Mayor and City Council of Baltimore v. BP P.L.C.

PUBLISHED

UNITED STATES COURT OF APPEALS
FOR THE FOURTH CIRCUIT

No. 19-1644

MAYOR AND CITY COUNCIL OF BALTIMORE,

Plaintiff - Appellee,

v.

BP P.L.C.; BP AMERICA, INC.; BP PRODUCTS NORTH AMERICA, INC.;
CROWN CENTRAL LLC; CROWN CENTRAL NEW HOLDINGS LLC;
CHEVRON CORP.; CHEVRON U.S.A. INC.; EXXON MOBIL CORP.;
EXXONMOBIL OIL CORPORATION; CITGO PETROLEUM CORP.;
CONOCOPHILLIPS; CONOCOPHILLIPS COMPANY; PHILLIPS 66;
MARATHON OIL COMPANY; MARATHON OIL CORPORATION;
MARATHON PETROLEUM CORPORATION; SPEEDWAY LLC; HESS CORP.;
CNX RESOURCES CORPORATION; CONSOL ENERGY, INC.; CONSOL
MARINE TERMINALS LLC; SHELL PLC; SHELL USA, INC.

Defendants – Appellants

and

LOUISIANA LAND & EXPLORATION CO.; PHILLIPS 66 COMPANY;
CROWN CENTRAL PETROLEUM CORPORATION.

Defendants.

--------------------------------

CHAMBER OF COMMERCE OF THE UNITED STATES OF AMERICA;
NATIONAL ASSOCIATION OF MANUFACTURERS; NATIONAL
ASSOCIATION OF CONVENIENCE STORES; SOCIETY OF INDEPENDENT
GASOLINE MARKETERS OF AMERICA; ENERGY MARKETERS OF
AMERICA; ENERGY POLICY ADVOCATES; STATE OF INDIANA; STATE
OF ALABAMA; STATE OF ALASKA; STATE OF ARKANSAS; STATE OF
GEORGIA; STATE OF KANSAS; STATE OF KENTUCKY; STATE OF
MISSISSIPPI; STATE OF MONTANA; STATE OF NEBRASKA; STATE OF
SOUTH CAROLINA; STATE OF TEXAS; STATE OF UTAH; STATE OF
WYOMING,

Amici Supporting Appellant.

NATIONAL LEAGUE OF CITIES; U. S. CONFERENCE OF MAYORS;
INTERNATIONAL MUNICIPAL LAWYERS ASSOCIATION; PUBLIC
CITIZEN, INC.; SHELDON WHITEHOUSE; EDWARD J. MARKEY; STATE OF
MARYLAND; STATE OF CALIFORNIA; STATE OF CONNECTICUT; STATE
OF NEW JERSEY; STATE OF NEW YORK; STATE OF OREGON; STATE OF
RHODE ISLAND; STATE OF VERMONT; STATE OF WASHINGTON; MARIO
J. MOLINA; MICHAEL OPPENHEIMER; BOB KOPP; FRIEDERIKE OTTO;
SUSANNE C. MOSER; DONALD J. WUEBBLES; GARY GRIGGS; PETER C.
FRUMHOFF; KRISTINA DAHL; NATURAL RESOURCES DEFENSE
COUNCIL; ROBERT BRULLE; CENTER FOR CLIMATE INTEGRITY;
CHESAPEAKE CLIMATE ACTION NETWORK; JUSTIN FARRELL; BEN
FRANTA; STEPHAN LEWANDOWSKY; NAOMI ORESKES; GEOFFREY
SUPRAN; UNION OF CONCERNED SCIENTISTS; SCHOLARS OF FOREIGN
RELATIONS AND FEDERAL COURTS; STATE OF DELAWARE; STATE OF
HAWAI"I; STATE OF MAINE; STATE OF MINNESOTA; STATE OF NEW
MEXICO; COMMONWEALTH OF MINNESOTA; DISTRICT OF COLUMBIA,

Amici Supporting Appellee.

On Remand from the Supreme Court of the United States.
(S. Ct. No. 19-1189)

Argued: January 25, 2022 Decided: April 7, 2022

Before GREGORY, Chief Judge, THACKER, Circuit Judge, and FLOYD, Senior Circuit
Judge.

Affirmed by published opinion. Senior Judge Floyd wrote the opinion in which Chief
Judge Gregory and Judge Thacker joined.

2
ARGUED: Kannon K. Shanmugam, PAUL, WEISS, RIFKIND, WHARTON,
GARRISON, LLP, Washington, D.C., for Appellants. Victor Marc Sher, SHER EDLING
LLP, San Francisco, California, for Appellees. ON BRIEF: Theodore J. Boutrous, Jr.,
Los Angeles, California, Anne Champion, New York, New York, Joshua S. Lipshutz,
Thomas G. Hungar, GIBSON, DUNN & CRUTCHER LLP, Washington, D.C.; Ty Kelly,
Jonathan Biran, BAKER DONELSON, BEARMAN, CALDWELL & BERKOWITZ,
P.C., Baltimore, Maryland, for Appellants Chevron Corporation and Chevron U.S.A., Inc.
John B. Isbister, Jaime W. Luse, TYDINGS & ROSENBERG LLP, Baltimore, Maryland;
Philip H. Curtis, Nancy G. Milburn, New York, New York, Matthew T. Heartney, John D.
Lombardo, ARNOLD & PORTER KAYE SCHOLER LLP, Los Angeles, California, for
Appellants BP Products North America Inc., BP P.L.C., and BP America Inc. Craig A.
Thompson, VENABLE LLP, Baltimore, Maryland; Theodore V. Wells, Jr., Daniel J. Toal,
Jaren Janghorbani, PAUL, WEISS, RIFKIN, WHARTON, GARRISON LLP, New York,
New York, for Exxon Mobil Corporation and ExxonMobil Oil Corporation. David C.
Frederick, James M. Webster, III, Brendan J. Crimmins, Grace W. Knofczynski, Daniel S.
Severson, KELLOGG, HANSEN, TODD, FIGEL & FREDERICK, P.L.L.C., Washington,
D.C.; Daniel B. Levin, Los Angeles, California, Jerome C. Roth, Elizabeth A. Kim,
MUNGER, TOLLES & OLSON LLP, San Francisco, California, for Appellants Shell
USA, Inc. and Shell plc. Warren N. Weaver, Peter Sheehan, WHITEFORD TAYLOR
AND PRESTON LLP, Baltimore, Maryland; Nathan P. Eimer, Pamela R. Hanebutt, Ryan
Walsh, Raphael Janove, EIMER STAHL LLP, Chicago, Illinois, for Appellant Citgo
Petroleum Corporation. Michael A. Brown, NELSON MULLINS RILEY &
SCARBOROUGH LLP, Baltimore, Maryland; David B. Hamilton, Hillary V. Colonna,
Sarah E. Meyer, WOMBLE BOND DICKINSON LLP, Baltimore, Maryland; Sean C.
Grimsley, Jameson R. Jones, Daniel R. Brody, BARTLIT BECK LLP, Denver, Colorado,
for Appellants ConocoPhillips and ConocoPhillips Company. Steven M. Bauer, Margaret
A. Tough, LATHAM & WATKINS LLP, San Francisco, California, for Appellants
ConocoPhillips, ConocoPhillips Company, and Phillips 66. Jonathan Chunwei Su,
LATHAM & WATKINS LLP, Washington, D.C., for Appellant Phillips 66. Shannon S.
Broome, San Francisco, California, Shawn Patrick Regan, New York, New York, Ann
Marie Mortimer, HUNTON ANDREWS KURTH LLP, Los Angeles, California, for
Appellants Marathon Petroleum Corp. and Speedway, LLC. Scott Janoe, Houston, Texas,
Martha Thomsen, Megan Berge, Emily Wilson, BAKER BOTTS L.L.P., Washington,
D.C., for Appellant Hess Corp. Michelle N. Lipkowitz, Thomas K. Prevas, SAUL EWING
ARNSTEIN & LEHR LLP, Baltimore, Maryland, for Appellants Crown Central LLC and
Crown Central New Holdings LLC. Kathleen Taylor Sooy, Tracy Ann Roman,
Washington, D.C., Honor R. Costello, CROWELL & MORING LLP, New York, New
York, for Appellants CNX Resources Corporation, Consol Energy Inc., and Consol Marine
Terminals LLC. Noel J. Francisco, David M. Morrell, J. Benjamin Aguiñaga, Washington,
D.C., David C. Kiernan, JONES DAY, San Francisco, California, for Appellant CNX
Resources Corporation. Mark S. Saudek, GALLAGHER EVELIUS & JONES LLP,
Baltimore, Maryland; James Stengel, New York, New York, Robert Reznick, ORRICK,
HERRINGTON & SUTCLIFFE, LLP, Washington, D.C., for Appellants Marathon Oil

3
Corporation and Marathon Oil Company. Andre M. Davis, Suzanne Sangree, Sara Gross,
BALTIMORE CITY LAW DEPARTMENT, Baltimore, Maryland; Matthew K. Edling,
SHER EDLING LLP, San Francisco, California, for Appellee. Steven P. Lehotsky,
Michael B. Schon, Andrew R. Varcoe, Stephanie A. Maloney, U.S. CHAMBER
LITIGATION CENTER, Washington, D.C.; Peter D. Keisler, C. Frederick Beckner III,
Ryan C. Morris, Tobias S. Loss-Eaton, SIDLEY AUSTIN LLP, Washington, D.C.;
William M. Jay, Andrew Kim, GOODWIN PROCTER LLP, Washington, D.C., for
Amicus The Chamber of Commerce of the United States of America. Michael Burger,
Susan Kraham, MORNINGSIDE HEIGHTS LEGAL SERVICES, INC., New York, New
York, for Amici The National League of Cities, the U.S. Conference of Mayors, and the
International Municipal Lawyers Association. Scott L. Nelson, Allison M. Zieve, PUBLIC
CITIZEN LITIGATION GROUP, Washington, D.C., for Amicus Public Citizen, Inc.
Gerson H. Smoger, SMOGER & ASSOCIATES, P.C., Dallas, Texas; Robert S. Peck,
CENTER FOR CONSTITUTIONAL LITIGATION, P.C., Washington, D.C., for Amici
Senators Sheldon Whitehouse and Edward J. Markey. Brian E. Frosh, Attorney General,
Joshua M. Segal, Special Assistant Attorney General, Steven J. Goldstein, Special
Assistant Attorney General, OFFICE OF THE ATTORNEY GENERAL OF
MARYLAND, Baltimore, Maryland, for Amicus State of Maryland. Rob Bonta, Attorney
General, OFFICE OF THE ATTORNEY GENERAL OF CALIFORNIA, Sacramento,
California, for Amicus State of California. Kathleen Jennings, Attorney General, OFFICE
OF THE ATTORNEY GENERAL OF DELAWARE, Wilmington, Delaware, for Amicus
State of Delaware. William Tong, Attorney General, OFFICE OF THE ATTORNEY
GENERAL OF CONNECTICUT, Hartford, Connecticut, for Amicus State of Connecticut.
Karl A. Racine, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF THE
DISTRICT OF COLUMBIA, Washington, D.C., for Amicus District of Columbia. Clare
E. Connors, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF HAWAII,
Honolulu, Hawaii, for Amicus State of Hawaii. Maura Healey, Attorney General, OFFICE
OF THE ATTORNEY GENERAL OF MASSACHUSETTS, Boston, Massachusetts, for
Amicus Commonwealth of Massachusetts. Andrew J. Bruck, Acting Attorney General,
OFFICE OF THE ATTORNEY GENERAL OF NEW JERSEY, Trenton, New Jersey, for
Amicus State of New Jersey. Ellen F. Rosenblum, Attorney General, OFFICE OF THE
ATTORNEY GENERAL OF OREGON, Salem, Oregon, for Amicus State of Oregon.
Thomas J. Donovan, Jr., Attorney General, OFFICE OF THE ATTORNEY GENERAL
OF VERMONT, Montpelier, Vermont, for Amicus State of Vermont. Letitia James,
Attorney General, OFFICE OF THE ATTORNEY GENERAL OF NEW YORK, Albany,
New York, for Amicus State of New York. Aaron M. Frey, Attorney General, OFFICE
OF THE ATTORNEY GENERAL OF MAINE, Augusta, Maine, for Amicus State of
Maine. Keith Ellison, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF
MINNESOTA, Saint Paul, Minnesota, for Amicus State of Minnesota. Hector Balderas,
Attorney General, OFFICE OF THE ATTORNEY GENERAL OF NEW MEXICO, Santa
Fe, New Mexico, for Amicus State of New Mexico. Peter F. Neronha, Attorney General,
OFFICE OF THE ATTORNEY GENERAL OF RHODE ISLAND, Providence, Rhode
Island, for Amicus State of Rhode Island. Robert W. Ferguson, Attorney General, OFFICE

4
OF THE ATTORNEY GENERAL OF WASHINGTON, Olympia, Washington, for
Amicus State of Washington. William A. Rossbach, ROSSBACH LAW, PC, Missoula,
Montana, for Amici Mario J. Molina, Michael Oppenheimer, Bob Kopp, Friederike Otto,
Susanne C. Moser, Donald J. Wuebbles, Gary Griggs, Peter C. Frumhoff, and Kristina
Dahl. Peter Huffman, NATURAL RESOURCES DEFENSE COUNCIL, Washington,
D.C., for Amicus Natural Resources Defense Council. Mark A. Griffin, Amy Williams-
Derry, Daniel P. Mensher, Alison S. Gaffney, KELLER ROHRBACK L.L.P., Seattle,
Washington, for Amici Robert Brulle, Center for Climate Integrity, The Chesapeake
Climate Action Network, Justin Farrell, Benjamin Franta, Stephan Lewandowsky, Naomi
Oreskes, Geoffrey Supran, and the Union of Concerned Scientists. Theodore E. Roitka,
Attorney General, Thomas M. Fisher, Solicitor General, Kian J. Hudson, Deputy Solicitor
General, Julia C. Payne, Deputy Attorney General, OFFICE OF THE ATTORNEY
GENERAL OF INDIANA, Indianapolis, Indiana, for Amicus State of Indiana. Steve
Marshall, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF ALABAMA,
Montgomery, Alabama, for Amicus State of Alabama. Treg R. Taylor, Attorney General,
OFFICE OF THE ATTORNEY GENERAL OF ALASKA, Anchorage, Alaska, for
Amicus State of Alaska. Leslie Rutledge, Attorney General, OFFICE OF THE
ATTORNEY GENERAL OF ARKANSAS, Little Rock, Arkansas, for Amicus State of
Arkansas. Christopher Carr, Attorney General, OFFICE OF THE ATTORNEY
GENERAL OF GEORGIA, Atlanta, Georgia, for Amicus State of Georgia. Derek
Schmidt, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF KANSAS,
Topeka, Kansas, for Amicus State of Kansas. Daniel Cameron, Attorney General, OFFICE
OF THE ATTORNEY GENERAL OF KENTUCKY, Frankfort, Kentucky, for Amicus
Commonwealth of Kentucky. Lynn Fitch, Attorney General, OFFICE OF THE
ATTORNEY GENERAL OF MISSISSIPPI, Jackson, Mississippi, for Amicus State of
Mississippi. Austin Knudsen, Attorney General, OFFICE OF THE ATTORNEY
GENERAL OF MONTANA, Helena, Montana, for Amicus State of Montana. Doug
Peterson, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF NEBRASKA,
Lincoln, Nebraska, for Amicus State of Nebraska. Alan Wilson, Attorney General,
OFFICE OF THE ATTORNEY GENERAL OF SOUTH CAROLINA, Columbia, South
Carolina, for Amicus State of South Carolina. Ken Paxton, Attorney General, OFFICE OF
THE ATTORNEY GENERAL OF TEXAS, Austin, Texas, for Amicus State of Texas.
Sean D. Reyes, Attorney General, OFFICE OF THE ATTORNEY GENERAL OF UTAH,
Salt Lake City, Utah, for Amicus State of Utah. Bridget Hill, Attorney General, OFFICE
OF THE ATTORNEY GENERAL OF WYOMING, Cheyenne, Wyoming, for Amicus
State of Wyoming. Linda E. Kelly, Patrick Hedren, Erica Klenicki, MANUFACTURERS’
CENTER FOR LEGAL ACTION, Washington, D.C., for Amicus National Association of
Manufacturers. Philip S. Goldberg, Christopher E. Appel, SHOOK HARDY & BACON
L.L.P., Washington, D.C., for Amici National Association of Manufacturers, National
Association of Convenience Stores, Society of Independent Gasoline Marketers of
America, and Energy Marketers of America. Matthew D. Hardin, HARDIN LAW
OFFICE, Washington, D.C., for Amicus Energy Policy Advocates. Anna C. Haac, Hassan
A. Zavareei, TYCKO & ZAVAREEI LLP, Washington, D.C., for Amici Scholars of

5
Foreign Relations and Federal Courts.

6
FLOYD, Senior Circuit Judge:

This appeal returns to us on remand from the Supreme Court, and we are now

tasked with examining the entirety of the district court’s remand order to determine if the

climate-change lawsuit in question was properly removed to federal court. BP P.L.C. v.

Mayor & City Council of Balt., 141 S. Ct. 1532, 1538, 1543 (2021). To accomplish that

charge, we must evaluate eight distinct grounds for removal that twenty-six multinational

oil and gas companies (Defendants) 1 maintain provide federal jurisdiction over the Mayor

and City Council of Baltimore’s (Baltimore) climate-change action. Because we conclude

that none of Defendants’ bases for removal permit the exercise of federal jurisdiction, we

affirm the district court’s remand order.

I.

A.

In July 2018, Baltimore filed suit in the Circuit Court for Baltimore City against

Defendants. According to Baltimore, Defendants substantially contributed to greenhouse-

gas pollution, global warming, and climate change by extracting, producing, promoting,

1
Defendants consist of BP entities (BP P.L.C.; BP America, Inc.; and BP Products North
America Inc.); Crown Central entities (Crown Central Petroleum Corporation; Crown
Central LLC; and Crown Central New Holdings LLC); Chevron entities (Chevron Corp.
and Chevron U.S.A. Inc.); Exxon Mobil entities (Exxon Mobil Corp. and ExxonMobil Oil
Corporation); Shell entities (Shell PLC and Shell USA, Inc.); Citgo Petroleum Corp.;
ConocoPhillips entities (ConocoPhillips; ConocoPhillips Company; Louisiana Land &
Exploration Co.; Phillips 66; and Phillips 66 Company); Marathon entities (Marathon Oil
Company; Marathon Oil Corporation; Marathon Petroleum Corporation; and Speedway
LLC); Hess Corp.; and CONSOL entities (CNX Resources Corporation; CONSOL Energy
Inc.; and CONSOL Marine Terminals LLC).
7
refining, marketing, distributing, and selling fossil-fuel products (i.e., coal, oil, and natural

gas). Baltimore asserts that Defendants deceived consumers and the public about the

dangers associated with their fossil-fuel products when they knew, for nearly fifty years,

of a direct link between their products and climate-change threats. With that knowledge,

as Baltimore alleges, Defendants (1) employed a “coordinated, multi-front effort to conceal

and deny their own knowledge of those threats”; (2) discredited “publicly available

scientific evidence”; and (3) created persistent doubt within the public sphere about the

“reality and consequences of the impacts of their fossil[-]fuel pollution.” J.A. 43. But that

is not all. Baltimore’s Complaint emphasizes that Defendants’ other actions also

contributed to climate change and Baltimore’s own harms: “Defendants individually and

collectively manufactured, promoted, marketed, and sold a substantial percentage of all

fossil[-]fuel products ultimately used and combusted.” J.A. 139 (emphasis added).

Resulting from Defendants’ collective conduct, Baltimore avers it has suffered

“climate[-]change-related injuries,” including “sea level rise and associated impacts,

increased frequency and severity of extreme precipitation events, increased frequency and

severity of drought, increased frequency and severity of heat waves and extreme

temperatures, and consequent social and economic injuries associated with those physical

and environmental changes . . . .” J.A. 92, 140–41 (emphasis added). Within Baltimore’s

boundaries, these environmental events have purportedly caused, among other things,

infrastructure damage during floods, automobile accidents and power outages when winter

storms hit, and public-health illnesses amid heat waves.

