Linearizing the inverse quadratic almost ideal demand system

This article investigates the linear approximation to the inverse quadratic almost ideal demand system (IQUAIDS), a recently introduced flexible functional form with potential usefulness. Linearizing this nonlinear model is of practical importance because nonstationary data, which are likely to be used in inverse demand systems, can be handled more properly in linear models. The IQUAIDS is linearized by replacing the two aggregator functions with a number of alternative quantity indices and new composite variables. The results of an application to Japanese fresh food demand suggest that the IQUAIDS can be linearized with adequate accuracy in terms of elasticities.