We mustn’t have any blind spots in our official estimations of poverty

livemint.com:

As against the money-metric approach used earlier, the Niti Aayog prefers to use poverty estimates based on the UNDP suggested multi-dimensional poverty index (MPI). Unlike the former that was based on consumption expenditure, the MPI uses a number of indicators on different aspects of well-being. The national MPI is based on data from the National Family Health Survey (NFHS), of which the most recent one available is for 2019-21 (and there are previous surveys for 2005-06 and 2015-16).

In July 2023, the Niti Aayog released the National MPI based on 2019-21 NFHS data. Recently, a discussion paper released by it on multi-dimensional poverty claims that 248 million individuals moved out of it in the nine years from 2013-14 to 2022-23, almost all of them under the present government. This is almost double the 135 million decline between 2015-16 and 2019-21 revealed in the July 2023 report.

The difference owes largely to the extrapolation done by the paper’s authors, Niti Aayog member Ramesh Chand and senior advisor Yogesh Suri, using original estimates. While extrapolation for monetary variables is generally acceptable, it may not be a good idea for measuring MPI-based poverty, given the sensitivity of these indicators to overall economic data, more so with the last two years of that period affected by covid. The issue is not with the methodology of extrapolation or estimates of poverty, but which dimensions contributed to improvements in well-being and which indicators need more effort.