← The Constellation
The Docket · from the courts

What the Words Will Carry

Thursday · July 23, 2026 · A gas-appliance ban that splits the circuits, and a tugboat lien that survives a clause its holder never read: two courts on what a statute's exact words will carry.
I · EPCA preemption

Association of Contracting Plumbers v. City of New York

U.S. Court of Appeals for the Second Circuit · June 30, 2026

Two appellate courts this summer took ordinary-looking disputes and turned them into clean tests of what a statute's exact words will carry. The Second Circuit held that a federal appliance-efficiency law does not stop New York from barring gas stoves and furnaces in new buildings, and it said so while openly breaking with the Ninth Circuit, which had struck down a materially identical ban in Berkeley.[1] In a separate case decided in the parallel world of maritime law, the Fifth Circuit let a tugboat company enforce a lien against two barges it had towed, ruling that a no-lien clause hidden in someone else's contract counted for nothing because the tug company had not learned of it before its own deal closed. The sharper move in the first case is a reading-comprehension charge, that the rival court quoted half of a statutory definition and ignored the half that decided the question. The sharper move in the second is an inference from silence, that Congress struck a duty to investigate from the books in 1971 and thereby erased it.

The setup

New York decided that new buildings should stop burning fossil fuels. New York City's Local Law 154 forbids installing anything in new construction that combusts a substance emitting 25 or more kilograms of carbon dioxide per million BTUs, which in practice bars gas stoves, furnaces, and water heaters. The state passed a parallel measure directing its building-code council to prohibit fossil-fuel appliances in new construction.[1] Plumbing contractors, builders, construction unions, and a gas distributor sued, arguing that a 1975 federal statute overrides the local rules. That argument invokes preemption, the principle that under the Constitution's Supremacy Clause a valid federal law displaces a conflicting state or local one. The federal law is the Energy Policy and Conservation Act, or EPCA, which sets national efficiency standards for appliances and expressly forbids states from imposing their own requirements "concerning the energy efficiency, energy use, or water use" of covered products.[4]

Two suits ran in parallel. The Southern District of New York dismissed the challenge to the city law, and the Northern District granted summary judgment for the state, a ruling on the papers that no trial is needed. Both trial courts held that EPCA does not reach these bans. The plaintiffs appealed, the appeals were consolidated, and Judge Myrna Pérez wrote for the panel. Review was de novo, meaning the appeals court decided the legal question fresh and gave the trial courts' answers no deference.[1]

The question

The clause at the center is narrow on its face. EPCA blocks state rules "concerning" the "energy use" of a covered appliance. A ban on gas furnaces plainly affects how much energy gets used in a building, in the ordinary sense of those words, so the plaintiffs had a natural-language case that these laws concern energy use. The answer was not obvious for a second reason. The Ninth Circuit had already looked at the same clause and reached the opposite result, holding a Berkeley ordinance preempted, so a sister court of appeals had staked out the contrary position.[2]

The reasoning, walked

The panel's argument begins where the statute tells the reader to begin, with EPCA's own definition. Congress did not leave "energy use" to plain meaning. It defined the term as "the quantity of energy directly consumed by a consumer product at point of use, determined in accordance with test procedures."[5] The court treats that whole sentence as a term of art describing a laboratory metric, a fixed number assigned to a model through standardized federal test procedures before it is ever sold. On that reading, "energy use" is a property of the appliance measured in a lab, not a description of what a household happens to consume.[1]

From that definition the rest follows almost mechanically, and this first step carries the entire opinion. A local law forbidding a gas furnace cannot change the furnace's measured energy-use rating; the rating stays whatever the test procedure produced. A ban therefore does not regulate "energy use" as the statute defines it, because it leaves the regulated quantity untouched. The plaintiffs answered that the clause bars rules merely "concerning" energy use, and that "concerning" is broad. The court accepted the breadth but bounded it. Even read as "related to," a related-to link demands either an impermissible connection with EPCA's objectives or a direct reference to the regulated subject. Neither is present here. EPCA's objective is a uniform floor of efficiency performance, not a guarantee that consumers can buy any fuel type, and a ban sorted by fuel does not push manufacturers to beat or undercut the federal efficiency numbers. The ban also refers to appliance type, not to any quantity of energy consumed, so there is no direct reference either.[1]

