OPHI Director Sabina Alkire has written a blog describing the key findings of OPHI’s analyses of the global Multidimensional Poverty Index (MPI) data for 2013. Published on Oxford University’s Department of International Development blog, ‘Debating Development’, the article summarises three studies conducted by OPHI.
The first, How Multidimensional Poverty went down, used data from 22 countries in every geographical region of the developing world to analyse how MPI poverty has changed over time. Alkire and her co-author, José Manuel Roche, found that the greatest absolute reductions had occurred in Nepal, Rwanda and Bangladesh, followed by Ghana, Tanzania, Bolivia and Cambodia. Relative to its initial (much lower) level of poverty, Colombia also reduced poverty strongly, as did Armenia.
The study also shows the importance of using the MPI, and not just the percentage of poor people, as a measure of poverty.
‘Why? Poorer countries like Ethiopia, Senegal and Malawi reduced the average intensity of poverty strongly; in other words, poor people were deprived in fewer things than they had been before, even if they had not graduated out of poverty,’ writes Alkire. ‘The MPI captures and celebrates that progress, which is so important at a human level.’
The blog also describes work by Alkire and Suman Seth on Multidimensional Poverty Reduction in India 1999-2006. This study looks at how poverty changed across caste and religious groups, by different household characteristics and different states. It found, worryingly, that over 46% of people who were multidimensionally poor in 1999, and 40% in 2006, were actually not just poor but destitute.
A third study, by Alkire, Roche and Seth, identified the world’s poorest billion people using different disaggregations of the MPI 2013 data. The study shows in black and white the importance of going beyond national averages to find where and how people are poor; something it is very easy to do with the MPI, because it measures deprivations directly for each person, and for 663 subnational regions.