Poverty is a paradoxical state. Recognizable in the field for any sensitive observer who travels in remote rural areas and urban slums and meets marginalized people in a given society, poverty still remains a challenge to conceptual formalization and to measurement that is consistent with such formalization. The analysis of poverty is multidisciplinary. It goes from ethics to economics, from political science to human biology, and any type of measurement rests on mathematics. Moreover, poverty is multifaceted according to the types of deprivation, and it is also gender and age specific. A vector of variables is required, which raises a substantial problem for individual and group comparisons necessary to equity analysis. Multidimensionality also complicates the aggregation necessary to perform the efficiency analysis of policies. In the case of income poverty, these two problems, equity and efficiency, have benefited from very significant progress in the field of economics. Similar achievements are still to come in the area of multidimensional poverty.
Within this general background, this book has a very modest and narrow-scoped objective. It proposes an operational methodology for measuring multidimensional poverty, independent from the conceptual origin, the size and the qualitative as well as the quantitative nature of the primary indicators used to describe the poverty of an individual, a household or a sociodemographic entity. It is my view that the proposed methodology should allow to integrate into the analysis of multidimensional poverty the sets of techniques already available or forthcoming in the area of income poverty. Despite this, I do not want to avoid the issue of the conceptual foundations of poverty. Thus, I propose from the start a quite comprehensive definition of poverty, whose ethical basis is briefly presented in an annex which can be skipped by readers not interested in such issues. The core of the methodology rests on a solution to the issue of the aggregation across the multiple subdimensions of poverty. The rationale of the proposed solution consists in exploring the internal structure of association between these subdimensions of poverty.
The methodology aims to be operational, by which I mean that it should be feasible with the use of computational tools that are easily accessible, as well as feasible without any specific limitation on the number and the nature of the poverty indicators used by the analyst. I also sincerely believe that the conceptual debates on the dimensions of poverty and on the measurement methodologies need to be lit with numerous empirical studies showing the strengths and weaknesses of different . . .