Please use this identifier to cite or link to this item:
Christensen, Mette
Year of Publication: 
Series/Report no.: 
IFS Working Papers, Institute for Fiscal Studies (IFS) 07/16
All micro studies of demand are based on using time series cross sectional data. Because in such data each household is only observed once, it is only under strong identifying restrictions that one can interpret the coefficients on consumer behavior. For example, if tastes are correlated with income, the usual estimates of income elasticities from cross sectional data are biased. In contrast, panel data allows identification of the coefficients on consumer behavior in the presence of unobservable correlated heterogeneity. In this paper we make use of a unique Spanish panel data set on household expenditures to test whether unobservable heterogeneity in household demands (taste) is correlated with total expenditures (income). We find that tastes are indeed correlated with income for half of the goods considered.
Engel curves
unobservable heterogeneity
panel data
Persistent Identifier of the first edition: 
Document Type: 
Working Paper

Files in This Item:
307.59 kB

Items in EconStor are protected by copyright, with all rights reserved, unless otherwise indicated.