CSR Misallocation Index — quantifying Gap 1
Method. Index = (state's share of state-attributable CSR) ÷ (state's share of India's MPI-poor). Index 1.0 = CSR share matches need share. Attributable CSR FY23-24 = ₹27,566 Cr (excludes ₹8,352 Cr of Pan-India/unattributed spend). MPI-poor ≈ 209 mn (NITI headcount ratios × 2023 population projections).
Headline numbers
- 61% of India's MPI-poor live in 7 states (Bihar, UP, MP, Jharkhand, Tripura, Meghalaya, Nagaland) that together receive 10.7% of attributable CSR.
- Bihar: Index 0.05 — 20.5% of India's poor, 0.95% of CSR. ₹61 of CSR per poor person per year.
- Goa: Index 48.8 — ₹64,308 per poor person. Delhi: Index 20.8. A poor person in Goa attracts ~1,000× the CSR of a poor person in Bihar.
- Correcting the 7 under-served states to Index 1.0 would require redirecting ≈ ₹14,000 Cr/yr — almost exactly the annual "Pan India" unattributed pool plus Maharashtra's excess.
Full table (sorted by Index, worst-served first)
| State | CSR Cr | ₹/capita | ₹/poor | CSR % | Poor % | Index |
|---|---|---|---|---|---|---|
| Bihar | 262 | 21 | 61 | 0.95 | 20.48 | 0.05 |
| Tripura | 9 | 23 | 176 | 0.03 | 0.26 | 0.13 |
| Uttar Pradesh | 1,591 | 68 | 294 | 5.77 | 25.85 | 0.22 |
| Madhya Pradesh | 613 | 71 | 343 | 2.22 | 8.55 | 0.26 |
| Jharkhand | 417 | 105 | 363 | 1.51 | 5.50 | 0.28 |
| Meghalaya | 35 | 105 | 378 | 0.13 | 0.44 | 0.29 |
| Nagaland | 15 | 70 | 454 | 0.06 | 0.16 | 0.34 |
| Mizoram | 4 | 37 | 704 | 0.02 | 0.03 | 0.53 |
| Assam | 496 | 139 | 717 | 1.80 | 3.30 | 0.54 |
| West Bengal | 877 | 88 | 744 | 3.18 | 5.64 | 0.56 |
| Chhattisgarh | 436 | 144 | 881 | 1.58 | 2.36 | 0.67 |
| Rajasthan | 1,177 | 145 | 949 | 4.27 | 5.93 | 0.72 |
| Jammu & Kashmir | 104 | 77 | 1,595 | 0.38 | 0.31 | 1.21 |
| Arunachal Pradesh | 40 | 249 | 1,809 | 0.14 | 0.11 | 1.37 |
| Odisha | 1,404 | 303 | 1,934 | 5.09 | 3.47 | 1.47 |
| Punjab | 368 | 120 | 2,525 | 1.34 | 0.70 | 1.91 |
| Uttarakhand | 366 | 313 | 3,235 | 1.33 | 0.54 | 2.45 |
| Manipur | 84 | 264 | 3,259 | 0.31 | 0.12 | 2.47 |
| Gujarat | 2,780 | 389 | 3,335 | 10.09 | 3.99 | 2.53 |
| Andhra Pradesh | 1,176 | 221 | 3,647 | 4.27 | 1.54 | 2.77 |
| Haryana | 850 | 282 | 3,981 | 3.08 | 1.02 | 3.02 |
| Himachal Pradesh | 158 | 211 | 4,280 | 0.57 | 0.18 | 3.25 |
| Karnataka | 2,328 | 344 | 4,538 | 8.45 | 2.45 | 3.44 |
| Telangana | 1,103 | 289 | 4,912 | 4.00 | 1.07 | 3.73 |
| Maharashtra | 6,284 | 497 | 6,366 | 22.80 | 4.72 | 4.83 |
| Tamil Nadu | 2,055 | 268 | 12,164 | 7.46 | 0.81 | 9.23 |
| Kerala | 400 | 112 | 20,288 | 1.45 | 0.09 | 15.39 |
| Sikkim | 42 | 598 | 23,005 | 0.15 | 0.01 | 17.45 |
| Delhi | 2,004 | 941 | 27,434 | 7.27 | 0.35 | 20.81 |
| Goa | 86 | 540 | 64,308 | 0.31 | 0.01 | 48.77 |
Caveats
- MPI headcount ratios are 2019-21 (latest official state series); poverty has fallen since, but the relative ranking is stable.
- Populations are 2023 projections; small states carry higher relative error.
- "Pan India" spend (22.7%) excluded — if it flows proportionally to observed patterns, the skew worsens.
- MPI is one need-proxy; a composite (MPI + learning outcomes + health burden) would refine, not reverse, the picture.