A Simulation Model for Welfare Policy Analysis

This paper describes a new approach to large-scale simulation modeling for the analysis of public welfare policies. Whereas conventional microanalytic simulation models extrapolate the dynamic behavior of a representative sample of individuals to the corresponding target population, the new modeling technique disaggregates an exogenous demographic forecast into enough detail to evaluate the impact and cost of proposed social programs. This approach uses a multivariate extension of the Johnson translation system of probability distributions to estimate the percentage of individuals in a target population who satisfy given program eligibility criteria. The methodology is demonstrated in a prototype decision support system for analysis of the Institutional Care and Community Care programs of the Texas Department of Human Services.