Abstract:
We consider the issue of the dynamics of perceptions, as expressed in responses to survey questions on subjective wellbeing. We develop a simulated maximum likelihood method for estimation of dynamic linear models, where the dependent variable is partially observed through ordinal scales. This latent auto-regression model is often more appropriate than the usual state dependence model for attitudinal and interval variables. The paper contains an application to a model of households' perceptions of their financial wellbeing, demonstrating the superior fit of the latent auto-regression model to both the usual static model and the state dependence model. Copyright 2008 Royal Statistical Society.