Applying a Multidimensional Poverty Index for targeted interventions in Kolkata’s slums

In early 2019, Calcutta Rescue – a medium-sized non-governmental organisation (NGO) working in Kolkata (formerly Calcutta) slums - developed its own Multidimensional Poverty Index (CR-MPI). This index captures poverty in multiple dimensions beyond merely income.

The CR-MPI was built on the methodological foundations of the Global Multidimensional Poverty Index and my coauthored study on Indian slums. Based on my long-standing expertise in multidimensional poverty measurement and analysis, I guided Calcutta Rescue in developing the CP-MPI and coauthored the recent report.

The CR-MPI was tailored specifically to the realities of the slums where Calcutta Rescue provides services (e.g. street medicines, mobile health units, and skills training programmes) and to the NGO’s operational priorities. The objective was to develop a practical tool for identifying the poorest slums and deprivations for better targeting, guiding interventions, monitoring changes over time and demonstrating impact to partners and donors. The first CR-MPI survey was conducted during 2019 across 23 slums, and key results from the report were later externally published.

One of the slums where Calcutta Rescue provided services during 2019-2024

The CR-MPI measures poverty using three dimensions — health, education and standard of living — comprising 17 indicators. These three dimensions are weighted equally, and the weight of each dimension is distributed equally across indicators. Households are identified as multidimensionally poor when they are deprived in at least one-third of the weighted indicators.

The CR-MPI is a product of two components: the incidence of poverty and the intensity of poverty. By combining the incidence with the intensity of poverty, CR-MPI offers a nuanced picture of both how prevalent poverty is and how intense it is among poor households.

Using data to better target poverty interventions

The CR-MPI has helped Calcutta Rescue to make more informed decisions about where support is needed most. Rather than allocating resources uniformly, the NGO was able to target specific slums and specific deprivations with greater precision, which has strengthened its ability to deliver evidence-based programmes, use donor resources efficiently and communicate results with clarity.

Following the success of its first report, Calcutta Rescue committed to repeating the CR-MPI measurement every five years. In 2024, the NGO conducted its second survey, covering more than 1,100 households across 25 slums. Of these, 19 slums had also been surveyed in 2019, enabling direct comparisons over time.

The findings from the 2024 survey are significant:

• Across the 19 slums that were comparable over time, the proportion of residents experiencing multidimensional poverty fell from 70.3% in 2019 to 53.5% in 2024 — a reduction of 16.8 percentage points.
• In 2019, 13 of these slums were classified as experiencing very high poverty, but by 2024, the number had fallen significantly.
• Substantial improvements were recorded in 15 of the 19 slums, with some of the largest reductions occurring in areas that had been among the poorest in 2019.

At the indicator level, the greatest progress was seen in latrine facilities, living security, antenatal care, disease knowledge and cooking fuel. A particularly strong example of how the CR-MPI has informed action can be seen in the slum of Bagbazar (see the briefing). Identified as a slum facing very high poverty in the first report, Bagbazar showed high deprivations in sanitation, household materials, living security, disease knowledge and school attendance. Calcutta Rescue was already active in parts of the slum through mobile medical services and its education centre, but the CR-MPI findings in the first report enabled the organisation to identify gaps more clearly and expand its reach into additional parts of the slum.

Interviewers during the 2024 CR-MPI survey in one of the slums

Ongoing challenges in reducing multidimensional poverty

Challenges remain. Of the 25 slums that were covered by the 2024 survey, eight slums appear to be in very high poverty, where more than 70% of residents can be recognised as living in multidimensionally poverty, and nine slums are in moderately high poverty.

Of the 17 indicators, overall deprivation rates are low (i.e. less than 10%) for child labour, vaccination, antenatal care, whereas deprivation rates are high (i.e. 50% or higher) in living security, cooking fuel, hygiene knowledge, disease knowledge and housing material. Future interventions should focus on reducing deprivations in these indicators focusing on the poorest slums.

Turning data into impact: lessons for NGOs

For organisations working to reduce poverty, effective action depends on appropriate measurement and understanding. Poverty has many faces and cannot be fully captured by income alone. Poverty is also shaped by poor health, inadequate education, insecure housing, unsafe sanitation and limited access to essential services. Calcutta Rescue’s experience, which is relatively rare among NGOs, demonstrates the practical value of multidimensional poverty measurement. By generating robust, locally relevant evidence, the NGO has been able to better target support and adapt its programming.

More broadly, Calcutta Rescue’s novel attempt during 2019-2024 offers an important example of how medium-sized NGOs can use data not simply for reporting, but as a strategic tool for evidence-based management and resource allocation and for delivering more effective and accountable interventions.

Needless to say, involving the target population is essential for ensuring that interventions reflect community priorities, improve relevance, strengthen ownership, and support empowerment.

Read the report and the policy brief.