On Treating Unlike Cases Alike

33 Const. Comment. 437

settled versus right: A Theory of precedent. By Randy J. Kozel.[1] Cambridge University Press. 2017. Pp. x + 180. $99.99 (hardcover), $34.99 (paper).

Frederick Schauer[2]

Perhaps we should blame Aristotle. In his enduring discussion of justice in the Nicomachean Ethics, Aristotle offered the now-ubiquitous maxim that like cases should be treated alike.[3] Yet despite the fact that a raft of scholars over the years have exposed the almost complete emptiness of the “treat like cases alike” maxim,[4] the maxim persists,[5] often blinding those who use it, or see it applied, to the way in which some substantive criterion of likeness is necessary in order to make the maxim anything other than a largely useless tautology.[6] Given that any two items in the world share some but not all of the properties of the respective items, any two items can be deemed alike in some respects and unalike in others, thus making the mere idea of likeness or unlikeness singularly unhelpful.

Despite the necessity of locating some criterion of similarity (and, conversely, of dissimilarity) in order to fill in this fundamental emptiness of the “treat like cases alike” maxim, the maxim has far too often served as one of the principal justifications for a regime of precedent,[7] a regime in which decision-makers have an obligation to follow previous “like” decisions even if and when they find those decisions mistaken.[8] But although there may be rare instances in which the question presented to a current court is genuinely identical to the question presented to the same court on some previous occasion,[9] those instances are sufficiently rare as to make the “treat like cases alike” maxim an unsatisfactory basis for a norm or regime of precedential constraint. Rather, such a norm or regime is important when, in the interests of stability, predictability, decision-maker constraint, or, perhaps most importantly, systemic integrity or internal coherence, the norm (presumptively[10]) requires decision-makers to ignore even relevant differences, and thus to treat unlike cases alike. Or so I will suggest here, and in doing so I will be using Randy Kozel’s important justification for a precedential regime as the platform for my analysis.


In defending the desirability of a regime of stare decisis—horizontal precedent[11]—for the Supreme Court of the United States, Kozel offers an account that is generally more normative than descriptive.[12] As a matter of empirical description, strong precedential constraint has rarely characterized Supreme Court decision-making.[13] Saul Brenner and Harold Spaeth’s wonderfully titled Stare Indecisis[14] captures well the conclusion of most empirical studies of Supreme Court decision-making—that although the Court’s opinions are replete with alleged reliance on earlier decisions, and although most of the Justices most of the time purport to be constrained by precedent in reaching their conclusions,[15] in fact stare decisis only rarely constrains and only rarely explains Supreme Court outcomes, whether for the Court as a whole or for individual Justices.[16] There are, of course, exceptions to this generalization. Justice Potter Stewart’s concurrence in Roe v. Wade,[17] seen in light of his dissent in Griswold v. Connecticut,[18] can only be understood, as Justice Stewart made clear in his Roe opinion, as largely dictated by his acceptance of the precedential force of a decision with which he disagreed. And so too with some number of criminal procedure decisions over the years by Justices John Harlan,[19] Byron White,[20] and Anthony Kennedy.[21] But such cases are the exceptions and not the rule, and in general the justices appear to have been, for generations, far better at announcing the importance of precedent than they have been in actually being constrained by it.[22]

Kozel, however, is far less concerned with describing Supreme Court practice than he is in prescribing what it ought to be. And here he offers an important argument for stare decisis as a second-best coordinating device for achieving legal consistency against the background of disagreements among the Justices (and among courts) across time and among the Justices within particular courts (pp. 45-49, 171-172). If the Justices were willing to suppress their own individual views in deference to what has been decided before and decided over time, Kozel argues, the Court would be better able to achieve that degree of cross-individual, cross-court, and cross-temporal consistency that is so vital, he believes, to the Court’s sociological legitimacy.[23]

