Dynamic Microsimulation of Elders' Health and Well-Being
Douglas A. Wolf, Principal Investigator
National Institute on Aging Grant No. R01-AG11815-01A1
The overall goals of this project are:
- to develop a model in which the trajectories of health/disability status, family composition, and economic resources are jointly determined, and
- to produce a computer program that allows us to explore the dynamic implications of this model using microsimulation techniques.
While the importance of understanding the dynamics of health and associated behaviors within the elderly population is widely acknowledged, researchers and policy analysts have also come to recognize the importance of simultaneously taking account of developments in the domains of health and functional status, family composition, and economic resources. However, the complexity of the outcome space when these dimensions are considered jointly are such as to make attractive the microsimulation approach, while greatly restricting the scope for analytic approaches. The model to be embedded in our microsimulation program is intended to support both scholarly inquiry (of substantive and methodological interest) and, equally importantly, to inform current and future policy analysis in the areas of income security, health service use, and long-term care policy for the older population.
The specific aims are to:
- Specify and estimate equations for dynamics of functional status, nursing home occupancy, income and death among those 65+, using data from the 1982, 1984, and 1989 National Long-Term Care Survey (NLTCS) linked to Medicare data for 1982-1992, building on and extending the Grade of Membership (GoM) framework;
- Develop equations for year-to-year income streams, determined jointly with changes of marital status, for all ages represented in the cohort to be simulated;
- Estimate model parameters governing the dynamics of family composition (the existence and characteristics of spouse, parents and children);
- Integrate the results of the above modeling efforts in a microsimulation computer program with the capacity to dynamically simulate life histories, focusing on the elderly population;
- Validate the simulated output by comparing the model's projections to actual data where possible, analyze the uncertainty attached to the output from the microsimulation model, and conduct sensitivity analyses of model results to alternative underlying assumptions; and
- Using microsimulation, conduct a series of disaggregated projections of the elderly population and its characteristics, and associated policy analyses.
The computer program to be created will be flexible in design, in order to adapt to a broad range of potential applications, and will be written so as to be maximally portable. It will also be fully documented and made available to the research and policy analysis communities..
