This is a picture of the statistical data we have provided to the ABA for our initial application and our limited site visit:
In case you cannot see the ruler clearly, this is 9½ inches of data. To a certain extent, if you measure the quantity of data in the inches of paper necessary to report it, you know that it is massive.
Effectively, to generate this data, one full-time, mid-level administrative employee-equivalent has been needed in each of the last two years (and will be needed for the foreseeable future). In fact, this probably underestimates the amount of time needed to generate the data as the rest of us, from the Dean on down, have had to spend weeks worth of time on the project. You have to look up from the Dean’s Office, too, to capture the efforts of the various university-level administrators who were involved. If you put all of this time together, this has been the functional equivalent of dropping our enrollment by three or four students each year. Please understand that this is not a cost to run an accredited institution, it is merely the cost of trying to establish that you are running one.
To focus on just one statistic, consider the way the ABA measures the bar passage rate of each law school. The goal is to determine exactly how many students both took and past the bar. Consequently, each school has to spend a lot of time and effort to track down how each student did. Of course, this is not a local matter as students take bar examinations throughout the country. Some jurisdictions make obtaining this data an easy process, but many do not. The school must follow up with the other jurisdictions or, worse, with the individual students. Of course a failure to track sufficient students has draconian effects as it becomes impossible to establish compliance with the standards.
It does not have to be this way. For those of us who do empirical research about the law, we commonly build statistical models for what we are studying. My current research effort, for example, involves patent quality. For the vast majority of patents, there is no direct measurement of quality. Consequently, I will be looking at a variety of other factors that I hope to establish can successfully predicts quality. Bar passage can be handled the same way.
While it might be nice to have a precise measurement of every student’s performance on the bar examination, we may be able to do as well by just measuring the domestic bar exam. You would think that if a school’s students do well on its home-state’s bar, they would do well elsewhere, too. If, for example, law schools that have an acceptable overall bar passage rate (however the ABA wishes to define this) consistently show a passage rate of X or more on the domestic bar, why not dramatically lower the cost of accreditation compliance by only requiring X to be reported?
To be fair, I do not know that there is such a relationship, but this is, at a minimum, an easily testable hypothesis. If confirmed, it could dramatically lower the cost of establishing compliance as could other similar statistical measurements.