Saving and income smoothing: Evidence from panel data

In this paper we use data from a panel of Dutch households to investigate the behavior of savings. As in Campbell (1987), we examine whether the pattern of savings is consistent with the predictions of the rational expectations-permanent income model, We consider one important prediction of the model: whether savings anticipate future income changes. We test the predictions of the permanent income model in several ways, First we propose some informal tests that exploit directly the information on expected future income changes available in the data. We then perform a test of the model by setting up a vector autoregression (VAR) in terms of savings and income changes and by using Euler equations expressed in terms of savings, The empirical evidence shows that in all the specifications used in this paper, the permanent income model is rejected, Contrary to the evidence from U.S. macro data, we do not find that the 'excess sensitivity' of consumption to income is responsible for the rejection of the model.