George Mason
Law Review

Article III Lawmaking

Micah S. Quigley
Volume 30
Issue 1


Article III courts sometimes apply unwritten federal law to resolve cases.1See Texas Indus., Inc. v. Radcliff Materials, Inc., 451 U.S. 630, 641 (1981) (overviewing the kinds of cases that get this treatment and noting, “absent some congressional authorization to formulate substantive rules of decision, federal common law exists only in such narrow areas as those concerned with the rights and obligations of the United States, interstate and international disputes implicating the conflicting rights of States or our relations with foreign nations, and admiralty cases” (footnotes omitted)). This is called “federal common-law adjudication,” and it raises all kinds of problems.2But see infra note 25; Section II.A.1 (noting common law is not the only form of unwritten law). A popular one is the proper scope of federal common law. Scholars perennially debate which cases are fit for common-law treatment in the federal system and which are not.3See sources cited infra notes 4–8 (collecting articles, most of which focus on the scope issue). This Article instead addresses the often-overlooked problem of method. It is one thing to understand which cases are within the domain of federal common law. It is another thing, and equally important, to find the right answers in those cases. Using the wrong approach yields wrong answers. It can also cause the judiciary to violate the Constitution by exercising the legislative power.

Preliminarily, consider three open questions. First, how is federal common-law adjudication, widely regarded as a form of lawmaking, compatible with the separation of powers? “Lawmaking” is just another word for “legislation”—and that is Congress’s domain.4See, e.g., Paul J. Mishkin, Comment, Some Further Last Words on Erie—The Thread, 87 Harv. L. Rev. 1682, 1683 (1974) (“That Congress may have constitutional power to make federal law displacing state substantive policy does not imply an equal range of power for federal judges. Principles related to the separation of powers impose an additional limit on the authority of federal courts to engage in lawmaking on their own (unauthorized by Congress).”); J. Harvie Wilkinson III, Our Structural Constitution, 104 Colum. L. Rev. 1687, 1694 (2004) (“The famous maxim that ‘there is no federal general common law’ means that federal courts should stick to their basic task of interpreting the Constitution and statutes and not go about creating new rules on their own.” (alteration omitted) (footnote omitted)); Jay Tidmarsh & Brian J. Murray, A Theory of Federal Common Law, 100 Nw. U. L. Rev. 585, 618 (2006) (warning that “broad permissive theories” of federal common law “threaten to undermine separation of powers and federalism constraints on judicial power”). Scholars and courts have proposed several answers, none of them entirely satisfying. Some see federal common lawmaking as, essentially, a necessary exception to the separation of powers.5See, e.g., James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 549 (1991) (Scalia, J., concurring) (“I am not so naive (nor do I think our forebears were) as to be unaware that judges in a real sense ‘make’ law.”); Martha A. Field, Sources of Law: The Scope of Federal Common Law, 99 Harv. L. Rev. 881, 884 (1986) (“[L]imits on federal common law are incoherent.”). But this approach glosses over the problem rather than acknowledging it and offering a solution. Others take a middle ground, viewing some federal common-law adjudication as permissible and some as impermissible.6See, e.g., Tidmarsh & Murray, supra note 4, at 630 (“Federal common law can be created only when there exists a legitimate concern that, if state law were created to deal with the dispute, state lawmakers likely would discriminate in a systematic and pervasive way in favor of the state or its citizens, and against outsiders whose interests are not likely to be protected in the lawmaking process.”); Bradford R. Clark, Federal Common Law: A Structural Reinterpretation, 144 U. Pa. L. Rev. 1245, 1250 (1996) (explaining that the Supreme Court itself has tried to strike such a middle ground by limiting federal common law to “several well-recognized enclaves,” and explaining that this approach does not provide “any ‘underlying rationale other than grandfathering.’” (quoting Field, supra note 5, at 912)). See generally Ernest A. Young, Preemption and Federal Common Law, 83 Notre Dame L. Rev. 1639 (2008) (among other things, discussing and elaborating on Clark’s views). This tack, like the first, often fails to take seriously the notion that federal courts have no lawmaking power, not a little lawmaking power.7Another middle-ground view is that of Caleb Nelson. See Caleb Nelson, The Legitimacy of (Some) Federal Common Law, 101 Va. L. Rev. 1 (2015). Nelson argues, or at least suggests, that federal common law that does not involve “robust lawmaking” is legitimate. See id. at 6–9, 63–64. But Nelson clarifies that he is “not trying to revive the old rhetoric that judges simply ‘discover’ the common law and play no role in ‘making’ it.” Id. at 7. That concession is problematic, for it requires Nelson to accept that federal courts have a non-zero amount of lawmaking power. And in any case, Nelson’s article is mainly a (very helpful) exploration of three “different senses in which judges might ‘make’ the common law.” Id. at 10 (capitalization altered). The Constitution, I will argue, does not allow federal judges to make federal law in any sense. That is why this Article seeks to “revive the old rhetoric” Nelson mentions.

A notable exception, however, is Bradford Clark’s view. See Clark, supra note 6, at 1274–75. Clark argues that some so-called federal common law is not really common law at all, and not in fact judge-made law, but is instead a form of inference from the constitutional structure. See id. (listing two criteria for such law). For a little more detail, see infra Part IV (briefly discussing the way this Article’s claims might interact with Clark’s theory).
Still others have taken a hard line, reasoning as follows: the federal courts may never make law; federal common-law adjudication is lawmaking; federal common-law adjudication is therefore impermissible.8See Martin H. Redish, Federal Common Law, Political Legitimacy, and the Interpretive Process: An “Institutionalist” Perspective, 83 Nw. U. L. Rev. 761, 765–67 (1989). That approach is internally consistent but perhaps so radical as to be unappetizing.

Second, consider scholarship on finding unwritten law. Professor Stephen E. Sachs recently argued that, modern assumptions notwithstanding, common-law adjudication need not be a form of lawmaking.9See Stephen E. Sachs, Finding Law, 107 Calif. L. Rev. 527, 531 (2019). On his view, unwritten law might be a kind of social fact that “can be found, rather than made”—even within a purely positivist theory of law.10See id. at 529; see also Nelson, supra note 7, at 9. Nelson suggests a similar view, but with a somewhat narrower notion of customary law. Nelson seems to conceive of unwritten law as customary mainly to the extent the unwritten law directly borrows extra-legalcustoms. See Nelson, supra note 7, at 10 (“From at least the seventeenth century on, though, many authors associated the common law with customs followed by people in the real world.”). But see id. at 9 (suggesting that “traditional principles of common law, or the collective thrust of precedents from across the fifty states,” also could count as “sources outside the federal judiciary” upon which unwritten law might be “firmly grounded”). Contrast that with Sachs’s view, which says unwritten law can be robustly customary even if the customs in question are legal customs not derived from extra-legal sources. See Sachs, supra note 9, at 535 (discussing the rule against perpetuities); see alsoinfra note 20 and accompanying text; infra Part II (discussing in detail the nature of customary law and aligning more with Sachs’s view than with Nelson’s). Scholars of legal history, for their part, broadly agree that common lawyers both before and after the American Founding conceived of courts as finders rather than makers of the common law.11See infra Sections I.A–B. Even so, Sachs argues only that lawfinding is possible and stops short of claiming it is a feature of the American legal system.12Sachs, supra note 9, at 530 (“Whether any legal systems actually work this way is an empirical question, which this Article doesn’t address; the claim here is that it’s plausiblefor a legal system so to arrange things.”); see also Nelson, supra note 7, at 7; cf. Richard A. Epstein, The Private Law Connections to Public Nuisance Law: Some Realism About Today’s Intellectual Nominalism, 17 J.L. Econ. & Pol’y 282 (2022) (arguing that, despite much hand-wringing about the inherent limitations of natural language, common-law terms like “nuisance” can have real and objectively ascertainable content). That leads to the second unanswered question: Is finding law, more than just being conceptually possible or historically interesting, actually a part of our legal system? And if not, should it be?

Third, does the original meaning of Article III have anything to say about the proper method for federal common-law adjudication? Article III of the Constitution grants only the “judicial Power of the United States” to the federal courts—not the executive or legislative powers.13U.S. Const. art. III, §§ 1–2. This axiom is presumably what spurs many thinkers to question the legitimacy of federal common law in the first place. Yet originalist scholarship on Article III’s relationship to federal common law has been hard to come by.14Scholarship on the original meaning of “judicial Power” in particular often aims to justify doctrines like stare decisis and judicial review. See Randy E. Barnett, The Original Meaning of the Judicial Power, 12 Sup. Ct. Econ. Rev. 115 (2004) (arguing “judicial Power” includes a grant of the power of judicial review); John O. McGinnis & Michael B. Rappaport, Reconciling Originalism and Precedent, 103 Nw. U. L. Rev. 803 (2009) (arguing stare decisis is included in the original meaning of “judicial Power”). Stewart Jay has also discussed the issue. See Stewart Jay, Origins of Federal Common Law: Part Two, 133 U. Pa. L. Rev. 1231, 1254–62 (1985) [hereinafter Jay, Part Two]. Jay’s analysis is thorough, but it focuses almost exclusively on the scope problem, to the exclusion of method.

This Article advances one thesis that can help answer all three questions. It argues that, on originalism, federal courts have no power to make common law. They can only find it.15As is obvious throughout, this Article (specifically Part I) aims to do originalism rather than defend it. It therefore assumes originalism is true. I would still like to acknowledge Jonathan Gienapp’s recent piece because, as far as I know, it is the first novel attack on originalism in some time. See Jonathan Gienapp, Written Constitutionalism, Past and Present, 39 Law & Hist. Rev. 321 (2021). Roughly, Gienapp argues that the original Constitution extended past the text and that most originalism unthinkingly assumes otherwise. See id. at 321–22. Instead of a purely textual and positive Constitution, says Gienapp, the original Constitution included fundamental rules of written, customary, and natural law. See id. at 339–43.

Gienapp does acknowledge the obvious fact that the Constitution is at least partly a written thing. Id. at 342. And more specifically, he admits that “[t]here was no existing legal referent in customary or fundamental law that could specify the composition of the various branches of government, such as the number of executives or senators, the federal ratio of representation, or the number of legislative bodies. These kinds of rules, and many more, needed to be positively stipulated.” Id. The separation of powers was one of America’s foremost constitutional innovations, so it seems like the “kind[] of rule” that would have “needed to be positively stipulated.” Id. Thus, because this Article focuses on the separation of powers, its thesis may be compatible with Gienapp’s. Even if not, I wonder whether the state of pre- and post-Founding common-law practice suggests even the extratextual Constitution (if there is one) bars judicial lawmaking in common-law cases. But that is a question for another day.
This is because Article III grants federal courts the “judicial Power,” a phrase whose original meaning did not include lawmaking power of any sort.16See Stewart Jay, Origins of Federal Common Law: Part One, 133 U. Pa. L. Rev. 1003, 1007 (1985) [hereinafter Jay, Part One]. Instead, Article III simply empowered courts to do what common-law courts had always done: find law, written or unwritten, and apply it.17I take no stance on how broad the federal judiciary’s common-law power is; I ask only whether it includes lawmaking power. Those interested in Early Republic arguments about the common law’s scope might consider Jay, Part One, supra note 16; Jay, Part Two, supra note 14 at 1254–71. This was a controversial topic at the time. See Jay, Part One, supra note 16, at 1012.

That thesis answers the first question by taking literally the notion that federal courts are not lawmaking entities. At the same time, it recognizes that lawfinding is entirely consistent with the judicial function—there is no need to axe all federal common law to restore the separation of powers. It answers the second question by taking Sachs’s and Professor Caleb Nelson’s scholarship to its natural, highly traditional conclusion—even if that conclusion seems unintuitive at first glance. Finding law is not just possible; it is not just historically pedigreed; it is constitutionally required in federal common-law cases. And it answers the third question by carefully assessing what the Constitution, as originally understood, empowers federal courts to do by granting them the “judicial Power.”

Part I argues for the thesis. The traditional view, long dominant among English common lawyers, saw courts as “lawfinders” and not lawmakers.18See generally Sachs, supra note 9. Founding-era Americans adopted that view, and they conceived of lawmaking as entirely outside the judicial realm.19See Frederick Schauer, Do Cases Make Bad Law?, 73 U. Chi. L. Rev. 883, 885 (2006). Thus, there is strong reason to believe Article III’s grant of “judicial Power,” as originally understood, excluded the power to make law.

Part II develops a concrete and practical method for finding law. The central idea is that, at heart, common-law rules are legal customs.20Legal customs should not be confused with extra-legal customs adopted as the rule of decision in a case. E.g., Ghen v. Rich, 8 F. 159, 162 (D. Mass. 1881). An extra-legal custom—a merely social custom— is not a “legal custom” in the sense I use the term here because it is not binding as a matter of unwritten positive law. Indeed, even after a court adopts such a custom as the rule of decision in a common-law case, the rule nonetheless becomes a part of the common law—and thus a “legal custom” in the robust sense—only if it also bears the characteristics of customary law. See infra Part II. Legal customs bear certain characteristics long ago identified by English thinkers: they are continuous with the existing body of customary law, they are received and approved by the people they govern, and they are accommodated to the extra-legal customs of those same people.21See Jay, Part One, supra note 16, at 1055. Courts can find law by finding rules that bear those characteristics. And when a court refuses to do so, that is a good sign the court is (wittingly or not) playing the legislator.22This indication is subject to a natural-law-based caveat, discussed in detail in Section II.A.1. The end of Part II uses this method to distinguish between judge-made and judge-found law in real cases.

Part III, using habeas corpus as an example, discusses recent Supreme Court cases that touch on judicial lawmaking. The Court and its justices (regardless of ideology) are often conceptually unclear about judge-made law. Worse, entire opinions sometimes rest on the assumption that the Court can make common law.23See Edwards v. Vannoy, 141 S. Ct. 1547, 1563 (2021); see also infra Section III.A. That assumption is incorrect and incompatible with the Constitution. This Article explains why. And by looking to the past for guidance, it may help provide a better path forward.

I.     Article III Grants No Lawmaking Power

This Part has three sections. The first outlines the traditional English understanding of the common law. The second explains that view’s migration to America and its survival here until well past the Constitution’s ratification—the critical moment for purposes of original meaning. The third section plugs the history into the Constitution’s text, showing that the original meaning of the phrase “judicial Power” had nothing to do with lawmaking and everything to do with lawfinding. Put differently, an ordinary reader in 1788 would read the phrase as conferring the power “to say what the law is,” even in common-law cases.24Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803). And because the federal courts have the “judicial Power” and no other, federal courts (on originalism) have no power to make common law.25The American system has two main modes of unwritten law: common law and equity. Ever since the merger of the two procedures in 1938, the Federal Rules of Civil Procedure have provided for “one form of action—the civil action.” Fed. R. Civ. P. 2. Thus, the federal courts routinely adjudicate issues of both common law and equity in the very same lawsuit. Yet the two are distinct. A court deciding whether to grant a temporary restraining order, for example, applies equitable, not common-law, principles. See, e.g., Winter v. Nat. Res. Def. Council, Inc., 555 U.S. 7, 24 (2008) (applying equitable principles in the preliminary injunction context). Conversely, equity plays no role when a court decides whether and how to apply a contributory negligence scheme. See, e.g., Alvis v. Ribar, 421 N.E.2d 886, 896–97 (Ill. 1981) (resolving a contributory negligence case by way of common-law rules).

This Article deals only with customary forms of law, like common law and its close cousin admiralty. It does not deal with equity, whose status as customary law is less obvious.

That is not to say Article III really grants courts lawmaking power in the realm of equity. Indeed, it is hard to see how such a power could exist if “judicial Power” includes no power to make law at all, as the evidence in this Article suggests. See infra Section I.B.1 (laying out evidence of a Founding-era consensus that all lawmaking is out of bounds for the judiciary). A limited scope, however, keeps things from getting too unwieldy. Constructing arguments about both common law and equity would require two historical inquiries, two parallel textual arguments, and two separate methods of distinguishing judge-made law from judge-found law. Focusing on common law avoids the mess. This narrowed scope does not allow evasion either: even if federal courts have power to make the law of equity, they certainly lack the power to transmute issues of common law into issues of equity at will. Thus, courts and commentators cannot dodge the Constitution’s strictures by resorting to equity. See infra Section III.B (applying that idea).

A.     Pre-Founding English Views

 1.     The Old View: Common Law as Found Law

Pre-Founding English common lawyers held that the common law was found and not made.26An intrepid reader might already be wondering: what concrete, practical difference is there between finding and making common law—now or in the past? In answer, I direct the curious to infra Section II.A, which explains a historically rooted lawfinding method and its practical import. For an early-Republic example of a court recognizing the difference between lawfinding and lawmaking, consider Livingston v. Jefferson, 15 F. Cas. 660, 663 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,411), discussed infra Section II.B.1. For a later example, consider The Harrisburg, 119 U.S. 199, 213 (1886), discussed infra Section II.B.4. They may have been wrong, but they really did believe it.27For scholarly discussions not detailed here, see Allan Beever, The Declaratory Theory of Law, 33 Oxford J. Legal Stud. 421 (2013) (discussing the ways in which equity could modify the common law by appealing to the natural law, also without any judicial lawmaking); Schauer, supra note 19, at 885–86 (“Common law decisionmaking was widely understood prior to the twentieth century as the process of discovering the rules and principles immanent in the existing law . . . . The common law judge was engaged, in the eyes of many, in locating and articulating a preexisting principle, no less preexisting for never before having been formally articulated. The approach was thus one of principle-finding rather than principle-making—of discovery rather than creation. Making law was rarely thought to be part of the process.”). See generally Philip Hamburger, Law and Judicial Duty (2008). As a general matter, examples are easy to come by. In a 1609 case at common law, Lord Coke explained that “neither have Judges power to judge according to that which they think to be fit, but that which out of the laws they know to be right and consonant to law.”28Calvin’s Case (1608) 77 Eng. Rep. 377, 409; 7 Co. Rep. 1 b, 27 a. As Beever explains, this quotation was part of Coke’s response to an argument that his judgment constituted judicial lawmaking. Beever, supra note 27, at 428. Coke’s response was to deny that he was engaging in such lawmaking—all while accepting that, if he were making law, he would be doing something illegitimate. See id. A while later, Sir Matthew Hale described the common law as a kind of customary law, accommodated to the English way of life and continually “received and approved” by the English people.29See Gerald J. Postema, Philosophy of the Common Law, in The Oxford Handbook of Jurisprudence and Philosophy of Law 588, 590–92 (Jules L. Coleman et al. eds., 2004) [hereinafter Postema, Philosophy of the Common Law] (discussing this view); see also Alan Cromartie, Introduction to Thomas Hobbes, A Dialogue Between a Philosopher and a Student, of the Common Laws of England, in 11 The Clarendon Edition of the Works of Thomas Hobbes, at xxviii–xxx (Alan Cromartie & Quentin Skinner eds., 2005) (ascribing a very similar view to the sixteenth- and seventeenth-century figures Edmund Plowden (a reporter), Sir John Doddridge (a common-law judge), and Sir Henry Finch (a common lawyer)). And Sir William Blackstone, shortly before the American Revolution, likewise saw the common law as a specific kind of “custom[ ] or maxim[ ],” discovered mainly by judges.301 William Blackstone, Commentaries, *69. Those judges were “not delegated to pronounce a new law, but to maintain and expound the old one.”31Id. Blackstone also echoed Hale by resting the common law’s authority “entirely upon general reception and usage” and not on the power of the sovereign.32Id. at *68.

More specifically, consider seventeenth-century English views on the nature of the common law. That timeframe is especially relevant for American purposes because “it was [the] common law practice in its seventeenth-century form . . . that spread with English colonialism to the New World and beyond.”33Postema, Philosophy of the Common Law, supra note 29, at 589; see also sources cited infra note 129. By that time, Professor Gerald J. Postema explains, the English conception of judge-found law revolved around custom and reason in equal measure.34See 590 (calling custom and reason “the twin foci of this conception of law”). For a slightly different view, placing more emphasis on the natural law than on artificial reason, see Cromartie, supra note 29, at xxx (“An influential line of thought thus treated the unwritten law as a blend of natural law and positive custom—and one whose evolution continued to be shaped by an appeal to natural principles. It would be most misleading to imply that any of these writers were expressing an established orthodoxy; it is, however, helpful for present purposes, to see them as the middle of a spectrum.”); see also Holden Tanner, How to Recover Conservative Judging, Anchoring Truths(Aug. 24, 2021), (“[Natural] [r]eason and tradition thus ground the common law in pre-existing legal obligations.”). For an explanation of why the difference is not important for present purposes, see infra note 41; see also infra Section II.A.

