No. 240: How Much and From Whom? An Information Experiment on Tax Misperceptions and Redistribution Preferences

Why do some studies find little appetite for “more redistribution” in aggregate despite perceived rises in inequality? We argue (i) that general statements about “redistribution preferences” do not capture the two distinct dimensions of redistributive tax design, namely progressivity and the revenue implied by the tax rate schedule and (ii) that misperceptions along each dimension shape preferences. We field a survey experiment on Germany’s personal income tax that elicits respondents’ perceived and fair average tax rates (ATRs) for the whole income distribution and then decompose preferences into a progressivity and revenue dimension. To study misperceptions, we provide information intended to correct beliefs about the current level of progressivity and revenue of the system. We find that revenue misperceptions inflate ATRs but leave progressivity intact: under a revenue-neutrality constraint, respondents rescale rates without changing progressivity. Second, providing progressivity information raises desired levels, consistent with a shift in reference points; these shifts are largest where baseline misperceptions are greater and are not systematically moderated by partisan identity. Methodologically, the decomposition is essential: single ATRs are easily biased by revenue beliefs, whereas progressivity captures distributional intent. For policymakers, separating these two dimensions clarifies what information campaigns should target, how to communicate reforms, and how to read survey responses as inputs to policy design.