Symposium - 2025 - Promises of Sovereignty
Can the Roberts Court Find Federal Indian Law?
Seth Davis *
Introduction
Imagine the lost world of “lawfinding.” 1 It’s easy if you try. See Loper Bright Enterprises v. Raimondo, 144 S. Ct. 2244, 2285 (2024) (Gorsuch, J., concurring) (referring to “a very old idea, one that constrains judges to a lawfinding rather than lawmaking role”); Stephen E. Sachs, Finding Law, 107 Calif. L. Rev. 527, 529 (2019) (describing the view that unwritten law is made, not found, as the “modern approach” and tracing it to Erie Railroad Co. v. Tompkins, 304 U.S. 64 (1938)). In that world, there was a general common law for federal judges to find. 2 See, e.g., Swift v. Tyson, 41 U.S. (16 Pet.) 1, 18-19 (1842) (finding general common law). And in that world, each statute had a “single, best meaning” 3 Loper Bright, 144 S. Ct. at 2266 (majority op.). for judges to unearth with the traditional tools of statutory interpretation.
Of course, we are not going back to that world. Too much has happened. Take Erie Railroad Co. v. Tompkins, a battle over corporate power that taught us there is no “federal general common law” out there to be found. 4 Erie, 304 U.S. at 78; see id. at 79 (rejecting the “assumption that there is ‘a transcendental body of law outside of any particular State’” (quoting Black & White Taxicab & Transfer Co. v. Brown & Yellow Taxicab & Transfer Co., 276 U.S. 518, 533 (1928) (Holmes, J., dissenting)). On the history of corporate litigants’ exploitation of diversity jurisdiction leading up to Erie, see Edward A. Purcell, Jr., Litigation and Inequality: Federal Diversity Jurisdiction in Industrial America 1870-1958, at 226-30 (1992). And then there was the battle of the bubble: Chevron USA Inc. v. Natural Resources Defense Council. 5 467 U.S. 837, 865 (1984) (holding that federal courts should defer to permissible agency interpretations of ambiguous statutes that they administer). On the controversy about the Environmental Protection Agency’s bubble policy and Chevron, see Thomas W. Merrill, The Story of Chevron: The Making of an Accidental Landmark, 66 Admin L. Rev. 253 (2014). Its lesson was that interpreting ambiguous statutes involves policymaking, which is for agencies to do when ambiguities appear in statutes they administer. 6 Chevron, 467 U.S. at 865 (“Courts must, in some cases, reconcile competing political interests, but not on the basis of the judges’ personal policy preferences. In contrast, an agency to which Congress has delegated policymaking responsibilities may, within the limits of that delegation, properly rely upon the incumbent administration’s views of wise policy to inform its judgments.”).
And yet, there are signs of that lost world all round us. Chevron is gone now. The problem with Chevron deference, which the Roberts Court overruled in Loper Bright Enterprises v. Raimondo, was that it was a judge-made “fiction” that confused construing the law with lawmaking. 7 Loper Bright, 144 S. Ct. at 2268-69, 2273 (overruling Chevron, 467 U.S. 837). In other words, Chevron failed at “lawfinding,” which, as Justice Gorsuch’s concurring opinion in Loper Bright put it, was the “traditional” task of judges. 8 See id. at 2277 (Gorsuch, J., concurring). Somewhere out there is Swift v. Tyson, 9 41 U.S. (16 Pet.) 1 (1842). still lumbering along, long after Erie’s impact. 10 See Abbe R. Gluck, Overruling Chevron Without a Coherent Theory of Statutory Interpretation and the Court-Congress Relationship, 62 Harv. J. Legis. 20, 26 (2024) (“Justice Gorsuch writes [in Loper Bright] as if Erie was never decided.”).
Administrative law is not the only field in which something Swiftian can be found in the jurisprudence of one (or more) of the Roberts Court justices. Federal Indian law, the subject of this symposium, is another. Federal Indian law is a field full of fundamental principles that were unwritten when the Supreme Court first recognized them. Debates about these principles have pitted members of the Roberts Court’s conservative majority against each other, with even Justice Gorsuch facing an accusation that he has failed at the traditional task of lawfinding. 11 Compare Oklahoma v. Castro-Huerta, 142 S.Ct. 2486, 2504 (2022) (Kavanaugh, J., writing for the Court) (“[Justice Gorsuch’s] dissent employs extraordinary rhetoric in articulating its deeply held policy views . . . . But this Court’s proper role under Article III of the Constitution is to declare what the law is, not what we think the law should be.”), with id. at 2505 (Gorsuch, J., dissenting) (“Where our predecessors refused to participate in one State’s unlawful power grab at the expense of the Cherokee [Nation], today’s Court accedes to another’s.”).
Can the Roberts Court find federal Indian law? The answer, I think, is “no,” if by finding law, we mean a process of decisionmaking that makes no normative judgments but instead engages in an empirical survey of written texts, common law decisions, and customary practices. Any effort to find federal Indian law should take seriously the overflowing well of its sources, including treaties and other agreements between tribal nations and the United States, the Constitution, statutes, regulations, executive orders, judicial decisions, customs, and practices, not to mention tribal laws. Selecting among these sources, and deciding which to treat as authoritative, is no mean task.
In some opinions, most notably Haaland v. Brackeen, the Roberts Court has recognized the seriousness of the problem. 12 See 143 S.Ct. 1609, 1627-29 (2023). In others, particularly Oklahoma v. Castro-Huerta and Arizona v. Navajo Nation, the Roberts Court has published opinions in the style of a modest science of lawfinding that are actually manifestos for radical legal change. 13 See generally 142 S.Ct. at 2504; 143 S.Ct. 1804, 1813-14 (2023) (concluding that no “duty on the United States to take affirmative steps to secure water for the Tribe” could be found in positive law). This science, I will argue, fails as an exercise in lawfinding and as a theory of legal change in federal Indian law. 14 On the distinction between law as a “normal science” and moments of a “paradigm shift,” see Hanoch Dagan, The Real Legacy of American Legal Realism, 38 Oxford J. Legal Stud. 123, 134-35 (2018).
My argument proceeds in three parts. Part I sketches the Roberts Court’s renewed interest in lawfinding through the lens of Loper Bright. Part II focuses upon the problem of sources. It maps the multiple modes of the Roberts Court’s Indian law jurisprudence, which show that the Court has no settled answer to the fundamental question of the role of the Article III “judicial Power” 15 U.S. Const. art. III, § 1. in federal Indian law. Part III builds upon this description of the sources of federal Indian law and the role of the federal courts as it focuses upon the problem of legal change. It discusses Castro-Huerta and Arizona v. Navajo Nation, which radically changed the law by changing the sources that the Court searched for authoritative rules of decision. In both cases, the Court did not treat the judicial task as one of working through “the seeming commands of the authorities and the felt demands of justice.” 16 .Karl N. Llewellyn, The Common Law Tradition: Deciding Appeals 37-38 (1960). It was none other than Justice Gorsuch—the “lawfinder” of Loper Bright—who dissented and gave voice to the impulse to harmonize “vision with tradition,” 17 See id. at 37; see also Dagan, supra note 14, at 136 (arguing that legal realists understood adjudication, “at its best,” in these terms). recognizing that finding federal Indian law requires more than modest legal science.