Kozel wisely recognizes that his normative account of the value of adherence to precedent hinges on an understanding of what is a precedent for what—of what potential decisions in the case now before a court in fact fall under the umbrella of some past decision, and thus undergird the consistency across decisions that is the hallmark of precedential constraint. And here Kozel offers an account that relies substantially on the distinction between holdings and dicta (pp. 70­-94), making his account in significant ways traditional.[24] There is nothing wrong with being in the vicinity of the traditional account, of course, however much Legal Realist commentators on precedent have been disturbed by the traditional account. The Realist picture of precedent, which includes contributions by Jerome Frank,[25] Herman Oliphant,[26] and Karl Llewellyn,[27] among many others,[28] and which reaches its pinnacle of sophistication in Edward Levi’s An Introduction to Legal Reasoning,[29] is largely focused on the ability of judges in the case before them—the instant case—to locate something in the vast array of available precedent cases[30] that will justify the instant case judge’s outcome preference, a preference that, according to the Realists, is determined for reasons standing apart from and typically predating the precedents.[31] According to the Realist picture, therefore, precedent is largely about law-creation or about the manifestation in legal terminology of first-order outcome preferences rather than being about constraint by law.

The Realist picture of precedent, like much of the Realist program more generally, is far more empirical than normative.[32] Although prescriptions about how judges ought to decide cases, how they ought to justify their decisions, and how lawyers ought to advise their clients pervade Realist writings, much of this comes out of the Realist concern—hence the self-appellation “Realist”—to describe legal and especially judicial decision-making as it actually exists. And thus the Realist view of what judges “really” do turns out to be not very much of an impediment to Kozel’s predominantly normative or prescriptive argument about what judges ought to be doing. Kozel is plainly not so naïve as to imagine that judges disinclined to follow precedent will be much influenced by his normative arguments. But he nevertheless likely believes that judges with broad antecedent sympathy with the very idea of precedential constraint will find his guidance valuable. More importantly, Kozel’s arguments might be understood as directed to that amalgam of judges, lawyers, legal scholars, casebook editors, pundits, bloggers, politicians,[33] and the elite non-legal media, a group whose collective opinion about which forms of judicial behavior ought to be praised and which condemned not only reflect but also constitute the norms of judicial decision-making.[34] And thus Kozel is best, and importantly, understood as offering a plea for the creation of precedent-based norms of legal reasoning and judicial justification, norms that might, or so Kozel can be understood to argue, be especially important in and for the Supreme Court of the United States.[35]


Even apart from the more foundational Realist challenge, however, there do remain alternatives to the distinction between holding and dicta as a way of determining the scope of the constraint imposed by a precedent case. A few generations ago, for example, Arthur Goodhart challenged the prevailing English view that it was the ratio decidendi—the holding, more or less, or the reason behind the holding—that determined the ambit of application of an earlier precedent case.[36] For Goodhart, the “meaning” of a precedent case lay not in its ratio, but rather in the material facts of that case, when combined with the earlier case’s result with respect to those facts, and Goodhart’s argument generated a lively debate.[37] More recently, Larry Alexander,[38] followed by others,[39] has insisted that precedents operate much like canonically formulated rules, and thus that it is the exact language of the earlier case that operates in rule-like fashion to constrain subsequent outcomes, at least for those judges committed to the very idea of precedential constraint. As with Goodhart’s in an earlier generation, Alexander’s arguments have themselves been the subject of criticism and subsequent debate.[40]

These analyses and debates about the scope of a precedent are important, for unless we know what a precedent is we cannot begin to understand just how precedent operates and thus just why a precedential regime might—or might not—be desirable. Still, it is crucial to recognize that all of the competing accounts operate within the realm of what we might call non-identity. When a formulated legal question in the instant case is for all practical purposes identical to the formulated legal question in a precedent case, the constraints of precedent, when accepted by judges or other decision-makers,[41] create a presumption in favor of deciding the same question the same way, even in the face of the instant court’s belief that the decision in the precedent case was mistaken. For example, if the question in some not-very-hypothetical instant case is whether a state may, consistent with the Constitution, enact a total prohibition on abortion, then Roe v. Wade[42] can be understood as a precedent case answering that exact question in the negative. It is not that the questions presented in the precedent and instant cases are similar, or that they are analogous. It is that they are the same question.[43] Just as it would seem odd to say that your 2008 dark blue Subaru Outback is similar to my 2008 dark blue Subaru Outback—much more commonly you would say “I have the same car,” even though of course it is not literally or logically or empirically the same car—it seems odd to say that the 2018 question whether a state can completely prohibit abortion is similar to the question presented in 1973 in Roe v. Wade.[44] No, the questions are not similar. They are the same question.

