Three Numbers
Every spring, applicants play the same game. They line up their admit letters, sort them by rank, and figure the higher rank is the safer bet. That instinct is mostly right. Once in a while it’s a disaster, and this chapter is about 1 of those times.
I’m going to use 2 made-up schools, built from real patterns. I’m calling them Greystone Law and Westfield Law. The numbers are the kind real schools have filed, real deans have approved, and real students have bet their careers on. They’re here to show how the machine works without real names getting in the way.
Greystone Law. Private. Ranked in the low 60s. Class of about 170. 3 years ago they opened a new $38 million building, paid for with revenue bonds (loans tied to tuition income). The dean gave a speech about momentum at alumni weekend, and there’s a donor’s name on the glass. Applications are up. The incoming class has the highest median LSAT in a decade. On paper, a winner.
Westfield Law. Private. Ranked in the mid-70s. Class of about 200. Old building, nothing fancy. Same dean for 12 years, and she used to run the budget. Walk the halls and you’d think this school isn’t investing in itself.
On paper, Greystone looks better. But 3 numbers separate them, and they point the other way.
Estimating the discount rate from a 509
You need the 509 and 4 minutes, no spreadsheet. The 509 is a public report every law school files each year with the ABA. It shows grants, employment, and bar passage data.
Here’s a worked example, using a made-up composite school.
Sticker tuition is $55,000. The 509 shows 80% of students get grants, and the median grant is $30,000. Those 2 numbers are all you need to estimate what the school actually collects per seat.
The 20% who get no grant pay close to full sticker, so about $55,000. The other 80% pay roughly $55,000 minus the $30,000 median grant, so around $25,000. Weighted average: (0.20 x $55,000) + (0.80 x $25,000) = $11,000 + $20,000 = $31,000 per student per year.1
So the school collects $31,000 per seat on a $55,000 sticker. Its discount rate is about 44%. For every dollar of sticker, it keeps 56 cents.
Now pull last year’s 509. Same school, 12 months earlier: 80% of students got grants, but the median grant was $25,000, not $30,000. Run the same subtraction: the grant-getters paid about $55,000 minus $25,000, so around $30,000 each. Average net per seat: (0.20 x $55,000) + (0.80 x $30,000) = $11,000 + $24,000 = $35,000. Back then the school kept 64 cents on the dollar.
The sticker price didn’t change, so to a casual observer the school looks the same. But the discount moved 8 points in 1 year, from 36% to 44%. On a class of 180, that’s an extra $720,000 in scholarship aid just to fill the same seats. Across all 3 enrolled classes, that’s about $2.16 million a year the school no longer has, while its costs still expect that money to be there.2
The trend is the signal, not the sticker price or the ranking. A school that holds enrollment flat by giving bigger discounts is in a quiet price war with itself. The class stays full while the margin gets thinner, and the 509 lets you see it before the school says a word, because the law makes them file the grant data every year.
The 509 gives you 1 more thing. Compare the school’s median grant to its own sticker over several years. If the median grant is growing faster than sticker, the net per seat is falling even when enrollment holds. A school raising sticker 3% a year while raising median grants 8% a year is moving the wrong way, and it’s doing it in public, in a document that took me about 4 minutes to read.3
Number 1: Debt
A revenue bond is a loan, and the contract comes with rules. The most common is a covenant, a binding promise: tuition revenue has to stay above a set floor every year. Drop below it and the lender can speed up the debt, reset the terms, or trigger a default. Greystone owed $38 million on its building, and the covenant required tuition revenue to clear a specific number. As long as classes came in on schedule, fine.
Westfield’s dean killed a $30 million building plan about 4 years before this story starts. In the board meeting she asked 1 question: “What happens to the loan payments if we lose 30 students next year?” Nobody had a clean answer, so she said no. For about 18 months she was the least popular person on campus, and then she looked like the smartest.
Greystone owed $38 million. Westfield owed nothing on its buildings.
Number 2: The discount rate
Almost no law student outside the top 20 pays sticker. Schools discount hard to fill seats, handing out merit scholarships to applicants they want but can’t land without a better offer. The gap between sticker and the average actually collected per student is the discount rate. A school running a 45% discount is giving away nearly half its sticker revenue to fill chairs.
When the market was good, both schools ran similar numbers, Greystone around 40%, Westfield around 42%, close enough that it didn’t matter much.
Then the crisis hit. Lending caps shrank the applicant pool, outcome data spread everywhere, and the demographic cliff hit earlier than anyone expected. A smaller pool meant schools fought harder for fewer students.
Greystone’s class fell from 170 to 145, and to land even those 145 in a tighter market, they had to sweeten every offer. The discount jumped from 40% to 55%. A smaller class plus lower revenue per student dropped Greystone’s collections well below its covenant floor, and the default clock started.
Westfield’s dean had spent a decade targeting the 158-164 LSAT band. She skipped the 170s, who were headed to schools ranked 40 spots higher anyway, and the borderline admits who’d push the discount past the breaking point. She recruited the middle, students she could land without wrecking the budget. When the market tightened, Westfield’s discount moved from 42% to 48%, a bump rather than a collapse.
Number 3: The full-pay percentage
At nearly every private school outside the top 20, about 25-30% of each class pays close to full sticker.4 Schools don’t advertise it. These students have the scores and GPAs to get merit money at lower-ranked schools. They picked a higher-ranked school instead, and they pay for it. Their full-price tuition funds the discounts the school hands the students it’s competing for. Nobody says that at admitted-students day.
