UNANTICIPATED AGGREGATE DISTURBANCES AND TESTS OF THE LIFE-CYCLE CONSUMPTION MODEL USING PANEL DATA

Several recent studies have used data on food consumption from the Panel Study of Income Dynamics (PSID) to test the rational expectations life-cycle hypothesis against the alternative of prevalent liquidity constraints. These tests invoke the rational expectations restriction that income forecast errors are independent of lagged information. This restriction, however, applies only in a time-series context. it is possible that unanticipated macroeconomic disturbances cause forecast errors to be correlated with lagged information in a cross-section of families. We present evidence of such a correlation in the PSID data, show that this correlation biases previous tests of the life-cycle model toward rejection, and derive a proper test of the life-cycle model using panel data.