Dispatch №2: college costs lagged the CPI benchmark at ~88% of eligible schools

We pulled the IPEDS 2023-24 academic-year charges file. It reports cost schedules for 2020-21 and 2023-24. For 2,224 IPEDS four-year institutions with total cost of attendance above $5,000 at both endpoints, we added in-state tuition and fees, on-campus room and board, books, and other on-campus expenses.

For the eligible IPEDS cohort, we compared reported cost increases with a BLS CPI-U annual-average benchmark: the index moved from 258.811 in calendar 2020 to 304.702 in calendar 2023, a ~17.7% increase.

1,955 of 2,224 eligible IPEDS four-year institutions (87.9%, ~88%) had reported 2020-21-to-2023-24 total-cost growth below the 2020-to-2023 calendar CPI-U benchmark. The other 269 schools, or 12.1% of that cohort, were at or above the benchmark. In this eligible IPEDS cohort, median reported cost growth from 2020-21 to 2023-24 was +8.5% at public four-year schools and +9.6% at private nonprofits.

That is a sharp finding. It also needs a sharp boundary: these are academic-year price schedules compared with a calendar-year CPI start-year proxy. This is not an exactly matched academic-year inflation adjustment, and it does not prove that every student's bill fell in real dollars.

Why the headline feels backward

Sticker prices can rise while rising more slowly than a benchmark. A school moving from $40,000 to $44,000 raised its nominal price 10%; that is below a ~17.7% calendar CPI-U benchmark. The comparison tells you about the pace of change, not whether $44,000 is affordable.

And tuition is only one part of total cost. Room and board, books, and other expenses also move. Calling one component “the cost of college” can miss the full bill.

What a family can pay is another question. We did not compare these prices with household income or wages, so this dispatch cannot tell you whether college became more affordable for a particular family.

What we are NOT claiming

College is not suddenly affordable. Loans are still loans. We measured reported cost schedules for a defined group of schools against a stated inflation proxy. A different time alignment, aid package, housing choice, or school can change the answer for a student.

What this changes for someone deciding right now

Ask for the current total bill and the aid offer at your school. For the separate question of long-run return, run the college-ROI calculator with your own assumptions. Its FREOPP earnings model does not classify schools by this historical cost-growth benchmark.

Footnotes for skeptics

• Source and periods: NCES IPEDS 2023-24 Institutional Characteristics academic-year charges and HD files, with cost schedules for academic years 2020-21 and 2023-24. BLS CPI-U, US city average, all items, annual averages, calendar 2020 (258.811) and 2023 (304.702). The calendar-year index ratio is 304.702 / 258.811 − 1, or ~17.7%; it is a proxy, not an academic-year-matched deflator.
• Universe: HD four-year institutions with reported total cost of attendance strictly above $5,000 at both endpoints; 2,224 schools across 58 IPEDS jurisdiction codes. No 50-states-plus-DC filter was applied: the cohort includes territories and freely associated states, while the US-city CPI-U proxy was applied to all of it. A missing in-state tuition-and-fees value makes the endpoint missing. The source generator treats missing non-tuition components as zero, which can understate a school's reported total.
• Total cost: IPEDS CHG2AY + CHG5AY + CHG4AY + CHG6AY: in-state tuition and fees, on-campus room and board, books and supplies, and other on-campus expenses. We use these reported values without adjusting for changes in a school's reporting method.
• Counting and medians: 1,955 eligible IPEDS institutions are strictly below the unrounded CPI index ratio; 269 are at or above it, including any exact ties. The +8.5% public and +9.6% private-nonprofit medians use unrounded school-level endpoint ratios and the standard middle-two average for even-sized groups, rounded once to one decimal place.

Correction — 2026-09-28

The version first published on 2026-05-09 incorrectly claimed 1,796 of 2,224 schools (80.8%) were below an approximately 18% inflation comparison. That count does not reproduce from the original IPEDS snapshot; the data were not revised after publication. Recomputing against the unrounded CPI annual-average index ratio gives 1,955 (87.9%) below and 269 (12.1%) at or above. A literal 18.0% cutoff would give 1,971 below; 16 schools lie between the unrounded CPI proxy and that rounded threshold. We classify schools using the unrounded index ratio throughout this corrected issue. We also corrected the public median: the standard median of unrounded school-level changes, rounded once, is +8.5%; the private-nonprofit median is +9.6%. The previous “college got cheaper in real dollars” headline overstated what this calendar-to-academic comparison can establish.

We'll see you next dispatch.