Far more commonly, of course, the question presented in the instant case is not identical to the question presented in the precedent case, or perhaps the facts in the instant case are in some potentially relevant way different from the facts in the precedent case. When such differences exist, the issue then is whether the question in the instant case is relevantly similar—analogous[45]—to the question asked and answered in the precedent case, or whether the facts in the instant case are relevantly similar to the facts in the precedent case. But similarity is not identity, and thus a determination that the question or facts in the instant case are relevantly similar to the question addressed or the facts presented in the precedent case is a determination about which non-identical questions (or issues, or facts, or whatever) will be treated as identical for the purposes of imposing or applying precedential constraint.

The basic point, which should be controversial only for those who believe that there are natural similarities between different acts, different events, different artifacts, and different non-natural kinds, is that precedential constraint, if and when it exists, and if and when it ought to exist, is not about treating like cases as alike. On the contrary, precedential constraint is about treating cases that are somewhat alike and somewhat different as being alike for purposes of precedential constraint. It is about treating the similarities as relevant and the differences as irrelevant, and about deciding which similarities matter and which do not. Thus, identifying what is a precedent for what is about attributing or ascribing likeness; and it is not about discovering, locating, or unearthing likeness. Determining precedential similarity, and thus determining which different events or acts or questions will in spite of those differences be treated as similar, entails the question of what a decision-maker in the instant cases deems to be similar, and not about what is actually similar in some deep ontological sense.[46]


Understanding that determinations of similarity are far more ascriptions than extractions allows us to reframe, but not to disagree with, Kozel’s normative argument for the adoption or reinforcement of a norm of stare decisis. At the heart of Kozel’s normative argument is the belief that different Justices may have different views, different perceptions, different politics, and thus different outcome preferences. But by requiring those divergent Justices to follow the decisions in previous cases even when they disagree with them, Kozel argues, we dampen differences and disagreements both among the nine Justices sitting at any one time, but also across Justices sitting at different times, and thus across differently constituted courts. To use a slightly tendentious word, although a word that would appeal to stare decisis skeptics such as Justice Scalia,[47] a regime of precedent in general and stare decisis in particular suppresses differences. And it suppresses not only irrelevant differences, but also, and importantly, relevant ones. There are plainly differences between abortion and contraception, but when Justice Stewart in Roe based his decision on Griswold[48] for reasons of stare decisis (although he did not use the exact phrase), he is best understood as saying that abortion is to be deemed similar to—in the same category as—contraception despite the apparent differences between the two.

For Kozel, this suppression of difference is, on balance, a good thing. Not only does suppression of even relevant differences produce the familiar virtues of precedent—reliance, stability, predictability, and their cousins[49]—but it also reinforces and increases the Supreme Court’s legitimacy. And it does so not, as some would have it, by sending out the (false) message that the Court discovers but does not make law.[50] Rather, a strict stare decisis regime would send out what Kozel sees as the (true) message that the Court, if and when operating under a regime of stare decisis, would in deciding cases of second (and more) impression be in fact discovering and bound by law, at least if we consider previously decided cases as law.

Kozel’s more important and more novel message, however, is that a regime of stare decisis serves a goal closely related to Ronald Dworkin’s conception of integrity.[51] “Integrity” is of course a term as slippery as it is vague, and this is certainly not the occasion or location for engaging in Dworkinian exegesis, but the basic idea here is that integrity is a form of coherence or cohesiveness among various different particulars. Consider, for example, the statement in the Declaration of Independence that “all men are created equal.” Putting aside, albeit only in this particular context, the reference to “men,” and putting aside the belief of the Declaration’s authors that “all men are created equal” was compatible with the persistence of slavery, the statement nevertheless was made against the background of the numerous ways in which people are not (and were not created) as factually or empirically equal. Some are smarter, nicer, fitter, more industrious, morally better, and much else than others, and the authors of the Declaration plainly knew that. But the ascriptive dimension of the statement is that people should be treated as equal not because they are equal in a descriptive sense, but despite the fact that they are not.

