Score contribution per author:
α: calibrated so average coauthorship-adjusted count equals average raw count
Distinguishing household groups when estimating the linear expenditure system (LES) has substantial advantages. Using an Australian household survey to estimate the LES, we show that allowing for household heterogeneity improves the model's predictions and its economic properties. Computable general equilibrium (CGE) model simulations reveal that allowing for heterogeneity of consumers opens up a new possible driver of long-term change in the economy's composition. The simplicity and parsimony of the LES account for its popularity, particularly in CGE models. The results underscore the practical importance of the disaggregated approach.