Aggregation and consumer behaviour: some recent results

Abstract

The availability of household-level data covering long periods of time makes it relatively easy to assess likely sources of aggregation bias. In this paper, we illustrate a promising methodology, which requires computing aggregation factors across households. If these factors are stable over time, aggregate data can be used to estimate micro-parameters.

Our application covers consumer demand equations, and dynamic consumption relations, but the methodology could be applied in a variety of different contexts.