Growth Volatility Indices

Abstract

We study the determinants of growth rate volatility in a multisector economy where sectors are heterogeneous in their individual volatility. We propose a model where aggregate volatility is explained by structural change and the size of the economy. We present a first attempt to test these predictions measuring growth volatility by indices based on Markov transition matrices. Growth volatility appears to (i) decrease with total GDP and (ii) increase with the share of the agricultural sector on GDP, although some nonlinearities appear. Trade openness, which we relate to the size of the economy, also plays a role. In accordance with our model, the explanatory power of per capita GDP, a relevant variable in other empirical works, vanishes when we control for these variables.