Nobody in EADA has a bad year
Zero. That is how many athletics deficits appear in the Equity in Athletics Disclosure Act file across 40,829 institution-year records, AY 2005-06 through AY 2024-25. Twenty years of submissions from roughly two thousand institutions a year, and not one of them reports having spent more on athletics than athletics brought in.
The latest year has the same shape. In AY 2024-25, 2,037 institutions report both a revenue total and an expense total. 1,288 of them, 63.23 percent, report the two figures exactly equal, to the dollar. The other 749 report revenue above expense. None report revenue below it. Across all 2,037 institutions the smallest revenue-minus-expense gap is exactly $0, and the median gap is exactly $0 as well.
A column that has never once gone negative is not measuring profit. It is measuring an allocation. The institution decides how much of the money that reached athletics gets booked as athletics revenue, and for most institutions the answer is: as much as athletics spent.
If you have ever opened the EADA file, pulled the two grand-total columns and subtracted, you were doing the obvious thing and the file invites it. The rest of this is what the subtraction produces, and the short answer to give a provost who wants to know whether athletics makes money.
Which two columns, and why
Two columns carry the whole argument: grnd_total_revenue and grnd_total_expense on eada.fact_institution_yearly, which holds one row per athletics department per academic year. The catalog defines them as grand total athletic revenue across all sports and sources, and grand total athletic expenses across all sports and categories. They exist only at the department grain, so any question about whether athletics pays for itself has to start here rather than at the sport table.
They are also defined arithmetically, and the definition is itself a tell. Department revenue equals the sum of every sport's revenue plus a bucket attributed to no sport at all. We tested that identity on all 2,037 institutions in AY 2024-25 and it holds on both sides for every one of them. That bucket is large: $7,443,059,737 of revenue, 26.3 percent of the national total, and $8,390,142,873 of expense, 30.4 percent. Roll the sport table up to the institution and you silently drop all of it.
The finding does not hinge on that column choice. Do the rollup anyway and you get 1,239 institutions exactly balanced (60.82 percent), 798 above, and again zero below. A third department-total definition, revenue_menall plus revenue_womenall against the matching expense pair, returns the same 1,239 and the same zero. Two other column pairings do produce shortfalls, and neither is a department total: total_rev_menwomen excludes coed teams and puts 9 institutions below, and the il_ operating pair puts 139 below. If you reproduce a deficit, check which pair you used.
One block to avoid on the way in. 111 of the 160 institution-level columns carry an il_ prefix, and the revenue and expense members are operating figures only. SUM(il_total_revenue_all) is $8,570,550,288 against a grand total of $28,289,565,477, or 30.3 percent. Read it as a department total and you understate athletics revenue by roughly 70 percent.
The one year the file does record shortfalls
The reasonable objection is that the loader cannot hold a negative number, so the zero says something about the pipeline rather than about athletics. It does not, and the file settles the question itself.
AY 2004-05 is the one academic year where institutions do report shortfalls. 70 of 1,978 institutions report revenue below expense, and the largest reported shortfall is $8,473,903. Negative gaps are storable. From AY 2005-06 onward they stop appearing, and that is a property of what institutions report.
Two earlier years cannot be used at all. grnd_total_revenue and grnd_total_expense are null for all 1,965 rows in AY 2002-03 and all 1,973 rows in AY 2003-04. Any financial series out of EADA begins at AY 2004-05.
Seventeen years, zero deficits
This is not a recent artifact either. The exact-balance rate has climbed from 53.58 percent in AY 2008-09 to 63.23 percent in AY 2024-25, a rise of 9.65 percentage points, with small dips along the way. The deficit column is zero in all seventeen years.
