What are the main limitations of the MPI?
The MPI has some drawbacks, due mainly to data constraints. First, the indicators may not reflect capabilities but instead reflect outputs (such as years of schooling) or inputs (such as cooking fuel). Second, the health data are relatively weak and overlook some groups’ deprivations, especially for nutrition, though the patterns that emerge are plausible and familiar. Third, in some cases careful judgments were needed to address missing data. But to be considered multidimensionally poor, households must be deprived in at least six standard of living indicators or in three standard of living indicators and one health or education indicator, or in two health or education indicators. This requirement makes the MPI less sensitive to minor inaccuracies. Fourth, intra-household inequalities may be severe, but these could not be reflected. Fifth, while the MPI goes well beyond a headcount ratio to include the intensity of poverty, it does not measure inequality among the poor, although decompositions by groups can be used to reveal group-based inequalities. Finally, the estimates presented here are based on publicly available data and cover various years between 2006 and 2016-17, which limits direct cross-country comparability.