Does Starting at a Community College Affect Mid-Career Wages and Occupational Outcomes?

Swiderski and McDonnell (2026) examined whether students who begin at a community college have different long-term educational, occupational, and wage outcomes than similar students who start at a four-year college. They analyzed data from the National Longitudinal Survey of Youth 1997 (NLSY97), linked to O*NET occupational data, and compared similar students using propensity score weighting. They found that community college starters were more than 20 percentage points less likely to earn a bachelor’s degree. However, wage differences were generally small—typically 2% to 8%—and often not statistically significant. Community college starters also entered occupations with similar math demands but somewhat lower social skill requirements.

Why This Article Was Selected for The Policy Scientist

College entry decisions shape educational attainment, career opportunities, and lifetime earnings for millions of students, making this an enduring policy issue. As tuition costs rise and workforce demands evolve, understanding whether beginning at a community college changes long-term career trajectories has become increasingly important. Swiderski and McDonnell have contributed repeatedly to this research area, and this study extends the literature by examining occupational characteristics rather than wages alone. Published in Education Finance and Policy, one of the leading peer-reviewed journals in education policy, the article uses a high-quality national longitudinal dataset and propensity score weighting, a credible causal inference approach that substantially strengthens the analysis over conventional multivariate regression. Although unobserved selection cannot be eliminated, the findings are likely to generalize broadly to similar U.S. postsecondary systems.

Full Citation and Link to Article

Swiderski, T., & McDonnell, J. W. (2026). Diverted to what? The impact of community college attendance on mid-career occupational characteristics and wages. Education Finance and Policy. Advance online publication. https://doi.org/10.1162/EDFP_a_00459

Central Research Question

The central question examined by Swiderski and McDonnell is whether students who begin their postsecondary education at a community college experience different long-term educational, occupational, and wage outcomes than otherwise similar students who begin at a four-year institution. Prior research has consistently demonstrated that community college starters are substantially less likely to complete a bachelor’s degree. However, evidence regarding the labor market consequences of this educational pathway has remained mixed. The authors seek to determine whether differences in educational attainment ultimately translate into meaningful differences in mid-career earnings or whether community college entrants follow alternative occupational pathways that partially offset lower rates of bachelor’s degree completion. To answer this broader question, they examine not only educational attainment and wages but also occupational sectors, educational requirements of occupations, and the math and social skill demands associated with the jobs individuals hold during their thirties.

Previous Literature

The study builds upon a long line of research examining the so-called diversionary effect of community colleges. Early scholarship argued that beginning at a two-year institution reduced students’ chances of completing a bachelor’s degree, while more recent studies have documented similar findings using stronger empirical methods. Meta-analytic evidence indicates that community college entrants are more than 20 percentage points less likely to earn a bachelor’s degree than comparable students who begin at four-year institutions.

Although the literature on educational attainment is remarkably consistent, research examining wages has produced less definitive conclusions. Several propensity score matching studies have found little or no statistically significant difference in earnings between comparable two-year and four-year starters during early adulthood. In contrast, regression discontinuity and instrumental variables studies examining admission to selective four-year institutions have sometimes reported substantially larger wage gains. The authors explain that these studies estimate somewhat different treatment effects because many comparison students in regression discontinuity designs would not necessarily have attended community colleges.

The present study extends this literature by examining dimensions of occupational outcomes that previous research has largely overlooked. Rather than focusing exclusively on educational attainment and earnings, the authors investigate occupational sectors, educational requirements of occupations, and job skill requirements. This broader perspective provides new insight into how community college attendance may influence career trajectories throughout adulthood.

Data

The analysis uses data from the National Longitudinal Survey of Youth 1997 (NLSY97), one of the highest-quality longitudinal datasets available for studying educational and labor market outcomes in the United States. The survey follows nearly 9,000 nationally representative individuals who were between 12 and 17 years old in 1997 and collects detailed information on education, employment, income, and family background over more than two decades.

The authors restrict their analysis to approximately 3,100 individuals who entered either a two-year or four-year college shortly after high school and who remained in the survey through 2019, when most participants were between 34 and 39 years old. This lengthy follow-up period allows the researchers to examine outcomes after educational attainment and occupational trajectories have largely stabilized.

