Efficiency Loss and Preferences for Moderate Redistribution - Some Experimental Evidence -

Income redistribution with an efficiency loss is expected to have a twofold negative effect on voters' support for redistribution, as it lowers aggregate egoistic support for redistribution and activates voters' efficiency preferences. The paper is dedicated to test whether such a negative relationship (i) exists, (ii) increases with the size of the efficiency loss and (iii) interacts with the individual income position. We present a laboratory experiment in which subjects receive a randomly allocated income and then must communicate using a novel communication tool, which keeps track of the iterative communication process until the group reaches a common tax rate. The rate of money "lost" as a part of the redistribution process is manipulated as a treatment variable (efficiency losses: 0%, 5%, 20% and 60%). Experimental evidence shows that efficiency loss exerts a robust negative effect on individual and collective support for redistribution. The effect shows a tipping point pattern, outlives group communication, is stronger at the lower end of the income distribution and is not fully explained by egoistic preferences. The average level of individually preferred and collectively agreed redistribution in societies for each efficiency loss is "moderate" (65% - 40%) which means that inefficiencies are less important in the context of redistribution