The Loophole Is the Policy: How Accounting Rules Hollow Out the Draft CAFE-III Norms

The Lift Line

The target is written in the headline. The stringency is written in the annexure, and the annexure is where the cut disappears.

Why This Editorial Matters for Your Exam

Vehicular emissions usually enter a GS3 answer as a single line about Bharat Stage VI. CAFE norms are a different instrument entirely, and the distinction is examinable: BS-VI regulates pollutants per vehicle (particulate matter, oxides of nitrogen), while CAFE regulates carbon dioxide and fuel consumption across a manufacturer’s fleet. One is a health standard applied to every car; the other is a climate and energy-security standard applied to an average.

More valuable than the facts is the method. This editorial is a clean worked example of how a policy can be tightened and weakened in the same document, and the technique of reading past the headline to the compliance rules transfers to forest cover definitions, poverty lines, renewable energy targets and fiscal deficit arithmetic alike.

The consultation window on the draft closed on 6 August 2026, the day this piece was published, which is why the argument was made when it was.

GS Paper 3: Conservation, environmental pollution and degradation, environmental impact assessment; infrastructure and energy; awareness in the field of science and technology; indigenisation of technology.

Background and Context

CAFE norms in India are notified under the Energy Conservation Act, 2001, and administered by the Bureau of Energy Efficiency (BEE) under the Ministry of Power. This matters constitutionally and administratively: emission standards for pollutants sit with the Ministry of Road Transport and Highways and the environment ministry, but fuel efficiency is treated as an energy-conservation subject, which is why the Ministry of Power is the issuing authority.

The rationale is threefold: cutting oil import dependence, reducing carbon dioxide emissions from a rapidly motorising economy, and pushing manufacturers toward efficient powertrains.

The draft CAFE-III document was released on 16 July 2026 with stakeholder comments invited until 6 August 2026, and the norms are proposed to take effect from 1 April 2027. The fuel-consumption equivalents in the draft run from 3.996 litres per 100 km in FY2027-28 to about 3.327 litres per 100 km in FY2031-32.

An earlier consultation draft of September 2025 had proposed a concession of about 3 gCO2/km for petrol cars under 909 kg kerb weight, under four metres in length and below 1200 cc. It was widely read as favouring a single dominant manufacturer, was opposed on the ground that rewarding lighter cars would work against crash safety in a Bharat NCAP era, and was reportedly revised out. The weight question then reappeared in a subtler form, as the reference weight and slope of the limit curve itself.

The Analysis

1. A fleet standard is an arithmetic, and the arithmetic is negotiable. Because compliance is judged on a sales-weighted average rather than model by model, every rule that decides how a sale is weighted is a rule about stringency. A regulator who wishes to appear ambitious and act cautiously has a menu: keep the headline target and adjust the weighting. That is what critics say has happened here.

2. Super-credits invert the incentive they were designed to create. A multiplier of 3.0 for a battery electric vehicle means one such sale is entered into the fleet average as three sales of a zero-emission vehicle. The manufacturer’s measured average falls sharply on a small volume. The intended effect was to seed an electric vehicle market; the practical effect, once the market exists, is that a token electric portfolio buys permission to keep selling high-emission vehicles. Retaining static multipliers into FY2031-32, rather than tapering them, is the specific charge.

3. Carbon-neutrality factors substitute assumption for measurement. An 8 per cent discount for E20 to E30 petrol, 22.3 per cent for flex-fuel ethanol vehicles and 5 per cent for compressed natural gas are applied to reported emissions on the reasoning that biogenic carbon was recently absorbed from the atmosphere. The reasoning is not baseless, but it is a full life-cycle claim being applied at the tailpipe, and it ignores the land, water and fertiliser intensity of ethanol feedstock. Whatever the merits, the atmosphere responds to emitted carbon dioxide, not to reported carbon dioxide.

4. The weight curve is the quietest and possibly the largest lever. Indexing targets to vehicle mass has a defensible logic: a seven-seat vehicle cannot be held to the same absolute limit as a hatchback. But two adjustments between the September 2025 and July 2026 drafts, raising the industry reference weight from 1,170 kg to 1,229 kg and flattening the target slope, both move permitted emissions upward for heavy vehicles. With utility vehicles at roughly 65 per cent of passenger vehicle sales in FY26, this is not a marginal adjustment. It removes the regulatory pressure to downsize in exactly the segment where downsizing would matter most.

5. Off-cycle credits and a buyout complete the picture. Up to 9.0 gCO2/km may be claimed through approved technologies whose benefits do not appear in the test cycle. Separately, the draft contemplates a mechanism under which a non-compliant manufacturer may purchase credits at an indicated Rs 2,500 to Rs 4,500 per gCO2/km, against a statutory penalty of the order of Rs 5,000 per gCO2/km. Pricing a compliance instrument below the penalty for non-compliance is a design decision with a predictable consequence: rational firms buy the cheaper option, and the standard becomes a tariff.

