There is an unsettled debate in experimental economics literature regarding the consistency of individuals' risk preferences in varying institutions. Much of this debate stems from observations of subjects' bids in sealed-bid auctions and the implications of those bids. In this paper, I have subjects participate in a sealed-bid auction experiment and then examine if the ostensible risk parameter that one can back out from subjects' bids matches up with their elicited risk preference from a separate task in the experiment. I find that subjects do exhibit consistent risk preferences. The aggregate measure of the subjects' risk parameter is stable across both tasks, the estimated numeric values of subjects' risk parameters are stable across tasks, and the ranking of subjects (most risk averse to least risk averse) is stable across both tasks.
Available at: http://works.bepress.com/russell_engel/4/