In a diverse federation like India, the state is the primary agent of development. Yet its sub-national governments carry roughly two-thirds of the country’s expenditure responsibilities while commanding only a fraction of its revenue powers. That mismatch is the root of the fiscal squeeze the states repeatedly lament, and it has deepened since 2017, when the GST folded much of their tax autonomy into a shared pool and left them more transfer-dependent than before. Borrowing ceilings cap what they can spend, a transfer mechanism tilting steadily from untied grants towards conditional, tied schemes narrows what they can choose. The states, in short, have far less room to manoeuvre than their budgets suggest.The usual debate about that room stops too early. It fixates on committed expenditure, and treats whatever is left as freely spendable. The irony is that the real constraint begins precisely where the committed-expenditure discussion ends. India’s fiscal squeeze is structural, not merely a matter of political choice. Incumbents everywhere want to spend to retain power, but the federal system leaves so little headroom that profligacy is more constrained than the rhetoric admits, which is exactly why states reach for increasingly innovative fiscal workarounds.The deeper story is not about any one government’s politics, it is about how India raises and spends public money. We profess fiscal decentralisation, the structure delivers something closer to its opposite. Most of what the states spend is not freely chosen but imposed, by the constitutional division of subjects, Finance Commission awards, the grants-in-aid framework, and a thicket of statutory and central mandates. The recent white papers released by the incoming Kerala and Tamil Nadu governments make the point in plain numbers.What committed concealsThe standard gauge of a state’s fiscal space is the ratio of committed expenditure, salaries, pensions and interest, to revenue receipts. In Tamil Nadu these three heads now absorb ₹1.89 lakh crore, or 64.4 per cent of revenue receipts, with interest alone at ₹67,050 crore, eating up 22.8 per cent. In Kerala the same trio consumes 77.6 per cent. Read conventionally, the story ends here. a state spending two-thirds to three-quarters of its income on staff, retirees and creditors has little left to deploy.But the committed-expenditure frame flatters reality. It implies that everything after salaries, pensions and interest is discretionary, money a finance minister can direct at will. That is the illusion the white papers, read carefully, dismantle.The non-discretionary layerBeneath the headline ratio lies a thick stratum of obligations no government can switch off, yet which appear nowhere in the committed-expenditure number. Grants to local bodies, mandated by State Finance Commission awards and constitutional design, typically claim 4-6 per cent of revenue receipts. The state’s matching share in Centrally Sponsored Schemes is effectively compulsory, since to decline your share is to forfeit the Centre’s, costing another 4-6 per cent. Add the bill for honouring judicial orders, regulatory compliance, the recurring cost of conducting elections and mandated disaster-management spending (together roughly 1-2 per cent), and the government’s own contribution to the National Pension System (1-2 per cent). Individually modest, together these add an estimated 10-15 per cent of revenue receipts on top of committed expenditure for an average state, and considerably more for a stressed one (Chart 1). None of it is committed in the accounting sense; all of it is non-negotiable in practice. Layer on the migration of transfers from untied to tied schemes, and the freedom to allocate shrinks further still.The squeeze on prioritiesWhatever sliver survives must still finance the promises that define a government. The strain already shows: developmental expenditure has slipped to 61 per cent of total spending, from 67 per cent before the pandemic (Chart 2). Food and welfare subsidies, pensions for the vulnerable, scheme outlays for education, health and housing, grants for creating capital assets, and the unglamorous but unavoidable cost of maintaining a vast stock of public infrastructure, these are not luxuries but the substance of governance, and the basis on which both governments were elected. Yet they must be funded from a discretionary space squeezed to a fraction of what the headline numbers imply.Capital formation bears the brunt. When the inflexible layer swallows the revenue account, capital spending becomes the residual, the first thing cut. And as Chart 2 shows, the squeeze is uneven: states carrying larger revenue deficits feel the heat most, their effective fiscal space compressed hardest.Two states, one destinationTamil Nadu and Kerala reach that point by different routes. Tamil Nadu’s debt has nearly doubled, from ₹5.13 lakh crore in 2020-21 to about ₹10 lakh crore in 2025-26, and ₹13.18 lakh crore once public-sector liabilities are counted. Its revenue deficit, at ₹78,324 crore, is the highest in the state’s history. Kerala carries ₹5.07 lakh crore in outstanding liabilities, a heavier 35.5 per cent of GSDP.In relative terms Kerala’s 77.6 per cent committed ratio is the more alarming, reflecting a large pension and salary bill against a slow-growing revenue base. Tamil Nadu’s bigger economy gives it marginally more headroom on paper (64.4 per cent), but its scale of borrowing and its ₹67,050-crore interest bill expose the weight of its debt. The figure that should anchor the debate, though, is neither committed ratio. It is the share of revenue locked up by the full non-discretionary layer, closer to 87 per cent in Tamil Nadu and an estimated 90 per cent in Kerala.Reframing the conversationThree conclusions follow. First, fiscal space is mis-measured: the effective residual, not the committed ratio, is the governance reality. Second, that residual is squeezed from two directions, a uniform, imposed structural layer common to all states, and a variable committed core that drives the differences between them. Structure and discipline are not rivals but entangled: structure sets the pressure, discipline shapes the response. Third, capital formation is the shock absorber that takes the strain, which makes the cost of today’s welfare spending intertemporal, the state keeps paying for welfare while ceasing to build the capacity to deliver it.The rupee a state government spends, in other words, was rarely free to begin with. Until the debate moves from the committed ratio to the effective residual, we will keep misreading both the squeeze and the states that live inside it.The writer is Assistant Professor, Gulati Institute of Finance and Taxation. Views are personalPublished on September 2, 2026