IdEP Seminar, Matthias Rodemeier - "Positive versus Normative Disagreement in Optimal Income Taxation"

Optimal incomes taxes depend on the efficiency costs arising from behavioral responses to taxation (the positive dimension), and the gains from redistribution according to society’s equity preferences (the normative dimension). This paper quantifies how much each component drives uncertainty over tax rates. We combine a large meta-analysis of 1,200 elasticity estimates of the Elasticity of Taxable Income (ETI) with incentivized experiments with 7,500 U.S. adults who make real redistributive transfers to households at different income levels. Our first key finding is that normative disagreement across the population about redistribution explains up to 85% of uncertainty in optimal tax rates, substantially more than positive uncertainty. Our second finding is that government aversion impedes support for progressive taxation: many people have strong redistributive preferences but do not view taxation by the government as the solution to reduce inequalities. Our findings demonstrate that uncertainty over optimal income taxes stems primarily from normative disagreement about redistribution and the role of government, rather than from behavioral responses to taxes.

This paper is joint work with Gregory Sun (Washington University).

Matthias Rodemeier
Assistant Professor of Finance at Bocconi University