The impact of gender and ethnic discrimination on redistribution and productivity

This study analyzes the impact of gender and ethnic discrimination on redistributive preferences and productivity using a large online experiment with US citizens on Amazon's Mechanical Turk. Participants are randomly allocated to different payment schemes for a real-effort task. Four payment schemes discriminate against women, men, whites, or people of color. Men's productivity slightly increases when they are discriminated against whereas productivity slightly decreases for women and people of color when they are discriminated against. After the task and revealing earnings, participants are given the chance to redistribute earnings by voting on a tax rate for the group. Discrimination against women highly increases preferred tax rates and discrimination against whites or people of color moderately increases the demand for redistribution. The results indicate that different forms of economic discrimination - even if financially indistinguishable - lead to very different reactions amongst the entire population and sub-groups. The results also indicate that brute luck of gender or ethnicity occurring at birth is perceived as different from brute luck experienced later in life.