Every kind of educational institution is judged on one number that moves real money: an appropriation, a grant formula, an aid budget, a loan eligibility ruling, a contract renewal. What that number is changes by sector. A private college board reads the tuition discount rate first. A superintendent in an attendance-funded state reads average daily attendance, the count of students in seats each day, because the state pays on that count. A research university watches what it collects of the overhead rate it negotiated on federal grants. Twenty-four other institution types sit alongside those three, each with its own version, and a plan built for one of them mismeasures the rest.
The map below covers all twenty-seven, grouped into six sectors. Each row names the metric in plain words, says who watches it and what money is attached, and carries the sharpest figure we can publish, with its source and year in the cell. Where no defensible public figure exists, the row says so and stays qualitative.
Figures we computed from public data are labelled computed, and the arithmetic sits in the methodology at the end. Rows point to our full research pieces where one exists; the row is the summary and the piece is the argument. Where two published numbers conflict, the row carries the one that survives checking, and a widely quoted number with no traceable source does not appear at all.
Undergraduate institutions
The undergraduate map splits on who pays. States fund the publics through formulas, and private colleges fund their own aid out of tuition they chose to forgo.
Community colleges
Judged on
The start rate: the share of admitted students who ever register for a class.
Who watches it, and what moves
There is no deposit to watch. The loss happens between admission and the first day of class, and the state pays on the students who enroll.
The sharpest figure
13.3% of state appropriations at two-year public colleges flow through performance formulas, a larger share than the 9.0% at four-year publics, and net tuition runs $2,668 per full-time student (SHEEO, FY2025).
Independent colleges
Judged on
The tuition discount rate: the share of gross tuition given back as institutional aid.
Who watches it, and what moves
The board reads the discount rate before anything else. Every point of it is aid the college funds from its own budget, about $450,000 a year at a 1,000-student entering class (computed).
The sharpest figure
The first-year discount was 54.5% in 2024-25, with 57.1% estimated for 2025-26 (NACUBO Tuition Discounting Study, 2025, n=258).
Liberal arts colleges
Judged on
Net tuition revenue from each entering class.
Who watches it, and what moves
A small residential college is tuition-dependent enough that one under-filled class reaches the operating margin the same year, and lenders read that margin against debt covenants.
The sharpest figure
Moody's projects 16% of private colleges will run a negative operating margin in 2026 (Moody's Ratings, 2026 outlook).
Public universities
Judged on
Completions and credit-hour progression.
Who watches it, and what moves
The budget office watches completions the way a private college watches deposits. Legislatures pay part of the appropriation on credit-hour progression and degrees, and several state formulas pay an outcomes premium for students from focus populations; Tennessee's current formula sets that premium at 40 percent (Tennessee Higher Education Commission).
The sharpest figure
9.0% of state appropriations at four-year publics run through performance formulas (SHEEO, FY2025), and one point of first-year retention at a 4,000-student entering class is about $780,000 in year one (computed from SHEEO FY2025 revenue per student).
Research universities
Judged on
The effective overhead recovery rate: what the university collects of the facilities and administrative rate it negotiated on federal grants.
Who watches it, and what moves
The negotiated rate averages 56.9% while the effective rate collected is 44.3% (COGR 2023 survey, published December 2024), and one point of effective recovery on a $200 million research base is $2 million a year, recurring (computed).
The sharpest figure
$7.07 billion of indirect cost went unrecovered in FY2024, inside $30.2 billion of research that institutions funded themselves (NSF Higher Education Research and Development survey, FY2024).
Full analysis: Universities now pay for a quarter of their own research
Historically Black colleges and universities (HBCUs)
Judged on
First-year retention.
Who watches it, and what moves
Title III-B, the federal strengthening program for these colleges, allocates money on three counts: students receiving Pell grants (the federal grant for lower-income students), graduates, and graduates who go on to graduate or professional study in fields where Black students are underrepresented. A student kept and graduated appears in the formula more than once.
The sharpest figure
One retention point is worth about $30,000 in year one at the median institution (computed from fiscal 2017 IPEDS finance data), and the program awarded $396 million in FY2023 at an average award near $4.0 million (U.S. Department of Education program funding tables, FY2023).
Hispanic-Serving Institutions (HSIs)
Judged on
The designation itself: the enrollment share and spending tests that make a college eligible for Title V.
Who watches it, and what moves
Title V, the federal grant program the designation unlocks, tests eligibility three ways, and two of the tests are poverty tests, one of them low educational and general spending per student. By federal definition an HSI spends less per student than its peers.
The sharpest figure
On September 10, 2025, the Department of Education ended discretionary awards for seven minority-serving programs, including Developing Hispanic-Serving Institutions: roughly $350 million discontinued, with about $132 million of mandatory funding continuing (U.S. Department of Education, 2025).
