An analyst using household survey datato construct a welfare metric is often confronted withonfronted with An analyst using household survey data toconstruct a welfare metric is often confronted with a numberof theoretical and practical problems. What componentsshould be included in the overall welfare measure? Shoulddifferences in tastes be taken into account when makingcomparisons across people and households? How best shoulddifferences in cost-of-living and household composition betaken into consideration? Starting with a brief review ofthe theoretical framework underpinning typical welfareanalysis undertaken based on household survey data, thispaper provides some practical guidelines and advice on howbest to tackle such problems. It outlines a three-partprocedure for constructing a consumption-based measure ofindividual welfare: 1) aggregation of different componentsof household consumption to construct a nominal consumptionaggregate; 2) construction of price indices to adjust fordifferences in prices faced by households; and 3) adjustmentof the real consumption aggregate for differneces inhousehold composition. Examples based on survey data froeight countries--Ghana, Vietnam, Nepal, the Kyrgyz Republic,Ecuador, South Africa, Panama, and Brazil--are used toillustrate the various steps involved in constructing thewelfare measure, and the STATA programs used for thispurpose are provided in the appendix. The paper alsoincludes examples of some analytic techniques used toexamine the robustness of the estimated welfare measure tounderlying assumptions.
PDF
download
878987797
028-85220240
