An independent comparison of … bachelor's-granting institutions in the United States across cost, graduation, earnings, debt, selectivity and class size. Source: U.S. Department of Education, College Scorecard, most recent cohorts (institution-level file, June 2026). Institutions with fewer than 500 undergraduates are excluded.
| Rank | Institution | State | Control | Admission rate | Net cost / yr | Graduation (6 yr) | Median earnings, 10 yr | Median debt | Students per faculty | Score |
|---|
Net cost is the average annual cost after grants and scholarships for students receiving federal aid. Graduation is the share of first-time full-time students completing within 150% of normal time. Earnings are the median for students who entered ten years earlier and are working and not enrolled. Debt is the median federal loan debt at graduation. Missing values are shown as — and excluded from that variable's normalisation. Score is the weighted sum of standardised variables under the weights in Build your ranking; the default weights are listed in Methodology.
What matters to you? The table above recomputes as you move these.
Sources. All institution-level variables are from the U.S. Department of Education College Scorecard, most recent cohorts file (published June 2026), which combines IPEDS, the National Student Loan Data System and Treasury earnings records. No proprietary survey is used. No institution is asked to rate another.
Coverage. Institutions whose predominant degree is the bachelor's and with at least 500 undergraduates: …. Non-U.S. institutions are not covered because no comparable public data exists for them; we do not estimate them.
Weighting. Each variable is standardised (mean 0, standard deviation 1, higher is better; cost, debt, admission rate and students per faculty are reversed) and combined with the weights below. Default weights:
| Variable | Weight |
|---|---|
| median earnings ten years after entry | 25% |
| graduation rate | 20% |
| net cost | 20% |
| median debt at graduation | 15% |
| students per faculty | 10% |
| admission rate (selectivity) | 10% |
| Total | 100% |
Missing data. A variable that is missing for an institution contributes zero to its score (the population mean), not a penalty; the count of missing variables is available in the data file. Earnings are suppressed by the Department for small cohorts.
Inflation. Cost, debt and earnings are as published; the Department reports them in the dollars of the year collected. Earnings are for a cohort that entered ten years before measurement, so they describe people who chose in the past under conditions that differed from yours.
Uncertainty. Score differences of less than 0.1 standard deviation are within the noise of the inputs and should not be read as an ordering. Rank positions near each other are interchangeable.
Graduate earnings can be measured. Educational quality can be approximated. Debt can be counted. Selectivity can be observed.
Whether you should spend four years of your life at any of these institutions is not present in the source data.
| Variable | Status |
|---|---|
| Will I meet people I love? | unavailable |
| Will I become more curious? | unavailable |
| Will I still care about this subject at 27? | unavailable |
| Would I choose this if nobody were impressed? | unavailable |
| What else might I have done? | no comparison group |
Methodology last revised: June 2026.
Answers are not stored and do not change the table. They are for you.
Common objection #1: "You need a degree."
DadYou should get the degree first. Then you can do whatever you want.
YouWhy first?
DadBecause you'll always have it.
YouDo I always want it?
DadThat's not the point.
YouWhat is the point?
DadSecurity.
YouAgainst what?
DadCan we not do this tonight?
YouDo you want another historical example of someone questioning whether schooling and education are the same thing?
DadNo.
YouOne?
DadFine.
Sort by:
| Name | Dates | Criticised | Attended university | Taught in one | Quotation status | Would tell you to go |
|---|
Results for parent would accept this example:
No historical critic met the criterion.
| Famous people who left university and became extremely successful | People who left university and later thought "that was probably a mistake" |
|---|---|
| Steve Jobs (Reed College, one semester; kept auditing) Bill Gates (Harvard, left; funds schools) … | insufficient public biography |
Historical examples are an unusually bad randomized controlled trial.
MumI just don't want you to make your life harder.
YouI know.
MumI had fewer choices than you do.
YouI know.
MumThat's why I'm trying to tell you which one to choose.
…
MumFine. What would you do instead?
Compared with the institutions above, under the same headings where data exists:
| Option | Cost | Learning | People | Opportunity | Reversibility | You actually want it |
|---|
Cost for institutions is net cost per year from the data above. Every other cell is not in any dataset. The last column is not in any dataset for any row.
YouDo you want another historical example?
MumNo.
YouWhy not?
MumBecause eventually you have to decide what you are going to do.
you are the remaining data point.
The table is built from data.json (1,520 rows, extracted from the Department's file) and the archive from list.json (every entry carries its sources). Both files are public domain except where an entry's source says otherwise. This site is not affiliated with any institution or ranking publisher.