I. Finding Law in Loper Bright
Judicial independence, Chief Justice Roberts wrote in his 2024 year-end report, is America’s gift to the “art of government.” 18 Actually, Chief Justice Roberts’ former boss, Chief Justice Rehnquist, wrote those words, which his successor quoted. Chief Justice John Roberts, 2024 Year End Report on the Federal Judiciary 3 (Dec. 31, 2024) [hereinafter Roberts Report] (quoting William H. Rehnquist, Judicial Independence, 38 U. Rich. L. Rev.579, 579-80 (2004)), https://perma.cc/2W8B-BCTX. Independent judges do not cater to politicians or buckle before them. Rather, independent judges “construe the law with ‘[c]lear heads . . . and honest hearts.’” 19 Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244, 2268 (2024) (Roberts, C.J., writing for the Court) (quoting 1 The Works of James Wilson 363 (J. DeWitt Andrews ed. 1896)).
Call this the Friday Night Lights theory of judging. 20 As contrasted with the just-calling-balls-and-strikes theory. See Confirmation Hearing on the Nomination of John G. Roberts, Jr. to be Chief Justice of the United States: Hearing Before the S. Comm. on the Judiciary, 109th Cong. 56 (2005) (statement of John G. Roberts, Jr., J., D.C. Cir.). In Friday Night Lights, a TV drama about high-school football in small town Texas, Coach Eric Taylor inspires his players with a saying: “clear eyes, full hearts, can’t lose.” See Emily St. James, Friday Night Lights’ Jason Katims on the Long Life of “Clear Eyes, Full Hearts, Can’t Lose,” Vox, Mar. 29, 2018, https://perma.cc/7X4G-MKCK. After Loper Bright, it is the law of the land. 21 Loper Bright, 144 S. Ct. at 2273 (“Courts must exercise their independent judgment in deciding whether an agency has acted within its statutory authority, as the APA requires.”). Indeed, according to Chief Justice Roberts’ opinion in Loper Bright, it has always been the law: The original understanding was that Article III “courts must exercise their independent judgment” when construing the law. 22 Id. at 2262.
Independent judges construe written law to find the best reading. Every statute has—”in fact, must” have 23 Id. at 2266 (emphasis added). Must it, though? See Ryan D. Doerfler, Can a Statute Have More Than One Meaning?, 94 N.Y.U. L. Rev. 213, 228-42 (2019) (arguing that a statute can have more than one meaning). —a “single, best meaning.” 24 Loper Bright, 144 S. Ct. at 2266. The judge’s job is to find it. To do so requires “independent judgment,” not deference to another’s (say, an administrative agency’s) interpretation. 25 Id. at 2261-62 (holding that APA requires reviewing courts to “exercise independent judgment,” not to defer to agency constructions of ambiguous statutes). That is the holding of Loper Bright and a statement of confidence in judicial competence. 26 Namely, the Roberts Court’s competence, not the competence of the Burger Court that decided Chevron. See id. at 2273 (overruling Chevron U.S.A. Inc. v. Nat. Res. Def. Council, 467 U.S. 837 (1984)).
But what is the independent judge to do when they conclude that the written law does not answer the question before them? Here too, Loper Bright suggests, the independent judge must confidently try to find the law, not make it up. According to Loper Bright, Chevron was a judge-made “fiction” at odds with the true meaning of the Administrative Procedure Act, not mention Marbury v. Madison, the urtext of judicial review. 27 Id. at 2268; see id. at 2257 (“Chief Justice Marshall famously declared that ‘[i]t is emphatically the province and duty of the judicial department to say what the law is.’”) (quoting Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803)); id. at 2261 (“The APA . . . codifies for agency cases the unremarkable, yet elemental proposition reflected by judicial practice dating back to Marbury: that courts decide legal questions by applying their own judgment.”). The Loper Bright Court held that the APA, which directs reviewing courts to decide “all relevant questions of law,” 28 5 U.S.C. § 706. requires them to do so independently without deferring to the agency under Chevron. 29 Loper Bright, 144 S. Ct. at 2265-66. In this, the APA codified the “traditional understanding” of the Article III “judicial Power,” which is a power to exercise judgment to decide cases. 30 Id. at 2262.
Among its many striking features, Loper Bright held that rules of statutory interpretation like Chevron are law—that is, they have stare decisis effect. 31 See Gluck, supra note 10, at 32 (“Loper Bright treats Chevron as a precedent—as conmon law to be overruled.”). That is why Chief Justice Roberts’s opinion had to address at length the grounds for dispensing with stare decisis. 32 See Loper Bright, 144 S. Ct. at 2270. Thus, Chevron was part of the positive law for interpreting agency-administered statutes, at least until Loper Bright overruled it.
Thus, in treating the law for interpreting statutes as a body of positive law, and in tracing the requirement of “independent judgment” to the APA and the original understanding, the Roberts Court implied that that judges must find law, including law that constrains how they will construe the law.
Lurking here are knotty, longstanding questions of jurisprudence, including questions about what counts as “the law” that is being construed and what the status of methods for construing law is given pervasive disagreement about them. 33 See, e.g., H.L.A. Hart, The Concept of Law 3, 273-74 (3d ed. 2012); Ronald Dworkin, Law’s Empire 3-11 (1986). There has been something of a revival in the idea of finding law in recent years. Erie, it seems, may have been “wrong.” 34 See Sachs, supra note 1, at 530 (arguing that if Erie is interpreted to deny that lawfinding is possible, “then Erie is wrong”). Finding law is possible. And it is desirable. Trying to find law, rather than make it, wields judicial power modestly: Indeed, it might be thought to define the “judicial Power” in Article III. 35 Cf. Loper Bright, 144 S. Ct. at 2277 (Gorsuch, J., concurring) (suggesting that the Framers meant to refer to the idea of “lawfinding” “when they spoke of the ‘judicial Power’”). It is desirable too because it may bring coherence to the doctrine, especially in the hard cases of constitutional law that occupy so many pages of the federal reporters and the law reviews. 36 Danielle D’Onfro & Daniel Epps, The Fourth Amendment and General Law, 132 Yale L.J. 910, 945 (2023).
That is not to suggest lawfinding is always easy. Stephen Sachs describes the task as analogous to derermining “whether a given outfit would be out of place at an important business meeting, or whether an ordinary English phrase would be rule out-of-bounds in English class.” 37 Sachs, supra note 1, at 535. While it takes diligence, “the social facts are ‘out there’ for diligent jurists to find.” 38 Id. at 531. One implication is that “[t]he law is one thing, the decisions of courts another.” 39 Id. at 532.
Consistent with this last idea, and quoting Edmund Burke, Justice Gorsuch offered criteria which test any exercise of lawfinding that relies upon past cases:
Burke offered five principles for the evaluation of past judicial decisions: “They ought to be shewn; first, to be numerous and not scattered here and there;—secondly, concurrent and not contradictory and mutually destructive;—thirdly, to be made in good and constitutional times;—fourthly, not to be made to serve an occasion;—and fifthly, to be agreeable to the general tenor of legal principles.” 40 Loper Bright, 144 S. Ct. at 2277 (2024) (Gorsuch, J., concurring) (quoting Speech of Dec. 23, 1790, in 3 The Speeches of the Right Honourable Edmund Burke 513 (1816)).
