Revised 6 August 2026 · v2.2.0

The claim

Everyone knows diesel drives inflation. The data does not.

The argument is intuitive enough that it rarely gets tested. Diesel moves freight; freight moves food; food moves the consumer price index. Therefore holding diesel prices down holds inflation down. It appears in market commentary as settled fact — one widely-read explainer walks through the entire chain, from tractors to irrigation to the Reserve Bank’s rate stance, without offering a single pass-through estimate, basket weight or citation.

So we tested it against fifteen years of official series: wholesale price indices for diesel, urea and DAP from DPIIT’s Office of the Economic Adviser, consumer price indices from MoSPI, April-anchored, FY2012 through FY2026.

The +0.92 is not evidence of transmission. Diesel sits inside the wholesale index, so the index partly measures diesel by construction. The number that would matter is diesel against retail inflation, and it is +0.11 — indistinguishable from noise. Against food inflation specifically, where the freight story is supposed to be strongest, the sign is negative. Lagging diesel by a year, which is how a cost shock would plausibly reach the shelf, gives +0.09.

What is actually happening

Two prices for one fuel, forty rupees apart

If diesel’s wholesale signal is not reaching retail prices, something is absorbing it. PPAC’s Industry Consumption Report for May 2026 says what.

Diesel sold directly by oil marketing companies to bulk buyers — industry, railways, mining, shipping — is priced more than ₹40 per litre above the same fuel at a retail pump. Bulk buyers have responded exactly as anyone would: they buy at the pump instead.

A ₹40 gap between two channels for an identical commodity is not a market outcome. It is the visible edge of an administered retail price, measured against a channel left closer to cost. Private marketers, who have no obligation to sell below cost and no PSU balance sheet to absorb the difference, lost more than a third of their volume in a single year.

This is the strongest available evidence that Indian retail diesel is held below cost-reflective levels. It is a current, official, unambiguous number, and it does not depend on a correlation. It also reframes the policy question. Price control is real and large. Its observable effect is not a lower CPI print — it is channel arbitrage and the retreat of private competition.

Where the fuel goes

Diesel is a freight fuel, and the freight is growing faster than the fuel

Diesel remains India’s dominant petroleum product at 43.3% of the market. May 2026 consumption was 8.73 million tonnes, up 1.6% year on year; the April–May total was 17.06 MMT, up 1.2%.

That is modest growth, and it sits oddly beside every proxy for freight activity in the same report.

Goods are moving considerably faster than diesel is being burnt. Some of that is efficiency and some is a shift to rail — railways rose from 9% to 12% of direct-sales diesel over the year. But the gap is wide enough to be worth naming as an open question rather than explaining away, and it is the natural place to ask what else could be carrying freight.

The leakage

A price gap on a bulk liquid is an invitation

Arbitrage on a commodity does not stay theoretical. When the same litre is worth ₹40 more depending on who is buying it, and a second subsidised liquid exists that looks similar in a tank, the incentive to move fuel between categories is substantial.

Both halves of that incentive appear in the same PPAC report. The ₹40 gap is one. The other is kerosene: subsidised PDS kerosene accounted for 73% of kerosene consumption in May 2026, against 27% non-PDS. Kerosene is the classic diesel adulterant.

The enforcement record is consistent with it. In Bengaluru in 2025, the Central Crime Branch arrested four people siphoning diesel destined for BMTC and KSRTC bus depots, seizing roughly 60,000 litres from three tankers; the method was to draw about 50 litres from each tanker in transit using duplicate keys and report the shortfall to depot managers as spillage. In Udupi, a tanker driver was arrested with 1,020 litres recovered, charged under the Bharatiya Nyaya Sanhita, the Essential Commodities Act 1955 and the Petroleum Act. The central government has separately directed state chief secretaries to act against fuel adulteration.

The recommendation

Convert the depot-based fleet; close the wedge; do not cut the retail price

Three independent lines of evidence point at the same segment of demand.

Structure. Heavy and light commercial vehicles plus buses were roughly 38% of diesel in the last economy-wide survey — the largest single block, and the one least likely to have shrunk since.

Feasibility. That segment is depot-based and route-bound. Refuelling infrastructure for compressed natural gas is tractable where vehicles return to a known point each night, and intractable where they do not. Agriculture, at around 13% of diesel, is dispersed and seasonal; it is the wrong target. Private diesel cars are solving themselves.

Leakage. Compressed gas is physically far harder to divert than a liquid. It cannot be siphoned into a can, decanted at the roadside, or blended with kerosene. It is dispensed under pressure into a sealed, metered vessel. For state transport undertakings and municipal fleets — exactly the buyers in the Bengaluru case — conversion removes a whole category of loss alongside the emissions case.