The balance test by academic year, department grand totals
| Academic year | Institutions reporting | Revenue exactly equals expense | Revenue above expense | Revenue below expense |
|---|---|---|---|---|
| AY 2008-09 | 2,053 | 1,100 (53.58%) | 953 | 0 |
| AY 2012-13 | 2,090 | 1,218 (58.28%) | 872 | 0 |
| AY 2016-17 | 2,062 | 1,234 (59.84%) | 828 | 0 |
| AY 2019-20 | 2,073 | 1,266 (61.07%) | 807 | 0 |
| AY 2020-21 (COVID year) | 1,844 | 1,114 (60.41%) | 730 | 0 |
| AY 2022-23 | 2,040 | 1,284 (62.94%) | 756 | 0 |
| AY 2024-25 | 2,037 | 1,288 (63.23%) | 749 | 0 |
AY 2020-21 loses 229 institutions against the year before, down to 1,844 from 2,073, and the sport table loses 3,231 rows. Coverage recovers the following year. Say so whenever 2020 appears in a series you publish.
Across the whole span, 23,237 of the 40,829 institution-year records, 56.91 percent, report the two figures exactly equal. Nationally the reported gap has run in favor of revenue in every year sampled: $369,899,417 in AY 2008-09, $626,042,823 in AY 2016-17, $740,773,123 in AY 2022-23, $663,934,941 in AY 2024-25. Reported national athletics revenue grew from $11,237,185,346 to $28,289,565,477 over that span, a 2.52-fold increase. None of those gap figures is a profit number, and the rest of this piece is about why.
The big-money divisions balance more often, not less
The expected story is that the big programs look different because they have real money moving through real contracts. They do look different, in the opposite direction from the one people guess.
Group the 358 NCAA Division I institutions (FBS, FCS, and Division I without football) against everyone else. Division I reports exact balance 74.9 percent of the time, 268 of 358, against 60.8 percent for the other 1,679 institutions. Those 358 institutions carry $21,246,969,740 of the $28,289,565,477 in national athletics revenue, 75.1 percent of it. The best-resourced athletics finance offices in the country, with the most dollars at stake, are the ones most likely to close the year at exactly zero.
Inside Division I the ordering is FCS at 80.9 percent (106 of 131), Division I without football at 77.8 percent (77 of 99), and FBS at 66.4 percent (85 of 128). The three least balanced of the 19 classifications are all small-college or community-college groups: NWAC at 46.9 percent (15 of 32), CCCAA at 49.6 percent (56 of 113), NCAA Division III with football at 52.3 percent (123 of 235).
So the reading that small colleges balance because they lack the accountants to do anything else runs backwards. And across all 19 classifications in the file, the count of institutions reporting a deficit is zero.
Exact balance rate, the three Division I tiers against the three least balanced classifications of 19, AY 2024-25
What does not explain it: sector, size, or luck
Three obvious explanations all fail on the data, which is useful, because each one is a memo somebody is about to write.
Sector explains almost nothing. Public 4-year institutions balance exactly 63.7 percent of the time (378 of 593), public 2-year 63.1 percent (365 of 578), private nonprofit 4-year 62.0 percent (516 of 832). That is a 1.7 point spread across the three sectors holding 2,003 of the 2,037 institutions. Whatever produces exact balance is not a governance or ownership effect.
Size explains nothing either. The median athletics budget of an exact-balance institution is $3,747,301 against $4,084,248 for the institutions reporting a surplus, a difference of 9.0 percent. The two groups are the same size of operation doing two different things with the last line.
And it is not year-to-year variation. 2,024 institutions appear in both AY 2023-24 and AY 2024-25, and 89.0 percent of them report the same balance behavior in both years: 1,168 balance exactly twice, 634 report a surplus twice, 222 switch. Institutions adopt a convention and keep it. That is also why ranking institutions by reported surplus produces a league table of reporting conventions rather than of financial health.
Where surpluses do get reported, they are small relative to the budget. Among the 749 institutions reporting one, the median surplus is $141,466, the quartiles are $33,533 and $400,708, and the largest in the country is $65,693,046.
A quarter of a billion dollars, matching to the dollar
Five of the fifteen largest athletics revenue lines in the country report revenue and expense identical to the dollar, not close to it.