8
 Essentially, Baltimore’s Complaint seeks to shift the burden of its climate-change

costs onto Defendants: “[Baltimore] seeks to ensure that the parties who have profited

from externalizing the responsibility for sea level rise, extreme precipitation events,

heatwaves, other results of the changing hydrologic regime caused by increasing

temperatures, and associated consequences of those physical and environmental changes,

bear the costs of those impacts on . . . [Baltimore] . . . .” J.A. 47. Baltimore, however,

“does not seek to impose liability on Defendants for their direct emissions of greenhouse

gases and does not seek to restrain Defendants from engaging in their business operations.”

J.A. 47.

Baltimore brings eight causes of action against Defendants, all under Maryland law:

(1) public nuisance; (2) private nuisance; (3) strict liability for failure to warn; (4) strict

liability for design defect; (5) negligent design defect; (6) negligent failure to warn; (7)

trespass; and (8) violations of the Maryland Consumer Protection Act (MCPA), Md. Code

Ann., Com. Law §§ 13-101 to -501. Each of Baltimore’s claims are factually premised on

Defendants’ “superior knowledge” of the negative, climate-change impacts attributable to

their fossil-fuel products. J.A. 150. And that “superior knowledge” stems from

Defendants’ control over the “extraction, refining, development, marketing, and sale of

[their] fossil[-]fuel products.” J.A. 150 (public nuisance); see also J.A. 156 (alleging, for

private nuisance, Defendants possessed “extensive knowledge” of their fossil-fuel

products’ hazards); J.A. 157, 166 (maintaining, for strict liability for failure to warn and

negligent failure to warn, Defendants breached a duty of care by failing to adequately warn

about the “climate effects that inevitably flow from the intended use of their fossil[-]fuel

9
products” when they had “information passed to them from their internal research

divisions”); J.A. 160 (asserting, for strict liability for design defect, “Defendants had

control over . . . the manufacturing and distribution processes”); J.A. 163 (contending, for

negligent design defect, Defendants allowed their fossil-fuel products to enter the stream

of commerce “despite knowing them to be defective”); J.A. 168 (asserting, for trespass,

flood waters entered Baltimore’s real property because of Defendants’ fossil-fuel products

and their knowledge of those products); J.A. 171 (alleging Defendants violated the MCPA

by making (1) false and misleading statements, and (2) false representations and misleading

omissions about their fossil-fuel products).

To remedy its harms, Baltimore seeks compensatory and punitive damages,

disgorgement of profits, civil penalties under the MCPA, and equitable relief, including the

abatement of the alleged nuisances and an injunction against future nuisances.

B.

After Baltimore’s suit was filed in state court in July 2018, Defendants Chevron

Corp. and Chevron U.S.A. timely removed Baltimore’s Complaint to the United States

District Court for the District of Maryland. Chevron asserted eight different grounds for

removal under statutory grants of federal jurisdiction and various legal theories, including:

(1) federal common law; (2) substantial issues of federal law, as well as foreign affairs,

under Grable & Sons Metal Products, Inc. v. Darue Engineering & Manufacturing, 545

U.S. 308 (2005); (3) complete preemption under the Clean Air Act (CAA), 42 U.S.C. §§

7401–7671q; (4) federal enclaves; (5) the Outer Continental Shelf Lands Act (OCSLA),

10
43 U.S.C. § 1349(b)(1); (6) the bankruptcy removal statute, 28 U.S.C. § 1452(a); (7) the

admiralty jurisdiction statute, 28 U.S.C. § 1333(1); and (8) the federal officer removal

statute, 28 U.S.C. § 1442(a)(1).

Responding to Chevron’s removal, Baltimore filed a motion to remand its

Complaint back to state court. After considering the parties’ filings, on June 10, 2019, the

district court granted Baltimore’s Motion to Remand in a forty-five-page order and opinion,

rejecting each of Defendants’ eight grounds for removal. Mayor & City Council of Balt.

v. BP P.L.C., 388 F. Supp. 3d 538 (D. Md. 2019). Defendants appealed the district court’s

remand order to this Court. Mayor & City Council of Balt. v. BP P.L.C., 952 F.3d 452 (4th

Cir. 2020). We reasoned that, under 28 U.S.C. § 1447(d), we could only analyze the

propriety of removal under the federal officer removal statute and lacked appellate

jurisdiction over the remaining seven grounds for removal. Id. at 461. We ultimately held

that federal officer removal was improper and affirmed the district court’s remand order

on that sole ground. Id. at 461–70.

Defendants appealed to the Supreme Court. The Supreme Court held that,

under § 1447(d), this Court is not divested of appellate jurisdiction over Defendants’ other

theories of removal and may consider all the bases for removal included within the district

court’s remand order. BP P.L.C., 141 S. Ct. at 1538, 1543. When vacating our opinion

and remanding for further proceedings, the Court declined Defendants’ invitation to

analyze their remaining removal bases and found that the “wiser course” was for this Court

to examine them in the first instance. Id. at 1543. The Court did not address our rejection

of Defendants’ invocation of the federal officer removal statute. Id. at 1543.

11
 Following the Court’s holding and mandate, we now evaluate the remaining theories

of removal Defendants proffer. Since Defendants relied upon the federal officer removal

statute as a path to federal court, we possess appellate jurisdiction to review the entirety of

the district court’s remand order under § 1447(d). BP P.L.C., 141 S. Ct. at 1538.

II.

Under the general removal statute, “any civil action brought in a State court of which

the district courts of the United States have original jurisdiction, may be removed by the

defendant or the defendants, to the district court of the United States for the district and

division embracing the place where such action is pending.” 28 U.S.C. § 1441(a). In turn,

federal district courts typically have original jurisdiction over cases involving federal

questions “arising under the Constitution, laws, or treaties of the United States.”

Id. § 1331. We refer to jurisdiction under § 1331 as federal-question jurisdiction. See

McCormick v. Am. Online, Inc., 909 F.3d 677, 679 (4th Cir. 2018).

“We review de novo issues of subject matter jurisdiction, including removal.”

Common Cause v. Lewis, 956 F.3d 246, 252 (4th Cir. 2020) (citation omitted); see also

Prince v. Sears Holdings Corp., 848 F.3d 173, 176 (4th Cir. 2017). “The party seeking

removal bears the burden of showing removal is proper.” Prince, 848 F.3d at 176 (citation

omitted); see also Strawn v. AT & T Mobility LLC, 530 F.3d 293, 297 (4th Cir. 2008).

“Because removal jurisdiction raises significant federalism concerns, we must strictly

construe removal jurisdiction.” Mulcahey v. Columbia Organic Chems. Co., Inc., 29 F.3d

12
148, 151 (4th Cir. 1994) (citation omitted). “If federal jurisdiction is doubtful, a remand is

necessary.” Id. (citations omitted).

III.

Before considering the merits of Defendants’ grounds for removal, we must look to

two legal doctrines that inform any removal inquiry before a federal court: (1) the well-

pleaded complaint rule, and (2) complete preemption. See Rivet v. Regions Bank of La.,

522 U.S. 470, 475–76 (1998). We recount them in turn.

First, “[t]he well-pleaded complaint rule applies to the original jurisdiction of the

district courts as well as to their removal jurisdiction.” Franchise Tax Bd. of Cal. v. Constr.

Laborers Vacation Tr. for S. Cal., 463 U.S. 1, 10 n.9 (1983) (citations omitted). When

applying the well-pleaded complaint rule to removal and federal-question jurisdiction, we

have stated that “courts ‘ordinarily . . . look no further than the plaintiff’s [properly

pleaded] complaint in determining whether a lawsuit raises issues of federal law capable

of creating federal-question jurisdiction under 28 U.S.C. § 1331.’” Pinney v. Nokia, Inc.,

402 F.3d 430, 442 (4th Cir. 2005) (alteration in original) (quoting Custer v. Sweeny, 89

F.3d 1156, 1165 (4th Cir. 1996)). In numerous cases, the Supreme Court has reiterated

that federal-question jurisdiction “must be determined from what necessarily appears in the

plaintiff’s statement of his own claim in the bill or declaration, unaided by anything alleged

in anticipation of avoidance of defenses which it is thought the defendant may interpose.”

Aetna Health Inc. v. Davila, 542 U.S. 200, 207 (2004) (quoting Taylor v. Anderson, 234

U.S. 74, 75–76 (1914)); see also Beneficial Nat’l Bank v. Anderson, 539 U.S. 1, 6 (2003);

13
Okla. Tax Comm’n v. Graham, 489 U.S. 838, 840–41 (1989); Metro. Life Ins. Co. v. Taylor,

481 U.S. 58, 63 (1987).

Using this well-known principle in practice, federal courts must first decide whether

federal or state law creates the cause of action by viewing the face of a plaintiff’s complaint.

Pinney, 402 F.3d at 442. If federal law, as opposed to state law, creates a plaintiff’s cause

of action, then removal is proper. Id. “The general rule, of course, is that a plaintiff is the

‘master of the claim,’ and he may ‘avoid federal jurisdiction by exclusive reliance on state

law’ in drafting his complaint.” Id. (quoting Caterpillar Inc. v. Williams, 482 U.S. 386,

392 (1987)). A plaintiff’s complaint “may not be removed to federal court on the basis of

a federal defense, including the defense of pre-emption, even if the defense is anticipated

in the plaintiff’s complaint, and even if both parties concede that the federal defense is the

only question truly at issue.” Caterpillar, 482 U.S. at 393 (citation omitted); see also

Burrell v. Bayer Corp., 918 F.3d 372, 381 (4th Cir. 2019).

Second, the doctrine of complete preemption, unlike ordinary preemption, is a

recognized exception to the well-pleaded complaint rule. See Aetna Health, 542 U.S. at

207–08. Complete preemption is an “independent corollary” to the well-pleaded complaint

rule. Franchise Tax Bd., 463 U.S. at 22. As a jurisdictional doctrine, “[it] has the effect

of ‘transform[ing]’ a state-law cause of action into one arising under federal law because

Congress has occupied the field so thoroughly as to leave no room for state-law causes of

action at all.” Johnson v. Am. Towers, LLC, 781 F.3d 693, 702 (4th Cir. 2015) (second

alteration in original) (quoting Caterpillar, 482 U.S. at 399). Complete preemption thus

treats a state-law cause of action as based on a federal statute “in reality,” meaning that,

14
under such circumstances, there is “no such thing” as that state-law claim. Beneficial Nat’l

Bank, 539 U.S. at 8, 11. In contrast, ordinary preemption is not a jurisdictional doctrine

because it “simply declares the primacy of federal law, regardless of the forum or the

claim.” 2 Lontz v. Tharp, 413 F.3d 435, 440 (4th Cir. 2005) (citation omitted). Ordinary

preemption is a federal defense to a plaintiff’s claims, and it cannot serve as a valid basis

for removal. Caterpillar, 482 U.S. at 393; see also Skidmore v. Norfolk S. Ry. Co., 1 F.4th

206, 211–12 (4th Cir. 2021). Indeed, we have recognized that civil defendants “may not

defend [their] way into federal court” as a way to bypass the well-pleaded complaint rule.

In re Blackwater Sec. Consulting, LLC, 460 F.3d 576, 584 (4th Cir. 2006). On the other

hand, if a court instead decides that a state-law cause of action is completely preempted by

federal law, removal is proper. Lontz, 413 F.3d at 439–40; see also Rivet, 522 U.S. at 475

(noting that the “artful pleading doctrine” permits removal when a federal statute

“completely preempts” a state-law claim).

2
Under the Supremacy Clause, federal law is the “supreme Law of the Land.” U.S. Const.
art. VI, cl. 2. There are three types of ordinary preemption: (1) express preemption; (2)
conflict preemption; and (3) field preemption. See Murphy v. Nat’l Collegiate Athletic
Ass’n, 138 S. Ct. 1461, 1480 (2018); see also W. Star Hosp. Auth. Inc. v. City of Richmond,
986 F.3d 354, 360 (4th Cir. 2021). In contrast to complete preemption, these three
preemption doctrines serve as substantive defenses and do not implicate federal
jurisdiction. See Whitehurst v. 1199SEIU United Healthcare Workers E., 928 F.3d 201,
206 n.2 (2d Cir. 2019); Retail Prop. Tr. v. United Brotherhood of Carpenters & Joiners of
Am., 768 F.3d 938, 948–49 (9th Cir. 2014); see also Lontz, 413 F.3d at 440. Because we
are only concerned with removal jurisdiction and complete preemption’s application, we
need not to delve into these defenses at Defendants’ disposal. See Beneficial Nat’l Bank,
539 U.S. at 6 (noting that federal defenses do not oust jurisdiction).

15
 As we have explained, “[c]omplete preemption applies only when ‘Congress has

clearly manifested an intent to make causes action . . . removable to federal court.’”

Johnson, 781 F.3d at 702 (quoting Metro. Life, 481 U.S. at 66). The congressional intent

to displace state law “must be clear in the text of the statute.” Lontz, 413 F.3d at 441 (citing

Metro. Life, 481 U.S. at 65–66). Given that exacting requirement, there is a rebuttable

presumption against finding the complete preemption of state-law claims. Id. at 440. And

the presumption can only be overcome if the removing party satisfies its “significant

burden” of “establish[ing] [a] congressional intent to extinguish similar state claims by

making the federal cause of action exclusive.” Id. at 441. In sum, we thus permit complete-

preemption findings when: “(1) the preempting statute displays a clear congressional intent

to ‘entirely displace’ state law; and (2) the preempting statute creates an exclusive federal

cause of action in an area of ‘overwhelming national interest.’” Norfolk S. Ry., 1 F.4th at

212 (quoting Lontz, 413 F.3d at 441).

IV.

With those legal principles informing our removal inquiry, we now examine, in turn,

Defendants’ eight grounds for removal: (1) federal common law; (2) substantial issues of

federal law, including foreign affairs, under Grable; (3) complete preemption under the

CAA, 42 U.S.C. §§ 7401–7671q; (4) federal enclaves; (5) the OCSLA, 43

U.S.C. § 1349(b)(1); (6) the bankruptcy removal statute, 28 U.S.C. § 1452(a); (7) the

admiralty jurisdiction statute, 28 U.S.C. § 1333(1); and (8) the federal officer removal

statute, 28 U.S.C. § 1442(a)(1).

16
 A.

As their primary vehicle for federal jurisdiction, Defendants insist that Baltimore’s

Complaint is “necessarily and exclusively governed by federal common law.” Defs.’

Suppl. Br. 3; see also Defs.’ Opening Br. 15–22. According to Defendants, even though

the Complaint never says anything about federal common law, Baltimore’s claims are

“inherently federal and necessarily arise under federal law because they seek to impose

liability based on the production and sale of oil and gas abroad.” Defs.’ Suppl. Br. at 8.

They specifically characterize Baltimore’s claims as “interstate-pollution claims” that arise

under federal common law. Id. at 16; see also id. at 3, 7–8, 19 (likening Baltimore’s causes

of action to interstate and/or international pollution). Defendants never point to the specific

cause of action under federal common law. Unsurprisingly, Baltimore stresses its suit has

“nothing to do with any body of federal common law.” Baltimore’s Suppl. Br. 7. For the

reasons set forth below, we resoundingly agree with Baltimore and reject Defendants’

attempts to invoke federal common law.

1.

At the outset, we note that Baltimore’s Complaint never expressly asserts any claim

under federal common law. And Defendants do not contest otherwise. Because

Baltimore’s Complaint does not propose a new federal cause of action, never alleges an

existing federal common law claim, and only brings claims originating under Maryland

law, the district court never had subject-matter jurisdiction under the well-pleaded

17
complaint rule. See Columbia Gas Transmission v. Singh, 707 F.3d 583, 588–89 (6th Cir.

2013) (holding that a federal district court did not possess subject-matter jurisdiction via

federal common law when a plaintiff did not “clearly seek recovery under federal law” and

merely brought a “state-law claim with a federal ingredient”). Although we discern no

federal common law claim from Baltimore’s Complaint, we nevertheless consider whether

the creation of a federal rule of decision is justified and if a pre-existing federal rule of

decision already applies because of Defendants’ unusual arguments. 3 See O’Melveny &

Myers v. Fed. Deposit Ins. Corp., 512 U.S. 79, 87 (1994).

2.

We begin with a well-known principle: “There is no federal general common law.”

Erie R. Co. v. Tompkins, 304 U.S. 64, 78 (1938). In decision after decision, the Supreme

Court has reiterated that federal common law does not exist wholesale. See Rodriguez v.

Fed. Deposit Ins. Corp., 140 S. Ct. 713, 717 (2020) (“As this Court has put it, there is ‘no

federal general common law.’” (quoting Erie, 304 U.S. at 78)); Hernandez v. Mesa, 140 S.

Ct. 735, 742 (2020) (“[F]ederal courts today cannot fashion new claims in the way that

they could before 1938.” (citation omitted)); Alexander v. Sandoval, 532 U.S. 275, 287

(2001) (“Raising up causes of action where a statute has not created them may be a proper

3
We use the term “federal rule of decision” synonymously with “federal common law.”
See Baker, Watts & Co. v. Miles & Stockbridge, 876 F.2d 1101, 1106 (4th Cir. 1989); see
also McGurl v. Trucking Emps. of N. Jersey Welfare Fund, Inc., 124 F.3d 471, 480 (3d Cir.
1997).

18
function for common-law courts, but not for federal tribunals.” (citation omitted)). But

despite that understanding, federal common law still exists in narrow areas involving: (1)

“the rights and obligations of the United States”; (2) “interstate and international disputes

implicating the conflicting rights of States or our relations with foreign nations”; and (3)

“admiralty cases.” Tex. Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 641 (1981)

(citations omitted); see also Rodriguez, 140 S. Ct. at 717 (noting that federal common law

exists in “admiralty disputes and certain controversies between States”); Atherton v. Fed.

Deposit Ins. Corp., 519 U.S. 213, 215–16, 225–26 (1997) (collecting cases where the Court

has created federal common law and ultimately declining to fashion a general standard of

care for officers and directors of federally insured institutions).

Federal courts should be reluctant to displace state law through federal common law

because displacement is typically a legislative decision for Congress. Atherton, 519 U.S.

218. Regardless, before a federal court can promulgate a federal rule of decision, a dispute

must satisfy two strict conditions: (1) there must be “uniquely federal interests” at play,

and (2) a party must show “a ‘significant conflict’ . . . between an identifiable ‘federal

policy or interest and the [operation] of state law . . . or the application of state law would

‘frustrate specific objectives’ of federal legislation.” Boyle v. United Techs. Corp., 487

U.S. 500, 504, 507 (1988) (second alteration in original) (citations omitted); see also

Rodriguez, 140 S. Ct. at 717–18 (declining to find a federal interest in how a consolidated

corporate tax refund is distributed among corporate group members); Atherton, 519 U.S.

at 219–26 (finding no “significant conflict with, or threat to, a federal interest” after

comparing the federal interests to state laws addressing corporate governance);

19
O’Melveny, 512 U.S. at 88 (refusing to create a federal rule of decision when a party

“identified no significant conflict with an identifiable federal policy or interest” and there

was no federal interest in “uniformity”); Radcliff Materials, 451 U.S. at 642 (rejecting a

federal interest in creating a federal common law right to contribution). The showing of a

“[significant] conflict” is normally a “precondition” to creating federal common law.

Atherton, 519 U.S. at 218 (quoting O’Melveny, 512 U.S. at 87); see also Kamen v. Kemper

Fin. Servs., Inc., 500 U.S. 90, 98 (1991); United States v. Kimbell Foods, Inc., 440 U.S.

715, 728 (1979).

Generally, if these requirements for expanding federal common law are satisfied,

then a federal district court possesses original jurisdiction over a well-pled federal common

law claim under § 1331, making removal jurisdiction proper. Nat’l Farmers Union Ins.

Cos. v. Crow Tribe of Indians, 471 U.S. 845, 850 (1985) (“It is well settled that [§ 1331’s]

grant of ‘jurisdiction will support claims founded upon federal common law as well as

those of a statutory origin.’” (quoting Illinois v. City of Milwaukee, 406 U.S. 91, 100

(1972))).

3.