Which leaves the price question, and here the evidence cuts against the obvious ask. Arguing for a cut in retail diesel prices is arguing against PPAC’s own report. Retail diesel already sits ₹40 below the unregulated channel. Cutting it further would widen the arbitrage that drives diversion, deepen the fiscal absorption, and weaken the economics of every fleet considering a switch.

The coherent ask is different: close the tax wedge on commercial fleet fuel while converting those fleets to gas. That shrinks the arbitrage from both ends — a smaller gap to exploit, and a fuel that is harder to exploit it with — and it preserves the relative advantage that makes conversion worth financing.

The stationary segment

Generators and telecom towers burn diesel that never moves

The freight argument above concerns diesel that travels. A second block does not move at all, and it is the easier conversion.

A generator is the best conversion candidate in the entire fuel system. It sits on a known site, it has a fixed address for a gas connection or a cylinder cascade, it does not need a refuelling network, and its duty cycle is predictable. Gas gensets are a mature product; the constraint is piped-gas availability at the site, not technology.

Telecom towers deserve separate mention because the sector has spent fifteen years trying to leave diesel and still burned 1.54% of national consumption in the study period — more than aviation and shipping combined. Towers are the clearest case of diesel used purely as backup for an unreliable grid, and the improving power position weakens the case for it: peak demand of 270.82 GW in April–May 2026 was met in full, with a deficit of 0.0%.

Doorstep delivery

The arbitrage, formalised and licensed

There is now a legal industry built on the gap this article has been describing. Doorstep diesel delivery brings fuel to the consumer in a licensed mobile dispenser rather than requiring the consumer to reach a pump.

The operators are real businesses at scale. FuelBuddy, founded in 2020, acquired MyPetrolPump in May 2021. Humsafar, founded in 2016, supplies non-automotive diesel to industrial, healthcare and hospitality customers. Repos Energy, backed by Ratan Tata, reports a presence in 130 cities. Pepfuels operates in the same segment, and the public-sector oil marketing companies have themselves invited start-ups into doorstep delivery.

The regulatory shape matters more than the company list. Delivery is licensed by the Petroleum and Explosives Safety Organisation, which specifies the tankers, requires periodic inspection of mobile dispensers, and mandates documentation of quantity, quality and delivery location for every transaction. Critically, it is permitted for approved business locations and generators, not for private vehicles or personal use.

Two things follow. First, the customer base is precisely the stationary segment above — gensets and industrial sites — so doorstep delivery and genset conversion are competing answers to the same problem. Second, a licensed channel with per-transaction documentation of quantity and location is the opposite of the siphoning described earlier. The same demand that fuels diversion also supports a formal channel; policy determines which one grows.

The running-cost gap

Where CNG’s advantage is actually measured

The clearest argument for switching is not environmental — it is arithmetic. A companion piece on this blog models cost per kilometre from 303 ARAI-certified SIAM fuel-efficiency declarations and OMC-published city fuel boards, on a Delhi July 2026 basis.

A quarter off the per-kilometre running cost is the switching case, and it does not depend on any claim about inflation or emissions. Two features of the comparison are worth drawing out. The CNG gain comes from combustion efficiency and per-unit price together, and — unlike ethanol blending, which dilutes one liquid with another and costs 2–6% in mileage on the ARAI/SIAM range — a petrol-to-CNG conversion on the same nameplate vehicle shows a mileage gain. The two fuel shifts move in opposite directions on efficiency.

Sources. Cost per kilometre — vehicle_fuel_mileage repository, calibrated on 303 ARAI-certified SIAM BS-VI fuel-efficiency declarations and OMC city fuel boards. Wholesale price indices — DPIIT, Office of the Economic Adviser, rebased 2022-23 = 100; consumer price indices — MoSPI; correlations are the author’s calculation on April-anchored annual growth rates, FY2012–FY2026. Consumption, channel shares, the direct-versus-retail price gap, sectoral direct-sales breakdown, kerosene shares, e-way bills, port cargo and vehicle sales — PPAC, Industry Consumption Report POL & NG, May 2026. Economy-wide diesel end-use — PPAC / Nielsen India, All India Study on Sectoral Demand of Diesel & Petrol, fieldwork January 2012 to June 2013, report dated 20 December 2013. Import dependency and self-sufficiency — PPAC, Snapshot of India’s Oil & Gas Data, March 2025. Peer-reviewed study cited but not read: P. P. Ghosh, “Impact of India’s diesel subsidy reforms and pricing policy on growth and inflation”, Energy Economics 113 (2022), doi:10.1016/j.eneco.2022.106195. Diesel siphoning cases — Deccan Herald reporting on the Bengaluru and Udupi arrests; fuel-adulteration directive — Autocar India. Press reports are secondary sources and are labelled as such in the text.

About this article: Researched, written and edited by Umashankar Triplicane Dwarakanathan, with AI research assistance; every figure is meant to trace to the primary source cited. See the Editorial Policy for how sourcing, AI use and corrections work.