On Friday 2 October, twenty states, the District of Columbia and the cities of Chicago, Denver, New York and San Francisco filed suit against the National Highway Traffic Safety Administration in the US Court of Appeals for the First Circuit.
Twenty-six plaintiffs. One target: the final CAFE rule that sets the US fleet average at 34.9 mpg by model year 2031, down from the roughly 50.4 mpg the 2024 rules projected for the same year.
The legal theory is narrow, old, and considerably more dangerous to the rule than a general climate argument would be.
The Statute Is the Whole Case
The coalition is not arguing that the rule is bad policy. It is arguing that NHTSA lacked the authority to write it.
The Energy Policy and Conservation Act of 1975 does not give the agency open discretion to pick a number. It directs NHTSA to set fuel-economy standards at the maximum feasible level. That phrase is a statutory command with fifty years of case law behind it, and it constrains the agency in one direction only: toward stringency that technology and cost make achievable.
The states' claim is that 34.9 mpg is not the maximum feasible level, and that NHTSA reached it by distorting the inputs rather than by finding that more was impossible.
The specific methodological objection is the sharp end of this. The coalition alleges the agency disregarded the existing EV fleet when calculating baseline fleet performance. If you model what the industry can achieve while excluding the most efficient vehicles the industry already builds and sells, "maximum feasible" comes out lower. The plaintiffs say that is not a judgement call — it is an input error that predetermined the output.
Layered on top is the Administrative Procedure Act claim: that the rule is arbitrary and capricious. That is the standard vehicle for arguing an agency did not rationally connect its evidence to its conclusion.
The 1975 statute does not ask NHTSA what target it prefers. It asks what is feasible. A rule that answers the first question when the law asked the second is vulnerable regardless of who is in office.
What the Rule Actually Changed
The final rule, signed by NHTSA Administrator Jonathan Morrison on 26 September and issued on 28 September, covers model years 2022-2031 — three of which are already complete. Its mechanics matter more than its headline number.
The stringency ramp is shallow:
- Passenger cars: 0.90% annually through MY2029, then 1% through MY2031
- Non-passenger vehicles: 0.51% annually through MY2029, then 1% through MY2031
Two structural changes do more work than the ramp:
Credit trading ends. From MY2028, manufacturers can no longer trade CAFE credits between each other. Credits earned in MY2027 stay tradable through 2032, then the market closes. The rule also removes air-conditioning efficiency and off-cycle fuel consumption improvement adjustments.
The fleet gets reclassified. From MY2030, a vehicle no longer counts as a light truck merely by offering three fold-flat seat rows. It must show payload plus towing capacity of at least 8,500 pounds, or meet tightened off-road criteria including a minimum 28-degree approach angle. The effect is to flip the fleet split from roughly 70% light trucks / 30% cars to roughly 70% cars / 30% trucks.
That reclassification is the part the industry will feel. Crossovers that have spent a decade being regulated as trucks — against a softer curve — move onto the passenger-car curve. A rule that lowers the headline target while moving a hundred-odd nameplates into the stricter category is not a straightforward loosening for every manufacturer.
The Two Dollar Figures Do Not Describe the Same Thing
Both sides are quoting money, and the numbers are not in conflict so much as they are measuring different ledgers.
NHTSA's own estimates: $60.6 billion in technology cost savings across MY2027-2031, roughly $1,290 saved per vehicle in MY2031, and 122 billion additional gallons of gasoline consumed through 2050. The Transportation Department puts the headline benefit at $138 billion over five years.
The states claim nearly $220 billion in lost fuel savings for drivers.
Both can be true. The agency is counting what manufacturers do not have to spend on compliance technology, which flows partly to buyers as lower sticker prices. The states are counting what drivers spend on the extra fuel those less-efficient vehicles will burn over their service lives. Cheaper to buy, more expensive to run — and the agency's own 122-billion-gallon figure is the states' case stated in the agency's own arithmetic.
Attorney General Rob Bonta, whose office led the filing, put it as the administration "attempting to gut fuel economy standards and force Americans to spend billions more on gas while poisoning the air in our communities."
The industry's position is not uniform but it is audible. Shane Karr, Stellantis' Senior Vice President for Public Affairs, welcomed "the Administration's efforts to reset the CAFE regulations to more achievable targets."
Why It Matters
For planning purposes, the rule is now contingent. Product cycles for MY2028-2031 are being locked in now, and the credit-trading shutdown and MY2030 reclassification are exactly the kind of provisions a court can vacate while leaving the rest standing. Any manufacturer building a compliance plan on the assumption that credits disappear in 2028 is building on a provision under active challenge.
For anyone watching how the US regulatory pendulum behaves, the venue choice is the thing to note. The First Circuit is not where an administration would want this heard, and petitioners get to pick. A rule written to be durable across administrations would have been written to survive the "maximum feasible" test on a clean record. This one is being defended on the record NHTSA built in a hurry.
The number to watch is not 34.9. It is whether the court accepts that the agency was entitled to exclude the EV fleet from its own baseline.
