Abstract

The paper analyzes the role of current income in providing new information about future income and thus signalling changes in permanent income. Using time-series analysis to quantify the revision in permanent income induced by an innovation in the current income process, a structural econometric model of consumption is developed. The rejection of the joint rational expectations-permanent income hypothesis is both statistically and quantitatively significant. The paper also shows that the test of the rational expectations-permanent income hypothesis proposed by Hall is based on the reduced form of this structural model and reconciles Sargent's consumption paper with Hall's.