For Justice Gorsuch, it followed that Chevron could and should be overruled because it failed to comport with “mainstream currents” in American law as well as “the law” of the APA. 41 See id. at 2281. Chevron deference was not the law but, instead, a “stray and unconsidered disgression” from the path of the law. 42 Id. at 2292.
Justice Gorsuch’s link between lawfinding and legal change is telling. Legal commentators, trained in the world that Erie and Chevron made, have called Loper Bright a change in the law—indeed, a radical change. 43 See, e.g., Ian Millhiser, The Supreme Court Just Made a Massive Power Grab That it Will Come to Regret, Vox (June 28, 2024), at https://perma.cc/76K6-WDBK. Others have suggested that it is not radical at all. See, e.g., Adrian Vermeuele, Chevron by Any Other Name, The New Digest (June 28, 2024), at https://perma.cc/V9RC-CLZT (“Please, everyone, take a deep breath. . . . It is entirely possible that much or most of what was (somewhat misleadingly) called ‘Chevron deference’ can be and will be recreated under a different label: ‘Loper Bright delegation.’”). Not so, Justice Gorsuch suggested. The Court was not changing the law so much as it was finding what the law already was. 44 Loper Bright, 144 S. Ct. at 2285 (Gorsuch, J., concurring) (“Chevron deference undermines [various] aspects of our settled law.”).
This all seems to suggest that lawfinding is a modest science. Judges who find the law are modest about judicial opinions, including their own, which are “evidence” of the law but not the law itself. The law itself can be written down. It can also be unwritten. The problem of where to look to find the law, especially the unwritten law, is a problem, but a soluble one. 45 See Sachs, supra note 1, at 538 (“[T]he mystery dissolves once we remember how good we are at identifying unwritten rules.”). And the project of lawfinding can accommodate itself quite easily to legal change, even the sort of rule reversal that occurred with Loper Bright.
II. The Problem of Sources and the “Judicial Power” in Federal Indian Law
Federal Indian law is full of principles that were unwritten when the Supreme Court first recognized them. The doctrine of discovery is one. 46 See, e.g., Johnson v. McIntosh, 21 U.S. (8 Wheat.) 543, 573 (1823). The Court did not say it was creating this doctrine, but rather purported to find it. The plenary power doctrine is another example of an unwritten principle. Not surprisingly, one modern criticism is that the plenary power doctrine is something judges made up. 47 See, e.g., Haaland v. Brackeen, 143 S. Ct. 1609, 1662 (2023) (Thomas, J., dissenting) (arguing that plenary power was “born of loose language and judicial ipse dixit”).
The sources of federal Indian are myriad. Cohen’s Handbook of Federal Indian Law is worth quoting at length on this point:
Researching a federal Indian law issue may require reference to multiple sources of law extending over a broad span of time. In any given case, these sources may encompass federal, state, and tribal law, and require reference to treaties that both predate and postdate the United States Constitution, federal laws adopted as long ago as the first Congress, and tribal law of ancient or recent origin. . . .
The sources of federal Indian law include the United States Constitution, principles of international law, treaties with Indian tribes, federal statutes and regulations, executive orders, and judicial opinions. Consulting the laws of specific tribes governing the resolution of disputes involving Indians and their territory may be necessary to resolve some disputes. Tribal law may also serve as the contextual background informing the development of specific doctrines in federal Indian law. 48 .Cohen’s Handbook of Federal Indian Law § 1.04 (Nell Jessup Newton & Kevin Washburn, eds. 2025).
These many sources raise hard questions, including which sources are authoritative for any particular legal issue, how to resolve conflicts among sources over time, how to decide when to consult unwritten customs and practices in addition to or in the absence of written law, and, obviously and importantly, whose customs or practices to consult.
This Part shows that solving these problems is not easy. Much of the difficulty stems from the sheer multiplicity and number of sources. But, as we shall see, some of the difficulty stems from the changes over time in how lawyers think about what counts as a legitimate source of law and the persistence of techniques of legal reasoning from, among others, the once (and future?) world of lawfinding.
A. Constitutional Interpretation and Law from the Atmosphere
Plenary power began as an unwritten principle of sovereignty that the Court found outside the U.S. Constitution. In the earliest case, United States v. Kagama, the Court had to decide whether the Major Crimes Act (MCA), a federal statute, was within Congress’s authority to enact. 49 See United States v. Kagama, 118 U.S. 375, 376 (1886). The Court concluded that the plenary power was a sufficient basis for the Act. 50 Id. at 383 (holding that the MCA was “within the competency of Congress”). This plenary power was not contained in the Constitution. Rather, it was an incident of sovereignty. A general principle of the law was that sovereigns own their territories, “and the right of exclusive sovereignty which must exist in the National Government” provided Congress with authority to enact criminal law for Indian lands, which were “within the geographical limits of the United States.” 51 Id. at 379-80.
More recent cases have insisted that plenary power must be tied to written law. In Haaland v. Brackeen, the Court was especially insistent on this point. 52 143 S. Ct. 1609, 1627 (2023). Erie could not have put the point better: the plenary power doctrine cannot come from “the atmosphere.” 53 Id. at 1629. Recognizing that “our precedent is unwieldy, because it rarely ties a challenged statute to a specific source of constitutional authority,” the Roberts Court engaged in a mixture of legal interpretation and lawfinding to tie the plenary power doctrine to the Constitution and conclude that constitutional law constrains its exercise. For example, it interpreted the Indian Commerce Clause and the Treaty Clause as bases for the plenary power doctrine. 54 Id. at 1627-28. At the same time, the Court found the plenary power in general principles, namely, “the Constitution’s adoption of preconstitutional powers necessarily inherent in any Federal Government” and the general “trust relationship between the United States and the Indian people.” 55 Id. at 1628 (quoting United States v. Curtiss-Wright Export Corp., 299 U.S. 304, 318 (1936) and United States v. Mitchell, 463 U.S. 206, 225-26 (1983)).
Thus, the Roberts Court’s members agree that Congress’s plenary power is something that must be based on more than general notions about sovereignty from the late nineteenth century or judge-made federal common law. At the same time, the Roberts Court relied upon unwritten sources to support its interpretation of the plenary power’s source, scope, and limits in Brackeen. Justice Barrett’s opinion for the Court was frank about the “unwieldy” nature of “our precedent,” which “rarely ties a challenged statute to a specific source of constitutional authority.” 56 Id. at 1629. As a result, “it is difficult to categorize cases and even harder to discern the limits on Congress’s power.” 57 Id.
Ultimately, Justice Barrett appealed to rule of law values to address the problem of sources, resting upon a normative principle that ruled out the way that the Court approached the problem in Kagama. Contrary to way the Court reasoned in 1886, Justice Barrett stipulated in Brackeen that “[a] power unmoored from the Constitution would lack both justification and limits.” 58 Id. at 1627. That is why plenary power could not be left “free-floating” around the “atmosphere,” 59 Id. where, one worries, it might someday collide with Swift’s general common law.
