Policy analytic implications for a theory of prediction and decision

Abstract

Decision theory developed to prescribe the consistent choice behavior under uncertainty of an ideal individual. The policy analysis context of organizational decision-making, however, necessitates certain modifications to the fundamental decision theory in order to deal with the problems of group consensus, ill-defined objectives and disparate information sources. The argument is presented that more attention must be given to the explicit formulation of judgment and that, in particular, there has to be an integration of the methodology of forecasting into the decision-analytic framework. A consequence of adopting a decision-theoretic approach to forecasting is the apparent need to develop ways of synthesizing the set of available predictive methods in order to satisfy the subjectivist principle of total evidence utilization. The synthetic approach to forecasting is in contrast to the conventional selective method. A brief review of some operational methods of pursuing the synthetic approach is given.