Can Experiments Falsify Expected Utility Theory?

Some recent articles debate the implications of the results of experimental economics. It is claimed by some that these results challenge the assumptions of expected utility theory. Others deny this. Both sides presume that the assumptions of rationality or expected utility-maximization are potentially falsifiable by empirical tests. This article contests this assumption. Building on previous work in economic methodology, it is argued that non-falsifiable assumptions, such as the standard rationality postulates, are potentially universal. Hence they can embrace any empirical phenomenon. That is both their strength and their weakness. The article concludes that the debate among experimental economists can only proceed if both sides accept the non-falsifiability of key propositions under dispute. That done, the protagonists can turn to the more pertinent questions such as the criteria for choosing theories. In particular, experimental economics may have a role in suggesting more narrow and context-specific behavioural assumptions for economic theory