"Should we redistribute grades like we do income?" Response to Megan McArdle

Megan McArdle does not think much of my response to Robin Hanson's implied argument that redistributing income is no more rational or right than redistributing students' GDP would be.  Let's take her rebuttals one by one (leaving out the final one, "hand-waving," which is mainly a repetition). The italics are hers, generally followed by a quote from my original post, shaded blue. My responses are below each indented excerpt from McArdle.

a) Comparing apples to oranges:  "GPA is more tightly tied to individual performance than earnings are. Granting that a) grading is an imperfect measure of the quality of student input, and b) earnings bear some relationship to performance, it's still true that student performance bears a closer relationship to grade than the social utility of the average person's work does to that person's earnings"  So the correct metric by which we assess grades is accuracy (does it correlate with performance) while the correct measure of income is whether it serves some larger social purpose.  Obviously, comparisons should be of like to like: either how accurately income/GPA are tied to performance, or how accurately they are tied to social utility.  It's not clear that GPA would win on either score.

Income is a measure of a performance that's valued by social means. Pay is generally awarded by a collective entity or a community of customers, which gets the ability to pay by its relationship to the larger community; the value pay confers is by definition social. We collectively pay athletes, mainly by watching commercials with which broadcasts of their games are interspersed and buying branded gear; we collectively pay Goldman Sachs investment bankers by buying stuff from the companies that see fit to pay their M&A fees; we collectively pay teachers by allocating tax money for education. We may pay contractors individually, but their value is determined by how large a subset of the community decides to hire them. Money gets its value by being equally valued by others; every "distribution" is a redistribution; and taxes are simply part of the mix.  Grades, in contrast, we at least attempt to tie to only one factor: the student's work.  Money is meant to be redistributed, student performance assessments are not.

b)  Begging the question: Of course, the latter comparison simply assumes what it wants to prove: that there is "social utility"; that this quantifiable concept differs in some way that we can know conclusively differs from income distribution; and that it differs in a way that can most reasonably be corrected by redistributing income (rather than, say, changing bad rules which result in the accumulation of great incomes from activities that generate no or little social utility.)  These are big questions, and no attempt is made to prove them, or even to simply say "I believe that . . . " rather, they are simply asserted as fact.  The author goes on to make similarly sweeping claims about the purpose and right of taxation.

Once again: Money is,by definition, a (very imperfect) measure of social utility. The larger society creates it, and creates/fosters/countenances the institutions that bestow it, creating a legal and regulatory framework in which they exist; the larger society also has the right to tax a portion of it back, particularly if the tax level is determined by democratic means.  If we want to tax-cut ourselves into a society with large cantons of third-world poverty, we can do that; if we want to go to Scandinavian levels of taxation we can do that too.  Yes, to some extent the argument is circular: in a democracy, the level of "redistribution" is a collective determination of what's just.  Agreed that fixing "bad rules" might make for a better, er, redistribution (though if McArdle doesn't believe in "social utility," by what standard are rules that enable rapine bad?), but the same elected officials make those rules as make our tax code.  We can't jump off our shadow; we might argue that many of us are to varying degrees disenfranchised by monied interests that pay to get "bad rules" (and also, of course, pay to minimize redistribution), but our only means of changing the rigged game is to organize democratically to change the larger rules by which "bad rules" and unjust or inefficient tax codes become law. So while we may not be able to "conclusively" quantify "social utility" -- and in we capitalism, agree for the most part not to try -- we are perfectly right to make, and abide by, and accept our right to make, enough of a judgment about social utility to tax varying portions of income to serve the common good.

McArdle complains that I make "no attempt to prove" the right to tax, i.e. redistribute some income. I think that charge boils down to the fact that I accept the society's collective right to make its own rules by democratic means, and she does not.

Overgeneralizing  "As Hanson intimates, developing one's value in the workplace is largely dependent on learning things in school.  You cannot progress adequately through our educational system without having your performance measured."  Even assuming this is true--and the second statement is at least open to question--it is not true of all education; it's true of a fairly limited set of skills in language, math, and science.  Virtually everyone needs basic reading and math in order to earn income.  Virtually no one needs American history or art.

First of all, if no one needs history or art to earn income, then redistributing GPA in those subjects has no point. But the claim that these studies have no effect on income is wrong on several levels. Since we're dealing with grades' implicit financial value, let's limit discussion to the way performance in these studies may directly affect income. First, collective GPA matters for advancement. Second, excelling in either or both subjects has a direct effect on many people's incomes, e.g.,  art teachers, fine artists, commercial artists, graphic designers, anyone involved in designing the visual component of anything made,  history teachers and college professors, the disproportionate number of lawyers and government officials who major in history, journalists, political bloggers -- and indeed, anyone who learned to write and research and critically assess hypotheses in part through studying history.  And presumably, in their class work, they need some assessment of whether they've developed the research and interpretive skills required for the next level of study or for the workplace, and institutions at the next level need the measure so they can determine who is capable of performing at the level they require. Of course the methods of assessment may be terribly imperfect, but we do the best we can. "Sharing" the degree of measurement above the mean earned by a given student is a deliberately absurd proposition. It assumes that the assessment has no meaning.

Post-hoc rationalization of the status quo Moreover, this is not a good description of how the educational system actually works, making this sound like exactly the sort of post-hoc reasoning that Hanson describes in his post.  We put the least focus on GPAs at the age when students are learning the most important skills: basic math and reading.  We put the most emphasis on them in junior and senior year, when it is already too late for students who have not learned basic skills. If the purpose of GPAs were to help students learn skills that would be useful in the workplace, we should see the opposite effect.

I don't see how grades are more or less important at either stage. Or rather, they're "more important" in the lower grades only for those whose performance is problematic. In any case, because grades are treated in Hanson's thought experiment as a quasi-monetary reward I was thinking primarily of those positioned to profit most by continuing to earn high ones.  That reward increases in value as you go up the food chain. A's at Harvard help you get into Stamford electrical engineering programs or Yale Law.

Non sequitur "Income redistribution in the U.S. is very moderate. No one who makes a lot of money has cause to complain that the state is depriving him or her of the fruits of labor.  And by multiple measures of national well-being, humans thrive best in societies that invest in commonwealth, that is, do their utmost to provide the conditions for a decent life to all and to moderate without eliminating disparities of wealth."  Of course, this doesn't tell us whether we should redistribute income, (or grades), only whether we are....[snip]

...the poor quality of the arguments for difference [between redistributing income and redistributing GPA]...suggest that most of us just want to redistribute income because, well, we wanna . . . not because we have any particularly good reason.  Which was Robin's point in the first place.

McArdle has no respect for the argument from collective experience and collective judgment.  I think we "want to redistribute income" because we've learned by collective trial and error that the right measure of redistribution (which we fight over) produces the most good for the most people. McArdle ignores the measures of national well-being I alluded to, which include reported happiness, trust, healthcare, and social mobility. Developed democracies that redistribute more than the U.S. does top out all those charts.  This is not to suggest that the U.S. should aspire to European tax levels, but only to point out that there is empirical grounding for the notion that a certain level of income redistribution (perhaps a different level for each country) promotes the general welfare.