B. The General Law and the New Federal Common Law in Federal Indian Law
The criticism that the plenary power doctrine is illegitimate judge-made law has traction in part because of what the Roberts Court has said about federal common law authority. Federal common law, in a word, is “disfavored.” 60 Cf. Ashcroft v. Iqbal, 556 U.S. 662, 675 (2009) (stating that implied rights of action, one type of federal common law, are “disfavored”). Summarizing this view, Justice Gorsuch has written that, “[s]ince Erie . . . , federal courts have largely disclaimed the power to develop federal common law outside of a few reserved areas.” 61 Turkiye Halk Bankasi A.S. v. United States, 143 S. Ct. 940, 955 (2023) (Gorsuch, J., concurring). Indian affairs has been one of those areas.
While they have worked to ground the plenary power doctrine in constitutional interpretation, the Justices of the Roberts Court have continued to decide federal Indian law cases through common law modes of reasoning. One mode of reasoning looks like Swift’s general common law, while another is the “new federal common law” that followed after Erie. 62 See generally Henry J. Friendly, In Praise of Erie—And of the New Federal Common Law, 39 N.Y.U. L. Rev. 383, 384 (1964) (arguing that Erie “cleared the way for the truly uniform federal common law on issues of national concern”).
1. General Law
Erie did not end reasoning from the general law, if, pace Professor Caleb Nelson, we define that concept as “rules that are not under the control of any single jurisdiction, but instead reflect principles or practices common to many different jurisdictions.” 63 Caleb Nelson, The Persistence of General Law, 106 Colum. L. Rev. 503, 505 (2006). Federal courts have continued to apply rules of decision that they justify by pointing to patterns and customs that not written laws adopted by a single sovereign. 64 See id. at 506-07. Such uses of general rule—shall we call it the “new federal general common law”?—are common in federal Indian law jurisprudence.
Not long after Erie, in fact, the Supreme Court looked to the general law of trusts as a source of judicially enforceable duties under the Indian trust doctrine. The Indian trust doctrine holds that the United States is a fiduciary for Indians and owes them duties of loyalty and care. In Seminole Nation v. United States, a landmark decision from 1942, the Court cited the Restatement of the Law of Trusts as well as Bogert’s Trusts and Trustees and quoted Meinhard v. Salmon, itself a landmark decision from the New York Court of Appeals, to conclude that the federal executive may have breached fiduciary duties it owed to the Seminole Nation in how it managed Seminole funds. 65 316 U.S. 286, 296-97 & n.12 (1942) (citing Restatement (First) of the L. of Trs. § 321 (Am. L. Inst. 1935); 4 George Gleason Bogert, The Law of Trusts and Trustees §§ 901, 955 (1935); and Meinhard v. Salmon, 164 N.E. 545, 546 (N.Y. 1942)). None of these sources, of course, were binding positive law for the case, but the Court cited them as evidence of the general law of trusts.
More recently, the Court has looked to find the general law of contracts in cases arising under the Indian Self-Determination and Education Assistance Act (ISDEAA). 66 25 U.S.C. § 5301. This Act is the backbone of the modern tribal self-determination era. Through self-determination contracts between the United States and tribal nations, the federal government provides funding to support tribal self-government and tribal government services, such as healthcare. 67 See Cherokee Nation v. Leavitt, 543 U.S. 631, 634 (2005). In a series of decisions, the Court has held that the federal government is bound by general principles of contract law to meet its obligations under the ISDEAA contracts, even when congressional appropriations fall short of what is needed to meet all of the government’s contractual obligations. Writing for the Roberts Court in Cherokee Nation v. Leavitt, Justice Breyer used general law to establish a crucial premise of the Court’s holding that the government had an enforceable obligation to pay: He defined the statutory term “contract” by looking to the Restatement (Second) of Contracts, which supported the conclusion that Congress “meant to treat alike promises made under the [ISDEAA] and ordinary contractual promises.” 68 Id. at 639 (citing Restatement (Second) of Contracts § 1 (Am. L. Inst. 1981)).
There are other examples and not all of them have led to the conclusion that tribal nations have judicially enforceable rights. In City of Sherrill v. Oneida Indian Nation, for example, Justice Ginsburg wrote for the Court in applying general principles of laches and impossibility to preclude the Oneida Indian Nation of New York from asserting its immunity to state and local taxation through a federal court injunction. 69 544 U.S. 197, 217-219 (2005). And in Upper Skagit Indian Tribe v. Lundgren, Justice Thomas wrote a dissenting opinion that drew upon four versions of the Restatement of Foreign Relations Law, treatises, and commentary on the law of nations to argue that there is an in rem exception to tribal sovereign immunity. 70 584 U.S. 554, 567-70 (2018) (Thomas, J., dissenting) (citing Hersch Lauterpacht, The Problem of Jurisdictional Immunities of Foreign States, 28 Brit. Y.B. Int’l L. 220, 244 (1951); Francis Wharton, Conflict of Laws § 273 607 (George Parmele ed., 3d ed. 1905); Restatement of Foreign Rels. L. of the U.S. § 71 cmt. c (Am. L. Inst., Proposed Official Draft 1962); Restatement (Second) of Foreign Rels. L. of the U.S. § 68(b) (Am. L. Inst. 1965); Restatement (Third) of Foreign Rels. L. of the U.S. § 455(1)(c) (Am. L. Inst. 1986); and Restatement (Fourth) of Foreign Rels. L. of the U.S. § 456(2) (Am. L. Inst., Tentative Draft No. 2, 2016)).
2. Federal Common Lawmaking
All of this suggests that there is something like the (old) general law in some Roberts Court decisions when written texts run out and the justices purport to find rules of decisions in unwritten law. There is also something like the new federal common law in other Roberts Court decisions, including those involving the implicit divestiture doctrine.
The Burger Court created the implicit divestiture doctrine in Oliphant v. Suquamish Indian Tribe. 71 435 U.S. 191, 209 (1978). See generally Alex Tallchief Skibine, Incorporation Without Assimilation: Legislating Tribal Jurisdiction over Nonmembers, 67 UCLA L. Rev. Discourse 166, 171 (2019) (“The implicit divestiture doctrine has its roots with the 1978 Oliphant decision where the Court ruled that Indian tribes had been implicitly divested of the inherent sovereign power to prosecute non-Indians.”). In an opinion by then-Justice Rehnquist, the Burger Court held that Indian tribes may not enforce their criminal laws against non-Indians who violate tribal criminal codes within Indian Country. This limit on federal recognition of tribal criminal jurisdiction was not a written principle of law. That is, the Court was not construing written law that applied to the case but either finding a general principle of law or making new law. Generally, the opinion reads as if the Court was finding law: Justice Rehnquist identified bits and pieces of evidence in a wide variety of sources, including citations to historical practice, treaties that did not apply, and proposed federal legislation that was never enacted. 72 See Oliphant, 435 U.S. at 197-199 & n.8. Thus, if anywhere, the implicit divestiture doctrine was somewhere in the “atmosphere.” 73 See Haaland v. Brackeen, 143 S.Ct. 1609, 1627 (2023).
