Multidimensional poverty measurement : making the identification of the poor count?

6th Meeting of the Society for the Study of Economic Inequality (ECINEQ), celebrat a Luxemburg, 13-15 de juliol de 2015The success of any poverty eradication program crucially depends on its ability to identify who is poor and who is not. In this paper, we show that the state-of-the-art methodology that is used to identify the poor in multidimensional contexts -the dual cutoff method suggested by Alkire and Foster- is insensitive to many of the subtle considerations that should be incorporated when making such delicate decisions. The simplicity of the counting approach that underlies the dual cutoff method precludes the possibility of generating "poor-identification rules" that are sensitive to interactions between the different dimensions of poverty. To go beyond the apples-and-oranges aggregation procedures characterizing the dual cutoff method, we suggest a much broader identification approach that contains the latter as a particular case. Our empirical findings using 48 Demographic and Health Surveys across the developing world suggest that the percentage of households that are inconsistenly identified as "poor" according to the dual cutoff and some of the methods suggested in this paper is around 30% -a result with enormous implications for the identification of the potential beneficiaries of poverty eradication programs worldwide