Faith-based colleges
Judged on
The same discount arithmetic as any independent college, plus enrollment as evidence of mission health.
Who watches it, and what moves
The board watches whether net tuition covers the operating budget, and the sponsoring body watches enrollment as the sign that the mission still draws students.
The sharpest figure
This row is qualitative: no faith-specific public benchmark exists, and the private-college discount figures in the independent colleges row are the numbers these boards use.
Art and design colleges
Judged on
Net tuition per enrolled student and the portfolio funnel that produces it.
Who watches it, and what moves
For every enrolled student, 8.6 portfolios are reviewed and 4.0 admits are produced, and 99% of full-time first-years receive institutional grants at a median of $20,234 against a $46,343 sticker price (computed from IPEDS data across 42 specialized private nonprofit colleges, 2026). The seat's value is what remains after the award.
The sharpest figure
Net tuition is $28,202 per student and 64% of all revenue; one seat carries $112,808 across four years (computed from IPEDS, 2026).
In all nine rows the money follows a count of students, and each row counts a different moment. The largest figure in the group is the roughly $9.46 billion states distribute through performance formulas (SHEEO, FY2025 tables, computed).
Graduate, professional and career programs
Above the bachelor's degree, the unit of account shifts from the institution to the program. The program's price, its ranking inputs and now its federal borrowing cap are all set at program level.
Graduate schools
Judged on
Net tuition revenue by program.
Who watches it, and what moves
Graduate programs produce 39% of institutional net tuition while the graduate enrollment office runs on about $912,000 against undergraduate admissions' $4.37 million (EAB, December 2024).
The sharpest figure
Only 6% of graduate enrollment leaders cannot say whether they hit headcount; 34% cannot say whether they hit their net tuition revenue goal (EAB and NAGAP, 2025).
Business schools
Judged on
The surplus the school transfers to its parent university, with placement as the public gauge.
Who watches it, and what moves
The dean's year is graded on two tables at once: the transfer to the parent university, which about three-quarters of business schools make (AACSB, 2026), and the placement outcomes that carry 50% of the ranking weight (U.S. News).
The sharpest figure
The median net transfer to the parent university is $7.2 million (AACSB, 2026), and the projected U.S. MBA median starting salary fell to $120,000 (GMAC Corporate Recruiters Survey, 2026).
Law schools
Judged on
The entering class's median score on the Law School Admission Test (LSAT), and the scholarship spend that buys it.
Who watches it, and what moves
Rankings run on entering credentials, and schools bid for them with institutional aid. The national median rose 1.79 points in one year, and 28 schools now hold a median of 170 or higher against 5 a decade ago (ABA 509 data, 2025).
The sharpest figure
80% of full-time law students receive institutional grants, median $24,176 a year (ABA 509 data, AY2024-25), roughly $57,000 of discount across a three-year degree (computed from the NACUBO and AccessLex 2016 study, n=36).
Engineering programs
Judged on
How many students reach upper-division standing, where the premium tuition attaches.
Who watches it, and what moves
Because differential tuition at public universities almost always begins at upper-division standing, the program teaches its most expensive introductory years at the standard rate and collects the premium only from students who reach the upper division.
The sharpest figure
ABET, the engineering accreditor, lists 4,863 accredited programs at 950 colleges and universities (ABET, 2026). No comparably sourced public figure prices the differential itself, and the tuition mechanics stay qualitative for that reason.
Teacher preparation
Judged on
Completer counts and state program approval.
Who watches it, and what moves
States approve programs and count completers, and accreditors review on multi-year cycles. Approval status and reputation move enrollment; very little money rides on any single metric.
The sharpest figure
This row is qualitative: there is no published revenue per candidate, no published cost per candidate and no published cost of placement coordination anywhere in the public record.
Trade and technical colleges
Judged on
The earnings premium: whether completers out-earn a high-school graduate.
Who watches it, and what moves
Since July 1, 2026, undergraduate completers must out-earn a high-school graduate aged 25 to 34, measured in year four, and a program failing the test in two of three years loses Direct Loan eligibility (STATS and Earnings Accountability rule, U.S. Department of Education, 2026).
The sharpest figure
The test is universal across every sector. An institution with more than half of its federal aid recipients in failing programs loses Pell grants, the federal aid for lower-income undergraduates, for those students, and twenty tipped-income program codes, cosmetology among them, are delayed to 2028 (U.S. Department of Education, 2026).
