7/07/2008

hey chicago schoolers

if upon graduation from jc, you were to give modular credits to everyone and asked them to bid for university courses, there wouldn't be any surprise if everybody dumped everything onto the course of their choice. after all, you can't take 2 courses at once, and even if allocating points to give you favour for your second choice were allowed, it probably would not be game-theoretically efficient to put anything for your second choice. so clearly, you can't have an allocation market with everyone having equal modular credits. in these scenarios, academic course are rationed based on grades (which are your "hard work" or incentive points, and also serve an additional benefit of signalling your academic aptitude and willingness to work). in addition there are interviews etc to generally determine your level of passion. it is, in general, what people see as a fair way to allocate, or at least the fairest of all unfair alternatives.

now i'm pretty interested in the modular credit market. the knowledge that some people are willing to put all their points into one course signals either (a. poor allocation of resources, or b. that there are certain bottleneck courses which people feel they cannot do without, e.g. clinical psych) b. implies that having these bottleneck courses is clearly much more highly valued than any possible combination of courses, such that given the limited MCs you have, there is no combination of courses you could have which would dominate you taking that one course. why might this be the case?

the reason MCs don't work as well as you would like them to do is because at the end of the day, it is not a market integrated into the real world. Let me give you an example. You are playing Second Life, or some MMORPG. The god of this virtual world decides to dispense 100 second life credits to everyone, with which anyone can use to bid for certain virtual goods, like virtual pie, virtual hamburger. the price of virtual goods will be determined by the total money supply and the relative tastes of the virtual citizens. and let us not kid ourselves, if red is a nice colour online, then red virtual clothes would probably cost more. so the virtual money system works well in allocating virtual resources (assuming of course server space is limited and that is why we have to ration item tags).

now a parallel market exists in which there is a demand for virtual pie in the real world, such that one is able to find people willing to give you 10 REAL dollars for a virtual pie. But for some reason, no one wants any of the other virtual goods. Then I would put all my virtual credits into virtual pie of course! Unless my taste for having virtual hamburger is 10 real dollars. The key is the mismatch between the free modular credits and the very tangible and very large real value of taking a bottleneck course. You see if one had to work for virtual money, then eventually the labour time involved on second life to earn the money to get (maybe you have to do 10 quests say, taking 1 hour), would equalize with the 10 real dollars. So you've traded your online labour for 10 bucks. But what the fuck do you trade when you have modular credits. there is no price you could make which would clear the market, because the real value of MCs is close to zero (maybe not if you allow borrowing from other years, or the other more insidious one, trading MCs for money/sex). If you gave everyone 10,000 MCs it wouldn't matter. it's like printing money.

So perhaps you could consider identifying bottleneck courses and selecting them the same way you would students for medicine. through some objective criteria you have to work for (and thus have some labour cost). but this would perpetuate inequality in education (people who by luck do better in one year will have everything going for them). in a modular system though i would say one could try again and take the module in another year.

expand the bottleneck course. probably the best idea. this cost resources, so it's like giving student's a free lunch, then you could charge. but this is also perceived as unfair. but you needn't have had an MC system to determine this. if the course is oversubscribed, fill it. but maybe the way to go is really to get rid of all the crap courses nobody wants and fire those professors. but what if someone wants to take latin next time?

of course, this is based on my limited understanding of MCs. I need to go read up more.

2 comments:

Anonymous said...

Your method of allocation is quite elitist. If we were to go through tests and interviews, I think it will be the same bunch of students that will get all the modules of their choice.

By allocating a fixed number of MCs, everyone will more or less have a fair chance at learning. (Personally, I think a chance at learning should be given to everyone and not made exclusive to a group of people.)

Actually, it matters. If everyone stick strictly to their budget, one will be force to rank the utilities of taking a popular module versus securing enough modules to graduate. Also, there is a choice between securing a less desirable time slot for a popular module against a more desirable time slot for the same module. In fact, it is like tests and intereviews as the MC you bid represents your desire for the module with one major advantage- everyone is equal.

That is provided that everyone sticks to their budget and are good economists who understands how to allocate resources according to utilities and across time even in the face of uncertainty. (There is no guarantee that you will secure the module or the module may be offer the next semester etc.) Of course, we also need good arbitragers to make the market "efficient".

Modules that are not popular may not be irrelevant. Maybe, no one know that they are interested in Latin. If they are "forced" to take it, they might develop an interest in it or even become a great Latin scholar. These situations are hopefully few and pleasant. The fixed MC will also ensure that some of these courses are preserved.

MC bidding is an art in itself and it makes a good economic study.

Just ask any NUS student.

Jesse said...

that is a fair enough comment. when i first discussed this issue that was the very first issue that someone rebutted me with. that education should perhaps be free of the inequalities that we see in the rest of life.

for example, in france, entry to universities or courses except for the "grandes ecoles" is not really dependent on your results in the baccalaurate. of course, this merely pushes the selection problem through the university system, where passing through to the next year is extremely selective, such that completing university in 3 years in rare. selection must kick in at some point, though you could say it's probably fairer (and a lot more fun) to give people more time to prepare.

yes, i was talking today to some people and it really doesn't seem as inequitable, and that they had many MCs left over. then again, most of them did not take too popular modules. And the appeal system and all that.

And i didn't know the choice extended to timeslots too, which would something LSE would do much to learn from. At least here you can spend MCs to clump courses together then to use the time productively to work part-time etc.

So, good to learn. I just enjoy poking at imperfections in systems.