At the outset, although Baltimore argued that Defendants failed to show a

“significant conflict” with any federal interest, we note that neither Defendants nor the

district court truly grappled with the federal common law inquiry using the cited precedents

20
from the Supreme Court. 4 Compare Baltimore’s Resp. Br. 27 n.4, and Baltimore’s Suppl.

Br. 7–9, with BP P.L.C., 388 F. Supp. 3d at 554–58, and Defs.’ Opening Br. 19–30, and

Defs.’ Suppl. Br. 12–13. Other than passively referencing Boyle and Radcliff Materials,

Defendants and the district court never mentioned Rodriguez, Atherton, or O’Melveny.

Instead, they immediately proceeded to the Court’s authorities dealing with global

warming and interstate pollution. Given the novelty of Baltimore’s Complaint about fossil-

fuel products and the Court’s precedents addressing federal common law for interstate and

greenhouse-gas pollution, see infra Part IV.A.4, we deem it prudent to first address whether

it is even appropriate to create federal common law for the issues raised in Baltimore’s

Complaint. See Martinson v. Kinney Shoe Corp., 104 F.3d 683, 686 (4th Cir. 1997) (noting

a district court’s error and explaining “we can affirm if its decision was correct for any

other reason” (citation omitted)); AFA Distrib. Co., Inc. v. Pearl Brewing Co., 470 F.2d

1210, 1211 (4th Cir. 1973) (vacating a district court’s decision interpreting a statute but

affirming the dismissal of a complaint for “other reasons”). We do not believe that it is.

Defendants identify three “uniquely federal interests” at play: (1) the control of

interstate pollution; (2) energy independence; and (3) multilateral treaties. Defs.’ Opening

Br. 15. Assuming these qualify as “uniquely federal interests,” Defendants’ request for

federal common law still fails because they do not satisfy the necessary “precondition” of

4
Baltimore’s Motion to Remand identified that a significant conflict needed to be shown
for the district court to create federal common law. See Baltimore’s Motion to Remand at
17, Mayor & City Council of Balt. v. BP P.L.C., 388 F. Supp. 3d 538 (D. Md. 2019) (No.
1:18-cv-02357-ELH), ECF No. 111-1.

21
creating federal common law—the recognition of a significant conflict between a federal

interest and state law’s application. Atherton, 519 U.S. at 218 (quoting O’Melveny, 519

U.S. at 87). Defendants, who bear the removal burden, never establish a significant conflict

between Baltimore’s state-law claims—which purport to impose liability on Defendants

for their marketing and use of their fossil-fuel products—and any federal interests within

either their Notice of Removal or Opening Brief. See J.A. 12–14 (explaining, in

Defendants’ Notice of Removal, that Baltimore’s claims only “implicate inherently

national and international interests” or “implicate[] inherently federal concerns” without

identifying any conflict); Defs.’ Opening Br. 19–30 (acknowledging the Supreme Court’s

significant-conflict requirement but failing to identify a significant conflict). And in both

of their submitted Replies, Defendants do not even use the word “conflict.” See Defs.’

Reply 8–12; Defs.’ Suppl. Reply 2–8. Under our precedents, Defendants’ failure to argue

a “significant conflict” between Baltimore’s causes of action and its identified federal

interests constitutes a waiver. See Grayson O Co. v. Agadir Int’l LLC, 856 F.3d 307, 316

(4th Cir. 2017) (“A party waives an argument by failing to present it in its opening

brief . . . .” (citation omitted)); United States v. Caldwell, 7 F.4th 191, 212 n.16 (4th Cir.

2021) (“Any other arguments raised for the first time in [a] Reply Brief, however, we deem

waived.” (citation omitted)). But Defendants’ neglect is more than just a waiver of an

argument, it substantively precludes the creation of federal common law. As the Supreme

Court put it, failing to identify a significant conflict when requesting a court to create

federal common law is “fatal” to a party’s position. See O’Melveny, 512 U.S. at 88 (“What

is fatal to respondent’s position in the present case is that it has identified no significant

22
conflict with an identifiable federal policy or interest.”). Given these failures, we see no

reason to fashion any federal common law for Defendants.

From what we can discern, Defendants seem to rely upon City of New York v.

Chevron Corp., 993 F.3d 81, 92 (2d Cir. 2021), to now suggest that Baltimore’s claims

present a conflict between the rights of States and the federal government’s international

relations. 5 See Defs.’ Suppl. Br. 11. In City of New York, the Second Circuit affirmed the

dismissal of a climate-change suit brought by New York City, under Federal Rule of Civil

Procedure 12(b)(6), after characterizing the complaint as seeking liability for “global

greenhouse gas emissions” and “the effects of emissions made around the globe over the

past several hundred years.” 993 F.3d at 89, 91–92, 103. City of New York does not pertain

to the issues before us.

First and foremost, City of New York was in a completely different procedural

posture. Id. at 93–94. As the decision itself concedes, the court was not required to

consider a “heightened standard unique to the removability inquiry” because New York

City initially filed suit in federal court as opposed to state court. Id. at 94. Indeed, the

Second Circuit confined itself to Rule 12(b)(6) and never addressed its own subject-matter

jurisdiction. Id. at 89. The suit before us was initiated in state court, so we are bound by

the well-pleaded complaint rule or “heightened standard” that did not apply in City of New

5
Because City of New York is included in Defendants’ Supplemental Brief and Reply, and
since that decision was decided while the parties were litigating in the Supreme Court, we
exercise our discretion to evaluate any holding that Defendants might be relying upon from
City of New York. Caldwell, 7 F.4th at 212 n.16; see also O’Melveny, 512 at 88
(considering the “closest” that a party “[came] to identifying a specific, concrete federal
policy or interest that [was] compromised by California law”).
23
York. Id. at 94. And under the well-pleaded complaint rule, we find Baltimore’s suit

centers on Defendants’ fossil-fuel products and misinformation campaign, not any federal

common law. See J.A. 150–71 (setting out allegations about Defendants’ fossil-fuel

products). Second, City of New York suffers from the same legal flaw as Defendants’

arguments: It fails to explain a significant conflict between the state-law claims before it

and the federal interests at stake before arriving at its conclusions. See id. at 90–93. For

instance, after recognizing federalism and the need for a uniform rule of decision as federal

interests, City of New York confusingly concludes that federal common law is “most needed

in this area” because New York’s state-law claims touch upon the federal government’s

relations with foreign nations. 6 Id. at 91–92. But it never details what those foreign

relations are and how they conflict with New York’s state-law claims. See id. at 92. The

same is true when City of New York declares that state law would “upset[] the careful

balance” between global warming’s prevention and energy production, economic growth,

foreign policy, and national security. Id. at 93. Besides referencing statutes acknowledging

policy goals, the decision does not mention any obligatory statutes or regulations

explaining the specifics of energy production, economic growth, foreign policy, or national

security, and how New York law conflicts therewith. See id. It also does not detail how

those statutory goals conflict with New York law. See id. City of New York essentially

evades the careful analysis that the Supreme Court requires during a significant-conflict

6
We note that uniformity—in and of itself—is not always a federal interest, and City of
New York discounts contrary precedent. See Empire Healthchoice Assurance, Inc. v.
McVeigh, 547 U.S. 677, 691 (2006); Atherton, 519 U.S. at 220; O’Melveny, 512 U.S. at
88.
24
analysis. Cf. Atherton, 519 U.S. at 219–26 (declining to find a significant conflict between

a federal policy and state-law standards of care after thoroughly examining uniformity,

history, the internal-affairs doctrine of corporate governance, agency opinions and

statements, and whether the federal government’s interest was being pursued); Boyle, 487

U.S. at 511–12 (finding, in the context of government procurement, that a federal statutory

provision demonstrated “the potential for, and suggest[ed] the outlines of,” a significant

conflict between federal interests and state law when the federal provision implied that the

“appropriate design for military equipment . . . [was] assuredly a discretionary function”).

It is for these two critical reasons that we cannot follow City of New York and find federal

jurisdiction at this juncture.

In short, we decline to create a federal rule of decision that would apply to

Baltimore’s claims since Defendants do not point to any significant conflict existing

between Maryland law and their purported federal interests, which is a complete abdication

of their removal burden. Prince, 848 F.3d at 176. Setting aside Defendants’ misstep, even

if they provided us with a significant conflict between Maryland law and a federal interest

that would justify a new federal rule of decision, the well-pleaded complaint rule would

still forbid the removal of Baltimore’s Complaint because it pleads no express invocation

of federal common law. Pinney, 402 F.3d at 442.

4.

Rather than grappling with the threshold inquiry above, Defendants invoke the

Supreme Court’s older authorities that once (or possibly) recognized federal common law

25
in the context of interstate pollution and greenhouse-gas emissions. Defendants present a

perplexing argument that Baltimore’s claims must be resolved by federal common law

because it is the source of the underlying claims. 7 Defs.’ Opening Br. 16–19; Defs.’ Suppl.

Br. 17–18. Baltimore responds that any federal common law in this area is nonexistent

because the CAA statutorily displaced federal common law claims. Baltimore’s Resp. Br.

24–28. We cannot conclude that any federal common law controls Baltimore’s state-law

claims because federal common law in this area ceases to exist due to statutory

displacement, Baltimore has not invoked the federal statute displacing federal common

law, and, as we later find, the CAA does not completely preempt Baltimore’s claims.

We begin with the precedents available to us. In 1972, the Supreme Court famously

stated that “[w]hen we deal with air and water in their ambient or interstate aspects, there

is a federal common law . . . .” City of Milwaukee, 406 U.S. at 103. Employing federal

common law, City of Milwaukee specifically recognized public nuisance claims for

disputes involving interstate and navigable waters. Id. at 103–04. The Court approved of

a public nuisance claim for Illinois when it sued municipalities and public sewerage

7
Defendants rely upon United States v. Standard Oil Co. of California, 332 U.S. 301
(1947), arguing that we must first determine the source of law for Baltimore’s claims.
Standard Oil speaks to the threshold question of whether to even create a federal common
law claim: “And the answer to be given necessarily is dependent upon a variety of
considerations always relevant to the nature of the specific governmental interests and to
the effects upon them of applying state law.” Id. at 310 (emphasis added). And as we have
decided, Defendants do not meaningfully grapple with the significant-conflict inquiry to
invoke the governance of federal common law. See supra Part IV.A.3. In any event,
Standard Oil does not aid Defendants. The decision did not turn on federal-question
jurisdiction, and the Court declined the opportunity to create a federal cause of action
sounding in indemnity for the federal government. 332 U.S. at 310–17.
26
commissions located within Wisconsin. Id. at 93–94. Unlike the case here, private

defendants were not being sued when the Court initially recognized public nuisance as a

federal common law claim. Nine years after City of Milwaukee, Congress passed the

Federal Water Pollution Control Act of 1942 (FWPCA, Clean Water Act, or CWA), and

the Court then held that the FWPCA displaced the federal common law claim of public

nuisance it had previously recognized for water pollution. City of Milwaukee v. Illinois,

451 U.S 304, 312–20 (1981); see also Middlesex Cnty. Sewerage Auth. v. Nat’l Sea

Clammers Ass’n, 453 U.S. 1, 21–22 (1981) (noting that the FWPCA “entirely preempted”

the federal common law of nuisance for water pollution involving ocean waters). These

issues were not grappled with again until decades later.

In American Electric Power Co., Inc. v. Connecticut, 564 U.S. 410 (2011), the Court

considered whether a group of plaintiffs, including eight States, New York City, and three

private land trusts, “[could] invoke the federal common law of nuisance to abate out-of-

state pollution” and impose liability on five electric companies for global warming. Id. at

418, 422. Recognizing that its precedents had once approved of federal common law for

suits “brought by one State to abate pollution emanating from another State,” the Court

emphasized that federal courts should not create federal common law simply because a

subject is amenable to governance under federal law, especially when there is no

“demonstrated need for a federal rule of decision.” Id. at 421–22 (emphasis added)

(collecting cases). And it also noted that it had not yet decided whether a political

subdivision, like Baltimore here, could even invoke federal common law to abate out-of-

state pollution. Id. at 422. While the Court expressed uncertainty about the existence of

27
federal common law for the plaintiffs, it ultimately left this “academic question” for

another day. Id. at 423. In the alternative, the Court found that any “federal common-law

claim for curtailment of greenhouse gas emissions because of their contribution to global

warming” was “displaced by the federal legislation authorizing [the Environmental

Protection Agency] to regulate carbon-dioxide emissions.” Id. The federal legislation

displacing the federal common law at issue was the CAA. See id. at 424 (“We hold that

the Clean Air Act and the EPA actions it authorizes displace any federal common-law right

to seek abatement of carbon-dioxide emissions from fossil-fuel fired powerplants.”).

There are a few things we learn from these cases that the parties rely upon. First,

although the terms have been used interchangeably by federal courts, there is a significant

distinction between the statutory displacement of federal common law and the ordinary

preemption of a state law. 8 City of Milwaukee, 451 U.S. at 316–17; see also Merrick v.

Diageo Americas Supply, Inc., 805 F.3d 685, 693–94 (6th Cir. 2015); United States v. Am.

Com. Lines, L.L.C., 759 F.3d 420, 422 n.1 (5th Cir. 2014) . When a federal statute displaces

federal common law, the federal common law ceases to exist. See City of Milwaukee, 451

U.S. at 317, 332 (concluding that “no federal common-law remedy was available” when it

was statutorily displaced by congressional amendments to the FWPCA); Nat’l Sea

8
Defendants cannot argue that federal common law completely preempts Baltimore’s
claims because the Supreme Court has only applied complete preemption in the context of
federal statutes, not federal common law. See Metro. Life, 481 U.S. at 63–64 (“Congress
may so completely pre-empt a particular area that any civil complaint raising this select
group of claims is necessarily federal in character.” (emphasis added)). There cannot be a
congressional intent to completely preempt via federal common law since it is created by
federal judges and not Congress. See id.

28
Clammers, 453 U.S. at 21–22 (noting that federal common law was “entirely preempted”

by the FWPCA and dismissing a federal common law claim for having no “underlying

legal basis”); see also Native Vill. of Kivalina v. ExxonMobil Corp., 696 F.3d 849, 857 (9th

Cir. 2012) (“Judicial power can afford no remedy unless a right that is subject to that power

is present. If a federal common law cause of action has been extinguished by Congressional

displacement, it would be incongruous to allow it to be revived in another form.”). But

when a state law is ordinarily preempted by a federal statute, the federal statute supplants

and supersedes the state law without extinguishing it. See Am. Elec. Power, 564 U.S. at

429; Am. Com. Lines, 759 F.3d at 422 n.1; S. Blasting Servs., Inc. v. Wilkes Cnty., 288 F.3d

584, 590 (4th Cir. 2002). Preemption requires a “clear and manifest purpose” from

Congress, while displacement does not. City of Milwaukee, 451 U.S. at 316–17. Secondly,

federal common law claims of public nuisance, at least for water and air pollution, have

been displaced by the CWA and CAA. See Am. Elec. Power, 564 U.S. at 423; Nat’l Sea

Clammers, 453 U.S. at 21–22; City of Milwaukee, 451 U.S at 312–20; see also City of

Oakland v. BP PLC, 969 F.3d 895, 906 (9th Cir. 2020). This all means something simple.

Public nuisance claims involving interstate pollution, including issues about greenhouse-

gas emissions, are nonexistent under federal common law because they are statutorily

displaced. In other words, a federal statute is the legal source of those claims, and a federal

common law remedy is unavailable. City of Milwaukee, 451 U.S. at 332.

Defendants seek removal through an extraordinary means in their attempt to use

federal common law. Essentially, Defendants believe that removal is proper based on

federal common law even when the federal common law claim has been deemed displaced,

29
extinguished, and rendered null by the Supreme Court. We believe that position defies

logic. The Court has previously emphasized that “federal courts are without power to

entertain claims otherwise within their jurisdiction if they are so attenuated and

unsubstantial as to be absolutely devoid of merit; wholly insubstantial; obviously frivolous;

plainly unsubstantial; or no longer open to discussion.” Hagans v. Lavine, 415 U.S. 528,

536–37 (1974) (citations omitted) (cleaned up). Again, Defendants repeatedly characterize

Baltimore’s claims as interstate-pollution claims. But due to statutory displacement,

federal common law claims concerning interstate pollution and the regulation of

greenhouse-gas emissions are now obsolete. See Am. Elec. Power, 564 U.S. at 423; Nat’l

Sea Clammers, 453 U.S. at 21–22; City of Milwaukee, 451 U.S at 312–20. Contrary to

Defendants’ insistence, it is “no longer open to discussion” that federal common law claims

even exist to govern Baltimore’s claims. Hagans, 415 U.S. at 537; see also City of

Oakland, 969 F.3d at 906 (noting that federal common law claims for public nuisance are

displaced by the CAA in a similar suit). Since those claims are defunct, and invoking them

is “devoid of merit,” a federal court cannot exercise federal-question jurisdiction on that

basis. See Dixon v. Coburg Dairy, Inc., 369 F.3d 811, 817 n.5 (4th Cir. 2004). Tellingly,

Defendants cite no authority justifying removal for nonexistent claims that have been

displaced by federal statutes. If anything, case law suggests that the displacement of

federal common law deprives federal courts of jurisdiction. See Native Village of Kivalina,

696 F.3d at 855, 858 (affirming a district court’s dismissal for lack of subject-matter

jurisdiction because the “federal common law addressing domestic greenhouse gas

emissions has been displaced by Congressional action”). Thus, we will not provide

30
Defendants with the unprecedented opportunity to obtain removal based on a nonexistent

theory of federal common law when its viability is “no longer open to discussion” as a

means of federal relief. Hagans, 415 U.S. at 537; see also Fitz Gerald v. Thompson, 222

U.S. 555, 557 (1912) (dismissing a writ of error when there was no federal jurisdiction

because “[t]he right to remove from the state court which was asserted had no legal

foundation” and was “manifestly frivolous and devoid of merit”).

In sum, we do not accept the governance of federal common law when the CWA

and CAA have statutorily displaced any federal common law that previously existed, and

Baltimore’s Complaint does not desire relief under either of those federal statutes. See

Pinney, 402 F.3d at 442. If we found federal common law as a valid removal basis in this

case, we would first undercut the well-pleaded complaint rule by ignoring Baltimore’s

pleaded claims and then undermine complete preemption by disregarding what that

separate inquiry later requires of us. See Bd. of Cnty. Comm’rs of Boulder Cnty. v. Suncor

Energy (U.S.A.) Inc., 25 F.4th 1238, 1257–65 (10th Cir. 2022) (rejecting Defendants’

invocation of federal common law as a basis for removal and considering complete

preemption under the CAA). We decline to endorse those outcomes.

5.

And finally, Defendants insist that we are bound to follow Caudill v. Blue Cross &

Blue Shield of North Carolina, 999 F.2d 74 (4th Cir. 1993), and North Carolina

Department of Administration v. Alcoa Power Generating, Inc., 853 F.3d 140 (4th Cir.

31
2017). Both of those decisions are readily distinguishable from the circumstances before

us.

In Caudill, a federal employee brought a breach-of-contract action against her

insurer concerning her coverage under a contract between the insurer and federal

government. 999 F.2d at 76–77. We held that federal common law governed the contract,

and removal was thus proper under federal-question jurisdiction. Id. at 77–79. But the

Supreme Court rejected Caudill’s use of federal common law as a basis for federal-question

jurisdiction in Empire Healthchoice Assurance, Inc. v. McVeigh, 547 U.S. 677, 689–93

(2006), when it concluded there was no showing of a conflict between a federal interest

and state law that mandated the formulation of federal common law. In light of Empire,

Caudill’s holding about the existence of federal common law for interpreting federal

contracts involving health insurance is now abrogated. Compare id., with Caudill, 999

F.2d at 76–77. We thus decline to follow reasoning that is outright overturned.

In Alcoa Power, North Carolina sought a declaration against a power company

concerning its ownership of a forty-five-mile segment of the Yadkin River. 853 F.3d at

143–45. According to North Carolina, the power company acquired the segment by deed

and, on the segment, constructed four hydroelectric dams to power its smelting plant. Id.

North Carolina maintained that the power company was only using the riverbed segment

of the Yadkin River with its permission and that permission was withdrawn after the

company decided to permanently close its smelting plant and layoff its employees. Id.

