working papers




Older working papers


A widely held opinion identifies the Principle of Dynamic Consistency (PDC) as a cornerstone of the idea of individual rationality. An immediate implication of this view is that non-Expected Utility models lead to irrational behavior, owing to their fundamental incompatibility with the PDC. Consequently, a return to the Bayesian paradigm is strongly advocated. We observe that the "rationality" content of the latter consists of the principle that the decision maker does the best he/she can given the available information. The principle, which is embodied in the concept of conditional expectation, is expressed in the form of an optimization problem involving an L 2 - norm. We show that alternative decision making paradigms can be generated by retaining the same optimality principle but with other norms or divergence functions in the place of the L 2 - norm. As there is no reason to claim the superiority of a norm over any other, after observing that these paradigms violate the PDC, we conclude that if the Bayesian Paradigm is to be considered rational, then Dynamic Consistency has nothing to do with rationality.


The Center on Capitalism and Society Working Paper No. 75, 2012.


CIREQ WP 13-2009.


Columbia University DP 2006.


 Columbia University, DP No. 0203-22.


Columbia University, DP No. 0102-73.


Columbia University DP 0506-17.