Internships: Who Gets Access and Who Benefits Most?

Share

Internships are now a normal part of university life, but they are not shared equally. A national U.S. study by Carrie Shandra found that participation rose from 29.6 percent of graduating seniors in 1994 to 1995 to 69.2 percent in 2015 to 2017. Over the same period, the gap between the lowest and highest family-income groups widened. Meanwhile, a 2026 evaluation of a paid internship program found that the students least likely to finish a degree gained the most from interning. Together, these studies suggest a mismatch: the students who could benefit most are often the least likely to get internships. This article reviews four recent studies on who takes part in internships and placements, who benefits, and what universities and families can do about it.

Why the question matters now

Universities, employers and governments increasingly treat work experience during a degree as part of being ready for a job. When an experience becomes expected, unequal access turns into unequal opportunity. Students who cannot afford an unpaid placement, who need paid work to stay enrolled, or who lack professional contacts may graduate without the experience that employers and graduate programs look for. For institutions, the question is no longer only whether to offer internships, but who actually reaches them.

What four recent studies found

Participation has more than doubled, and the income gap has grown

Shandra (2026), in Sociological Science, analyzed College Senior Survey data from 1994 to 2017, with a final analytic sample of 329,949 U.S. graduating seniors. Students were asked whether they had taken part in an internship program since entering college. Participation rose from 29.6 percent in 1994 to 1995 to 69.2 percent in 2015 to 2017, more than doubling.

The growth was uneven, even after controls for grades, major and demographics. Students from the lowest family-income quartile had a predicted probability of interning of 0.493 in 2006 to 2008, compared with 0.520 for the highest quartile, a gap of 2.7 percentage points. By 2015 to 2017, the gap had grown to 7.3 percentage points, and the author reports that this later gap is statistically larger. The gap between first-generation and continuing-generation students was 3.6 percentage points from 2009 to 2017, which the author describes as persistent. Students at private colleges were more likely to intern than those at public colleges. Shandra concludes that internships follow "similar patterns of stratification as formal credentials." The study reports associations, not causal effects, and the author notes that the survey sample is not nationally representative.

A paid internship helped most the students least likely to finish

Neylon and Attewell (2026), in the Journal of Postsecondary Student Success, evaluated a paid public-service internship program at a multi-campus urban university in the United States. The university has both two-year associate's and four-year baccalaureate campuses. Interns who started between fall 2013 and fall 2017 worked 11 to 12 hours a week for 24 weeks. They were paid $12 an hour from 2013 to 2016 and $13 an hour in 2017.

The authors used three methods to adjust for differences between interns and non-interns: coarsened exact matching, a doubly robust weighting model and Bayesian additive regression trees. Across these methods, interning was associated with a 15.0 to 17.8 percentage-point higher probability of completing a degree within 16 semesters, at both the associate's and baccalaureate levels.

The central finding concerns who benefited most. All subgroups benefited. However, the authors report that effects "were significantly larger for men, for Black and Hispanic men in particular, for students with a low first-year GPA, and for low likelihood of completion students." At the baccalaureate level, students with the lowest predicted probability of completion had a benefit 16.3 percentage points above the average effect of 16.9 points, which the authors describe as nearly double. Yet students with stronger first-year grades and more credits were more likely to become interns, which the authors attribute to institutional barriers to entry, including advertised GPA and credit requirements. The authors call this a "Matthew Effect" and argue that colleges should recruit the students at highest risk of not completing. They note that unobserved differences between interns and non-interns could still affect the estimates.

UK placement students earn higher degrees, but access is uneven

Traynor and colleagues (2024), in the Journal of Learning Development in Higher Education, analyzed records for 31,159 undergraduates who graduated from a UK Russell Group university between 2016 and 2023. Of these, 1,389 (4.5 percent) completed a Year in Industry placement. These placements normally last 26 to 52 weeks of full-time work and are expected to be paid by the employer.

Placement students were much more likely to earn a first-class degree (70.1 percent, against 28.5 percent of non-placement students) and a good degree (97.7 percent, against 83.6 percent). Access, however, was uneven. Women made up 55.9 percent of non-placement students but only 45.1 percent of placement students. Students with a known disability made up 14 percent of non-placement students but 9.8 percent of placement students. White students made up 83.7 percent of placement students, compared with 73.7 percent of non-placement students. Among students from areas with low participation in higher education, the share earning a good degree was 96.4 percent for placement students and 81.2 percent for those without a placement.

The authors are candid about selection. Placement students already had higher average Year 1 marks (69.3 percent against 63.4 percent) before going on placement. The authors write that it is unclear whether placements raise final-year performance or whether stronger students are better at securing placements. The study is descriptive and covers one institution.

One university's audit: very low participation and clear barriers

Choe-Smith and colleagues (2025), in Metropolitan Universities, report a self-study at California State University, Sacramento, a large public university. They used ten years of enrollment data for for-credit internship courses, a logistic regression on 122,576 observations, and student surveys. In fall 2023, very small shares of students in each group were enrolled in for-credit internship courses, for example 1.10 percent of Black students and 1.91 percent of White students. In the regression, the coefficients for Black and Latinx students were negative and statistically significant, indicating lower participation.

In the 2025 student survey, 166 students answered a question about barriers. The most common were:

  • scheduling (59 percent);
  • financial barriers, such as needing to work or a lack of paid opportunities (55 percent);
  • institutional barriers, such as navigating services and processes (51 percent);
  • a lack of social or professional networks (49 percent);
  • practical barriers, such as transportation, childcare or accessibility (33 percent).

