The National Economic Mobility Index (NEMI): A Framework for Income-Based Assistance

The problem NEMI is trying to solve

Any system of assistance needs an address. It has to answer the question: who is this rupee for? India's current answer is a category tag inherited at birth. That tag was chosen in the 1950s because it was the only proxy for deprivation that could be administered without a computer, a bank account or a verifiable income record. All three now exist for the overwhelming majority of households.

The National Economic Mobility Index is an attempt to write a better address. It is not a moral argument about who deserves help; it is an engineering argument about targeting error. Every targeting system produces two failures — help that reaches a household that does not need it, and a household in need that receives nothing. A good index minimises the second while keeping the first small enough to be politically and fiscally tolerable.

How the index is constructed

NEMI scores a household on four verifiable inputs. Verified household income, drawn from filed returns and bank-linked records rather than a self-declared certificate. Landholding and property, since income alone understates rural wealth. Parental education, which captures first-generation-learner status — the single strongest predictor of educational outcome in the Indian data. And district development, which corrects for the fact that a given income buys very different opportunity in Gurugram and in Bastar.

Each input is normalised and weighted, producing a continuous score rather than a binary in/out decision. Continuity matters: binary thresholds create cliffs where a rupee of extra income costs a family an entire scholarship, and cliffs are exactly what people learn to game. A continuous score means assistance tapers, so there is no incentive to understate income at the margin.

The index is designed to be recomputed annually. A family that has moved into stability stops drawing support gradually and automatically, without needing a committee to declare it. This is the mechanism the current system lacks: nothing in the existing framework can ever tell a household that it no longer qualifies.

Objections worth taking seriously

The strongest objection to any income-based model is that discrimination is not only economic. A family may be prosperous and still face exclusion in housing, marriage or hiring. This is true, and NEMI does not pretend to address it. Discrimination is a matter for enforcement — anti-discrimination law, complaint mechanisms with teeth, and prosecution — not for a scholarship formula. Confusing the two has left India with a scholarship formula doing the work of a police function, and doing it badly.

The second objection is administrative: income can be concealed. It can, but far less easily than in 1950, and concealment is a solvable enforcement problem with penalties, random audit and cross-verification against property and GST records. The current system does not have a concealment problem because it does not ask the question at all — that is not an advantage.

The third objection is transitional. Moving from tags to an index cannot be done in a single admission cycle without chaos. The framework therefore proposes a decade-long phase, running both systems in parallel with published outcome data each year, so the transition can be halted or corrected on evidence rather than on argument.