Colombian Growth Determinants: What Do We Really Know?

Bayesian Model Averaging of Classical Estimates (BACE) approach is undertaken to analyze the short and long run determinants of Colombian economic growth in the context of time series. The methodological approach uses an Auto Regressive Distributed Lag (ARDL) scheme to derive the short and long run impacts of an array of growth determinants suggested in theoretical and empirical literature. The results point up to show that short and long run path of growth is robustly explained for the amount of workers and stock of capital. Aside from this correlates the effect of other variables are negligible and lack statistical significance in a classical sense. Trade openness, real depreciation and bank liquidity shows a positive though negligible relationship with short and long run growth, public debt and debt service apparently reduce short run growth but improve the long run path of growth. In the short run framework, changes in economic growth are positively associated with public deficit and negatively with growth volatility. Results for other correlates remain ambiguous.