In this paper, we test for the direct wealth effect in aggregate data on U.S. households over four distinct sub-periods from 1952 through 2011. We use recent time series techniques to distinguish between the direct wealth effect from indirect channels which may operate through personal disposable income or liabilities. We find evidence of a direct wealth effect for housing wealth, in particular, from 1998 to 2011. The responses of consumption in the 1998 to 2011 period are in contrast to an indirect or “common cause” explanation of the wealth effect. For financial wealth, there is some evidence of a direct wealth effect for the 1998 to 2011 period, but the effect overall is smaller than for tangible wealth. Also, before 1998 evidence for a direct wealth effect from either housing wealth or financial wealth is weak.
