A data story
Is your degree
worth it?
We followed 62,000+ real college programs all the way through, to the jobs graduates enter, what they earn against the debt they carry, how exposed those jobs are to AI, and whether the paycheck covers the rent. Every number is a real, published figure. Nothing is generated.
Explore your major ↓Where does it lead?
A major isn't a job; it's a spray of them. We trace each of 226 fields to the 800+ occupations its graduates actually enter.
What does it pay?
Five years out, against the debt it took to get there. Some degrees pay it back in two years; some never quite do.
Will AI come for it?
How much of the day-to-day could a language model already do? The uncomfortable answer: the best-paid fields are often the most exposed.
Can you afford the life?
A six-figure salary in one city is a stretch in another. We put the paycheck against the rent, metro by metro.
Now find yours
Pick a school and major below to trace it the whole way through: the jobs, the payoff, the AI exposure, and whether the pay covers the rent where you want to live.
The explorer
Where Your Degree Takes You
Pick a major and a school. Follow it all the way through: the jobs graduates enter, what they earn against the debt they carry, how exposed those jobs are to AI, and whether the paycheck covers the rent.
What it pays vs. what it costs
Every dot is a school offering this major. Right is better pay; lower is less debt. Colour shows how fast the debt pays off. Source: U.S. Dept. of Education College Scorecard, Field of Study (5-year earnings; Title IV completers).
Loading programs…
Where this degree leads
The occupations graduates of this field most commonly enter, weighted by an even split (the crosswalk gives no native weights, so the method is disclosed per row). AI exposure is the published Eloundou “GPTs are GPTs” β measure: the share of an occupation’s tasks generative AI plus tools could do substantially faster. GPT-4-era (2023) task overlap, not a forecast of job loss.
No occupation mapping available for this major.
What this can and can't tell you
- Earnings are Title-IV only. They cover federally-aided graduates (those who took loans/grants). Wealthier non-borrowers are excluded, which can bias a program's figures.
- The numbers describe past graduates. 5-year earnings are 2014 to 2016 cohorts measured in 2020 to 2021, a defensible comparison, not a forecast of what you'll earn.
- A major isn't one job. Degree→occupation flows are weighted by employment (disclosed per row), not observed individual outcomes. Geography drops generic catch-all occupations to stay meaningful.
- AI exposure ≠ job loss. It's GPT-4-era (2023) task overlap from published measures: how much could be assisted, not a prediction that the job disappears.
- The premium is observational. The selection-adjusted edge controls for who enrolls, but it is correlational, not causal proof that the major caused the earnings.
- Rent and pay are medians. Affordability uses metro-median rent vs. graduate-weighted pay; OEWS covers fewer metros than Zillow, so unmatched metros are omitted, never estimated.
Sources, method & downloads
Only real, source-traceable numbers are shown. Privacy-suppressed cells are left out, never estimated. Earnings reflect federally-aided completers and a multi-year cohort lag; “years to pay off” assumes 10% of earnings goes to debt, a disclosed formula, not a prediction. Everything below is reproducible end to end.
Processed datasets
All ~2,200 JSON files ship in the repo.
Pipeline & code
Download → run → reproduce every number.