When Welfare Works but Poverty Persists: India's Structural Divide

In the last decade, India has come a long way in alleviating poverty and improving human development, though stark disparities still persist across states and regions. The story of poverty is truly remarkable: the nation has lifted hundreds of millions of people out of multidimensional deprivation over the past twenty years. But aggregate figures can be misleading. As soon as one disaggregates the same official data to the state level, a different — and far more uncomfortable — story emerges.

Three government data sources, when read together, reveal a consistent pattern. The first is NITI Aayog's National Multidimensional Poverty Index (MPI), based on NFHS-5 (2019-21) survey data. The second is the SDG India Index (2023-24). The third and the most recent is the Ministry of Food & Public Distribution's Food Grain Bulletin for January 2026, which contains state-wise data on food entitlement dependence under the National Food Security Act. Together, they do not merely describe poverty; they also reveal its structural patterns.

The NFHS-5 survey puts Bihar's MPI headcount at 33.8% — meaning roughly one in three people in the state were multidimensionally poor in 2019-21. At the other extreme, Goa, Kerala, and Puducherry record MPI headcounts below 1%. On the SDG composite, Bihar scores 57, while Kerala and Uttarakhand score 79. The 22-point SDG gap and the 33-percentage-point MPI gap, taken together, suggest that these states are not merely slow and fast runners in the same race. They are operating on fundamentally different development trajectories.

The table below presents NFHS-5 actual survey data (2019-21), NITI Aayog's 2022-23 projected headcounts, and the SDG India Index 2023-24 composite score for every state and union territory.

Sources: MPI NFHS-5 headcount ratios from NITI Aayog National MPI Progress Review 2023. 2022-23 projections from NITI Aayog Discussion Paper (January 2024). SDG composite scores from SDG India Index 2023-24 (PIB). Status classification is editorial, not an official GoI categorisation.

The January 2026 edition of Food Grain Bulletin provides a third and complementary lens on India's poverty geography: it shows not just who is poor, but who is heavily dependent on the state for basic food security.

Under the Pradhan Mantri Garib Kalyan Anna Yojana (PMGKAY), the government provides free food grain to all beneficiaries of the National Food Security Act — 5 kg per person per month for Priority Households and 35 kg per household per month for Antyodaya Anna Yojana families, the poorest of the poor. The scheme has been free since January 2023, for a period of five years, at a budgeted cost of ₹1.63 trillion for 2025-26 alone.

The January 2026 Bulletin reveals a striking geographic concentration of this dependence. The states with the highest NFSA rural coverage rates are almost precisely the states with the highest MPI headcounts — a convergence that suggests a strong correlation. It reflects a structural reality: where market opportunities and state capacity remain limited in their ability to build productive livelihoods and human capital, the PDS becomes the last line of economic support.
 

Source: Ministry of Food & Public Distribution, Food Grain Bulletin January 2026 (Page 33 — Statement indicating State-Wise number of Persons/Families covered under NFSA as on January 2026). Annual allocation from Page 27. Fair Price Shops from Page 48.

The numbers are revealing. Bihar alone accounts for 87.1 million NFSA beneficiaries, or 85.1% of its rural population — the second-largest absolute dependency pool in the country after Uttar Pradesh. Its 51,965 fair price shops are the second-largest network in India. Its annual free grain allocation of over 5.5 million tonnes under NFSA exceeds the allocations of all southern states. In Jharkhand, 86.5% of the rural population is covered by NFSA.

The contrast with the southern states is instructive. Kerala covers 52.6% of its rural population under NFSA, not because it is stingy with entitlements, but because a larger share of its population has sufficient economic security to not require them. Tamil Nadu, despite being a high-coverage state by design (it runs a universal PDS), does so from a base of dramatically lower poverty. The intent of coverage in Tamil Nadu is universal provisioning; the necessity of coverage in Bihar is economic survival.

Three economic frameworks converge on the same diagnosis for Bihar, Jharkhand, Meghalaya, Assam, Uttar Pradesh, and Madhya Pradesh — and none suggest that transfers alone are sufficient to address the problem.

Poverty traps (Azariadis & Drazen, 1990; Banerjee & Duflo, 2011):  When households cannot invest adequately in nutrition, health or education, deprivation becomes self-reinforcing across generations. Bihar's NFHS-5 MPI data shows the classic pattern: bank accounts and electricity have improved, while nutrition, housing quality and clean cooking fuel remain constrained, limiting human capital accumulation.

Institutional quality (Acemoglu, Johnson & Robinson, 2001; North, 1990):  The SDG India Index reflects, in part, the effectiveness of state capacity. Bihar's score of 57 and Jharkhand's of 62 indicate not just poverty but also the limited ability of state systems to translate financial resources into delivered outcomes. This helps explain why high NFSA offtake does not translate into proportionate reductions in multidimensional poverty.

Economic geography and structural transformation (Krugman, 1991; Kaldor, 1966):  Industrial and services growth has been spatially concentrated, particularly in the South and West. Agglomeration effects reinforce this pattern, making lagging regions difficult to revive. The NFSA coverage map is, in this sense, a near mirror image of India's structural transformation map.

Policy Implication
The combined reading of NFHS-5 MPI, SDG India Index 2023-24, and the January 2026 Food Grain Bulletin points to a clear implication: the relevant unit of analysis for India's remaining poverty challenge is a cluster of states that are simultaneously high on MPI headcount, low on SDG scores, and deeply dependent on food entitlements. Bihar, Jharkhand, Meghalaya, Uttar Pradesh, Madhya Pradesh, and Assam form this cluster.

Their problem is not insufficient transfers. It is insufficient structural transformation — and that requires a qualitatively different, multi-decade policy toolkit: school quality reform, primary healthcare system strengthening, industrial cluster development, and agricultural value-chain investment of the kind applied by South Korea in the 1960s-70s and China in the 1990s-2000s.

The data to make this diagnosis at the district level already exists across three ministries. The question is whether policymakers will move beyond all-India averages and treat India's poorest states not as laggards on a single national curve, but as distinct structural problems, each demanding its own long-duration prescription.

Data & Methodology
MPI headcount ratios: NFHS-5 (2019-21) actual survey estimates from NITI Aayog National MPI Progress Review 2023 and the NITI Aayog discussion paper 'Multidimensional Poverty in India since 2005-06' (January 2024). 2022-23 projections use compound rates of decline from NFHS-4 to NFHS-5 and are projections, not survey estimates. SDG composite scores from SDG India Index 2023-24 (PIB), covering 113 indicators across 16 SDGs (MoSPI National Indicator Framework). NFSA coverage, beneficiary counts, allocation data, and Fair Price Shop numbers from the Ministry of Food & Public Distribution, Food Grain Bulletin January 2026 (Pages 27, 33, 48, 49) — the most current official data available as of March 2026. PMGKAY subsidy figures from the same bulletin (Page 45). The 'Status' classification (Strong/Middle/Laggard) is an editorial synthesis and does not represent an official GoI categorisation. Regional groupings are geographic, not official GoI statistical zones. Economic frameworks cited: Azariadis & Drazen (1990), Banerjee & Duflo (2011), Acemoglu, Johnson & Robinson (2001), North (1990), Krugman (1991), Kaldor (1966).