Custom first. “All [seventeenth-century English] general discussions of the common law started with the claim that common law was common custom of the whole realm.”35Postema, Philosophy of the Common Law, supra note 29, at 590 (first citing Thomas Hedley, Speech in Parliament on Royal Impositions, in 2 Proceedings in Parliament 1610 at 175 (Elizabeth Read Foster ed., 1966); then citing Matthew Hale, The History of the Common Law of England 17 (Charles M. Gray ed., 1971); and then citing 1 Blackstone, supra note 30, at *68 (beginning his discussion of the “grounds or foundations” of the common law in this way)); see also Postema, Philosophy of the Common Law, supra note 29, at 592 (qualifying this statement slightly). For a deeper synthesis of the classical view of common law as customary law, see Gerald J. Postema, Bentham and the Common Law Tradition 3–13 (2d ed. 2019) [hereinafter Postema, Bentham]. Understandings of “custom” varied somewhat. Some enthusiastically held that the common law’s substance had remained largely “unchanged since Roman times.”36Postema, Philosophy of the Common Law, supra note 29, at 591 (describing Lord Coke’s view). Hale’s more nuanced and mainstream view, discussed in depth later in this Article, was that the common law could be robustly customary even while evolving over time.37See id. at 591–92. This view called the common law “customary” not in the sense that it remained substantively static for centuries, but rather in the sense that it was stable, fitted to the English way of life, and accepted throughout the land.38See id. at 591 (describing Sir Matthew Hale’s view and calling it “a more moderate and common view” than Lord Coke’s view); see also Postema, Bentham, supra note 35, at 6 (“[T]he dominant view is that of Hale.”); David J. Bederman, Custom as a Source of Law 31 (2010) (describing Blackstone’s view of custom as highly influential in America and noting that (as with Hale), “[f]or Blackstone, ‘general usage’ and ‘universal reception’ were the hallmarks of the English common law as a customary regime”).

As for reason, “[t]he mantra of classical common lawyers was: ‘the common law is no other than common reason.’”39Postema, Philosophy of the Common Law, supra note 29 at 592; see also Edward Coke, First Part of the Institutes of the Laws of England at 97b (Garland ed., 1979) (1628)). But see Cromartie, supra note 29, at xxvii–xxxii (discussing reason and describing Coke’s view as “extreme” because of its heavy reliance on reason). So perhaps it is not quite right to say, as Postema does, that custom and reason were the “twin foci” of the common law. Partly for that reason, this Article stresses the customary aspect of common law more than the reason aspect. Common lawyers, according to Postema, appealed directly to the natural law only rarely.40See Postema, Philosophy of the Common Law, supra note 29, at 592–93 (“Natural law was too abstract and theoretical for their pragmatic, concretely focused minds; it was too often silent, or contested, or simply out of touch with concrete human affairs.”). For a helpful discussion of the ways in which, at least as of the late Middle Ages, English common lawyers relied on “reason”—as well as discussion of that word’s content and its relation to the natural law—see Norman Doe, Fundamental Authority in Late Medieval English Law 108–31 (J.H. Baker ed. 1990). The common law’s “reason” was not primarily based in natural law.41But see R.H. Helmholz, Natural Law in Court: A History of Legal Theory in Practice 96–100 (2015) (marshaling authorities to the effect that English common law was indeed seen as rooted in and even incorporating the natural law, though without contradicting Postema’s claim that common lawyers directlyappealed to the natural law only rarely); Cromartie, supra note 29, at xxx. For helpful discussion of the role of the natural law in the High Middle Ages (1100–1350), see J.M. Kelly, A Short History of Western Legal Theory 134–46 (1992) (describing the dominance of natural law theory as a backdrop for positive law). For discussion of the role of natural law in the seventeenth century, see id. at 222–27 (describing the continuing vitality of natural law theory, as well as objections to natural law theory and its (partial) relabeling as “reason” rather than “natural law”). For discussion of the same topic in the eighteenth century, see id. at 258–62 (describing many references to the natural law—including those by the likes of Blackstone—as “purely perfunctory”). Regardless of the extent to which the common law incorporated natural law, nobody seems to suggest common lawyers saw the natural law as judge-made. So if anything, perhaps the view expressed in this Article gives judges too little leeway to turn to the natural law. Even so, natural lawyering is not lawmaking. See discussion infra Section II.A.1. It was instead “an artificial perfection of reason gotten by long study, observation, and experience, and not every man’s natural reason.”42Postema, Philosophy of the Common Law, supra note 29, at 593 (emphasis added) (quoting Coke, supra note 39, at 97b). As Professor Michael W. McConnell has put it, artificial reason was “based on a deep, intuitive, almost aesthetic, sense of the way in which the new case ‘fits’ into the rich body of the law. This is an undertaking that can be performed only by one specially trained and thoroughly versed in the common-law tradition.”43Michael W. McConnell, Tradition and Constitutionalism Before the Constitution, 1998 U. Ill. L. Rev. 173, 181 (1998); see also id. at 181–83 (discussing artificial reason further and explaining Coke’s use of the concept in an actual—and quite heated—dispute with King James I). As a whole, then, common-law judging was not seen as a mode of lawmaking but rather a dual exercise in custom and (artificial) rationality.44This was true even when a common-law court was confronted with a truly novel situation. These situations were the closest the common law came to allowing judicial lawmaking. As Postema puts the point, “[c]ommon [l]aw theory empower[ed] judges to fashion solutions to novel problems and so empower[ed] them, in effect, to fashion new legal rules, albeit subject to the constraint that rules to cover new situations must be fashioned out of the materials of the existing law.” Postema, Bentham, supra note 35, at 11. Postema also quotes Calebresi on this point:

The most powerful engine of change in the common law was, strangely enough, the great ‘principle’ that like cases should be treated alike. Courts acting on that principle could change the law, indeed make law, without arrogating to themselves undue power because they always seemed to apply past precedents or principles in new ways to situations made new by the world around them.

Id. (quoting Guido Calabresi, A Common Law for the Age of Statutes 13 (1982)). Postema’s approving quote of Calebresi, and in particular Calebresi’s assertion that common-law courts could sometimes “make law,” offers little support for the view that common-law theory allowed judicial lawmaking. This is for four reasons. First, Postema himself nowhere else asserts that common-law courts could make law. Indeed, as I have discussed throughout this Section, he often says exactly the opposite. See, e.g., id. (explaining, immediately after quoting Calebresi, that “[i]t is the case, not the judge, that extends the law.”); id. at 9 (explaining that common law theory allowed no judicial lawmaking). Second, Postema immediately goes on to clarify that the common lawyers themselves viewed this process as one of lawfinding—sometimes with the aid of legal fictions—rather than lawmaking. Id. at 11–13. And the common lawyers’ view is self-evidently more relevant than Calabresi’s later gloss on that view. Third and related, Postema cites no historical examples to support Calabresi’s claim that judges could make law. Fourth and perhaps most important, Calebresi’s claim is limited to the narrow context of entirely novel situations—and even there, judges’ only option was to apply existing rules to the new situation as best they could.

Not every English thinker of yesteryear agreed that the common law was a body of custom, discovered by reason rather than made. Thomas Hobbes and Jeremy Bentham are notable examples.45It is not quite clear when Hobbes wrote his Dialogue expositing this idea, but it was likely sometime between 1661 and the mid-1670s. Cromartie, supranote 29, at xiv. Bentham’s Of Laws in General was written in the 1780s but was not published until 1945. See H.L.A. Hart, Bentham’s Of Laws in General, in Essays on Bentham: Studies in Jurisprudence and Political Theory 105, 107 (1982). My framing of Hobbes’s and Bentham’s views draws heavily on Postema’s account of them. Postema, Philosophy of the Common Law, supra note 29, at 616; Postema, Bentham, supra note 35, at 280–83. I am especially indebted to Postema on this score. These two subscribed to early versions of the command theory of law, later popularized by the likes of John Austin.46See Gerald J. Postema, Law as Command: The Model of Command in Modern Jurisprudence, 11 Phil. Issues 470, 470–71 (2001). That theory says all law is the command of a sovereign.47See John Austin, The Province of Jurisprudence Determined 18–19 (London, John Murray 1832) (going on to give a fuller command-theory definition of law). But see Postema, Bentham, supra note 35, at xiii (arguing, in a vein not relevant for present purposes, that Bentham did not really adopt this view). Taking that theory to its extreme, Hobbes denied that the common law was true law at all; instead, he called it “[p]hilosophy” and a “disputable Art[].”48Thomas Hobbes, A Dialogue Between a Philosopher and a Student, of the Common Laws of England, reprinted in 11 The Clarendon Edition of the Works of Thomas Hobbes, supra note 29, at 29. No sovereign decree, no law. Bentham viewed the very idea of a common law handed down by no sovereign as “a thing merely imaginary.”49Jeremy Bentham, A Comment on the Commentaries and a Fragment on Government 119 (J.H. Burns & H.L.A. Hart eds., 1977). And Professor John Chipman Gray quotes Bishop Benjamin Hoadly as hinting at a similar view in 1717: “Whoever hath an absolute authority to interpret any written or spoken laws, it is he who is truly the Law-giver to all intents and purposes, and not the person who first wrote or spoke them.”50John Chipman Gray, The Nature and Sources of the Law 102 (Roland Gray ed., 2d ed. 1972).

These critics, however, were clear outliers. Professor Alan Cromartie, in his introduction to one of Hobbes’s works, describes the legal backdrop in Hobbes’s day this way: “[F]ew if any [common lawyers] would have disagreed that their unwritten principles were customary and rational in nature: customary in having existed time out of mind; rational in that the wisdom of professional tradition had adapted them to English circumstances.”51Cromartie, supra note 29, at xxvii. So it was precisely the dominance of Hobbes’s target that made his criticisms so famous. And later, Bentham likewise played the role of external critic—hurling himself against the longstanding, majority-endorsed edifice of the common law.52See Postema, Bentham, supra note 35, at 3–37 (laying out the “Elements of Classical Common Law Theory” as the target of Bentham’s attacks); id. at vii (“Bentham’s most important jurisprudential work has lain buried under a great rubble. In this rubble we find the ruins of much of the practice and ideology of the Common Law system he so mercilessly and effectively attacked.”). To be sure, Hobbes and Bentham instigated and prosecuted a war against classical common-law theory. Eventually, of course, their side seems to have won (for now). But in their own day, they were in the minority by far.

Also noteworthy is that Blackstone and Coke themselves were sympathetic to something like the command theory.53See 1 Blackstone, supra note 30, at *44 (defining municipal law as “a rule of civil conduct prescribed by the supreme power in a state, commanding what is right and prohibiting what is wrong”); John Underwood Lewis, Blackstone’s Definition of Law and Doctrine of Legal Obligation as a Link Between Early Modern and Contemporary Theories of Law, 3 Irish Jurist 336, 338 (1968) (explaining that this was Blackstone’s blanket definition of law and noting “[t]hat [Blackstone] means his readers to understand that law is a command, an imperative, is plain even at first glance”); see also id. at 344–47 (delving into exactly what Blackstone meant by this); Kelly, supra note 41, at 237 (ascribing a similar view to Coke). That is instructive, for it suggests mainstream common lawyers considered the command theory compatible with common-law-as-custom.54See Kelly, supra note 41, at 289–90 (“We have seen virtually throughout the history of Western legal thought that, when law is defined, it receives a definition in which the imperative element, in other words the element of command, is central.” (emphasis added)). Hobbes and Bentham were unique, then, not because they favored the command theory but because they took the theory to its extremity. Only that extreme form of the command theory contradicted the majority view of the common law. And in any case, as discussed below, the colonists took Blackstone and company—not Bentham and company—with them when they crossed the Atlantic.

2.     Separation of Powers in England

A word about the separation of powers is in order. This Part ultimately argues that the original meaning of Article III’s “judicial Power” phrase included no lawmaking power—nobody at the time thought judges could make law. It concludes that judges violate the separation of powers by trying to do so. But the English system really had no separation-of-powers doctrine. So, if the English system mixed powers in a way that gave judges lawmaking power, then maybe “judicial Power” just meant “the power judges have always had, which includes the power to make law in some situations.” This objection is worth heading off at the outset.

The first important point is that the English system sometimes vested the legislative and judicial powers in the same actors.55This is not the same as “mixing” the legislative and judicial powers. That is because the two powers are conceptually and metaphysically distinct things. Even when vested in the same person or entity, they remain so. The second is that this admixture seems not to have given judges lawmaking power. James Madison’s description of the English system of government, in Federalist 47, is helpful:

On the slightest view of the British constitution, we must perceive, that the legislative, executive, and judiciary departments, are by no means totally separate and distinct from each other. The executive magistrate forms an integral part of the legislative authority. He alone has the prerogative of making treaties with foreign sovereigns, which, when made, have, under certain limitations, the force of legislative acts. All the members of the judiciary department are appointed by him; can be removed by him on the address of the two houses of parliament, and form, when he pleases to consult them, one of his constitutional councils. One branch of the legislative department, forms also a great constitutional council to the executive chief; as, on another hand, it is the sole depository of judicial power in cases of impeachment, and is invested with the supreme appellate jurisdiction in all other cases. The judges again are so far connected with the legislative department, as often to attend and participate in its deliberations, though not admitted to a legislative vote.56The Federalist No. 47, at 250 (James Madison) (George W. Carey & James McClellan eds., 2001) (emphasis added).

The final sentence is the most pointed: one might read it to suggest that, aside from all the other mixtures of power, judges did indeed take on some legislative functions in the English system.

But the fact that English judges sometimes advised the legislature does not entail that those judges, much less courts, ever made law. To the contrary, as Madison himself said, “[t]he judges can exercise no executive prerogative, though they are shoots from the executive stock; nor any legislative function, though they may be advised with by the legislative councils.”57Id. at 251 (emphasis added).

3.     A Brief Example—Posthumous Children

Before turning to Founding-era American views, here is an example of the seventeenth-century English approach working in practice. The following Section draws heavily on a discussion in C’Zar Bernstein’s recent article, Fetal Personhood and the Original Meanings of “Person.58C’Zar Bernstein, Fetal Personhood and the Original Meanings of “Person, 26 Tex. Rev. L. & Pol. 485, 537–50 (2022).

The English common law needed a way to apply inheritance law to “posthumous children,” i.e., children whose fathers died before they were born.59See id. at 539 n.285. The general rule was that an interest in an estate must vest instantly at the moment of the testator’s death.60See id. at 539. Further, an inherited interest could vest in a child only if the child had already been born by the moment of his father’s death.61See id. That meant unborn children were out of luck: no instantaneous vesting was possible, so no vesting would happen at all.62See id.; Millar v. Turner (1748) 27 Eng. Rep. 907, 907–08; 1 Ves. Sen. 85, 86 (explaining that this had once been the rule); Nurse v. Yerworth (1674) 36 Eng. Rep. 993, 994; 3 Swans. 608, 611 (“[S]eeing the contingency did not happen at the time of the testator’s death, at which time the will ought to take effect, ergo for want of an heir of the body then in being, the remainder in fee to the Defendant vested presently by purchase, which no after-born son could devest.”).

But in the 1694 case Reeve v. Long,63(1694) 87 Eng. Rep. 395; 4 Mod. 282. the House of Lords—acting as a court of last resort—deviated from that original rule. To simplify slightly, the Lords characterized the will in question as an “executory devise” rather than a “contingent remainder” in a self-proclaimed effort not “to disinherit an heir upon a nicety, against the plain intention of the testator.”64Id. at 395; 4 Mod. at 282–83 (emphases omitted). Or as a subsequent case put it, the Lords refused to apply the then-existing rule because they “thought [the rule] hard.”65Millar, 27 Eng. Rep. at 907–08; 1 Ves. Sen. at 86. Bernstein explains that this was not a full-on abolition of the old rule so much as the creation of a new exception.66Bernstein, supra note 58, at 539–40. It was nevertheless a significant change.67See id. at 540–41.

If English common law were judge made, we should expect something like the following. Because the House of Lords was sitting as an unreviewable court of last resort, it was free to make whatever common-law rule it wished. Just as a legislature cannot “get statutory law wrong” when it passes a statute (because the statute just is statutory law), the Lords could not “get the common law wrong” when they decided Reeve (because the holding of a court of last resort just is common law). So Reeve necessarily would have been a correct statement of the common law. The Lords would not be “final because [they were] infallible,” but rather “infallible . . . because [they were] final.”68Brown v. Allen, 344 U.S. 443, 540 (1953) (Jackson, J., concurring).

That is not how things went. Rather than accepting the House of Lords’ pronouncement as an alteration of the common law, “[a]ll the Judges . . . were much dissatisfied with this reversal, and did not change their opinions” about what the common law said on this issue.69Bernstein, supra note 58, at 539 (quoting Reeve, 87 Eng. Rep. at 397; 4 Mod. at 285 (reporter notes)). Gray later characterized the Reeve decision as “misunderstanding the existing law.”70Gray, supra note 50, at 39 n.2 (“The House of Lords . . . misunderstanding the existing law, and to the great disgust of the Judges, allowed a child who was begotten but not born at the end of a life estate to take . . . .”). It seems to have taken a 1698 parliamentary enactment to effect the Lords’ desired change.71An Act to enable Posthumus Children to take Estates as if borne in their Fathers Life time 1698, 10 Will. 3 c. 22 (Eng.); see also Reeve, 87 Eng. Rep. at 397; 4 Mod. at 285 (reporter notes) (“All the Judges, however, were much dissatisfied with this reversal, and did not change their opinions . . . And therefore to take such cases out of the old law, Fearn. C. R. 236, the statute of 10 & 11 Will. 3 c. 16, was made . . . .” (emphasis added)); 4 Matthew Bacon, A New Abridgment of the Law 311–12 (1759) (describing the old rule and Reeves, then explaining that “all the Judges were much dissatisfied with [Reeves] and did not change their Opinions, but blamed the Judge who permitted it to be found specially where the Law was so certain and clear”); id. (going on to explain that Parliament’s “Act was made by Reason of this Case, and of the Strictness of the Law herein,” which implies that the “Strict[]” pre-Reeves rule remained the “Law” until Parliament changed it by Act).

This illustrates that the English were not just spouting platitudes when they claimed courts were lawfinders. They understood that the House of Lords’ court of last resort status prevented other courts from reviewing its judgments, even the erroneous ones. But even so, the House of Lords was tasked with finding and applying the common law just like any other court. If it arrived at what would otherwise be the wrong answer, the answer did not become right ipse dixit.

B.     Views During the Founding Era

 1.     The Endurance of the English View

The English understanding of the common law is relevant, but only secondarily so. The key question is this: Did the American Founding generation agree with their English forebears that courts were not lawmaking bodies? If there was a consensus in 1788 America that judges could make common law, there would be every reason to think the Constitution’s grant of “judicial Power” included some lawmaking power.

The Founding generation decidedly endorsed Hale and Blackstone’s view. McConnell states the point succinctly: “Hale’s conception of the common law was adopted as orthodoxy by American lawyers of the founding period.”72McConnell, supra note 43, at 188; see also id. at 186–89 (elaborating, including helpful 1790–1792 quotations from James Wilson); id. at 189 (specifically characterizing the Founding-era American orthodoxy as a rejection of Hobbes’s view). Professor Jonathan Gienapp explains that Founding-era lawyers centered the common law on both custom and reason: “custom was . . . considered the best evidence of what reason required.”73Gienapp, supranote 15, at 340; see also id. at 339–41 (elaborating). In other words, American lawyers agreed wholeheartedly with their English forebears that the common law was a body of custom that judges discovered by the use of right reason. On this view of the history, though Americans eventually gave up on finding common law, they did so long after the Founding.74See Swift v. Tyson, 41 U.S. (16 Pet.) 1, 18 (1842) (case predicated on the notion that “the principles established in the general commercial law,” a close relative or perhaps a variant of common law, are out there to be found); Anthony J. Bellia Jr. & Bradford R. Clark, General Law in Federal Court, 54 Wm. & Mary L. Rev. 655 (2013) (discussing general law). Indeed, one can even find cases after Erie Railroad Co. v. Tompkins, 304 U.S. 64 (1938), employing a method that looks suspiciously similar to the traditional lawfinding discussed infra, Part II. Consider Hinderlider v. La Plata River & Cherry Creek Ditch Co., 304 U.S. 92 (1938), a case handed down the same day as Erie, where the Court (at least arguably) applied traditional common-law principles and perhaps a dose of natural law to resolve a dispute about interstate water rights. See id. at 102 (applying a “rule of equitable apportionment” that it considered “settled” law); see also Richard A. Epstein, Federal Preemption, and Federal Common Law, in Nuisance Cases, 102 Nw. U. L. Rev. 551, 576 (2008) (putting Hinderlider in the class of cases “in which the use of the federal common law is preferable to that of the state” because uniformity and neutrality are necessary). But see Tidmarsh & Murray, supra note 4, at 596–97 (asserting that Hinderlider was a matter of judicial lawmaking).

Scholars do, however, debate the exact timing of the ideological shift toward judicial lawmaking.75In addition to the late-shift view described above and the earlier-shift view described in the following paragraphs, interested readers might consider Helmholz, supra note 41, at 142–72, discussing the role of specifically natural law in American courts. See also id. at 170 (coming to an ambivalent conclusion about “[t]he actual extent of its influence,” but suggesting the natural law did “help[] shape the law of the United States”). Specifically, Professors Morton J. Horwitz and Stewart Jay have argued the shift happened earlier than one might suspect.76See Morton J. Horwitz, The Transformation of American Law, 1780-1860, at 1–30 (Stanley N. Katz ed., 1977); Jay, Part One, supra note 16, at 1060. I will here assume, for the sake of argument, that the Horwitz–Jay view is correct. Even on that assumption, there is little or no reason to believe the Founding generation saw the common law as judge-made. All agree that lawfinding was the dominant norm until years after 1788.

