De Gustibus non est Taxandum

Bryan Caplan quotes a passage from  Daniel Kahneman's Thinking, Fast and Slow (which I have not read, but plan to):

A large-scale study of the impact of higher education... revealed striking evidence of the lifelong effects of the goals that young people set for themselves. The relevant data were drawn from questionnaires collected in 1995-1997 from approximately 12,000 people who had started their higher education in elite schools in 1976. When they were 17 or 18, the participants had filled out a questionnaire in which they rated the goal of "being very well-off financially" on a 4-point scale ranging from "not important" to "essential."...

Goals make a large difference. Nineteen years after they stated their financial aspirations, many of the people who wanted a high income had achieved it. Among the 597 physicians and other medical professionals in the sample, for example, each additional point on the money-importance scale was associated with an increment of over $14,000 of job income in 1995 dollars!

In other words, one reason that people differ in their incomes is that some people care more about having a high income than others. To put it in geekspeak, preferences over pecuniary goods (say, consumption) and nonpecuniary goods (say, leisure) are heterogeneous. Bryan goes on to suggest that to the extent this is true, it weakens the case for income redistribution.

He is absolutely right.  Most of the literature on optimal taxation and redistribution, following Mirrlees, assumes homogeneous preferences.  But Matthew Weinzierl has a recent paper on preference heterogeneity, which shows " to the extent that variation in income is due to preference differences rather than productivity differences, the optimal extent of redistribution is lower, and the neglect of preference heterogeneity biases the results of conventional optimal tax analyses in favor of redistribution of income."

By the way, the title of this post is taken from the Weinzierl paper.