From Ideas:

Most law students study law in order to practice it. For most of them, practice requires passing the bar. Naturally enough, they want information on how likely they are to pass according to what school they go to. There is currently no way they can get that information. …

To provide applicants the information they want, schools need to publish bar passage rates as a function of LSAT. A simple way of doing so would be to break LSAT scores into groups — 176-180, 171-175, … — and report bar passage rates for each group, perhaps summed over a period of two or three years to provide enough data for a meaningful figure. It might turn out that the elite schools did a worse job for everyone. More plausibly, it might well turn out that the elite schools did a better job for students with high LSATs and a worse job for students with low LSATs—useful information for the latter in deciding where to go.

(Hat Tip: Ben Cunningham.)