Consumer demand in EU member states

By combining econometric estimations with the calibration of consumer expenditure data, this study estimates a linear expenditure system (LES) for the 27 EU countries. We use the EU Household Budget Survey (HBS) data for the reference years of 2010 and 2015, which provide data for about 275, 000 and 276,100 EU households, respectively. In addressing the challenge of insufcient price variability in cross-sectional data, we first construct household-specific pseudo unit values, which enable us to derive the estimates of countryspecific Frisch parameters by expenditure decile for all EU countries. The obtained variability of the Frisch parameters across countries and household income groups underscores the importance of properly accounting for and incorporating this heterogeneity in modelling exercises that are based on the parameter in question. Subsequent estimations of the LES model differentiate households by three urbanization degrees of household residence and ten income groups. To take full advantage of consumer microdata, the resulting income elasticities for 23 commodity groups are calibrated within a relative entropy optimization framework to obtain consistent estimates of household-specific income and price elasticities. Finally, the study illustrates the richness of this combined approach in accounting for heterogeneity and nonlinearities in household consumption behavior.