It is shown how one can effectively use cross-section data in modelling the change over time in aggregate consumption expenditure of a heterogeneous population. The starting point of our aggregation analysis is a dynamic behavioral relation on the household level. Based on certain hypotheses on...