Why CAG Still Measures Output While World Measures Impact
Image credit X @OmBirla
Until the country shifts from counting audit paragraphs to counting audit payoffs, the gap between effort and effect will continue to widen.
By P SESH KUMAR
New Delhi, November 17, 2025 — The last two Activity Reports of the Comptroller and Auditor General of India-2023–24 and 2024–25-read like two fat epics with widening battlefields but eerily static victories. Each report celebrates the expanding territory of audit: more reports tabled, more paragraphs drafted, more recommendations issued, more Action Taken Notes (ATNs) vetted and more ministries nudged. But beneath the flood of numbers lies a quieter, more troubling truth: the Indian audit ecosystem still measures effort, not effect.
It celebrates the generation of audit material even when that material does not translate into legislative scrutiny, administrative correction or financial consequence. It is an audit system that produces evidence without ensuring afterlife.
As India’s Supreme Audit Institution (SAI) celebrates its Annual Audit Diwas this year in a glittering function in the prestigious Bharat Mandapam, it is appropriate to critically look at some of the challenges it faces in measuring and reporting its impact on the spectrum of public financial accountability.
Yes, it has announced MoUs with IITs, IIM, ICAI and IPAI- come out with its AI audit strategy framework, is working towards greater digitization through CAG- Connect and One IAAD ONE System(OIOS) etc. These are indeed laudable initiatives to equip the SAI with latest professional and technological tools to keep up with developments and perhaps even catch up with international peers in SAI world. Only time will tell their real impact on the ground in terms of improvement in timeliness, spread and quality of its reports and acceptance level of its recommendations. Hopefully CAG would reflect demonstrable achievements in its ensuing reports going beyond narration of initiatives.
Let us, however, start the story of effectiveness of its somewhat recent outputs, that begins in 2023–24, a year in which the CAG approved 115 audit reports, each dense with paragraphs flagging leakages, misappropriations, omissions and systemic faults. These reports produced 2,145 recommendations-an impressive number until one realises that barely 65 audit paragraphs from this entire corpus were actually selected by the Public Accounts Committee (PAC) of Parliament for discussion. That means the legislative pipeline absorbed less than ten percent of the audit universe.
Many of us would recall that as recently as August 2024, no less than the Supreme Court of India had declared the CAG reports acquired finality only on PAC examining these and giving its recommendations to Parliament. Till such time, it may not be incorrect to state that these reports are only views and ‘non- binding’ findings.
Therefore, even this ten percent is deceptive, for the Activity Report does not reveal how many of the discussed paragraphs resulted in specific administrative action, how many were accepted only notionally, and how many recommendations eventually died in the bureaucracy’s long corridors.
Acceptance of recommendations by ministries stood at 1,115, which superficially appears encouraging until divided by the total population of recommendations: barely half found acceptance and, more importantly, no data exists on the far more critical question of how many of these were implemented fully, partially or not at all. Audit acceptance without action is merely polite acknowledgment dressed up as compliance.
The situation becomes starker in 2024–25. The CAG’s outputs rise dramatically: 156 audit reports, 2,706 recommendations, and 3,010 ATNs vetted. In a country where audit is often dismissed as slow or conservative, these figures suggest an audit engine firing at full power.
Yet the engine’s torque remains curiously weak. Only 1,178 recommendations were accepted, a sharp drop in acceptance percentage from the previous year. Meanwhile, the number of paras actually discussed by the PAC shows no corresponding rise, meaning Parliament’s absorptive capacity has remained flat despite a 35 percent jump in audit outputs. Interestingly or rather alarmingly, if one may say, PAC in understood to have submitted its reports only in 41.26% cases (CAG reports) during 2015-24.
CAG’s Activity Reports of 2021-22 showed the acceptance levels of its recommendations at 19.66% for Union Government and 27.58% for States. These figures hovered around 17.87% and 47.96% respectively for 2022-23. Average acceptance level was 40.42% during 2021-22 to 2023-24 which very marginally increased to 43.53% in 2024-25.
Can one say that it is the paradox of Indian accountability: the SAI produces supposedly more evidence every year, but Parliament continues to digest the same small teaspoonful. CAG has unfortunately not been able to find a way out of this rather sad continuing trend of unconsidered reports.