Oliphant was a mess as an exercise in lawfinding, much less legal interpretation. This point has been made ably elsewhere. 74 See Matthew L.M. Fletcher, Muskrat Textualism, 116 Nw. U. L. Rev. 963, 1014-1015 (2022). Nevertheless, once written down in Oliphant, the implicit divestiture doctrine became the basis for what looks like federal common lawmaking.
Indeed, the Court itself has described applications of the doctrine as federal common law. This happened in United States v. Lara. 75 541 U.S. 193, 207 (2004). The background to Lara was Duro v. Reina, which extended Oliphant’s holding to shield nonmember Indians from the criminal jurisdiction of tribes. 76 495 U.S. 676, 679 (1990). Congress responded with the Duro-fix, which recognized and reaffirmed the authority of tribal nations to apply their criminal laws to all Indians within their territories. 77 25 U.S.C. § 1301(2). In Lara, the Court held that Congress had authority to abrogate Duro’s holding, treating it as “judicially made’ federal Indian law.” 78 Lara, 541 U.S. at 207 (quoting Oliphant v. Squamish Indian Tribe, 435 U.S. 191, 206 (1978).
The Court’s most recent application of the implicit divestiture doctrine reads like lawmaking. As the Court explained in United States v. Cooley, the doctrine extends beyond the enforcement of tribal criminal law to the civil jurisdiction of tribal governments. 79 United States v. Cooley, 141 S. Ct. 1638, 1643 (2021). The key case was Montana v. United States, which held as a “general proposition” that tribes do not have authority to regulate nonmember activities on fee land within tribal territories. 80 Montana v. United States, 450 U.S. 544, 565 (1981). There are two exceptions to this general principle, the second of which permits a tribe to exercise civil jurisdiction in response to threats to the “political integrity, the economic security, or the health or welfare of the tribe.” 81 Id. at 566. In Cooley, the Roberts Court applied this second Montana exception to hold that a tribal police officer may “search and detain for a reasonable time any person he or she believes may commit or has committed a crime.” 82 Cooley, 141 S. Ct. at 1643. Its opinion reads as an exercise in federal common lawmaking, one that emphasizes the practical need for tribal officers to have this authority and the unworkability of the limits that the lower court had designed. 83 See id. at 1643-45. The Court rejected the argument that its role was solely to construe federal cross-deputization statutes to find written law authorizing the tribal police to search and detain individuals. 84 Id. at 1645. Rather, it concluded that under Montana, the requisite “authority . . . rests upon a tribe’s retention of sovereignty.” 85 Id.
As with the plenary power doctrine, so too with the implicit divestiture doctrine, critics have argued that what is going on is illegitimate federal common lawmaking. 86 See Grant Christensen, Article III and Indian Tribes, 108 Minn. L. Rev. 1789, 1794 (2024); Frank Pommersheim, Tribal Courts and the Federal Judiciary: Opportunities and Challenges for a Constitutional Democracy, 58 Mont. L. Rev. 313, 328 (1997). There is an apparent conflict between the Court’s conclusion that Congress’s plenary power to expressly divest tribal nations of jurisdictional authority must be found within the Constitution and the apparent absence of any judicial power to implicitly accomplish the same outcome. As there is no obvious source of law for the implicit divestiture doctrine, the question remains what, if anything, justifies this application of the Article III judicial power to enjoin tribes from applying their own laws within their own federally-recognized territories.
III. The Problem of Legal Change and the Failures of Modest Legal Science
In this Part, I will discuss two cases that show how the problem of sources may become a problem of legal change within federal Indian law. Oklahoma v. Castro-Huerta, which dismissed a centuries-old “general notion” of law and replaced it with a new general notion, was about state authority in Indian Country. 87 142 S. Ct. 2486, 2493 (2022) (5-4 decision). Arizona v. Navajo Nation, which created a new rule for construing the law of obligations, was about the trust responsibility of the federal government to tribal nations. 88 143 S. Ct. 1804, 1815-16 (2023).
As exercises in lawfinding, and as opinions purporting to justify radical legal change in the style of a modest science of lawfinding, both were failures. Having made these choices about sources, the Court then treated the questions before it as “straightforward” 89 Castro-Huerta, 142 S. Ct. at 2502. ones of applying the law the Court had found. These moves fail on their own terms as exercises in the modest science of lawfinding, for reasons that are made clear by consulting the Burkean criteria that Justice Gorsuch quoted in Loper Bright. Moreover, the Court’s modest science in these federal Indian law cases fails to meet the problem of legal change.
There are, Hanoch Dagan has argued, “two genres of interpretive legal theories”: one for the “day-to-day operation” of law and another for moments that challenge “on-the-wall understandings” of law. 90 Hanoch Dagan, Two Genres of Interpretive Legal Theories, in Understanding Private Law: Essays in Honour of Stephen A. Smith 49, 57 (Evan Fox-Decent et al. eds., 2025). The science of the day-to-day operations of law skirts normative debates, and for good reason, as its aim is to be workable without resolving deep disagreements about justice. 91 Id. at 55. But the conventional understandings of law can persist for only so long, and when they wear down and are challenged, “the normal science of legal discourse [becomes] . . . useless.” 92 Id. at 56. What count as adequate legal reasons for paradigm shifts are different, and require serious engagement with the well-worn conventions and normative arguments challenging them 93 See id. at 58 (arguing that legal reasons for paradigm shifts “expose the contingency of the implicit background understandings . . . ; explain why they are mistaken—why they render the existing law normatively indefensible; and offer a different understanding of the very same pedigreed sources and concepts”). —the sort of serious engagement that Justice Gorsuch offered in his dissenting opinions in both Castro-Huerta and Arizona v. Navajo and that, ironically enough, elicited from the Court’s majority the charge that Justice Gorsuch was not taking the task of lawfinding seriously. 94 See infra notes 117-119and accompanying text.
A. One General Notion Yields to Another
The question in Oklahoma v. Castro-Huerta was whether a state has authority to prosecute a non-Indian person for a crime committed against an Indian person in Indian Country. 95 Oklahoma v. Castro-Huerta, 142 S. Ct. 2486, 2491 (2022). One year before, the Court held in McGirt v. Oklahoma that the State of Oklahoma did not have authority to prosecute an Indian person for a crime committed within the reservation of the Muscogee (Creek) Nation, which was Indian Country. 96 140 S. Ct. 2452, 2459 (2020). McGirt rested upon principles of federal Indian law recognized in Worcester v. Georgia, the third in the Marshall trilogy, a case that, not incidentally, held that the State of Georgia did not have authority to prosecute a non-Indian person for being in Indian Country without permission from the state government. 97 See id. at 2476-77(quoting Worcester v. Georgia, 31 U.S. (6 Pet.) 515, 557 (1832)). Worcester’s holding was based on the principle, which was recognized in a treaty between the Cherokee Nation and the United States, that an Indian tribe “is a distinct community occupying its own territory.” 98 Worcester, 31 U.S. (6 Pet.) at 561. The question presented in Castro-Huerta, in short, was answered by settled law.
Except, apparently, it wasn’t, because the Roberts Court held that the State of Oklahoma had authority to prosecute a non-Indian person for a crime committed within Indian Country. Justice Kavanaugh’s opinion conceded that its holding was inconsistent with the principles of federal Indian law recognized in Worcester. 99 See Castro-Huerta, 142 S. Ct. at 2493. No matter, however, because the “general notion” in Worcester was not law. 100 Id.(quoting Organized Village of Kake v. Egan 369 U.S. 60, 72 (1962)).
