A Bayesian Approach to Modelling Longitudinal Employment Status of Immigrants.

This paper investigates a Bayesian hierarchical model for the analysis of transitions in employment status using data from a large longitudinal social survey of immigrants to Australia. Data for each subject is observed on three separate occasions, or waves, of the survey. A model for the employment status of immigrants is developed by introducing, at the first stage of the hierarchy, a multinomial model for the response. Six different models of varying degrees of complexity are considered with subsequent terms introduced to explain wave effects and to capture overdispersion in the form of between-subject variability. To estimate the model we use the Gibbs Sampler, a Markov Chain Monte Carlo (MCMC) algorithm. This algorithm allows time-varying transition effects and within-subject random effects to be estimated with relative ease, and the imputation of missing values according to an appropriate prior distribution. The six alternative models are compared using the Deviance Information Criteria (DIC) which is appropriate for assessing the fit of complex models containing random effects. After accounting for significant covariate effects in the model, results show that the relative probability of remaining unemployed diminished with time following arrival in Australia. [ABSTRACT FROM AUTHOR]