6. The test is the arithmetic of ambition. The claim that matters is quantitative. On the Centre for Science and Environment analysis, the combined effect of these provisions is that a manufacturer can reach full compliance with an electric vehicle sales share of only about 8 per cent by FY2031-32, and that the real-world emission reduction falls to roughly 13.8 per cent against a headline tightening of about 31 per cent. India’s stated national ambition is 30 per cent electric vehicle penetration by 2030. A fuel-efficiency regulation that can be satisfied at a quarter of that share is not merely neutral toward the transition; it removes the regulatory reason to accelerate it.

Data and Institutions Vault

Prelims-grade facts:

• CAFE = Corporate Average Fuel Efficiency; notified under the Energy Conservation Act, 2001; administered by the Bureau of Energy Efficiency (BEE), Ministry of Power
• Applies to passenger vehicles of gross vehicle weight below 3,500 kg, across petrol, diesel, LPG, CNG, hybrid and electric
• CAFE-I (2017 to 2022): 130 gCO2/km · CAFE-II (2022 to 2027): 113 gCO2/km
• Draft CAFE-III: 94.76 gCO2/km in FY2027-28 tightening to 78.90 gCO2/km in FY2031-32, effective from 1 April 2027; draft released 16 July 2026, comments closed 6 August 2026
• Fuel-consumption equivalents: 3.996 litres/100 km to about 3.327 litres/100 km
• Super-credit multipliers: battery electric and range-extended electric 3.0, plug-in hybrid 2.5, strong hybrid 1.6 (reduced from 2.0), flex-fuel 1.1 (reduced from 1.5)
• Carbon-neutrality factors: E20 to E30 petrol 8 per cent, flex-fuel ethanol 22.3 per cent, CNG 5 per cent
• Industry reference weight raised from 1,170 kg (September 2025 draft) to 1,229 kg (July 2026 draft); target slope flattened
• Off-cycle technology credits capped at 9.0 gCO2/km
• Test cycle: Modified Indian Drive Cycle (MIDC) at present; migration to WLTP referenced but the compliance date is not notified
• Utility vehicles were about 65 per cent of Indian passenger vehicle sales in FY26 (SIAM)
• European Union comparator: a mass-indexed limit curve with a fleet target of 93.6 gCO2/km (WLTP) for 2025 to 2029 and a 100 per cent reduction, that is zero tailpipe carbon dioxide, from 2035 under Regulation (EU) 2023/851
• Author: Anumita Roychowdhury of the Centre for Science and Environment, writing in Down To Earth, 6 August 2026

Watch the trap: do not write that CAFE-III sets a target of 70 g/km. That figure belongs to the indicative CAFE-IV phase. CAFE-III runs from 94.76 to 78.90 gCO2/km. Equally, do not confuse CAFE with Bharat Stage VI: BS-VI is a per-vehicle pollutant standard under motor vehicle rules, CAFE is a fleet-average carbon dioxide and fuel consumption standard under the Energy Conservation Act. And note that super-credits are not subsidies, they are accounting multipliers, so they cost the exchequer nothing and cost the atmosphere a great deal.

The Debate

Argument FOR the flexibilities. India is not Europe. Per capita vehicle ownership is low, personal mobility is still aspirational, and a regulation that raises the price of an entry-level car suppresses a legitimate welfare gain. Charging infrastructure and battery supply chains are immature, so a standard calibrated to a mature electric vehicle market would simply be missed rather than met, producing penalties instead of cleaner cars. Super-credits were designed precisely to bridge that gap. Carbon-neutrality factors are internally consistent with a national policy that mandates E20 blending, and it would be incoherent for the state to require ethanol blending and then refuse to recognise it. A weight-indexed curve is standard international practice, including in the European Union. Predictable, gradual tightening keeps the industry inside the process rather than litigating outside it.

Argument AGAINST. Every one of those defences is an argument for a transition, and a transition requires a sunset. What the draft does instead is keep static multipliers to FY2031-32, expand off-cycle credits, raise the reference weight and flatten the slope in the same revision that claims to tighten the standard. The cumulative effect is measurable and it is large: an 8 per cent electric vehicle share suffices for compliance, and the real-world cut collapses from about 31 per cent to about 14 per cent. Meanwhile the weight curve subsidises the fastest-growing and least efficient segment in a market where utility vehicles are already two of every three cars sold. A buyout priced below the statutory penalty tells manufacturers, in the plainest possible terms, that non-compliance is a budget line rather than a breach.

Balanced verdict. Flexibility mechanisms in a fleet standard are not a scandal; they are how such standards are made politically survivable everywhere, including in the European Union. The serious objection is not that they exist but that they are static, uncapped in aggregate effect, and applied simultaneously with a weight-curve adjustment that pushes in the same direction. The correct design is a tapering schedule announced in advance, an aggregate cap on credits any single manufacturer may claim, neutrality factors grounded in measured life-cycle evidence and revised periodically, a steeper rather than flatter weight curve, and publication of the modelled real-world trajectory alongside the headline target. None of that requires abandoning gradualism. It requires only that the announced ambition and the achievable outcome be allowed to face each other in public.

