India’s 65-utility AT&C loss headline is 15.04% for FY25. It is arithmetically correct. It is also a weighted mean doing substantial concealment work over a 16-percentage-point state dispersion, two distinct failure modes, and three simultaneously live national figures.
The 15.04% national average is a weighted mean of a 16-point state spread (RAOSCAFF arithmetic). The published state floor is 6.61% (Kerala) and the ceiling 22.76% (Madhya Pradesh) — both sit in the same MoP Integrated Rating FY25 report. The AT&C identity arithmetic proves both the MoP and PFC headline figures correct to the decimal from their respective billing and collection inputs.
The Ministry of Power’s 14th Annual Integrated Rating and Ranking of Power Distribution Utilities (FY25) — covering 65 rated utilities (54 discoms comprising 42 state-owned and 12 private + 11 power departments) — reports an all-India AT&C loss of 15.04% for FY2024-25, down from 15.97% in FY24. Prepared with PFC and REC as nodal and rating agencies.
AT&C — Aggregate Technical and Commercial — measures the share of all units handled by a utility that is neither billed nor collected as revenue. The direction of travel is established: the MoP Integrated Rating series (65-utility universe) shows 15.04% for FY25; a separate national-level series — the Rajya Sabha written reply by MoS Power — shows 16.16% for FY25 declining from 21.91% in FY21. These are two different defined quantities measuring one AT&C-loss idea; the FY21 trajectory point belongs to the parliamentary/national series, not to the 65-utility Integrated Rating universe. But the national figure is a weighted mean and a weighted mean conceals.
Three things the 15.04% does not reveal: the 16-point spread (RAOSCAFF arithmetic) across states and utilities; the two economically distinct leakages (billing gap and collection gap) that compose the single figure; and the three simultaneously-current correct values for “India’s AT&C loss” that three publishers report for the same concept. Each is decomposed below.
A 15.04% national average is a weighted mean of a 16-point state spread (RAOSCAFF arithmetic) — the published state floor of 6.61% (Kerala) and the ceiling of 22.76% (Madhya Pradesh) exist in the same FY25 report.
The private-vs-state gap is 5.35 percentage points (10.05% vs 15.40% — RAOSCAFF arithmetic), meaning the national figure blends two structurally different ownership regimes. Rajasthan, at 22.13% in FY24, retreated to 15.18% in FY25 — a single annual data point can mask violent intra-period reversals at the utility level. (Qualitative context on contributing factors appears in The Core; causal attributions are kept outside scope per the brief’s §F discipline.)
Sources: T&D India (state-level figures, ownership aggregates), attributing to MoP 14th Integrated Rating FY25. The Core (Rajasthan trajectory, Punjab smart-meter status) provides qualitative analysis and context; causal attributions are kept outside scope per §F discipline.
AT&C loss is the product of two separable failure modes. And three publishers simultaneously report different correct values for the same concept.
The AT&C identity: AT&C loss = 1 − (billing efficiency × collection efficiency). Billing efficiency measures metered-but-unbilled losses — metering gaps, energy accounting, tariff compliance. Collection efficiency measures billed-but-uncollected — payment enforcement and liquidity. They are economically distinct: a utility with high collection but low billing has a different problem profile, and a different intervention set, from one with the reverse. The composite number suppresses that diagnostic.
The arithmetic reconciles to the decimal in both MoP and PFC published universes: for MoP FY25, 1 − (87.59% × 97.00%) = 15.04% (RAOSCAFF arithmetic — matches MoP headline). For PFC FY24, 1 − (86.91% × 96.51%) = 16.12% (RAOSCAFF arithmetic — matches PFC headline). The two figures are for different rated sets and different years; neither corrects the other.
| Figure | Period | Publisher / Source | Rated universe |
|---|---|---|---|
| 15.04% | FY25 (2024-25) | MoP 14th Integrated Rating (PFC/REC nodal) | 65 rated utilities |
| 16.16% | FY25 (2024-25) | Rajya Sabha written reply, MoS Power Shripad Yesso Naik (session date not pinned by secondary coverage) | Broader national (definition unstated in secondary coverage) |
| 16.12% | FY24 (2023-24) | PFC Report on Performance of Power Utilities 2023-24 | 63 utilities |
The 15.04% vs 16.16% gap for the same FY25 year is not a data error: the Integrated Rating covers a curated 65-utility set; the parliamentary figure likely encompasses a broader or differently weighted aggregate. The 16.12% PFC figure is FY24 — from the PFC Report on Performance of Power Utilities 2023-24, a separately branded PFC publication. (The PFC Report on Performance of Power Utilities 2024-25 (February 2026) has since been released by PFC; it was not directly fetched for this brief.) Each value is accurate within its own definition. A citation without naming the publisher and rated universe is incomplete. 21 utilities achieved the upper threshold on billing efficiency in FY25; 17 utilities achieved 100% collection efficiency.