To enrich the analysis, the researchers link participants’ occupations to the Occupational Information Network (O*NET), which provides standardized information about educational requirements and occupational skill demands. This linkage enables the authors to examine whether individuals work in occupations requiring bachelor’s degrees, whether their jobs typically require postsecondary education, and the levels of math and social skills associated with their occupations. The combination of nationally representative longitudinal survey data and detailed occupational information substantially expands the range of outcomes that can be evaluated.

Methods

The study employs propensity score weighting, an established causal inference technique designed to compare individuals with similar observable characteristics who differ primarily in whether they began college at a two-year or four-year institution. Unlike conventional multivariate regression, propensity score methods explicitly construct more comparable treatment and comparison groups, thereby reducing bias arising from observable differences between students.

The authors estimate numerous weighting specifications and conduct extensive sensitivity analyses to assess the robustness of their findings. Their preferred specification estimates the average treatment effect for students who actually began at community colleges while weighting four-year entrants to resemble those students. Alternative specifications reverse the weighting strategy to evaluate whether results differ for students who more closely resemble traditional four-year entrants.

The researchers examine a broad range of outcomes, including educational attainment, employment, annual wages, household income, occupational sector, educational requirements of occupations, job satisfaction, educational debt, and occupational math and social skill requirements. They also estimate multiple robustness checks using alternative observation years and alternative weighting approaches. Although propensity score weighting cannot eliminate bias from unobserved factors, it represents a considerably stronger design than conventional regression analysis for estimating treatment effects in observational data.

Findings/Size Effects

The study confirms previous evidence that beginning at a community college substantially reduces bachelor’s degree attainment. Community college starters were more than 20 percentage points less likely to earn a bachelor’s degree than otherwise comparable students who began at four-year institutions.

The wage findings are considerably more nuanced. Under the authors’ preferred specification, community college starters earned approximately 4% lower wages than comparable four-year starters, although this difference was not statistically significant. When restricting the analysis to employed individuals, the estimated wage difference declined to roughly 2%. However, alternative weighting specifications produced statistically significant wage advantages for four-year starters approaching 8%, suggesting that wage effects may vary depending on which student population is being examined.

Occupational outcomes differed less dramatically than educational attainment. Community college starters were only about 6 percentage points less likely to work in occupations that typically require a bachelor’s degree, despite the much larger difference in bachelor’s degree completion. This finding suggests that many four-year graduates ultimately enter occupations that do not require bachelor’s degrees, while many community college starters obtain positions that value postsecondary training without strictly requiring bachelor’s degrees.

The study also finds no meaningful differences in the mathematical skill requirements of occupations held by the two groups. Community college starters, however, worked in occupations requiring somewhat lower levels of social skills. Occupational sorting also differed across sectors. Community college entrants were more likely to work in trade, industrial, and applied science occupations, while four-year starters more frequently entered professional service occupations such as teaching and counseling. The former occupations often combine relatively high wages with comparatively modest educational requirements, helping explain why wage differences remain much smaller than educational attainment differences.

Additional outcomes reinforce this pattern. Community college starters exhibited somewhat lower household incomes, although these differences were generally not statistically significant. They reported similar levels of job satisfaction and carried lower educational loan balances by their thirties than comparable four-year starters.

Conclusion

The study demonstrates that lower bachelor’s degree completion among community college entrants does not automatically translate into proportionally lower mid-career earnings. Instead, occupational sorting appears to partially offset educational differences by directing many community college starters into relatively well-paying technical and trade occupations while many four-year graduates enter professional occupations requiring higher educational credentials but offering only moderate wage premiums.

By expanding the analysis beyond educational attainment and wages, the authors provide a richer understanding of how initial college choice influences long-term occupational trajectories. Their findings suggest that occupational characteristics deserve greater attention when evaluating postsecondary pathways because educational credentials alone provide an incomplete picture of career outcomes.

Methodologically, the study represents a meaningful contribution through its use of high-quality longitudinal data, detailed occupational measures, and propensity score weighting. Although unobserved selection remains a limitation common to observational research, the extensive robustness analyses strengthen confidence in the findings. Future research employing randomized interventions where feasible or stronger quasi-experimental causal inference designs could further clarify the mechanisms through which community college attendance influences long-term occupational and economic outcomes.