Don’t shoot yourself in the foot! A (real-effort task) experiment on income redistribution and voting.

According to the World Inequality Database 2020, income inequality has increased in almost all countries in recent decades. Nevertheless, top tax rates on upper-income earners have declined during the same period. This empirical evidence seems not to support models based on the median voter theorem as the one proposed by Meltzer and Richard (1981). To shed some light on this contradictory evidence, this paper reports results from a laboratory experiment that investigates the Meltzer–Richard model of equilibrium tax rates in which individuals face a real-effort task that includes leisure at the work place. We classify participants into high-skilled and low-skilled workers, according to their performance in a tournament at the beginning of the experiment. They vote over two (high and low) exogenous tax rates and their pre-tax income is determined according to their performance in a real-effort task and also to the skill premium. We find that a large proportion of low-skilled workers vote for the lowest tax rate (the one that gives them the lowest payoff), especially when the alternative tax rate is very high. However, this proportion is significantly reduced in treatments in which the subjects are given extra information about how the tax operates in redistributing income. This result suggests that the lack of information about the role of taxes in income redistribution may be an important factor in explaining the counter-intuitive voting behavior of low-income voters over income redistribution. We also find some support that the prospect of upward mobility and the belief in the negative effect of taxes on productivity make low-income voters support low tax rates, especially when the alternative tax rate is very high.