The authors write that "many students responded that they were unable to give up paid work for an unpaid internship." This is a single-campus, descriptive report, and the authors note that the university had no central system for recording all internships.

The four studies compared

Four recent studies on internship access and benefits
StudySetting and dataDesignMain finding
Shandra (2026)U.S. College Senior Survey, 1994 to 2017; 329,949 seniorsRegression with controls (associations)Participation rose from 29.6 to 69.2 percent; the income gap widened from 2.7 to 7.3 points
Neylon and Attewell (2026)Paid public-service internship at a multi-campus U.S. university; interns starting 2013 to 2017Matching, weighting and Bayesian modelsDegree completion 15.0 to 17.8 points higher; the largest gains were for students least likely to complete
Traynor et al. (2024)One UK Russell Group university; 31,159 graduates, 2016 to 2023Descriptive comparisonPlacement students earned higher degrees; women, students with disabilities and ethnic minority students were under-represented
Choe-Smith et al. (2025)California State University, Sacramento; ten years of course data and surveysInstitutional self-studyLow for-credit internship enrollment; the need for paid work and missing networks were leading barriers

Our interpretation

Read together, the four studies point in the same direction, although none of them settles the question alone. First, internships have become close to standard in some systems, so the students who miss out are now the exception rather than the rule. This makes the missing experience easier for employers to notice. Second, the barriers the studies report are mostly practical and financial: needing to earn money, unpaid or low-paid roles, scheduling, transport and a lack of contacts. Third, the Neylon and Attewell results suggest that a paid, structured program can reach students whom unpaid placements usually exclude, and that these students may gain the most.

We would not conclude from these studies that internships cause higher grades or better jobs in every setting. The UK results are descriptive, and the authors themselves note that stronger students may be the ones who secure placements. The U.S. evaluation uses careful methods, but it covers a single paid public-service program. The safer conclusion is narrower. Selective entry criteria and unpaid roles push internships toward students who already have advantages, and the evidence gives institutions good reason to test more open, paid models and to measure who takes part.

Practical steps for universities

  • Measure participation by student group. Record every internship and placement, paid or unpaid, for credit or not, and report participation by income, first-generation status, sex, disability and ethnicity. Choe-Smith and colleagues recommend disaggregated, longitudinal data that separate paid from unpaid roles.
  • Review entry requirements. GPA thresholds and essay requirements can screen out the students who might benefit most. Neylon and Attewell recommend actively recruiting students at high risk of not completing, with outreach and pre-internship training.
  • Fund paid and flexible formats. Stipends, part-time placements and roles that fit around paid work address the most common financial barrier.
  • Recognize the work of applying. Traynor and colleagues suggest giving academic credit for the effort of applying for placements, and targeting support at under-represented groups.
  • Build internships into the curriculum. Course-based or embedded work experience reduces reliance on personal networks and on unpaid summer roles.

Practical steps for students and families

  • Ask universities what share of their students complete an internship or placement, whether placements are paid, and what support exists for finding one.
  • Look for paid, part-time or for-credit options early, so that work experience does not compete with earning money.
  • Use the university's career service to build contacts if family networks are limited. Missing networks were a leading barrier in the Sacramento survey.
  • Treat high placement grades or outcomes with care. As the UK study notes, students who secure placements may already be stronger academically.

Limitations

This review draws on four studies, all read in full: Shandra (2026), Neylon and Attewell (2026), Traynor and colleagues (2024) and Choe-Smith and colleagues (2025). None of the four was read from the abstract only. Three are set in the United States and one in the United Kingdom, so the findings may not transfer to other higher-education systems. Only Neylon and Attewell attempt causal estimation, and they acknowledge possible unobserved confounding. The other three studies are descriptive or associational. Two are single-institution studies. Shandra's survey data are not nationally representative and do not record whether internships were paid. Several other relevant studies were set aside because their publisher pages could not be opened and verified during this review.

Related reading on UNIRANKS

For more on graduate outcomes, see the Employability & Career Success hub and the article on enhancing graduate employability through experiential simulation learning. For teaching and learning quality, visit the Quality of Education hub.

Editorial disclosure

UNIRANKS publishes university rankings, and its methodology includes an Employability & Career Success dimension. This article reviews independent research on access to university internships and placements. It does not describe how UNIRANKS scores institutions.

Research References

  1. Shandra, C. (2026). More Common, Less Equal: Disparities in College Internship Participation Over Time. Sociological Science, 13, 476-500. https://doi.org/10.15195/v13.a19 (verified 22 November 2026)
  2. Neylon, S., & Attewell, P. (2026). Undergraduate Internships, Degree Completion, and the Matthew Effect. Journal of Postsecondary Student Success, 5(2), 111-134. https://doi.org/10.33009/fsop_jpss139836 (verified 22 November 2026)
  3. Traynor, K., Evans, K., Barlow, C., et al. (2024). Year in Industry: who gets access and what difference does it make? Access and awarding gaps in UK university undergraduate placement programmes. Journal of Learning Development in Higher Education, (30). https://doi.org/10.47408/jldhe.vi30.1155 (verified 22 November 2026)
  4. Choe-Smith, C., Repa, M., Jackson, J., et al. (2025). Equitable Access to For-Credit and Paid Internships. Metropolitan Universities, 36(2). https://doi.org/10.18060/28676 (verified 22 November 2026)
Share this article