Incentivizing motivation and self-control preferences

In this paper we develop a theoretical model concerning self-control, motivation and commitment. The model, based on Gul and Pesendorfer (2001), studies a two-period decision problem of an agent who faces a given menu and might experience temptation in the second period. In this case, the agent is tempted by a choice that is, from an ex ante normative point of view, inferior and has to exercise self-control to resist temptation. A random, time-variant degree of motivation is introduced to influence his cost of self-control. We introduce an investment-payoff combination as a commitment device that can help the agent pre-commit to his normative choice before he faces temptation. Depending on how accurately the agents predict their future self-control costs we distinguish between sophisticated and (partially) naive agents. The theoretical results show that our mechanism can help agents to commit successfully, and also can explain why certain agents with a preference for commitment might fail - behavior that is usually attributed to a preference reversal. This commitment failure is associated with underestimation of future self-control costs.