Abstract
We use randomized pay experiments among Uber drivers, paired with a natural experiment in access to a competitor, to examine how outside options shape labor supply to the firm. When hours are flexible, the firm-specific labor supply elasticity combines a market-hours component and a firm-substitution component. Access to a single competing platform nearly doubles drivers’ firm-specific elasticity and cuts the implied monopsony markdown from 68% to about 50%. The same experiments identify sex differences: women are about twice as elastic to the market as men, yet no less elastic to their employer.