Poor households or poor individuals? The issue of the unit of identification when designing an MPI
Maria Emma Santos, Pablo Villatoro and Xavier Mancero
“The statistic for global poverty is typically stated as x million people living in poverty, but the correct statement about the figures (…) is that there are x million people living in households that are in poverty” (Atkinson, 2019, p.73). Selecting the unit of identification in poverty measurement is a much debated fundamental step, in which most frequently the desire for individual-identification meets the reality of data constraints. This paper explores in depth the possibilities and limitations for an individual-based multidimensional poverty measure using regular household surveys of Latin America. We perform a series of empirical exercises comparing a household-based Multidimensional Poverty Index (MPI), with a set of individual-based MPIs, using in all cases the Alkire and Foster (2011) methodology. The sequence of exercises exposes a number of important tensions in the decision of the unit of analysis. First, a series of methodological constraints needs to be met for a set of individual-based MPIs to be aggregated into an overall poverty measure, representative of the whole population. However, those same constraints represent a limitation for each individual-based measure to be representative of each particular population group. Second, we highlight that household-based poverty measures do not necessarily obscure intra-household deprivations (i.e. they can respect a rights-based approach), but they can obscure intra-household inequalities. Third, while under the refereed certain methodological constraints household-based measures have no exclusion discrepancy with individual-based measures, they maximize inclusion discrepancy, which can be understood as an overstatement of intra-household externalities. However, at the same time, individual-based measures naturally lead to a somehow uncomfortable condition of ‘partially poor’ households, which restates the question of to what extent is wellbeing exclusively an individual characteristic with no presence of external effects. This opens up a third – hybrid – possibility in terms of the unit of identification in which, while considering primarily the individual, it allows considering some external effects in the identification of poverty for certain groups.
Maria Emma Santos is an Assistant Professor at Dept. of Economics at Universidad Nacional del Sur (UNS) and a CONICET Research Fellow at the Instituto de Investigaciones Económicas y Sociales del Sur (IIESS), Bahia Blanca, Argentina. She is also a Research Associate to the Centro de Estudios para el Desarrollo Humano (CEDH) of Universidad de San Andres in Argentina, and to the Oxford Poverty and Human Development Initiative, at the University of Oxford (OPHI), UK. Her research interests include the measurement and analysis of multidimensional and chronic poverty. She is particularly interested in Latin American countries.
This event will be held on Zoom (registration: bit.ly/2GPQbUA)