The court then stacks structural and historical confirmation on top of the textual core. The preemption provision sits under a heading about energy conservation standards, the statute's separate definition of "energy conservation standard" tracks the same efficiency-metric language, and building-code carve-outs elsewhere in EPCA show Congress contemplated a narrower displacement than the plaintiffs urged. A 1987 rewording, the plaintiffs' best history, tightened the provision rather than expanding it, and the Department of Energy had long read it narrowly.[1] The panel closes with a consequences check. If "energy use" swept in any law that affects how much fuel a building burns, EPCA would federally preempt fire-safety bans on kerosene heaters, zoning limits on furnaces, and even noise ordinances, a wholesale takeover of local safety regulation that Congress never signaled.[1]

The Ninth Circuit had walked a shorter path to the opposite end. In California Restaurant Association v. City of Berkeley it read the same clause to preempt an ordinance banning natural-gas piping in new buildings, reasoning that a law which effectively eliminates the use of an energy source is a rule concerning energy use.[2] The Second Circuit's rebuttal is a close-reading charge. The Ninth Circuit, it says, seized on the phrase "at point of use" and stopped there, dropping the rest of the sentence, "determined in accordance with test procedures," which is exactly the language that anchors the metric to pre-sale testing rather than to what a resident does with an appliance. Reading only half the definition, the panel argues, let the other court treat everyday consumption as the statutory subject when the statute meant a lab value. The disagreement was not one-sided even inside the Ninth Circuit, where eight active judges dissented from the refusal to rehear Berkeley en banc, that is, before the full court rather than a three-judge panel.[2]

What turns on it

There are now two federal rules on the same recurring question. Inside the Second Circuit, outright fossil-fuel bans in new construction survive an EPCA challenge, while inside the Ninth, materially identical bans fall.[2] Builders, gas utilities, and municipalities face compliance that depends on which coast they build on, and a clean split between two circuits on a question of federal statutory interpretation is the standard trigger for Supreme Court review.[3] On the reasoning itself, the Second Circuit has the more disciplined reading, because it accounts for the full definition rather than a fragment of it, and the eight-judge en banc dissent suggests the Ninth Circuit's own bench doubts the broader view. The vulnerability is that the whole result rides on the defined-term move, and that move can read as formalism, since in plain English a gas ban obviously concerns energy use and the court's answer is that the statute's private dictionary governs instead. This is an assessment of how the opinion reasons, not legal advice about any building or project.

II · Maritime liens

Trailer Bridge v. Louisiana International Marine

U.S. Court of Appeals for the Fifth Circuit · June 17, 2026

The setup

A towing company left holding unpaid bills reached past its bankrupt customer and went after the barges themselves. Trailer Bridge owned two barges and chartered them, that is, leased them out, to a shipper named Work Cat to move containers between Tampa and Brownsville. Work Cat hired two tugboats from Louisiana International Marine to tow those barges. Work Cat paid for about two weeks, then stopped, ran up roughly 1.5 million dollars in towing charges, and went into bankruptcy.[1] The tug company responded by asserting a maritime lien on the barges. A maritime lien is a security interest that attaches to the vessel itself rather than to any person who owes money, and it is enforced by suing the vessel directly, an action styled in rem, meaning the case runs against the property, "the barge," rather than against a human defendant.[4]

The complication was a clause the tug company had never seen. The lease between the barge owner and Work Cat contained a no-lien provision, in which Work Cat promised to keep liens from arising on the barges. The tug company received a copy of that lease only on December 20, 2020, more than a month after it had already signed its own towing contract on November 12.[1] The owner sued first, seeking a declaratory judgment, a court ruling that fixes the parties' rights, to establish that no lien existed. The tug company counterclaimed and filed in rem against the barges. After a two-day bench trial, the district court found valid liens of about 1.49 million dollars and refused to award attorney's fees. The owner appealed the liens, the tug company cross-appealed on fees, and Judge Edith Jones wrote for the panel. The existence of a lien was reviewed de novo, a fresh legal call with no deference to the trial judge, while the underlying facts were reviewed only for clear error.[1]

The question

The dispute reduces to who bears the risk of a hidden no-lien clause. Does such a clause, sitting in a contract the tug company never read before doing the work, defeat the lien? And behind that, does a company supplying a vessel owe any duty to go hunting for such clauses before it signs? The statute grants liens generously to anyone providing necessaries, the goods and services a vessel needs to operate, a category the law defines to include towage, repairs, and supplies.[6] Yet the same body of law lets owners guard their vessels, and the friction between those two impulses is what the case had to resolve.

