College students are facing one of the most precarious labor markets in years. A recent New York Federal Reserve study pointed to a sharp increase in the post-college unemployed compared with other workers, with explanations ranging from artificial intelligence job displacement to a broader slowdown in hiring.

One response has been a boom in trade-school and apprenticeship programs that launch young people into “AI-proof” careers. But students at community and four-year colleges can also beat these headwinds and launch a career through on-the-job internships.

We’re a group of multidisciplinary scholars at The Ohio State University who study two paid internship programs in science, technology, engineering and math, or STEM, with a proven record of landing jobs for students. Although the two initiatives use different models, they both illustrate four lessons that colleges and employers can use to turn internships into genuine career pathways.

Just as notable is that a large share of these students come from diverse and underrepresented backgrounds, with many being the first in their families to go to college. These programs can serve as a powerful equalizer for young people who may not otherwise have such career opportunities.

Quality over quantity

Internships remain one of higher education’s most powerful tools for preparing students for work. Decades of research show that students who complete high-quality internships are more likely to secure job interviews and high-earning positions post-graduation.

At the same time, the number of internship opportunities is diminishing, even though that preparation matters more than ever. Artificial intelligence is transforming entry-level work, leading employers to increasingly value the human skills that machines cannot easily replicate: judgment, adaptability, collaboration and the ability to solve unfamiliar problems.

Academic leaders are recognizing these pressures. More than one-third of university presidents and two-thirds of community college presidents told an Inside Higher Ed survey that they plan to add or expand apprenticeship pathways over the next three years.

But it’s not just the number of internships that matters – internship quality counts just as much.

Our research at the College Impact Laboratory examined learning through experience with two multi-institutional studies. One is Bridging Academic Training Through Experiential Research and Innovation, or BATTERI, led by principal investigator Jay Sayre and project manager Joanna Gardner. The other is Drive to Thrive.

Across both initiatives, students benefited when their experiences combined substantive work with mentoring and professional development, rather than being treated as inexpensive labor. This distinction suggests that merely placing students in workplaces isn’t enough. The structure and quality of those experiences help determine whether internships lead to meaningful, high-earning careers.

Learning by doing

Drive to Thrive deploys an approach called earn-as-you-learn. Through partnerships with Ford dealerships, students alternate between classroom instruction and paid workplace experience. They work alongside experienced technicians, diagnose and repair customers’ vehicles and receive regular coaching.

In this sense, it’s similar to the European model of traineeships for young people.

This initiative reached 635 students at 18 community colleges, including 345 first-generation students and 300 students of color. Through interviews and focus groups, a majority told us the internship helped them see themselves as true professionals. And several said the internship confirmed they had chosen the right career.

The numbers also backed up our approach. Students in the earn-as-you-learn program were almost 50% more likely to graduate compared with students in a matched comparison group.

Our other program brings together clean energy manufacturing students from three different types of institutions – a major research university, a historically Black university and a community college – for a 10-week internship. The program combines workplace learning with professional development workshops, peer networking, community engagement and projects with industry partners such as Honda, Stellantis and Schaeffler.

This initiative took a distinctive approach by assigning multiple mentors to each student. The idea is to help students develop relationships with faculty members, technical experts, employers and classmates. This network, in turn, strengthened students’ confidence, professional identity and sense of belonging while complementing the technical skills they develop.

To date, the formula has worked. More than half the participants in 2025 have already lined up either a new internship or job interviews. Industry partners also plan to rehire several members of the 2026 cohort as interns, extending their training beyond our program and opening pathways into clean-tech careers.

Based on this record, we believe that colleges should build broader support networks around internships rather than expect a single supervisor to shoulder the entire responsibility for a student’s professional development.

Making work matter

In both programs, four valuable principles emerged.

First is the importance of meaningful work. We believe interns should tackle real organizational challenges with deliverables that matter beyond the classroom. For example, two interns spent the summer addressing a safety challenge for Honda – how to prevent battery cells from catching fire and cooling them down when they do.

After developing potential solutions, they presented their work to industry representatives. A senior engineer from the Honda Research Institute was so impressed that he contacted them about a post-internship opportunity.

That sense of consequence changed how the students approached the work. Rather than viewing it as an assignment to complete, they were making a contribution that others depended on.

Second is the constellation of support for students. The liaison between the institution and industry partners is essential, serving as the connective tissue that holds the internship together and ensuring that students don’t fall through the cracks.

But that liaison alone isn’t enough. Students also need access to faculty, industry supervisors, technical experts and peers, since each brings a different kind of guidance. One strength of the clean energy program, for example, is its inclusion of industry experts across institutions and organizations. Students benefit most when they can draw on multiple mentors and peer support to develop key collaborative skills.

Third, students should learn how to normalize struggle. Internships aren’t a test of whether students can already function as professionals. They’re environments for students to try, fail and retry. When educators emphasize struggle as part of the learning process, students are encouraged to avoid perfectionism and instead ask for help when difficulties arise.

For example, students submit weekly reports during the clean energy program that give them the chance to discuss challenges, troubleshoot problems together and recognize that uncertainty is part of becoming a professional.

Finally, it’s vital to make interns’ contributions visible. Presentations, project showcases and recognition from industry professionals reinforce that students’ work matters while strengthening their professional identities.

We believe universities and industries alike should heed these lessons, which extend well beyond clean energy manufacturing and automotive technology. Successful programs build ecosystems that often determine whether students simply complete an internship or launch a career.

As colleges respond to changing labor markets, they should resist measuring success just by how many internships they create. Instead, they should ask whether those opportunities help students become confident, capable professionals. Internship quantity opens the door to better outcomes after college. Internship quality changes lives.

Matthew J. Mayhew receives funding from the National Science Foundation, the U.S. Department of Education, the Ewing Marion Kauffman Foundation, the ECMC Foundation, the Arthur Vining Davis Foundations, the Templeton Religion Trust, the Andrew W. Mellon Foundation, the Alfred P. Sloan Foundation, and the Fetzer Institute.

Courtney DeRoche receives funding from the ECMC Foundation (PI: Matthew J. Mayhew).

Yun-Han Weng is an evaluator on the National Science Foundation-funded grant (co-PI: Matthew J. Mayhew).