Multidimensional Poverty Rankings based on Pareto Principle: A Practical Extension

This paper proposes a ranking method of multidimensional poverty and extends it aiming to enhance its practical utility. While our original ranking method that assumes non-comparability among different dimensions of poverty succeeds in eliminating some implicit arbitrariness in existing ranking, it also confronts a disadvantage that a non- negligible number of objectives (countries) are ranked at the same level. In order to improve this disadvantage, we propose an extended ranking method, where we allow the data to have a certain range of bandwidth. The introduction of bandwidth improves the usefulness of our ranking in the sense that it decreases the number of countries with the same rank. In addition, a simulation exercise shows that this extension also improves the robustness of the ranking against measurement errors.