We develop a Bayesian approach that performs variable selection in panel regression models that are subject to breaks. Our variable selection approach enables deactivation of pervasive regressors and activation of weak regressors for short periods. Allowing the coefficients on individual...