North Carolina originally sought its declaration in state court as a state-law claim to quiet

title, but the power company removed the suit to federal court, contending that the issue of

32
navigability for title was a federal question. Id. at 145–46. On appeal, we held that the

district court possessed federal-question jurisdiction, warranting removal, because state

ownership of the beds of navigable waters relies on the Constitution. Id. at 147–50. When

the Supreme Court emphatically declared that “questions of navigability for determining

state riverbed title are governed by federal law,” we reasoned that the Court was

“reaffirming the federal nature of the issue of navigability for title” and recognizing its

precedents from over 150 years. Id. at 148 (quoting PPL Mont., LLC v. Montana, 565 U.S.

576, 591 (2012)). Alcoa Power is inapplicable here. Unlike Alcoa Power, Defendants do

not rely on any constitutional provision suggesting federal law applies to or governs

Baltimore’s claims. And particularly given American Electric Power’s holding,

Defendants certainly cannot point this Court to over 150 years of precedent recognizing

the federal character of Baltimore’s claims. Even more, Baltimore’s Complaint is not

concerned with a declaration of title to any navigable water owned or occupied by

Defendants. Alcoa Power provides little in the way of principle that governs here.

6.

Compared to state common law, federal common law is extremely limited. Radcliff

Materials, 451 U.S. at 640 (observing that areas involving federal common law are “few

and restricted” (quoting Wheeldin v. Wheeler, 373 U.S. 647, 651 (1963))). State law has

traditionally governed the realm of products liability and continues to do so here. See Pac.

Atl. Trading Co., Inc. v. M/V Main Express, 758 F.2d 1325, 1130 n.1 (9th Cir. 1985); In re

“Agent Orange” Prod. Liab. Litig., 635 F.2d 987, 991 n.9, 993–94 (2d Cir. 1980).

33
Defendants have failed to show that federal common law truly controls this dispute

involving their fossil-fuel products and misinformation campaign. At the second oral

argument in this case, that failure became even more evident when Defendants did not even

identify what federal common law claim is at Baltimore’s disposal and what their own

defense to it would be. At most, Defendants present us with an ordinary preemption

argument that does not warrant removal. See Caterpillar, 482 U.S. at 393. Regardless of

how they frame their invocation, we decline to permit Defendants to rely upon federal

common law as a theory for removal and affirm the district court’s sound rejection thereof.

B.

Defendants next seek to establish federal jurisdiction under Grable and its progeny,

arguing “[s]everal aspects of [Baltimore]’s claims” present substantial and disputed federal

issues. Defs.’ Opening Br. 33. Here, Defendants argue those federal issues include

national security, foreign affairs, energy policy, and environmental regulation. Baltimore,

however, posits that Defendants “dramatically overread[]” Grable’s scope. Baltimore’s

Resp. Br. 33. We agree with Baltimore and find Defendants’ invocation of Grable

jurisdiction and the foreign-affairs doctrine fails to pass legal muster.

1.

There is a “‘slim category’ of cases . . . in which state law supplies the cause of

action but federal courts have jurisdiction under § 1331 because ‘the plaintiff’s right to

relief necessarily depends on resolution of a substantial question of federal law.’” Burrell,

34
918 F.3d at 380 (citations omitted). Federal courts must be “cautious” in exercising this

form of jurisdiction because it lies at the “outer reaches of § 1331.” Id. (quoting Merrell

Dow Pharms. Inc. v. Thompson, 478 U.S. 804, 810 (1986)). The Supreme Court has

emphasized that the “mere presence of a federal issue in a state cause of action does not

automatically confer federal-question jurisdiction.” Merrell Dow, 478 U.S. at 813

(emphasis added) (citations omitted).

To ensure complaints alleging only state-law claims are not in federal court when

they merely implicate federal issues, the Supreme Court established a four-prong test for

determining the existence of federal-question jurisdiction. See Grable, 545 U.S. at 314.

Federal-question jurisdiction exists over a state-law claim if a federal issue is: “(1)

necessarily raised, (2) actually disputed, (3) substantial, and (4) capable of resolution in

federal court without disrupting the federal-state balance approved by Congress.” Gunn v.

Minton, 568 U.S. 251, 258 (2013) (emphasis added). A federal question is “necessarily

raised” under § 1331 “only if it is a ‘necessary element of one of the well-pleaded state

claims.’” Burrell, 918 F.3d at 381 (quoting Franchise Tax Bd., 463 U.S. at 13). A federal

issue is “actually disputed” when the parties disagree about the effect of federal law. See

Gunn, 568 U.S. at 259. The substantiality question “looks . . . to the importance of the

issue to the federal system as a whole.” Id. at 260. And, to strike a balance between federal

and state judicial responsibilities, a federal court must ensure that it gives leeway to States

in areas where they possess “special responsibilit[ies].” Id. at 264 (quoting Ohralik v. Ohio

State Bar Ass’n, 436 U.S. 447, 460 (1978)).

35
 2.

Looking at the face of Baltimore’s Complaint, Grable jurisdiction cannot lie

because a federal issue is not “necessarily raised.” Id. at 258. Grable jurisdiction thus fails

on the very first prong. 9

Again, a federal issue is “necessarily raised” only when a federal question is a

“necessary element” of one of the pleaded state-law claims within a plaintiff’s complaint.

Burrell, 918 F.3d at 381 (quoting Franchise Tax Bd., 463 U.S. at 13). The Court’s

precedents indicate when this requirement is satisfied. For instance, in Grable, the

Supreme Court held that a quiet-title action under Michigan law “necessarily raised”

federal issues because the plaintiff premised its state-law claim on the Internal Revenue

Service’s failure to comply with notification requirements established by federal law. 545

U.S. at 310, 314–15. Similarly, the Gunn Court held that a legal-malpractice claim under

Texas law “necessarily require[d] application of [federal] patent law to the facts of [the]

case” since the state-law claim required a showing of prevailing in a federal patent

infringement action. 568 U.S. at 259.

We have adhered to the Court’s guidance by looking for federal ingredients that are

“necessary” for the state-law claim’s success. For example, in Pinney, we examined seven

state-law claims, all under the laws of different States, to conclude that federal law was not

a “necessary element” for any of the state-law claims and they only required the “resolution

9
Because a federal issue is not “necessarily raised” by Baltimore’s Complaint, we need
not address the remaining factors of Grable jurisdiction. See Burrell, 918 F.3d at 384, 386.

36
of questions of state law.” 402 F.3d at 442–46. Most recently, in West Virginia State

University Board of Governors v. Dow Chemical Co., a historically black college sued a

chemical company for contaminating the groundwater beneath the land it owned. 23 F.4th

288, 292–94 (4th Cir. 2022). The university’s claims were brought exclusively under West

Virginia common law. Id. at 296. We held that federal issues were not “necessarily raised”

or even “substantially raised” because the college did not challenge a “cleanup” order under

the Comprehensive Environmental Response, Compensation, and Liability Act, 42 U.S.C.

§ 9613(b), (h), and its state-law claims were not preempted by the Resource Conservation

and Recovery Act’s savings clause, id. § 6972. 23 F.4th at 307–12. These cases

demonstrate that state-law claims must “hinge on the determination of a federal issue” to

fulfill Grable’s first prong. Vlaming v. W. Point Sch. Bd., 10 F.4th 300, 306 (4th Cir. 2021).

Defendants never identify what federal question is a “necessary element” for any of

Baltimore’s state-law claims. 10 See Defs.’ Opening Br. 33–40. All of Baltimore’s claims

are brought under Maryland law, and none of them invoke federal law as a necessary

requirement for imposing liability upon Defendants. See J.A. 149–72. Thus, Defendants’

liability does not turn or “hinge” upon interpreting federal law. Cf. Bauer v. Elrich, 8 F.4th

291, 297 (4h Cir. 2021) (holding that plaintiffs “necessarily raised” a federal issue because

they sought to enforce a federal statute and did not advance a state-law right). Failing to

10
We note that City of Oakland considered federal common law in the context of Grable
jurisdiction in a similar suit involving Defendants. 969 F.3d at 906. In the case before us,
Defendants have not asked us to analyze federal common law under Grable, so we deviate
from City of Oakland in this respect. See Defs.’ Opening Br. 33–40.

37
carry their removal burden, Defendants provide us with no federal question Baltimore has

alleged that is “essential to resolving” its claims under Maryland law. Burrell, 918 F.3d at

383 (citation omitted).

Read most generously, Defendants’ Opening Brief maintains that federal agencies

typically weigh the costs and benefits of fossil-fuel extraction, so Baltimore’s nuisance

claims “invite a state court factfinder [to] adjudicate the reasonableness of . . . federal

agencies’ balancing of harms and benefits.” Defs.’ Opening Br. 34–35. But this argument

first rests on a misunderstanding of Baltimore’s Complaint. Baltimore essentially

challenges the efficacy and safety of Defendants’ fossil-fuel products and sales practices

promoting them. See J.A. 150–71. The Complaint is not solely about the initial act of

fossil-fuel extraction, nor is it concerned with setting and regulating greenhouse-gas

emissions. See J.A. 47.

Defendants’ argument then misapprehends the elements of public and private

nuisance under Maryland law. “A public nuisance is an injury to the public at large or to

all persons who come in contact with it,” while “[a] private nuisance is injury to an

individual or a limited number of individuals only.” Adams v. Comm’rs of Trappe, 102

A.2d 830, 834 (Md. 1954). Typically, “[a p]rivate nuisance is ‘a nontrespassory invasion

of another’s interest in the private use and enjoyment of land.’” Exxon Mobil Corp. v.

Albright, 71 A.3d 30, 94 (Md. 2013) (quoting Wietzke v. Chesapeake Conf. Ass’n, 26 A.3d

931, 943 (Md. 2011)). The distinction between a private and public nuisance turns on

whether the impacted rights are “confined to private ownership or are cast broadly across

the general public . . . .” Wietzke, 26 A.3d at 943 (citation omitted).

38
 Adopting the Second Restatement of Torts, Maryland courts require a public

nuisance to involve an “unreasonable interference” with the public’s rights. Tadjer v.

Montgomery Cnty., 479 A.2d 1321, 1327 (Md. 1984) (quoting Restatement (Second) of

Torts § 821(B) (Am. L. Inst. 1979)). Circumstances showing an “unreasonable

interference” may include: (1) “[w]hether the conduct involves a significant interference

with the public health, the public safety, the public peace, the public comfort or the public

convenience”; (2) “whether the conduct is proscribed by a statute, ordinance or

administrative regulation”; or (3) “whether the conduct is of a continuing nature or has

produced a permanent or long-lasting effect, and, as the actor knows or has reason to know,

has a significant effect upon the public right.” Id. (citation omitted). Claimants can point

to any or all of those three circumstances when attempting to prove the “unreasonable-

interference” element of a public nuisance. They can avoid federal law entirely, for

example, if they show harmful conduct either involving a “significant interference” with

the public’s safety or producing a “permanent or long-lasting effect.” See id. Neither of

those avenues require federal law as a “necessary element.”

It is true that the Second Restatement of Torts indicates that the “unreasonable-

interference” question may be fulfilled by showing the conduct at issue is proscribed by “a

statute, ordinance or administrative regulation.” Id. (emphasis added). So claimants may

invoke a federal law or regulation to show that there is an “unreasonable interference” with

the public’s rights. But that is discretionary and not a “necessary element.” Without

resorting to any federal law, Plaintiffs can also utilize a state law or regulation when

showing an “unreasonable interference” with the public’s rights. Maryland courts agree.

39
See Raynor v. Md. Dep’t of Health & Mental Hygiene, 676 A.2d 978, 990–91 (Md. Ct.

Spec. App. 1996) (finding a Maryland regulation “[did] no more than prohibit or abate a

public nuisance” when examining the Takings Clause under the Fifth Amendment); see

also Comm’rs of Trappe, 102 A.2d at 836–37.

A private nuisance also requires a claimant to establish an “unreasonable and

substantial interference” with the use of his or her private property. Exxon Mobil, 71 A.3d

at 94. Like public nuisances, that element of a private-nuisance claim may also be proven

by exclusively using state statutes or regulations. 11 See Wietzke, 26 A.3d at 942, 947. Since

neither public nor private nuisances “necessarily raise” federal law as a “necessary

element,” we find that federal agencies’ balancing of the harms and benefits of fossil-fuel

extraction is not “necessary” for proving either claim. See Pinney, 402 F.3d at 449 (“The

Supreme Court has been quite clear that for removal to be proper under the substantial

federal question doctrine, a plaintiff’s ability to establish the necessary elements of his state

law claims must rise or fall on the resolution of a question of federal law.” (citing Merrell

Dow, 478 U.S. at 813)); see also Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1266–

67.

11
We agree with Defendants that Maryland law permits a factfinder to balance competing
property interests for the “unreasonable-interference” question for private nuisances. See
Wietzke, 26 A.3d at 942–47. However, we disagree that a Maryland factfinder, during its
context-specific inquiry, would have to “necessarily” consider federal law when balancing
property interests and deciding if Defendants’ conduct was an “unreasonable-interference.”
See id. at 944 (“And it is equally true, that the mere lawfulness of the act is not in itself a
test in all cases, of exemption from liability for [private nuisance].” (quoting Short v. Balt.
City Passenger Ry. Co., 50 Md. 73, 81 (1878))); id. at 947 (describing the multiple factors
to consider for an “unreasonable-interference” determination for private nuisance).
40
 Defendants also assert that Baltimore’s “promotion claims implicate federal duties

to disclose . . . .” Defs.’ Opening Br. 37 (emphasis added). We must reject this argument.

Despite possessing the removal burden, Defendants do not tell us which of Baltimore’s

causes of action under Maryland law is a “promotion claim.” See id. But even if we assume

Defendants are referring to Baltimore’s claims involving strict liability for failure to warn,

negligent failure to warn, and the MCPA, they have not identified how any of those claims

require federal law as a “necessary element” for their resolution. 12 See id. And we have

not found any federal law or issue that is raised by the elements of those state-law claims.

See Pinney, 402 F.3d at 446–49 (finding a federal issue was not “necessarily raised” by

design-defect claims when federal radiation standards were only “one factor” for

establishing liability, and liability could still be found even with regulatory compliance);

see also Carmine v. Poffenbarger, 154 F. Supp. 3d 309, 312 n.1, 314–17 (E.D. Va. 2015)

(declining to find federal issues were “necessarily raised” by failure-to-warn and design-

defect claims because a federal question was not an element of either claim).

We cannot find that Baltimore’s Complaint “necessarily raises” any question of

federal law as envisioned by the Supreme Court. It is a far cry from what the Court has

deemed sufficient to satisfy the “necessarily raised” prong. Cf. Gunn, 568 U.S. at 259;

Grable, 545 U.S. at 310, 314–15. Accordingly, federal-question jurisdiction does not lie

since Baltimore’s Complaint is not one of those “slim category” of cases warranting Grable

12
Defendants have waived any such arguments since they were not raised in their Opening
Brief. See Grayson O Co., 856 F.3d at 316.

41
jurisdiction. See Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1265–71 (rejecting

Defendants’ invocation of Grable jurisdiction); City of Oakland, 969 F.3d at 906–07

(holding Grable jurisdiction was improper because a federal issue was not “necessarily

raised” by a complaint originally filed in state court against Defendants); Flying Pigs, LLC

v. RRAJ Franchising, LLC, 757 F.3d 177, 182–83 (4th Cir. 2014) (finding Grable

jurisdiction was unwarranted because a federal issue was not “necessarily raised” when a

claimant never sued under the Lanham Act but brought a state-court proceeding to enforce

an equitable lien). We thus affirm the district court’s rejection of Grable jurisdiction as a

basis for removal.

3.

Defendants wrongly rely on the foreign-affairs doctrine in the Grable context for

federal jurisdiction. 13 Stating that “[t]he question of how to address climate change has

long been and remains the subject of international negotiations[,]” Defendants assert that

Baltimore wants to “replace . . . international negotiations and congressional and executive

decisions with Maryland common law and private litigation in state court.” Defs.’ Opening

13
It is unclear whether Defendants intend to invoke the foreign-affairs doctrine. See Defs.’
Opening Br. 38–39. But they seem to appeal to it by citing to American Insurance
Association v. Garamendi, 539 U.S. 396 (2003). See id.

42
Br. 38–39. According to Defendants, Maryland law must yield to the federal government’s

international policies. We are not persuaded.

Under our Grable inquiry, there is nothing in Baltimore’s Complaint indicating that

foreign affairs are “necessarily raised” by its state-law claims. See City of Oakland, 969

F.3d at 906–07 (concluding Defendants’ argument about foreign policy did not raise a

substantial question of federal law for Grable jurisdiction). While the Complaint contains

historical references to international treaties in a brief section, see J.A. 114, 123, there is

no indication that Baltimore’s state-law claims either rise or fall based on any foreign

policies, international treaties, or relationships with foreign nations. “The most one can

say is that a question of [foreign affairs] is lurking in the background . . . .” Gully v. First

Nat’l Bank, 299 U.S. 109, 117 (1936).

In any case, Defendants suggest that the foreign-affairs doctrine preempts

Baltimore’s Complaint. “Under the foreign[-]affairs doctrine, state laws that intrude on

this exclusively federal power are [constitutionally] preempted.” Movsesian v. Victoria

Versicherung AG, 670 F.3d 1067, 1071 (9th Cir. 2012). This is so because the power to

conduct international affairs is solely vested with the federal government, not the States.

See United States v. Pink, 315 U.S. 203, 233–34 (1942); see also Hines v. Davidowitz, 312

U.S. 52, 63 (1941) (“The Federal Government . . . is entrusted with full and exclusive

responsibility for the conduct of affairs with foreign sovereignties.”). The foreign-affairs

doctrine may constitutionally preempt state laws through conflict preemption or field

preemption. Movsesian, 670 F.3d at 1071–72. For a state law to give way under conflict

preemption, there must be a “sufficiently clear conflict” between the state law and an

43
express foreign policy. See Am. Ins. Ass’n v. Garamendi, 539 U.S. 396, 420–24 (2003);

see also Gingery v. City of Glendale, 831 F.3d 1222, 1228–29 (9th Cir. 2016). Field

preemption applies “in the absence of a treaty” and when a state law or policy “disturb[s]

foreign relations” or if a State attempts to “establish its own foreign policy.” Zschernig v.

Miller, 389 U.S. 429, 441 (1968). Field preemption asks whether the state law “has more

than ‘some incidental or indirect effect in foreign countries[]’ . . . .” Id. at 434.

Taking up conflict preemption, Defendants do not identify any express foreign

policy from the federal government that conflicts with Baltimore’s state-law claims. See

Defs.’ Opening Br. 38–39. At best, Defendants reference the Kyoto Protocol, signed in

1997 by President Clinton but never ratified by the United States Senate. Nathan

Richardson, The Rise and Fall of Clean Air Act Climate Policy, 10 Mich. J. Env’t & Admin.

L. 69, 75 (2020); see also Massachusetts v. Env’t Prot. Agency, 549 U.S 497, 509 (2007).

At worst, they point us to a slew of remarks from Presidents Ford, Carter, Reagan, H.W.

Bush, Clinton, W. Bush, Obama, and Trump. See Defs.’ Opening Br. 39 (citing J.A. 265–

69). In and of themselves, those remarks are not explicit foreign policies that may create

a conflict. See Garamendi, 539 U.S. at 401, 413–24 (considering the preemption of

California’s Holocaust Victim Insurance Relief Act of 1999, Cal. Ins. Code §§ 13800–07,

when the Executive Branch signed agreements with foreign nations). And those statements

do not establish that Maryland common law, or even the common law of States generally,

is an obstacle to the federal government’s dealings with foreign nations. See Crosby v.

Nat’l Foreign Trade Council, 530 U.S. 363, 366–68, 383–85 (2000) (noting and relying

on the Executive Branch’s direct remarks about a Massachusetts statute barring the

44
purchase of goods from those “doing business with Burma”). Once again, in so much as

Defendants fail to identify a conflict, the Court cannot find that they have carried their

removal burden. Prince, 848 F.3d at 176.

