Abstract
Despite legal protections, discrimination based on sexual orientation remains difficult to detect and measure in labor markets. We present evidence from a field experiment (N = 1,128) conducted in Australia on a major gig economy platform that allows users to outsource everyday tasks. By manipulating both the sexual orientation of job posters (gay vs. straight male) and physical proximity requirements of tasks (inside vs. outside the home), we identify discrimination against gay profiles and demonstrate a specific mechanism for its emergence: physical proximity. Gay profiles received fewer offers and less engagement, and attracted workers with significantly lower quality ratings, but only for tasks requiring close physical proximity. Additional exploratory analysis further suggests that the results are driven by an anti-gay bias rather than an anti-men or pro-women bias and that discrimination is stronger for tasks with a higher degree of interaction between worker and job poster. Our findings demonstrate how digital labor platforms can perpetuate traditional forms of discrimination, while shedding light on physical proximity as one important mechanism. These results have important implications for users of the gig economy, a sector of increasing importance due to its rapid growth and comparatively lax regulatory frameworks.