Ely Lectures

Draft for the Overview Lecture on 7th March There are three key objectives of these lectures: First, to provide a powerful test of integrability conditions on individual household data without the need for para-metric models of consumer behavior. Second, to provide tight bounds on welfare costs of relative price and tax changes. Third, to provide tight bounds on demand responses. The aim is to allow price response parameters to vary across points in the income distribution. The idea is to fully exploit micro data on consumer expen-ditures and incomes across a finite set of discrete relative price or tax regimes. This is achieved by combining the theory of revealed preference with the seimparamet-ric estimation of consumer expansion paths (Engel curves). Recent developments on semiparametric estimation of shape invariant models with endogenous regres-sors are exploited. Empirical examples are given from the analysis of cost of living indeces and from tax policy reform