Washington Asks for Preemption From a Scheme It Is Busy Narrowing

The Supreme Court released its October 2026 argument calendar on August 4. The first case of the sitting, set for Monday, October 5, is Suncor Energy (U.S.A.) Inc. v. Commissioners of Boulder County, No. 25-170. The Court granted certiorari on February 23, 2026, from the Supreme Court of Colorado's decision in 2025 CO 21, decided May 12, 2025, and directed the parties to brief a jurisdictional question alongside the merits. Petitioners filed their merits brief on May 14; respondents filed theirs on July 27, corrected August 3.

The underlying litigation is eight years old. Boulder County and the City of Boulder sued in Colorado state court in 2018, pleading five Colorado common-law claims against fossil-fuel producers: public nuisance, private nuisance, trespass, unjust enrichment, and civil conspiracy. A Colorado Consumer Protection Act claim was dismissed without prejudice for failure to plead with particularity and is not before the Court. Respondents' brief separates the surviving claims into deception claims, alleging decades of misleading advertising, and sales and production claims, alleging that producing, promoting, refining, marketing, and selling fossil fuels caused and contributed to climate alteration. Respondents seek damages only, and state that they do not ask any court to enjoin oil and gas operations or to enforce emissions controls. The complaint alleges property damage valued in the billions of dollars.

Two questions are presented. First, whether the Court has statutory and Article III jurisdiction. Second, whether federal law precludes state-law claims seeking relief for injuries allegedly caused by the effects of interstate and international greenhouse-gas emissions on the global climate. The United States filed as amicus curiae supporting petitioners in May; the Solicitor General is recused, and Deputy Solicitor General Sarah M. Harris signed as counsel of record. That brief reports that 60 States and localities have brought near-identical suits, and discloses that the United States has itself sued to block two of them, in United States v. Minnesota, No. 26-cv-2456 (D. Minn.), and United States v. Vermont, No. 25-cv-463 (D. Vt.).

For anyone pricing this docket, the operative figure is not the damages number in any one complaint. It is the count of parallel suits. Exposure spread across dozens of separately venued cases, each with its own causation record and its own trial calendar, is a different item entirely from exposure that resolves in a single ruling. A merits ruling for petitioners on the preclusion question would retire the whole set at once. A ruling that stops short of it leaves every one of those cases exactly where it sits.

The government's route to that result runs through two arguments. The constitutional one is that Boulder is using Colorado law to regulate conduct that occurred almost entirely elsewhere, and that structural limits on state authority plus the federal foreign-affairs power forbid it. The statutory one is that the Clean Air Act makes EPA the decisionmaker on whether and how to regulate stationary-source emissions under 42 U.S.C. 7411, leaves source States a cooperative role under Section 7416, and cannot tolerate Colorado judges setting a competing standard by damages award. Both rest on the premise that common-law liability is regulation, which is correct, and which the Court said plainly enough in Kurns v. Railroad Friction Products Corp.

The difficulty sits in the government's own footnote. It acknowledges that EPA concluded in February that it lacks authority under 42 U.S.C. 7521(a)(1) to prescribe greenhouse-gas emissions standards for new motor vehicles on climate-change grounds, at 91 Fed. Reg. 7686, and distinguishes that conclusion as limited to mobile sources rather than to Section 7411. The distinction is real and the brief is entitled to make it. But the posture should be said out loud: the United States is arguing that a federal scheme is exclusive enough to extinguish state tort claims at the same time that the agency running it is narrowing what it believes the scheme authorizes. Preemption is an assertion of federal power, and a desk that objects when agencies claim authority Congress did not grant them owes the same discipline here. Field exclusivity is earned from statutory text, not from the convenience of a single national answer, and the States hold the police power until that text says otherwise.

Boulder leads with jurisdiction, and that is the strongest part of its brief. On Article III, it argues that ASARCO Inc. v. Kadish supplies the rule and defeats review: the Colorado decision was not a final judgment altering tangible legal rights but a refusal to dismiss on the pleadings, and the increased risk of an eventual adverse judgment plus the cost of continued defense are byproducts of being sued rather than cognizable injuries. On 28 U.S.C. 1257(a), it argues that the fourth exception in Cox Broadcasting Corp. v. Cohn does not apply because federal defenses survive and reversal would end no claim outright, and that Atlantic Richfield Co. v. Christian does not stretch to this posture, because petitioners sought no writ, named no court as a respondent, asked in terms for interlocutory review, and obtained a remand.

The government's answer is that Colorado Appellate Rule 21 creates a self-contained original proceeding, that the Colorado Supreme Court said it was exercising original jurisdiction, and that under Atlantic Richfield the nature of the proceeding rather than the issues reviewed controls finality. That is a serious answer, and Atlantic Richfield is close on its facts. It is also self-limiting in a way the brief does not quite face: if the nature of the proceeding governs rather than the label, the state court's own use of the phrase cannot be dispositive either, and the inquiry turns back to what the proceeding actually did. Reasonable justices can land on either side of that.

On the merits, respondents have one strong argument and one weak one. The strong one is doctrinal housekeeping: American Electric Power Co. v. Connecticut and International Paper Co. v. Ouellette both treated the availability of state-law claims after federal common law was displaced as an ordinary question of statutory preemption, and Ouellette permitted nuisance claims under source-state law rather than barring state law wholesale. If that is right, the second question presented is a Clean Air Act question, not a constitutional one, and the constitutional theory is doing work the statute should be doing. The weak one is scope. Respondents characterize their claims as reaching upstream production and marketing rather than emissions, but the damages they seek are measured by the downstream effects of emissions, and American Trucking Associations v. Los Angeles cautions against escaping a preemptive scheme by moving up the supply chain. That characterization is doing more work than the damages theory can support.

Argument is October 5, and the first signal will be how much of the hour goes to jurisdiction. If the questioning is dominated by Section 1257(a) finality and by whether a Rule 21 proceeding is self-contained, the realistic outcomes narrow toward a jurisdictional dismissal or a dismissal as improvidently granted. Either would leave 2025 CO 21 in place and return Boulder to the Colorado trial court, with the parallel suits proceeding on their own calendars and the preclusion question returning years later on a final judgment. If jurisdiction is disposed of briskly, watch whether the merits argument sounds in the Clean Air Act or in constitutional structure. That choice determines how far the eventual ruling travels beyond climate claims.

The operational point for counsel and finance leadership on either side of this docket is that outcome and timing are separate variables, and reserving practice tends to conflate them. A jurisdictional disposition is not a merits loss, but for planning purposes it behaves like one: it restores the full calendar of separate cases without resolving the question they share. Any model that treats October 5 as a single binary event is understating the range.

Bennet's exposure-modelling desk tracks the state and local climate docket, the federal suits filed against those programs, and the reserving and disclosure practice that follows them.