October 2026 release

Data: College Scorecard, October 15, 2026.

• School earnings now use College Scorecard's figure for five years after graduating instead of four. Why: the Department of Education found that its four-year school figure counted some graduate students as undergraduates, at schools that also offer graduate degrees, and it did not update that figure in this release (the Department's errata). The five-year figure was updated. At Harvard, for example, the five-year figure is $107,987 and the four-year figure on Scorecard's own page is $134,794. Part of that gap is a different group of graduates, so not all of it is the error.
• Starting pay taken from a school's figure now backs out four years of 2% raises instead of three, because the figure is measured a year later.
• Program earnings stay on the four-year figure, which the errata did not affect. Programs are compared only with other programs.
• The earnings benchmark for each kind of school is now calculated from College Scorecard data: the median five-year figure across operating schools of that kind. Scorecard publishes a benchmark only for the four-year figure.
• Schools that award only graduate degrees no longer show undergraduate earnings, and the guided interview no longer offers them for a certificate, associate or bachelor's degree. Why: the errata says those figures were an error.
• Net price, full cost, graduation rates, earnings ten years after starting, the share earning more than a high school graduate, and program debt are updated to the new release.
• Three programs that College Scorecard no longer reports left the "not scored" list, among them Penn's nursing certificate, which we had shown as a post-graduate certificate.
• The Standard Calculator's benchmarks now come from the same figures as the rest of the site. Before this release they still held the March 2026 values.
• How the national summary moved: at bachelor's schools, the share that falls short of the no-degree path at full cost went from 79.0% to 80.9%, and at net price from 57.3% to 59.0%. Across all levels it stayed near 68% at full cost. About 180 schools that were scored on the four-year figure have no five-year figure, so they are no longer scored, and about 60 are scored for the first time.

October 15, 2026 update

Data: College Scorecard, June 10, 2026 (unchanged).

• Some certificate, associate and master's programs pay more than a higher degree in the same field at the same school, usually because students arrived already holding a degree or years of experience. We now show their earnings but no Lifetime Value Added, and leave them out of rankings (54 programs). Why: the model assumes a student who starts right after high school, or right after a bachelor's degree, so it can't measure what these programs added.
• Penn's nursing certificate was also shown as a post-graduate certificate.
• Branch campuses that report one shared earnings figure are counted once.
• Programs College Scorecard lists as non-credential were removed.
• In the guided interview, the master's and doctorate paths can now use a specific school. The card uses that program's earnings and debt.
• The national summary did not change.

October 14, 2026 update

• Site maintenance: an analytics tag on every page, and sitemap links fixed. No change to data or results.

October 12, 2026 release

Data: College Scorecard, June 10, 2026 (the book uses March 23, 2026).

• School data moved from the March 2026 release to June 2026. June changed the school list, the benchmarks and some schools' degree classes. It did not change school earnings, prices or graduation rates.
• Federal taxes use the 2026 tables by default. The 2025 tables stay as an option, with the standard deduction as amended in July 2025.
• Program length follows the degree a school mostly awards: one year at certificate schools, two at associate schools, four otherwise. Why: community colleges that add a few bachelor's programs were being priced as four-year schools against their school-wide earnings.
• Each school is compared with peers at the level it is priced at.
• Major pages use the American Community Survey 2020 to 2024 and the BLS job projections for 2025 to 2035.
• A warning now marks schools whose single earnings figure mixes in longer programs, and one age cutoff (25) applies on every page.
• How the national summary moved: across all levels, the share that falls short at full cost went from 71.9% to 67.7%. At bachelor's schools it went from 79.2% to 79.0%.

Starting point: the book edition

Data: College Scorecard, March 23, 2026, with the 2025 tax tables. The book's edition of this tool stays online at collegeroi.org/roi-book.