Interpreting Saving-Investment Correlations

Abstract

Feldstein and Horioka (1980) argued that the cross-sectional correlation of saving and investment provides a test of global capital mobility. We argue that the long-run correlation is determined by the intertemporal budget constraint, limited capital mobility and current account targeting. The short-run correlation reflects limited capital mobility and adjustment to supply and demand shocks. Our empirical analysis shows that the short-run correlation varies across countries and not over time, which suggests that it is a country-specific business cycle fact. The long-run correlation has substantially decreased over time, which suggests that limited capital mobility is partly responsible for its high value in the past.