This conception of grouping the dissimilar—of treating differences as irrelevant and treating the different as if they were similar—exists in numerous dimensions of life. Consider, for example, a pass/fail grading system. Even assuming, counterfactually, that some number of students actually fail when a pass/fail system is in place, it is still the case that there are great differences among those who receive passing grades. Some of the passes are better than others, perhaps even much better, but the very good passes and the barely passing passes all receive the same grade. There are often good reasons for such an approach, but it is another example of treating unlikes alike for often valuable normative or institutional purposes.[52]

It would not take very much imagination to add to these few examples, but the point should now be clear. Once we appreciate the extent to which a regime of precedent is about grouping dissimilar items, we can see that adopting a regime of precedent is about creating a community[53] of decisions and community of decision-makers, just as the creation of any other community is about the suppression of differences among community members in the service of the advantages of community itself.


The account offered here is not a challenge to Kozel. Rather, it is an explication of what appears to be Kozel’s own justifications for stare decisis in the Supreme Court. Although Kozel puts the point in language somewhat different from mine, his argument about disagreement is entirely compatible with the account I offer here about difference more generally. Far too much of the literature on precedent generally and stare decisis in particular has been saddled for too long with the “treat like cases alike” paradigm, a paradigm that invites those who accept or follow it to imagine that they are identifying likenesses rather than creating them. The idea of precedent, whether vertical or horizontal, is fundamentally ascriptive and not descriptive, and the ascriptiveness—the act of “deeming”—enables those who make decisions under a regime of precedent to participate in the creation of a community of decisions. To call a group of similar decisions about dissimilar facts and dissimilar questions a “community” is perhaps more than a trifle metaphorical, but a community of decisions may bear some relationship to the existence of a political community in a less metaphorical sense, and thus also to a community of decision-makers. The community is not a community of like decision-makers making like decisions, but instead a community of unlike decision-makers with unlike views being compelled by a regime of precedent to make like decisions. The community of decisions and the community of decision-makers is thus an example and product of the treatment of unlikes as being alike, as Kozel has valuably helped us to understand.

        [1].    Professor of Law, Notre Dame Law School.

        [2].    David and Mary Harrison Distinguished Professor of Law, University of Virginia. This paper was prepared for the symposium honoring and discussing Randy Kozel’s Settled Versus Right: A Theory of Precedent, held at the University of Richmond School of Law on April 20–21, 2018.

        [3].    Aristotle, Ethica Nicomachea 5.1131a10–b15 (W. D. Ross trans., Oxford Univ. Press 1925) (c. 350 B.C.E.). See also Aristotle, Magna Moralia 1.1193b–1194b (W. D. Ross trans., Oxford Univ. Press 1925) (n.d.); Aristotle, Politics, 3.1280a8–16, 1282b18–23 (Ernest Barker ed. & trans., Oxford Univ. Press 1946) (c. 350 B.C.E.); W. von Leyden, Aristotle on Equality and Justice: His Political Argument (1985).

        [4].    E.g., Larry Alexander & Emily Sherwin, The Rule of Rules: Morality, Rules, and the Dilemmas of Law 136–56 (2001); H. L. A. Hart, The Concept of Law 159 (Penelope A. Bulloch et. al. eds., 3d ed. 2012) (1961); David Lyons, On Formal Justice, in Moral Aspects of Legal Theory: Essays on Law, Justice, and Political Responsibility 13 (1993); Joseph Raz, The Morality of Freedom 239–44 (1986); Christopher J. Peters, Foolish Consistency: On Equality, Integrity, and Justice in Stare Decisis, 105 Yale L.J. 2031 (1996); Adam M. Samaha, Randomization in Adjudication, 51 Wm. & Mary L. Rev. 1 (2009); Peter Westen, The Empty Idea of Equality, 95 Harv. L. Rev. 537 (1982); Kenneth I. Winston, On Treating Like Cases Alike, 62 Calif. L. Rev. 1 (1974); Ben Johnson & Richard Jordan, Should Like Cases be Decided Alike? A Formal Analysis of Four Theories of Justice (Feb. 21, 2018) https://ssrn.com/abstract=3127737. My own contributions to this literature are Frederick Schauer, Precedent, in The Routledge Companion to Philosophy of Law 123 (Andrei Marmor ed., 2012); Frederick Schauer, Precedent, 39 Stan. L. Rev. 571 (1987).

