Good Tags, Bad Tags
Chris William Sanchirico
University of Pennsylvania Law School; University of Pennsylvania Wharton School - Business Economics and Public Policy Department
July 18, 2013
U of Penn, Inst for Law & Econ Research Paper No. 13-24
In the literature on optimal taxation, a “tag” is a government-observable taxpayer attribute that is effectively immutable – like blindness, race, gender, or even height. Conventional optimal tax theory prescribes that tags should be included in the tax base so long as they are in some way correlated with “ability” or “endowment” (more precisely, with “social welfare weight”: the marginal social welfare of transferring resources to the taxpayer). Such correlation is a weak requirement. And it has recently been pointed out that the list of seemingly absurd taxes and subsidies that are thereby deemed optimal poses a challenge to the basic framework of optimal tax theory in the form of a reductio ad absurdum. This paper attempts to draw a principled distinction between two ideal types of tags – those that are directly welfare-relevant and those that are welfare-relevant only when and if they are taxed or subsidized, and only through such taxation or subsidy. A stark distinction arises between these types when the optimal tax model is extended to include the realism-enhancing feature that the government is uncertain regarding the association in the population of taxpayers between the incidence of the tag and social welfare weight. It is shown that such uncertainty generally decreases the impetus for taxation or subsidy when the attribute is non-welfare relevant, but not when it is directly welfare-relevant.
Number of Pages in PDF File: 14
Keywords: Tags, optimal taxation, policy uncertainty
JEL Classification: H2, H20, H21, H24, K34
Date posted: July 20, 2013 ; Last revised: July 27, 2013
© 2015 Social Science Electronic Publishing, Inc. All Rights Reserved.
This page was processed by apollo7 in 0.297 seconds