Selected large athletics programs, AY 2024-25
| Institution | Reported revenue | Reported expense | Gap |
|---|---|---|---|
| Ohio State University-Main Campus | $295,265,883 | $295,265,883 | $0 |
| The University of Tennessee-Knoxville | $285,364,027 | $285,364,027 | $0 |
| University of Oklahoma-Norman Campus | $234,426,470 | $234,426,470 | $0 |
| University of Southern California | $234,029,848 | $234,029,848 | $0 |
| University of Miami | $230,484,608 | $230,484,608 | $0 |
| The University of Texas at Austin | $343,106,827 | $277,413,781 | $65,693,046 |
Texas at Austin reports the largest gap of any institution in AY 2024-25. Ohio State reports $295,265,883 on both lines. Neither of those is a statement about which department is in better shape. They are two different answers to the question of how much revenue to book, filed under the same instructions.
Deficits exist. They just do not reach the total line
Drop one level down and the shortfalls appear. In AY 2024-25 the sport table holds 18,045 institution-sport rows. 11,107 of them (61.6 percent) report revenue exactly equal to expense, 5,113 report revenue above expense, and 1,825 (10.1 percent) report revenue below expense. Roughly one sport row in ten shows a shortfall.
Follow those rows up to the institution and the shortfalls disappear. 492 institutions report at least one sport with revenue below expense. Of those, 141 report a department that balances exactly and 351 report a department surplus. Zero report a department shortfall. Whatever the individual sports did, the total line closes at or above zero in every submission. That is where the balancing happens, and it happens before the last row is written.
At the other end of the range, 1,148 institutions (56.4 percent) report revenue exactly equal to expense for every sport they field. 570 institutions report no sport balancing exactly.
One structural note before anyone reads a sport number. eada.fact_sport_yearly carries one row per institution per sport covering both the men's and the women's team, not one row per sex. All 1,942 basketball rows in AY 2024-25 belong to 1,942 distinct institutions, and 1,841 of them report participants of both sexes on that single row. Gender splits come from the _men and _women column pairs, never from a row filter.
Sport rows reporting revenue below expense, the three highest classifications of 19 and the other two Division I tiers, AY 2024-25
FBS is the only group that routinely lets individual sports show losses: 28.4 percent of its sport rows report revenue below expense, more than double any other group and about four times the 6.9 to 7.7 percent in the other two Division I tiers. FBS also reports exact balance on only 58.3 percent of its sport rows, against 81.4 percent for Division I without football and 81.0 percent for FCS.
And zero FBS institutions report a department deficit. The division that does the most visible cross-subsidy accounting internally still closes the total line in the black every time.
The football surplus reconciles to zero inside the same file
Football is where this gets expensive to get wrong, because the football number is the one that ends up in a board deck. Nationally in AY 2024-25, 876 institutions report football: $7,991,396,676 of revenue against $5,839,248,699 of expense, a gap of $2,152,147,977. Basketball, reported by 1,942 institutions, shows $4,284,558,776 against $4,124,411,515, a gap of $160,147,261. Volleyball, soccer, baseball, softball, golf, tennis, lacrosse, wrestling, swimming, field hockey and beach volleyball all report aggregate expense above aggregate revenue.
Read literally, that is the familiar story: football and basketball pay for everything else. The same file shows where that money lands. Take the 128 FBS institutions on their own.
The FBS department line, AY 2024-25, reported revenue minus reported expense
| Line in the FBS submission | Revenue minus expense |
|---|---|
| Football | plus $2,123,624,779 |
| Basketball | plus $110,106,170 |
| All other sports (1,259 rows) | minus $825,058,107 |
| Attributed to no sport at all | minus $985,498,055 |
| Sum of the four lines | plus $423,174,787 |
| Department total as filed | plus $423,174,787 |
The FBS department line is $15,086,433,507 of revenue against $14,663,258,720 of expense. Every dollar of the football surplus is already spoken for by shortfalls booked elsewhere in the same submission, and $4,101,098,529 of FBS athletics revenue, 27.2 percent of it, is attributed to no sport at all.
