Consumption Dynamics and Panel Data: A Survey

Abstract

Since the permanent income hypothesis was posed by Friedman [1957] and by Modigliani and Brumberg [1954] stating that consumption is a function of the flow of income (“permanent income”) that, if sustained across one’s life time would just compensate expected earnings and wealth, the question of the sensitivity of consumption to current income has focussed the attention of three decades of econometricians. Yet, until the end of the seventies the analysts always came up against the problem that permanent income is unobservable. Then Hall [1978] showed that by incorporating rational expectations a household maximizing expected intertemporal utility subject to the budget constraint behaves such that the marginal utility of current consumption next year is expected to be proportional to the marginal utility this year (see also Sargent’s [1978] contribution). He also found empirical evidence (on macro data) which suggested that lagged real disposable income and other variables dated t — 1 or earlier had little explanatory power on present (aggregate) consumption so long as lagged consumption was a regressor.