The Real Effects of Uncertainty Shocks: New Evidence from Linear and Nonlinear SVAR Models

This paper applies both linear and nonlinear structural vector autoregressive (SVAR) models using two distinct identifications methods to disentangle the macroeconomic effects of uncertainty shocks for a developing country. As an application, we use macroeconomic data for the Democratic Republic of Congo (Congo), one of the largest and least developed countries in the world with a rich history of domestic political instability and high macroeconomic volatility. Our measure of uncertainty is the world uncertainty index for the Congo recently developed by Ahir et al. (2018) for a panel of developed and developing countries. Using a standard SVAR model with sign restrictions, we provide evidence that an unexpected increase in uncertainty triggers contractions in GDP and investment on impact in the Congo. We show that uncertainty shocks are among the greatest drivers of economic fluctuations. Our results are robust across alternative linear and nonlinear SVAR specifications using the Cholesky decomposition.