Justice Kavanaugh’s rhetoric trades on the notion, which we have from Erie, that law does not consist of general notions. Law must be based on positive authority. Legal rules must stem from applicable positive law. 101 In 1L private law courses, for instance, we say that the Restatements of Law are not actually law. (And then we let students cite the Restatements on their final exams on the law they’ve learned.). The problem with Swift v. Tyson 102 41 U.S. (16 Pet.) 1 (1842). was that it placed great stock in notions drawn from general law. Justice Kavanaugh would not make that mistake in Castro-Huerta.
Yet all Justice Kavanaugh did was replace one general notion with another. The notion was that “Indian [C]ountry is part of the State, not separate from the State.” 103 Castro-Huerta, 142 S. Ct. at 2493. This notion is the opposite of what Worcester held. 104 See 31 U.S. (6 Pet.) at 542, 561-62.
The first, and most telling, citation for Justice Kavanaugh’s general notion was the Tenth Amendment, which he did not quote. 105 See Castro-Huerta, 142 S. Ct. at 2493. Citing the Tenth Amendment begs the question in Castro-Huerta. The Tenth Amendment does not actually say that “Indian Country is part of the State, not separate from the State.” It says, “[t]he powers not delegated to the United States by the Constitution, nor prohibited by it to the States, are reserved to the states respectively, or to the people.” 106 .U.S. Const. amend. X.
On the question of whether non-Indian persons in Indian Country are subject to state criminal jurisdiction, Justice Kavanaugh’s opinion reads like a postmodern pastiche of quotations yanked from an hour-long Westlaw search. In six seriatim paragraphs, he presented quotations from six opinions as free-standing statements whose meaning was self-evident (or, one might think, self-evidently empty). 107 See Castro-Huerta, 142 S. Ct. at 2493-94. One of these was a dictum from a 1962 decision, to wit, “the ‘general notion drawn from Chief Justice Marshall’s opinion in Worcester v. Georgia’ ‘has yielded to closer analysis.’” 108 Id. at 2493 (quoting Organized Village of Kake v. Egan, 369 U.S. 60, 72 (1962)).
Others have shown that none of these cases supported the Roberts Court’s holding that Oklahoma could prosecute a non-Indian for a crime committed against an Indian within the Cherokee reservation. 109 Dylan R. Hedden-Nicely, The Reports of My Death Are Greatly Exaggerated: The Continued Vitality of Worcester v. Georgia, 52 Sw. L. Rev. 255, 257-58, 260-68 (2023); Gregory Ablavsky, Too Much History: Castro-Huerta and the Problem of Change in Indian Law, 2022 Sup. Ct. Rev. 293, 318-19, 326-32, 336, 341-343. Indeed, the closest the Court could come was the assertion that an 1882 case, United States v. McBratney, was “the leading case in the criminal context.” 110 Castro-Huerta, 142 S. Ct. at 2494 (citing United States v. McBratney, 104 U.S. 621, 623-624, (1882)). Even that case, however, did not hold that a state could prosecute a non-Indian person for a crime against an Indian person within Indian Country, for the simple reason that there was no Indian victim in McBratney. 111 See McBratney, 104 U.S. at 624. But, in a classic case of the-exception-becomes-the-rule, McBratney rose to the fore in Castro-Huerta.
Notwithstanding his dismissal of a “general notion” found in Worcester, Justice Kavanaugh’s opinion amounts to a simple shift in the sources to be treated as authoritative sources of general rules of decision. What happened to the original understanding? Superseded, apparently. 112 Cf. Castro-Huerta, 142 S. Ct. at 2493 (“[T]he ‘general notion drawn from Chief Justice Marshall’s opinion in Worcester v. Georgia’ ‘has yielded to closer analysis.’”) (quoting Organized Village of Kake, 369 U.S. at 72). The Constitution’s references to Indian tribes? Not mentioned, much less interpreted for what light they might shed on the Tenth Amendment’s relevance for Indian Country. Federal statutes that might be read to codify the original understanding that Indian Country is separate from the states? A few are discussed but interpreted through the lens of the general notion that Justice Kavanaugh took for granted, namely, that Indian Country is not separate from the states. 113 Id. at 2494-95.
What about the case law in Castro-Huerta? Justice Kavanaugh failed the test of lawfinding, at least if the test is the one offered by Edmund Burke and repeated by Justice Gorsuch in Loper Bright. 114 See Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244, 2277 (2024) (Gorsuch, J., concurring); see also supra note 40 and accompanying text. Justice Kavanaugh’s assemblage of quotations to cases was scattered by design and made no effort to address contradictory opinions. Most strikingly, at least for experts in the field, Justice Kavanaugh did not acknowledge that the two most important case citations in his opinion—McBratney and Organized Village of Kake—were decided in the Allotment and Assimilation Era and the Termination Era, respectively. 115 See McBratney, 104 U.S. at 623-624 (Allotment and Assimilation Era); Organized Village of Kake v. Egan, 369 U.S. 60, 72 (1962) (Termination Era). These were, to put it mildly, not “good and constitutional times” 116 .Burke, supra note 40, at 513. for the federal government’s treatment of Native people.
Justice Gorsuch, who was joined by Justices Breyer, Sotomayor, and Kagan, dissented from the majority’s methodology and its conclusions. Methodologically, Justice Gorsuch wrote, the Court’s “declaration comes as if by oracle, without any sense of the history recounted above and unattached to any colorable legal authority.” 117 Castro-Huerta, 142 S. Ct. at 2511(Gorsuch, J., dissenting). Substantively, “a more ahistorical and mistaken statement of Indian law would be hard to fathom.” 118 Id.
“From start to finish,” Justice Kavanaugh replied for the Court, Justice Gorsuch’s “dissent employs extraordinary rhetoric in articulating its deeply held policy views about what Indian law should be.” 119 Id. at 2504 (majority opinion). Not just rhetoric, which is bad enough, 120 See Aristophanes, The Clouds 92 (William Arrowsmith trans. 1962). but extraordinary rhetoric, the sort of rhetoric unbecoming of a judge who agrees that the judicial task is to find “what the law is.” 121 Castro-Huerta, 142 S. Ct. at 2504 (emphasis added). What was so extraordinary about Justice Gorsuch’s rhetoric? Well, “the dissent emphasize[d] the history of mistreatment of American Indians,” and this history, Justice Kavanaugh suggested, was of no matter for the modest science of lawfinding. 122 Id. at 2502.
Justice Gorsuch’s dissent considered an array of diverse, and sometimes contradictory, sources across a long arc of history. 123 Id. at 2505-11 (Gorsuch, J., dissenting). It began with Cherokee Nation history from before the creation of the American Republic and carried forward an historical story to the Supreme Court’s decision in Denezpi v. United States, which the Court decided “just weeks” before Castro-Huerta. 124 See id. at 2505 (Cherokee Nation); id. at 2511 (citing Denezpi v. United States, 142 S. Ct. 1838, 1843(2022)). In Justice Gorsuch’s telling, Worcester’s general principle excluding states from applying their criminal laws in Indian Country could be traced through many legal sources over the centuries. 125 Id. at 2511.
