The American Bar Association is apparently considering increasing the bar exam passage rate required for law school accreditation, as has been reported here and here. The new standard would reportedly require a law school to achieve a 80% passage rate within two years of graduation. Sounds reasonable, right?
Actually a uniform nationwide passage rate requirement would be crazy! As I have written here and here, there is a tremendous difference among the states in the difficulty of their bar exams. Some bars are so easy you could accidentally get admitted to them by walking too close to the testing center. Some bars are so difficult that the former Stanford Law School dean could fail them.
If you don't believe my calculations, just look at the cut rates for MBE scores, which are the minimum passing scores for bar exams on a multistate bar exam scale. The states' passing scores range from 128 to 145, which encompasses a huge swathe of test-takers nationwide. To illustrate, below is a plot of the distribution of MBE scores from the National Conference of Bar Examiner's 2012 Statistics. The two red lines show the range from the easiest state to the most difficult state.That's a range of over 30% of the bar-taking population nationwide who could pass one bar and fail another with the same score.
How could an ABA committee conclude that a uniform passage rate would be appropriate under those circumstances? I suppose if you're from Missouri (to choose a state randomly) the 80% pass rate makes a lot of sense. The Missouri bar requires a MBE score of 130, which is what a person with a LSAT score of less than 140 (!) would expect to score on the MBE, as detailed here. Even more egregiously, in Wisconsin, if you graduate from an in-state school you automatically pass the bar! I could see why if you couldn't achieve a 80% pass rate after two years in such states one might say your law school had a serious problem. In Delaware, however, people actually fail the bar and the same analysis is not at all true.
If the 80% uniform bar exam proposal flows from concerns about consumer protection for prospective law students, the ABA should focus its attention elsewhere-- perhaps on employment data! After all, the bar passage rate is one of the few reliable indicators prospective law students have about a law school (largely because the reporting is out of the law school's control).
On the one hand, it's hard to complain about requiring a 80% pass rate after two years of trying. After all that's as many as five attempts. I think I could pass a pregnancy test if I took it five times! On the other hand, allowing so many attempts favors law schools in states with bars that are unreliably and randomly graded, and the idea of a uniform passage percentage is just oblivious to the data. It would be better to base the passage rate on first-time takers and to modulate it for different states.
I am not a big fan of the bar exam, but for those who are, there is an easy solution to this. The ABA could simply scale the bar passage requirement to the MBE cut scores for the various states. I doubt that will happen, however, because that could expose some rather unpleasant realities in states with easy bars. Let's hope that these early reports have oversimplified what the ABA committee is really considering, which will undoubtedly turn out to be much more sensible...