DETERMINANTS OF ECONOMIC GROWTH: A BAYESIAN PANEL DATA APPROACH

Due to model uncertainty, there is no consensus about which are the most salient determinants of economic growth. To address model uncertainty, Bayesian Model Averaging approaches have recently been used in cross-country cross-sectional analyses. This paper extends these approaches to panel data models with country specific fixed effects. We find that when using this approach the strongest evidence of robustness is for the trade openness, the size of the labor force, the government consumption level and the Latin America dummy. Moreover, our results are robust to different prior assumptions on the expected model size.