A new fuzzy unidimensional poverty index from an information theory perspective

Abstract

This article proposes to use an information theory approach to design a fuzzy unidimensional poverty index. In order to avoid the usual binary definition of poverty, a fuzzy set approach has been used. The total population is partitioned into three mutually exclusive groups around the poverty line. This builds on fuzzy set theory whereby the definition of the threshold of who is poor or non-poor is fuzzy. This article proposes a method using a membership function and a confidence interval of poverty line to identify to what extent households can be considered as poor or non-poor. According to this membership function, a relative entropy measure is adopted to assess an aggregation method of fuzzy individual poverty. An application using individual well-being data from Tunisian households is presented.