As to field preemption, we do not believe that Baltimore’s claims are precluded on

this basis either. Defendants have not articulated how Baltimore’s common law claims

serve as Baltimore’s assertion of its own foreign policy. See Defs.’ Opening Br. 38–39.

In Zschernig, an Oregon statute permitted escheat 14 when “nonresident alien[s] claim[ed]

real or personal property” and three conditions were satisfied. 389 U.S. at 430–31. The

Court held that the Oregon statute unconstitutionally intruded upon the field of foreign

affairs, reasoning that the statute required state courts to delve into the “actual

administration of foreign law, . . . credibility of diplomatic statements, and . . . speculation

[concerning] the fact that some received delivery of funds should ‘not preclude

wonderment as to how many may have been denied ‘the right to receive’ . . . .” Id. at 435

(citations omitted). The Court went on to examine how Oregon courts were addressing

foreign relations, holding that their statute was impacting foreign relations “in a persistent

and subtle way.” Id. at 435–41.

Despite bearing the removal burden, Defendants have not provided us with even

one decision from Maryland courts showing how any of Baltimore’s state-law claims entail

14
Generally, escheat permits a State to take custody or assume title of abandoned property
when a person dies without leaving the property to any heirs and that property is located
within that State. See Delaware v. New York, 507 U.S. 490, 497 (1993); Texas v. New
Jersery, 379 U.S. 674, 675 (1965).

45
foreign relations. Even more importantly for field preemption, Defendants have not at all

explained how common law claims under state law meaningfully “disturb foreign

relations,” nor have they delineated how Baltimore’s claims are an attempt to “establish its

own foreign policy.” Id. at 441. Baltimore’s Complaint does not contain any allegations

that develop foreign policies with other countries, and nor does it undermine the federal

government in the international arena. At best, it involves an intersection between

Maryland law and private, international companies. See Md. Code Ann., Corps. & Ass’ns

§ 7-105 (“By doing intrastate, interstate, or foreign business in this State, a foreign

corporation assents to the laws of this State.” (emphasis added)). Thus, we find no

persuasive reason to apply field preemption.

At bottom, we decline to apply the foreign-affairs doctrine as either a constitutional

bar to Baltimore’s Complaint or a valid means for removal under Grable jurisdiction. Our

conclusion neatly aligns with our sister circuits’ approach of applying the foreign-affairs

doctrine to disputes only with direct impacts on foreign relations. Cf. City of Glendale,

831 F.3d at 1229–31 (holding the foreign-affairs doctrine does not preclude a local

government’s expression, in the form of a monument, about foreign affairs); Movsesian,

670 F.3d at 1070, 1075–77 (concluding the foreign-affairs doctrine applied when a state

statute permitted state courts to entertain insurance claims of “Armenian Genocide

victim[s]” against insurers covering persons and property in Europe and Asia between 1875

and 1923); Deutsch v. Turner Corp., 324 F.3d 692, 703–04, 708–16 (9th Cir. 2003)

(applying the foreign-affairs doctrine to a state statute creating a cause of action against

46
corporations for employing slave labor during the Second World War because it intruded

upon the federal government’s power to resolve war claims for committed wrongs).

C.

Next is Defendants’ argument that Baltimore’s Complaint is completely preempted

by the CAA. This argument fails as well.

As we have already stated, complete preemption requires “the congressional intent

that state law be entirely displaced . . . be clear in the text of the statute.” Lontz, 413 F.3d

at 441 (citing Metro. Life, 481 U.S. at 65–66). The federal statute must show “Congress

intended it to ‘provide the exclusive cause of action’ for claims of overwhelming national

interest.” Id. (quoting Beneficial Nat’l Bank, 539 U.S. at 9, 11). To date, the Supreme

Court has only applied complete preemption to three federal statutes: (1) §§ 85 and 86 of

the National Bank Act of 1863; (2) § 502 of the Employee Retirement Income Security Act

of 1974 (ERISA); and (3) § 301 of the Labor Management Relations Act of 1947 (LMRA).

See N.J. Carpenters & the Trs. Thereof v. Tishman Constr. Corp. of N.J., 760 F.3d 297,

302 (3d Cir. 2014) (collecting cases); see also Lontz, 413 F.3d at 441 (same). We have

extended complete preemption to certain state-law claims implicating § 301(a) of the

Copyright Act and § 10501(b) of the Interstate Commerce Termination Act. See

Rosciszewski v. Arete Assocs., Inc., 1 F.3d 225, 230–33 (4th Cir. 1993) (section 301(a) of

the Copyright Act); Skidmore, 1 F.4th at 212–17 (section 10501(b) of the Interstate

Commerce Termination Act).

47
 We turn to the history and text of the CAA as required by our complete-preemption

inquiry. The CAA was enacted in 1963, and Congress declared that its express purpose

was to “protect the Nation’s air resources so as to promote the public health and welfare

and the productive capacity of its population . . . .” Clean Air Act, Pub. L. No. 88-206, § 1,

77 Stat. 392, 393 (1963) (codified as amended at 42 U.S.C. § 7401(b)(1)). In 1990,

Congress further recognized that “air pollution prevention (that is, the reduction or

elimination, through any measures, of the amount of pollutants produced or created at the

source) and air pollution control at its source is the primary responsibility of the States and

local governments . . . .” 15 Clean Air Act Amendments, Pub. L. No. 101-549, § 108, 104

Stat. 2399, 2468 (1990) (emphasis added) (codified at 42 U.S.C. § 7401(a)(3)). The CAA

regulates air pollution from stationary sources, emission standards for moving sources,

noise pollution, acid raid, and stratospheric ozone protection. 42 U.S.C. §§ 7401–7515,

7521–7590, 7641–7642, 7651–7651o, 7671–7671q. It also provides a means for citizen

suits and outlines a permitting process for emission standards. Id. §§ 7604, 7661–7661f.

The Supreme Court has suggested that the CAA has force under ordinary preemption

principles and not under complete preemption principles. See Am. Elec. Power Co., 564

U.S. at 429 (“In light of our holding that the Clean Air Act displaces federal common law,

the availability vel non of a state law depends, inter alia, on the preemptive effect of the

federal Act.” (citing Int’l Paper Co. v. Ouellette, 479 U.S. 481, 489, 491, 497 (1987))).

15
This language appeared in the CAA when it was first passed, excluding the parenthetical
addressing the “reduction or elimination” of pollutants. See Clean Air Act § 1, 77 Stat. at
393.

48
 First, primarily relying on § 7607(b)(1), Defendants correctly point out that the

CAA allows parties to challenge various actions, regulations, and standards promulgated

by the Administrator of the Environmental Protection Agency (EPA).

However, § 7607(b)(1) has nothing to do with lawsuits against private parties as it

explicitly authorizes a means for judicial review of the EPA’s final actions and rulemaking.

Baltimore’s Complaint does not ask a court to review the legality of any of the EPA’s

decisions or regulations. 16 And as we have explained above and more fully below, its state-

law claims do not involve the regulation of emissions. Baltimore’s causes of action simply

do not “duplicate[], supplement[], or supplant[]” § 7607(b)(1)’s procedures for obtaining

judicial review of actions by the EPA. See Aetna Health, 542 U.S. at 209. Rather than

establishing an exclusive federal scheme through § 7607(b)(1), Congress

intended § 7401(a)(3) to vest state and local governments with the “primary responsibility”

of controlling and preventing air pollution. State and local governments could not have

any “primary responsibility” over air pollution if their laws were completely preempted

and replaced by a federal regime. § 7401(a)(3).

Second, “[t]he presence of a savings clause counsels against a finding that Congress

intended to sweep aside all state claims in a particular area.” Pinney, 402 F.3d at 450. The

16
According to Defendants, the State of Maryland and other municipalities have availed
themselves of § 7607(b)(1)’s procedures by bringing actions against the EPA. See Defs.’
Opening Br. 49 n.14. This suggests that Baltimore is only required to do the same if it
wants a federal court to review the EPA’s actions or rulemaking process. § 7607(b)(1).
Baltimore is not seeking such relief. If anything, this indicates that Baltimore’s Complaint
is seeking a different relief since it has not joined the State of Maryland in those lawsuits.

49
CAA contains two savings clauses that preserve state and local governments’ legal right to

impose standards and limitations on air pollution that are stricter than national

requirements. See §§ 7416, 7604(e). One concerns citizen suits under the CAA and

specifically counsels: “Nothing in this section shall restrict any right which any

person . . . may have under any statute or common law to seek enforcement of any emission

standard or limitation or to seek any other relief[.]” § 7604(e) (emphases added). The

second savings clause provides:

[N]othing in this chapter shall preclude or deny the right of any State or
political subdivision thereof to adopt or enforce (1) any standard or limitation
respecting emissions of air pollutants or (2) any requirement respecting
control or abatement of air pollution; except that if an emission standard or
limitation is in effect under an applicable implementation plan . . . , such
State or political subdivision may not adopt or enforce any emission standard
or limitation which is less stringent than the standard or limitation under such
plan or section.

§ 7416 (emphases added). Under §§ 7604(e) and 7416, except to the extent a state law

falls below a federal requirement under a limitation plan, the plain language of the CAA’s

savings clauses evidence no congressional intent for the CAA to be the exclusive cause of

action for air pollution claims. Section 7604(e) permits parties to resort to state statutes

and state common law to enforce emission standards or “to seek any other relief.” Section

7416 permits States and political subdivisions to “adopt or enforce . . . any” standard,

limitation, or requirement about air pollution that are more demanding than federal

provisions, and this broad language encompasses state-law claims that may be used to rein

in air pollution. See City of Oakland, 969 F.3d at 907–08 (holding that the CAA did not

satisfy the requirements of complete preemption for state-law claims involving public

50
nuisance after considering § 7416’s impact). These sweeping clauses—both of which show

respect for state law—fail to show how Congress clearly and manifestly intended to

completely preempt the types of claims Baltimore presents here, tipping the scale against

complete preemption. See Johnson, 781 F.3d at 703 (holding that a savings clause

counseled against finding complete preemption of § 332 of the Communications Act); Her

Majesty the Queen in Right of Ont. v. City of Detroit, 874 F.2d 332, 342–43 (6th Cir. 1989)

(noting that § 7604(e) of the CAA “clearly indicates that Congress did not wish to abolish

state control” and declining to apply complete preemption to a Michigan statute).

In the face of the CAA’s savings clauses, Defendants posit that the CAA “authorizes

states to impose additional restrictions only on in-state emissions[] and . . . provide[s]

remedies only for localized injuries stemming from in-state air pollution.” Defs.’ Opening

Br. 50. But Defendants’ argument continues to rest on a fundamental confusion of

Baltimore’s claims. None of Baltimore’s claims concern emission standards, federal

regulations about those standards, or pollution permits. Their Complaint is about

Defendants’ fossil-fuel products and extravagant misinformation campaign that

contributed to its injuries. Indeed, since we are operating under the well-pleaded complaint

rule, Aetna Health, 542 U.S. at 207, we take Baltimore at its word when it claims that it

“does not seek to impose liability on Defendants for their direct emissions of greenhouse

gases and does not seek to restrain Defendants from engaging in their business operations.”

J.A. 47.

In sum, there is simply nothing within the “text of the statute” suggesting that state-

law claims are completely displaced by the CAA. See Lontz, 413 F.3d at 441. And more

51
specifically, we do not see anything within the CAA requiring the complete preemption of

state-law claims that seek to impose liability upon fossil-fuel products that are allegedly

harmful to the public at large. We join our sister circuits and reject this complete-

preemption argument from Defendants, especially when Defendants do not identify any

statutory sections that indicate the complete preemption of Baltimore’s state-law claims.

See Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1263–65 (noting that the CAA

“does not provide an exclusive cause of action for suits against private polluters, nor does

it completely displace all state law in that area” and rejecting Defendants’ argument); City

of Oakland, 969 F.3d at 907–08 (holding that the CAA did not satisfy the requirements of

complete preemption for state-law claims involving public nuisance). Thus, we affirm the

district court’s rejection of complete preemption under the CAA and find no federal-

question jurisdiction on this basis.

D.

Relying upon a federal-enclaves theory, Defendants assert that federal jurisdiction

is appropriate because a “substantial portion” of their operations occurred on federal land,

including the Elk Hills Naval Petroleum Reserve in Kern County, California, and multiple

naval installations. Defs.’ Opening Br. 46–47. Defendants are correct that naval

installations are generally considered federal enclaves. See Allison v. Boeing Laser Tech.

Servs., 689 F.3d 1234, 1235 (10th Cir. 2012) (observing that federal enclaves include

“military bases, federal facilities, and even some national forests and parks”). However,

federal-question jurisdiction is not conferred merely because some of Defendants’

52
activities occurred on military installations. We decline to endorse Defendants’

overreaching approach to federal-question jurisdiction premised on federal enclaves.

Congress possesses the power to “exercise exclusive Legislation in all Cases

whatsoever, over such District (not exceeding ten Miles square) as may, . . . become the

Seat of the Government of the United States, and . . . exercise like Authority over all Places

purchased by the Consent of the Legislature of the State in which the Same shall be, for

the Erection of Forts, Magazines, Arsenals, dock-Yards, and other needful Buildings[.]”

U.S. Const. art. I, § 8, cl. 17. The federal government thus possesses “sole jurisdiction”

over its enclaves. Surplus Trading Co. v. Cook, 281 U.S. 647, 652 (1930). Accordingly,

“[f]ederal courts have federal question jurisdiction over tort claims that arise on ‘federal

enclaves.’” Durham v. Lockheed Martin Corp., 445 F.3d 1247, 1250 (9th Cir. 2006)

(citations omitted); see also Akin v. Ashland Chem. Co., 156 F.3d 1030, 1034 (10th Cir.

1998) (“Personal injury actions which arise from incidents occurring in federal enclaves

may be removed to federal district court as a part of federal question jurisdiction.”).

When deciding if a federal enclave confers jurisdiction, this Court has considered

whether the injury itself was sustained within the federal enclave. See Stokes v. Adair, 265

F.2d 662, 663, 665–66 (4th Cir. 1959). In Stokes, we were tasked with deciding whether a

district court possessed federal-question jurisdiction over an automobile accident that

caused a plaintiff to sustain injuries “on the United States Military Reservation of Fort

Leavenworth in the State of Kansas . . . .” Id. at 663. We held that the district court had

federal-question jurisdiction because Kansas ceded Fort Leavenworth to the federal

government, and we specifically noted that personal-injury actions occurring “on a federal

53
reservation” were not precluded from trial in state courts. Id. at 665–66. Other federal

courts have similarly reasoned that federal-question jurisdiction only lies over federal

enclaves when personal injuries are sustained within an enclave’s boundaries. See Akin,

156 F.3d at 1034–36 (holding that federal-question jurisdiction was not ascertainable from

a complaint when it stated that a plaintiff’s injuries were “sustained ‘while working at’” a

federal enclave, but jurisdiction was conferred when an interrogatory eventually provided

“sufficient notice” that the relevant conduct took place “wholly within the enclave”);

Mater v. Holley, 200 F.2d 123, 123–25 (5th Cir. 1952) (holding there was federal-question

jurisdiction when a claimant suffered personal injuries “within the boundaries” of Fort

McPherson, Georgia, and Fort McPherson was ceded to the United States by Georgia).

On the Complaint’s face, Baltimore specifically states that “‘Baltimore’ refers to

Baltimore City’s geographic area, and specifically to non-federal lands within its

boundaries, unless otherwise stated.” J.A. 43. Baltimore therefore excludes federal

enclaves from its Complaint “unless otherwise stated.” J.A. 43. As to where Baltimore’s

climate-change injuries have occurred, the Complaint emphasizes they have taken place

within Baltimore’s borders and not on a federal enclave. For instance, the Complaint

maintains that climate change, resulting from Defendants’ fossil-fuel products and

marketing campaign, has damaged Baltimore’s internal infrastructure, including its

railways and roads, and increased the costs of maintaining, repairing, and replacing

infrastructure within Baltimore. See J.A. 144–45. It also describes the adverse impacts

that climate change will have on the health of Baltimore’s citizens as opposed to those

living on federal enclaves. See J.A. 146–47. None of these allegations suggest that

54
Baltimore’s injuries occurred or will occur on federal enclaves. All of Baltimore’s harms

are pleaded within the confines and boundaries of Baltimore City. See J.A. 139–48. So

given Baltimore’s alleged injuries have not occurred on a federal enclave, it seeks relief

for harms sustained on non-federal land, which precludes the exercise of federal-question

jurisdiction. See Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1271–72 (rejecting

Defendants’ broad theory of federal enclaves and holding federal-question jurisdiction was

an improper basis for removal when injuries on federal lands were expressly disclaimed).

Again, federal-question jurisdiction tied to federal enclaves “generally requires ‘that

all pertinent events t[ake] place on a federal enclave.’” Id. at 1271 (quoting Rosseter v.

Indus. Light & Magic, No. C 08-04545 WHA, 2009 WL 210452, at *1 (N.D. Cal. Jan. 27,

2009)). The district court reasonably concluded that “the claims appear to arise in

Baltimore, where the City allegedly suffered and will suffer harm.” BP P.L.C., 388 F.

Supp. 3d at 566. We agree with the district court and affirm its firm rejection of jurisdiction

based on this doctrine.

E.

Continuing on their quest for federal jurisdiction, Defendants invoke the OCSLA’s

jurisdictional grant to reach federal court. They believe “[Baltimore]’s claims as alleged

encompass all of Defendants ‘exploration and production’ of fossil fuels on the OCS . . . .”

Defs.’ Opening Br. 43. Rejecting Defendants jurisdictional invocation of the OCSLA, the

district court held that Defendants failed to show a but-for connection between Baltimore’s

causes of action and the Outer Continental Shelf (OCS). BP P.L.C., 388 F. Supp. 3d at

55
566–67. Defendants do not believe a but-for connection is a requirement under the

OCSLA’s jurisdictional grant, and, even if it is, they maintain it is satisfied. We disagree

with Defendants on both fronts.

1.

We first consider whether the OCSLA’s jurisdictional grant requires a but-for

connection between a cause of action and the OCS. In full, the OCSLA provides federal

district courts with original jurisdiction to hear cases involving the OCS 17:

[T]he district courts of the United States shall have jurisdiction of cases and
controversies arising out of, or in connection with (A) any operation
conducted on the outer Continental Shelf which involves exploration,
development, or production of the minerals, of the subsoil and seabed of the
outer Continental Shelf, or which involves rights to such minerals, or (B) the
cancellation, suspension, or termination of a lease or permit under this
subchapter.

43 U.S.C. § 1349(b)(1) (emphasis added). Section 1349(b)(1) is a “broad” grant of federal

jurisdiction. Texaco Expl. & Prod., Inc. v. AmClyde Engineered Prods. Co., Inc., 448 F.3d

760, 768 (5th Cir. 2006). When assessing jurisdiction under this provision, we consider

whether “(1) the activities that caused the injury constituted an ‘operation’ ‘conducted on

the outer Continental Shelf’ that involved the exploration and production of minerals, and

(2) the case ‘arises out of, or in connection with’ the operation.” In re Deepwater Horizon,

17
The OCSLA defines “outer Continental Shelf” as “all submerged lands lying seaward
and outside of the area of lands beneath navigable waters as defined in section 1301 of this
title, and of which the subsoil and seabed appertain to the United States and are subject to
its jurisdiction and control.” 43 U.S.C. § 1331(a). “Lands beneath navigable waters” has
three different definitions that are irrelevant to our jurisdictional inquiry. Id. § 1301(a)(1)–
(3).
56
745 F.3d 157, 163 (5th Cir. 2014) (citation omitted). Since Baltimore does not contest

Defendants’ “operations” on the OCS, we are only concerned with the meaning of “arising

out of, or in connection with,” which the OCSLA does not define. Departing from the Fifth

Circuit, Defendants preliminarily contest whether those phrases impose a but-for

connection at all. We thus resort to our tools of statutory construction to determine if those

phrases require a but-for connection between a plaintiff’s case and operations on the OCS.

Those tools include: (1) statutory text; (2) statutory structure; (3) legislative history; (4)

judicial interpretations; (5) related statutes; and (6) congressional purpose. See Brown &

Williamson Tobacco Corp. v. Food & Drug Admin., 153 F.3d 155, 162 (4th Cir. 1998);

United States v. Jackson, 759 F.2d 342, 344 (4th Cir. 1985).

