Methodology · CollegeCostWise

The rules we never break

No filled-in gaps. Scorecard hides figures when too few graduates are behind them. We show "not enough graduates to report" and build no page for that school and major. About 70% of school-and-major rows have no earnings at all.
Banded figures stay banded. Where Scorecard publishes a rate only as a band (for example "20% or less"), we show the band, never a made-up exact number.
Verdicts come only from the data. No school or partner can pay for a ranking, an order or a verdict.
Not advice. The numbers describe what past graduates typically earned and owed. They are not a prediction for any one student, and nothing here is personalised financial or loan advice.
Independent. We aren't affiliated with or endorsed by any college or the Department of Education.

Where each number comes from

Earnings 1 year after graduating: Median earnings of graduates who got federal aid (Scorecard EARN_MDN_1YR).; Treasury AY2020-21, AY2021-22 pooled earnings cohort measured in CY2022, CY2023, inflation adjusted to 2024 dollars
Earnings 4 years after graduating: EARN_MDN_4YR, same group.; Treasury AY2017-2018, AY2018-2019 pooled earnings cohort measured in CY2022, CY2023, inflation adjusted to 2024 dollars
Typical debt: Median federal debt at graduation for the program (DEBT_ALL_STGP_EVAL_MDN).; NSLDS pooled AY2022-23, AY2023-24 cohort
Net price (undergraduate): Average yearly cost after grants for students who got federal aid, overall or by family income band (NPT4 and NPT41–45). Out-of-state at a public school adds the published gap between out-of-state and in-state tuition.; AcadYr 2024-25, reported in IPEDS DCY2025-26
Cost of a graduate degree: Scorecard publishes no net price for graduate programs, so the cost used is the program's median debt. This undercounts what students paid from savings, family or employers.; NSLDS pooled AY2022-23, AY2023-24 cohort
Odds of finishing: The school's share of first-time, full-time students finishing within 150% of normal time (C150). It is school-wide, not per major.; Fall 2019 or AcadYr 2019-20 cohort (differs by institutional calendar), reported in IPEDS DCY2025-26
Borrowers in default or making progress: Share of federal borrowers in default, or paying down their balance, two years into repayment (BBRR2). We don't show the official 3-year default rate: the payment pause set it to 0% at almost every school.; NSLDS AY2016-17, AY2017-18 pooled cohorts measured in 2018-2019, 2019-2020
Comparison earnings (no degree): For undergraduate credentials: median earnings of high-school graduates aged 25–34 in the school's state (Scorecard EARN_THR_STATE). For graduate credentials: the median 4-year earnings of the state's bachelor's programs.; 2023 ACS 5-year pooled, adjusted to 2024 dollars
Loan interest rate: Federal Direct Loan rates for 2026-27: 6.52% undergraduate, 8.07% graduate unsubsidized, 9.07% PLUS, for loans first disbursed from 2026-07-01. You can change the rate on the comparison.; Federal Student Aid electronic announcement GENERAL-26-33 (posted June 4, 2026), checked 2026-11-04
Today's dollars: Scorecard figures are restated to today's prices with the BLS CPI-U (all items, U.S. city average).; BLS series CUUR0000SA0

Payback and return, as ranges

Each comparison runs 800 scenarios and shows the middle result with a likely range (the 10th to 90th percentile). The scenarios vary only what we are honestly unsure about:

Typical earnings. A median from few graduates is less certain, so it is varied more: by about 0.55 × 1.25 ÷ √(number of graduates).
Growth after year 4. Scorecard stops at about 5 years. Later yearly growth is drawn between 0 and the growth seen from year 1 to year 4, at most 4% a year, and fades to zero over 15 years.
Time to finish. Between the normal length and 1.5 times it, weighted by the school's on-time and 150%-time completion rates.
Repayment term. Between 10 and 20 years, which sets the interest paid.

Earnings bump is graduates' earnings minus the comparison earnings, which grow 0.5% a year after inflation. Payback is how many years of bump it takes to cover the net cost plus loan interest. 10- and 20-year return is the total bump over that time minus net cost, interest and the earnings given up while studying. Nothing is discounted and taxes aren't modelled.

Confidence is high with 100 or more graduates behind the earnings, medium with 40–99, and low below 40.

What these numbers can't tell you

Earnings only cover students who got federal aid, so graduates who never borrowed or got a Pell Grant are missing.
They show what graduates earned, not what the school caused. Students who pick a school or major differ in many ways.
Figures lag by several years, and the cohort years are shown beside each one.
Students who don't finish aren't in the earnings at all. That's why odds of finishing are shown on their own.

Partner links

Some pages show "Next steps" links. Until a partner program is set up they go to federal pages and earn nothing. If we ever earn a commission, the note next to the links says so. Partners never change a number, a verdict or the order of anything.