Unpacking the MPI : A Decomposition Approach of Changes in Multidimensional Poverty Headcounts

Pé ; rez Pé ; rez, Jorge Eduardo ; Rodriguez Castelan, Carlos ; Trujillo, Jose Daniel ; Valderrama, Daniel

World Bank, Washington, DC

关键词: WASTE;    UNEMPLOYMENT;    EMPLOYMENT;    EQUITY;    HOUSEHOLD SIZE;   

DOI  :  10.1596/1813-9450-7514
RP-ID  :  WPS7514

学科分类：社会科学、人文和艺术（综合）

来源: World Bank Open Knowledge Repository

【 摘 要 】

Multidimensional measures of povertyhave become standard as complementary indicators of povertyin many countries. Multidimensional poverty calculationstypically comprise three indices: the multidimensionalheadcount, the average deprivation share among the poor, andthe adjusted headcount ratio. While several decompositionmethodologies are available for the last index, lessattention has been paid to decomposing the multidimensionalheadcount, despite the attention it receives from policymakers. This paper proposes an application of existingmethodologies that decompose welfare aggregates--based oncounterfactual simulations--to break up the changes of themultidimensional poverty headcount into the variationattributed to each of its dimensions. This paper examinesthe potential issues of using counterfactual simulations inthis framework, proposes approaches to assess these issuesin real applications, and suggests a methodology based onrank preservation within strata, which performs positivelyin simulations. The methodology is applied in the context ofthe recent reduction of multidimensional poverty inColombia, finding that the dimensions associated witheducation and health are the main drivers behind the poverty decline.