In this paper we present empirical evidence of the importance of aggregation bias in Euler equations for consumption. The main result is that estimates of the elasticity of intertemporal substitution for consumption are consistently lower for aggregate data than for average cohort data. In trying to explain these differences we find that a major role is played by the non-linearity of the estimable equation and by omitted demographic factors (normally unobservable on aggregate data). Even when these sources of aggregation bias are corrected for, however, the estimates of the elasticity of intertemporal substitution obtained from aggregate consumption growth depend on expected productivity and income growth. This suggests that entries and exits from the consumption pool, or more generally finite horizons and incomplete markets, might explain some of the difference. In the second half of this paper we investigate theoretical explanations of our empirical results. We develop an overlapping generations model that enables us to evaluate analytically the effects of entries into and exits out of the consumption pool. This results in a negative bias caused by an omitted variable correlated to productivity growth. Such a model could be interpreted as a formalization of a missing market environment. We also discuss some other aggregate problems that can result in a biased estimate of the elasticity of intertemporal substitution.
