Tuesday, November 18, 2003
Presenter: Johan A. Mistiaen, DECRG, World Bank
Johan Mistiaen presented a new technique for estimating first-order approximations of policy-induced distributional impacts at exceptionally detailed levels of geographic and/or socio-economic disaggregation. The gainers and losers of many contemporary policy reforms-both macro, sectoral and programmatic-are often heterogeneously distributed across space and/or concentrated within population sub-groups. Thus, a dissagregated "profile" of ex-ante welfare impacts can be used to improve the design of interventions and/or mitigate adverse effects.
Welfare analysis relies on the detailed income and/or expenditure information collected via household survey data. The scope for disaggregation is therefore limited by survey sample size. This data constraint can be circumvented by applying small area estimation techniques that enable combining the detailed welfare estimation from household surveys with the complete coverage provided by population census data.
The approach was tested using data from Madagascar to simulate the first-order household-level welfare effects of rice food price changes induced by eliminating rice import tariffs. The results indicate there is considerable heterogeneity in welfare impacts across geographic areas, even within and among relatively small administrative units both in rural and urban neighborhoods.
Back to Training Events and Materials