Examining the text alongside relevant case law resolves this matter. To “arise”

means “[t]o originate; to stem (from)” or “[t]o result (from).” Arise, BLACK’S LAW

DICTIONARY (11th ed. 2019). “Connection” denotes a “contextual relation or association”

or “relationship in fact.” Connection, MERRIAM-WEBSTER, https://www.merriam-

webster.com/dictionary/connection (last visited Feb. 27, 2022). Under their plain

meanings, “arising out of” and “in connection with” both require a causal relationship to

determine if a given controversy actually “result[s] (from)” or possesses a “relationship in

fact [with]” activities conducted on the OCS. See Burrage v. United States, 571 U.S. 204,

214 (2014) (holding that “results from” imposes but-for causation); Levine v.

Diamanthuset, Inc., 950 F.2d 1478, 1485–86 (9th Cir. 1991) (noting that “in connection

with” requires a causal relationship for actions under Rule 10b-5, which implements

section 10(b) of the Securities and Exchange Act of 1934); Arthur Young & Co. v. Reves,

57
937 F.2d 1310, 1327–28 (8th Cir. 1991) (stating that “in connection with” is satisfied under

Rule 10b-5 when but-for causation is shown). We are not alone in this conclusion. Federal

courts interpreting “arise out of, or in connection with” under the OCSLA have consistently

determined that it imposes a but-for relationship between a party’s case and operations on

the OCS. See Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1272–75; In re

Deepwater Horizon, 745 F.3d at 163; Tenn. Gas Pipeline v. Hous. Cas. Ins. Co., 87 F.3d

150, 155 (5th Cir. 1996); Recar v. CNG Producing Co., 853 F.2d 367, 369 (5th Cir. 1988).

We decline to disrupt this settled and sensible trend.

Accordingly, we join our sister circuits and find that invoking jurisdiction

under § 1349(b)(1) requires a but-for connection between a claimant’s cause of action and

operations on the OCS. See Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1272–75;

In re Deepwater Horizon, 745 F.3d at 163.

2.

When applying a but-for test, we must ask if Baltimore’s injuries “would not have

occurred” but for Defendants’ conduct on the OCS. See Wright v. Lassiter, 921 F.3d 413,

419 (4th Cir. 2019) (explaining but-for causation); see also Martin v. Duffy, 977 F.3d 294,

299 (4th Cir. 2020) (same); Talkington v. Atria Reclamelucifers Fabrieken BV (Cricket

BV), 152 F.3d 254, 264 (4th Cir. 1998) (same). Requiring a but-for connection still

“implies a broad jurisdictional grant under [the OCSLA] . . . .” Tenn. Gas Pipeline, 87 F.3d

at 155. But a “mere connection” between a claimant’s case and operations on the OCS is

insufficient to show federal jurisdiction if the relationship is “too remote.” In re Deepwater

58
Horizon, 745 F.3d at 163 (citation omitted); see also Plains Gas Sols., LLC v. Tenn. Gas

Pipeline Co., LLC, 46 F. Supp. 3d 701, 704–05 (S.D. Tex. 2014) (noting that the but-for

test is not “limitless” under the OCSLA).

Here, Baltimore’s allegations and injuries are not confined to Defendants’ fossil-

fuel activities on the OCS. Defendants are also being sued for unlawfully marketing,

promoting, and ultimately selling their fossil-fuel products, which includes their collective

failure to warn the public of the known dangers associated with their fossil-fuel products.

See J.A. 43, 150–71. Defendants’ marketing practices, which led to increased consumption

of their fossil-fuel products and then climate change, are far removed from their OCS

activities and their tort liability. In other words, irrespective of Defendants’ activities on

the OCS, Baltimore’s injuries still exist as a result of that distinct marketing conduct.

Regardless, Defendants concede that some of their fossil-fuel production occurred outside

of the OCS. Defs.’ Opening Br. 41, 46–47; J.A. 41 (Defendants’ noting that their oil and

gas activities occurred within the National Wildlife Refuge System). And Baltimore’s

Complaint contains examples of Defendants’ land-based activities that contributed to its

injuries and are hundreds of miles away from the OCS. See J.A. 66 (mentioning

CONSOL’s coal mines in Appalachia). Because Baltimore’s injuries remain even after we

disregard “whatever slice” of Defendants’ fossil-fuel production occurred on the OCS, we

cannot find a but-for connection satisfying the OCSLA’s jurisdictional grant. See Bd. of

Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1272–75 (rejecting Defendants’ invocation of

jurisdiction under the OCSLA using the but-for cause inquiry).

59
 Case law supports this finding since Baltimore’s Complaint only has a “mere

connection” to the OCS. In Tennessee Gas Pipeline, a barge physically “allided” with a

fixed platform on the OCS, and the Fifth Circuit found a but-for connection, concluding

“there would not have been an accident had Tennessee Gas not built its platform to extract

minerals from the OCS.” 87 F.3d at 152, 155. Most recently, in In re Deepwater Horizon,

the Fifth Circuit found a but-for connection when it was “undeniable” that contaminants

would not have entered the State of Louisiana’s territorial waters but for drilling and

exploration on the OCS. 745 F.3d at 163–64. Additional case law features tort claims

involving personal injuries with a direct connection to an OCS operation. See Barker v.

Hercules Offshore, Inc., 713 F.3d 208, 213 (5th Cir. 2013) (“Barker’s employment on the

jack-up rig was directly related to the development of minerals or other natural resources

on the OCS.” (citation omitted)); Hufnagel v. Omega Serv. Indus., Inc., 182 F.3d 340, 350

(5th Cir. 1999) (“Hufnagel’s injuries occurred on a stationary drilling platform involved in

the ‘exploration, development, or production’ of minerals on the shelf.” (citation omitted));

Ronquille v. Aminoil Inc., C/A No. 14-164, 2014 WL 4387337, at *2 (E.D. La. Sept. 4,

2014) (finding the but-for test satisfied where a plaintiff alleged he was exposed to asbestos

when he provided direct support for “Shell Oil’s rigs within [the OCS]”). Baltimore’s

Complaint does not align with cases finding connections to the OCS. Instead, the

allegations contained therein turn on Defendants’ deceptive marketing practices as well as

the resulting impacts of their fossil-fuel products, both of which are far removed from any

production occurring on the OCS. See Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at

1273–75; see also Plaquemines v. Total Petrochemical & Refining USA, Inc., 64 F. Supp.

60
3d 872, 898 (E.D. La. 2014) (declining to find jurisdiction under the OCSLA when there

were “injuries sustained in state waters from activities that occurred off the shelf” and

finding a relationship to the OCS as “too remote and attenuated”). Contrary to dangerous

contaminants entering a State’s territorial waters as a result from drilling on the OCS, or

the worker who is injured on an OCS platform, Baltimore’s claims of Defendants’ trickery

and deceit, conduct leading to climate change, bear a weak relationship to the OCS.

Plaquemines, 64 F. Supp. 3d at 898. We find that this is all “too remote and attenuated”

for a but-for connection under OCSLA. Id.

Ignoring the OCSLA’s text and judicial decisions applying it, Defendants argue that

the OCSLA’s policy aims will be frustrated and a parade of horrible outcomes will ensue

if we decline federal jurisdiction. To them, if Baltimore is ultimately granted relief, “[s]uch

relief would substantially discourage OCS production and jeopardize the future viability of

the federal OCS leasing program, potentially costing the federal government hundreds of

millions of dollars in revenues.” Defs.’ Opening Br. 45. Maybe so. But under our laws,

“a defendant cannot establish removal jurisdiction by mere speculation and conjecture,

with unreasonable assumptions.” Ibarra v. Manheim Invs., Inc., 775 F.3d 1193, 1197 (9th

Cir. 2015); see also Bd. of Cnty. Comm’rs of Boulder Cnty., 25 F.4th at 1275. Even if such

speculative and policy-laden arguments were permitted, the authorities Defendants rely

upon for their jurisdictional theory are markedly different from Baltimore’s suit because

they involve the intersection of commercial disputes satisfying the “operation” element of

the OCSLA, not injurious torts impacting a municipality’s citizenry and internal

infrastructure. See Defs.’ Opening Br. 45 (citing EP Operating Ltd. P’ship v. Placid Oil

61
Co., 26 F.3d 563, 566 (5th Cir. 1994) (a party seeking partition by licitation); United

Offshore Co. v. S. Deepwater Pipeline Co., 899 F.2d 405, 406 (5th Cir. 1990) (a claimant

requesting the compulsion of arbitration and dissolution of an injunction); Amoco Prod.

Co. v. Sea Robin Pipeline Co., 844 F.2d 1202, 1203–04 (5th Cir. 1988) (a party disputing

its take-or-pay obligations in contracts for the purchase and sale of natural gas)). As it

concerns physical damages to persons and property, we thus join other courts in rejecting

this policy-based theory of jurisdiction under the OCSLA. See Bd. of Cnty. Comm’rs of

Boulder Cnty., 25 F.4th at 1275; Plaquemines, 64 F. Supp. 3d at 896–98.

For the foregoing reasons, we affirm the district court’s thoughtful rejection of the

OCSLA’s jurisdictional grant as a basis for federal jurisdiction over Baltimore’s claims.

F.

Defendants next press removal jurisdiction under the bankruptcy removal statute.

The bankruptcy removal statute provides:

A party may remove any claim or cause of action in a civil action other than
. . . a civil action brought by a governmental unit’s police or regulatory
power, to the district court where such civil action is pending, if such district
court has jurisdiction of such claim or cause of action under section 1334 of
this title.

28 U.S.C. § 1452(a). In turn, § 1334(b) states that federal district courts shall have “original

but not exclusive jurisdiction of all civil proceedings arising under title 11, or arising in or

related to a case under title 11.” Defendants do not argue that Baltimore’s Complaint

involves a bankruptcy proceeding under Title 11. Instead, they maintain that Baltimore’s

Complaint is “related to” bankruptcy cases because it primarily seeks to hold them liable

62
for the “pre-bankruptcy conduct” of a Chevron subsidiary, Texaco, Inc. Defs.’ Opening

Br. 52. Defendants describe Texaco’s bankruptcy plan, along with those of Defendants’

other predecessors, subsidiaries, and affiliates, as confirmed. Id. But as with Defendants’

reliance upon the OCSLA, Baltimore’s suit is too remote for bankruptcy removal to lie.

Generally, at the pre-confirmation stage of a reorganization plan, we first stated that

the “related to” test addressed “whether the outcome of [a civil] proceeding could

conceivably have any effect on the estate being administered in bankruptcy.” In re Celotex

Corp., 124 F.3d 619, 625 (4th Cir. 1997) (quoting Pacor, Inc. v. Higgins, 743 F.2d 984,

994 (3d Cir. 1984)). However, we then endorsed a “close nexus” test at the post-

confirmation stage, allowing jurisdiction over “[m]atters that affect the interpretation,

implementation, consummation, execution, or administration of the confirmed plan . . . .”

Valley Historic Ltd. P’ship v. Bank of N.Y., 486 F.3d 831, 836–37 (4th Cir. 2007) (alteration

in original) (quoting In re Resorts Int’l, Inc., 372 F.3d 154, 167 (3d Cir. 2004)); see also

In re Kirkland, 600 F.3d 310, 317 (4th Cir. 2010) (same). Essentially, a plaintiff’s suit

“must affect an integral aspect of the bankruptcy process” for “related to” jurisdiction to

exist. In re Kirkland, 600 F.3d at 317 (citation omitted).

1.

Defendants explicitly rely upon a 1988 confirmed plan from Chevron’s subsidiary,

Texaco, to contend we possess bankruptcy jurisdiction. 18 First, we find it hard to fathom

18
The record does not appear to contain Texaco’s 1988 confirmed plan.

63
how Baltimore’s suit, filed thirty years later, has any “close nexus” to Texaco’s confirmed

planned because it is so far removed from the initial bankruptcy confirmation. See Nuveen

Mun. Tr. ex rel. Nuveen High Yield Mun. Bond Fund v. Withumsmith Brown, P.C., 692

F.3d 283, 294 (3d Cir. 2012) (noting that bankruptcy jurisdiction “wanes” after the

confirmation of a case). Secondly, Baltimore’s claims are completely independent and

distinct from Texaco’s bankruptcy plan, there is no indication that the bankruptcy plan

involved climate change, and Defendants do not explain how a judgment more than thirty

years later could impact Texaco’s estate. See Valley Historic, 486 F.3d at 837 (finding no

“related to” jurisdiction when a bankruptcy plan was “substantially consummated”); New

Horizon of N.Y. LLC v. Jacobs, 231 F.3d 143, 154–55 (4th Cir. 2000) (holding there was

no “related to” jurisdiction for state-law claims that were “completely unrelated to the . . .

administration of the bankruptcy estates”). For those reasons, we conclude Baltimore’s

suit is too far removed from Texaco’s 1988 confirmed plan for us to find a “close nexus”

warranting bankruptcy jurisdiction.

Citing to a powerpoint presentation they filed in the district court, Defendants

speculate that other corporate entities related to Defendants “may also be operating under

confirmed bankruptcy cases.” Defs.’ Opening Br. 52 (citing Ex. 20 to Decl. of Joshua S.

Lipshutz, Mayor & City Council of Balt. v. BP P.L.C., 388 F. Supp. 3d 538 (D. Md. 2019)

(No. 1:18-cv-02357-ELH), ECF No. 125-20 at 3)). Interestingly, Defendants’ filed

presentation includes 134 bankruptcy filings from energy companies from 2015 to 2017.

Yet, Defendants do not specify if any of those corporate entities are actually related to any

of them, nor do they indicate if or when those bankruptcy cases were confirmed by federal

64
courts. By failing to direct us to anything further, we find this is insufficient to carry any

burden for bankruptcy removal and decline to do counsel’s work. Prince, 848 F.3d at 176;

N.Y. Rehab. Care Mgmt., LLC v. Nat’l Lab. Rels. Bd., 506 F.3d 1070, 1076 (D.C. Cir.

2007) (“It is not enough merely to mention a possible argument in the most skeletal way,

leaving the court to do counsel’s work.” (citation omitted)).

Accordingly, we dispense with Defendants’ primary arguments because they have

failed to show that Baltimore’s suit has a “close nexus” or is “related” to any bankruptcy

plan involving any of its predecessors, subsidiaries, or affiliates under § 1452(a).

2.

Even were we to find bankruptcy jurisdiction is proper, removal is still inappropriate

if the proceeding is a civil action by a “governmental unit to enforce such governmental

unit’s police or regulatory power . . . .” § 1452(a). Baltimore’s suit is such an action, and

we note that Defendants only advance one sentence concerning whether this “police or

regulatory power” exception is inapplicable to their bankruptcy removal invocation. See

Defs.’ Opening Br. 52–53.

Baltimore clearly qualifies as a “governmental unit” since it is a municipality under

Title 11 and, thus, for the purposes of § 1452(a) as well. See 11 U.S.C. § 101(27) (defining

a “governmental unit” as a “municipality”). This Court has not yet interpreted a

governmental unit’s “police or regulatory power” under § 1452(a). But we have interpreted

“police and regulatory power” under § 362(b)(4) of the bankruptcy code, which is an

exception to the automatic stay of actions brought by creditors against debtors after

65
bankruptcy petitions are filed. See Safety-Kleen, Inc. (Pinewood) v. Wyche, 274 F.3d 846,

864–66 (4th Cir. 2001). In Safety-Kleen, this Court held that a “police and regulatory

power” is being exercised if the purpose of a state law is to effectuate public policy or

promote the public safety and welfare. Id. at 865 (citations omitted). We stated that this

is an objective analysis that requires a court to “determine the primary purpose of the law

that the state is attempting to enforce.” Id. (citations omitted).

As noted above, Baltimore brings eight different claims against Defendants, and all

of those claims seek to shift the costs of climate-change injuries onto Defendants as

opposed to burdening “local taxpayers, residents, or broader segments of the public.” J.A.

47. In its public nuisance claim, for example, Baltimore asserts that Defendants’

interference with its property, infrastructure, and public resources will be “borne by

[Baltimore’s] citizens” because they will purportedly suffer economic losses and negative,

public-health consequences. J.A. 151. This is easily said for Baltimore’s other claims as

well. Baltimore thus seeks to protect its citizens, property, and resources by suing

Defendants, all of whom are private parties, for the detrimental impacts of their fossil-fuel

products. See J.A. 150–71. We have no doubt this suit is a valid exercise of Baltimore’s

police power. See Crutcher v. Kentucky, 141 U.S. 47, 61 (1891) (“[T]he police power of

the state extends to almost everything within its borders,–to the suppression of nuisances;

[and] to the prohibition of manufactures deemed injurious to the public health . . . .”

(citations omitted)); Sansotta v. Town of Nags Head, 724 F.3d 533, 541 (4th Cir. 2013)

(“The Town’s actions to abate a nuisance were reasonable . . . uses of its police

power . . . .” (citation omitted)). Accordingly, we hold that the exception to bankruptcy

66
removal is applicable, precluding Defendants’ ability to remove under § 1452(a). See In

re Methyl Tertiary Butyl Ether (“MTBE”) Prods. Liab. Litig., 488 F.3d 112, 133 (2d Cir.

2007) (holding that “police or regulatory powers” were exercised under § 1452(a) when

California and New Hampshire sought to “remedy and prevent environmental damage with

potentially serious consequences for public health”); Safety-Kleen, 274 F.3d at 866

(holding there was a “clear exercise” of South Carolina’s regulatory power when it sought

to deter environmental misconduct through financial assurance regulations).

Thus, we find no federal jurisdiction under the bankruptcy removal statute and

affirm the district court in this regard.

G.

With few theories remaining, Defendants attempt to reach federal court by

appealing to our admiralty jurisdiction under the Constitution and 28 U.S.C. § 1333(1).

They believe that admiralty jurisdiction is conferred merely because “fossil-fuel extraction

occurs on vessels engaged in maritime commerce[.]” Defs.’ Opening Br. 53. We reject

Defendants’ far-reaching view of admiralty jurisdiction.

1.

The Constitution extends our judicial power to “all Cases of admiralty and maritime

Jurisdiction.” U.S. Const. art. III, § 2. Congress provides that “[t]he district courts shall

have original jurisdiction, exclusive of the courts of the States, of: . . . Any civil case of

admiralty or maritime jurisdiction, saving to suitors in all cases all other remedies to which

67
they are otherwise entitled.” 28 U.S.C. § 1333(1). The saving-to-suitors clause

of § 1333(1) “preserves remedies and the concurrent jurisdiction of state courts over some

admiralty and maritime claims.” Lewis v. Lewis & Clark Marine, Inc., 531 U.S. 438, 445

(2001) (citations omitted). The Supreme Court previously emphasized that claims brought

under the saving-to-suitors clause in state court are not removable to federal court based

on federal-question jurisdiction. See Romero v. Int’l Terminal Operating Co., 358 U.S.

354, 371–75 (1959). We have likewise stated that the savings-to-suitors clause “preserves

a maritime suitor’s election to pursue common-law remedies in state court.” Servis v.

Hiller Sys. Inc., 54 F.3d 203, 206 (4th Cir. 1995). Yet, “[a]dmirality and maritime cases

may . . . be removable to federal court when there exists some independent basis for federal

jurisdiction, such as diversity of citizenship or when federal jurisdiction is independently

established by a federal maritime statute.” Id. at 207 (citations omitted) (cleaned up); see

also In re Lockheed Martin Corp., 503 F.3d 351, 356 (4th Cir. 2007).

Adhering to those precedents, Baltimore argues that its state-law claims are not

removable under § 1441, the general removal statute, if they sound in admiralty unless

there is “some independent jurisdictional basis, such as diversity or federal question

jurisdiction.” Baltimore’s Resp. Br. 52–53. Defendants aptly point out that the Venue

Clarification Act of 2011 eliminated a portion of § 1441(b) that federal courts previously

believed blocked the removal of admiralty claims without another jurisdictional basis. Pub.

