Transfer Income Model (TRIM) Projection Capability: A Report on the Validation of the Housing Allowance Data Base

摘要

The Urban Institute's Transfer Income Model (TRIM) operates on a large, nationally representative microdata base used in various contexts to analyze alternative income - conditioned transfer plans. This validation report focuses on one of the existing TRIM capabilities -- its ability to represent different time periods, as applied to the Experimental Housing Allowance Program (EHAP). EHAP is a research project analyzing direct cash housing assistance to lower - income clients. The TRIM aging procedures involve updating the basic microdata, representative of 1969 incomes and 1970 demographic information, and synchronizing it with experimental data representative of more recent years. The validation of these aging procedures consists of two phases; this report constitutes an interim stage between them. It reports on the validity of earlier TRIM estimates of cost coverage and benefit patterns for various designs of a housing allowance program, appraises current procedures and identifies areas that need improvement, and provides information on the desirability of using a new data base in the second phase of the Integrated Analysis work on alternative housing allowance programs. Section 2 of this report compares a recently aged census public use sample data base with published estimates from the current population survey. Compared in Section 3 are factors used in aging the data base in phase one to data from the annual housing survey. The discussion in Section 2 focuses on the validity of the procedures, highlights areas requiring additional work, and provides information for deciding whether or not to continue using an aged data base. Section 3, on the other hand, focuses on the accuracy of the forecasts used in the first phase of the TRIM rather than on procedures validity. The projected growth rates of the earlier TRIM work for house value and gross rent are compared to the actual growth rates revealed by the annual housing survey, thus serving to validate previously provided estimates of cost, coverage, and benefit patterns. The final section discusses the weaknesses in the procedures and indicates how they could be overcome by the introduction of the 1976 survey of income and education, which is in a time frame closer to that of the experimental findings; thus, less chronological synchronization would be necessary between the data base and the behavioral information. Tabular data are included. Tables and footnotes are provided.