Multidimensional Poverty Index with Dependence-Based Weights

An important aspect of the multidimensional perception of poverty phenomenon is the dependence among the underlying indicators. However, the commonly applied approaches to multidimensional poverty assessment do not capture this interdependence. In this paper we propose a new multidimensional poverty index accounting for the dependence and innovate over the weighting approach. The weighting method proposed here incorporates the copula-based rank dependence among well-being dimensions and contains necessary normative parameters. In particular, the latter includes the elasticity of substitution among dimensions and the belief-adjusting parameter, which specifies the direction of relation between the dependence and the weights. The results of poverty evaluation in the selected European countries suggest that multidimensional poverty is driven not only by the individual shortfalls, but also by the degree of interdependence among well-being indicators. Moreover, multidimensional poverty is relatively higher, if dimensional weights are in direct proportion to the dependence compared to the cases of inverse relation and equal weighting. Considering the novel weighting approach, this paper contributes to the literature on composite indicators by suggesting a channel to enclose the dependence structure in the multidimensional poverty index.