In an effort to spice up the otherwise dreadfully boring section in my micro class on different industry types (price takers vs. searchers, oligopoly, etc.), I always have an "experiment" to demonstrate the difficulties of cartel formation. These aren't controlled experiments so I don't bother keeping accurate figures, but I can share some highlights.First, let me explain the setup. I tell everyone to rip off a small piece of paper and put his/her name on the top. Then on the bottom they will write a C or a D. After they will fold the paper so no one else can see the writing, I will collect, then I'll go to the front and open them up one by one, being careful to shield the name and only showing the letter. If everyone writes C, then the class gets some points on the next test. If one person writes a D, that person gets a large number of points on the next test, while everyone else gets 0. If two people write D, they each get a decent number of points, if three people they each get a few points, and if four or more write D then no one gets any extra points.
The point of course is to make it weakly dominant to write D; you always get more points (or the same) if you write D rather than C. But naturally if at least four people realize this, no one gets anything. So this is to represent cheating on a cartel agreement, where I represent the consumers.
After they understand the rules, I leave the classroom for 5 mins or so and let them "form a cartel." Then I come back and tell them to write their letters, then we see the results.
I've tried this 5 or 6 times in the past, and I never had problems getting people to defect; i.e. a handful of students always did the "rational" thing. But back then I had been using money as the prize; e.g. if everyone writes C the class splits up $40, but if one person writes D he alone gets $20, etc. As I say, there was no problem at all "getting away" with running such an experiment and not paying out money.
However, last year I tried using points on the next test, and I found a LOT more cooperation. I believe that 2 out of 4 times students have all put C when points are at stake. Now after this happened, people would explain that "football players threatened to kill us if we put D" etc., but nonetheless it is interesting that these extralegal incentives were NOT added when money was at stake.
So one thing I'm not sure of is: (1) Do students care more about the grades than the money, so they figure out how to alter the incentives? Or is it more (2) students feel a lot guiltier cheating their fellows out of points, but not money?
Last, let me mention what this semester's class did: I knew SOMEthing was up because no one was even nervous as I went through the pile; it seemed as if everyone knew full well there were all Cs on the slips of paper. (In contrast, in past semesters even when all cooperated, people were on the edge of their seats.) Someone explained that when I was out of the room, they had all swapped slips, so that when I came back people were writing the letter on someone ELSE's paper. I certainly have no problem giving them all 2 points on the next test...