The crisis of ATNs reveals the system’s structural flaw most clearly. In 2023–24, CAG vetted 1,973 ATNs and received 1,346. In 2024–25, it vetted over 3,000 and received close to 1,900. Yet these flows conceal the massive iceberg beneath-the many thousands of pending ATNs accumulated over a decade. The backlog continues to grow because ministries do not face any statutory deadline to respond, and because PACs cannot physically sit often enough to clear the pile.
To compound the problem, there were 6818 ATNs (85%) where not even first replies were received by CAG in 2024-25. This number was 6907 in 2023-24.
The ATN system, which was conceived as the bridge between audit observation and administrative action, has become a slow-moving administrative swamp in which paragraphs sink without trace. In the UK or the US such a situation would be unthinkable. The British National Audit Office (NAO) requires departments to respond swiftly, and the PAC follows up with rigorous hearings. The US Government Accountability Office (GAO) goes further, tracking every recommendation publicly, posting live implementation scores, and compelling federal agencies to respond within stipulated timeframes. Nothing comparable exists in India.
The most dramatic illustration of India’s impact deficit lies in the numbers on audit recoveries. In 2023–24, the CAG noted recoveries of ₹6,266.68 crore at the instance of audit against a departmental spend of ₹6,103.76 crore-almost a rupee returned for every rupee spent, which in any global audit system would be celebrated. But even in that year, “recoveries accepted” by departments were much higher than recoveries actually realised, pointing to the familiar problem that acceptance is not action- or the proverbial long winding path followed.
The next year, the picture deteriorates sharply. Recoveries accepted jump to ₹24,193.22 crore, an impressive mountain of audit persuasion. Yet actual realisation plummets to ₹2,599.03 crore, barely eleven percent of the commitments made. To call that an impact would be an exaggeration bordering on fiction. The GAO or NAO would classify the unrealised ninety percent as “benefits not yet achieved” and would publicly shame the agencies concerned; the Indian system quietly absorbs these unrecovered sums into the shadowlands of bureaucratic inertia.
The comparison with NAO and GAO is almost unfair-but instructive. The NAO’s 2024–25 report opens not with the number of paragraphs it wrote or the number of pages printed, but with the blunt declaration that it delivered £5.3 billion in financial impact, generating £53 for every £1 spent. It then narrates how its work led to specific policy corrections, service improvements and expenditure savings.
The GAO matches this with an even more astonishing figure: $75.9 billion in financial benefits in 2024, or $76 returned for every dollar invested, backed by implementation rates that consistently exceed seventy percent. Unlike India, these institutions do not appear to allow their recommendations to evaporate in bureaucratic humidity; they compel action through statutory timelines, parliamentary enforcement and relentless public tracking of implementation status.
India’s CAG, by contrast, looks like a brilliant scholar who produces a flood of research but has no publisher, no editor and no audience except a small committee that is perpetually overworked and perennially late. The problem is not that audit quality is lacking (it is a subject for exclusive examination) ; the problem is that the accountability ecosystem has not evolved to process and implement audit intelligence.
The CAG has become an output factory without a functioning conversion unit. Until India begins measuring the ratio of paras discussed to paras generated, the ratio of recommendations implemented to recommendations accepted, and the ratio of recoveries realised to recoveries claimed, the audit system will remain an exercise in documentation rather than transformation.
The way forward must therefore shift fundamentally from “audit as enumeration” to “audit as enforcement.” The CAG should move towards an annual “Impact Statement” that begins with a rupee-for-rupee return calculation, followed by a transparent account of rules amended, processes corrected and systemic reforms triggered by audit.
Parliament must expand PAC capacity, enforce strict timelines for ATN responses and hold ministries formally liable for non-compliance. And the audit reports themselves must evolve from long catalogues of paragraphs to focused examinations of high-impact areas with built-in follow-up mechanisms.
Audit means little unless it changes something. The CAG’s two recent reports show an institution capable of producing a tremendous volume of accountability material. What they also show-and what India must acknowledge-is that producing material is not the same as producing impact. Until the country shifts from counting audit paragraphs to counting audit payoffs, the gap between effort and effect will continue to widen. NAO and GAO have shown the world how an audit institution can become a powerhouse of administrative improvement. The challenge before India is not to admire them, but to emulate them.
(This is an opinion piece, and views expressed are those of the author only)
Follow The Raisina Hills on WhatsApp, Instagram, YouTube, Facebook, and LinkedIn
About The Author
