From class discussions I’ve had with various students over the years, I gather that only a few of them favor radical income redistribution, such as complete equalization of incomes. Most believe though that it’s fair to tax “the rich” at a higher rate than “the poor” and that the poor should really get some form of income subsidy, because the rich probably got that way owing not just to hard work but perhaps also to socioeconomic advantages (e.g., stable home life, good education). Moreover, they think the effects of progressive taxes are revenue neutral—that is, the total amount of income earned stays pretty much the same.

So in the spirit of experimental economics, which is all the rage in college economics courses these days, I thought I might try a little experiment with one of my classes this spring.

**The Experiment**

There are about 21 students in my intermediate microeconomics class. The average score on their midterm exams was 71 out of 100 possible points. Thus, the total number of points the class earned as a whole was 1,491. The highest score was 101.5 (the exam included a bonus question) and the lowest was 44.

I give letter grades according to the following table:

A: 90–100

B: 80–89

C: 65–79

D: 55–64

F: 54 and below

So a class average of 71 translates to an average letter grade of C. There were 3 As, 5 Bs, 8 Cs, 3Ds, and 2Fs.

As a lesson on the effects of redistribution on incentives, I’m thinking of proposing the following “progressive tax” on grades:

71 to 80 taxed at 20 percent

81 to 90 taxed at 30 percent

91 and above taxed at 40 percent

(These rates reflect, in a simplified way, the current United States tax code.)

For example, my top scorer, the one who scored 101.5 points out of 100, would be taxed 20 percent, or 1.8 points, for the first 9 points above the mean; 30 percent, or 3.3 points, for the next 9 points; and 40 percent, or 4.2 points, for the last 11 points. Thus, her total tax bill would be 9.3 points, leaving her with 101.5 - 9.3 = 92.2 points. My student who failed with only 44 points (with 55 needed to pass) would then receive those points as a subsidy, giving him44 + 9.3 = 53.3 points.

The next-highest scoring student received 94 points on the midterm. His tax would be calculated as follows: 1.8 and 3.3 points as with the highest scorer, plus 1.2 for the 3 points he earned above 91, giving him a total tax of 6.3, leaving him with a score of 87.7 points. These points would be redistributed to the next-lowest scorer, who got 50.5 points. That student then would have a total score of 56.8 points.

This system would continue for every other student in my class, half of whom would be taxed, the other half of whom would receive a subsidy.

#### The Results

With respect to the best and worst performers, the tax/subsidy would not change their letter grades: the former would still get an A, the latter an F. But in many cases, perhaps most (I haven’t actually done the calculations for everyone), there would be a change of grade. For instance, the next-highest performer would go from an A (94 points) to a B (87.85 points) while the next-lowest performer would go from an F (50.5 points) to a passing grade of D (56.55 points).

If any of them complained about the proposal’s fairness, I would point out that innate differences in intelligence, drive, and discipline unfairly create gaps in performance; “the smart” probably went to better high schools and had other socioeconomic advantages that were also driving the unequal outcomes, even though the rules were the same for everyone.

I would then ask my students, if I actually implemented my progressive tax scheme on their midterms, to give a general prediction of (1) the amount of effort they and their classmates would devote to studying and learning the material covered in the second half of the course and (2) the total number of points the class would earn and the mean score on the final exam. Would these figures go up or down?