You can also watch the cancellation happen institution by institution. 312 institutions report football revenue above football expense, and 64 of them (20.5 percent) then report the whole department balancing to the dollar, which means the football surplus is exactly offset elsewhere in the same filing. At FBS institutions football carries 43.4 percent of reported department revenue and 30.2 percent of reported department expense. That 13.2 point wedge is the entire "football pays for everything" claim, and it is a bookkeeping wedge. The spread behind it is wide: the median FBS football gap is $125,897, with a minimum of minus $8,609,425 and a maximum of $113,636,406. Of the 128 FBS institutions, 66 report football revenue above expense, 57 report the two figures identical, and 5 report a football shortfall.
Outside FBS, football is hard to distinguish from break-even. Division I-FCS reports plus $14,419,763 on $724,530,810 of revenue. Division II with football, plus $2,931,996 on $290,506,532. Division III with football, plus $6,144,529 on $156,352,623. NAIA Division II, minus $663,594 on $52,134,356. FBS alone accounts for $6,544,977,511 of the $7,991,396,676 in national football revenue, 81.9 percent. The 234 Division III football programs did not independently land within seven million dollars of zero. That is the signature of an allocation.
Men's basketball and women's basketball, same allocated frame
The national basketball aggregate hides a split worth naming. Men's basketball reports $2,892,464,009 of revenue against $2,574,457,372 of expense, a gap of $318,006,637. Women's basketball reports $1,392,094,767 against $1,549,954,143, a gap of minus $157,859,376.
Because both sides are allocations, that gap is not a measurement of what each team earned. It is a statement about how institutions choose to book revenue across two teams inside one submission, filed under a statute whose purpose is athletics equity disclosure. That makes it worth reporting, with the caveat attached.
What EADA is genuinely good for
None of this makes the file useless. It makes it a spending file rather than a financial-performance file, and spending files answer plenty of real questions. The test is simple: a comparison is valid when both sides of it were allocated by the same institution, under the same instructions, inside the same submission.
Relative scale holds up. Athletics department expense spans more than two orders of magnitude, from $543,468 at the 10th percentile to $3,907,990 at the median, $29,228,663 at the 90th and $186,428,633 at the 99th. That ordering is meaningful even though the levels are allocations.
Per-participant spending holds up, and it is the comparison the file was built for. In AY 2024-25, men's teams account for $12,535,046,571 of allocated athletics expense and women's teams $6,700,441,103. Men are 57.9 percent of participants and receive 65.2 percent of allocated spending, which is $28,267 per male participant against $20,784 per female participant, a ratio of 1.36 to 1. On the narrower operating-expense line the divisions separate cleanly: FBS spends $32,229 per male athlete and $14,749 per female athlete, a ratio of 2.19 to 1, while NCAA Division III with football spends $1,785 per male athlete and $2,060 per female athlete. The sign flips below Division I, with women ahead in ten of the fourteen classifications that have at least 50 institutions, and the FBS-to-Division III ratio on male athletes is 18.1 to 1.
Direction of travel within an institution holds up too, because the same filer applies the same convention on both ends. The 1,963 institutions present in both AY 2018-19 and AY 2024-25 went from $19,218,093,299 to $27,505,310,246 of athletics expense, up 43.1 percent, with the median institution up 36.6 percent. 1,751 grew and 212 shrank. AY 2018-19 is the right base year rather than AY 2019-20, which had its spring season cut short.
