Do workers discriminate against their out-group employers? Evidence from an online platform economy

Abstract

We study possible worker-to-employer discrimination manifested via social preferences. We run a well-powered, model-based experiment, wherein we recruit 6,000 white American workers from Amazon’s M- Turk platform for a real-effort task. We randomly (and unobtrusively) reveal the racial identity of their non-fictitious employer, who may either be white or black. We find evidence of race-based altruism towards black employers. However, the workers display significant racial discrimination in reciprocity - a small gift induces workers to put higher effort for white employers relative to black. Our results suggest that taste-based discrimination favoring ingroup can have significant adverse effects on outgroup employers.

Keywords: Discrimination, Worker-to-Employer, Social Preferences, Taste-based discrimination, Gig Economy, Mechanical Turk, Structural Behavioral Economics

JEL Classification: J71,D91, C93