Horwitz takes up the issue in The Transformation of American Law, 1780–1860.77See Horwitz, supra note 76, at 1–16; see also Gienapp, supra note 15, at 340; cf. id. at 341 (“Fundamental law [in Founding-era America] was distinct from positive law, common law, or natural law, but was also the site where aspects of each most clearly merged.”). Though Founding-generation Americans distrusted English statutes (which they conceived to be exercises of Parliament’s will), they were much more comfortable with the English common law, which they conceived as incorporating both natural law and custom.78Horwitz, supra note 76, at 5–6; see also id. at 12–13 (discussing colonial concerns with judicial discretion, but noting that these concerns applied, oddly, only in the realm of statutes rather than to common-law judging). Horwitz describes the idea as follows: “common law doctrines were derived from natural principles of justice, statutes were acts of will; common law rules were discovered, statutes were made.”79Id. at 7 (describing the view Virginia Attorney General John Randolph espoused during argument in Blackwell v. Wilkinson, Jeff. 73, 77 (Va. 1768)). Thus in 1772, a Massachusetts judge asked rhetorically whether common law courts are free to depart from the rules of the common law.80Id. at 8. His response? “No, surely, unless they take it upon them to be wiser than the Law.”81Id. (quoting Hooton v. Grout, Quincy 343, 362 (1772)). And four years after the ratification, Justice James Iredell declared for the Court “that the distinct boundaries of law and Legislation [should not] be confounded” because “that would make Courts arbitrary, and in effect makers of a new law, instead of being (as certainly they alone ought to be) expositors of an existing one.”82Chisholm v. Georgia, 2 U.S. (2 Dall.) 419, 448 (1793) (opinion of Iredell, J.) (emphases altered).

Jay’s two-part article, Origins of Federal Common Law,83Jay, Part One, supra note 16; Jay, Part Two, supra note 14. discusses and critiques Horwitz’s Transformation.84For Jay’s critiques of Horwitz, see Jay, Part One, supra note 16, at 1060 n.284, 1061 n.290, 1072 & n.339. Jay nonetheless agrees with Horwitz that—at least until 1790—the colonists thought of the common law as found and not made. Jay puts the point this way:

[H]ostility to English institutions was sufficiently intense to make it highly unpopular in many areas to contend that English law had been incorporated as a body of recognizable rules.

Notwithstanding these sentiments, Americans of this time did not generally reject that part of their English heritage of law that . . . consisted of a belief in “God-given, natural, inalienable rights, distilled from reason and justice,” and was expressed partially “in the common law of England.” Discovering the principles of this natural law required an investigation of inherited custom. . . . Cases were merely the best evidence of the principles of the common law, the same point made years later in [Swift v. Tyson].85Id. at 1056 (emphasis added) (footnotes omitted) (citing Swift v. Tyson, 41 U.S. (16 Pet.) 1, 18 (1842)).

Thus, Jay affirms that the colonists did indeed carry the conception of judges-as-finders over the Atlantic with them and into the new Republic. Though the common law was seen as rooted in natural law, it was discovered not by abstract reflection but by “investigation of inherited custom.”86Id.

That view did not fall out of favor until after the Constitution’s ratification.87The reasons for the shift are complex and not relevant here. For discussion, see Horwitz, supra note 76, at 9–30 (discussing contemporary politics and the need to square the common law with the consent of the people—both of which, in Jay’s view, may have strained the view of judges-as-law-finders). Though it is hard to pin down the timing with any precision, Horwitz’s sources indicate the shift started in the 1790s and was mostly finished by 1820.88Id. at 9, 30. For more discussion, see id. at 12 (“Although fear of judicial discretion had long been part of colonial political rhetoric, it is remarkable that before the last decade of the eighteenth century it was not associated with attacks on the common law jurisdiction of the judiciary.”). Though Horwitz does elsewhere describe the process of “dethroning the common law from the unchallenged place it had occupied in the . . . revolutionary generation” as beginning “sometime in the 1780s,” id. at 11, every one of Horwitz’s examples (aside from some of Jefferson’s complaints, from the 1780s, that Lord Mansfield’s common law decisions were too uncertain to be relied upon, see id. at 18) is from 1791 or later. Horwitz certainly does point out many significant changes in American conceptions of the law, but even Jay suggests Horwitz is wrong that the finding-law view was on its way out by 1820.89Horwitz may be sliding between the idea (quickly adopted in the Founding era) that the common law must be justified by the people’s consent and the very different idea that common law is judge-made. See Horwitz, supra note 76, at 20–23. The two are distinct. A people can consent to the common law precisely by accepting and endorsing its customs. See Gienapp, supra note 15, at 340–41 (describing how “the ‘common-law mind’ [had] so elegantly fused consent and reason in the domain of municipal law”). Elsewhere, Horwitz makes a similar slide: he observes the breadth of nineteenth-century precedents and jumps to the conclusion that “the process of common law decision making had taken on many of the qualities of legislation.” Horwitz, supra note 76, at 2. Horwitz does provide at least one example of judges who, in 1813, took themselves to be de facto legislators. Id. at 23. Jay concedes that the finding-law attitude “was under attack in [Justice James] Wilson’s time and would eventually be replaced by a view of judges as instrumental decisionmakers who were making rules, not discovering them.”90Jay, Part One, supra note 16, at 1056–57. Justice Wilson was one of the original Supreme Court justices. For historical background, see Kermit L. Hall, Introduction to 1 James Wilson, Collected Works of James Wilson, at xiii–xxviii (Kermit L. Hall & Mark David Hall eds., 2007). He goes on: “Still, the older conception of objectively discernable principles underlying judicial decisions persisted in many circles throughout the next century . . . despite its being at odds with the fundamental ideological shift occasioned by the Revolution.”91Jay, Part One, supra note 16, at 1057 (footnote omitted).

Even on the assumption that Horwitz is completely correct on this point and that every American man, woman, and child in 1820 thought the common law was judge-made, the change did not happen until after the Constitution’s ratification.92See id. At that time, judges were considered law-finders, not lawmakers.93Id. at 1057–58. And ratification is the relevant moment for most mainstream versions of originalism.94See, e.g., Lawrence B. Solum, The Fixation Thesis: The Role of Historical Fact in Original Meaning, 91 Notre Dame L. Rev. 1, 1 (2015) (“The meaning of the constitutional text is fixed when each provision is framed and ratified: this claim can be called the Fixation Thesis. This thesis is one of two core ideas of originalist constitutional theory . . . .” (emphasis omitted)). The shift was simply too late to affect the original public meaning of Article III.

 2.     Separation of Powers in America

This Section addresses the “mixing” of powers that occurred in pre-Constitution America, as well as some of the Constitution’s own exceptions to the separation of powers. Though pre-Constitution state governments sometimes vested legislative and judicial powers in the same actors, their strategy was just like that of the English: it never allowed courts to make law.95See The Federalist No. 47, at 251 (James Madison) (George W. Carey & James McClellan eds., 2001). Similarly, Article I of the Constitution gives Congress limited power to adjudicate disputes, but Article III gives no reciprocal lawmaking power to the federal courts.96See Jason Boatright, End Judicial Lawmaking, 24 Tex. Rev. L. & Pol. 355, 370 (2020). The line-crossing in both instances, then, features legislatures empowered to issue judgments. There is no hint of courts empowered to make law.97See id. at 370–71 (making a similar point on the way to applying the principle to Texas courts).

Federalist 47 gives a helpful overview of pre-Constitution separation-of-powers doctrine. It first points out that—with rare exceptions—state constitutions explicitly proclaimed that the executive, legislative, and judicial powers were to be kept separate.98The Federalist No. 47, at 252–55 (James Madison) (George W. Carey & James McClellan eds., 2001) (surveying state constitutions). But, says Madison, “notwithstanding the emphatical, and in some instances, the unqualified terms in which this axiom has been laid down, there is not a single instance in which the several departments of power have been kept absolutely separate and distinct.”99Id. at 252. Madison then surveys eleven of thirteen states, demonstrating that the government of every single state involved incomplete separation of powers.100Id. at 252–55. His point, of course, is that literally complete separation of powers is both impossible and undesirable.101See id. at 255; see also The Federalist No. 48 (James Madison) (using Virginia and Pennsylvania as examples of states where inadequate inter-branch checking led to abuses).

Yet Madison mentions not one instance where a state judiciary ended up with lawmaking power. The omission is striking, given that he catalogues almost every other kind of unorthodox power-allocation one could think of. In New Hampshire, for example, the legislature could sometimes exercise executive power: “Several of the officers of state [were] appointed by the legislature” rather than the executive.102The Federalist No. 47, at 252 (James Madison) (George W. Carey & James McClellan eds., 2001). Likewise in New Jersey, where judges were “appointed by the legislative department.”103Id. at 253. In Massachusetts, the executive could exercise a form of legislative power by way of a partial veto.104See id. at 252–53. And many states allowed individuals to serve as “members of the legislature” and “at the same time [as] justices of the peace.”105Id. at 254. This last arrangement—which allowed one individual to exercise legislative and judicial power concurrently, albeit in separate offices—might be the closest example we have to pre-Constitution judicial lawmaking. Yet despite all Madison’s instances of incomplete separation of powers in pre-Constitution state governments, he points to no instance of courts making law.

That does not quite end the inquiry, for Professor Christine A. Desan has shown that colonial-era legislative and judicial powers were more intertwined than sources like Madison may suggest.106See generally Christine A. Desan, The Constitutional Commitment to Legislative Adjudication in the Early American Tradition, 111 Harv. L. Rev. 1381 (1998). To be fair to Madison, he does anticipate Desan by mentioning that New York “[gave], nevertheless, to the executive magistrate a partial control over the legislative department; and what is more, gives a like control to the judiciary department, and even blends the executive and judiciary departments in the exercise of this control.” The Federalist No. 47, at 253 (James Madison) (George W. Carey & James McClellan eds., 2001). Desan focuses on the colony of New York, uncovering a regime that “confounded categories we now assume [and] produced political representation that included the power to adjudicate the cases of constituents and rights that could not be enforced in courts.”107Desan, supra note 106, at 1384. Desan later suggests that her findings generalize to other colonies. See 1495–503. More specifically, the New York legislature sometimes “exercised the authority to interpret public law,” and “the legislature[] as well as the courts adjudicated cases.”108Id. at 1384. And it was not until after the ratification of the federal Constitution that New Yorkers shifted away from this conception of legislator-as-adjudicator.109See id.; see also id. at 1384 n.7 (collecting sources to this effect). That being said, there is evidence that the Constitution intentionally broke from the colonial habit of investing legislatures with judicial power. For a brief but helpful overview, see Chief Justice John Roberts’s discussion, in dissent, in Patchack v. Zinke, 138 S. Ct. 897, 914–15 (2018) (Roberts, C.J., dissenting). This may reduce the relevance of Desan’s article for present purposes. This historical situation, Desan says, “belies the arresting affect” of conventional separation-of-powers doctrines by showing that “the line that separates law from politics” was not always as clean as we tend to assume today.110Desan, supranote 106, at 1386.

And of course, the Constitution’s system of checks and balances gives Congress itself a quantum of judicial power. Article I gives the House of Representatives “the sole Power of Impeachment” and the Senate “the sole Power to try all Impeachments.”111U.S. Const. art. I, § 2, cl. 5; id. art. I, § 3, cl. 6. Impeachment is a judicial process rather than a legislative one.112See The Federalist No. 65, at 337 (Alexander Hamilton) (George W. Carey & James McClellan eds., 2001) (discussing the Senate’s power to act in a “judicial character as a court for the trial of impeachments”); Hayburn’s Case, 2 U.S. (2 Dall.) 409, 413 (1792) (“[N]o judicial power of any kind appears to be vested [in Congress], but the important one relative to impeachments.”); Kilbourn v. Thompson, 103 U.S. 168, 191 (1880) (“The Senate also exercises the judicial power of trying impeachments, and the House of preferring articles of impeachment.”). Like the English system and New York’s colonial practice, then, the Constitution sometimes vests both the legislative and judicial functions in a single actor.

But just as in the English system, the line-crossing in the colonies and in the Constitution only goes one way: the legislature gets some power to wear the judicial hat and perform judicial functions. Neither Desan nor the Constitution mentions the converse situation, which would involve courts wearing the legislative hat and performing legislative functions. Thus, though the separation of powers may have been weaker in the colonies than is often thought, there is scant reason to believe common-law judges had any lawmaking power. Historical practices simply do not suggest that “judicial Power,” as used in Article III, would have been understood by contemporaries to include lawmaking power.

C.     Applying the History to the Text

Now for the text—Article III’s “judicial Power.” For clarity’s sake, here is a formalized argument for this Article’s thesis:


(1) Article III courts have only the powers Article III confers;


(2) If Article III confers a power to make common law, it does so via its grant of the “judicial Power,” as determined by that phrase’s original public meaning;


(3) Article III’s grant of the “judicial Power,” as determined by that phrase’s original public meaning, confers no power to make common law;


(4) Therefore, Article III courts have no power to make common law.113Jason Boatright argues that Texas courts have no lawmaking power. My argument has a similar logical structure to his. See Boatright, supra note 96, at 362–63. But see infra note 372 and accompanying text (explaining a point of seeming disagreement about what counts as judicial lawmaking).


As premise (1) suggests, I do not argue that Article III affirmatively bars federal courts from making law. That would presumably require showing that the phrase “judicial Power” includes such a prohibition in its original meaning.114See Stephen E. Sachs, Constitutional Backdrops, 80 Geo. Wash. L. Rev. 1813, 1863–66 (2012) (calling the “claim that stare decisis was part of the linguistic meaning of ‘judicial Power’ . . . a very aggressive claim in light of the evidence”). Instead, because ours is a government of enumerated powers, the judiciary by default lacks any power that cannot be traced to a constitutional grant.115This might be slightly underqualified. Even on some forms of originalism, it is possible for legal rules to survive the Constitution. The Constitution might even protect these rules from change. Sachs calls these kinds of pre-standing rules “constitutional backdrops.” See id. at 1815–16. But backdrops are not relevant for this Article. It is also possible, I suppose, for some other part of the Constitution (besides Article III) to empower judges. But that seems far-fetched. For example, consider the claim that federal courts have no power to declare war. It would be silly to demand proof that Article III affirmatively denies that power. It is enough that Article III does not grant it—no grant, no power, no problem. Likewise, as long as Article III does not grant the power to make law, the federal courts cannot do so.

Premise (2) is straightforward. If common lawmaking power is anywhere in Article III, it must be in the grant of “[t]he judicial Power of the United States” to “one Supreme Court, and . . . such inferior Courts as the Congress may from time to time ordain and establish.”116U.S. Const. art. III, § 1; see also id. art. III, § 2 (detailing jurisdiction—but only as a demarcation of how far the “judicial Power” can “extend” rather than an independent grant of power). Premise (2) also asserts that originalism is true. As explained above, this Article aims to do originalism, not defend it. More precisely, premise (2) asserts original public meaning originalism, which says the meaning of constitutional text is determined by what the original, ratifying public would have interpreted that text to mean.117The rest of this Article would not look much different on any other version of originalism.

Premise (3) is the hinge. But if the historical conclusions above are correct, it holds. Recall the main contention: The old conception of the customary-and-reasonable common law remained dominant in America well past the Constitution’s ratification. And on that conception, courts could find the common law but not make it.118This Article’s historical analysis focuses on common law because it is surpassingly obvious that judges interpreting and applying statutes were not considered lawmakers. Instead, they were “finding law” in the most straightforward sense possible: by reading the statutes in front of them. So it is hard to imagine an ordinary reader (lawyer or non-lawyer) at the time of the ratification who would have read “judicial Power” to include the power to make common law. Rather, such a reader would have taken “judicial Power” to include the power “to say what the law is,” and nothing more—even when it came to common law.119Marbury v. Madison, 5 U.S. (1 Cranch) 137, 177 (1803). And if that is so, then premise (3) is true. Finally, if premises (1), (2), and (3) are all true, the thesis follows.

An objector might argue like so: “What if the phrase ‘judicial Power’ had no definite meaning? Perhaps the grant of the ‘judicial Power’ was originally read to grant something like ‘the power to decide cases according to right principles of jurisprudence.’ And yes, at the time of the ratification, nobody would have taken this to let federal judges make common law. But now, we know better. We know that finding law is a loser’s game, and we know judicial lawmaking is the right way to decide cases. Therefore, even if originalism is true, premise (3) is unsupported.”

But this objection neglects the sharp line, already discussed, that Founding-era thought drew between judicial and legislative power.120See supra Section I.B. Again, though Desan shows the states sometimes crossed this line, they did not do so in a way that gave courts legislative functions. See supra notes 106–09 and accompanying text. Justice Iredell’s words are worth repeating as just one example: “the distinct boundaries of law and Legislation [should not] be confounded.”121Chisholm v. Georgia, 2 U.S. (2 Dall.) 419, 448 (1793) (opinion of Iredell, J.) (emphasis added). It should go without saying, but I assume that legislation is a form of lawmaking and not lawfinding. That conflation “would make Courts arbitrary, and in effect makers of a new law, instead of being (as certainly they alone ought to be) expositors of an existing one.”122Id. (emphases omitted). I do not, then, claim that “judicial Power” had a great deal of meaning packed into it. But I do believe the word “judicial,” as used in opposition to Article I’s “legislative,” referred to something entirely distinct from and mutually exclusive with the power to make law. The history backs up this claim.123See supra Sections I.A–B. Interpreting “judicial Power” to describe the power to decide cases any which way, including by exercising the legislative power decidedly not granted by Article III, contradicts the Founding-era understanding of the role of courts.

A second objector might say: “This argument, if sound, proves more than you bargained for. Because natural law was part of the English common-law tradition, your argument entails that federal courts are required both to find law and to constrain themselves to the bounds of the natural law. Thus, if your conclusion follows, it also follows that federal courts must apply highly controversial and archaic natural-law theory in common-law cases.”

Assume, for argument’s sake, that (a) natural law was part of the English common-law tradition and (b) it would be a bad thing for today’s courts to apply the natural law in common-law cases. This objection still fails. It reads much more into the phrase “judicial Power” than is plausible. It is one thing to infer that the original meaning of “judicial Power” did not include lawmaking power—as just mentioned, it was quite clear at the time of the Founding that the judicial and legislative powers were distinct. But it is quite another thing to infer that the phrase “judicial Power” imposed an affirmative requirement that courts look to the natural law when deciding common-law cases.

Think of it this way. If a member of the public of 1788 saw a judge hand down a ruling that violated the natural law, he might think the ruling to be a wrong, unjust, or even void exercise of power. But surely he would not consider the ruling to be an exercise in non-judicial power. The court in question may have violated the natural law, but that does not mean it was making law or otherwise doing something non-judicial. So there is little less reason to think “judicial Power,” as a matter of original semantic meaning, excluded natural-law-violating exercises of that power. The argument does not prove overmuch.

A third objection is as follows: “The founding generation may have believed that common-law courts were finding law, but perhaps they were mistaken in this. Perhaps courts have been lawmakers all along. If so, why should the original understanding of ‘judicial Power’ bind us at all? If even those courts were not law-finders, then it is hard to see why their mistaken belief that they were should have any constitutional import.”

The response is simple: there is no reason to believe early judges were making law. There would be reason to think so if finding unwritten law were impossible, or if anything more than a tiny minority of thinkers thought finding law was a sham. Neither is true.124See Sachs, supra note 9, at 530–31. Another possible response is that, even if the common law has been judge-made all along, what counts is the meaning of Article III’s language. Even if the Founding generation failed to practice what it preached, that failure does nothing to change the meaning of Article III’s language. So it does nothing to change the content of the law. Neither is there any reason, beyond an unjustified sense of chronological superiority, to assume judges and commentators were flat lying about what went on in common-law cases.

II.     How to Find Law

So federal courts cannot make common law. That alone is significant, for it shows that the dominant assumption underlying federal common law is not merely wrong but constitutionally unsound. Yet it might also be helpful to have a method for finding law. It is easy enough, after all, to complain about federal courts’ behavior. It is both harder and more helpful to suggest a real alternative. That is what I aim to do next.125Sachs and Professor Caleb Nelson have each discussed finding law, but neither has provided a lawfinding method. See Sachs, supra note 9, at 530 (cabining the article to the claim that finding law is possible); Nelson, supra note 7, at 9 (also cabining). Also, it is no secret that H.L.A. Hart suggested that “[custom], like statute, [can] be law before the court applies it.” H.L.A. Hart, The Concept of Law 47 (1961). For his general discussion of this topic, see id. at 43–48. The following account is based in this view, and I am therefore greatly indebted to Hart.

