We use information from the television game show with the highest guaranteed average payoff in the U.S., Hoosier Millionaire, to analyze risk-taking in a high-stakes experiment. We first characterize gambling behavior under alternative assumptions about contestant rationality and preferences, and derive testable restrictions on individual risk-attitudes based on these assumptions. We then use an extensive sample of decisions in these games to estimate distributions of risk-aversion parameters that are consistent with the theoretical restrictions and revealed preferences. We find that although most contestants display risk averse preferences, the extent of the risk aversion varies substantially with the stakes involved in the different decisions.
 

Click here to download a full PDF version of this paper.