Sources: Power Line (MoP FY25 and PFC FY24 billing/collection figures); Saurenergy (Rajya Sabha reply); T&D India (MoP universe). Derived figures labelled RAOSCAFF arithmetic.
If you read one thing: the 15.04% national figure is arithmetically correct — and it is a weighted mean that compresses a 16-point spread (RAOSCAFF arithmetic), two separable failure modes, and three simultaneously valid national values into a single headline.
The direction of travel is established: the MoP Integrated Rating series (65-utility universe) reports 15.04% for FY25; a separate national-level series (Rajya Sabha written reply, MoS Power) reports 16.16% for FY25 declining from 21.91% in FY21 — two different defined quantities measuring one AT&C-loss idea, with the FY21 point belonging to the parliamentary series only. The pace and uniformity are not: 26 of 65 utilities deteriorated in FY25 even as the national mean improved; the state floor-to-ceiling gap (16.15 pts, RAOSCAFF arithmetic) is wider than the national mean itself; and the AT&C identity arithmetic exposes two economically distinct failure modes compressed into a single published percentage. Which component (billing vs collection) dominates differs by utility; the composite figure does not reveal that diagnostic. Three publishers simultaneously report three different correct values for the same concept, each accurate within its own rated universe and reference year; a citation without naming the publisher and universe is structurally incomplete. The metric’s composition, not just its level, is the analytical object. This brief decomposes the published number as a measurement object and takes no position on policy, on any utility’s performance, on any government scheme, or on any minister or party.
Mirror format — RAOSCAFF anchors on named publishers and decomposes the figures they print. Mirror No. 24 draws on five corroborating trade sources (secondary citations) attributing to three distinct primary documents: (1) the MoP 14th Annual Integrated Rating and Ranking of Power Distribution Utilities (FY25), cited at 15.04% for 65 utilities, as reported by Power Line (16 Mar 2026), T&D India, and The Core, each independently attributing to that primary; (2) a Rajya Sabha written reply by MoS Power (16.16% FY25, trajectory from 21.91% FY21) as reported by Saurenergy — this figure is not from the MoP 14th Integrated Rating; and (3) the PFC Report on Performance of Power Utilities 2023-24 (16.12% FY24, billing/collection inputs) as reported by Power Line (Jun 2025) — a separately branded PFC publication. The canonical primary PDFs were not directly fetched; pfcindia.co.in returned a TLS error during direct fetch. The three publisher figures are not all from one source; each traces to its labelled publisher and period. Every state-level figure, billing/collection efficiency component, and ownership-sector aggregate used in this brief traces to one of five fetched source URLs listed in the Sources section below. Subsidy-receivable decomposition has no fetchable hard figure and is kept as qualitative texture only, not a quantified axis.
Four RAOSCAFF arithmetic figures appear in this brief: the state dispersion range (22.76 − 6.61 = 16.15 pts); the private-vs-state gap (15.40 − 10.05 = 5.35 pts); the AT&C identity check for MoP FY25 (1 − 0.8759 × 0.9700 = 15.04%, reconciles to MoP headline); and the AT&C identity check for PFC FY24 (1 − 0.8691 × 0.9651 = 16.12%, reconciles to PFC headline). All four are labelled “RAOSCAFF arithmetic” throughout. No figure is introduced from model memory or training data. The three-publisher divergence table uses figures cited directly from fetched sources.
All five sources are trade publications (secondary citations). Sources 1, 2, and 5 attribute to the MoP 14th Annual Integrated Rating and Ranking of Power Distribution Utilities (FY25). Source 3 (Saurenergy) attributes to a Rajya Sabha written reply — not to the MoP 14th Integrated Rating. Source 4 (Power Line) attributes to the PFC Report on Performance of Power Utilities 2023-24 — a separately branded PFC publication. The canonical primary PDFs were not directly fetched; pfcindia.co.in returned a TLS error during direct fetch. Full citation register including RAOSCAFF arithmetic inputs and outputs is in the companion FACTS.md, § H.