The 'ratchet effect' refers to a situation where a principal uses private information that is revealed by an agent's early actions to the agent's later disadvantage, in a context where binding multi-period contracts are not enforceable. In a simple, context-rich environment, we experimentally study the robustness of the ratchet effect to the introduction of ex post competition for principals or agents. While we do observe substantial and significant ratchet effects in the baseline (no competition) case of our model, we find that ratchet behavior is nearly eliminated by labor-market competition; interestingly this is true regardless of whether market conditions favor principals or agents.