OR Practice—Computer-Based Models for Policy Making: Uses and Impacts in the U.S. Federal Government

This paper examines factors influencing implementation and use of computer models that have been successfully institutionalized as part of federal policy-making. It is based upon detailed case studies of two kinds of models that have been widely assimilated and used in federal agencies—microanalytic simulation models and macroeconomic models. It concludes that the most important influences on successful modeling are the means to do modeling (supply factors) and the desire of bureaucrats and politicians to use model-generated information (demand factors). Of these, demand factors are most important. The availability and promotion of computer models alone is insufficient to generate use, whereas their political saliency is critical. Thus, where the means to do modeling are weak, but the desire to use models is strong, government agencies invest heavily in the creation of new modeling efforts.