Decision Making and Postdecision Surprises

Abstract

J. Richard Harrison and James G. March Most ideas of intelligent choice assume that decision making involves estimating the probable future values of currently available alternatives and choosing the best of them. Sometimes chosen alternatives turn out to be better than anticipated; sometimes they turn out to be worse. The difference between the predecision estimated value of a chosen alternative and its postdecision value, determined after some of its consequences have been experienced, can be defined as postdecision surprise. This paper examines a systematic bias in the distribution of postdecision surprises attributable to the structure of intelligent choice itself. It is shown that unbiased, random errors in estimation result in a structural tendency toward postdecision disappointment, which will be most characteristic of decision situations in which variation among the true values of alternatives is relatively small, the ambiguity or uncertainty in evaluation is relatively high, and the number of alternatives considered is relatively large. The decision dilemma is clear. Choosing apparently better alternatives will, on average, produce higher returns; however, in the absence of behavioral adjustments, higher expected benefits will be associated with greater expected disappointments. The effects are illustrated with results computed for the special case of normally distributed values and errors. Some implications are suggested for understanding postdecision surprise and the development of social norms of intelligent choice in individuals and organizations.*