The reasoning, walked

The governing statute is the Commercial Instruments and Maritime Liens Act, which grants a maritime lien to anyone who provides necessaries to a vessel on the order of the owner or of a person authorized by the owner.[4] The court works through the three elements and finds each satisfied. Towage is a listed necessary, a barge is a vessel, and a charterer in control of a vessel is presumed by the statute to be authorized to order necessaries for it. Work Cat's order for tugs therefore presumptively created a lien on the barges before any defense is considered.[5]

The owner's first defense was that the tug company had looked only to Work Cat's credit, never to the barges, which if true would sink the lien. The court rejects it on the burden of proof. Sending invoices to the charterer alone does not show exclusive reliance on the charterer, and to overcome the statutory presumption favoring liens an objector must prove a deliberate intention to look solely to the owner's personal credit and to give up the lien. Trial testimony showed the tug company had regarded the barges as security all along, so the presumption held.[1]

That brought the court to the no-lien clause, and here the operative rule is one of timing. In the Fifth Circuit a no-lien clause defeats a lien only if the party furnishing necessaries had actual knowledge of the clause, and actual knowledge means an affirmative communication reaching an employee able to negotiate the deal, before the deal is made. The tug company saw the lease on December 20, but its towing contract had closed on November 12. Knowledge that arrives after the contract is complete comes too late to matter, so the clause was unenforceable against the tug company.[2]

The owner's strongest argument was that the tug company should at least have investigated for such clauses, a duty of reasonable diligence. The court answers with an inference drawn from words that are no longer in the statute, and this is the opinion's sharpest step. The 1920 Maritime Lien Act contained exactly that duty, denying a lien where a supplier by the exercise of reasonable diligence could have discovered that the person ordering the work lacked authority. In 1971 Congress deleted the reasonable-diligence language. The court reads the deletion as a deliberate abolition of the duty, reasoning that Congress removed the words because it meant to remove the requirement. It also refuses a narrower reading under which the deletion helped only shoreside suppliers, holding that the older term for such suppliers was used generically to cover everyone the modern statute reaches, towage operators included.[3]

Two smaller moves round out the analysis. The owner argued in the alternative that the no-lien clause stripped Work Cat of authority, so the third element failed. The court declines to conflate two different things, since the statute requires authority to obtain necessaries, not authority to create liens, and treating a no-lien clause as an automatic bar to authority would make the actual-knowledge rule pointless, because the clause would always defeat the lien no matter what anyone knew.[1] The tug company then tried to expand its win, folding roughly 361,000 dollars of fuel and lubricants into the lien as part of the value of towage. The court refuses, invoking the rule that maritime liens are construed stricti juris, strictly and without extension by analogy or inference. Fuel and lubricants were necessaries supplied to the tugs, not to the barges, the contract itemized them separately, and counting them would open an unbounded category.[1] On fees, the tug company prevailed yet recovered none, because its claim was in rem against the barges while a fee award would run in personam against the owner, who had no separate liability of its own, so the denial was within the trial court's discretion.[1]

What turns on it

For the industry the practical lesson is blunt. A no-lien clause in an owner-charterer contract is close to worthless against a tug or supplier that signs first and hears about the clause afterward, and owners cannot bolt on protection once the services have begun.[3] Suppliers, for their part, owe no duty to go looking, so the decision hardens the circuit's lien-friendly posture while reaffirming that stricti juris keeps the lien from swallowing every downstream cost.[2] The opinion drew no dissent, and the reasoning is tightest where it leans directly on statutory text, the presumption of authority and the timing rule. Its most exposed joint is the same deleted-text inference, because a repeal of words can mean Congress abolished a duty or merely trimmed language it thought redundant, and that is exactly the step a dissent would have pressed had there been one. The clean, administrable timing rule is likely why every judge signed on. This is a walk through how the court reasoned, not guidance for anyone's shipping arrangements.