Long-term Employment Relations when Agents Are Present-biased

Abstract: We analyze how agents' present-bias affects optimal contracting in an infinite horizon employment setting absent moral hazard. Sophisticated agents, exactly aware of their present-bias, and fully rational time-consistent agents are optimally offered the same contracts. In contrast, naive and partially naive agents can be exploited by the principal to the same degree. These naive agents are offered a menu of contracts, consisting of a virtual contract - which they intend to choose in the future - and a real contract which they end up choosing. The results are robust to imposing limited-liability constraints. Moreover, under limited-liability implemented effort can be inefficiently high from a social planner perspective and inefficient employment relationships may not be terminated. The findings are also robust to learning or adverse selection and persist in a finite time horizon set-up.