Billions Spent, and the Gap Remains. The Cause Is Not the People — It Is the System.

When an institution spends billions, holds the data, employs capable people, and the gap survives anyway, the ordinary explanations collapse. It was not underfunded — the money was spent. It was not ignorance — the data existed. It was not incompetence — the people were, by every measure, able. And yet the readiness shortfall remained, the project slipped, the cost overran. When budget, information, and competence are all present and the outcome still fails, the cause is none of them. The cause is structural: the system captured the signal and could not carry it to a decision that closed.

India's own auditor has documented this pattern for more than a decade, across defence, public-sector undertakings, and national projects. Read the reports as a body of evidence rather than a sequence of scandals, and one shape appears again and again: a problem detected, a problem reported, and then a gap that persists — because between the detection and the corrected outcome, nothing owned it, tracked it, or learned from it. That gap is not a failure of the people inside these institutions. It is the absence of a closed loop.

The money was spent. So why did the gap survive?

Because spending is an input, and the gap is an outcome — and nothing in most systems connects the two after the money leaves. A budget is approved, an order is placed, a project is sanctioned. The system of record captures each of these perfectly: the sanction, the contract, the disbursement. What it does not do is carry that commitment forward to whether it produced the readiness, the capability, the delivery it was meant to. The record ends at the transaction. The outcome forms somewhere the record cannot see.

Consider what the Comptroller and Auditor General found on ammunition. In its audit, the CAG reported that 121 of 152 types of ammunition did not meet the minimum levels needed to sustain intense war, and that a large share of the stockpile would not last even ten days against a requirement measured in weeks. Money had been allocated. The requirement was known. The shortfall was documented. And still it persisted — because the signal that stock was falling below the operational line was never converted, in time, into an owned correction that closed the gap. The audit even noted that the situation was flagged to the ministry in 2013 and did not receive a reply for two years. That two-year silence is the open loop, made visible. The detection happened. The closing did not.

Isn't this just a human failure — someone who didn't act?

No — and this is the heart of it. It is tempting to read every audit finding as someone, somewhere, failing to do their job. But when the same pattern repeats across different institutions, different decades, different people, the individual explanation stops being credible. If capable people, given the data and the budget, still cannot close the gap, then the thing failing is not the people. It is the machinery they are forced to work inside — machinery that hands each person a signal and no mechanism to carry it through to a verified result.

A person can read a report. A person cannot, alone, hold a forward position across a vast supply chain, price the shortfall in days and capability before it binds, assign the correction to an accountable owner with a clock, watch whether it happened, and score the outcome to learn for next time. That is a system's job, and the system was never built to do it. So the work falls to individuals who do their best with the fragment they can see — and the intelligence to close the loop lives, if it lives anywhere, in their own experience, which walks out the door when they are posted elsewhere. The gap is not that people failed. It is that no system converted what they knew into a correction that held.

What does "the system cannot convert the signal" actually mean?

It means the data existed and never became a decision. Every one of these institutions runs systems of record. They capture the transactions, the inventories, the project milestones with precision. The signal — stock falling, a schedule slipping, a cost rising — is present in the data. What is missing is the layer that reads that signal forward and turns it into an owned, tracked, corrected outcome. The signal arrives. It is even reported. And then it sits, because the architecture ends at reporting.

The CAG's audit of DRDO's mission-mode projects shows the same mechanism in a different domain. It found that 67 per cent of the projects examined did not adhere to their deadlines, with time overruns ranging from 16 to 500 per cent. In one case, after more than nine years and roughly ₹942 crore, a system still did not meet the user's requirements. These are not projects that no one was watching. They were monitored, reviewed, and reported on throughout. The signals of slippage were captured at every stage. What was absent was a loop that carried each slippage forward to a correction, scored whether the correction worked, and learned which early signals truly preceded failure — so the next project was caught sooner. Detection ran continuously. The loop stayed open.

Does speeding things up fix it?

Not if the loop is still open — and the audit record proves it. When the gap becomes visible, the instinct is to act faster: waive the process, expedite the order, compress the timeline. But speed without a closed loop just produces a faster open loop. The CAG examined army emergency procurement contracts concluded under special waivers whose entire purpose was speed, and found that in 72 per cent of the contracts examined, items were not delivered within the stipulated timelines. Only a minority were completed within a year. The waiver removed the friction. It did not close the loop — nothing tracked each expedited order forward to the outcome the speed was meant to secure, so the delays reappeared inside the fast lane.

This is the lesson hidden in the numbers: the problem was never the speed of the process. It was that no system carried the decision through to a verified result and learned from the gap between what was promised and what was delivered. You can make an open loop faster. It is still open.

Is this only a defence problem?

No — and that is the clearest proof that it is structural rather than institutional. The same pattern appears wherever large sums meet complex delivery. Government monitoring of major infrastructure has repeatedly shown hundreds of projects running over cost simultaneously — in one official report, 443 projects each above ₹150 crore carrying cost overruns exceeding ₹4.92 lakh crore, alongside hundreds more delayed. These are different ministries, different agencies, different people from the defence examples entirely. The one thing they share is the architecture: systems that record the commitment and cannot carry it to the outcome.

If defence, DRDO, procurement, and public infrastructure — separate institutions, separate workforces — all produce the same shape of gap, the gap cannot be explained by any of them individually. It is not a defence problem or a PSU problem or an enterprise problem. It is an open-loop problem, and every institution that spends against a plan and cannot close the distance between the plan and the result has it. Private enterprises have it too; they simply do not have a Comptroller and Auditor General to document it in public.

What would closing the loop have changed?

It would have converted each detected signal into an owned, verified correction — which is the one thing every one of these cases lacked. A closed loop does not merely report that stock is falling or a project is slipping. It carries the position forward, prices the shortfall in money and time before it binds, assigns the correction to an accountable owner on a clock, tracks whether the correction happened, scores the outcome against what was predicted, and learns which early signals truly precede the failure — so the next one is caught earlier and handled better. The forecast, the risk, the owner, the action, the outcome, the score, the learning: one loop, closed, and auditable at every step.

Applied to the audit findings, the difference is concrete. The ammunition shortfall would have surfaced as a priced, owned correction while there was still time to act, not as a two-year silence. The slipping project would have been scored against its own forecast at each stage, and the pattern of slippage learned rather than repeated across 178 projects. The expedited contract would have been tracked to delivery, and the gap between promised and actual speed fed back so the next waiver was not the same false comfort. In every case, the data was already there. What was missing was the loop that turned it into a decision that held.

The uncomfortable truth the audits keep circling is that the institutions were not short of money, data, or capable people. They were short of a system that closes the loop — and no amount of the first three substitutes for the fourth. That is why the gap survives the spending. It is not a human failure to be corrected by exhortation, and not a budget failure to be corrected by more money. It is a structural absence, and it is corrected by one thing only: closing the loop, so the signal the institution already holds finally becomes the decision it was always meant to produce. The loop has to close.