The borrowing caps sit over this whole group. Grad PLUS, the federal loan program that let graduate students borrow up to the full cost of attendance, ended for periods of instruction beginning on or after July 1, 2026. The replacement caps are $20,500 a year for master's students and $50,000 a year for programs on the professional list, assigned by instructional program code (Public Law 119-21; Federal Student Aid, GENERAL-26-42, 2026).
Clinical programs
Clinical education breaks the enrollment pattern, because the scarce thing is the seat itself.
Medical schools
Judged on
The filled seat-year.
Who watches it, and what moves
The accreditor's transfer standard requires that transfer students be comparable and allows final-year transfer only in rare and extraordinary circumstances (LCME Standard 10.7). In practice many schools accept no transfers at all, and a vacated first-year seat stays empty for the remaining years.
The sharpest figure
A first-year departure costs about $224,000 at a private school and about $131,000 at a public one across the remaining program (computed from AAMC 2025-26 median tuition and fees: $74,661 private, $43,648 public in-state).
Full analysis: Every college buys its class twice
Nursing programs
Judged on
Program approval status with the state board of nursing.
Who watches it, and what moves
The state board of nursing grades the program itself, and clinical placement capacity caps enrollment. Approval status decides how many seats the program may fill.
The sharpest figure
13.3% of programs, 226 of 1,695, already operate under less than full approval status (NCSBN National Nursing Education Database, 2023-24).
K-12
Students arrive by geography here, so the money attaches to counting: who is enrolled, who is present, and on what day.
K-12 districts
Judged on
Average daily attendance: the count of students actually in seats each day, in the states that fund on it.
Who watches it, and what moves
California, the largest attendance-funded state, prices one instructional day at exactly one one-hundred-and-eightieth of the apportionment (Education Code 46200). Its funding floor is the highest of the current year, the prior year or the three-year average, which means a decline reaches the budget only after it has pulled the average down.
The sharpest figure
One day of one high-school student is $70.81 in 2025-26 (computed from the California Department of Education apportionment rate of $12,746 per student in average daily attendance), and 22.6% of students were chronically absent in 2024-25 against about 15% before the pandemic (American Enterprise Institute, June 2026, 45 states with 2025 data).
K-12 career and technical
Judged on
The Perkins program quality indicator the state chose: the credential, the dual credit or the work-based learning placement.
Who watches it, and what moves
Perkins, the main federal grant for career and technical education, scores districts against negotiated performance levels: a district that misses 90% must run an improvement plan, and two consecutive years below 90% let the state withhold the grant (Perkins V, 20 U.S.C. 2343). The evidence for the indicator lives with testing vendors, college registrars and employers, none of whom work for the district.
The sharpest figure
Federal Perkins money reaching school districts works out to $80.27 per secondary career and technical student per year (computed from U.S. Department of Education FY2025 distribution data over 2023-24 enrollment).
Charter and virtual networks
Judged on
Enrollment on the count day, and the authorizer's rating at renewal.
Who watches it, and what moves
In single-count-date states the whole year's funding is set by who is enrolled and counted on one day; the same absence is worth either nothing or a full year's funding depending on the date. A larger share of a network's revenue comes from the state than a district's does.
The sharpest figure
This row is qualitative: published tallies of single-count-date states conflict with each other, and no per-family recruitment cost has a public source.
States, systems and networks
The buyers above the institution are judged on numbers they mostly do not produce themselves.
States
Judged on
Whether the state can produce the outcome numbers its own statutes demand.
Who watches it, and what moves
Legislatures mandate the longitudinal data system, the state database that follows students across years and schools, and the governor's IT office often buys it. One published state procurement shows the shape: mandated by statute, a fourteen-month build, four one-year options.
The sharpest figure
70% of responding states report inadequate staffing for research and analytics, and fewer than half consider their longitudinal data system's funding sustainable (Education Commission of the States survey, published November 17, 2025).
Systems
Judged on
Total education revenue per student across the campuses.
Who watches it, and what moves
A system office is judged on the appropriation it wins and how it divides it, and outcome formulas are applied at the state and system level, upstream of any campus budget.
The sharpest figure
State appropriations per full-time student fell 1.0% in FY2025, the first per-student decline since 2012, while total education revenue reached $19,443 per full-time student (SHEEO, FY2025).
Community college districts
Judged on
District-level apportionment, and the discount applied for being a multi-college district.
Who watches it, and what moves
California computes the apportionment for the district as a whole, and the basic allocation pays a multi-college district less per college on the assumption that shared services save money; per-student spending data shows the assumed saving was never delivered (computed from IPEDS institutional support spending, 2026).
The sharpest figure
The multi-college discount is about $2.22 million a year per large college and about $33.3 million a year across the sixty-six California colleges inside districts (computed from California's community college funding formula, 2026), and first-year transfer-level math and English pay $1,510 to $2,463 per student in the success allocation (California funding formula, 2026), more than any credential that student earns three years later.
