Investment-saving comovement under endogenous fiscal policy

I expand Feldstein's (1983) model by including flexible exchange rate and by introducing endogenous fiscal policy. Using this model, I demonstrate how a positive investment-saving correlation can arise in a world with endogenous fiscal policy. I show that this correlation does not depend on capital mobility and therefore is compatible with any degree of capital mobility. This implies that the observed investment-saving comovement is not necessarily due to imperfect capital mobility. The model has a testable implication: it predicts a lack of Granger causality from private saving to private investment. Empirical examination of this prediction indicates that U.S. time series data is compatible with the hypothesis of endogenous fiscal policy during a flexible exchange rate period, but not during a fixed exchange rate period.