Modeling the Public Welfare Systems: Part I

Helman I. Stern
Attia I. Sweillam

DOI: 10.2190/LAC8-J9JT-Q9K3-FMRB


The purpose of this study is to construct an analytical model of caseload dynamics to be employed in the projection of potential N.Y.S. Public Welfare cases. The term "welfare" is used in this study to mean Public Assistance. Public Assistance is disaggregated into five major categories. The model is composed of three major components. Markovian transition rates are determined for two of these components through an analysis of historical data compiled by the Department of Social Services. The components of the system are integrated into a set of difference equations to predict quarterly caseload distributions by category. In addition, the Markov Chain Model is used to analyze the static structure of caseload transitions within the welfare system. The model has been calibrated on three years of quarterly data from 1969 to 1971. The model was tested ex post facto for 1972. This validation procedure indicated mean absolute errors for individual categories and total welfare caseloads of 4. and .2 per cent, respectively. Since the closing part of the model introduced an appreciable portion of errors, further work is required. Part two extends the model to include a detailed methodology for forecasting the openings and closings portions of the model.

Creative Commons License This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.