The Broken Yardstick
The LSAT did more than test applicants. For the lower half of American legal education, it set the price of everything, and almost nobody admitted it.
Here’s how it worked. Law schools compete for applicants with high LSAT scores. US News uses the median LSAT score of a school’s incoming class to build its rankings. A school’s rank affects who it can recruit, which affects tuition money, which affects whether the school stays open. So every school has a money reason to push its class median as high as it can.
The tool is scholarships. With a few exceptions, these are merit scholarships tied to test scores, not need-based aid. A 175 gets a full ride somewhere, because to the rankings math that 175 is worth more than the tuition the student would have paid. A 155 gets a partial discount, which still helps a school fighting in a tight score range. A 148 pays close to full price, because that score pulls the median down and the school can’t afford to give it away.
Paul Campos called this the Reverse Robin Hood Principle.1 The students who can least afford full price pay closest to it. On average, lower-income students score lower, so they fund the scholarships that buy high scorers, who skew wealthier. And everyone calls it meritocracy.
The whole system ran on a common yardstick. Every applicant has a score, and every school knows what each score is worth to its median, so every school can price scholarships with some discipline. A 170 is worth so many dollars at this school, a 162 less. The school makes an offer, the applicant decides, and the market settles.
Then the ABA pulled the yardstick. In November 2024, the ABA Council voted to let schools admit up to 100 percent of a class with no standardized test.2 The old rule, Interpretation 503-3, had capped no-test admits at 10 percent. That cap is gone. By 2025, 14 schools had taken variances, and about 60 more had applied for JD-Next variances, a related path that uses a different measure.3 The variances run 3 to 5 years, and the ABA says it’s tracking 1L GPA, dropout rates, and bar passage at those schools.
Good. It should. The pricing system is already starting to break.
Think about the scholarship math when an applicant has no LSAT score. The school has a GPA, maybe a personal statement, letters of recommendation. But as I’ll cover later in this book, AI-written text is already flooding those letters and making them almost impossible to tell apart. Without a test score, the school has to price a scholarship for someone it can’t measure on the scale every other school uses.
So it throws more money at the uncertainty. That’s already happening at test-optional schools. Without a common yardstick, schools fall back on the one lever that always works, bigger offers. Applicants walk into financial-aid offices with competing letters from schools that admit they can’t read each other’s applicants. Scholarships get bigger. Sticker prices stay high but mean less, since almost nobody pays what the school actually charges. The share of tuition given away as discounts, the discount rate, climbs.
Steven Harper found that across 189 schools, the median link between LSAT score and first-year grades was 0.36.4 Real but modest. That doesn’t prove the LSAT is a great predictor of legal ability. But it’s the best shared signal the market has, and the whole pricing system was built on it. Pulling it away replaces the signal with noise.
Now look at it from a first-generation applicant’s seat. They don’t have alumni contacts, can’t read a school’s finances, and no parent went to law school. But the LSAT was something they could study for, improve on, and use to pull money from schools that needed their score. The path was clear. Test-optional admissions helps a different applicant, the one with a great story, strong activities, and the polish to present well on paper. That polish tends to come with advantage.
The downstream problem lands 3 years after the enrollment decision.
Schools are tracking bar passage among test-optional admits, which tells you the ABA already expects it could drop. The LSAT is linked to bar passage. The link isn’t perfect, but it shows up in accreditation standards, in bar examiners’ studies, and in the risk warnings schools must give students about their odds of passing.5 A school that stops using the LSAT is betting that something else in its process does the same predictive work. Maybe it does. But you won’t know until the bar results post, and by then the students who didn’t pass have already borrowed the money, spent 3 years in school, and taken the hit to their careers.
The ABA tracking window is 3 to 5 years, so the first real bar passage data for test-optional classes shows up in the late 2020s. If those numbers are bad, schools with variances face a hard choice: bring testing back, watch bar passage stats wreck their reputation, or spend more on bar-prep programs with money and time neither the school nor its students have. If the numbers are fine, more schools go test-optional. Either way, the pricing system stays broken in the meantime, because nobody waits 5 years for data before setting a scholarship budget.
Tamanaha showed the pricing effect in action back in 2012. The University of Illinois gave its whole entering class scholarships, and its median LSAT dropped 4 points in a year, from 167 to 163.6 One school made one decision and took a measurable hit, because it broke the usual score-for-scholarship trade. Now picture that at scale, with 14 schools running without the yardstick, competing for the same applicants against schools still using it.
The schools still using the LSAT know what their applicants are worth. The schools that dropped it don’t, not on the same scale. So the test-optional schools bid up, the LSAT-using schools raise their offers to compete, everyone’s discount rate rises, and everyone collects less tuition per seat than planned. The doom loop turns another notch.
Of course, the schools don’t advertise the quiet part. The ones most likely to go test-optional aren’t the elite schools. Harvard and Yale aren’t dropping the LSAT; their medians already sit near the top of the 120-to-180 scale, packed so tightly the yardstick barely changes their pricing anyway. The schools going test-optional sit in the second and third tier, the exact schools already squeezed by the loan caps, running high discount rates, and watching their incoming medians soften.
These schools are breaking the yardstick to buy short-term enrollment relief. It lets them admit applicants they couldn’t reach in a test-required world, fill seats, and hit revenue targets. Understandable. It’s also the kind of short-term move that speeds up the long-term problem: it raises the discount rate without raising the quality of the class, and it creates bar-passage risk that will show up on their 509 filings, the yearly reports the ABA makes each school publish, in 3 years.
The sticker price is already fiction at most of these schools. The average student pays well under it, and test-optional makes the fiction worse. The school can’t tell what a student is worth to its market position, so it can’t price the discount with any care. More money gets thrown at uncertain bets, and the margin on every seat gets thinner.
One more point. The broken yardstick matters for applicants too, not just schools. The score was a signal they could use in their favor, a portable credential that let them play schools against each other in the open, with proof. School A offered me this. School B offered me that. Here’s my score. That worked because everyone was measuring the same thing.
Without the yardstick, applicants at test-optional schools bargain in a market where the currency is vague. They can’t trade on a score, because they don’t have one, or the school doesn’t use it the same way, or the school they’re comparing against measures something else. The edge a strong LSAT score gave a prepared applicant gets murkier.
Hold onto that. Part Two of this book is about what happens to the yardstick itself, to the score and the market around it, as these forces tighten. In a market where everything else has gotten noisy, the LSAT might be the last number left standing.
Notes
- Paul Campos, Don’t Go to Law School (Unless) (2012), ch. 7. back to text
- ABA Council of the Section of Legal Education, variance to Standard 503 (November 2024). Source: Inside Higher Ed (Nov. 15, 2024); 2Civility; ABA Journal. back to text
- 14 schools granted test-optional variances in 2025; approximately 60 more applied for JD-Next variances. Variances run 3-5 years; ABA tracking 1L GPA, attrition, and bar passage. Source: LegalNewsFeed (Aug. 2025); ABA Journal. back to text
- Steven J. Harper, The Lawyer Bubble (Basic Books, 2013), ch. 2: LSAT-to-1L-grade correlation across 189 schools, median r = 0.36, from 2010 validity studies. back to text
- ABA Standard 316 governs bar passage and requires that at least 75 percent of a school’s graduates who sit for a bar exam pass within 2 years, or the school faces accreditation consequences. Source: ABA Standard 316. back to text
- Brian Tamanaha, Failing Law Schools (University of Chicago Press, 2012), ch. 13: every student in the University of Illinois’s 2011 entering class, including every student admitted off the waitlist, received a scholarship; the class’s median LSAT fell 4 points, from 167 to 163. back to text