This Is a National Crisis. We Should Start Treating It Like One.

Hey there!

52%.

That is the share of college graduates who are underemployed within one year of getting their degree, according to a 2024 report called "Talent Disrupted" from the Burning Glass Institute and Strada Education Foundation, which tracked the career histories of more than 60 million workers. Underemployed means working in a job that doesn't typically require a college degree.

More than half of graduates. Within one year.

And here is what makes it worse: it doesn't get better quickly. Ten years after graduation, 45% of graduates are still underemployed. And of those who started underemployed, 73% remained so a decade later. Researchers call this the "scarring effect." Your first job after college sets a trajectory that is extraordinarily hard to escape.

The Federal Reserve Bank of New York tracks this quarterly. In Q4 2025, underemployment for recent college graduates hit 42.5%. The unemployment rate for grads aged 22 to 27 was 5.7%. The Federal Reserve economist Jaison Abel called 2025 "among the most challenging labor markets for recent college graduates in the last decade, outside of the pandemic period."

The Class of 2025 submitted more applications on average than the Class of 2024. They received fewer job offers.

This is not a bad stretch. This is a structural failure.

What the data actually tells us about what works

Before I get into what I've been hearing from career services, I want to name the one thing the Burning Glass report found that actually moves the needle.

Internships.

Graduates who had at least one internship were 48.5% less likely to be underemployed after graduation. The effect held across degree fields. It was one of the strongest signals in the entire dataset.

That finding matters because it reframes what career services is actually trying to do. The goal isn't just helping students write resumes or prepare for interviews. The goal is getting students into substantive work experience before they graduate, because that experience is what separates the students who land well from the ones who don't.

And right now, too many students are not getting there.

What I've heard from 70+ career services conversations

Over the past several months I've been on calls with career services directors, career coaches, and employer relations professionals at universities across the country. Big research institutions, small liberal arts colleges, business programs, engineering departments, HBCUs, Hispanic-serving institutions.

What I've found is a system that is genuinely trying, staffed by people who care deeply, and structurally unable to deliver what the moment requires.

The bandwidth is not there.

At NYU's Tandon School of Engineering, a former Google recruiter turned career advisor supports eight thousand students with a team of three. At the University of Wisconsin, one advisor serves the largest college on campus: eighteen thousand students. These are not outliers. They are typical.

The consequence is predictable: help goes to the students who seek it out. Multiple advisors used the same phrase independently. They give the squeaky wheel more grease. The students who are quietly struggling, who don't know what to ask, who disappear into the noise, go largely unserved. And if Burning Glass is right that the first job determines a decade of outcomes, those are the students who pay for it longest.

The data system is broken in a way that hides the problem from the people who need to fix it.

This is the part that I keep coming back to. It is not just that career services doesn't track the right things. It is that the way they collect the data they do have almost guarantees a distorted picture.

Universities track graduation outcomes: where did students end up six months after they crossed the stage? They collect this through surveys sent to recent graduates. And here is the structural flaw: graduates who are proud of where they landed fill out those surveys. Graduates who are struggling, underemployed, unemployed, or too demoralized to respond largely do not.

At USF, the advisor told me she gets placement data by personally chasing down graduates on LinkedIn and calling them repeatedly until they give her a title and a salary. At Texas State, the data is only as good as what students voluntarily report back. At NYU, internship placements aren't tracked at all. At the University of Miami, the director of employer engagement described the graduation survey as the main data source, and acknowledged that response rates collapsed after the pandemic, when they stopped requiring students to pick up their regalia in person. At Bucknell, the senior career advisor has been requesting a dedicated data analyst for two years. At one school after another, I heard some version of: we know our data is incomplete, but it's what we have.

So the universities that report 80%, 85%, 90% placement rates are, in most cases, reporting on the subset of graduates who responded to the survey. Which is disproportionately the subset who got good jobs.

The students who are underemployed, in jobs that have nothing to do with their degree, working in retail or food service or taking anything they could find while still applying, are the least likely to fill out a form announcing where they ended up. They are invisible to the data system. And because they are invisible, universities don't feel the urgency of the real number, which Burning Glass found by tracking actual career histories of 60 million workers, not self-reported surveys. That number is 52%.

