The Sanction Is Not the Squadron — India’s Defence Money Now Moves; Its Timelines Do Not

Pritiraj BrahmaFounding Partner, Pritiraj & Partners · 16 August 2026

India’s defence budget has crossed ₹7.84 lakh crore and, for the first time in years, the capital account was spent almost in full. The binding constraint has moved from money to time — and time is the one variable the system does not publish.

ProvenancePublished 16 August 2026. Budget figures are 2026-27 Budget Estimates as presented on 1 February 2026. Utilisation figures are 2025-26 revised-estimate-stage outturns as reported by the Ministry of Defence in April 2026. Committed liabilities have not been publicly disclosed since 2019-20.

In 2025-26 the Ministry of Defence spent 99.62 per cent of what it was given and granted Acceptance of Necessity to proposals worth ₹6.81 lakh crore — roughly three and a half times the entire modernisation budget it has for 2026-27. The money problem is closing. The clock problem is not being measured.

For most of the last decade the standard criticism of Indian defence budgeting wrote itself: the capital account was too small, and what little there was could not be spent. The first half of that sentence is still arguable. The second half has stopped being true.

The Ministry of Defence reports that it fully utilised its capital outlay on Defence Services for 2025-26 — ₹1.86 lakh crore at the revised-estimate stage, raised during the year from a Budget Estimate of ₹1.80 lakh crore to meet requirements arising after Operation Sindoor — and that utilisation of the ministry’s whole budget, including civil expenditure and pensions, was 99.62 per cent. In the same year the ministry granted Acceptance of Necessity to 109 proposals worth ₹6.81 lakh crore, against 56 proposals worth ₹1.76 lakh crore the year before, and signed contracts for 503 proposals worth ₹2.28 lakh crore.

The 2026-27 Budget, presented on 1 February 2026, gives the ministry ₹7,84,678 crore. That is 15.19 per cent above the 2025-26 Budget Estimate of ₹6.81 lakh crore, about 2 per cent of estimated GDP, and 14.67 per cent of central government expenditure — the largest allocation to any ministry. PRS Legislative Research points out that measured against the 2025-26 revised estimate rather than the budget estimate, the increase is 7 per cent. Both numbers are correct. Which one is quoted tells you what the quoter is arguing.

Capital Outlay on Defence Services rises to ₹2,19,306 crore, about 22 per cent above last year’s Budget Estimate. Within the ministry’s own composition statement, 27.95 per cent of the allocation is capital, 26.40 per cent is pay and allowances, 21.84 per cent is defence pensions, 20.17 per cent is revenue expenditure on sustenance and operational preparedness, and 3.64 per cent is civil organisations. PRS puts salaries and pensions together at 44 per cent of defence spending in 2026-27, growing at about 8 per cent a year since 2014-15.

Five findings anchor this analysis:

1. Absorption is no longer the binding constraint. A ministry that reports 99.62 per cent utilisation, and a capital account raised mid-year and then spent in full, is not a ministry that cannot spend. The lapse-and-surrender critique of Indian defence capital has been overtaken by events, and analysis that still leads with it is arguing about the last decade.
2. Sanctions now run several times ahead of the money that can follow them. ₹6.81 lakh crore of Acceptance of Necessity in a single year sits against a modernisation allocation of ₹1,85,467 crore for 2026-27 (MP-IDSA’s reading of the budget documents). At that ratio the approvals granted in one year represent something close to three and a half years of the entire modernisation budget — and by coincidence almost exactly the size of the previous year’s whole defence budget. An AoN is a decision to seek a capability, not a decision to pay for one; the gap between the two is where Indian defence acquisition actually lives.
3. The number that would let anyone audit that gap is not published. Committed liabilities — the payments falling due in a year on contracts already signed — have not been publicly disclosed since 2019-20, a point PRS makes in its 2026-27 analysis while citing the Standing Committee on Defence’s third report of December 2019. Without it, no outsider can say how much of the ₹1.85 lakh crore modernisation budget is already spoken for before a single new contract is signed.
4. The capital share remains below the level Parliament’s own committee has asked for. Capital is 27.95 per cent of the ministry’s allocation this year, against roughly 27 per cent averaged over the past decade; the Standing Committee on Defence has held that a 60:40 revenue-to-capital split would be healthier. The gap is not caused by parsimony. It is caused by a personnel and pension floor that no budget cycle can move.
5. Indigenisation is being measured at the point of order, not the point of delivery. ₹1.39 lakh crore — about 75 per cent of the capital acquisition budget — is earmarked for procurement from domestic industry, and research and development draws ₹17,250 crore of capital with ₹29,100 crore for DRDO. The earmark is an input target with a published number. There is no comparable published series for what was delivered against contract, on time, and accepted by the service.

