Despite a large amount of evidence about discrimination in the gig economy, the mechanisms that result in lower earnings and fewer opportunities for women, racial minorities and other groups remain poorly understood. This project examines the underlying mechanisms of hiring and matching in digital labor platforms. In particular, it seeks to understand how platform architecture and user behavior interact to affect opportunities for different groups. The methodology combines big data analytics with traditional survey and experimental methods. The results seek to contribute to theorizing about discrimination and inequality in the gig economy.