The authors use high-frequency data on the productivity of individual supermarket cashiers to study the duration effects of unimployment insurance benefits. Since the magnitude and duration of unimployment insurance alters the expected cost of job loss, do employed workers respond to benefit changes? The authors find a negative relationship between unimployment insurance benefits and worker productivity, suggesting evidence of this kind of ex-ante moral hazard effect. This effect is strongest for more experienced and less productive cashiers, for whom unimployment insurance expansions are especially relevant.