Why the sanction is the wrong milestone

An Indian defence acquisition passes through a long chain of gates: Acceptance of Necessity by the Defence Acquisition Council, a request for proposal, technical evaluation, field trials, staff evaluation, commercial negotiation, approval by the competent financial authority or the Cabinet Committee on Security, contract signature, and then delivery over years against a payment schedule.

Two points on that chain are published as annual totals — AoNs granted and contracts signed. Everything between them, which is where years are lost, is not published as a series at all. There is no public dataset of median time from AoN to contract, no public distribution of how long trials take, no published count of proposals that have been carrying an AoN for more than three years. The Ministry of Defence is not hiding these numbers so much as not being required to keep them in a form anyone can inspect.

That is an institutional design failure of the most ordinary kind. The system has an owner for each gate and no owner for the interval between gates. A file that moves slowly through six offices, each of which acted correctly, produces a delay for which nobody is accountable — because delay is a property of the chain and the chain has no custodian.

What a full capital account conceals

Full utilisation is a genuine achievement and it is also a weaker signal than it appears. Capital spending in defence is dominated by instalments on contracts signed years earlier. A ministry can spend its entire capital budget without concluding a single new competition, because committed liabilities from past contracts absorb most of the account. This is precisely why the committed-liability figure matters: it is the difference between a budget that is buying new capability and one that is servicing old decisions.

Because the number is not published, the honest position is that we do not know which of those two descriptions fits 2026-27 — and the reader should be suspicious of anyone, in either direction, who claims to.

The counter-case, honestly stated

Three arguments run the other way, and they are strong.

First, the system demonstrably compressed under pressure. Emergency procurement after Operation Sindoor moved money and equipment on timelines the peacetime process does not achieve, and the mid-year increase in capital outlay was absorbed rather than surrendered. A process that can compress has capacity; the question is why the compressed mode requires an operation to trigger it.

Second, a large AoN pipeline is deliberate, not accidental. A wide funnel is how a buyer keeps competition alive and options open; the number of approvals is not a promise, and treating every AoN as an unfunded commitment misreads what the instrument is for. Doubling AoN count in a year is a statement of intent by a ministry that has been criticised for indecision.

Third, 503 contracts worth ₹2.28 lakh crore in one year is real throughput by any comparable standard, and the 75 per cent domestic earmark has coincided with a defence industrial base that did not exist at this scale a decade ago.

All three are true. None of them requires the timeline to stay unpublished.

What we would do

1. Restore publication of committed liabilities, annually, by service. One table, in the budget documents, showing payments due in the coming year on contracts already signed. It was published until 2019-20. Restoring it costs nothing and immediately makes every claim about the modernisation budget checkable.
2. Put a published clock on every acquisition gate. For each live AoN: date granted, current gate, days at that gate, and the officer accountable for moving it. Aggregate the distribution publicly; keep the platform-level detail classified where it must be. What is secret is the capability, not the calendar.
3. Report capital performance against the revised estimate, not the budget estimate. The 15 per cent and the 7 per cent in this year’s budget describe the same allocation. A ministry serious about its own record should publish both, in the same table, every year.
4. Convert the indigenisation earmark into a delivery measure. Publish, against the 75 per cent domestic target, the value actually delivered and accepted in the year — not the value contracted. An earmark that is only measured at signature will be met at signature.
5. Give the interval an owner. A single accountable authority for elapsed time across the acquisition chain, reporting quarterly to the Defence Acquisition Council on ageing proposals, with the power to close proposals that have carried an AoN past a stated limit. A pipeline nobody prunes is not a pipeline; it is a backlog with better manners.

India has done the hard fiscal part. The money is there, it is larger than it has ever been, and it is being spent. What remains is the part that costs nothing and is therefore never done: publishing how long things take, and naming who is answerable when they take longer. A sanction is a sentence in a file. A squadron is a squadron.

Companion essay: Ten Lakh Uniforms and Three Different Answers — the same architecture approached through a different instrument.