Accepted paper:

Do citizen tax policy preferences reflect efficiency and inequality considerations? Experimental evidence from Tanzania

Authors:

Arne Wiig (Chr Michelsen Institute)
Odd-Helge Fjeldstad (Chr. Michelsen Institute)
Ivar Kolstad (Chr. Michelsen Institute)

Paper short abstract:

Based on a discrete choice experiment, citizen preferences for taxation do not seem to reflect considerations of efficiency and inequality. People rather respond to simple arguments related to the nationality of company owners and local employment effects.

Paper long abstract:

Optimal tax theories typically include efficiency and inequality in the objective function of the government. However, it is unclear how well this objective function reflects how ordinary citizens think about taxation. In a field experiment in Tanzania, respondents randomly assigned to two treatment groups were asked to explicitly state the tax implications of focusing on efficiency and inequality, respectively. Their tax decisions were measured using a discrete choice experiment, and responses compared to those of a control group stating their own tax preferences. The results show that the responses in the three conditions are virtually indistinguishable. In other words, in a situation of some complexity, citizen preferences for taxation do not seem to reflect considerations of efficiency and inequality. The result holds even if we control for respondents with a poor understanding of the task given. Results from the control group instead suggests that people respond to simple arguments related to the nationality of company owners and local employment effects. This has implications for the understanding of the arguments used in tax debates, and for the political economy of taxation more generally.

panel J05
Preferences for redistribution: Experimental evidence