We propose a dynamic principal-agent model characterized by a state variable given by the agent’s initial bargaining power and by a law of motion that governs the changes of their bargaining power. Through a numerical exercise, we find that agents with the same relative risk aversion show...