A.     A Lawfinding Method

As Part I discussed, America inherited a customary common law.126As likewise discussed above, custom was only one of the common law’s central aspects; the other was artificial reason. See Postema, Philosophy of the Common Law, supra note 29, at 590. However, law-as-reason is not particularly helpful for finding law. This is because artificial reason did not define the content of the common law. Instead, it simply characterized the way of thinking common-law judges employed to reach their holdings. As Postema puts it, artificial reason “by itself provided no contentful tests by which to assess the legitimacy of a given legal rule or doctrine.” Id. at 593; see also Postema, Bentham, supra note 35, at 7 (“The principles of Common Law are not themselves validated by reason; but they are the products of a process of reasoning, fashioned by the exercise of the special, professional intellectual skills of Common Lawyers over time refining and co‑ordinating the social habits of a people into a coherent body of rules. In this sense . . . it is in the nature of law to be reasonable, but at the same time the law, emerging from this unique intellectual process, constitutes the standards by which the community judges the reasonableness or unreasonableness of actions.” (emphases omitted) (footnote omitted)). Because custom, unlike reason, did determine and circumscribe the content of the common law, I focus on custom here. Indeed, it strikes me as possible that if present-day lawyers were to return to finding law—by way of the custom-based method I describe here—that change in our way of thinking might itself be a return to the “artificial reason” of the common law tradition.

Also, this sentence is slightly imprecise. Depending on how one reads the history, it is possible that the common law was not just customary but also incorporated at least parts of the natural law by reference. See discussion supra note 41. For an objection based on this point, see infra Part II.A.1.
The method centers on this point. We can find rules of the common law by (a) determining the essential127See infra note 152 for discussion of this word. characteristics of customary law and (b) asking whether a would-be legal rule has those characteristics. If the rule does not, then it is no part of the customary law and therefore no part of the common law. The converse is also true: if a would-be rule has the essential characteristics of customary law, then it is part of the common law.128This assumes that all American customary law is ipso facto common law, which may be imprecise. It also leaves unspecified whose common law we are talking about (federal common law, general law, or state law—and if so, which state?). For simplicity’s sake, I will not delve too deeply into those choice-of-law issues here. Suffice it to say that each jurisdiction’s common law will be determined by the customs of that jurisdiction.

Sir Matthew Hale is a useful resource for understanding the essential characteristics of customary law. I have four justifications for relying on Hale’s account. First, Hale was writing in the seventeenth century. And it was seventeenth-century English legal views which the colonists took with them to America—views they continued to endorse through the Founding era.129“[I]t was [the] common law practice in its seventeenth-century form . . . that spread with English colonialism to the New World and beyond.” Postema, Philosophy of the Common Law, supra note 29 at 589; see also Jay, Part One, supra note 16, at 1053–1057 (discussing, among other things, the perseverance of such views as Hale’s in America); id. at 1057–60 (discussing Connecticut Judge Zephaniah Swift’s early eighteenth-century endorsement of a conservative, customary form of common law very much like Hale’s); Alan Cromartie, Sir Matthew Hale, 1609–1676: Law, Religion, and Natural Philosophy 30–41 (1995) (discussing the similar views of John Selden, another seventeenth-century common lawyer). Further, the case of Livingston v. Jefferson, 15 F. Cas. 660 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,411), showcases one early American court (in an opinion by Justice Marshall, riding circuit) using at least two of the three Hale-derived principles I am about to discuss. Second, Hale articulated his views on customary law more clearly and consistently than both Coke (whom Hale succeeded as Chief Justice of the Court of King’s Bench) and many of Hale’s own contemporaries.130See supra Section I.A.1 for brief discussion of Coke’s view, which was ecstatic and historically dubious; see also McConnell, supra note 43, at 183–86 (discussing the historical problems with Coke’s view of custom-as-immemorial). Third, Hale’s views on the topic were mainstream: when it came to customary law, “the dominant view [was] that of Hale.”131Postema, Bentham, supra note 35, at 6; see also 1 James Wilson, Of the General Principles of Law and Obligation, reprinted in The Works of James Wilson 97, 102 (Robert Green McCloskey ed., 1967) (series of lectures given in 1790–92, in which Wilson expounded the view that “custom [is] the most significant, and the most effectual” source of law). In this sense, it is a little like a functional restatement of the law as it stood at the time—and as it was later endorsed in America.132See McConnell, supra note 43, at 186–89 (explaining the Founding-era American acceptance of Hale’s view and discussing James Wilson in particular). And fourth, as described below, Hale’s account is workable. It provides three principles which can easily be put to use in the lawfinding project.133That being said, I do not claim the final word on what counts as customary law and what does not in the American context. My hope is to propose one possible method of lawfinding, not to have the last word on the subject. Others are more than welcome to propose their own methods. Perhaps these methods would boast stronger historical pedigrees or draw even sharper, more easily administrable lines between lawmaking and lawfinding.

The customary law’s first characteristic is its internal continuity.134This Section discusses only “customary law.” This is because I am concerned only with common-law-as-custom at the moment. However, because all American common law is customary law, the distinction between “common law” and “customary law” makes little difference. This is a small-c conservative concept: customary law can and does change over time, but only very gradually. That affords some much-needed flexibility, unlike Lord Coke’s view that the common law remained substantively unchanged for centuries.135See Postema, Bentham, supra note 35, at 6. For Blackstone’s discussion of the common law’s flexibility, which sometimes involved legal fictions in the interest of gradual change, see 3 Blackstone, supranote 30, at *265–69 (discussing the ability of the common law of property to adapt itself even to the rather drastic shift away from feudalism); see also Postema, Bentham, supra note 35, at 11–13 (discussing this part of Blackstone).At the same time, the requirement of continuity means that “through[out] [its] changes [the customary law] maintained its integrity as a single, coherent body of law.”136Postema, Philosophy of the Common Law, supra note 29, at 591 (citing Hale, supra note 35, at 40). Hale also compared the common law to the ship of the Argonauts: over the course of the journey, every plank of the ship had been replaced—but the ship remained the same. See Hale, supra note 35, at 40. James Wilson elaborated on Hale’s view in 1790, explaining that “though the [common-law] system suffer . . . partial and successive alterations, yet it continues materially and substantially the same.”1371 James Wilson, Of the Common Law, reprinted in The Works of James Wilson, supra note 131, at 334, 353. In the same passage, Wilson (again drawing on Hale) described common-law change as so slow that it happened only “insensibly.”138Id.; see also McConnell, supra note 43, at 186–88 (discussing a passage from Hale and emphasizing that change must happen, if at all, “insensibly,” with the conclusion that “‘[i]nsensible’ adaptation to ‘insensible’ change was, as Hale said, more likely to produce a ‘reasonable’ answer than [other] alternative[s]” (citing Hale, supra note 35, at 44)).

We can operationalize the principle as follows: a would-be legal rule that changes the customary law gradually, or not at all, is continuous with existing customary law.139Where does one look to find the existing customary law? The casebooks are not entirely dispositive, but past cases are very good evidence of it. See Jay, Part One, supra note 16, at 1056 (collecting sources in support of the proposition that, on the English and early American majority view, “[c]ases were merely the best evidence of the principles of the common law”); Hale, supra note 35, at 45 (“[Judicial decisions] do not make a Law properly so called . . . yet they have a great Weight and Authority in Expounding, Declaring, and Publishing what the Law of this Kingdom is especially when such Decisions hold a Consonancy and Congruity with Resolutions and Decisions of former Times; and tho’ such Decisions are less than a Law, yet they are a greater Evidence thereof than the Opinion of any private Persons, as such, whatsoever.”); Wilson, supra note 137, at 353. For this reason, the principle of continuity ends up looking a bit like a commitment to a form of stare decisis. A rule that purports to change the law too radically is discontinuous and does not become part of the customary law at all.140As the below discussion of Livingston v. Jefferson will illustrate, a rule that breaks continuity at the time of its announcement might nonetheless obtain a place in the customary law if it is received in the realm over a long period of time. See infra Section II.B.1.

The customary law’s second characteristic is its reception in the realm. Hale’s common law was “‘received and approved’ in the kingdom.”141Postema, Philosophy of the Common Law, supra note 29, at 591. This principle means that, no matter what a court may say, a rule is not part of the customary law until it is accepted by both the people who will apply the rule and the people affected by it. Wilson opined that “custom” is “the most significant, and the most effectual” source of law precisely because custom “involves in it internal evidence, of the strongest kind, that the law has been introduced by common consent; and that this consent rests upon the most solid basis—experience as well as opinion.”142Wilson, supra note 131, at 102.

Or as Postema explains:

To claim that common law existed “time out of mind” meant that the validity and binding force of any rule of law depended not on who made it or when, but on its being “received and approved” in the kingdom. This “reception” was manifested in their integration into the body of doctrines and practices that makes up the common law. Integration, not origin, was the key. This integration of custom, statute, or judicial decision is not simply a matter of logical consistency or coherence. It is a practical and historical matter: practical, because it is a matter of whether the rule is “taken up,” practised, and used (by its subjects and by officials who must assess their actions in light of the law); and hence, historical because only time can tell whether a rule, however it happened to be introduced, is thus integrated and becomes part of the common law.143Postema,Philosophy of the Common Law, supra note 29, at 591 (emphases omitted) (citing Hale, supra note 35, at 3, 6, 8).

In more concrete terms: even after a court announces a rule of would-be law, there is a wait-and-see period. If the people and relevant governmental actors144Other courts can presumably also factor into this analysis. And Livingston suggests that reception by the courts is indeed relevant. See Livingston v. Jefferson, 15 F. Cas. 660, 664–65 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,441). approve of the rule and adopt it, the rule has been “received.” A rule that has not been thus received is not part of the customary law, even though the court’s ruling still binds the parties before it.145See Postema, Bentham, supra note 35, at 5 (describing the traditional common law theory as holding that, “[s]ince the [common] law exists only in, and is known only through, practice, a rule becomes law or a decision marks a new departure in the law only if it is taken up into the practice of the community. Only time can tell whether a rule becomes a law, because only time—i.e. practice and use over time—validates”). One could imagine a requirement that the relevant actors give explicit approval of the rule in question, but more realistically to “approve” might just entail tacit assent to the rule.

And finally, the customary law’s rules are accommodated to the people’s way of life. Wilson again: “It is the characteristick [sic] of a system of common law, that it be accommodated to the circumstances, the exigencies, and the conveniencies [sic] of the people, by whom it is appointed.”146Wilson, supra note 137, at 353. And once more, Postema elaborates:

[F]or Hale, and many of his contemporaries, integration involved accommodating the rule or maxim to the nature of the nation . . . . [T]he process of accommodation worked two ways. The rules of law, at first rough and clumsy, are refined over time, softened to fit the contours of the community’s daily life. Simultaneously, following the rules and practices shapes the dispositions, beliefs, and expectations of the people. Thus, what they took to be reasonable and practicable solutions to the problems of social interaction depended on a sense of continuity of present practice with the past; but also, what counted as continuous with the past depended heavily on what were regarded by participants as reasonable projections from the arrangements and practices of the past to present conditions and problems.147Postema, Philosophy of the Common Law, supra note 29, at 591–92.

Thus, the customary law has a symbiotic relationship with the life of the community. On one hand, the community’s character is influenced by the law. On the other hand, a rule can become part of the customary law only if the rule takes adequate account of the people’s customs. So a rule superimposed upon the political community without regard for that community’s relevant148This qualifier simply means that not every one of a community’s customs will be relevant to every would-be addition to the customary law. For example, if a case concerns the proper standard of care for bargees, the holding must accord with the customs of the nautical world. See infra Section II.B.2. But the customs of, say, landowners and their guests need not come into it. See infra Section II.B.3. peculiarities is not a part of the customary law. This principle is similar to reception in the realm.149See Postema, Philosophy of the Common Law, supra note 29, at 592. But the two are not identical. Reception involves a conditional, probationary period before a rule becomes part of the customary law.150See id. at 591–92. In contrast, the accommodation principle requires at the outset that any change to the customary law take due account of extra-legal customs.151Wilson suggests that in some extraordinary cases—in “some great eras”—circumstances can change so drastically that the common law must undergo “considerable changes” to accommodate the people’s customs. See Wilson, supra note 137, at 354. (“At some of those eras, the improvement is as rapid as the change is great.”). Wilson had the American Revolution in mind—hence, “[w]hy should not the present age in America, form one of those happy eras?” Id. Thus, it seems possible that accommodation can allow (require?) a non-continuous shift in law. But given the context—the Revolution—it also seems likely that this kind of rapid shifting is permissibly in only very rare, truly extraordinary circumstances.

These three principles, it bears emphasizing, are characteristics of customary law itself. This is what makes them so useful for finding law. Find a rule and ask: Would adopting this rule be continuous with existing customary law? Does the rule adequately accommodate the customs of the people it would be affecting? And after a rule has been adopted, has the rule been “received in the realm”? If the answer to all three questions is “yes,” then the rule is part of the customary law. If the answer to one of the questions is “no,” then the rule is ineligible for customary-law status.152These characteristics are essential properties of customary law. In other words, they are not mere signifiers of customary-law status, but instead necessary aspects of the very nature of customary law. This assumption lines up nicely with Hale’s own view, on which customary law seems to have attained the status of law in virtue of exhibiting these characteristics. See Hale, supra note 35, at 43 (“[T]he Strength and Obligation, and the formal Nature of a Law, is not upon Account that the Danes, or the Saxons, or the Normans, brought it in with them, but [rather that] they became Laws, and binding in this Kingdom by Virtue only of their being received and approved here.” (emphasis added)). For discussion of the nature of essential properties, as opposed to accidental properties, see Robertson Ishii, Teresa Atkins & Philip Atkins, Essential vs. Accidental Properties, Stan. Encyc. Phil. (Oct. 26, 2020), This process may not be entirely determinative—there may be several possible rules that would bear all three characteristics of customary law. Thus, this method of finding law cabins judicial discretion without entirely eliminating it.153This does not mean the judge who picks among the possible rules is making law for two reasons. First, though the traditional process might leave some discretion in the end, the process itself is fundamentally one of finding and not making. The lawfinding judge is obliged to examine the existing customary law and the customs of the people. The mere fact that this process may yield several possible rules is of no import. To put it bluntly, why believe that the mere existence of discretion entails lawmaking? See Jay, Part One, supra note 16, at 1059–60 (discussing Zephaniah Swift’s view of the common law, which included judicial “discretion [to shape] their rules to every possible variety of circumstance” even while maintaining that judges engaging in this process were not making law). On this point, I partially agree with Caleb Nelson. Nelson suggests that, on the “possible view” that “unwritten law does indeed rest partly on sources that exist outside of the courts, such as real-world customs and other social practices, but these sources are only partially determinate,” judges are engaging in “only a subsidiary type of ‘lawmaking’” when they exercise their discretion. Nelson, supra note 7, at 13 (footnote omitted) (going on to suggest that this form of decision-making is “less analogous to legislation than to a species of interpretation”). I would go further and say this is not lawmaking at all—I am not quite sure why scholars (including Nelson) tend to assume, without argument, to the contrary.

And second, even after the judge picks a rule and applies it in the case before her, the body of customary law has not been enlarged. The principle of reception means that, though the rule binds the parties to the case, it does not become law until it is received in the realm. For further discussion of ways in which customary law can maintain flexibility without judicial lawmaking, see Postema, Bentham, supra note 35, at 10–13 (noting that even in entirely novel cases, the principle that “like cases should be treated alike” meant that “[it was] the case, not the judge, that extend[ed] the law” (quoting Calabresi, supra note 44, at 13)); see also Sachs, supra note 9, at 561–63 (discussing “As-if law” along similar lines).

Similarly, the characteristics of customary law can help distinguish judge-found holdings from judge-made ones. To bring out the point, here is a formalized argument:


(1) If a holding is judge-found, then the holding successfully picks out either a rule of customary law or a rule of natural law;


(2) If a holding successfully picks out a rule of customary law, then the holding is continuous with existing customary law; it is received in the realm; and it accommodates itself to the everyday customs of the relevant people;


(3) Therefore, if a holding is judge-found, then the holding either exhibits these three characteristics or it successfully picks out a rule of natural law;


(4) Therefore, if a holding does not exhibit these three characteristics, the holding is either judge-made or successfully picks out a rule of natural law.


The idea behind premise (1) is that, if a holding is judge-found law, the holding must really find something. In the common-law context, the only two realistic candidates are rules of customary law and rules of natural law. Premise (2) simply restates the core attributes of customary law: attributes without which a would-be legal rule is no part of the customary law at all. This is subject to the qualification that (as established in the discussion of Livingston v. Jefferson,15415 F. Cas. 660 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,441). below) a warm reception can eventually allow a non-continuous rule to be grandfathered into the customary law. In any case, if the first two premises are both true, (3) follows. (4) is just a way of reshuffling (3) to emphasizes the consequences of failing to fulfill the principles.155(4) follows from (3) as long as one implicit premise holds: that a holding is either judge-found or judge-made. This seems self-evident.

 1.     What About the Natural Law?

Here is an objection. “The above argument leaves open the possibility that a holding can be dissonant with the customary law even while conforming with the natural law. A court that sets forth such a holding need not be making law; it might merely be finding and applying a rule of natural law. Thus, simply asking whether a holding conforms with the principles of customary law will not tell us whether the holding is judge-made.”

That is fair. If a federal court really does find and apply natural-law rules to decide cases, the court is not making law. That being said, I want to make two comments. First, this Article takes no stance on the proper relationship, if any, between the common law and the natural law. All it has to say on that matter is that using natural law principles as rules of decision is not illegitimate on judicial-lawmaking grounds. It does not follow that using natural-law principles as rules of decision is legally permissible or a good idea. Appealing to natural law might be wise or unwise; it might be constitutional or unconstitutional; it might be illegal on non-constitutional grounds; or it might be mandatory on moral grounds. Those questions are for another day.

Second, the principles of customary law can still be used to tell the difference between judge-made and judge-found law. Despite its recent resurrection from total neglect, natural-law jurisprudence is very far from the mainstream at present.156See Diarmuid F. O’Scannlain, The Natural Law in the American Tradition, 79 Fordham L. Rev. 1513, 1514–20 (2011) (lamenting this and discussing the views of prominent natural-law scholars); see also id. at 1516, 1518–20 (discussing the views of Professors Hadley Arkes and Robert P. George); J. Joel Alicea, The Moral Authority of Original Meaning, 98 Notre Dame L. Rev. (forthcoming 2022) (discussing the relationship between natural law and originalism). Mentioning the idea to a legal scholar is more likely to invoke a sneer than thoughtful discussion.157This is not conjecture. So while it may be possible in principle that any given deviation from customary law is really a swerve into the natural law, the fact is that American courts have not conceived of themselves as finding and applying the natural law for quite a long time.158See generally Stuart Banner, The Decline of Natural Law: How American Lawyers Once Used Natural Law and Why They Stopped (2021). Thus, if a federal court’s common-law ruling deviates from the customary law, the court is almost certainly—in reality and by its own lights—attempting to make a new legal rule rather than applying the natural law to the facts.

The objector might respond that, in a hypothetical world where everyone accepts that federal judges cannot make common law, there is nothing stopping willful judges from making things up and then claiming, disingenuously, that their preferred rules are part of the natural law. I beg to differ. There certainly is something getting in such a judge’s way. Other people would be able to call him out on grounds that these made-up rules did not really conform with the natural law.159See Antonin Scalia, Sovereign Immunity and Nonstatutory Review of Federal Administrative Action: Some Conclusions from the Public-Lands Cases, 68 Mich. L. Rev. 867, 867 (1970) (calling it “inherited wisdom” that “responsible professional comment and criticism are the principal restraints upon judicial arbitrariness at the highest level”). If the principles of the natural law are discoverable, after all, there is nothing stopping others from discovering them and pointing out that the disingenuous judges are making things up. In other words, natural law—again, on the assumption it has a role to play—is a set of discoverable rules and legal principles just like any other.160Hadley Arkes, a staunch defender of the natural law, explains his view of how this can work in the following way:

It frequently happens that some of our friends who are most skeptical of natural law discover that they have been practicing it handsomely for many years without quite realizing it—much like that character in Molière who discovers that he has been speaking prose all his life. It is rather like the man who asks, ‘Can I order coffee without using syntax?’ He may not realize that of course he is using syntax and speaking prose without quite recognizing the conceptual world he inhabits, or the understandings that are woven into his own nature. It is no wonder, then, that we find some of our best natural lawyers among the distinguished jurists who have been the most skeptical of natural law. They may go on to discover, as a late colleague once said, that we have principles we have not even used yet.

Hadley Arkes, Constitutional Illusions and Anchoring Truths: The Touchstone of the Natural Law 78 (2010).
Those rules and principles can be used well or poorly, honestly or dishonestly. The problem of disingenuous judges need not be more dramatic in the realm of natural law than in any other area of law.

A final word on natural law: imagine the hypothetical world where most lawyers and judges agree that Article III grants no power to make common law. In that world, suppose judges end up resorting (in good faith or not) to natural law in lieu of making their own common law. This would likely give judges back much of the discretion that judicial lawmaking gives them in the status quo. Yet this hypothetical legal world would be drastically different from the real world. Yes, federal judges would still have discretion in common-law adjudication—but that is about the only similarity. The entire project of common-law judging would be flipped on its head. Judges would no longer be acting as legislators, but instead as designated law-finders, seeking customary and natural law and using it to decide cases. So, though it is perhaps tempting to say that resort to natural law would leave the legal world unchanged, that sounds to me like saying baseball and football are the same sport just because they both involve grass.161Indeed, maybe that is the way our legal system was really supposed to be. As I keep reiterating, this Article does not rest on natural-law theory—it works even on the view that common law is purely customary and positive. But I also want to be clear that I am not rejecting natural-law theory either. It is simply a topic for another day.

