Evaluating multidimensional poverty at the individual level in a middle-income country: a comprehensive approach using Peruvian data

Abstract

I develop a comprehensive, policy-relevant and individual-based measure of multidimensional poverty applied to a middle-income Latin American country: the Comprehensive Multidimensional Poverty Index for Peru (C-MPI-P). The C-MPI-P is conceptually grounded in Amartya Sen’s capability approach and is based on the axiomatic counting method of poverty identification and aggregation developed by Alkire and Foster. I estimate the C-MPI-P taking advantage of a well-being specialised survey collected in late 2018 in Peru, which included novel data on individual achievements across 12 well-being dimensions, information on the value that respondents place upon each of these dimensions, and on self-perceived poverty. I use this information to calibrate two key parameters of the C-MPI-P: the dimensional weights (w), and the poverty threshold (k). The C-MPI-P identifies a person as experiencing poverty if she suffers deprivations in approximately four out of 12 weighted dimensions (k=34%). I estimate that in 2018 almost one in four Peruvian people aged 18 or over lived in multidimensional poverty. I also find significant disparities in terms of poverty incidence –proportion of people living in multidimensional poverty–, especially between the rural and urban population, and indigenous and non-indigenous people. My main results are robust to different values for k and to weighting schemes that affect the dimensional structure of the C-MPI-P.