When the federal lending caps took effect, these students were the first the model lost. A student paying close to sticker at a school ranked 65 carries a huge private debt load, because federal aid only covers part of the gap. And private lenders, unlike the federal government, actually think about whether the degree pays off. At schools with weak employment numbers, they started saying no, tightening terms, or pricing loans high enough to make the bet feel stupid. The full-pay group didn’t vanish overnight, but it shrank fast at schools that couldn’t show clean outcome numbers.
Greystone lost close to half its full-pay students over 2 years, a structural collapse in revenue. The students who stayed needed bigger scholarships, the budget shrank, and the doom loop from Chapter 2 started running.
Westfield’s dean had rebuilt the scholarship grid 3 years before the crisis. She cut the full-ride packages at the top, since full rides for the highest scorers were expensive and those students usually had better options anyway, and moved that money into partial scholarships for the 158-164 band. The shift made the school a little less attractive to the very top of the pool and much more attractive to the middle. When the full-pay squeeze hit, Westfield’s full-pay group shrank about 15% instead of 50%.
The student who saw it
Maya was deciding between 2 schools. The first was ranked a bit higher, had the newer building, and was offering about $30,000 a year in scholarship. The other was Westfield: ranked lower, older building, a little less scholarship.
She asked 2 questions.
Full-time legal employment, meaning bar-required or JD-advantage work, 10 months after graduation. The higher-ranked school said 78%. Westfield said 91%.5
First-attempt bar passage. The higher-ranked school: 82%. Westfield: 93%.6
You don’t expect the cheaper school to have better outcomes. Maya did the math and picked Westfield.
The higher-ranked school crossed its covenant floor in 2027. It cut clinics, froze hiring, and killed the AI law practice clinic, the same kind Maya had taken at Westfield. It closed in 2029, and Westfield climbed 9 spots the next cycle as competing schools shut down. The students at the school that closed spent their last year inside an institution visibly dying around them, chasing bar-exam results from a school that would be gone before they sat.
Where to find the 3 numbers
Look, none of this needs insider access. The data is public. Most applicants just never look.
Debt. Universities file audited financial statements. Look for long-term debt and bond covenant language. Revenue bonds, debt-service coverage requirements, and maintained-tuition covenants are all in there. Finding it takes about 20 minutes. Nobody at the reception will show it to you.
The discount rate. The ABA 509 shows the percentage of students on grants, the median grant, and how grant aid breaks down by size.7 From those numbers you can estimate how much of its sticker price the school actually collects. A school with 70% of students on grants and a median grant equal to half of sticker is running a very aggressive discount.
Employment outcomes and bar passage. Also in the 509. Look for full-time, long-term, bar-required employment at 10 months out, and first-time bar passage in the school’s home state. A school in the 75-80% range on either number isn’t a crisis, but you should ask why it’s 15 points below Westfield.
The questions to ask on admitted-students day
The school is marketing to you. Fine. You’re doing due diligence, and for a 3-year, 6-figure commitment, none of these questions is rude.
What percentage of your operating budget comes from tuition? A school over 90% has almost no cushion. Most law schools lean heavily on tuition, but even a few points of other income matter a lot when enrollment dips.
Has the school taken on any new long-term debt in the last 10 years? If yes, what are the covenant terms?
What happens to merit scholarships if enrollment drops? Specifically, does a scholarship get cut if the school’s overall discount rate crosses a threshold, or if the entering class shrinks? Some schools bury language like this in the fine print.
What is your employment outcome for the most recent graduating class, and how does it compare to 5 years ago?
Sometimes you’ll get a non-answer, and a non-answer is data.
The real lesson
Greystone looked stronger on the metric most applicants use, ranking. The building, the dean’s speech, and the scholarship offer all confirmed the story. The 3 numbers, debt, discount rate, and full-pay percentage, told the opposite story 4 years before the closing announcement. And the outcome data, employment and bar passage, told you which school was actually graduating lawyers.
Most applicants pick schools the way people pick hotels: look at the star rating and assume the price reflects quality. For hotels, that works fine, but a law school bet follows you into your career, or doesn’t, depending on how it lands.
Notes
- This calculation uses the median grant as the representative discount, which slightly understates the distribution spread but gives a fast, useful floor estimate. The 509 also breaks out grant amounts by quartile (less than 1/4 tuition, 1/4 to 1/2, 1/2 to full), so a more precise estimate is possible; the median-based shortcut is a practical starting point. back to text
- Illustrative composite. The $720,000 figure assumes exactly 180 first-year students at the $4,000 per-seat discount increase ($30k median vs $25k median). Actual classes vary in size and grant distribution. The mechanism is the point; the number is illustrative. back to text
- The ABA 509 disclosure is filed annually and publicly available at the ABA required-disclosures site (abarequireddisclosures.org). The school example here is an illustrative composite. back to text
- The 25-30% full-pay figure is an illustrative composite range consistent with ABA 509 disclosure patterns at mid-ranked private schools. back to text
- Employment figures here are illustrative composites; real per-school figures are in each ABA 509 report. back to text
- Bar passage figures here are illustrative composites; real per-school figures are in each ABA 509 report and from state bar examiners. back to text
- ABA 509 disclosures are filed annually and publicly available at the ABA’s website. The disclosure includes grant aid distribution, employment outcomes, and bar passage data, as of mid-2026, at the ABA required-disclosures site (abarequireddisclosures.org). back to text