Six questions, and whether EADA can answer them
| Question | EADA answer | Why |
|---|---|---|
| How big is this athletics department relative to that one | Yes | Department expense runs from $543,468 at the 10th percentile to $186,428,633 at the 99th, and the ordering is meaningful |
| What do we spend per participant, and how does that split by sex | Yes | Both sides come from one submission under one set of instructions: $28,267 per male participant against $20,784 per female participant |
| Is our athletics spending growing, and how fast | Yes | Same filer, same convention on both ends. 1,963 institutions grew athletics expense 43.1 percent from AY 2018-19 to AY 2024-25 |
| Does athletics pay for itself | No | The revenue total is balanced to the expense total by construction. Zero of 40,829 institution-year records report a shortfall |
| How large is the institutional subsidy to athletics | No | Not one of the 281 columns across the two EADA fact tables names a revenue source |
| Which institution is in better athletics financial shape | No | The surplus column sorts by reporting convention: 89.0 percent of institutions repeat the same balance behavior year to year |
"EADA tells you how much the institution decided to book against athletics and how much athletics spent. For roughly two thirds of institutions those are the same number by construction, and no institution has ever filed a shortfall. Use it for scale, for spending per participant, and for direction of travel. Do not use it to answer whether athletics pays for itself."
Running this on your own submission
Five checks, in the order that gets the fastest read on your own filing and your peers. All of them run off the two grand-total columns plus the unallocated pair, and none of them takes longer than an afternoon.
Five checks on your own EADA filing
Subtract your own two grand totals, then look at the last three years
Thirty seconds, and it tells you which convention your office uses
Pull grnd_total_revenue and grnd_total_expense from fact_institution_yearly for your institution across several years. If the difference is $0 every year, your submission balances by construction and no internal narrative about athletics profitability should ever cite it. If you report a surplus, check that it is stable: 89.0 percent of the 2,024 institutions present in both AY 2023-24 and AY 2024-25 reported the same behavior in both years.
- Do not use the il_ prefixed columns for this. They are operating figures only and total 30.3 percent of the grand total nationally.
- Do not roll the sport table up instead. That drops the unallocated bucket, which is 26.3 percent of national revenue.
What you learn: whether your institution is one of the 1,288 that balance to the dollar or one of the 749 that report a surplus, and whether that is a choice your office made once and kept.
Measure how much of your revenue you attribute to no sport at all
tot_revenue_all_notalloc divided by grnd_total_revenue is the share of your athletics revenue that never touches a sport line. The median institution leaves 24.0 percent unattributed, and every one of the 2,037 institutions in AY 2024-25 leaves something: the national unallocated revenue line runs from $1,250 to $136,706,540. The larger that share, the less any sport-level figure from your file can support.
What you learn: the ceiling on how much detail your own sport-level numbers can carry before someone over-reads them.
Check whether any of your sports report a shortfall
At the sport grain, 1,825 of 18,045 rows nationally (10.1 percent) report revenue below expense, and 492 institutions report at least one. If yours is one of them, note what your department line then says: of those 492 institutions, 141 report the department balancing exactly and 351 report a department surplus. None report a department shortfall. The absorption between those two levels is the thing worth understanding before anyone quotes either number.
What you learn: where in your own submission the cross-subsidy is being booked, which is usually the question the athletics director is asking.
Benchmark on scale and per-participant spending, never on surplus
Build the peer comparison from department expense, expense per participant, and the gender split of both. Those are allocated by each institution under the same instructions, so they compare. The reported surplus does not compare: exact-balance and surplus institutions have almost identical budgets, with medians of $3,747,301 against $4,084,248.
- Weight per-participant figures by participants before averaging. Averaging the ratio directly gives a 12-athlete golf team the same weight as a 120-athlete football team.
- EADA counts participation slots, not people. An athlete on two teams is counted twice, and the AY 2024-25 total of 765,836 slots is not a headcount and is never enrollment.
What you learn: a peer table that survives the first sceptical question, because every column in it is a spending comparison rather than a profitability claim.
Write the provost sentence before the question arrives
The deliverable, not the analysis
The question is almost never "what is the reported revenue". It is "does athletics pay for itself", and the honest answer has three parts: here is what we spend, here is how that compares to peers and to our own past, and here is why the federal file cannot tell you whether it pays for itself. Having those three sentences drafted turns a two-day scramble into a forwarded email.
