Abstract: Many studies have compared individual measures of health expectancy across older populations by time-invariant variables. However, very few have included time-varying variables when calculating health expectancy. Since events in the life course are likely to be changing over time in related ways, it is valuable to incorporate time-varying socioeconomic factors. This paper proposes a Multiple Multistate Method (MMM) that situates the multistate model within the broader family of Vector Autoregression (VAR) models. When estimating multistate models with sample survey data, sparseness in the transition matrices often makes such models unfeasible should two or more time-varying variables be built into the state spaces. This approach allows for the estimation of more complex state spaces (including the modeling of time-varying covariates) by reducing less important interactions in the model. We then demonstrate the MMM in two empirical applications, showing the flexibility of the approach to explore health expectancies with complex state spaces.
Key words: Multistate model, discrete-time Markov processes, microsimulation, health expectancy, VAR model
Download (1.23 MB)