Rational and Empirical Play in the Simple Hot Potato Game

We define a “hot potato” to be an exchange good which may be traded a finite number of times, and which becomes a bad if and when it can no longer be exchanged. Under standard assumptions of rationality and common knowledge, it is easy to show that no rational actor would freely accept a hot potato good, irrespective of the magnitude of the good’s exchange payoff or its last-holder penalty. Boundedly rational actors, however, may accept the good for a number of reasons. Here, we show results from three experiments which indicate that human subjects are willing to accept hot potato goods under a variety of conditions. Willingness to accept the hot potato is positively related to the magnitude of the potential payoff and to the absolute length of the remaining chain; the presence of a greater number of potential exchange partners is found to reduce willingness to accept the good. Qualitative data from subject responses suggest a trichotomy between subjects’ actual behavior, subjects’ attempted behavior, and the explanations of subjects for their own decision processes. Implications of these findings for rational choice theory and economic sociology – including altruism and the emergence of fiat currency – are discussed.