 2.     What About the Text?

The semantic meaning of the phrase “judicial Power” does not include the three above-described characteristics. Instead, that meaning merely excludes lawmaking, and it would be a mistake to read too much more into the phrase.162See supra Part I. So how can the three principles of customary law be constitutionally binding? I have two answers.

Most straightforwardly, there is no need to locate continuity, reception, and accommodation in the Constitution at all. The Constitution rules out judicial lawmaking, and that means our job is to figure out how to do common-law adjudication without making law. The way to do so is by recovering the traditional rules of customary common law. If anybody knew how to find customary law, English common lawyers like Hale did.

For those who find that unsatisfying, the second answer is that these principles may be a matter of construction. As Professor Randy Barnett puts it, interpretation is the act of “discovering the semantic meaning of the words in the text of the Constitution,” and construction is the act of “putting that meaning into effect by applying it in particular cases and controversies.”163Randy E. Barnett, Interpretation and Construction, 34 Harv J. L. & Pub. Pol’y 65, 65 (2011). Construction is sometimes necessary to alleviate constitutional problems of vagueness (situations where a word or phrase “needs to be applied to an object that may or may not fall within its penumbra”).164Id. at 68–69. Another possibility is that the meaning of a vague constitutional word or phrase was liquidated by early practice. As Professor William Baude describes, constitutional liquidation requires “three key elements”:

First, there had to be a textual indeterminacy. Clear provisions could not be liquidated, because practice could “expound” the Constitution but could not “alter” it. Second, there had to be a course of deliberate practice. This required repeated decisions that reflected constitutional reasoning. Third, that course of practice had to result in a constitutional settlement. This settlement was marked by two related ideas: acquiescence by the dissenting side, and “the public sanction”—a real or imputed popular ratification.

William Baude, Constitutional Liquidation, 71 Stan. L. Rev. 1, 1 (2019). Here, there is simply not enough evidence of “a course of deliberate practice” that would have liquidated “judicial Power” to include Hale’s three principles of customary law in particular.

On that view, judges faced with constitutional vagueness or other textual under-determinacies should “identify the original functions or spirit of the provision” and “then specify a construction—an implementing doctrine—that resolves the case at hand in a manner that is consistent with the relevant original function, and susceptible of application to future cases of a similar kind.”165Randy E. Barnett & Evan D. Bernick, The Letter and the Spirit: A Unified Theory of Originalism, 107 Geo. L.J. 1, 35 (2018) (footnote omitted). Not all originalists countenance any construction at all.166See generally, e.g., John O. McGinnis & Michael B. Rappaport, Original Methods Originalism: A New Theory of Interpretation and the Case Against Construction, 103 Nw. U. L. Rev. 751 (2009). And a full-scale defense of constitutional construction is beyond the scope of this Article. But I maintain that implementing the vague phrase “judicial Power” by drawing a line between judge-found and judge-made law falls well within the purview of “good-faith construction.”167Barnett & Bernick, supra note 165, at 33. The phrase “judicial Power” has the function of excluding judicial lawmaking, but it is vague about exactly which judicial holdings count as lawmaking. The above-described method does nothing more than implement the phrase “judicial Power,” alleviating the vagueness problem by providing an administrable line between lawfinding and lawmaking. And it does so by appeal to history. This construction is therefore consistent with the “original function” of the phrase (i.e., to grant some powers, but not others, to federal courts).168Id. at 35. And as the next Section shows, the construction is indeed “susceptible of application” to a range of cases.169Id.

B.     Applying the Method

This Section applies the lawfinding method to four cases (in one instance, a series of cases) and asks whether each court was finding or making law. Fortunately, finding law is not just constitutionally required. It is also practically workable.

1.     Livingston v. Jefferson

Livingston v. Jefferson showcases real lawfinding in an early Republic case. This was a trespass action in the Circuit Court for the District of Virginia, which was the equivalent of a modern-day district court.170Livingston v. Jefferson, 15 F. Cas. 660, 660 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,441). The plaintiff was Edward Livingston.171Id. He sued former President Thomas Jefferson for trespasses that Jefferson, while still president, had allegedly committed in New Orleans.172Id. But Jefferson was not in New Orleans when the trespasses happened; he had merely ordered troops to do the alleged trespassing.173Id. Because Jefferson resided in Virginia,174Id. at 663 (Marshall, Circuit Justice). Livingston sued in diversity in Virginia federal court.175See id. at 660. If Livingston had sued in New Orleans, the court would have lacked personal jurisdiction over Jefferson. This explains why Livingston sued where he did. See id. at 665 (Marshall, Circuit Justice).

The court dismissed for lack of jurisdiction, and Justice John Marshall (riding circuit) and District Judge John Tyler Sr. each wrote an opinion.176Livingston, 15 F. Cas. at 663 (Tyler, District Judge); id. at 665 (Marshall, Circuit Justice). Jurisdiction depended on whether an action for trespass was a “local” or a “transitory” action under the common law.177See id. at 661–62 (Tyler, District Judge) (“But to proceed in the examination of the point before us; and that is, to inquire, whether this court has jurisdiction over this cause? And how it comes to be made a question at this day, I confess myself entirely at a loss to say; but as it is made, we must determine it. By the common law, which was adopted by an act of convention of this state, so far as it applied to our constitution, then formed, this point has been settled uninterruptedly for centuries past, and recognized by uniform opinion and decisions, both in England and America.”). A local action had to be brought, if at all, in a district of the same state where the facts giving rise to the action had occurred.178See id. at 663 (Marshall, Circuit Justice). A transitory action could be brought in any district in which there was otherwise jurisdiction.179In this case, that question turned on the Judiciary Act of 1789, 1 Stat. 73, which allowed suit in any district where the defendant resided or could be found. See Livingston, 15 F. Cas. at 665. Roughly speaking, the plaintiff was asking for a change in the common law: trespass was generally considered a local action, and the plaintiff was arguing it should be made transitory.180Id. at 661–62 (Tyler, District Judge) (describing the plaintiff’s argument and criticizing it as, among other things, too innovative).

As a preliminary matter, Justice Marshall’s opinion explained that the distinction between local and transitory actions was a matter of common law.181Id. at 665 (Marshall, Circuit Justice). The opinions seem to suggest Tyler saw this as a question of pure Virginia law, and Marshall saw it as a question of federal law—perhaps general law. On Tyler’s view, the English common law was binding only in virtue of Virginia’s legislative enactment of the common law. Id. at 661–62 (Tyler, District Judge). But Marshall said, “I am decidedly of opinion, that the jurisdiction of the courts of the United States depends, exclusively, on the constitution and laws of the United States.” Id. at 665 (Marshall, Circuit Justice) (also noting that even if the common law of England had not been adopted by Virginia statute, he “should have thought it in force”). For discussion of the general law, see Bellia & Clark, supra note 74. Because general law was itself a form of common law, see id. at 658–60, I do not belabor the difference here.

I call this a “preliminary matter” because it is logically prior to the rest of Marshall’s opinion. Even though he made the point partway through, the rest of the opinion rests on it.
For the content of that common law, Marshall looked to the ancient rules of English common law. Any decision before the moment of the American Revolution, he said, remained authoritative in America by default.182See Livingston, 15 F. Cas at 664–665 (Marshall, Circuit Justice). English common-law decisions made after the Revolution, Marshall said, are persuasive authority but not part of American common law and thus not binding. See id. Additionally, Marshall took a Coke- and Hale-inspired view of the common law, seeing it as “human reason applied by courts, not capriciously, but in a regular train of decisions, to human affairs, according to the circumstances of the nation, the necessity of the times, and the general state of things.”183Id. at 663. He noted that the common law “was thought susceptible of modification—and judges have modified it.”184Id. (emphasis added).

The story of modifications goes like this: The original rule said “all actions were local.”185Id. But, finding this rule too constrictive, English judges created exceptions. They made “all personal torts, and . . . all contracts wherever executed” into transitory actions.186Id. at 664. Then, despite some failed attempts by Lord Mansfield to make even more kinds of actions transitory, the exceptions went no further.187See id. (disregarding Mansfield’s opinions on this matter because one was in dicta and two were at nisi prius (tried before the King’s Bench) and thus had no precedential authority).

So as of Livingston, the rule was that if the events giving rise to the action “might have taken place anywhere,” as with a contract or a tort, the action was transitory.188Livingston, 15 F. Cas. at 664 (Marshall, Circuit Justice). But if those events were “in [their] nature necessarily local,” the action was local.189Id. The question, then, was whether to keep with this distinction or disregard it and create yet another exception to the original rule.

Marshall (with Tyler) accepted past judges’ modifications of the original rule, but concluded “it would require a hardihood which I do not possess” to modify the rule further.190Id. Strikingly, Marshall admitted the rule he was applying made little sense. Actions concerning real estate contracts are precisely as locality-bound as trespass actions: both may require investigation of titles and boundaries and even a land survey.191Id. But because, in land-contract actions, these considerations are “more than countervailed by the opposing consideration, that if the action be disallowed, the injured party may have a clear right without a remedy in a case where the person who has done the wrong, and who ought to make the compensation, is within the power of the court,” land-contract actions were held transitory rather than local.192Id. (emphasis added). In other words, past judges had attempted to make land-based contracts transitory because keeping them local would let defendants off the hook on a mere technicality. And Marshall acknowledged that this same “total failure of justice” would likely happen if trespass cases (land-based as they are) remained local.193Id. at 665. Yet he concluded that his hands were tied: “If, however, this technical distinction be firmly established, if all other judges respect it, I cannot venture to disregard it.”194Livingston, 15 F. Cas. at 664 (Marshall, Circuit Justice).

Livingston was a hard case not just because of fairness considerations, but also because it involved an originally discontinuous rule which had nonetheless been received in the realm for centuries. Thus, on one hand, Marshall was skeptical of the judges who changed the ancient rule to begin with. Throughout the opinion, he used hedging, sometimes disapproving, language to describe those modifications.195Id. at 663 (“This . . . principle of unwritten law . . . was thought susceptible of modification.” (emphasis added)); id. at 664 (“They have, without legislative aid, applied this fiction to all personal torts, and to all contracts wherever executed.” (emphasis added)); see also id. at 663 (referring to the scheme of modifications as “the creature of the court”). Though Marshall would not have seen the common law as entirely static, he seemed to suspect that prior judges had deviated too sharply from the relevant “principle of unwritten law” by introducing their deep-cutting exceptions to the local-transitory doctrine.196Id. at 663. But see id. (“This however being not a statutory regulation, but a principle of unwritten law, which is really human reason applied by courts, not capriciously, but in a regular train of decisions, to human affairs, according to the circumstances of the nation, the necessity of the times, and the general state of things, was thought susceptible of modification—and judges have modified it.”). I am slightly unsure whether Marshall thought these modifications were illegitimate—but it would be puzzling if he considered the earlier modifications to be legitimate and simultaneously thought himself incapable of carrying them further. In that case, why not make one more change? Regardless, what isclear, as discussed in the text above, is that (a) Marshall considered the modifications binding because they had been long accepted and (b) he considered further modifications unlawful. But he also recognized that the modifications—even though originally discontinuous with existing common law—had long been received in the realm. And Marshall was unwilling to disregard even improper modifications that had been accepted “for a long course of time.”197Id. at 664. To repeat his words: “If . . . this technical distinction be firmly established, if all other judges respect it, I cannot venture to disregard it.”198Id. (emphasis added). Though this Article’s framing of reception focuses on reception by other branches of government and by the people, I see no reason it should not also extend to other courts (and indeed, this quote provides strong evidence it should be extended to other courts).

Finally, even though these ill-advised changes had been received as part of the common law, Marshall concluded he was not allowed to modify them further. He noted that “for a long course of time, under circumstances which have not changed, [judges] have determined this to be the limit of their fiction.”199Id. So, because the limit of the exceptions had by then long been settled, Marshall concluded that “it would require a hardihood which I do not possess, to pass this limit.”200Livingston, 15 F. Cas. at 664 (Marshall, Circuit Justice); see also Sachs, supra note 9, at 547–48 (citing Livingston for the proposition that “[l]iving with past judicial shenanigans doesn’t mean giving carte blanche to future ones”).

In short, Marshall’s view was that judges should not have made these drastic modifications in the first place. But they did make the changes, and the realm accepted them over a long period of time. It was too late, by Marshall’s time, to right the wrong. Illegitimate-yet-successful modifications, like illegitimate-yet-successful political revolutions, demand acceptance with a shrug after the fact.201Cf. Sachs, supra note 9, at 547 (making a similar point). Given longstanding reception, Marshall considered himself unable either to do away with the modifications or extend them further.

This raises the interesting question, as an aside, of whether a court may ever make an exception to an existing rule without ipso facto engaging in lawmaking. To answer, we might draw a distinction between two kinds of common-law rules. Some rules are hard and fast. The traditional distinction between local and transitory actions, at least by Marshall’s lights, is an example. Because the rule was a categorical one, any exception automatically flew in the rule’s face. So exception making necessarily amounted to lawmaking.

But other rules are by their very nature more flexible. As an example, consider the English common-law rule against murder. We might think of the rule against “murder” as really being a rule against killings of a human beingwithout excuse or justification. Indeed, that is how English jurists sometimes formulated it.202E.g., 4 Blackstone, supra note 30, at *177–78. Coke once said, “[m]urder is when a man of sound memory, and of the age of discretion, unlawfully killeth within any county of the realm any reasonable creature . . . with malice fore-thought.”203Edward Coke, The Third Part of the Institutes of the Laws of England 47 (E. and R. Brooke 1797). That formulation has the concept of exceptions baked right in: A killing might be prima facie murder, but if an exception applies (e.g. the killer was not “a man of sound memory,” was below “the age of discretion,” or most broadly, did not act “unlawfully”), the killing is not murder after all. So if a court confronted a new homicide fact pattern, the court might be able to hold that an “exception” to the rule applies. That need not be lawmaking because the court, properly speaking, would simply be holding that the defendant never violated the rule against killings of a human beingwithout excuse or justification at all.

Livingston is a sterling example of early Republic lawfinding in federal court. Some of the same elements of customary law Hale had identified centuries earlier drove Marshall’s opinion. And though Marshall did not frame this as a constitutional case, he did see himself as powerless to radically alter the body of received common law.204Put another way, this Article’s constitutional claim depends on the original meaning (sense) of the phrase “judicial Power.” That original meaning does not depend on all early Americans, not even John Marshall, perceiving all of the sense’s legal consequences. See generally Christopher R. Green, Originalism and the Sense-Reference Distinction, 50 St. Louis U. L.J. 555 (2006).

 2.     United States v. Carroll Towing

Now consider whether United States v. Carroll Towing Co.,205159 F.2d 169 (2d Cir. 1947). in which Judge Learned Hand gave the first algebraic statement of his famous Hand formula, was an instance of judicial lawmaking. This was an admiralty case, not a common law case per se, but the same principles apply because admiralty law is another form of customary law.206See Am. Ins. Co. v. Canter, 26 U.S. (1 Pet.) 511, 545–46 (1828) (“A case in admiralty does not, in fact, arise under the Constitution or laws of the United States. These cases are as old as navigation itself; and the law admiralty and maritime, as it has existed for ages, is applied by our Courts to the cases as they arise.”). The key issue was whether the plaintiff, the owner of a sunk barge, was contributorily negligent on the ground that the bargee was absent at the time of the accident.207A bargee is an employee who sits on the barge and makes sure nothing bad happens. See Carroll Towing, 159 F.2d at 172. It is unremarkable that Judge Hand found the owner to be negligent. What is noteworthy, of course, is the way he reached his conclusion.

Judge Hand’s opinion first reviewed existing Second Circuit cases, searching for a “general rule to determine when the absence of a bargee . . . will make the owner of the barge liable for injuries to other vessels if she breaks away from her moorings.”208Id. at 172–73. But he found no broad rule dictating that bargees are always (or never) required. In some circumstances, courts had found a bargee’s absence negligent; in others, not.209Compare The Sadie, 57 F.2d 908 (S.D.N.Y. 1932), aff’d sub nom. New York & Cuba Mail S.S. Co. v. Sadie, 62 F.2d 1076 (2d Cir. 1933) (per curiam), with The Trenton, 72 F.2d 283, 285 (2d Cir. 1934). The reason for the lack of uniformity, Hand concluded, was not that courts were being inconsistent. In his words, “there can be no such general rule” because negligence in these circumstances is not determined solely by a bargee’s presence or absence.210Carroll Towing, 159 F.2d at 173. Negligence, instead, “is a function of three variables: (1) The probability that [the vessel] will break away; (2) the gravity of the resulting injury, if she does; (3) the burden of adequate precautions.”211Id. Before giving the algebraic Hand formula itself (negligence if and only if “B [the burden of the untaken precaution] is less than L [the harm caused] multiplied by P [the injury’s ex ante probability]”), Hand suggested this rule would apply “in other similar situations,” not just in admiralty.212Id.

At first glance, it seems like Hand was making law here. This was famously the first time the algebraic Hand formula appeared. So if the formula was a drastic departure from then-standing rules of negligence law in admiralty, then the formula would plainly be out of continuity with the existing customary law. Though it may be hard to tell how sharply a rule can change without counting as “discontinuous,” it is surely discontinuous to fabricate a formula out of whole cloth and in isolation from existing rules.

On the other hand, if the formula was nothing more than a reformulation of a pre-existing and unstated admiralty-law rule, Hand may not have run aground (so to speak). Judge Richard A. Posner has argued that the formula was an implicit part of admiralty law for many years, and all Hand did was make it explicit.213See Richard A. Posner, A Theory of Negligence, 1 J. Legal Stud. 29, 29, 32 (1972) (using “a sample of 1528 American appellate court decisions from the period 1875-1905” to argue that the Hand formula was “an attempt to make explicit the standard that the courts had long applied”). Professor Richard W. Wright, in contrast, contends that “Posner’s arguments are based on speculative and implausible assumptions, overbroad generalizations, selective quotations, and superficial descriptions of cases.”214Richard W. Wright, Hand, Posner, and the Myth of the “Hand Formula, 4 Theoretical Inquiries L. 145, 145 (2003). Rather than being implicit in admiralty law, Wright says the formula suffered from a “lack of support . . . in the preexisting cases.”215Id. at 148.

Regardless of who is right, the crucial idea is that whether the formula veered off the beaten path is an empirical question, to be determined by assessing how closely the formula approximated earlier admiralty negligence tests. If the formula just summarized existing rules, then Hand was not necessarily making law. But if the formula turned the law of negligence on its head, he was—and incidentally, he was exceeding the Article III judicial power by doing so.216This depends on the idea that each characteristic of customary law is an essential characteristic. For support of this proposition, see supra note 152.

Even if Carroll Towing was discontinuous with the law of admiralty and was therefore impermissible lawmaking,217I am setting aside the possibility that Hand was finding a rule of natural law. Livingston suggests that reception for “a long course of time” can render an otherwise-illegitimate holding a bona fide part of the customary law.218See Livingston v. Jefferson, 15 F. Cas. 660, 664 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,411) (Marshall, Circuit Justice). So, if Hand was impermissibly making law in Carroll Towing, has his formula since been so well received as to become part of the law nonetheless? One might think so: as most 1Ls can attest, law schools love the Hand formula. And law schools influence hordes of future lawyers. It also figures prominently in the field of law and economics.219Neil Komesar, The Logic of the Law and the Essence of Economics: Reflections on Forty Years in the Wilderness, 2013 Wis. L. Rev. 265, 312 (discussing the connections between tort law and economic analysis).

But we should be careful here: mere scholarly interest, or even consensus, is probably insufficient to establish reception in the realm. That is because many scholarly consensuses have little influence on legal actors. This situation may illustrate the point. Wright argues that the Hand formula, despite much discussion, has almost never been applied.220At least as of 2003. Wright, supra note 214, at 273. He concludes that “[i]f one . . . looks carefully . . . at what the courts are actually doing rather than (merely) at what they are saying, one finds that the courts almost never attempt to apply the [formula]; instead, the [formula] is merely trotted out as dicta or boilerplate separate from the real analysis.”221Id. Assume he is right. If so, it would seem implausible to declare the formula received—no matter how many law professors wish it had been.

We can speculate a bit about accommodation. Customary law must accommodate itself to the everyday customs of the people who live under, and with, the law. Thus, assessing a would-be rule of customary law requires careful specification of the relevant “people.” For this task, the key fact is that the Hand formula did not apply to all areas of negligence law, within all areas of life, the moment Carroll Towing was handed down. Though Hand hinted that the formula would apply beyond admiralty,222See United States v. Carroll Towing Co., 159 F.2d 169, 173 (2d Cir. 1947). the case was about when a company is liable for failing to have a bargee on board. So in order to assess whether the formula has the requisite degree of fit with underlying customs, one must consider the customs of the barging business in particular. I will eschew detailed industry research on this point (which would be a poor fit anyway because, as Hand himself pointed out, whether a bargee is needed is highly circumstance-specific)223See id. and instead do a little armchair economics. If barge-owning firms are profit-maximizing, they will put bargees on board if and only if doing so is cost-effective with respect to their own bottom lines. We cannot assume at the outset that firms will follow the Hand formula; the formula is a choice of legal rule, and the legal rule influences firms’ decisions whether to use bargees. In other words, there is an endogeneity issue here. But still, if the formula functions properly, it merely holds barge-owning firms liable when the firms fail to take cost-efficient precautions. The formula just forces firms to internalize certain of their externalities, and it does so in a predictable way. So while endogeneity stops us from doing a pre-legal analysis of barge-business customs, we can guess that determining whether a precaution is cost-efficient—including the cost of externalities the firm must internalize—will be second-nature to profit-maximizing firms. Firms are in the business of calculating cost-efficiency. So as a first guess, the formula is pretty well accommodated to the customs of the barge industry.

