Non-Technical Summary:
Multidimensional poverty measures based on counts of dimensions in which individuals are deprived have gained prominence in recent decades. Poverty measures of this sort are currently used by many governments and international organisations to monitor poverty trends in developed and developing countries.
Defining poverty in this way is a very simple and intuitive approach. Yet when constructing counting poverty measures analysts face multiple methodological choices that can influence poverty levels and comparisons. These choices include the function linking individuals’ level of deprivation with the number of poverty dimensions, the threshold specifying the minimum number of dimensions individuals need to be deprived to be deemed as multidimensionally poor, and the weights assigned to each of the wellbeing indicators.
While the sensitivity of poverty estimates to these choices is generally acknowledged, the common approach involves evaluating the sensitivity of poverty orderings considering a limited and usually arbitrarily set of alternative individual poverty functions, cut-offs values and dimensional weights. Although easy to implement, this approach is inferior to classical approaches used in the income poverty literature.
This paper proposes new dominance criteria for multidimensional counting poverty measures. We derived conditions that are both necessary and sufficient to guarantee the robustness of multidimensional poverty orderings to the choice of the poverty index, the multidimensional poverty cut-off, and the vector of dimensional weights used to construct counting poverty scores. The new conditions are easy to test empirically, and the new criteria apply to a broad class of contemporary counting poverty measures.
We also derived a set of useful necessary conditions that allow the analyst to rule out the robustness of poverty comparisons to changes in poverty functions, identification cut-offs, and dimensional weights. These conditions are easy to implement, as they only require comparing the proportion of people deprived in each of the dimensions and the proportion deprived in all dimensions.
We illustrate our method through an empirical assessment of poverty trends in Australia in the 2000s using a framework based on three indicators of economic deprivation. Our findings indicate that poverty comparisons based on counting measures can be highly sensitive to changes in dimensional weights, cut-offs and poverty functions. Given the growing prominence of this type of measures in social policy and academic debates, it is crucial to have dominance conditions that allow the systematic evaluation of poverty orderings to changes in those methodological choices. This papers constitutes as important step in this direction.
Supplementary files for this LCC Working Paper, including beta versions of the algorithms used in the paper, can be downloaded by clicking here.
An updated version of this paper has been published as Azpitarte, F, Gallegos, J and Yalonetzky, G. (2020) On the robustness of multidimensional counting poverty orderings. Journal of Economic Inequality, 18, 339–364, DOI: 10.1007/s10888-019-09435-5