In this paper we study ultimatum games with incomplete information on the side of the proposer, which are repeated against changing opponents. The games have the same subgame equilibrium outcome as its complete information version. A proposer has to decide on an offer for the responder without knowing the exact pie size. A responder can accept or not the offer. If the offer is rejected both get nothing. If the offer is greater than the pie size there are two versions of outcome. Either the outcome is not feasible (called ER) or the proposer receives a negative payoff (called NP) if his offer is accepted. We distinguish two negative payoff versions: with starting capital allowance (called NP10) to prevent that a proposer can get total negative payoffs; or without any starting capital allowance (called NP0). We use the strategy method developed by Selten (1967). We find that offers in NP0 are closer to the subgame perfect equilibrium than ever observed in any ultimatum game. We interpret the behavior as loss avoidance. Responders typically don�t reject offers greater than the pie size, which lead to negative payoffs for the proposers. Half of them accept the smallest money amount for all possible pies and the other half respond with strategies that indicate that payoff comparisons are relevant, although proposers don�t know the pie size.
© 2001-2024 Fundación Dialnet · Todos los derechos reservados