~Why Do We Redistribute so Much but Tag so Little? Normative Diversity, Equal Sacrifice and Optimal Taxation

Tagging is a free lunch in conventional optimal tax theory because it eases the classic tradeoff between efficiency and equality. But tagging is used in only limited ways in tax policy. I propose one explanation: conventional optimal tax theory has yet to capture the diversity of normative principles with which society evaluates taxes. I generalize the conventional model to incorporate multiple normative frameworks. I then show that if the principle of equal sacrifice--a classic, comprehensive criterion of fair taxation proposed by John Stuart Mill and associated with the Libertarian normative framework--is given some weight in the social objective function, tagging generates costs that must be weighed against the benefits it generates through conventional channels. Only tags that are sufficiently predictive of ability, such as disability status, will be used. Calibrated simulations using micro data from the United States show that optimal policy may simultaneously include substantial redistribution across income-earning abilities, as in the standard model, and reject three prominently-proposed tags--gender, race, and height--as in actual policy. This explanation for limited tagging also implies that optimal marginal tax rates at high incomes are lower than in standard analysis and closer to those observed in policy. NBER Working Paper #18045, HBS WP #12-064.