Worldwide Persistence, Business Cycles, and Economic Growth

Abstract

We study the time series properties of aggregate data drawn from the Penn World Tables using numerical Bayesian procedures which facilitate inference with small samples. We find substantial persistence in world aggregates, and some evidence for a world business cycle. Across economies, there is great dispersion in our measure of persistence of shocks to real gross domestic product. That we also find no evidence of a relationship between growth and persistence sheds light on which of two competing models of endogenous growth is likely to be able to explain the PWT data.