Precautionary saving has engendered much interest, both because of the possibility that it can explain why, contrary to the basic Lifecycle/Permanent-Income Hypothesis, consumption roughly tracks income over the lifecycle and because of speculation that precautionary saving might account for a large fraction of aggregate saving. However, recent findings have indicated that the effects of precautionary saving are much less significant than was first thought. Early researchers did not take into account general equilibrium effects. Furthermore, most research into precautionary saving has employed buffer-stock saving models that also incorporate a borrowing constraint (either imposed exogenously or resulting endogenously from assumptions about the income process). After separating out the effects of uncertainty from the effects of the borrowing constraint, one finds that most buffer-stock saving can be accounted for by the borrowing constraint. However, buffer-stock models make the simple but unrealistic assumption that absolutely no borrowing can occur. While it has generally been assumed that borrowing constraints and precautionary saving are complementary, when these two frictions operate simultaneously a tight no-borrowing constraint will dominate. Here, we show that when borrowing is unconstrained, precautionary saving can, indeed, have significant effects. Moreover, a general equilibrium model of precautionary saving and unconstrained borrowing can better explain the lifecycle consumption profile than the corresponding buffer-stock saving model
