This article describes and assesses the development of a major evaluation and policy research tool-microdata simulation modeling-over the past two decades. The method of microdata simulation modeling is described and its reliance on the development of large, nationally weighted household surveys is noted. After a brief tracing of the history of this evaluation method from the first-generation static models to dynamic models, to models with economic behavior imbedded in them, and to applied general equilibrium models, the problems and prospects of the technique are assessed. The analytic gains in richness and reliability are cited, as are the offsetting problems of cost, complexity, and the reliability of the underlying microdata and behavioral response junctions. The discussion concludes with an examination of the effect of this War on Poverty-stimulated research method on the development of the social sciences.