Behind this effort to harmonize history was Justice Gorsuch’s frankly expressed sense of justice. As he understood it, the United States had “exile[d]” the Cherokee Nation to Oklahoma, but it made a promise “that [the Nation] would remain forever free from interference by state authorities.” 126 Id. at 2505. In Worcester, the Court had worked to hold the United States to its promises to the Cherokee Nation. 127 Id. But in Castro-Huerta, it broke that promise. 128 See id.
Justice Gorsuch was right that the Court’s majority had made a “category error.” 129 Id. at 2511. It was not just a doctrinal error, however. Castro-Huerta was a paradigm-shift masquerading as “normal science.” 130 See Dagan, supra note 90, at 60. Justice Kavanaugh presented the holding as a simple application of the already-existing paradigm, not as the paradigm-shift it actually was. The Roberts Court’s jurisprudential failure in Castro-Huerta was to mistake, or, worse, to refuse to acknowledge, that its decision was a radical reversal from the day-to-day operation of federal Indian law.
B. You Didn’t Think That “Trust” Really Meant Trust, Did You?
Justice Kavanaugh also wrote for the Roberts Court in Arizona v. Navajo Nation in an opinion that repeated the same jurisprudential errors. 131 143 S. Ct. 1804 (2023). The case was about the rights of the Navajo Nation to water from the Colorado River. 132 Id. at 1811-12. The issue was not whether the Navajo Nation has rights in those waters. The issue instead was quantifying the water rights that the Navajo Nation reserved through its treaties with the United States. 133 See id. at 1819 (Gorsuch, J., dissenting) (describing the Navajo Nation’s complaint).
For decades in the twentieth century, the federal government litigated against western states about the Navajo water rights while successfully opposing the Navajo Nation government’s motions to intervene. 134 Id. at 1823-24 (Gorsuch, J., dissenting). In Winters v. United States, a 1908 decision involving water rights to support the Fort Belknap Indian Reservation, the Court held that the reservation of Indian lands implies the reservation of surface water rights to support the Indian reservation. 135 207 U.S. 564, 575-77 (1908). Since at least the 1960s, the Navajo Nation has been trying to get a judicial quantification of its reserved rights to the Colorado River. 136 Arizona, 143 S. Ct. at 1824 (Gorsuch, J., dissenting). Without it, there is a significant risk that plans about the Colorado River will divert Navajo water to non-Navajo people. Given the acute water crisis on the Navajo Nation, where approximately 30 percent of homes do not have running water, 137 Navajo Water Project, About the Project, https://perma.cc/VS93-3N44. this risk of loss of water rights threatens Navajo families.
Because the federal government had resisted quantification of the Navajo rights, the Nation sued under the APA for injunctive and declaratory relief to obtain an accounting of its rights to Colorado River water. The Ninth Circuit held that the Navajo Nation could plead a claim for an accounting of its rights and that the APA provided a right of action for equitable relief to order the federal government to quantify the water rights. 138 See Navajo Nation v. U.S. Dep’t of Interior, 26 F.4th 794, 803-04, 808-09, 812-14 (9th Cir. 2022).
The Roberts Court reversed and held that there was no legal basis for the relief that the Navajo Nation requested. 139 Arizona, 143 S. Ct. at 1816. Writing for the Court, Justice Kavanaugh construed the complaint as requesting “affirmative steps” from the United States to protect Navajo water rights and found no legal source of any obligation to take such steps. 140 Id. at 1810. To impose an affirmative obligation on the federal government without a basis in positive law would be to veer from the Court’s “constitutional lane.” 141 Id. at 1814.
Deference to the democratically accountable branches and positivism were the two main themes of Justice Kavanaugh’s opinion. 142 See id. at 1814. Water rights are complex and a hot button issue in the West. It follows, Justice Kavanaugh reasoned, that the political branches should decide disputes about them. 143 Id. Judicially enforceable obligations to tribal nations must exist in positive law. The general trust relationship between the United States and tribal nations is not enough to establish a legal duty that a federal court may enforce. 144 Id. at 1813. Nor are general principles of trust law enforceable against the United States when it acts as a trustee for tribes—as the Court acknowledged the United States is doing with respect to the Navajo Nation’s water rights. 145 See id. at 1813 n.1. Rather, the federal government’s judicially enforceable obligations must be found, if at all, in specific provisions of positive law.
Moreover, the federal government’s positive legal obligations must be express. That is, a tribe suing the federal government must show that the United States has “expressly accepted” an obligation in a “treaty, statute, or regulation.” 146 Id. at 1813 (quoting United States v. Jicarilla Apache Nation, 564 U.S. 162, 177 (2011)). Judges should not be implying obligations from the purposes of positive enactments or creating obligations that are not express in those enactments.
The Navajo Nation’s APA suit failed this test. 147 Id. at 1814.. The 1868 Treaty, which the Nation emphasized, did not expressly require “affirmative steps” from the United States, at least with respect to water from the Colorado River. 148 Id. at 1810. And the Court would not imply an “affirmative” obligation from the purposes of the 1868 Treaty. 149 Id. at 1814. Rather, it would defer to the political branches’ decisions about how to implement the treaties and the United States’ trust responsibility to the Navajo Nation.
Justice Kavanaugh’s opinion treated the case as one of the straightforward application of settled principles. But the rule of decision he applied cannot be traced to any specific provision of the constitution or a treaty, statute, or regulation. Justice Gorsuch pointed this out in his dissenting opinion, to no avail. 150 See id. at 1830-31 (Gorsuch, J., dissenting).
Prior to Arizona v. Navajo Nation, the Supreme Court had never held that a tribal nation suing the federal executive branch for equitable relief under the APA had to identify a specific provision of positive law that “expressly accepted” an obligation to the tribe. The APA, which played such an important role in Loper Bright but which Justice Kavanaugh never cited in his opinion for the Court, provides a right of action and waives sovereign immunity for suits for equitable relief against federal agencies. 151 See 5 U.S.C. § 702. It expressly authorizes judicial relief for agency action “unlawfully withheld.” 152 Id. § 706. Nowhere in the APA does it state that a suit for equitable relief to require an agency to act must be based upon an express requirement in a treaty, statute, or regulation. 153 See 5 U.S.C. §§ 702, 704, 706. Thus, even if one accepts Justice Kavanaugh’s framing of the Navajo Nation’s complaint—that it was demanding “affirmative steps” from the federal executive branch—the APA provided a right to sue for unlawfully withheld action.
In fact, the Navajo Nation’s complaint was a paradigmatic one in equity. All nine members of the Roberts Court agreed that the United States was a trustee of the reserved water rights of the Navajo Nation. 154 See Arizona, 143 S. Ct. at 1814; id. at 1829 (Gorsuch, J., dissenting). They also all acknowledged that the United States executive branch was making decisions about the potential future allocation of water from the Colorado River, which would impact the exercise of those rights. And, as all recognized, the federal executive branch was making those decisions about a trust asset without ever quantifying the value of the asset. 155 Id. at 1812-13 (majority op.); id at 1824 (Gorsuch, J., dissenting).