 3.     Rowland v. Christian

Rowland v. Christian,224443 P.2d 561 (Cal. 1968). a California decision, is an apt example of judicial lawmaking. Though it was a state case, the court departed starkly enough from the common law to merit attention. It all started when Christian invited Rowland over as a guest to her apartment.225Id. at 562–63. While using the bathroom, Rowland cut his hand badly on Christian’s cold-water faucet, which had been cracked for some time.226Id. Rowland sued, alleging Christian had been negligent by failing either to warn him about the broken faucet or to fix it.227Id. At the time, California common law divided landowners’ (or, as here, possessors’) duties into two broad categories. Landowners, in the first category, owed invitees (business visitors invited or permitted onto the land) a full duty of care.228Id. at 565. But landowners in the second category owed licensees (typically, social guests) and trespassers “only the duty of refraining from wanton or willful injury.”229Id. This is a rough presentation of the common law duties, but it should suffice.

Under the old rule, Christian would likely not have been liable: Rowland was her licensee, and even if Christian had been negligent, her failure to warn or repair was probably neither “wanton [nor] willful.”230Rowland, 443 P.2d at 563, 565. But the court did away with the common-law rule, choosing instead to impose the full duty of care on all landowners.231Id. at 567–68. The court pointed to a statute as part of its justification for this switch (the statute implemented the full standard of care as a blanket rule).232Id. at 563–64, 568 (“Every one [sic] is responsible, not only for the result of his willful acts, but also for an injury occasioned to another by his want of ordinary care or skill in the management of his property or person, except so far as the latter has, willfully or by want of ordinary care, brought the injury upon himself.” (quoting Cal. Civ. Code § 1714 (1872))). But the court could not and did not claim that the statute, which was passed in 1872, superseded all common law deviations from the full duty of care.233Instead, the court took the stance that courts should make “exception[s]” to the full, statutory duty of care only if “clearly supported by public policy.” Id. at 564. This was unavoidably a common-law decision.234If nothing else, it was a common law decision in that, by the court’s own lights, courts get to decide when to apply the statute and when to apply common law rules, based on “policy” considerations. Id.

The Rowland decision broke sharply from the existing common law. More or less, the Rowland court discarded the common-law categories because it saw them as bad, confusing policy. The court did point to California cases criticizing those categories.235Id. at 566–67. It also cited cases where the courts had bent the rules to impose the full standard of care more broadly.236See Rowland, 443 P.2d at 565–66 (discussing a judicial “retreat from” the traditional rule that “a trespasser and licensee or social guest are obliged to take the premises as they find them” and going on to explain the doctrinal minutiae by which the courts had effected this retreat). But the court never argued the categories were not currently part of California’s common law.237See id. Indeed, the court acknowledged a whole category of cases (“trap” cases) where the categories applied. Id. at 566. Instead, the court acknowledged the categories’ status as law but discarded them anyway: “[W]e are satisfied that continued adherence to the common law distinctions can only lead to injustice or, if we are to avoid injustice, further fictions with the resulting complexity and confusion. We decline to follow and perpetuate such rigid classifications.”238See id. at 568. Thus, the Rowland court did “find law” in a sense; it just decided not to follow that law. It even refused to modify the rule in some less drastic way, perhaps by altering the categories without erasing them entirely.239For example, after Rowland was decided, while many states followed suit, many others “abolished the invitee/licensee classification while maintaining the trespasser distinction.” Hall v. Cagle, 773 So. 2d 928, 931–32 (Miss. 2000) (McRae, J., concurring) (collecting cases). Whether such a move would be “continuous” with the existing common law is unclear. On one hand, that would still involve a rejection of a longstanding classification. On the other, it is much less drastic than doing away with the classification scheme entirely. Not only does it keep one of the old distinctions, but in so doing, it avoids the most drastic change of Rowland: suddenly lumping trespassers together with invitees and licensees. Additionally, if the common law is to avoid stagnation, the law must have some way to move away from an old classification scheme without violating the principle of continuity. Perhaps moving away from the old classifications one small step at a time (here, by abolishing the invitee/licensee line and only then, after waiting for reception in the realm of California, considering eliminating the trespasser/non-trespasser distinction) would be the way to effect this change while maintaining continuity.

It is also possible, however, that because the old rule seems not to have featured baked-in exceptions (unlike the English formulation of murder, discussed supra Section II.B.1), there was no way to change it without judicial lawmaking. If California’s constitution does not allow judicial lawmaking, that result seems unappetizing—the courts would have been stuck with the old rule. But that kind of situation is no different from what happens when courts are stuck with written law they dislike: the legislature, and only the legislature, has the power to change the rules. The fact that courts cannot solve all ills is not a reason to embrace judicial lawmaking.
This discontinuity means the Rowland court was not finding customary law. Instead, the court was making law on its own.

Further, post-Rowland developments suggest the case’s holding was not received in the realm of California. In 1985, the California legislature passed a statute abrogating Rowland—specifically, the part of Rowland that had treated trespassing criminals just like invitees and licensees for purposes of landowner liability.240Act of Oct. 2, 1985, ch. 1541, § 1, 1985 Cal. Stat. 5656 (codified at Cal. Civ. Code § 847). Gone was Rowland’s idea of one duty of care for entrants. The new statute provided instead that “[a]n owner . . . of any estate or any other interest in real property . . . shall not be liable to any person for any injury or death that occurs upon that property during the course of or after the commission of any” enumerated felonies.241Id. Owners would be liable only if they engaged in “willful, wanton, or criminal conduct, or . . . willful or malicious failure to guard or warn against” dangers on the property.242Id. 1985 Cal. Stat. at 5657.

Calvillo-Silva v. Home Grocery243968 P.2d 65 (Cal. 1998). further illustrates that California did not receive Rowland’s rule. There, the California Supreme Court considered the duties of landowners under the new statute.244Id. at 69. Unsurprisingly, the statute trumped Rowland.245See id. at 72. More interesting is the Calvillo-Silva court’s account of the statute’s impetus:

In evaluating the matter, the Legislature specifically considered two controversial cases in which plaintiffs had sought substantial sums for injuries they incurred while trespassing on the property of others. In one case involving public property, a plaintiff sued a school district for $3 million after he fell through a skylight during an attempt to illegally remove floodlights from the roof of a school gymnasium. The plaintiff, who was rendered a quadriplegic from the fall, obtained a settlement of $260,000 plus monthly payments of $1,200 for life. In another case, a motorcycle thief who trespassed and went joyriding across a farmer’s field received nearly $500,000 in damages from the farmer for injuries he sustained after hitting a pothole on the field. The bill to enact section 847 was viewed as proposing a partial reversal of [Rowland], which in effect had permitted such lawsuits to be maintained. As noted by various legislative committees, the sentiment providing the impetus for the legislation was reflected in the following statements of the bill’s author: “[W]hatever may be said in defense of the alleged right of a trespasser to sue a landowner for the trespasser’s injuries sustained while trespassing, there is almost nothing to be said on behalf of the thief, a cattle rustler or other felon who is injured in the course of his felony. Such a wrongdoer should not be allowed by the law to add still more injury to insult.”246Id. at 71–72 (alteration in original) (footnote omitted) (citations omitted).

The lesson seems clear: there was a “sentiment”—presumably among the people of California—that Rowland had gone too far.247Id. at 72. More concretely, the statute itself voiced the legislature’s disapproval of (part of) Rowland. In a system where legislative enactments trump common-law rulings, Rowland became a moot point as soon as the legislature spoke up. But the combination of popular and legislative disapproval means that—even aside from legislative preemption—Rowland was never “received” in California. Since a non-received holding does not become customary law, it follows that Rowland did not become part of the customary law of California. Thus (again setting aside discontinuity and legislative preemption), if Rowland became law at all, it was only as judge-made law, not judge-found law.

There are also implications for accommodation. It seems that Rowland’s sweeping protection of criminals struck a nerve. Possibly, the people of California subscribed to and practiced the following custom: as a general rule, we don’t reward people for doing wrong. So if someone gets hurt while committing a serious crime, we won’t feel too bad for him, and we won’t go out of our way to help him. May his blood be upon his own head. This custom need not have been a legal custom at all. It could have been an extra-legal social norm, perhaps manifesting as a reluctance to show sympathy for people injured while committing crimes. Imagine, for instance, an ordinary Californian who is in the habit of organizing fundraisers for her injured relatives. Perhaps she would refrain upon finding out that this time, Uncle Joe got hurt while robbing a bank.

Suppose this was indeed Californians’ prevailing attitude around the time of Rowland. (Whether it was is another empirical question.) If so, the custom could easily explain the public’s dissatisfaction with the case: Californians heard that injured evildoers were successfully suing their would-be victims for damages. This flatly contradicted their no-sympathy-for-injured-criminals custom. Hence the 1985 bill abrogating Rowland. If this speculative account is accurate, then Rowland was ineligible for inclusion in the customary law the moment it came down because it flew in the face of the people’s customs. That is the opposite of accommodation.

For those reasons, I conclude the Rowland court was attempting to make law. Because this was a state-court decision, whether the court had lawmaking power is beyond this Article’s scope—it depends on California’s constitution.248Cal. Const. art. VI, § 1 provides: “The judicial power of this State is vested in the Supreme Court, courts of appeal, and superior courts, all of which are courts of record.. I only claim that the court was not finding law.249Natural law excepted.

 4.     The Sea Gull, The Harrisburg, and Moragne v. States Marine Lines

This Section discusses three admiralty cases decided over the course of 105 years. The central issue in all three was whether the law of admiralty provides a cause of action for wrongful death, in the absence of a statutory cause of action.250The first two cases considered not just American cases, but cases from the maritime law more generally. As the Harrisburg Court put it, the law to be applied was “the general maritime law as administered in the courts of the United States.” The Harrisburg, 119 U.S. 199, 205 (1886), overruled by Moragne v. States Marine Lines, Inc., 398 U.S. 375 (1970). This choice-of-law issue (or perhaps general-law issue) is not directly relevant to this Article, so I leave it to one side. The first case is The Sea Gull,25121 F. Cas. 909 (Chase, Circuit Justice, C.C.D. Md. 1865) (No. 12,578). an 1865 case in the Circuit Court for the District of Maryland. There, Chief Justice Chase (riding circuit) joined a growing group of courts that—contrary to the then-longstanding rule—recognized a new wrongful death cause of action in admiralty.252See id. at 909–10 (discussing cases). Twenty-one years later, in The Harrisburg,253119 U.S. 199 (1886). the Supreme Court took up the same issue and held that The Sea Gull and its ilk were wrong.254See id. at 213–14. And finally, in the 1970 case of Moragne v. States Marine Lines, Inc.,255398 U.S. 375 (1970). the Court explicitly overruled The Harrisburg.256Id. at 378 (“The Harrisburg . . . somewhat dubious even when rendered, is such an unjustifiable anomaly in the present maritime law that it should no longer be followed.”).

We can take away at least three lessons from this series of cases. First, The Sea Gull, and cases like it, were almost certainly impermissible attempts at judicial lawmaking. Second, The Harrisburg was correct to overrule these cases—using vertical stare decisis to reinforce the old rule of admiralty The Sea Gull had tried to get rid of. And third, Moragne’s overruling of The Harrisburg was, just like The Sea Gull, an instance of unconstitutional judicial lawmaking.

Chief Justice Salmon Chase began in The Sea Gull by acknowledging that “[t]here are cases, indeed, in which it has been held that in a suit at law, no redress can be had by the surviving representative for injuries occasioned by the death of one through the wrong of another.”257The Sea Gull, 21 F. Cas. at 910. But Chase did not find this rule controlling because the common law’s “peculiar rules in relation to this subject, traceable to the feudal system and its forfeitures,” had lately been disapproved by several American courts and state legislatures.258Id. Further, refusing a cause of action would contravene “natural equity and the general principles of law.”259Id. (quoting Cutting v. Seabury, 6 F. Cas. 1083, 1084 (D. Mass. 1860) (No. 3,521). If someone is merely injured at sea, a relative has a cause of action against the tortfeasor—so why should that cause of action disappear when the injury is “aggravated by the death of the [victim]”?260Id. (quoting Cutting, 6 F. Cas. at 1084). Thus, Chase reasoned, “it better becomes the humane and liberal character of proceedings in admiralty to give than to withhold the remedy, when not required to withhold it by established and inflexible rules.”261Id. On these grounds, the court refused to follow, in the admiralty context, the common-law rule against wrongful death actions.262Id.

The problem with The Sea Gull is its discontinuity with existing customary law. Chief Justice Morrison Waite’s subsequent opinion in The Harrisburg spelled out this defect. Waite acknowledged Chase’s point that common law and admiralty law are separate bodies of custom—and that their rules may therefore differ.263The Harrisburg, 119 U.S. 199, 205 (1886) (explaining that “[w]e know of no English case in which it has been authoritatively decided that the rule in admiralty differs at all in this particular from that at common law,” but going on to allow that, in theory, “under the general maritime law as administered in the courts of the United States, a contrary rule [could have] been or ought to be established”). But Waite explained that the law of admiralty did not “differ[] at all in this particular from [the rule] at common law,” which did not allow a private cause of action for wrongful death.264Id. at 204–05. Thus, the default rule—even in admiralty cases—allowed no wrongful-death cause of action. Thus, Waite went on to ask whether American admiralty law had altered that default rule.265Id.

Then something quite interesting happened. Waite acknowledged that cases like The Sea Gull had attempted to alter the rule—but in his view, they had failed to do so.266See id. at 205–13. The problem, he said, was:

The argument everywhere in support of such suits in admiralty has been, not that the maritime law, as actually administered in common law countries, is different from the common law in this particular, but that the common law is not founded on good reason, and is contrary to “natural equity and the general principles of law.”267Id. at 213.

The Harrisburg Court got at the heart of the issue: cases like The Sea Gull were arriving at their answer, not because the common law demanded that answer, but because they disliked what the common law had to say. In this respect, The Sea Gull is similar to the much later Rowland case in California: both cases apprehended the common-law rule, and both cases chose not to apply that rule because of policy concerns. And deviating from the common law on the basis of policy, explained the Court, is impermissible:

Since, however, it is now established that in the courts of the United States no action at law can be maintained for such a wrong in the absence of a statute giving the right, and it has not been shown that the maritime law, as accepted and received by maritime nations generally, has established a different rule for the government of the courts of admiralty from those which govern courts of law in matters of this kind, we are forced to the conclusion that no such action will lie in the courts of the United States under the general maritime law. The rights of persons in this particular under the maritime law of this country are not different from those under the common law, and as it is the duty of courts to declare the law, not to make it, we cannot change this rule.268Id. at 213–14 (emphasis added).

This is very close to an outright pronouncement that courts violate the separation of powers by attempting to drastically change common-law rules. Note that the Court did not accuse other courts, like the Sea Gull court, of lawmaking. This may be simple politeness, or perhaps it is because the state of the common-law rule was somewhat uncertain at the time of The Sea Gull.269See The Sea Gull, 21 F. Cas. 909, 910 (Chase, Circuit Justice, C.C.D. Md. 1865) (No. 12,578) (engaging in discussion that might be construed as casting doubt on the state of the common-law rule). The Supreme Court, in fact, issued an opinion in the years between The Sea Gull and The Harrisburgclarifying that there was no wrongful-death cause of action at common law.270Ins. Co. v. Brame, 95 U.S. 754, 756–57, 759 (1878). Yet the only reason the common-law rule was in doubt in the first place is because courts, as in The Sea Gull, had been mistakenly assuming the power to discard that rule simply because it made for bad policy. The Sea Gull was probably an unconstitutional attempt at judicial lawmaking, as The Harrisburg suggested in the course of reaffirming the old rule.

The 1970 case Moragne v. States Marine Lines, Inc. overruled The Harrisburg.271Moragne v. States Marine Lines, Inc., 398 U.S. 375, 409 (1970). Because Moragne was completely discontinuous with the rule of admiralty expounded in The Harrisburg, the Court’s decision in Moragne amounted to unconstitutional lawmaking. Justice John Marshall Harlan’s opinion gave two main reasons for its rejection of the Harrisburg rule, but my view is that neither justifies the Court’s deviation from the customary law.272Another proffered reason was that, even if the common-law rule was to stay binding on land:

[I]t is not apparent why the Court in The Harrisburg concluded that there should not be a different rule for admiralty from that applied at common law. Maritime law had always . . . been a thing apart from the common law. It was, to a large extent, administered by different courts; it owed a much greater debt to the civil law; and, from its focus on a particular subject matter, it developed general principles unknown to the common law. These principles included a special solicitude for the welfare of those men who undertook to venture upon hazardous and unpredictable sea voyages.

Id. at 386–87 (footnote omitted). This argument distorts The Harrisburg’s reasoning. As discussed above, Justice Waite was not arguing that the two bodies of law are always identical. Rather, he argued that the rule of admiralty law on this issue had (as a matter of contingent, not necessary, fact) long been identical to the common-law rule. Because the common-law rule allowed no cause of action, and because the admiralty rule had not changed such that it deviated from its default of agreeing with the common law, the admiralty rule remained in its default position. The Harrisburg, 119 U.S. at 213–14.

First, Harlan argued that The Harrisburg’s rule made no sense and was inadequately justified, even when it was adopted.273Moragne, 398 U.S. at 378–81. This was partly a policy-based argument. Harlan claimed that there was simply no good reason for the Harrisburg Court (or the Insurance Co. v. Brame27495 U.S. 754 (1878).Court before it, in the common-law context) to adopt such a counterintuitive rule.275Moragne, 398 U.S. at 378–81, 386–88. In his words, the common-law rule “was based on a particular set of factors that had, when The Harrisburg was decided, long since been thrown into discard even in England, and that had never existed in this country at all.”276Id. at 381. The rule was based on the doctrine of felony merger, an English doctrine (never accepted in America) that disallowed wrongful death actions because wrongful death merged into the felony of homicide.277Id. at 382 (explaining the felony-merger doctrine). And Harlan pointed to the maxim that “while ‘[o]ur ancestors brought with them [the] general principles [of the common law] and claimed it as their birthright; . . . they brought with them and adopted only that portion which was applicable to their situation.’”278Id. at 386 (alterations in original) (quoting Van Ness v. Pacard, 27 U.S. (2 Pet.) 137, 144 (1829)). The Harrisburg’s rule, Harlan argued, violated this maxim and was consequently both unjust and unreasonable.279For the full ambit of Harlan’s policy arguments, see id. at 379–86.

This line of argument, however, runs into the same objection the Harrisburg Court gave to The Seagull: policy considerations are not enough to justify outright rejection of a settled rule of common law.280See The Harrisburg, 119 U.S. 199, 213–14 (1886). And even if The Harrisburg’s rule was unjustified at the time of its adoption (perhaps, in Hale’s terms, it was poorly accommodated to the underlying customs of the people), the rule had lain undisturbed for over eighty years between The Harrisburg and Moragne. This means that the proper course would have been treating the rule as settled—even if it was wrong when it was handed down. This was Justice Marshall’s approach in Livingston, and it was Justice Waite’s approach in The Harrisburg itself. A rule that has enjoyed long and uninterrupted acceptance is binding, even if courts never should have adopted it in the first instance.281The Moragneopinion did go on to ask whether it should uphold The Harrisburg “simply as a matter of stare decisis.” Moragne, 398 U.S. at 403. Unsurprisingly, this discussion focused on three familiar, pragmatic principles: first, “the desirability that the law furnish a clear guide for the conduct of individuals, to enable them to plan their affairs with assurance against untoward surprise;” second, “the importance of furthering fair and expeditious adjudication by eliminating the need to relitigate every relevant proposition in every case;” and third, “the necessity of maintaining public faith in the judiciary as a source of impersonal and reasoned judgments.” Id. What the Court’s discussion of stare decisis did not do, however, is take seriously the possibility that the common law is really and truly binding law—law that courts must follow and that they cannot discard on pragmatic grounds. See id. at 403–05. The Moragne Court’s first criticism of The Harrisburg does not work.

