Inside the USNWR law school rankings: when a median LSAT of 173 is better than a 174
The latest 509 data disclosed the LSAT medians of the entering 2024 students (Class of 2027) for law schools. Topping the list were Yale and Harvard, which each had a median LSAT of 174. Harvard’s 25th percentile was a 171, and Yale’s a 170. The next tranche of five schools had a median of 173—Chicago, Columbia, Cornell, Stanford and Washington University in St. Louis. Columbia had a 25th percentile of 170; Chicago, Cornell, and Stanford a 169; and Washington University a much lower 163 (for reasons why schools have significant 50th-25th percentile gaps, check out this post). Another band of schools below that has a 172 median.
For USNWR rankings purposes, you might look at this data and realize, well, first, it’s odd to only use the median LSAT rather than considering other bands (e.g., 75th and 25th percentiles) of admitted students. But if medians are used, it might look like Yale and Harvard are tied for first, and the next five schools tied for third.
But, very roughly, here’s what USNWR sees:
1. Chicago 0.9815
1. Washington University 0.9815
3. Harvard 0.9784
4. Columbia 0.9771
5. Cornell 0.9762
6. New York University 0.975
6. Penn 0.975
6. Virginia 0.975
9. Yale 0.9728
10. Stanford 0.9703
*
The first thing that USNWR does is measure the percentile equivalent of the LSAT score. Near the very top of the LSAT scores, there is a very high degree of compression. The increase of a median LSAT score from 173 to 174 is much less valuable than, say, the increase of a median score from 163 to 164. A 174 puts you higher than 98.66% of test-takers; a 173, higher than 98.15%. A small gap (0.51 percentage points), all things considered. Compare that to the 163-164 gap (83.08% to 85.58%, 2.5 points), or 153-154 (49.98%-53.87%, 3.89 points).
But there is now tremendous compression at the top of the law school rankings, so even small differences can mean a lot. And even though USNWR significantly devalued “inputs” (like admissions metrics—LSAT medians are just 5% of the overall rankings now), again, small changes can mean a lot.
In theory, Yale and Harvard should place first with a 0.9866 in the metric. But USNWR adds a caveat:
These are the combined median scores on the LSAT and GRE quantitative, verbal and analytical writing exams of all 2023 full- and part-time entrants to the J.D. program. Reported scores for each of the four exams, when applicable, were converted to 0-100 percentile scales. LSAT and GRE percentile scales were weighted by the proportions of test-takers submitting each exam. For example, if 85% of exams submitted were LSATs and 15% submitted were GREs, the LSAT percentile would be multiplied by 0.85 and the average percentile of the three GRE exams by 0.15 before summing the two values. This means GRE scores were never converted to LSAT scores or vice versa. If a school was not required to report GRE scores to the ABA for the disclosure requirements because a small number of incoming students reported these scores, they were not used in the rankings.
Many schools reports some number of students admitted with a GRE score. But you are only required to disclose the GRE scores of your students if you admit at least 10. Chicago (5) and Washington University (6) admit them in sufficiently small numbers that they do not need to disclose. But Harvard (43) and Yale (24), among others, do.
USNWR uses a nonsensical methodology for GRE scores. As I’ve chronicled here, the Educational Testing Service, which administers the GRE, has its own comparison tool to take GRE scores and offer a predicted LSAT score.
Let’s take Yale for a moment. Its 24 GRE enrollees have a median verbal score of 167 and a median quantitative score of 164. ETS predicts that to be a 172—quite close to the 174 median, and around the 97.5 percentile.
USNWR instead simply takes the percentile equivalents of three GRE sections (not just two), the verbal (for Yale, 167 is 97%); the quant (for Yale, 164, 66%), and the analytical (for Yale, 5.5, 98%). It appears USNWR weighs them equally, for an 87% score—not only much lower than 98.66% for the LSAT, but also lower than the ETS percentile equivalent of 97.5%.
More than 10% of Yale’s enrollees with an LSAT or GRE opted for the GRE (24 GRE, 204 LSAT), so that is factored into the Yale median LSAT score.
You can see the total enrollees, and the enrollees with an LSAT score or a GRE score. It includes the medians for these scores, the average for the GRE scores, and an “LSAT equivalency” score at the end.
How significant is this decision? As you can see, a pure 173 LSAT median is more valuable than an LSAT median of 174 with a non-trivial GRE cohort that includes the average of the GRE percentiles, rather than the ETS measure of GRE-LSAT comparison.
And for Yale, the difference is so significant that, in my projected USNWR rankings, all else being equal, if Yale could simply rely on its 174 median instead of this GRE-LSAT composite metric, it would be tied for second rather than third.