How College Scorecard Earnings Are Measured
Start with the cohort, not the headline
Every earnings figure on this site begins with a defined group of students: people who entered a college in a particular period, received federal financial aid, and were later observed in federal earnings records. The Department of Education matches education records with tax data, removes students who are enrolled again at the measurement point, and reports the median — the middle value — for students who are working. The ten-year figure used as our headline is therefore not the average salary of a college's alumni, not a survey response, and not a starting salary. It is the median annual earnings, a decade after entry, of an aid-receiving cohort whose members have had very different paths through and beyond college.
Why the median, and why it matters
A median is used because earnings are skewed: a handful of very high earners would drag an average upward and make a typical graduate's outcome look better than it is. The median says that half the measured cohort earned more and half earned less, which is a more honest summary — and also a reminder of how wide the hidden spread is. Around any median on this site sit graduates earning far more and graduates earning very little, in different occupations and places. When a college reports a median of, say, $60,000, planning as if $60,000 were guaranteed misreads the statistic completely; planning with the knowledge that outcomes straddle that figure widely is the correct use of it.
Working and not enrolled: a quiet filter with loud effects
Students who have gone back to school — for a master's degree, a professional qualification or a second credential — are excluded from the earnings measurement while enrolled. That filter flatters some colleges and penalises others in predictable ways. A college whose graduates routinely proceed straight to well-paid graduate programmes may show lower measured earnings at early measurement points precisely because its most ambitious graduates are in classrooms, not workplaces. Conversely, a college feeding directly into employment shows its earnings sooner. Comparing a six-year figure with a ten-year figure for the same college often reveals this effect, which is why our profiles publish both where the release provides them.
Inflation, timing and cohorts
A ten-year cohort entered college roughly a decade before measurement and graduated into the labour market of several years ago. Their earnings reflect that era's economy, that college's major mix at that time, and programmes that may since have opened, closed or changed. The Department adjusts and pools cohorts according to its published technical documentation, but no adjustment turns a historical cohort into a forecast. Treat the figure as the best comparable evidence about what a college's students have gone on to earn — evidence strong enough to rule out fantasy claims and to separate broadly different outcomes — and not as a prediction printed for your own cohort, whose economy nobody has seen yet.
What to do with an earnings figure
Use it in sequence. First check completion, because earnings describe people who largely got through. Then compare the figure with the same measure at peer colleges — same definitions, same release — rather than with advertised salaries, which use unknown populations and methods. Then go down a level to the field-of-study figures for the major you actually intend, where the spread within a college usually exceeds the spread between colleges. Finally, set the figure next to debt, because earnings are only good news relative to what they cost to reach. That sequence is the editorial spine of this site, and it is built into every profile we publish.
Further reading: return to the Learn index, apply these definitions in the college profiles and field pages, or read the full methodology for sources, files and suppression rules.
Note on this data: Figures on this page come from the U.S. Department of Education College Scorecard release retrieved in October 2026. Earnings cover students who received federal financial aid and were working and not enrolled at the measurement point; they are medians for a defined past cohort, not a promise of what any individual graduate will earn. Check the college and the Scorecard for the most current figures before you decide.