How to Think About This

The transferable pattern is: the target sets the headline, the accounting sets the outcome.

Whenever a regulator announces a numerical target, three questions follow before you accept the number.

What is the unit of compliance? A standard applied to an average behaves nothing like a standard applied to each item. Averages can be met by composition changes rather than by improvement, and the weighting rules decide everything.

What deductions are permitted before the measurement is compared to the target? Multipliers, discount factors, banking, trading, carry-forward and technology credits all sit between the measured quantity and the compared quantity. Each is individually reasonable; their product is often decisive.

What happens on failure, and what does that cost relative to compliance? If the price of non-compliance is set below the cost of compliance, the regulation has priced itself out of effect.

Apply the same three questions to renewable purchase obligations, to fiscal deficit targets that permit off-budget borrowing, to forest-cover targets that count plantations as forest, and to carbon-neutrality pledges that lean on offsets. The structure recurs exactly.

Diagram-in-Words

THE HEADLINE
CAFE-II: 113 gCO2/km  ──────────►  CAFE-III: 78.90 gCO2/km by FY2031-32
                                    (a tightening of about 31%)

THE ACCOUNTING LAYER (where the cut leaks out)

  Measured fleet CO2
        │
        ├─ x SUPER-CREDITS ─────────► BEV 3.0 | PHEV 2.5 | Strong hybrid 1.6
        │                             (a few clean sales outweigh many dirty ones)
        │
        ├─ minus CARBON-NEUTRALITY ─► E20-E30: -8% | Flex-fuel: -22.3% | CNG: -5%
        │        FACTORS              (discount by assumption, not measurement)
        │
        ├─ minus OFF-CYCLE CREDITS ─► up to 9.0 gCO2/km
        │
        └─ compared against a WEIGHT-INDEXED TARGET
                 reference weight 1,170 kg ──► 1,229 kg
                 slope FLATTENED
                 (heavier car = more permissive target = SUVs rewarded,
                  in a market that is already 65% utility vehicles)
                                    │
                                    ▼
        AND IF STILL SHORT: buy credits at Rs 2,500-4,500 per gCO2/km
                            (statutory penalty is about Rs 5,000 per gCO2/km)

THE OUTCOME
Real-world cut falls to about 14%          Compliance needs only ~8% EV share
                                            National ambition: 30% by 2030
                                    │
                                    ▼
        LEVER: taper the multipliers, cap total credits, steepen the
               weight curve, base CNFs on measured life-cycle data,
               notify the WLTP date, publish the real-world trajectory

Takeaway Box

Lift line for an answer:

A fleet emission standard is not tightened by lowering the target. It is tightened by refusing to lower the passing grade.

Prelims hooks: CAFE notified under the Energy Conservation Act, 2001, administered by BEE, Ministry of Power; CAFE-I 130 g/km (2017-22), CAFE-II 113 g/km (2022-27), draft CAFE-III 94.76 to 78.90 g/km (FY2027-28 to FY2031-32) effective 1 April 2027; draft released 16 July 2026; super-credits BEV 3.0, PHEV 2.5, strong hybrid 1.6; CNF E20-E30 8 per cent, flex-fuel ethanol 22.3 per cent, CNG 5 per cent; off-cycle credit cap 9.0 gCO2/km; reference weight 1,170 kg to 1,229 kg; MIDC to WLTP migration undated; utility vehicles about 65 per cent of PV sales in FY26; EU fleet target 93.6 g/km (WLTP) 2025-2029 and zero tailpipe CO2 from 2035.

Ethics and interview angle: a technically accurate regulation whose annexures ensure a far smaller effect than its headline implies has told no lie. Is there an ethical obligation on a regulator to disclose the expected real-world outcome of a rule, not merely its stated target? Consider this against the administrative values of transparency, non-arbitrariness and the public trust doctrine, and against the reality that the affected public here, breathing urban air, is not represented in the consultation while the regulated industry is.

PYQ linkage: UPSC has examined vehicular emission standards and the BS-VI transition, the relationship between energy security and import dependence, and the design of India’s electric mobility policy including FAME and the National Electric Mobility Mission Plan. This editorial gives you the current regulatory instrument and, more usefully, a critique of it that is quantitative rather than rhetorical.

Probable question: “The effectiveness of an environmental standard is determined less by its stated target than by its compliance accounting.” Examine this statement in the light of India’s draft CAFE-III norms, and suggest design reforms.

Sources: Down To Earth, Bureau of Energy Efficiency, Ministry of Power, SIAM

Source: The Loophole Is the Policy: How Accounting Rules Hollow Out the Draft CAFE-III Norms — Ujiyari.com | Free UPSC & State PCS Editorial Analysis