We develop a simple Bayesian method for heterogeneous variable selection in nonlinear panel data models, where heterogeneity implies that variable selection takes place at the individual level.
Each individual-specific parameter is either zero or comes from a Dirichlet process mixture of normals. For inference, we develop an efficient MCMC sampler. In a Monte Carlo study, we show that our approach is able to capture heterogeneous variable selection whereas a standard Dirichlet process mixture is not. An application on real choice data reveals that accounting for heterogeneous variable selection and non-normal continuous heterogeneity leads to an improved out-of-sample fit.