The gap between what universities report and what is actually happening is not the result of dishonesty. It is the result of a data methodology that was never designed to see the full picture. And it means that institutions that genuinely want to serve their students are making decisions based on a reality that is significantly rosier than the one their graduates are living.

What none of them track at all is anything upstream: interview rates, application volume, how well students' resumes match the roles they're targeting, whether students are engaging with resources before their senior year panic sets in. The signal that would allow career services to intervene earlier, to catch the student on application 400 with zero callbacks before another hundred go out the same way, simply does not exist in the system.

Students are making the same fixable mistakes, everywhere.

The University of Wisconsin advisor described what he sees constantly: students on application 300, 400, sometimes 700. With one interview. Or zero. When he looks at what they've been doing, the pattern is almost always the same.

A computer science student applying to data analyst roles whose resume doesn't mention a single data skill. A data science student who lists Python in a skills section but never integrates it into any project description. Students applying to roles that have nothing to do with what they actually want to do. Students not reading job descriptions before submitting.

One advisor who spent fifteen years in the field and had talked to thousands of students put it plainly: "Recruiting is fundamentally broken. There are people who are extremely talented, overqualified for positions, and the process is failing them."

The advisor from a large research university described the specific moment when students finally understand the mistake they've been making: when he shows them a strong resume from another student at a different school, removes the identifying information, and asks, "If you're the employer, who are you going to interview?" Then, he said, they get it. But that moment is usually arriving junior or senior year, when months of the job search have already been spent going in the wrong direction.

The system is built to react, not anticipate.

At SCSU, the student success coach described the pattern she sees every year without fail: students reach junior year and realize they need an internship, not knowing that applications for summer internships at many firms open in October or November of the prior academic year. At Wisconsin, the advisor described three categories of students: proactive freshmen, mid-search students who come in for help, and seniors who show up two months before graduation having never once used career services.

That last group needs the most help. By the time they arrive, the window is nearly closed.

One advisor at a research institution described the students they never reach at all: the ones tracking toward grad school who suddenly decide to pursue industry, the transfer students who feel behind before they start, the commuter students who are on campus for class and then gone. At SCSU, more than half of students are commuters. Building career readiness for students you rarely see in person is a problem the system has not solved.

Why this matters beyond the individual student

The Burning Glass researchers did not treat their findings as a personal finance story. They treated them as a policy crisis.

When 52% of graduates are underemployed, and 73% of those who start underemployed stay that way for a decade, the economic consequences compound. An underemployed graduate earns roughly 25% more than someone with only a high school diploma, rather than the 70% premium that a well-placed degree-holder earns. Student loans, which were borrowed against the expectation of the higher number, don't shrink because the outcome was worse than projected.

Senator Mark Warner described the trajectory plainly in late 2025: if we eliminate the front end of the career pipeline for young people, how do they ever reach the mid-career stage? He called the potential for disruption "unprecedented."

This isn't a blip. The gap between unemployment rates for recent college graduates and the overall population is now the smallest it has been in thirty years. The advantage of a degree, at least in terms of quickly finding good work, is narrowing. And career services, doing the best it can with what it has, cannot bend that curve on its own.

One model that is doing things differently

Not every conversation I had was a story of strain. At Hult International Business School in Boston, I spoke with a career director who has deliberately built a different kind of operation.

His core premise: the real problem in job searching is not resume formatting or ATS optimization. It is that students are unknown to the people they want to be known to. Everything else is noise layered on top of that fundamental gap.

His office monitors 72 cities for labor market and economic development data to surface opportunities students aren't looking for yet. They run a seven-hour career event on Saturdays that consistently draws over 130 students, not because they teach resume writing but because they teach market intelligence, relationship strategy, and how to navigate a hiring process as a human being rather than as an applicant number. He made a point I keep returning to: every career center tells students to network and access the hidden job market, then surfaces a job board to them. That is a contradiction. His office doesn't pretend otherwise.

The Burning Glass data backs up what he's building toward. The students who get college-level jobs out of the gate, and stay in them, are the ones who show up known, who applied with purpose, who had internship experience that made their resume legible to the people reading it.

The system, at scale, is not producing enough of those students. That is what needs to change.

Hit reply and tell me: what do you think it would actually take to fix this? I am genuinely asking.

Keep progressing,

Ford Coleman Founder & CEO, Runway

P.S. If you know someone who works in higher education, policy, or hiring, send this their way. This conversation needs more people in it.