Shrinkage in a Bayesian Panel Data Model with Time-Varying Coefficients

We consider regression models for panel data, where regression effects and within subject dependence are allowed to vary over time. We adopt a Bayesian approach with priors that allow shrinkage to constant and zero effects as well as to simpler dependence structures. The model is evaluated in a simulation study and applied to the analysis of yearly earnings of mothers in Austria who returned to the labour market after maternity leave.