Trade openness and economic growth: Bayesian model averaging estimate of cross-country growth regressions

•This paper investigates the relationship between openness and economic growth.•It utilises Bayesian model averaging techniques to account for model uncertainty.•The paper concludes that openness is not robustly correlated with long-run growth.•Economic institutions and macroeconomic uncertainties are key factors for growth.

In this paper, we investigate the robustness of the relationship between trade openness and long-run economic growth over the sample period 1960-2000, utilising Bayesian model averaging techniques to account for model uncertainty issues in a systematic manner. We find no evidence that trade openness is directly and robustly correlated with economic growth in the long run. We further check the robustness of this finding by employing a battery of proxies for trade openness, namely, current openness, real openness, the fraction of open years based on the Sachs and Warner (1995) criteria and the weighted averages of tariff rates, non-tariff barriers and the black market premium. The main result is robust to the inclusion of different trade openness proxies and none of the proxies is robustly associated with economic growth. The data evidence also indicates that economic institutions and macroeconomic uncertainties such as those induced by high inflation and excess government consumption are key factors in explaining economic growth.