The first job after college is often discussed as a starter role you will replace. A 2024 Strada report, Talent Disrupted, describes a stickier pattern. Among graduates who were underemployed a year out, most were still underemployed ten years later. The year on the report matters. The release is from 2024. That longitudinal picture was public in February 2024, and it is still the clearest public account of how long a mismatched first job can last.
Strada found that 52 percent of graduates were underemployed one year after college, and 45 percent were underemployed at the ten-year mark. Of those who started out underemployed, 73 percent remained so a decade later. An internship cut the likelihood of underemployment by about 49 percent. Inside Higher Ed’s 22 February 2024 account of the same report is a second door into those figures. Use either page, and say 2024 when you quote them. Do not paste them next to a newer unemployment rate and pretend they are the same statistic.
Underemployment here means work in a job that typically asks for less schooling than a bachelor’s degree. Unemployment counts people without a job. The two measures answer different questions. A graduate who is employed in a role that did not ask for a bachelor’s degree can be invisible in an unemployment headline and still be the person Strada is describing. If you only track whether you are “employed,” you can miss the finding that matters for the next ten years.
What “tends to stick” can and cannot plan
The 73 percent figure is about people who were already underemployed at the one-year mark. Your first offer is not sentenced by that share. The finding says the match at the start is hard to undo. A plan that says “I will take anything for a year and then switch” is betting against that persistence. Some people do switch. The report says most of the group that started underemployed did not.
The internship result is the practical half. A student who completes an internship, in this 2024 analysis, had sharply lower odds of landing in that persistent underemployment. The report does not say the internship must be at a famous employer. It says the experience itself changed the odds. That is a reason to treat a paid summer role, a co-op, or a substantial on-campus job you can describe as part of the academic plan. Waiting until classes feel easy gives the finding away.
@randomrecruiter wrote in April 2026 that the job wants three years of experience and the internship wants one. The line is a recruiter’s complaint about postings. It is not a measured average. Read it as a description of the circular ask students meet. Strada’s older finding is the reason the circle matters. If the first real workplace is what keeps people in college-level work, then a posting that demands experience before the first job is asking for the very thing the research says you need in order to avoid a long mismatch.
What to put on the plan this year
Open the plan for this academic year and write one workplace you can finish before you graduate. Name the term. A summer internship, a part-time role during school, or a campus job with a task you can explain all qualify if you can say what you did. A title with no task does not. The Strada internship result is about experience, which means someone could check that you did the work.
If you are already in a job that does not use the degree, do not wait for a magical second year. Write the skill the college-level roles in your field actually ask for, and spend the next month producing one proof of that skill while you keep the paycheck. The report’s warning is about staying in the mismatched work by default. A default is a calendar that never holds a different application.
Look at the 52 percent and the 45 percent as a pair, and keep the labels. One year out, about half. Ten years out, still close to half. The drop is real and it is modest. Quoting only the ten-year number makes the problem sound permanent in a vague way. Quoting only the internship reduction makes the problem sound solved. A plan needs both: the mismatch lasts, and an internship was the factor that cut it.
When you talk about this with a classmate, say the year. “Strada, 2024” takes four extra words and stops the figure from being laundered into a claim about this month’s market. The labor market has moved since February 2024. The persistence pattern is still the reason to care about the first role’s content.
A decision you can write on one card
Before you accept an offer, write what a stranger could learn about you from the job after six months. If the answer is a software tool, a lab method, a client type, or a population you served, you are building the experience Strada associates with staying in college-level work. If the answer is only that you were employed, ask whether a different offer on the table teaches more, even if the pay is close.
This is not an argument for unpaid work. The report’s internship finding should not be used to pressure you into a role that you cannot afford. A paid campus job you can describe beats an unpaid title you cannot survive. Record the pay and the hours next to the task so the plan stays honest.
If you have no offer yet, the card still works. Write the task you want the first role to contain, then search for postings that name that task. A search sorted by brand will not do this. A search sorted by the verb in the description will.
How Flightpath helps
The portal plan by year is where that first workplace belongs, in the year you are actually in. Pin roles that match the task you wrote down. The gap list compares those postings with facts you have confirmed, including work you have already done. A class project becomes usable there only after you confirm the line. A read-only GitHub connection stays partial until you confirm the repository. Bringing projects in as their own items, separate from a resume you upload, is coming soon.
The applications tracker keeps the roles you are choosing among, so “anything” does not become the plan by accident. Eligibility labels stay on the row. A role you cannot take is never ranked above one you can take. You approve anything before it is sent. Strada’s percentages stay in the 2024 report. The plan holds the one workplace you are willing to finish.