What Drives Global Uncertainty? Learning from Bayesian Approaches

This article examines the determinants of global uncertainty. In contrast to the instrumental approach, which assesses the effect of uncertainty on socio-economic conditions or the political-institutional environment, we adopt an intrinsic approach that focuses on the factors influencing the internal dynamics of uncertainty. Based on a global sample of 142 countries observed over the period 1996-2020, we adopt a Bayesian Model Averaging (BMA) approach, which combines a series of rival models and has the ability to reduce overconfidence in a particular model by incorporating uncertainty about the variables used. Of the 18 potential determinants selected, the results highlight the relevance of 13 of them, which are socio-economic and structural. Furthermore, the results show that the dynamics of overall uncertainty are sensitive to the level of economic development of the countries, and the sample selected. Also, these results are robust to some alternative methods such as MAMI, WALS and GETS