Second and independently, Harlan criticized The Harrisburg as “an . . . anomaly in the present maritime law.”282Id. at 378. Harlan argued that, even if The Harrisburg was correctly decided, “the wholesale abandonment of the rule [against wrongful-death causes of action] in most of the areas where it once held sway” had since made the rule obsolete.283Id. at 388. Harlan pointed to both judicial and legislative shifts in favor of wrongful-death causes of action. The English House of Lords, for example, de facto created a cause of action for wrongful death in 1937.284See id. at 388–89 (citing Rose v. Ford [1937] AC 826 (UK)). And by the time of Moragne, “every State [had] enacted a wrongful-death statute.”285Id. at 390. Likewise, Congress had by that time created wrongful-death liability in many circumstances, and some legislators had sharply criticized the rule of The Harrisburgas “a disgrace” and out of line with the law of “enlightened nations.”286Id. at 397 (quoting H.R. Rep. No. 66-674, at 4 (1920)); see also id. at 390 (first citing Jones Act, Pub. L. No. 66-261, 41 Stat. 988 (1920) (wrongful death cause of action for merchant seamen); then citing Death on the High Seas Act, Pub. L. No. 66-165, 41 Stat. 537 (wrongful death cause of action for certain deaths on the high seas); and then citing Federal Tort Claims Act, Pub. L. No. 79-601, 60 Stat. 842 (1946) (subjecting the United States to wrongful death liability in certain circumstances)). The death at issue in Moragne did not fit within the Death on the High Seas Act because that Act dealt only with deaths “occurring on the high seas beyond a marine league from the shore of any State.” § 1, 41 Stat. at 537. The death in Moragne occurred “in navigable waters within the State of Florida.” Moragne, 398 U.S. at 376.

Harlan acknowledged that none of these decisions or statutes applied to the case at hand but nonetheless concluded that “[t]hese numerous and broadly applicable statutes, taken as a whole, make it clear that there is no present public policy against allowing recovery for wrongful death.”287Moragne, 398 U.S. at 390. The idea is that, in his words, the general policy in favor of wrongful-death liability “has become itself a part of our law, to be given its appropriate weight not only in matters of statutory construction but also in those of decisional law.”288Id. at 390–91; see also id. at 392 (arguing that common-law courts have long “perceive[d] the impact of major legislative innovations and . . . [interwoven] the new legislative policies with the inherited body of common-law principles”). So these developments made the wrongful-death cause of action “the general rule of American law, and its denial the exception.”289Id. at 393.

Perhaps we can construe this as an argument that the rule of The Harrisburg lost its binding force because, as evidenced by later judicial and legislative developments in related areas, the rule had not been “received in the realm.” If that is Harlan’s argument, it fails for two reasons. The first and most obvious is that the rule of The Harrisburg was received in the realm, before The Harrisburg was even decided. That was, after all, the whole thrust of the Court’s reasoning in that case: the longstanding rule forbade wrongful-death causes of action in admiralty (as at common law), and that rule was a binding part of the law of admiralty.290The Harrisburg, 119 U.S. 199, 204–05, 213–14 (1886). Justice Harlan subtly evaded this issue by pointing out criticisms that arose after The Harrisburg—as if The Harrisburg had been laying down a new rule that a broad judicial and legislative trend then rejected in favor of wrongful-death recovery.291Moragne, 398 U.S. at 388–93. The one exception is Lord Campbell’s Act 1846, 9 & 10 Vict. c. 93, § 1 (Eng.), which partially abrogated the standard common-law rule. But The Harrisburg was merely defending and clarifying an old rule, not creating a new one. And because the old rule had been broadly received before The Harrisburg, Harlan’s argument misses the mark.

The other problem, related to the first, is that none of the cited judicial decisions or legislative enactments governed the case before the Court. Congress undisputedly could have created a wrongful-death cause of action for deaths, like that in Moragne, on waters within three miles of shore. But Congress did not do so. The Moragne opinion acknowledges this simple reality and goes to great lengths to avoid it.292See Moragne, 398 U.S. at 392–402. Harlan’s view, ultimately, seems to be that legislative enactments can indirectly change a particular rule of customary law by directly abrogating other rules of customary law.293“Customary law” here refers to any kind of customary law, including both common law and admiralty law. But this is a distortion of the reception principle. The reception inquiry asks whether a putative rule of customary law has been, as an initial matter, received in the realm. Once received, the rule becomes part of the customary law—as long as it is also continuous with prior customary law and accommodated to the customs of the people.294See supra Section II.A. It does not follow that a rule that is already a part of the customary law loses its legal force because of related, but not controlling, legislative enactments.

In Moragne, the Court correctly identified the governing rule of admiralty law (no cause of action for wrongful death) but refused to follow it. The Court could have waited for Congress to change the settled rule. But it decided not to. This was judicial lawmaking: the Moragne Court’s new rule was entirely discontinuous with the rule recognized in The Harrisburg. And the Court’s two justifications for this move boil down to a policy argument and a kind of reception argument. The former fails because judicial lawmaking is unconstitutional even when it would make for good policy. The latter fails mainly because Congress’s statutes—none of which actually governed the case at hand—did not undo the binding effect of The Harrisburg’s rule, which had already been a well-established part of admiralty law by the time Congress started to legislate on related issues. The best way to understand Moragne is as an instance of unconstitutional judicial lawmaking.

III.     Implications for the Court

The Supreme Court and its Justices are often unclear about Article III lawmaking. In pure common-law cases, even the most ardent separation-of-powers hawks sometimes assume the Court can make law.295See, e.g., Boyle v. United Techs. Corp., 487 U.S. 500, 504 (1988) (holding that, in areas of federal preemption and where no “explicit statutory directive” governs an issue, the Court must “prescribe[]” the content of governing law). Other times, Justices disapprove of that practice—but their discomfort lacks a basis in theory.296See, e.g., id. at 516 (Brennan, J., dissenting). Justice Brennan contended that “this Court lacks both authority and expertise to fashion [the majority’s] rule.” Id. And he concluded with gusto: “I would leave that exercise of legislative power to Congress, where our Constitution places it.” Id. Quite right, but this may have been more rhetoric than anything. Brennan’s dissent focused on the scope of federal common-law adjudication without seriously critiquing the majority’s methodologicalassumption: that the courts can make (“prescribe[]”) common law. See id. More perplexing still are the cases where written and unwritten law interact. When a statute seems to condone or even solicit judicial lawmaking, should courts accept Congress’s invitation?297See, e.g., 15 U.S.C. § 1; Andrew S. Oldham, Sherman’s March (in)to the Sea, 74 Tenn. L. Rev. 319, 367–79 (2007) (complaining that this broad text, which courts construe as a license to make common law, violates the nondelegation doctrine). When an otherwise-mandatory statute gives courts no orders on an issue, may they make law to fill the gap?298See infra Section III.B.

A comprehensive treatment of these questions is beyond this Article’s scope. So this Part zooms in on habeas corpus for the sake of illustration. My purpose here is not to condemn the Court’s doctrines, nor even to claim definitively that the Court made law in any given case. I merely want to highlight and discuss some areas of theoretical confusion. Lawfinding might be a first step toward clarity on issues that have vexed the Court for decades.

A.     Edwards v. Vannoy and Justice Gorsuch’s Critique of Brown v. Allen

1.     Edwards and the Basics of Retroactivity

In 2007, a non-unanimous Louisiana jury convicted Thedrick Edwards of several serious felonies.299Edwards v. Vannoy, 141 S. Ct. 1547, 1552–53 (2021). His conviction became final in 2011.300Id. at 1553; id. at 1568 & n.2 (Gorsuch, J., concurring) (“Everyone accepts that, in our criminal justice system today, a judgment becomes final only after the completion of a trial and the appellate process, including the opportunity to seek certiorari from this Court on questions of federal law.”). At that time, Supreme Court precedent allowed non-unanimous verdicts in state criminal proceedings.301Id. at 1551 (majority opinion) (citing Apodaca v. Oregon, 406 U.S. 404 (1972)). But in the 2020 case Ramos v. Louisiana,302140 S. Ct. 1390 (2020). the Court overturned its precedent and held the Sixth Amendment requires jury unanimity across the board.303Id. at 1397. So by the time the Court (in 2021) decided Edwards v. Vannoy,304141 S. Ct. 1547 (2021). the requirement of jury unanimity was well established. Nevertheless, the Court denied Edwards relief.305Id. at 1562.

To see how that could happen, compare Evangelisto Ramos and Thedrick Edwards. They had each allegedly committed serious felonies. A jury convicted each of them.306Ramos, 140 S. Ct. at 1393–94; Edwards, 141 S. Ct. at 1553. And in each case, the jury’s conviction verdict was nonunanimous.307Ramos, 140 S. Ct. at 1394; Edwards, 141 S. Ct. at 1553. Despite those similarities, Ramos and Edwards had one key difference: timing. Ramos was a direct-review case—Ramos appealed straight from the Supreme Court of Louisiana.308See Ramos, 140 S. Ct. at 1394–95, 1406. That means his conviction was not final when the Court decided his case, so no obstacle stood between him and the benefit of the new unanimity rule (the “Ramos rule”).309See supra note 300. In contrast, Edwards’s opportunity to appeal his conviction directly to the Supreme Court came and went in 2011.310Edwards, 141 S. Ct. at 1553. That meant Edwards was a collateral-review case—he was asking the Supreme Court to undo his already-final conviction by way of habeas corpus.311Id. Accordingly, Edwards could get the benefit of the Ramos rule only if the Court decided to apply the rule retroactively to petitioners, like him, whose convictions had already become final when the Court decided Ramos.

Unfortunately for Edwards, the Court generally disallows retroactivity. In Teague v. Lane,312489 U.S. 288 (1989). the Court held: “Unless they fall within an exception to the general rule, new constitutional rules of criminal procedure will not be applicable to those cases which have become final before the new rules are announced.”313Id. at 310 (plurality opinion). That means most habeas petitioners do not get the benefit of post-finality changes in law.

Teague allowed for two exceptions to the no-retroactivity rule.314Id. at 311–12. One of those was for “watershed rules of criminal procedure”—idea being that some new rules are so critical they simply must apply retroactively.315Id. Given the obvious significance of Ramos’s unanimity rule, Edwards urged the Court to dub that rule “watershed,” hold that it therefore applies retroactively, and apply it to his case.316See Edwards, 141 S. Ct. at 1553–54. The Court refused to do so. It instead held that the watershed exception (which it had never actually used since Teague) was dead, concluded that the Ramos rule was not retroactive, and affirmed the Fifth Circuit’s denial of Edwards’s petition.317Id. at 1560, 1562.

 2.     Brown v. Allen as the Source of All Ills

Now we turn to Justice Neil Gorsuch’s concurrence in Edwards, joined by Justice Clarence Thomas.318In Brown v. Davenport, 142 S. Ct. 1510 (2022), a full six Justices signed onto a Gorsuch opinion expressing much the same view as the Edwards concurrence. I have decided to focus on the Edwards concurrence rather than the Davenport majority opinion for two reasons. First, Edwards came first. Davenport does not showcase any great shift in Justice Gorsuch’s stance on these issues; it is significant because a majority of the Court endorsed that stance for the first time. And second, Davenport’s historical discussion seems to equivocate between discussing the common law and discussing statutory interpretation. Compare id. at 1520–21 (casting “the common-law rule” as a “limit[]” on the “writ” itself), with id. at 1521 (seeming to cast the very same issue as one of statutory interpretation). That makes Davenportunwieldy and rather hard to work with. Granted, Justice Gorsuch was not crystal-clear on this point in Edwards, either, but it seems marginally clearer here than in Davenport that his concurrence concerned the common law. See infra note 332.

This problem is by no means unique to Justice Gorsuch. Opinions and articles are chronically unclear about whether Brown was about the common law or statutory interpretation. I intend to address the issue soon in another article, which is currently a work in progress.
The pair made three main claims. First, habeas corpus doctrine is a jurisprudential disaster.319See Edwards, 141 S. Ct. at 1571–72 (Gorsuch, J., concurring). Second, discarding the watershed exception was a step toward mitigating that disaster.320See id. at 1573. Third and most important, the Court’s 1953 decision in Brown v. Allen321344 U.S. 443 (1953). is to blame for the entire mess.322See Edwards, 141 S. Ct. at 1568–70 (Gorsuch, J., concurring). The concurrence gave these three propositions in a different order.

Justice Gorsuch began his critique of Brown by explaining the traditional scope of the habeas writ.323Id. at 1566–67. At common law, a court could grant habeas relief only if either the executive was detaining someone who had not been convicted of a crime, or the convicting court had lacked jurisdiction.324Id. at 1567 (framing this as a common-law inquiry). But in 1915, in an attempt to curb mob-dominated and racism-influenced trials, the Court bent that rule by suggesting that “a state court’s extreme departure from ‘established modes’ of criminal trial practice, such as proceeding under the specter of mob violence, might be akin to the loss of ‘jurisdiction,’ at least if no corrective mechanism like an appeal existed.”325Id. at 1568 (quoting Frank v. Mangum, 237 U.S. 309, 326 (1915)); see also Wright v. West, 505 U.S. 277, 285 (1992) (plurality opinion) (explaining more fully the changes leading up to Brown).

Brown went beyond merely blurring the line between jurisdictional and nonjurisdictional defects. The Brown Court discarded the old common-law rule altogether.326Edwards, 141 S. Ct. at 1568 (Gorsuch, J., concurring); Brown, 344 U.S. at 464–65, 478–79 (addressing the merits of a constitutional challenge to a conviction without holding the state court lacked jurisdiction to convict). In Gorsuch’s words, this “effectively recast habeas as another way for federal courts to redress practically any error of federal law they might find in state court proceedings,” rather than only jurisdictional errors.327Edwards, 141 S. Ct. at 1568 (Gorsuch, J., concurring). Gorsuch went on to bemoan that the wrongly decided Brown suddenly “upended centuries of settled precedent,” “invited practical problems,” and “produced anomalies.”328Id. at 1569; see also infra note 332. This radical shift created the need for the Court’s retroactivity doctrine—a doctrine that has by all accounts become bewilderingly complex.329Edwards, 141 S. Ct. at 1570–72 (Gorsuch, J., concurring). Thus, the Court’s holding in Edwards was a regrettably necessary step toward undoing the problems Brown created.330Id. at 1572–73.

Justice Gorsuch’s argument sounded in common law.331See id. at 1566–67. Justice Gorsuch arguably framed this as a question of statutory interpretation once or twice. See id. at 1567–68 (“When called upon to interpret that statute, this Court defined the scope of habeas review by looking ‘to the common law.’” (quoting Ex parte Bollman, 8 U.S. (4 Cranch) 75, 93–94 (1807)). But Gorsuch did not analyze the statute’s text; instead, he focused on the common-law rule. See Steve Vladeck, Hiding Behind Habeas’s Hardness, JOTWELL (Oct. 6, 2021), Jonathan R. Siegel, Habeas, History, and Hermeneutics, 64 Ariz. L. Rev. 505 (2022)) (comment, Jonathan Siegel) (“Justice Gorsuch’s assertion that ‘[t]he writ of habeas corpus does not authorize federal courts to reopen a judgment issued by a court of competent jurisdiction once it has become final’ is an assertion about habeas tradition, not a textualist analysis of the habeas statute.” (alteration in original) (quoting Edwards, 141 S. Ct. at 1573 (Gorsuch, J., concurring))). His main claim was that, contra Brown, “[t]he writ of habeas corpus does not authorize federal courts to reopen a judgment issued by a court of competent jurisdiction once it has become final.”332Edwards, 141 S. Ct. at 1573 (Gorsuch, J., concurring). Justice Gorsuch never explicitly called for the Court to overturn Brown, but the above-quoted language directly contradicts Brown’s central holding. That is a statement about “the writ” itself—about what that writ does and does not “authorize” courts to do.333See id. In support, Gorsuch emphasized that Brown’s holding came entirely out of left field.334Id. at 1568–69. That does not count Justice Gorsuch’s policy-like points about “practical problems” and “anomalies.” See id. Given his well-known formalism, it seems safe to say he intended those points to be rhetorical rather than strictly legal. From time immemorial through 1952, the common law said habeas corpus extended only to cases involving jurisdictional defects. In 1953, after mucking about for a while with the jurisdictional versus nonjurisdictional distinction, the Court undid that ancient rule by fiat.

Suppose that empirical claim is correct: Brown suddenly reversed an ancient common-law rule.335See id. at 1568–69. Gorsuch’s telling of the history is the minority view. Many law review pages are filled with scholarly “debate” about just this historical issue. The scare quotes are warranted because hardly any of those articles take Gorsuch’s side and argue that Brown was a drastic deviation from historical practice—much less that it was actually wrong. See, e.g., Alan Clarke, Habeas Corpus: The Historical Debate, 14 N.Y.L. Sch. J. Hum. Rts. 375, 433–34 (1998) (stating, in conclusion, that “[t]hose who would eviscerate modern habeas corpus rely on a distorted view of that history to buttress their case”); see also id. at 376 n.3 (discussing the opposing side, but citing only four sources—two of which do not even go so far as to say Brown and its ilk were wrongly decided). For what it is worth, I do not find this scholarly consensus very convincing. But that is not the point. The point is that both sides of the debate would be better off with more theoretical clarity. We might wonder, “so what?” Without a theoretical framework that forbids sudden reversals of ancient rules, it is hard to see the problem with Brown. Indeed, if the Court can make common law, perhaps we should applaud the Brown Court for taking courageous legislative action in the face of changed circumstances. That is what good legislatures do, after all.336See Donald Chisum, In Defense of Modern Federal Habeas Corpus for State Prisoners, 21 DePaul L. Rev. 682, 685 (1972) (concluding that the proto-Gorsuch view of the history is essentially right but correctly noting: “That the Court departed from history in Brown v. Allen, of course, does not inevitably mean that it did so unwisely or without justification”).

This Article provides the missing theory. If federal courts cannot make common law, and if Hale’s lawfinding method is sound, Justice Gorsuch’s argument was essentially about continuity. Brown’s sudden deviation from the longstanding rule, on this view, was an unconstitutional attempt at Article III lawmaking. After years of muddying the distinction between jurisdictional and nonjurisdictional defects, the Court got rid of the traditional rule and its pesky line drawing altogether: find the common law, dislike the common law, and make your own common law. That is judicial legislation in exactly the same sense Rowland was.337If that is the case, maybe Teague was nothing more than a way of walking back Brown. That would mean Teague’s bar is not a judge-made doctrine at all and is instead a half-hearted—but likely permissible—attempt to undo Brown’s judicial lawmaking. Cf. Edwards, 141 S. Ct. at 1573 (arguing Brown created the need for Teague).

Maybe. Again, I am not claiming Brown actually was unconstitutional lawmaking. That would require decisive agreement with Justice Gorsuch’s historical analysis—a topic far beyond this Article’s scope.338It would also require accepting his initial premise—that Brown really was a case about the common law of habeas, rather than a case about statutory interpretation. In that case, the two competing sides are not, or should not be, arguing about judicial lawmaking at all. See supra note 318. And of course, even a wrongly decided Brown may have eventually become part of the common law by way of reception in the realm.339See Livingston v. Jefferson, 15 F. Cas. 660, 664–65 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,441) (Marshall, Circuit Justice). That being said, it is also possible that subsequent judicial decisions and statutory enactments limiting Brown’s scope, when considered in tandem, suggest disapproval of the decision. Regardless of which side is right, however, it would be helpful to begin with a clear understanding of what the debate is about: judicial lawmaking.

B.     Justice Scalia, Judge-Made Habeas Bars, and Brown v. Davenport

What happens when a statute empowers courts to grant habeas relief (i.e., for present purposes, to grant a writ of habeas corpus)—but does not mandate that they do so? 28 U.S.C. § 2241(a) does just that. It provides: “Writs of habeas corpus may be granted by the Supreme Court, any justice thereof, the district courts and any circuit judge within their respective jurisdictions.”34028 U.S.C. § 2241(a) (emphasis added). Despite the overwhelming number of provisions restricting federal courts’ power to grant the writ, not a single provision commands them to do so. Every one speaks in the negative.341See, e.g., Brief for Jonathan F. Mitchell and Adam K. Mortara as Amici Curiae in Support of Neither Party at *13 n.7, Edwards v. Vannoy, 141 S. Ct. 1447 (2021) (No. 19-5807) (explaining that habeas statutes, to this day, “erect[] necessary but not sufficient conditions for issuing the writ”); 28 U.S.C. § 2241(c) (“The writ of habeas corpus shall not extend to a prisoner unless . . . .” (emphasis added)); 28 U.S.C. § 2241(e)(1) (“No court, justice, or judge shall have jurisdiction to hear or consider an application for a writ of habeas corpus filed by or on behalf of” certain aliens. (emphasis added)); see also, e.g., 28 U.S.C. § 2254(a)–(b)(1), (d).

Meanwhile—generally without referencing that textual quirk—the Court has imposed various nonstatutory limits on the writ’s availability.342See infra Section III.A for an explanation of the broader changes in habeas doctrine that, perhaps, created the need for the Teague bar. And of course, one of those is retroactivity doctrine. The TeagueCourt justified its holding on functional grounds, casting its no-retroactivity rule as a way to balance “interests of comity and finality” against those of error-reduction and fairness.343See Teague v. Lane, 489 U.S. 288, 308, 312 (1989) (plurality opinion). Beyond those policy-type concerns, however, the Court offered little explanation for how exactly its rule was rooted in law, written or unwritten.

