This week, many of my former students will be undergoing the painful experience of taking the Virginia bar exam. My general view on bar exams is that they should be abolished, or at least that you should not be required to pass one in order to practice law. If passing the exam really is an indication of superior or at least adequate legal skills, then clients will choose to hire lawyers who have passed the exam even if passage isn't required to be a member of the bar. Even if a mandatory bar exam really is necessary, it certainly should not be administered by state bar associations, which have an obvious interest in reducing the number of people who are allowed to join the profession, so as to minimize competition for their existing members.
Somin also issues a challenge to those administering the bar: take it yourself, every year, or lose your positions. If the knowledge required on the bar is really necessary for practicing lawyers, then everybody should have to demonstrate proficiency and not just the new entrants. Via Jeff Ely at Cheap Talk, who chews on the possibility:
My bottom line is that banning the bar increases welfare but perhaps for different reasons than Somin has in mind. Routine services will become more competitive and this is good. Increased concentration at the high end is probably also good because market power means less output and for the kinds of lawyering they do, reduced output is welfare-improving.
This time next year I will be spending massive amounts of time preparing for comprehensive exams. I'll be reviewing my notes, reviewing other peoples' notes, memorizing citations, re-reading "foundational" papers, trying to fill in gaps in my knowledge of the literature, and generally freaking out over this massive exam that will determine whether I have just wasted several years of my life. All of this crammed knowledge will (hopefully) be forgotten the second the exam is over as I embark on a several-months-long drinking binge. Whatever remains will surely dissipate within months after. All for what? So I can spew several thousand words in a short amount of time displaying general and specific knowledge of my sub-discipline, including a lot of knowledge with no bearing on my research interests. Even worse: I have to do this twice in two different sub-disciplines within a 6 month period.
But all of that effort comes at a price: namely, my productivity during several major chunks of time during my third year in graduate school. I could spend that time working on conference papers, or refining my M.A. thesis in hopes of publication, or preparing a dissertation proposal, or collecting data, or doing any number of other things that could have a positive impact on my intellectual development and career prospects. Like blogging.
Instead I'm forced to spend my time on a largely fruitless pursuit. Twice. I have had tenured professors acknowledge to me that comprehensive exams are essentially worthless, but since they had to run through the gauntlet they're going to make damn sure that everybody else has to do it as well. Fair enough, I suppose, but surely there are Pareto-improving coordination possibilities that could make everyone better off? For example, instead of spending all that time preparing for and taking my comps, I could spend several months in my third year as an unpaid research assistant for a professor. I could gain hands-on experience in how to conduct research in a refined way and (possibly) an acknowledgement in a publication or even co-authorship credit to enhance my CV, the professor gets help for a project without sacrificing grant money, and both of us get experience working with the other, which could come in handy when determining who should be included on a dissertation committee. Everybody wins! Even the department looks better, if only slightly, since there is more research output than before.
And there are first-mover advantages for departments who adopt this new policy earliest: their grad students will have better CVs and more research experience (and so will perform better on the job market ceteris paribus), and their faculty will have more forth-coming publications. Departmental prestige does not come from grad students passing comps; it comes from published research and job placement. This system could improve both, if only marginally.
What is lost? The stamp of approval that comes from passing a comprehensive exam. But everyone seems to acknowledge that mastery of a subject is better determined by semester-long evaluations in substantive courses than in typo-ridden, caffeine-fueled mental diarrhea spewed out on one exam day. If a student does well in a course, as judged by semester-long interactions with the professor as well as written and oral assignments, then they know the material. If a student under-performs in a substantive course, then the course evaluator can require that (s)he make up the difference by writing a literature review or research design before receiving a full passing grade for the course (which is basically what happens anyway).
How would this not be a better system?