Preferences for Redistribution and Perception of Fairness : An Experimental Study

Why is there political support for progressive taxation and government transfers in western democracies? We study the importance of fairness preferences, risk aversion, and self-interest in determining support for redistribution by conducting a laboratory experiment that includes both elements not previously found together and novel features. Our design permits within-subject comparisons of the effects of earned vs. unearned sources of pre-tax-and-transfer inequality, within-subject comparisons of choice as a disinterested observer of inequalities among others versus that as an affected party, and within-subject comparisons of choice with and without uncertainty as to own pre-tax income. Between subject variation lets us study the effects of changing the direct cost of taxation to the decision-maker and the efficiency cost to the society. We use large subject groups and the actual U.S. pre-tax income distribution to create a more macro framing than in related past experiments. We show how our results can be used to calibrate individual utility functions defined over own expected income, own income variance, social inequality, and efficiency, and we examine what the estimates mean for the median preference for redistribution and the level of redistribution that maximizes additive social welfare. Most of our subjects prefer that there be less inequality among others and demand for redistribution responds in predictable ways to the cost of taxation and to the dead-weight loss associated with it. We also find that preferred levels of redistribution are highly responsive to whether or not pre-tax incomes are determined according to task performance, for male subjects, but much less so for female subjects, with this difference accounting for much of a displayed gender gap in which females prefer more redistribution