Inequality among the Multidimensionally Poor in over 100 Countries

This joint seminar with the Oxford Poverty and Human Development Initiative (OPHI) and the United Nations Development Program (UNDP) Human Development Report Office (HDRO) virtual event took place on Monday, March 1st, 2021. "Inequality among the Multidimensionally Poor in over 100 Countries", featured Maria Emma Santos of OPHI, with Heriberto Tapia (UNDP) and Hector Moreno (OPHI) providing discussant remarks. The discussion was moderated by OPHI Director Sabina Alkire and IIEP Co-Director James Foster.

Maria Emma Santos addressed how analyzing inequality among the poor is key to ensure that the poorest poor are not left behind, which is often the case since they are at the crossroads of marginalized groups. It is very difficult for policies -even at sub-national levels- to actually and effectively reach them. In her paper she examines inequality within over 100 countries among the multidimensionally poor, as measured by the Global Multidimensional Poverty Index (G-MPI). She compares two approaches proposed so far for incorporating inequality into the measurement of multidimensional poverty. First, she looks at the ‘assimilated approach’, by which the poverty measure incorporates sensitivity to inequality among the poor, as it is the case of the MGamma class of poverty measures proposed by Alkire and Foster (2016, 2019); this uses a relative inequality measure. The other method is the ‘complementary approach’, by which the poverty measure is complemented alongside the variance of deprivation scores among the poor, an absolute inequality measure.

Prof. Santos finds that country rankings by absolute vs. relative inequality among the poor differ quite substantially, suggesting that the selection of one or the other type of inequality matters when only that aspect of poverty is evaluated. However, the country ranking by the G-MPI, which considers poverty incidence and intensity, is highly robust to the incorporation of inequality into measurement of poverty, either using the MGamma2 measure or complementing the G-MPI with the variance among the poor. In other words: bad things seem to go together. Countries with a higher proportion of their population in multidimensional poverty tend to have higher average poverty intensity, and such higher average intensity tends to be more unequally distributed among the poor. This does not mean that it does not matter to know and measure inequality among the poor. A high inequality among the poor signals the need to develop different kinds of policies according to different poverty intensities. Our understanding is that it is the distribution of the deprivation scores alongside the dimensional decomposition that can be more illuminating for designing effective policies to leave no-one behind.

These seminars are organised jointly with the Oxford Poverty and Human Development Initiative at Oxford University and the UNDP Human Development Report Office. They are hosted by IIEP Co-Director James Foster, GWU.

[LalitSharma-fcvr]

how to calculate Inequality among the Multidimensionally Poor in Stata software