L. No. 112-63, § 103, 125 Stat. 758, 759 (2011); see also Lu Junhong v. Boeing Co., 792

F.3d 805, 817–18 (7th Cir. 2015) (holding that § 1441 permits removal when there is an

admiralty case under § 1333(1)). The parties, however, only devote one paragraph each to

68
this admittedly complicated issue that has divided federal courts. Thomas J. Schoenbaum,

Admiralty and Maritime Law § 4.3, Westlaw (database updated Dec. 2021) (“After 2011,

courts split on whether the working of the amended statute changes the rule for removal of

maritime claims.”). Because of their inadequate briefing, we decline to formally decide

whether § 1441 permits removal based on the admiralty jurisdiction grant under § 1333(1).

United States v. Banks, 884 F.3d 998, 1011 n.3 (10th Cir. 2018) (“Given the unsettled

nature of this question and the parties’ inadequate briefing, we decline to decide this

question here.”); see also Covey v. Assessor of Ohio Cnty., 777 F.3d 186, 198 n.11 (4th

Cir. 2015) (same); Brownfield v. City of Yakima, 612 F.3d 1140, 1149 n.4 (9th Cir. 2010)

(same). Instead, we press ahead under the assumption § 1441 permits removal because of

the original jurisdiction grant of § 1333(1).

2.

To invoke admiralty jurisdiction, a party must satisfy “conditions both of location

and of connection with maritime activity.” Jerome B. Grubart, Inc. v. Great Lakes Dredge

& Dock Co., 513 U.S. 527, 534 (1995). To satisfy the location test, a tort must either

“occur on navigable waters, or, if suffered on land, at least be caused by a vessel on

navigable water.” White v. United States, 53 F.3d 43, 45 (4th Cir. 1995) (citing Grubart,

513 U.S. at 534). For the connection test, a court must decide: (1) “whether ‘the general

features of the type of incident involved’ have ‘a potentially disruptive impact on maritime

commerce[,]’” and (2) “whether the general character of the activity giving rise to the

incident shows a substantial relationship to traditional maritime activity.” Id. at 46 (quoting

69
Grubart, 513 U.S. 534). Defendants never argue that either condition is satisfied in their

Opening Brief. See Grayson O Co., 856 F.3d at 316 (noting that parties waive arguments

not raised in their opening brief). But we nevertheless find that Defendants’ admiralty

invocation begins and ends with the location test.

3.

To begin, Baltimore pleads that its injuries involve damage to its “highways, rail

lines, emergency response facilities, waste water facilities, and power plants . . . .” J.A.

143. And, according to Baltimore, those land-based injuries stem from “sea level rise and

associated impacts, increased frequency and severity of extreme precipitation events,

increased frequency and severity of draught, [and the] increased frequency and severity of

heat waves and extreme temperatures . . . .” J.A. 140 (emphasis added). While Baltimore

alleges “sea level rise” as one of the many sources of its injuries, the actual torts involving

Baltimore’s property have occurred on land as opposed to navigable waters. Grubart, 513

U.S. at 533. Baltimore’s Complaint never mentions any tort that occurred on navigable

waters, and Defendants do not identify one.

Still, the location test may be satisfied when a land-based tort is caused by a vessel

on navigable waters. Grubart, 513 U.S. at 534 (citing 46 U.S.C. § 30101). Defendants

seem to argue that their “floating oil rigs” and “floating drilling platforms” are vessels

meeting the location test. Without giving us more, we disagree. Whether a craft or

structure qualifies as a “vessel” is typically a question of law, but it sometimes comes down

to the facts. Manuel v. P.A.W. & Drilling & Well Serv., Inc., 135 F.3d 344, 347 (5th Cir.

70
1998). Here, we deem it a question of law. The term “vessel” is “generally defined

broadly[.]” Bernard v. Binnings Constr. Co., Inc., 741 F.2d 824, 828 (5th Cir. 1984).

Usually, “[c]onventional ships and barges as well as such unconventional craft as

submersible drilling barges and floating dredges which are designed for navigation and

commerce are vessels within general maritime . . . jurisdiction and retain such status even

while moored, dry-docked, or otherwise immobilized and secured to land.” Cook v. Belden

Concrete Prods., Inc., 472 F.2d 999, 1001 (5th Cir. 1973) (emphasis added) (citations

omitted). Defendants never suggest that their floating rigs and platforms were either

designed for navigation or used for navigation when Baltimore suffered its injuries. They

exclusively posit that their structures are used for oil and gas production, not for any

navigation purposes. See Defs.’ Reply 28. This is fatal to a characterization of Defendants’

floating rigs and platforms as vessels for admiralty jurisdiction. See Bernard, 741 F.2d at

831 (noting that a structure is typically not a “vessel” under the Jones Act, 46

U.S.C. § 30104, when it is constructed primarily for use as a work platform); Gremillion

v. Gulf Coast Catering Co., 904 F.2d 290, 291 n.2, 294 (5th Cir. 1990) (holding a barge

used as a “floating hotel” was not a vessel under the Jones Act); Cook, 472 F.3d at 1102

(holding that there was no vessel for maritime jurisdiction when a “floating construction

platform was not designed for the purpose of navigation,” and it was “engaged in its

primary function as a stationary construction platform”). But see Barker, 713 F.3d at 215

(considering “jack-up drilling platforms” as vessels).

Even if we credit Defendants with having vessels, Baltimore never alleges that any

vessel on navigable waters caused any of its land-based injuries. Instead, Baltimore

71
repeatedly references “flood-associated damages” and “heavy rains” that have destroyed

its infrastructure and exacerbated the health and environmental risks of its citizens. J.A.

145. There are no allegations that its injuries were either “caused by the vessel itself or its

appurtenances.” Egorov, Puchinsky, Afanasiev & Juring v. Terriberry, Carroll & Yancey,

183 F.3d 453, 456 (5th Cir. 1999) (declining to find admiralty jurisdiction because neither

the vessel nor its appurtenances caused physical damage on land or the alleged tort of

tortious interference with contract); see also MLC Fishing, Inc. v. Velez, 667 F.3d 140,

141–42 (2d Cir. 2011) (holding that a claimant’s slip-and-fall on a ramp leading from a

floating dock was not caused by a vessel or its appurtenances); Corrigan v. Harvey, 951 F.

Supp. 948, 950, 952–53 (D. Haw. 1996) (holding admiralty jurisdiction did not lie when a

plaintiff failed to allege any facts about his injury being caused by a vessel, and the parties

agreed that the injury resulted from a physical fight on a pier). Since no vessel is alleged

to have caused any of the committed torts, we find this issue easily resolved under the well-

pleaded complaint rule. Metro. Life, 481 U.S. at 63.

Accordingly, Defendants have failed to carry their removal burden of showing how

their floating rigs and platforms qualify as vessels for the location test. We conclude that

they are not, as a matter of law, especially since Baltimore’s Complaint never invokes them

as the cause of its land-based torts. Cf. Grubart, 513 U.S. at 529–30, 534–35 (holding that

the location test was satisfied when a crane, attached to a barge, was used to lift and replace

pilings around a bridge pier and a tunnel flooded after an accident). As such, we find no

72
merit to Defendants’ invocation of admiralty jurisdiction and affirm the district court’s

rejection of this basis for federal jurisdiction. 19

H.

At this juncture, the only remaining path to federal court is Defendants’ theory of

federal officer removal. In their Supplemental Brief, which was filed after their litigation

in the Supreme Court, Defendants reiterate that Baltimore’s case is removable on the

grounds that it originally raised on appeal, including federal officer removal. Defs.’ Suppl.

Br. 2 n.1. Defendants do not present any new arguments or shortcomings concerning our

previous holding that rejected the propriety of federal officer removal. The Supreme Court

only required us to consider Defendants’ other removal grounds on remand and never

addressed our holding concerning federal officer removal. Compare BP P.L.C., 141 S. Ct.

at 1543, with BP P.L.C., 952 F.3d at 461–71. Nevertheless, because the Supreme Court

vacated the entirety of our prior opinion, it has no precedential effect. Accordingly, we

deem it appropriate to adopt and include our prior opinion and its reasoning, which rejects

Defendants’ ability to remove under the federal officer removal statute.

The federal officer removal statute authorizes the removal of state-court actions

filed against “any officer (or any person acting under that officer) of the United States or

19
Because we hold Defendants failed to satisfy the location test for admiralty jurisdiction,
we decline to address the connection test. Gruver v. Lesman Fisheries Inc., 489 F.3d 978,
982 (9th Cir. 2007) (“[A] party seeking to invoke federal maritime jurisdiction over a tort
claim must satisfy both a location test and a connection test.” (citation omitted)).

73
of any agency thereof, in an official or individual capacity, for or relating to any act under

color of such office.” 28 U.S.C. § 1442(a)(1). Its “basic purpose” is to protect against the

interference with federal operations that would ensue if a state were able to arrest federal

officers and agents acting within the scope of their authority and bring them to trial in a

state court for an alleged state-law offense. Watson v. Philip Morris Co., 551 U.S. 142,

150 (2007) (explaining that state-court proceedings may (1) “reflect ‘local prejudice’

against unpopular federal laws or federal officials”; (2) “impede [enforcement of federal

law] through delay”; or (3) “deprive federal officials of a federal forum in which to assert

federal immunity defenses” (citations omitted)).

Thus, to remove a case under § 1442(a)(1), a private defendant must show: “(1) that

it ‘act[ed] under’ a federal officer, (2) that it has ‘a colorable federal defense,’ and (3) that

the charged conduct was carried out for [or] in relation to the asserted official authority.”

Sawyer v. Foster Wheeler LLC, 860 F.3d 249, 254 (4th Cir. 2017) (first alteration in

original) (citations omitted). Here, Defendants assert that Baltimore’s state-court action is

removable under the federal officer removal statute “because the City ‘bases liability on

activities undertaken at the direction of the federal government.’” BP P.L.C., 388 F. Supp.

3d at 567 (citation omitted). It is the first and third prongs that are therefore in dispute.

See Baltimore’s Resp. Br. 14–21. We begin with the first, though the acting-under and

causal-nexus prongs often “collapse into a single requirement.” In re MTBE, 488 F.3d at

124; see also 28 U.S.C. § 1442(a)(1) (targeting for removal state-court actions “for or

relating to any act under color of [federal] office”).

74
 1.

The statutory phrase “acting under” describes “the triggering relationship between

a private entity and a federal officer.” Watson, 551 U.S. at 149. Although the words

“acting under” are “broad,” the Supreme Court has emphasized that they are not

“limitless.” Id. at 147. In cases involving a private entity, the “acting under” relationship

requires that there at least be some exertion of “subjection, guidance, or control” on the

part of the federal government. See id. at 151 (quoting Webster’s New International

Dictionary 2765 (2d ed. 1953)). Additionally, “precedent and statutory purpose” make

clear that “‘acting under’ must involve an effort to assist, or to help carry out, the duties or

tasks of the federal superior.” Id. at 152.

In Watson, the Supreme Court held that “simply complying with the law” does not

constitute the type of “help or assistance necessary to bring a private [entity] within the

scope of the statute,” id., no matter how detailed the government regulation or how

intensely the entity’s activities are supervised and monitored, see id. at 153. In doing so,

the Court distinguished several decisions cited by the defendant there in which lower courts

had held that private contractors fell within the terms of § 1442(a)(1), at least where the

relationship was “an unusually close one involving detailed regulation, monitoring, or

supervision.” Id. at 153 (citing Winters v. Diamond Shamrock Chem. Co., 149 F.3d 387

(5th Cir. 1998)). The difference between those cases and a case involving a highly

regulated private firm, the Court reasoned, was the fulfillment of a government need:

The answer to this question lies in the fact that the private contractor in such
cases is helping the Government to produce an item that it needs. The
assistance that private contractors provide federal officers goes beyond

75
 simple compliance with the law and helps officers fulfill other basic
governmental tasks. In the context of Winters, for example, Dow Chemical
fulfilled the terms of a contractual agreement by providing the Government
with a product that it used to help conduct a war. Moreover, at least arguably,
Dow performed a job that, in the absence of a contract with a private firm,
the Government itself would have had to perform.

Id. at 153–54.

The Supreme Court found these circumstances sufficient to distinguish Dow

Chemical (the contractor in Winters) from the regulated tobacco companies who sought

removal in Watson, and so it did not address “whether and when particular circumstances

may enable private contractors to invoke the statute.” Id. at 154. Nevertheless, in light of

the Court’s reasoning, we have relied on Watson to hold that certain private contractors

“act under” federal officials. See Sawyer, 860 F.3d at 255. In Sawyer, we observed that

“courts have unhesitatingly treated the ‘acting under’ requirement as satisfied where a

contractor seeks to remove a case involving injuries arising from equipment that it

manufactured for the government.” Id. Thus, in that case, we found that the defendant

“acted under” the United States Navy when it manufactured boilers to be used aboard naval

vessels per a detailed government contract. See id. at 252–53, 255.

2.

Here, Defendants collectively seek removal under § 1442 based on three contractual

relationships between certain Defendants and the federal government: (1) fuel supply

agreements between one Defendant (Citgo) and the Navy Exchange Service Command

(“NEXCOM”) from 1988 to 2012; (2) oil and gas leases administered by the Secretary of

the Interior under the OCSLA; and (3) a 1944 unit agreement between the predecessor of

76
another Defendant (Chevron) and the U.S. Navy for the joint operation of a strategic

petroleum reserve in California known as the Elk Hills Reserve. For the reasons that

follow, we agree with Baltimore that none of these relationships are sufficient to justify

removal under the federal officer removal statute in this case, either because they fail to

satisfy the acting-under prong or because they are insufficiently related to Baltimore’s

claims for purposes of the nexus prong.

a.

First, we have little trouble concluding that the NEXCOM fuel supply agreements

do not satisfy the “acting under” requirement. These agreements required Defendant Citgo

to advertise, supply, and distribute gasoline and diesel to NEXCOM, which NEXCOM

resold at a discount to “active duty military, retirees, reservists, and their families” at

“service stations operated by NEXCOM on Navy bases located in a number of states across

the country.” J.A. 216. Although Defendants contend that Citgo helped “the Government

to produce an item that it needs” by selling NEXCOM fuel for resale on Navy bases, see

Watson, 551 U.S. at 153, such logic would bring every seller of contracted goods and

services within the ambit of § 1442 when the government is a customer.

We refuse to adopt such a sweeping interpretation of Watson. In our view, the key

lesson from Watson is that closely supervised government contractors are distinguishable

from intensely regulated private firms because the former assist the government in carrying

out basic governmental functions. See 551 U.S. at 153–54 (“The assistance that private

contractors provide federal officers goes beyond simple compliance with the law and helps

77
officers fulfill other basic governmental tasks. . . . [And they are tasks that] the

Government itself would [otherwise] have . . . to perform.”). And the provision of means

to engage in chemical warfare, as in Winters, or even the provision of specific component

parts to be used aboard military vessels, as in Sawyer, is different in kind from the provision

of motor vehicle fuel for resale on Navy bases—both in terms of the nature of the “item”

provided and the level of supervision and control that is contemplated by the contract.

To be sure, other circuits have applied the Watson dictum beyond the military-

procurement-contract context, and we do not suggest that only defense contractors may

invoke the federal officer removal statute. 20 Yet none of those cases have confronted a

contract like the one we have here, which involves the sale of a standardized consumer

product. Indeed, the Ninth Circuit has held, albeit in an unpublished decision, that the fact

that the federal government purchases “off-the-shelf” products from a manufacturer “does

not show that the federal government [has] supervised [the] manufacture of [such products]

or directed [that they be] produce[d] in a particular manner, so as to come within the

meaning of ‘act[ed] under.’” Washington v. Monsanto Co., 738 F. App’x 554, 555 (9th

Cir. 2018) (sixth alteration in original) (quoting 28 U.S.C. § 1442(a)(1)).

Although Defendants strongly resist the off-the-shelf-products analogy by pointing

to particular provisions in the fuel supply agreements, we find those provisions unavailing.

20
For cases involving people other than defense contractors, see, for example, Goncalves
ex rel. Goncalves v. Rady Child.’s Hosp. San Diego, 865 F.3d 1237, 1245–49 (9th Cir.
2017); In re Commonwealth’s Motion to Appoint Couns. Against or Directed to Def. Ass’n
of Phila., 790 F.3d 457, 469 (3d Cir. 2015); Bell v. Thornburg, 743 F.3d 84, 89 (5th Cir.
2014); Jacks v. Meridian Res. Co., LLC, 701 F.3d 1224, 1232–35 (8th Cir. 2012); and
Bennett v. MIS Corp., 607 F.3d 1076, 1088 (6th Cir. 2010).
78
Defendants emphasize that the agreements: (1) “set forth detailed ‘fuel specifications’ that

required compliance with specified American Society for Testing and Materials standards,

and compelled NEXCOM to ‘have a qualified independent source analyze the products’

for compliance with those specifications”; (2) “authorized the Contracting Officer to

inspect delivery, site, and operations”; and (3) “established detailed branding and

advertising requirements.” Defs.’ Reply Br. 19–20 (footnotes omitted). But we have

reviewed the contractual provisions cited by Defendants, and they are a far cry from the

type of close supervision that existed in both Sawyer and Winters. See Sawyer, 860 F.3d

at 253 (noting that the Navy provided “highly detailed ship [and military] specifications”

that boilers were required to match and exercised “intense direction and control . . . over

all written documentation to be delivered with its naval boilers,” including warnings);

Winters, 149 F.3d at 398–99 (noting that the Department of Defense required Dow

Chemical to provide Agent Orange under threat of criminal sanctions, maintained strict

control over the chemical’s development, and required that it be produced according to its

specifications); cf. Isaacson v. Dow Chem. Co., 517 F.3d 129, 138 (2d Cir. 2008) (rejecting

“off-the-shelf argument” because “commercially available products did not contain the

Agent Orange herbicides in a concentration as high as that found in Agent Orange”).

Rather, the cited provisions seem typical of any commercial contract. They are incidental

to sale and sound in quality assurance. 21

21
In light of the misleading-marketing allegations that are at the center of Baltimore’s
Complaint, we pause to note that the “detailed branding and advertising requirements”
cited by Defendants have absolutely nothing to do with those allegations. They simply
(Continued)
79
 b.

Next up are the oil and gas leases. Defendants allege that Chevron and “other

Defendants” have extracted oil and gas on the federal OCS pursuant to a leasing program

administered by the Secretary of the Interior under the OCSLA. J.A. 212; see, e.g., J.A.

233–39 (boilerplate lease); see also Ctr. for Sustainable Econ. v. Jewell, 779 F.3d 588, 592

(D.C. Cir. 2015) (“The [OCSLA] created a framework to facilitate the orderly and

environmentally responsible exploration and extraction of oil and gas deposits on the OCS.

It charges the Secretary of the Interior with preparing a program every five years containing

a schedule of proposed leases for OCS resource exploration and development.”).

The leases grant lessees “the exclusive right and privilege to drill for, develop, and

produce oil and gas resources” in the submerged lands of the OCS in exchange for certain

royalties on production, see J.A. 233–34, and requires them to exercise diligence in the

development of the leased area by engaging in exploration, development, and production

activities in accordance with government-approved plans, see J.A. 234; see also 30 C.F.R.

§§ 550.200–.299 (expounding plans referenced in lease). The leases also place certain

conditions on the disposition of oil and gas that is produced. Defendants highlight two

such conditions. The first mandates that twenty percent of production be offered to “small

or independent refiners.” J.A. 235. The second gives the government a right of first refusal

address whether and when the government will market a branded product under a
contractor’s brand or trade name. See Exs. F and G to Decl. of Arnold Walton, Mayor &
City Council of Balt. v. BP P.L.C., 388 F. Supp. 3d 538 (D. Md. 2019) (No. 1:18-cv-02357-
ELH), ECF Nos. 127-6 at 23, 127-7 at 15).
80
to purchase all production “[i]n time of war or when the President of the United States shall

so prescribe.” J.A. 235.

Defendants argue that the foregoing provisions demonstrate that the Defendant

lessees were “acting under” the Secretary of the Interior in extracting, producing, and

selling fossil-fuel products on the OCS. We disagree.

