promote equitable postsecondary pathways.
Drawing on administrative application data from a selective public flagship university, combined with external datasets on community socioeconomic indicators and high school characteristics, this study investigates whether and how final test-score disclosure decisions among applicants who initially submitted test scores are associated with their educational contexts under test-optional admissions. Using linear probability models and subsample analyses, the findings show that final test-disclosure behaviors are associated with high school characteristics, peer reference patterns, and community income, after conditioning on students’ test scores, individual backgrounds, and other application materials. In most cases, these associations appear broadly consistent across the score distribution, with limited heterogeneity by test score level. These findings offer practical implications for admissions officers, emphasizing the importance of contextual awareness in evaluating applicants under test-optional policies.
US colleges and universities are increasingly recognizing and supporting first-generation students—those whose parents have not earned a bachelor’s degree and who are therefore the first in their families to attend college. To investigate the extent to which first-generation student support has expanded and what factors have facilitated the expansion, I draw on original data collected from a nationally representative sample of 234 US higher education institutions. I find that 35.9
Student evaluation response rates have declined sharply in recent years, raising concerns about the representativeness of data used for high-stakes personnel decisions. We provide evidence that a simple, system-level incentive substantially increases participation. That incentive is early grade access upon evaluation completion. When one large public university inadvertently removed this incentive in Fall 2016, response rates fell 29 percentage points, from 66 percent to 37 percent, over three years. Using panel data covering 438,504 student-course observations from 13,683 students observed both before and after the policy change, we estimate a within-student decline of nine percentage points. Comparison institutions with no policy change show stable response rates over the same period. We find only small differences in the policy’s impact across student demographic groups, suggesting the incentive removal affected participation relatively uniformly. Our findings demonstrate that grade-access incentives represent a low-cost, scalable intervention for increasing evaluation participation.
The role of social connections in supporting student retention in higher education is well-documented. Strong networks are associated with better academic outcomes and reduced dropout rates, but the COVID-19 pandemic disrupted traditional patterns of social interaction. This study examines the relationship between social networks and student retention, comparing cohorts from before the pandemic and during the pandemic in higher education. Drawing on Tinto’s theory of academic and social integration, as well as research on the role of social networks in fostering belonging and academic success, we analyze data from two cohorts of undergraduate university students: one beginning their studies in 2017 and the other in 2020 during COVID-19 restrictions. Using social network research, survey data, and student management system data collected over the three years of undergraduate studies, we identify significant differences in the size and strength of social networks across cohorts. While the 2020 cohort reported fewer connections at the beginning of the study, retention rates remained comparable. Findings show that early social ties strongly predict on-time study completion for both cohorts. In the 2017 cohort, the expected probability of students finishing their studies increased by around 10 percentage points for students with 1–3 connections compared to students with no prior connections, and by around 22 percentage points for students with 4 or more connections. In the 2020 cohort, the predicted probability of students finishing their studies increased by around 16 percentage points for students with one or more connections, compared to students with no prior connections (the difference between 1 and 3 connections and 4 or more was not significant). These insights highlight the need for higher education institutions to prioritize social and academic integration, especially during times of crisis, to ensure students can succeed despite challenging circumstances.
The increasing connection between higher education and the workforce may influence gender-based educational and occupational segregation. If men and women continue to gravitate toward fields historically dominated by their own gender, this segregation will persist. Furthermore, the chosen field of study carries implications for labor market mismatch, as certain majors are more prone to producing graduates who struggle to find positions aligned with their qualifications. Using a representative sample of employed graduates from the Spanish Labor Force Survey, this paper examines the relationship between gender, gender-dominated fields of study, and educational mismatch. Conditional on labor market participation, discrete choice models reveal that studying in a gender-dominated field is associated with lower vertical and horizontal mismatch (i.e., overeducation and field of study mismatch) —a finding supported by several robustness checks. This finding is especially pronounced in horizontal mismatch when gender and field-dominance coincide. Consequently, the higher risk of mismatch in gender-neutral fields may provide a rationale for the continued persistence of gender-segregated educational choices.
This study comprehensively examines the effect of the College Scorecard (the Scorecard), first launched in 2015, on student application and enrollment at 4-year universities in the United States by reported earnings level and sector. Using institution-year panel data from the Integrated Postsecondary Education Data System, we employed a regression discontinuity in time design and an event study analysis to examine whether application and enrollment trends at 4-year institutions shifted after the release of the Scorecard. Results indicate little evidence of shifts: students did not substantially move away from lower-earning institutions or sort into higher-earning ones, and the null findings remain across earning levels and sectors. These findings align with early studies showing limited influence of the Scorecard on college choice. Finally, we conclude our manuscript by pointing out that the National Center for Education Statistics’ federal, individual-level longitudinal surveys, which are currently suspended, will serve as critical data sources for future evaluation efforts on the Scorecard and other related federal policies, including the new accountability framework recently introduced by the One Big Beautiful Bill Act. We recommend lifting the suspension of these individual-level longitudinal surveys.
