Running somewhat in parallel to this post from Megan McArdle, Frances Woolley contemplates an educational system with one-for-one class transferability (“credit recognition”):
- Butterfly wings and meatgrinders (Worthwhile Canadian Initiative)
But universities aren’t entirely frank about the way that post-secondary education produces value. We say it’s about knowledge, human capital and all that. In fact, post-secondary has value in part because it acts as a signal: I took ECON 1000 and survived. It’s like climbing Mount Everest – it’s an impressive accomplishment solely because lots of people try to do it and fail.
“Meat-grinder courses” – courses like intermediate microeconomics that chew students up and spit them out – must exist if a bachelor’s degree is to have any value as a signal of ability.
Yet credit recognition would – if it was to serve any useful purpose – have to allow students to take meat-grinder courses at other universities.
But a university could make good money by delivering meat-grinder courses on a long-distance basis, offering a kinder, gentler, more nurturing version of, say, ECON 1000. For a credit recognition system to be effective, there would have to be some way of preventing grade inflation, and a competitive lowering of standards.
I’m not sure how much this differs between disciplines, but it seems to me that even meat-grinder prereqs are in fact prerequisite for a reason, and this gives us a reasonably pragmatic-seeming way to prevent that “competitive lowering of standards”. Every time regulations come up in Formula One*, I rant about how car performance should be controlled by behaviour rather than dimensional specs — if you want to limit the amount of downforce a car generates, then test its -L/D ratio and make sure it’s below a certain constant under test conditions rather than fuck around with wing dimensions. Similarly, if you want to make sure that Ed’s Interweb Collij isn’t selling a watered-down version of ECON 1000, track the students and see how many of them fail subsequent courses. If you expect no more than 10% of students who’ve passed your august institution’s terrifyingly carnassial ECON 1000 course to fail out of ECON 2100, then set an upper bound of, say, 15% or 20% and see how many of Ed’s students fail. If Ed’s graduates fail at higher rates, then Ed’s doing something wrong and doesn’t get certified. This won’t work for graduate seminars with single-digit enrolment, of course, but filter courses should have a large enough population to avoid statistical outliers.
And if it turns out that even students who’ve taken kinder, gentler meat-grinder courses excel in subsequent dependent courses, well, the higher-ed bubble needs to pop sooner or later.
* I’ll try to post some thoughts about the 2012 cars before the season starts