Justice Antonin Scalia joined the Teague majority opinion. But in a related case a few years later, he wrote separately to expound a more principled basis for Teague (and for other habeas bars).344Withrow v. Williams, 507 U.S. 680, 717–18 (1993) (Scalia, J., concurring in part and dissenting in part); see also infra note 366 and accompanying text (describing the other bars). Scalia, joined by Justice Thomas, gave two main justifications.345Justice O’Connor, Teague’s author, also wrote separately and advanced a similar justification. Withrow, 507 U.S. at 697–714 (O’Connor, J., concurring in part and dissenting in part). I focus on Scalia’s approach here because it eventually gave rise to the Court’s current framing of habeas bars. See Davenport v. Brown, 142 S. Ct. 1510, 1523–24 (2022) (echoing Scalia).

Note that this case was mainly about the doctrine of Stone v. Powell, 428 U.S. 465 (1976), which limits the availability of habeas relief for Fourth Amendment violations. See Withrow, 507 U.S. at 718–24 (Scalia, J., concurring in part and dissenting in part).
First, he claimed the statutes’ permissive voice leaves room for courts to employ “equitable discretion” when considering whether to grant the writ.346Withrow, 507 U.S. at 716–18 (Scalia, J., concurring in part and dissenting in part). The statutes have changed significantly since 1993, but as described just above, they still retain their permissive form. Second, he pointed to several historical sources that established habeas’s status as a “prerogative writ” and that courts sometimes retained discretion not to issue the writ.347Id. (“The federal habeas statute did not ‘deprive the court of discretion,’ which ‘should be exercised in the light of the relations existing, under our system of government, between the judicial tribunals of the Union and of the States.’” (quoting Ex Parte Royall, 117 U.S. 241, 251 (1886))).

Justice Scalia also pointed to a statutory provision directing courts to “dispose of the [habeas] matter as law and justice require.” Id. at 716 (quoting 28 U.S.C. § 2243); see also Danforth v. Minnesota, 552 U.S. 264, 278 (2008) (majority opinion framing Teague as a matter of interpretation and referencing this provision); Davenport, 142 S. Ct. at 1523–24 (also referencing this provision). The question of how that statute interacts with the unwritten law of habeas is for another day. But I strongly suspect Scalia, followed by the Davenport Court, overread the amount of freedom the statute confers on courts. I also suspect he overread the cases which (he claimed) supported his capacious reading. See Royall, 117 U.S. at 251–53 (acknowledging the discretion inherent in “law and justice,” but going on to find a rule implicit in the constitutional structure of federalism rather than treating the statute as a license to make a habeas bar of its own); Ex parte Watkins, 28 U.S. 193, 201–02 (1830) (acknowledging the permissive nature of the statute but going on to find a longstanding rule of common-law habeas corpus rather than making a rule).

On this view, courts may not grant the writ more freely than allowed by statute. But they may create bars to relief that restrict the writ more than the statutes do.348See Withrow, 507 U.S. at 716–18 (Scalia, J., concurring in part and dissenting in part). As for Teague’s seemingly policy-based justifications, they were really “equitable principles”—and those principles justified the Court’s decision to erect a non-statutory bar against habeas relief.349Id. at 716–20.

For a while, Justices Thomas and Gorsuch were this theory’s only vocal advocates on the Court.350See Edwards v. Vannoy, 141 S. Ct. 1547, 1570 (2021) (Gorsuch, J., concurring); see also id. at 1566 n.4 (Thomas, J., concurring). In Brown v. Davenport,351142 S. Ct. 1510 (2022). however, a majority endorsed Scalia’s account with enthusiasm. This development added little in the way of justifying unwritten habeas bars. The Court’s discussion was more or less a rehash of Scalia’s original arguments on this score.352See id. at 1523–24. But it seems that Davenport increased the significance of unwritten habeas bars. Over a dissent, the majority held that—even in a case where statutory requirements and unwritten habeas bars impose fairly similar requirements—a petitioner must satisfy both sets of requirements to get habeas relief.353Id. at 1524. In the Court’s words: “Today, then, a federal court must deny relief to a state habeas petitioner who fails to satisfy either this Court’s equitable precedents or [the habeas statutes]. But to grant relief, a court must find that the petitioner has cleared both tests.”354Id. This move arguably overruled then-existing precedents, which had suggested the two sets of requirements were not truly independent of one another.355See id. at 1535–37 (Kagan, J., dissenting) (arguing the majority’s view was at odds with both Fry v. Pliler, 551 U.S. 112 (2007), and Davis v. Ayala, 576 U.S. 257 (2015)). But see id. at 1526–28 (majority opinion).

For those who believe federal courts have the power to make unwritten law, the above likely makes perfect sense. And Justice Scalia—the main progenitor of these unwritten habeas bars—believed exactly that.356See Antonin Scalia, A Matter of Interpretation 10 (Amy Gutmann ed., 1997) (“[Madison] wrote in an era when the prevailing image of the common law was that of a preexisting body of rules, uniform throughout the nation (rather than different from state to state), that judges merely ‘discovered’ rather than created. It is only in this century, with the rise of legal realism, that we came to acknowledge that judges in fact ‘make’ the common law, and that each state has its own.”); James B. Beam Distilling Co. v. Georgia, 501 U.S. 529, 549 (1991) (Scalia, J., concurring in the judgment) (“I am not so naive (nor do I think our forebears were) as to be unaware that judges in a real sense ‘make’ law.”). But if federal courts cannot make law, things start to look more suspicious. Start with Scalia’s bedrock assumption: federal courts have equitablediscretion in habeas matters. That is a counterintuitive place to start. Habeas is a common-law writ.357See William F. Duker, The English Origins of the Writ of Habeas Corpus: A Peculiar Path to Fame, 53 N.Y.U. L. Rev. 983, 983 (1978). And even though equitable courts could grant habeas relief in England, the English common-law courts had come to dominate habeas practice long before the Founding.358See id. (discussing the use of the writ by “the common law courts in the fifteenth, sixteenth, and seventeenth centuries to defeat the encroaching jurisdictions of the equity and ecclesiastical courts and of the various executive councils” (emphases added)). But see Erica Hashimoto, Reclaiming the Equitable Heritage of Habeas, 108 Nw. U. L. Rev. 139 (2014). Hashimoto says habeas is de facto equitable. See id. at 139–42 (arguing that this justifies some habeas bars, but not Teague); id. at 143–44 (citing Paul D. Halliday, Habeas Corpus: From England to Empire 87 (2010) for historical support). The idea is that the King’s Bench, a common-law court, often treated the writ as if it were equitable—even if it was in fact a common-law writ. See id. Hashimoto infers that contemporary American courts can and should treat habeas as a matter of equity. See id.

Assume arguendo that English courts treated habeas more flexibly than other common-law writs. And assume early American courts also treated habeas as de facto equitable—a point Hashimoto assumes but does not argue for. Even on those assumptions, it does not follow that equitable rules govern habeas. The stubborn fact is that habeas corpus came to this country as part of the common law. See, e.g., Ex parte Bollman, 8 U.S. 75, 93–94 (1807) (opening the opinion by explaining that “for the meaning of the term habeas corpus, resort may unquestionably be had to the common law”); id. at 79–80 (reporter’s notes) (Bollman’s counsel pointing out that all superior courts in England had power to issue the writ, but noting the power came from common law and not mentioning equity); Ex Parte Watkins, 28 U.S. 193, 202 (1830) (explicating habeas as a creature of the common law, with no reference to equity, and explaining that “the celebrated habeas corpus act of the 31st of Charles II . . . . enforces the common law”). No amount of King’s Bench rule-fudging changes that, and it follows that the ordinary rules of common-law change—not the rules of equitable change—govern habeas. To illustrate, consider this analogy: The English courts blurred the line between local and transitory actions. That was probably judicial lawmaking, but those mistakes have been in place for so long that we must now live with them. Yet we are not free to carry the lawmaking still further. See Livingston v. Jefferson, 15 F. Cas. 660, 664 (Marshall, Circuit Justice, C.C.D. Va. 1811) (No. 8,411) (Marshall, Circuit Justice). Similarly, grant the King’s Bench fudged the rules of habeas by granting relief more broadly and flexibly than ordinary common-law rules really allowed. If so, we may have to shrug and accept the now-longstanding doctrinal errors of King’s Bench. That means accepting that the rules they made are part of today’s common law of habeas. Thus, Scalia and Hashimoto would have a stronger argument if they could show the Court’s habeas bars themselves were in fact old rules. They do not claim to do so; instead, they borrow principles from other areas of equity and import them into habeas. See Hashimoto, supra, at 151. What we may not do is throw up our hands, dub habeas close-enough-to-equitable, and go about making new law—our own rules—as we see fit.

Justice Scalia also noted that habeas was originally a prerogative writ. See Withrow v. Williams, 507 U.S. 680, 716–17 (Scalia, J., concurring in part and dissenting in part). True, “[i]n English law, ‘prerogative writ’ was the name given to certain judicial writs issued by the courts only upon proper cause shown, but never as a mere right.” Duker, supra note 357, at 984 n.2. But it does not follow that today’s courts retain broad discretion to refuse the writ without a basis in law. Prerogatives are matters of royal power—and the writ of habeas corpus is significant precisely because it changed, over many years, from a tool of royal power into a means of restraining royal power. See id. at 1015. Again, strong historical claims about habeas are beyond this Article’s scope. But it is hard to see how the prerogative factor, any more than the foibles of King’s Bench, could teleport habeas from the domain of common law to the domain of equity.
Scalia cited various twentieth-century cases that imputed an equitable component to habeas review.359Withrow, 507 U.S. at 717–18 (Scalia, J., concurring in part and dissenting in part). But he cited only two nineteenth-century cases: Ex parte Watkins36028 U.S. (3 Pet.) 193 (1830). and Ex parte Royall,361117 U.S. 241 (1886). neither of which mention the word “equity” a single time. So it seems strange to locate habeas in the (putatively) less restrictive realm of equity rather than in the common law’s domain.362Even if Scalia were correct that habeas is an equitable area of law, that would not end the inquiry. The historical discussion in Part I of this Article suggests federal courts cannot make any law at all, including the law of equity. See supra note 25. Perhaps, at most, habeas is a particularly flexible part of the common law—able to make case-by-case accommodations for exceptional circumstances.363See Royall, 117 U.S. at 247–53. But that kind of flexibility is exactly the opposite of the kind of categorical, per se bar the Court laid down in Teague.

As for Scalia’s statutory argument, appeals to silence can only go so far. Statutory silence simply means the statute does not govern the issue in question. And when no statute governs, the default rule in the American legal system is for courts to look to unwritten law. The default rule is not that courts get free reign to fill in the gaps by making their own law. Accordingly, the fact that no statute commands courts to grant habeas relief cannot amount to a permission slip for judicial lawmaking. That is doubly true in an area where preexisting common-law rules abound.364See Watkins, 28 U.S. at 201–03 (finding and applying such rules); Royall, 117 U.S. at 249–53 (similar, though perhaps finding the rule in the constitutional structure). See generally Clark, supra note 6 (discussing the practice of finding so-called “federal common law” rules by making inferences from the Constitution’s structure).

Justice Scalia’s theory rests on questionable foundations that may be in tension with the separation of powers. At the least, defenders of unwritten habeas bars should be able to point to some source of law—presumably the common law—to justify these rules. (This is especially true in light of Davenport, which was essentially a steroid shot into the already-brawny arm of these unwritten bars.) That would require a showing that the Court’s bars are continuous with existing common law, received in the realm, and accommodated to the people’s customs.365But see supra Section III.A (suggesting that, on one view of history, Teague may have been nothing more than a licit, partial undoing of another act of judicial lawmaking).

For other unwritten bars to habeas relief, see Withrow v. Williams, 507 U.S. 680, 718–24 (Scalia, J., concurring in part and dissenting in part) (discussing the unwritten Stone rule limiting habeas relief for Fourth Amendment violations); Slack v. McDaniel, 529 U.S. 473, 480–82 (2000) (Court’s opinion, in which Scalia joined, recognizing or creating an unwritten necessary condition for the issuance of Certificates of Appealability in habeas cases); Miller-El v. Cockrell, 537 U.S. 322, 335–38 (2003) (Court adding another unwritten requirement onto the Slack rule); id. at 349 (Scalia, J., concurring) (Scalia extending his theory to explain both the Slack and Miller-El rules); Brown v. Davenport, 142 S. Ct. 1510, 1523 (2022) (locating the doctrines of Stone v. Powell, 428 U.S. 465 (1976), Wainwright v. Sykes, 433 U.S. 72 (1977), McCleskey v. Zant, 499 U.S. 467 (1991), and Brecht v. Abrahamson, 507 U.S. 619 (1993) in the unwritten-bar category).
Whether they meet that standard is a question for another day, but it is at least a little odd that nobody seems to be seriously interrogating the issue.

IV.     Further Possibilities

If the Constitution’s original meaning bars courts from making common law, that raises at least six questions. First, what happens if Article III judges openly attempt to make law? Maybe these decisions are just legally wrong, like when a judge misinterprets a statute. If so, lawmaking decisions still bind the parties unless subsequently vacated. Or maybe they are more like decisions made without jurisdiction. And it might follow that they are not just legally wrong, but legally void—not binding even on the parties before the court.366See William Baude, The Judgment Power, 96 Geo. L.J. 1807, 1810 (2008) (“[C]ourts only have the power to issue binding judgments in cases where they have jurisdiction. That means that if the controversy is not one that the court is authorized to resolve, the judgment binds nobody.”); see also id. at 1807 (arguing that “the President is entitled to ignore a judgment if the issuing court lacked jurisdiction over the case in question”).

Second, what about nondelegation? It seems intuitive that the nondelegation doctrine applies to the judiciary as well as to the executive.367See, e.g., Margaret H. Lemos, The Other Delegate: Judicially Administered Statutes and the Nondelegation Doctrine, 81 S. Cal. L. Rev. 405 (2008); see also Aaron Nielson, Erie as Nondelegation, 72 Ohio St. L.J. 239 (2011). But see, e.g., Eric A. Posner & Adrian Vermeule, Interring the Nondelegation Doctrine, 69 U. Chi. L. Rev. 1721, 1731 (2002). If so, it might follow that Congress cannot word statutes so broadly as to delegate lawmaking power to the courts. On the other hand, scholars and courts generally agree that nondelegation concerns are less pressing in situations where the delegee branch already has some lawmaking power adjacent to the power Congress wants to hand over.368See, e.g., Gundy v. United States, 139 S. Ct. 2116, 2137 (2019) (Gorsuch, J., dissenting) (contending that a would-be delegation is likelier to be constitutionally permissible in situations where “Congress’s legislative authority . . . overlaps with authority the Constitution separately vests in another branch”). Accordingly, if federal courts have the power to make federal common law, then perhaps even the most open-ended statutes merely encourage courts to exercise lawmaking power they already possess. And that seems unobjectionable. Indeed, Professor Alexander Volokh has argued that most delegations to the judiciary are permissible for just this reason.369See Alexander Volokh, Judicial Non-Delegation, the Inherent-Powers Corollary, and Federal Common Law, 66 Emory L.J. 1391, 1395–­­96 (2017) (basing his conclusion on the idea that “[f]ederal courts have a lot of inherent powers . . . from their inherent power to make procedural rules for themselves to their inherent power to make federal common law in particular circumstances”). But if federal courts lack lawmaking power in toto, that justification falls apart entirely—Congress cannot delegate “additional” lawmaking power to a judiciary that has no lawmaking power to begin with. This could put many broad, court-empowering statutes in serious danger of violating the nondelegation doctrine.

Third, do federal courts have the power to make law in the context of equity? The historical evidence in Part I of this Article focused specifically on the common-law tradition. But much of that same evidence suggests a blanket rule against judge-made law of any kind—even the law of equity.370See supra note 25.

Fourth and related, how exactly would a court go about finding the unwritten law of equity? Maybe the method would look similar to the method for finding common law. Or maybe it would allow for a great deal more flexibility and even explicit recourse to natural law.

Fifth, do state courts have the power to make common law? Most common law is state law, after all. So one could imagine a fifty-state survey, exhausting as that would be, inquiring into the original meaning of each state’s constitution. Find the state-constitutional provision empowering the courts, ask whether that provision’s original meaning included the power to make law, and you have your answer. The analysis for the early states might look similar to this Article, given the timing of their state-constitutional ratifications. But the analysis for later additions to the Union might look very different.

Jason Boatright did exactly this in a 2020 article about judicial lawmaking in Texas.371Boatright, supra note 97; cf.Holden Tanner, States, Courts, and Common-Good Conservatism, Anchoring Truths (Jan. 3, 2022), (“I find Erie troubling [because] [t]he pronouncement that federal courts will treat state court decisions as on par with legislative enactments is nothing short of federally mandated legal realism.”). Boatright argued that no legal instrument has ever given Texas courts the power to make law—and that, in fact, “the Texas constitution prohibits them from doing so.”372Boatright, supra note 97, at 361. Boatright also stated that “[a]lmost all state constitutions and the U.S. Constitution forbid judges from making law.” Id. at 360. Obviously, I agree with at least the second part of that claim. Boatright, however, gave fulsome consideration only to the question of Texas courts’ power (vel non) to make law; he implicitly leaves the question of federal courts’ lawmaking power for another time. See id. at 361. Boatright also seemed to imply that every change in the common law is judicial lawmaking (and therefore unlawful in Texas).373See id. at 391–92, 394 (“In Texas, the common law is a fixed body of law that can be

changed by statute but not by judicial decision.”).
That view is in tension with this Article’s lawfinding method, which allows for certain gradual changes without legislative intervention. Despite that apparent disagreement, I believe state courts and legal scholarship would both benefit greatly from more articles like Boatright’s.

Sixth, how does this theory fit with Professor Bradford Clark’s proposal that some of what we call “federal common law” really amounts to the practice of inferring from the constitutional structure?374See Clark, supra note 6, at 1251 (explaining that the article “attempts to provide an alternative explanation for at least a portion of the rules that fall within the traditional enclaves of federal common law identified by the Supreme Court. This attempt rests largely upon the neglected method of inference from the structures and relationships created by the constitution in all its parts or in some principal part” (internal quotation marks and alterations omitted)). Clark’s theory, in brief, subjects any given rule to two criteria. First is “whether the matters governed by the rule in question fall within, or beyond, the legislative competence of the states.”375Id. at 1274. If the latter, the rule does not raise serious federalism concerns.376Id. For rules that pass that test, the second criterion “asks whether judicial application of the rule in question constitutes either the application of rules implied directly from the constitutional structure, or adherence to customary rules of decision necessary to implement a basic feature of the constitutional scheme.”377Id. at 1275. If yes, the rule is implied in the Constitution’s structure itself—and thus it is not judge-made and raises no specter of judicial legislation.378Id.

I suspect that, in some ways, this Article’s proposal explains the legitimacy of more federal common law than does Clark’s. I agree, subject to the caveat below, that Clark’s theory parsimoniously explains how and why rules that dosatisfy both his criteria are constitutionally permissible. A court that finds law in the Constitution’s structure is doing just that—finding law.379See id. at 1275 n.139 (explaining that properly speaking, rules that satisfy the two criteria are not judge-made). But Clark’s theory says nothing about rules that do not satisfy both criteria.380Cf. Clark, supra note 6, at 1275 n.40 (“This Article does not attempt to establish that federal common law rules that fail to satisfy the proposed criteria are necessarily unconstitutional. Such an attempt is beyond the scope of this Article.”). This Article’s lawfinding method could explain the legitimacy of some rules that flunk the second criterion: if a rule bears the three essential characteristics of customary law, a court can adopt that rule without making law—even if the rule is neither “implied directly from the constitutional structure” nor “necessary to implement a basic feature of the constitutional scheme.”381Id. at 1275. There is nothing illegitimate about finding rules of customary law.

Further, this Article’s method can help clarify what it means to “adhere[] to customary rules of decision”—part of Clark’s second criterion.382Id. Clark’s allowance for “customary rules of decision necessary to implement a basic feature of the constitutional scheme” seems to be doing some heavy lifting—he sometimes leans on the idea without explaining exactly what a rule must do to count as “customary.”383See id. at 1274–75, 1368–75 (applying this second criterion with a fair bit of generosity to the case of Boyle v. United Technologies Corp, 487 U.S. 500 (1988)). But see id. at 1287 (helpfully discussing customary law under the Swift doctrine). The traditional lawfinding method provides clarity: a rule is a “customary rule[] of decision” only if the rule is continuous with customary law, received in the realm, and accommodated to the people’s customs.384Id. at 1275 (emphasis added). Thus, even in cases where the Constitution’s structure demands some rule of decision and provides someguidance as to what such a rule should be, courts are not free to fill in the gaps with judge-made law. They must instead find the customary common law and apply it, as usual.


“Federal courts have no power to make law.” This is one of originalism’s core tenets. Yet many originalists assume that federal courts can make common law. And as the above discussion of habeas corpus illustrates, the Court has used this putative authority as a sort of wedge, slowly but perceptibly nudging bodies of doctrine this way or that.

We cannot have it both ways. The Constitution’s original meaning gives federal courts no power to make any kind of law. Common law is no exception. It is past time to remember the Constitution’s bounds, foreswear Article III lawmaking forever, and rediscover the English lawfinding tradition.


Share this article


Share it