How the Multidimensional Poverty Index works
Published 6 November 2026
NITI Aayog’s headline is that 24.82 crore people escaped multidimensional poverty between 2013-14 and 2022-23. The index behind that headline does not look at income at all. It looks at twelve things a household may or may not have, from nutrition to a bank account. Here is how it is built, and why a bank account counts.
Twelve indicators, three dimensions
The National Multidimensional Poverty Index (MPI) measures deprivation across three dimensions, each given equal weight of one third: health, education, and standard of living. Twelve indicators sit inside those dimensions.
Health has three indicators. Nutrition carries a weight of one sixth. Child and adolescent mortality carries one twelfth. Maternal health, measured as antenatal care, carries one twelfth. Together they add up to one third.
Education has two indicators. Years of schooling carries one sixth. School attendance carries one sixth. Together they add up to one third.
Standard of living has seven indicators, each carrying one twenty-first: cooking fuel, sanitation, drinking water, electricity, housing, assets, and bank account. Seven times one twenty-first is one third.
A person is counted as multidimensionally poor if they are deprived in one third or more of the weighted indicators. The index itself is the product of two numbers: the headcount, which is the share of the population that is poor, and the intensity, which is the average share of indicators in which poor people are deprived. The data comes from the National Family Health Survey.
India’s national MPI has twelve indicators. The global MPI, built by the Oxford Poverty and Human Development Initiative and the United Nations Development Programme, has ten. India added two: antenatal care and bank account, to reflect national priorities.
The headline numbers come from NITI Aayog’s discussion paper of January 2024. The multidimensional poverty headcount fell from 29.17 percent in 2013-14 to 11.28 percent in 2022-23, and about 24.82 crore people moved out of multidimensional poverty in that period.
Why a bank account counts
Of the twelve indicators, the bank account is the one people ask about most. It is also the one that moved the most.
A household is deprived on this indicator if no member of the household has a bank account or a post office account. The reasoning is that a bank account is a gateway. It is how a household receives government transfers directly, saves safely, and borrows from a formal lender instead of a moneylender. In a country where welfare is increasingly paid by direct transfer, having no account means being cut off from the way the state delivers help.
That is also why the indicator moved so fast. The share of Indians living in a household with no bank account fell from 58.11 percent in 2005-06 to 9.66 percent in 2015-16, and under 4 percent by 2019-21, largely because of the Jan Dhan Yojana, which opened hundreds of millions of accounts. When that many households cross from “no account” to “account”, the MPI headcount falls, because the bank account indicator is one of the twelve and it flipped for a very large number of people at once.
Critics say this drove too much of the improvement. A 2024 investigation by the Reporters’ Collective argued that the bank-account indicator did much of the work in the MPI’s decline, and questioned whether opening an account should count as heavily as, say, nutrition or schooling. It is a fair question. An account is a means to other things, not an end in itself, and an account with no money in it does not feed a child.
The honest answer is that the bank account belongs in the index because financial access is a real deprivation, and it moved the index because Jan Dhan was a very large programme. Both things are true at once. The MPI is not a trick, but its headline is sensitive to this one indicator in a way that is worth knowing.
What the MPI does and does not do
The MPI measures deprivation, not income. A household can be above the income poverty line and still be multidimensionally poor, for instance if it has no sanitation or no electricity. And a household can be below the income line but not multidimensionally poor, if it has schooling, clean water, and a solid house.
That is the point of it. Income is one dimension of poverty, and the MPI is an attempt to count the others. It is a complement to the income measure, not a replacement for it.
It has limits. It counts whether a household is deprived on each indicator, not how deprived. A household without any schooling and a household where one child missed a year both count as deprived on schooling, though their situations differ. It also depends on the survey: the MPI is only as fresh as the latest National Family Health Survey, and it cannot be updated every year the way consumption data can.
Read it for what it is: a count of simultaneous deprivations, built from twelve concrete things, that tells you whether the basics of a decent life are reaching people at the same time. Read alongside the income numbers, it gives a fuller picture than either measure alone.
Sources
- NITI Aayog, “Multidimensional Poverty in India since 2005-06” (discussion paper, January 2024). 24.82 crore people escaped multidimensional poverty between 2013-14 and 2022-23; the headcount ratio fell from 29.17 percent to 11.28 percent.
- NITI Aayog, National Multidimensional Poverty Index: A Progress Review 2023. Multidimensional poverty fell from 24.85 percent in 2015-16 to 14.96 percent in 2019-21; the 12 indicators and their weights across the three dimensions.
- The Reporters’ Collective, “Modi Govt’s Grand Illusion of Financial Inclusion and Poverty Reduction” (December 2024). The bank-account indicator drove much of the MPI improvement; the share of Indians in households without a bank account fell from 58.11 percent in 2005-06 to 9.66 percent in 2015-16, and under 4 percent by 2019-21, largely because of Jan Dhan.
These sources are also listed on the Sources page.
Updated 9 November 2026