Empirical evidence suggests that dismissal costs are negatively associated with R&D intensity across countries but positively within countries over time. This evidence can be rationalized theoretically with a negative entry effect and a positive innovation effect for incumbent firms.