To put it plainly, that is not how a trustee is supposed to manage a trust asset. Compare, for example, the Navajo Nation’s request for a quantification of its water rights to an equitable action for accounting. Equity provides the beneficiary of a trust the right to sue for an accounting of the trust assets. The equitable action for an accounting has a centuries-long history, 156 See Joshua Getzler, Rumford Market and the Genesis of Fiduciary Obligations, in Mapping the Law: Essays in Memory of Peter Birks 577, 580 (Andrew Burrows & Alan Rodger eds. 2006). and is not a disfavored action that requires specific and express authorization in law. To the contrary, the purpose of equitable relief is to provide for a remedy when law has run out, that is, when there is no adequate remedy at law. Elsewhere, the Roberts Court has recognized this basic point of equitable practice and said that equity’s historical role is an important guide to equity today. 157 See, e.g., US Airways, Inc. v. McCutchen, 569 U.S. 88, 94-95 (2013) (citing Mertens v. Hewitt Associates, 508 U.S. 248, 256 (1993)); eBay Inc. v. MercExchange, L.L.C., 547 U.S. 388, 391 (2006).
Thus, the “expressly accepted” requirement is not to be found either in the APA, which expressly authorized the suit but which Justice Kavanaugh did not cite, or in equity, which provides for the remedy that the Navajo Nation was seeking. Here too, if Loper Bright is any guide, the Roberts Court failed the lawfinding test.
Justice Kavanaugh’s opinion cited neither the APA nor equitable decisions, which were both applicable sources of law, but instead cited snippets of text from an opinion involving Rule 501 of the Federal Rules of Evidence and a few opinions applying the Indian Tucker Act, a statute that applies only to actions seeking damages, a form of legal not equitable relief. 158 Arizona, 143 S. Ct. at 1812-14 (citing United States v. Jicarilla Apache Nation, 564 U.S. 162, 173-174, 177-178 (2011); United States v. Navajo Nation, 537 U.S. 488, 506-507 (2003); and United States v. Mitchell, 445 U.S. 535, 542, 546 (1980)). Only one of these opinions—the one applying the Federal Rule of Evidence to figure out the scope of the attorney-client privilege between the Department of Justice and the Department of the Interior—contained the “expressly accept[ed]” statement that became the rule of decision in Arizona v. Navajo Nation. 159 Jicarilla Apache Nation, 564 U.S. at 177. The other opinions interpreting the Indian Tucker Act did refer to the need for a tribe suing for damages to point to a specific provision of positive law. 160 See Navajo Nation, 537 U.S. at 506; Mitchell, 445 U.S. at 542, 546. This was because a tribe suing for damages—unlike a tribe suing under the APA—needs a waiver of sovereign immunity under the Indian Tucker Act. 161 28 U.S.C. §§ 1491, 1505. But even the Indian Tucker Act cases say that as long as the tribe has pointed to specific provisions that make the United States the trustee of a tribal resource and create rights or impose duties, the courts must specify the fiduciary obligations of the federal government by reference to general principles of trust law and may “infer[]” the “availability” of damages remedies. 162 Navajo Nation, 537 U.S. at 506; United States v. White Mountain Apache Tribe, 537 U.S. 465, 475-76 (2003); United States v. Mitchell, 463 U.S. 206, 226-27 (1983). In the end, then, Justice Kavanaugh’sbest evidence for the “expressly accepted” requirement was a single sentence applying the federal rules of evidence in an opinion about the government’s attorney-client privilege. 163 In its limitations on the Indian trust doctrine, Justice Kavanaugh’s opinion shares something with Justice Thomas’s radical critique of the trust responsibility and the Indian canons of construction. In his concurring opinion in Arizona v. Navajo Nation, Justice Thomas argued that the general principle that the United States is a trustee for Native people is not a valid source of legal rules. Arizona, 143 S. Ct. at 1817-18 (Thomas, J., concurring).
The obvious question is why a statement in an opinion about an evidentiary privilege, one that involved the express common law authority in FRE 501, should control the availability of equitable relief under the APA. Justice Kavanaugh answered that question with a footnote that cited the separation of powers. 164 Arizona, 143 S.Ct. at 1813 n.1 (majority op.). Thus, it appears, the Roberts Court found a rule of decision in a sentence of an opinion about the attorney-client privilege that, it has turned out, is based on the structure of the constitution and therefore applies whenever a tribe seeks relief, whether legal or equitable, against the federal executive branch.
In the traditional understanding of lawfinding, recall from Justice Gorsuch’s opinion in Loper Bright, judges did not find law from a single opinion. 165 Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244, 2276 (2024) (Gorsuch, J., concurring). Rather, law consisted of general principles and customary practices that had to be proven through multiple sources of evidence. Any single opinion could err in its restatement of those principles or practices. A judge setting out to find law would therefore be cautious about staking too strong a claim based on any one opinion. But that is precisely what Justice Kavanaugh did in Arizona v. Navajo Nation. His opinion staked a strong claim about the separation of powers based on a statement in a single opinion. It then applied a rule of decision based upon that claim even though it was inconsistent with multiple other pieces of evidence, including the APA’s text and general principles of equity that have been restated for centuries.
In his dissenting opinion, Justice Gorsuch again identified the paradigm-shift that the Court had accomplished by shifting the sources it would treat as authoritative. 166 Arionza, 143 S. Ct. at 1831 (Gorsuch, J. dissenting). As he saw it, the Navajo Nation’s case had to be understood within the long and tortured history of the Nation’s relations with the United States. 167 Id. at 1824 (Gorsuch, J. dissenting) (“With a view of this history, the proper outcome of today’s case follows directly.”). And again, as in Castro-Huerta, Justice Gorsuch sought to harmonize the doctrinal authorities and a felt sense of justice, hoping that, the Court’s holding notwithstanding, “some measure of justice will prevail in the end.” 168 Id. at 1833 (Gorsuch, J., dissenting).
Conclusion
Perhaps justice will indeed prevail. That is, one possible lesson of the Roberts Court’s federal Indian law jurisprudence is that the legacy of the legal realists still holds sway, all the talk of lawfinding notwithstanding. If nothing else, it is striking to hear echoes of Karl Llewelyn’s talk of “the felt demands of justice” in Justice Gorsuch’s hopes of “some measure of justice.” 169 Compare Llewellyn, supra note 16, at 38, with Arizona, 143 S. Ct. at 1833 (Gorsuch, J., dissenting).
There are other possibilities. Perhaps the lesson of the Roberts Court’s jurisprudence is that we have never been as modern as Erie told us we were. 170 Cf. Bruno Latour, We Have Never Been Modern 10-12 (Catherine Porter, trans. 1993). Or, for a third, more troubling possibility, perhaps modernity is a choice, and the right to choose is part of what the Roberts Court is celebrating with its “independent judgment.” 171 Loper Bright, 144 S. Ct. at 2262.
* Professor of Law, University of California, Berkeley School of Law. I want to thank Greg Ablavsky, Jud Campbell, Hanoch Dagan, Judge William Fletcher, José Argueta Funes, Michael McConnell, and Liz Reese for helpful comments and conversations, as well as participants at the Stanford Law Review’s 2025 symposium and at the Berkeley Law Faculty Workshop for their helpful feedback.