What you learn: nothing new about the data. It saves you the week you would otherwise spend rediscovering all of this the next time a trustee asks.
What would bound the subsidy question better
The natural follow-up is: fine, so which federal source does separate earned athletics revenue from institutional support? Within the public data available to us, the honest answer is none of them, and it is worth knowing that before you spend a week looking.
All 281 columns across the two EADA fact tables were regex-tested for revenue-source words: ticket, media, donation, gift, contribution, subsidy, fee, support, sponsor, conference, NCAA, endowment, guarantee. Zero matches. There is no ticket-sales, media-rights, donations, institutional-support or student-fee column. EADA records how much revenue was booked to athletics and never where it came from, which is the structural reason it cannot size a subsidy.
IPEDS Finance does not rescue it. Its four finance tables carry 1,346 columns between them and not one names athletics or auxiliary enterprises. IPEDS reports revenue by source for the whole institution, which does not isolate the athletics department, and EADA carries an OPEID rather than a UNITID, so joining the two takes a crosswalk we did not build. Other athletics financial collections exist outside the federal files; we did not query them, so we are not going to quote numbers from them.
That leaves the subsidy question where it belongs: as a limitation you name in the memo rather than a number you estimate around. If your institution wants a subsidy figure, it comes from your own general ledger, not from a federal file.
Method, vintages, and what these numbers do not say
Everything above comes from direct queries against the loaded EADA data, logged so they can be re-run, plus two confirmation queries against IPEDS. The full query set, including the analyses we discarded and why, is in the evidence files listed under Sources.
EADA covers academic years 2002-03 through 2024-25. The year key equals the starting year, so academic_year_id = 2024 is the year labelled 2024-2025, written AY 2024-25 throughout. The unit of analysis for every headline figure is the institution-year on eada.fact_institution_yearly, and the sport-level figures come from eada.fact_sport_yearly, one row per institution per sport.
Four things these numbers do not say:
- No institution in this analysis "made money" or "lost money" on athletics. Every figure describes what an institution reported: revenue above expense, the two equal, or a shortfall. The distinction is the whole point, and dropping it converts a spending file into a profit claim it cannot support.
- Participation figures are slots, not people. EADA counts an athlete on two teams twice, so the AY 2024-25 total of 765,836 (443,453 men, 322,383 women) is a participation count and never an enrollment or headcount figure.
- Operating expense is a narrow line. The $3,969,154,060 of total operating expense in AY 2024-25 is 14.4 percent of the $27,625,630,536 in total expense, because operating expense excludes athletic aid, coaching salaries and recruiting. Do not use it as a proxy for total athletics spending.
- Sport aggregates should not be summed across overlapping sport labels. The sport dimension mixes granularities: Cross Country, Track and Field (Indoor), Track and Field (Outdoor) and Track and Field and Cross Country (combined) all coexist, as do Swimming, Diving and the combined Swimming and Diving. This analysis treats each as a separate row and never adds them together.
Get the caveat along with the number
Clema reads EADA alongside IPEDS and the other federal files, and answers athletics questions in plain English, including the ones where the useful answer is why the file cannot support the claim.
Book a demoSources
Dataset: U.S. Department of Education, Equity in Athletics Disclosure Act data, academic years 2002-03 through 2024-25, with two confirmation queries against the IPEDS finance component. Column definitions come from the loaded data catalog, since the EADA tables carry no database-level column comments.
Reproducible query log: docs/data-evidence/athletics-revenue-is-not-revenue-what-eada-reports/queries.md (42 numbered queries, the discarded analyses and why they were dropped, and a dated corrections log) and docs/data-evidence/athletics-revenue-is-not-revenue-what-eada-reports/schema.md (table grains, column definitions, join paths and eight confirmed gotchas).
More on what Clema does with this file on the EADA data source page. Related reading on this site: why three quarters of programs have no earnings number, which programs colleges are quietly closing, and the federal data release calendar for when the next EADA vintage lands.