An economics professor at Texas Tech said he had never failed a single student before but had, once, failed an entire class.
The class had insisted that socialism worked and that no one would be poor and no one would be rich, a great equalizer. The professor then said, "OK, we will have an experiment in this class on socialism."
All grades would be averaged and everyone would receive the same grade so no one would fail and no one would receive an A. After the first test the grades were averaged and everyone got a B. The students who studied hard were upset and the students who studied little were happy.
But, as the second test rolled around, the students who studied little had studied even less and the ones who studied hard decided that since they could not make an A, they studied less. The second Test average was a D! No one was happy.
When the 3rd test rolled around the average was an F.
Students remember very little from specific classes. Even though this experiment is obviously going to generate horrible incentives, the lesson learned probably would be more long lasting and profound than any mastery of problem sets with indifference curves and linear programming.