We provide a critical assessment of the ambiguity aversion literature, which we characterize in terms of the view that Ellsberg choices are rational responses to ambiguity, to be explained by relaxing Savage's sure thing principle and adding an ambiguity-aversion postulate. First, admitting Ellsberg choices as rational leads to behavior, such as sensitivity to irrelevant sunk cost, or aversion to information, which most economists would consider absurd or irrational. Second, we argue that the mathematical objects referred to as 'beliefs' in the ambiguity aversion literature have little to do with how an economist or game theorist understands and uses the concept. This is because of the lack of a useful notion of updating. Third, the anomaly of the Ellsberg choices can be explained simply and without tampering with the foundations of choice theory. These choices can arise when decision makers form heuristics that serve them well in real-life situations where odds are manipulable, and misapply them to experimental settings. * We are grateful to Drew Fudenberg, Edi Karni, Bart Lipman, Marciano Siniscalchi, Costis Skiadas, and Rakesh Vohra for detailed comments that substantially improved the paper. We also thank

No abstract

No abstract