For starters, we note that many of lease terms are mere iterations of the OCSLA’s

regulatory requirements. Though OCS resource development is highly regulated,

“differences in the degree of regulatory detail or supervision cannot by themselves

transform . . . regulatory compliance into the kind of assistance” that triggers the “acting

under” relationship. See Watson, 551 U.S. at 157. Of course, the presence of a contractual

relationship (here, a lease) is an important distinction. But we are skeptical that the

willingness to lease federal property or mineral rights to a private entity for the entity’s

own commercial purposes, without more, could ever be characterized as the type of

assistance that is required to trigger the government-contractor analogy. See, e.g., Bd. of

Cnty. Comm’rs of Boulder Cnty. v. Suncor Energy (U.S.A.) Inc., 405 F. Supp. 3d 947, 977

(D. Colo. 2019) (“At most, the leases appear to represent arms-length commercial

transactions whereby ExxonMobil agreed to certain terms (that are not in issue in this case)

in exchange for the right to use government-owned land for their own commercial

purposes.”), aff’d, 25 F.4th at 1250–54.

Moreover, we need not decide whether the OCSLA leases are distinguishable from

other more run-of-the-mill natural-resources leases because they implicate national energy

needs. Either way, we are not convinced that the supervision and control to which OCSLA

81
lessees are subject connote the sort of “unusually close” relationship that courts have

previously recognized as supporting federal officer removal. See Watson, 551 U.S. at 153–

54; see also supra Part IV.H.1 (discussing Winters and Sawyer). As Baltimore points out,

the leases do not appear to dictate that Defendants “extract fossil fuels in a particular

manner.” Baltimore’s Resp. Br. 18. Nor do they appear to vest the government with

control over “the composition of oil or gas to be refined and sold to third parties,” let alone

purport to affect “the content or methods of Defendants’ communications with customers,

consumers, and others about Defendants’ [fossil-fuel] products.” Id.; accord Suncor

Energy, 405 F. Supp. 3d at 976–77. 22

Finally, even to the extent that the OCSLA leases toe the “acting under” line, we

still agree with the district court’s analysis as to § 1442’s third prong. Any connection

between fossil-fuel production on the OCS and the conduct alleged in the Complaint is

simply too remote.

To satisfy the third prong, the conduct charged in the Complaint need only “relate

to” the asserted official authority. See Sawyer, 860 F.3d at 257–58; see also 28 U.S.C.

§ 1442(a)(1) (“for or relating to any act under color of such office” (emphasis added)).

22
Defendants do not seriously contend otherwise. Instead, in their documents here and
below, they repeatedly point to the same lease provisions that we cite above, without
further explanation. This is a complex case, and we do not intend to suggest that
Defendants were required to outline the leases’ requirements in painstaking detail in order
to satisfy their burden of justifying federal officer removal. But they must provide
“‘candid, specific and positive’ allegations that they were acting under federal officers.”
In re MTBE, 488 F.3d at 130 (citation omitted) (quoting Willingham v. Morgan, 395 U.S.
402, 408 (1969)). Here, the lack of any specificity as to federal direction leaves us unable
to conclude that the leases rise to the level of an unusually close relationship, as required
by the first “acting under” prong.
82
That is, there must be “a connection or association between the act in question and the

federal office.” Sawyer, 860 F.3d at 258 (emphasis omitted) (quoting Papp v. Fore-Kast

Sales Co., 842 F.3d 805, 813 (3d Cir. 2016)). We elaborated upon this requirement in

Sawyer. There, we held that the district court imposed “a stricter standard of causation

than that recognized by the statute” by demanding a showing of “specific government

direction” as to whether the defendant manufacturer should have warned shipyard workers

who assembled boilers for use aboard naval vessels about the dangers of asbestos, which

was a component of the boilers manufactured by the defendant under a contract with the

Navy. See id. at 252, 258. Notably, the Navy required the use of asbestos in boilers despite

its known dangers and dictated the content of the warnings that accompanied the boilers.

The defendant manufacturer complied with those requirements. Accordingly, we

concluded that the defendant’s performance of the contract was “sufficient to connect the

plaintiffs’ claims, which fault[ed] warnings that were not specified by the Navy, to the

warnings that the Navy specified and with which [the defendant] complied.” Id. at 258

(emphasis added); see also id. (“These claims undoubtedly ‘relat[e] to’ all warnings, given

or not, that the Navy determined in its discretion.” (alteration in original)).

In this case, the district court held that even if the “acting under” and “colorable

federal defense” requirements were satisfied, Defendants did not plausibly assert that the

charged conduct was carried out “for or relating to” the alleged official authority, given the

“wide array of conduct” for which they were sued. See BP P.L.C., 388 F. Supp. 3d at 568–

69. Specifically, the court explained that Defendants were sued “for their contribution to

climate change by producing, promoting, selling, and concealing the dangers of fossil[-

83
]fuel products,” and yet failed to show that a federal officer “controlled their total

production and sales of fossil fuels,” or “directed them to conceal the hazards of fossil fuels

or prohibited them from providing warnings to consumers.” Id. at 568.

On appeal, Defendants take issue with primarily two aspects of the district court’s

analysis. First, they argue that the lack of direction as to concealment or warnings is

irrelevant to some of Baltimore’s claims, namely, strict liability for design defect. Second,

they contend that a lack of control as to total production and sales is not dispositive under

Sawyer’s relaxed reading of the third “nexus” prong.

We disagree with Defendants on both fronts. When read as a whole, the Complaint

clearly seeks to challenge the promotion and sale of fossil-fuel products without warning

and abetted by a sophisticated disinformation campaign. Of course, there are many

references to fossil-fuel production in the Complaint, which spans 132 pages. But, by and

large, these references only serve to tell a broader story about how the unrestrained

production and use of Defendants’ fossil-fuel products contribute to greenhouse gas

pollution. Although this story is necessary to establish the avenue of Baltimore’s climate-

change-related injuries, it is not the source of tort liability. Put differently, Baltimore does

not merely allege that Defendants contributed to climate change and its attendant harms by

producing and selling fossil-fuel products; it is the concealment and misrepresentation of

the products’ known dangers—and the simultaneous promotion of their unrestrained use—

84
that allegedly drove consumption, and thus greenhouse gas pollution, and thus climate

change. 23

For this reason, the lack of federal control over the production and sale of all fossil-

fuel products is relevant to the nexus analysis, and the district court did not err in relying

upon that fact when finding that any connection between the charged conduct and the

asserted official authority was even further diminished. If production and sales went to the

heart of Baltimore’s claims, we might be inclined to think otherwise. After all, the alleged

government-directed conduct (here, the production and sale of fossil fuels extracted on the

23
The same holds true for Baltimore’s strict-liability design-defect claim. As Defendants
point out, design-defect claims generally focus on “the product itself,” rather than “the
conduct of the manufacturer.” Phipps v. Gen. Motors Corp., 363 A.2d 955, 958 (Md.
1976). But that is not how Baltimore has framed its claim. Instead, Baltimore relies on
the same misleading-marketing and denialist-campaign allegations cited above, averring
that Defendants not only failed to warn the public about the climate effects they knew
would result from the normal use of their products, but also took affirmative steps to
misrepresent the nature of those risks, such as by disseminating information aimed at
casting doubt on the integrity of scientific evidence that was generally accepted at the time
and by advancing their own pseudo-scientific theories. According to Baltimore, these
tactics “prevented reasonable consumers from forming an expectation that fossil-fuel
products would cause grave climate changes.” J.A. 161; see also Maryland v. Exxon Mobil
Corp., 406 F. Supp. 3d 420, 461 (D. Md. 2019) (explaining that Maryland applies a
consumer-expectation test in design-defect cases, and only applies the risk-utility test when
the product malfunctions in some way (citing Halliday v. Sturm, Ruger & Co., 792 A.2d
1145 (Md. 2002)). Under Baltimore’s own theory of liability, then, its design-defect claim
hinges on its ability to demonstrate that Defendants’ promotional efforts deprived
reasonable consumers of the ability to form expectations that they would have otherwise
formed. Though we agree with Defendants that Baltimore’s theory appears to be a novel
one, at least in the design-defect context, this may be a function of the unique circumstances
that have allegedly given rise to this litigation. For our purposes, it is sufficient that
Baltimore has limited its design-defect theory to one that turns on the promotion
allegations, which have nothing to do with the action purportedly taken under federal
authority. The viability of such a theory under Maryland law is a question for the Maryland
courts to decide.
85
OCS) need only “relate to” the conduct charged in the Complaint. But given the foregoing

allegations, we agree with the district court’s conclusion that the relationship between

Baltimore’s claims and any federal authority over a portion of certain Defendants’

production and sale of fossil-fuel products is too tenuous to support removal under § 1442.

In sum, we hold that the Defendants who participated in the OCSLA leasing

program were not “acting under” federal officials in extracting and producing fossil fuels

on the OCS, and any connection between such activity and Baltimore’s claims is too

attenuated in any event.

c.

That leaves the 1944 unit agreement governing the operation of the Elk Hills

Reserve. Because the agreement has a complicated history, we begin with its origin and

purpose, followed by a general overview of its terms (or at least those in dispute). In the

end, however, we decline to pass on the question of whether it satisfies the “acting under”

prong. Like the OCSLA leases, we hold that the agreement fails to meet the third prong.

i.

The Elk Hills Reserve is located in Kern County, California, and originated from a

1912 Executive Order.

At the turn of the [twentieth] century, Government lands in the West were
rapidly being turned over to private ownership. At the same time, there was
a growing realization of the importance of oil for the Navy, which was then
changing its ships from coal to oil burning. In response to arguments that the
Government should preserve oil for Naval purposes, President Taft withdrew
large portions of land in California and Wyoming from eligibility for private

86
 ownership, and in 1912 set aside [the Elk Hills Reserve] by an Executive
Order. . . .

The establishment of the Reserve was expressly made subject to pre-existing
private ownership. There are approximately 46,000 acres within the Reserve,
approximately one-fifth [was] owned by [the Standard Oil Company of
California] and the remainder, approximately four-fifths by Navy. The
Standard lands [were] not in one block, but [were] checker-boarded
throughout the Reserve. The Executive Order establishing the Reserve
affected the Government lands in the field as far as future use and disposition
were concerned, but it had no effect on the privately owned lands, and the
owners of those lands were free to use and dispose of them as they saw fit.

United States v. Standard Oil Co., 545 F.2d 624, 626–27 (9th Cir. 1976). 24

Because production from one part of the Elk Hills Reserve could have reduced the

amount of oil underlying another part of the Reserve, the Navy and Standard Oil (a Chevron

predecessor) initially “had an understanding to the effect that neither would drill

wells . . . without six months’ notice to the other.” Id. at 627; see also id. (explaining that

underlying both parties’ lands were “separate accumulations of hydrocarbons,” which,

“unlike solid minerals, do not remain in place but move because of changes in underground

pressure and [thus] move toward producing wells”). But the tension between Standard’s

legitimate goal of producing oil on its land and the Navy’s duty to conserve its

hydrocarbons in the ground until needed in an emergency became untenable on the brink

of World War II. So the parties began negotiations over “an exchange, purchase or

condemnation of Standard’s land in the Reserve on the one hand, or their operation as a

unit with the Navy land,” on the other. Id.

24
Standard Oil involved a prior dispute over the same agreement, in which the Ninth
Circuit endorsed the foregoing summary agreed upon by the parties in a pretrial statement.
87
 These negotiations ultimately resulted in the 1944 Unit Plan Contract (“UPC”). 25 A

“unit agreement” is “a common arrangement in the petroleum industry where two or more

owners have interests in a common pool,” which is operated as a “unit.” Id. The parties

share production and costs in agreed-upon proportions, and, ordinarily, the objective is “to

produce currently, at minimum expense and pursuant to good engineering practices.” Id.

The UPC involved here, however, was unique in that “its purpose was not to produce

currently, and its effect was to conserve as much of the hydrocarbons in place as was

feasible until needed for an emergency.” Id. “This required curtailing production of

Standard’s hydrocarbons along with that of Navy, for which Standard would have to

receive compensation.” Id. Accordingly, “in consideration for Standard curtailing its

production plus giving up certain other rights,” id. at 627–28, the UPC gave Standard the

right to take specified volumes of oil from certain zones in the pool—namely, an average

of 15,000 barrels per day, or a lesser amount fixed by the Secretary of the Navy, with (1) a

ceiling of 25,000,000 barrels or one-third of Standard’s total share, whichever was less,

and (2) a floor of an amount sufficient to cover Standard’s out-of-pocket expenses in

maintaining the Reserve in good oil-field condition, see id. at 628; J.A. 245–46, 250–52.

25
The parties entered into an earlier contract in 1942, but it was voluntarily terminated in
1943 due to doubts expressed by the Attorney General as to its legality. Id. The parties
entered into the UPC in 1944, after Congress passed enabling legislation. See id. The UPC
governed the joint operation and development of three initial “commercially productive
zones” underlying the Elk Hills Reserve, two of which contained oil (the Stevens Zone and
Shallow Oil Zone). Only the latter zone is at issue here, and all of the provisions discussed
in this opinion pertain to that zone.
88
 ii.

With this background in mind, we turn to the specific UPC provisions relied upon

by Defendants to establish that one of their predecessors (Standard) “acted under” the Navy

when it engaged in fossil-fuel production during the twentieth century.

In the main, Defendants stress that the UPC gave the Navy “exclusive control over

the exploration, prospecting, development, and operation of the [Elk Hills] Reserve,” and

the “full and absolute power to determine . . . the quantity and rate of production from[]

the Reserve.” Defs.’ Reply Br. 18 (second alteration in original) (citation omitted); accord

J.A. 249–50. In particular, they note that the UPC “obligated” Standard “to operate the

Reserve in such manner as to produce ‘not less than 15,000 barrels of oil per day,’” and

allowed the Navy to suspend or increase the rate of production in its “discretion.” Defs.’

Reply Br. 18–19 (quoting J.A. 250) (citing J.A. 250–51).

Baltimore counters that these provisions do not establish that Standard was

producing oil at the direction of a federal officer. According to Baltimore, these provisions

merely required that the pool be maintained in a manner that would have made it capable

of producing at least 15,000 barrels per day until Standard received its share under the

contract. See J.A. 250 (“Until Standard shall have received . . . its share of production . . . ,

the Reserve shall be developed and operated in such manner and to such extent as will, so

far as practicable, permit production . . . to be maintained at a rate sufficient to produce

therefrom not less than 15,000 barrels of oil per day . . . .”). As a result, Baltimore argues

that Standard could have complied with the contract by producing no oil at all, unless and

89
until the Navy elected to increase the rate of production via congressional authorization. 26

And even then, Baltimore says, the contract did not necessarily make Standard responsible

for production on the Navy’s behalf. See generally J.A. 249 (“Navy shall, subject to the

provisions hereof, have the exclusive control over the exploration, prospecting,

development, and operation of the Reserve, and Navy may, in its discretion, explore,

prospect, develop, and/or operate the Reserve directly with its own personnel or it may

contract for all or any part of such [activities] with competent and responsible parties[,

including] . . . Standard . . . .” (emphasis added)).

At our first oral argument, Defendants shifted their focus away from whether the

15,000-barrels-per-day provision actually required Standard to produce any oil, as they

argued in their briefs. Instead, Defendants pointed to the Naval Petroleum Reserves

Production Act of 1976 (“1976 Act”), which “authorized and directed” the Secretary of the

Navy to produce the Elk Hills Reserve “at the maximum efficient rate consistent with

sound engineering practices for a period not to exceed six years . . . .” Naval Petroleum

Reserves Production Act, Pub. L. No. 94-258, § 201(3), 90 Stat. 303, 308 (1976); see also

supra note 26 (discussing UPC’s congressional-authorization requirement). Congress

26
See generally J.A. 246 (“[The UPC] does not and cannot, in and of itself, authorize the
production of any of Navy’s share of the oil, . . . as distinct from that portion of Standard’s
share hereinafter permitted to be produced and received by Standard under the terms of
[the above-cited provisions]. The production of the remainder of Standard’s share and of
all of Navy’s share must, except for the purpose of protecting, conserving, maintaining, or
testing the Reserve, be preceded by and based upon [congressional] authorization . . .; and
references hereinafter to an authorization or election by Navy to order the production of
any such oil are intended to be limited to action by the Navy within the terms of any such
[authorization].”).

90
authorized this increase in production after determining that “the Navy’s intent to maintain

a petroleum reserve, in case of national emergency in 1944, was no longer relevant,”

Chevron U.S.A., Inc. v. United States, 71 Fed. Cl. 236, 244 (2006), and in response to the

1973 oil crisis, J.A. 214. The 1976 Act also gave the Secretary the authority “to sell or

otherwise dispose of the United States share of such petroleum produced from” the Elk

Hills Reserve. See 90 Stat. at 308.

Shortly thereafter, in 1977, Congress transferred authority over the Elk Hills

Reserve to the Department of Energy and assigned to it the Navy’s interest in the Reserve

as well as the UPC. Chevron, 71 Fed. Cl. at 244–45. Standard, and later Chevron as a

successor, “continued its interest in the joint operation” of the Reserve until 1997. J.A.

214.

iii.

The parties’ dispute about the UPC and its significance for purposes of federal

officer removal thus can be distilled to two main issues. First, was any oil ever produced

from the Elk Hills Reserve at the Navy’s direction? And second, if so, was it Standard

who carried out those orders?

In light of the 1976 Act, we think the answer to the first question is yes. But as to

the second, we simply have no idea whether production authorized by Congress was carried

out by Standard. At our first oral argument, counsel for Chevron merely stated that it was

his “understanding” that Standard extracted oil on the Navy’s behalf under the unit

agreement, and, more generally, that the government relies upon private companies

91
because it does not have its own oil and gas engineers or drilling equipment. And although

counsel later submitted a Rule 28(j) letter stating that the government had final authority

over all production, “which was carried out by Standard, and later Chevron,” Defs.’ Letter

of Suppl. Authorities 1, ECF No. 133, the letter merely cites the UPC as a whole in support

of this assertion. In other words, it does not explain why Baltimore’s reliance on the

operational-control provision cited above is misplaced, see J.A. 249, nor does it point to

any other provision or provisions that support a different reading. 27 Thus, we are left

wanting for pertinent details about Standard’s role in operating the Elk Hills Reserve and

producing oil therefrom on behalf of the Navy, which might bear directly upon the “acting

under” analysis. Indeed, if Standard was not responsible for producing the oil authorized

by Congress in 1976, the upshot is that any extensive government control contemplated by

the UPC only affected the parties’ relative shares and the development of the Reserve, not

Standard’s duties with respect to any production carried out for the Navy’s benefit.

Nevertheless, even if we were to conclude that Standard was responsible for such

production under the UPC—and that this responsibility transformed Standard into a person

“acting under” the Navy for purposes of § 1442—the production of oil from the Elk Hills

Reserve by the predecessor of one of the twenty-six Defendants, like the production of

fossil fuels on the OCS, is not sufficiently “related” to Baltimore’s claims. See supra Part

27
Because Baltimore only claimed that Standard was not responsible for production at oral
argument—in response to Defendants’ reliance on the 1976 Act, which Defendants, in turn,
did not rely upon in their briefs on appeal—this issue is not addressed in Defendants’
briefing, either. Nor can we find any relevant explanation in the federal-officer allegations
in the Notice of Removal.

92
IV.H.2.b. Accordingly, the district court was correct in concluding that the UPC cannot

support federal officer removal in this case.

V.

The impacts of climate change undoubtably have local, national, and international

ramifications. See Massachusetts, 549 U.S at 521–53 (noting that the harms associated

with climate change are “serious and well recognized”). But those consequences do not

necessarily confer jurisdiction upon federal courts carte blanche. In this case, a

municipality has decided to exclusively rely upon state-law claims to remedy its own

climate-change injuries, which it perceives were caused, at least in part, by Defendants’

fossil-fuel products and strategic misinformation campaign. These claims do not belong

in federal court. Given the jurisdictional inquiry before us, we take no view on whether

Baltimore will ultimately fail or succeed in proving its claims under Maryland law. We

cannot decide those questions. But we are confident that Maryland courts can capably

adjudicate claims arising under their own laws that fail to otherwise provide any federal

jurisdiction. Because we do not discern a proper basis for removal that permits a federal

court to entertain Baltimore’s action, the district court’s order granting Baltimore’s Motion

to Remand is

AFFIRMED.

93