A0505
Title: Grouped heterogeneity in Markov-switching panel models
Authors: Bernhard van der Sluis - Erasmus University Rotterdam (Netherlands) [presenting]
Abstract: Grouped heterogeneity has become a popular way of characterizing heterogeneity in panel data. Similarly, regime-switching is often used to parsimoniously characterize the instability of economic relationships. Both features are combined in a single panel data model. The panel model contains per individual a separate finite-state Markov process with different coefficients per regime. Different ways of grouping are considered, ranging from grouping coefficients only and leaving the regimes unrestricted, to grouping the latent regimes and coefficients at the same time. A two-step estimation procedure is proposed that combines the grouped fixed effects approach with the expectation-maximization algorithm. It is shown that estimators for the slope coefficient and the group membership structure are consistent, also when the regimes follow a latent Markov process. Monte Carlo simulations demonstrate good finite sample performance of the proposed procedure, even when some assumptions are relaxed. The methods are applied to examine similarities in business cycle patterns across the U.S. states.
