About the
index
Larger Indian companies must spend part of their profits on corporate social responsibility (CSR) and report where the money went. The index sets that reported spending against poverty in each district, to show where need is high and recorded spending is low.
The data and code are published, and the scores can be rebuilt from them.
Spending and need
Companies reported over twenty-one thousand crore rupees of CSR spending in FY2023-24 that can be traced to a named district, according to Ministry of Corporate Affairs filings. Further spending was filed as "Pan India" or named no district, so it cannot be placed. The traceable money is spread very unevenly between states, as the comparison of Bihar and Maharashtra shows.
Poverty and CSR per person
Figure 01Share poor
CSR per person, relative
Share poor
CSR per person, relative
The share of Bihar's people who are multidimensionally poor is more than four times Maharashtra's (33.76% against 7.81% in 2019-21, by NITI Aayog's count), yet Maharashtra receives about 20 times as much CSR spending per person. The index measures the same gap for each district.
What the index does
The CSR Opportunity Index ranks 651 districts by how much poverty they carry against how little CSR spending they receive. Each district gets a score from 0 to 100, the Philanthropic Opportunity Score (POS), and a higher score means more poverty relative to recorded spending. The index also flags 44 whitespace districts, where CSR spending per person is in the bottom 25% of all districts and the share of people in poverty is in the top 25%. On the index page you can change the weights behind the score or recompute it for a single CSR sector, and download a research brief on any district.
How the data is joined
NITI Aayog MPI
Poverty data for 653 districts
MCA CSR Data
Ten fiscal years of spending by district, via Dataful
Census 2011
District population
Joined and scored
651 districts with a POS
We matched districts across the three datasets by name, allowing for differences in spelling where names did not agree exactly. Districts are then split into three equal groups by population, and each district's spending is compared with the median of its own group, so that small districts are measured against other small districts.
What the score measures
Poverty
The share of the district's people who were multidimensionally poor in 2019-21, which NITI Aayog calls the headcount ratio.
Funding gap
How far the district's CSR spending per person falls below the median for districts of similar population. A district at or above that median has a gap of zero.
Persistence
The 2019-21 headcount divided by the 2015-16 one, which gives the share of the earlier poverty still present. A district where poverty has not fallen scores 1 or more.
POS = (0.40 × N̂ + 0.40 × Ĝ + 0.20 × Û) × 100
Each measure is rescaled to run from 0 to 1 across all 651 districts (min-max normalisation, marked by the hat over each letter), then weighted and added. The letters stand for need, gap and unresolved poverty. For districts with no 2015-16 figure, persistence is set to the national median. The weights have default values, and you can change them on the index page.
Who it is for
Programme officers
They can compare districts before deciding where a new grant round should go.
CSR heads
They can show a board how the company's district spending lines up with poverty.
Philanthropy advisers
They can set a client's giving against need in the districts it reaches.
Researchers
They can study how CSR spending is spread across districts and states, using the published data.
Key numbers
Sources
NITI Aayog, National Multidimensional Poverty Index 2023, based on NFHS-5 (2019-21) and computed with the Alkire-Foster method of the Oxford Poverty and Human Development Initiative (OPHI) at the University of Oxford.
CSR filings to the Ministry of Corporate Affairs, Government of India, as compiled by Dataful.in (dataset ID 1612), with ten fiscal years of district-level spending.
Census of India 2011, Office of the Registrar General of India. District populations are the denominator of CSR per person and set each district's population group.
Questions and corrections
Souhard Sharma built and maintains the index. You can reach him on LinkedIn about the method or the data, or about using the index in a foundation's work.