Consortia
Judged on
Budget per member: whether the consortium runs programs or only shared infrastructure.
Who watches it, and what moves
From public tax filings for FY2023, one consortium runs $53.9 million across seven members, about $7.7 million each, while another runs $2.2 million across fourteen members, about $156,000 each; the difference is whether it operates programs.
The sharpest figure
The per-member spread is roughly fiftyfold (computed from IRS Form 990 filings, FY2023; the consortia are unnamed here by design).
International groups
Judged on
Whether the network itself adds anything beyond its campuses.
Who watches it, and what moves
Cross-border groups are judged on portfolio enrollment and on whether shared scale ever shows up in a campus's results, and the sector's own history runs against the synergy story: listed portfolios have shrunk to a fraction of their size within a decade while enrollment held up far better than the campus count did.
The sharpest figure
This row carries no figures by design; the evidence for this segment is portfolio history, and it is qualitative.
Online divisions and operators
Two players share the online student and keep separate books. The institution owns persistence, and the operator owns acquisition. No figure crosses between the two rows.
Online and adult divisions
Judged on
Term-over-term re-registration.
Who watches it, and what moves
An adult learner mostly has no deposit, and the loss shows up term over term. What blocks the return is often small: the median past-due balance behind a registration hold in Ohio was $735 (Ithaka S+R, November 2023).
The sharpest figure
A student who starts at 25 or older has a 43.6% chance of reaching a second fall, against 77.1% for the entering cohort as a whole (National Student Clearinghouse, fall 2024 cohort, published June 2026).
OPMs
Judged on
Payback: the years a partner program takes to return the operator's upfront investment.
Who watches it, and what moves
The operator's own capital buys the recruiting and the services for a program it does not own, and the money comes back only as a fixed share of tuition across the life of the contract. The contract term is set by the payback period.
The sharpest figure
Initial contract terms run seven to fifteen years, and the operators' own securities filings put program payback at roughly three to five years (SEC filings of publicly reporting operators, 2014-2026; operators are unnamed here by design).
Methodology
Every row carries the sharpest figure with a public source that survived checking. Where no defensible public figure exists, the row says so plainly and stays qualitative; that is a deliberate standard, and it applies to faith-based colleges, teacher preparation, charter and virtual networks, engineering tuition mechanics and the whole international groups row.
Figures labelled computed are our arithmetic on published inputs, shown here so a reader can reproduce them. The twenty-eight-fold seat spread divides the AAMC private medical median ($74,661) by SHEEO's two-year net tuition per full-time student ($2,668); the two inputs measure tuition differently, net revenue on one side and posted tuition and fees on the other, and the finding names both. The $450,000 discount point assumes gross tuition near $45,000 a student across a 1,000-student entering class; one percentage point of that is $450,000 a year. The $780,000 retention point multiplies 40 students by SHEEO FY2025 education revenue per full-time student; the HBCU $30,000 point multiplies roughly 3 students by $9,539 of student-attached revenue per year, a fiscal 2017 figure, because the public IPEDS finance series available through the Urban Institute portal ends at fiscal 2017. Pairing that revenue figure with current retention practice mixes vintages, and the sentence on the page says so. The $80.27 Perkins figure divides $693,728,790 of FY2025 secondary distributions by 8,642,099 secondary participants reported for 2023-24, a one-year offset we disclose. The $70.81 school day divides California's 2025-26 adjusted 9-12 rate of $12,746 by the 180-day divisor that Education Code 46200 itself encodes. The medical departure figures multiply the AAMC 2025-26 median tuition by three remaining years. The consortium spread divides $7.7 million per member by $156,000 per member.
The Title III-B pair is stated as the coherent FY2023 pair ($396 million, $4.0 million average award), because the widely circulated $363 million belongs to a different fiscal year than the $4.0 million average. The professional borrowing list is described as an interim list of program codes, without a count, because the announcement itself states no count. All HBCU finance figures carry their fiscal 2017 vintage. The combined performance-funding total is stated as about $9.46 billion, computed from SHEEO's live FY2025 tables ($3.71 billion at two-year colleges plus $5.75 billion at four-year institutions); a smaller figure in circulation conflates the total with SHEEO's net-tuition-per-student figure. Each of these constructions comes from the verification stage, which checks built pages against live primary sources.
Some strong numbers are absent on purpose. No law school melt rate exists, so none appears. Nursing licensure exam pass rates and attrition figures are excluded. National per-pupil expenditure dollars do not appear on the district row because they describe no actual district. The widely quoted HBCU six-year completion figure appears nowhere here, because it misleads unless its transfer-out detail travels with it, and this page had no room to carry both.