Robustness of Shared Prosperity Estimates: How Different Methodological Choices Matter

Abstract

This paper is the first to systematically test the robustness of shared prosperity estimates to different methodological choices using a sample of countries from all regions in the world. The tests that are conducted include grouped versus microdata, nominal welfare aggregate versus adjustment for spatial price variation, and different treatment of income with negative and zero values. The empirical results reveal an only minimal impact of the proposed tests on shared prosperity estimates. Nevertheless, there are important caveats. First, spatial adjustment can change the ranking of households, affecting the distribution of the population in the bottom 40 percent. Second, the negligible impact of spatial deflation holds only if price adjustments are carried out consistently over time. Finally, the treatment of negative and zero income numbers can potentially lead to substantial differences in shared prosperity, depending on the magnitude of negative income and the share of households with negative and zero numbers across years.

Keywords: Poverty Impact Evaluation, Poverty Monitoring & Analysis, Poverty Assessment, Small Area Estimation Poverty Mapping, Poverty Diagnostics, Poverty Lines