Projecting U.S. Household Changes with a New Household Model

Leiwen Jiang, Brown University
Brian C. O'Neill, International Institute for Applied Systems Analysis (IIASA)

Anticipating changes in number, size, and composition of households is an important element of many issues of social concern. To facilitate continued progress in these areas, an efficient household projection model with moderate data requirements, manageable complexity, explicit representation of demographic trends and output that includes the most important household characteristics is needed. None of the existing models meets all the needs. This study aims to specify appropriate models for size-specific headship rates as functions of overall population changes and demographic events. Based on the PUMS data of 1900-2000 and future household projection output using dynamic household model ProFamy, we empirically test the validity of the new proposed approach. Comparing the simulation results under various headship rate models with different levels of complexity, we analyze the importance of considering demographic events in projecting household changes with the new approach and provide recommendations for the development of the most efficient household model.