        [5].    See, e.g., Marvin Ammori, First Amendment Architecture, 2012 Wis. L. Rev. 1, 74 (2012); Joseph William Singer, Normative Methods for Lawyers, 56 UCLA L. Rev. 899, 948 (2009).

        [6].    See Frederick Schauer, Profiles, Probabilities, and Stereotypes 199–223 (2003). To the same effect, although more obliquely, are Hans Kelsen, Aristotle’s Doctrine of Justice, in What is Justice? Justice, Law, and Politics in the Mirror of Science 110–136 (1957), and Richard A. Wasserstrom, The Judicial Decision: Toward a Theory of Legal Justification 112–13 (1961).

        [7].    See, e.g., Rupert Cross, Precedent in English Law 4 (3d ed., 1977); Neil MacCormick, Legal Reasoning and Legal Theory 73–99 (1978); Alfonso Ruiz Miguel, Equality Before the Law and Precedent, 10 Ratio Juris 372 (1997). An important rejoinder is Lyons, supra note 4.

        [8].    On the importance of understanding that a genuine argument from precedent is not dependent on the precedent-follower’s belief in the correctness of the precedent decision, see Kimble v. Marvel Entm’t, LLC, 135 S. Ct. 2401, 2409 (2015); Wasserstrom, supra note 6, at 52.

        [9].    See Causeway Med. Suite v. Ieyoub, 109 F.3d 1096, 1113 (5th Cir. 1997) (Garza, J., concurring) (regretting that he was compelled by a directly applicable precedent to make a decision he believed was “inimical to the Constitution”).

      [10].    In theory, the force of a precedent might be conclusive, such that ignoring (or overturning) a precedent was simply impermissible. This may well have been English practice prior to 1966, when the view was that a court could not overrule or override even its own precedent—that was for Parliament to do, and not even for the court that had originally decided it. This understanding of the infinite weight of a precedent was changed with the famous 1966 Practice Statement (see Cross, supra note 7, at 109­–13; R. W. M. Dias, Jurisprudence 127 (5th ed., 1985)), and Great Britain has now joined the rest of the common law world in considering even plainly applicable precedents as presumptive and not absolute. Nevertheless, if the presumptivity of a precedent does not require that there be better reasons for overriding, ignoring, or overruling a precedent than there would have been for reaching a different result in the first instance, then the precedential force of a presumptive precedent disappears. See Frederick Schauer, Playing By the Rules: A Philosophical Examination of Rule-Based Decision-Making in Law and in Life 181–87 (1991); Wasserstrom, supra note 6, at 52–53.

      [11].    On the distinction between vertical precedent (the obligation of judges to follow the decisions of those above them in the judicial hierarchy) and horizontal precedent (or stare decisis), and the obligation of judges to follow the previous decisions of their own court, even when that court is now populated by different judges, see Frederick Schauer, Thinking Like a Lawyer: A New Introduction to Legal Reasoning 36–41 (2009) [hereinafter, Schauer, Thinking]; Larry Alexander, Constrained By Precedent, 63 S. Cal. L. Rev. 1 (1989).

      [12].    Kozel describes Chapter 1 of his book as, in part, “descriptive” (p. 7), but Chapter 1 is in reality a description of the history, theory, and doctrine of precedent. It is not a description of the extent to which that theory and doctrine actually influence judicial decisions.

      [13].    Much of the empirical research is summarized in Frederick Schauer, Stare Decisis and the Selection Effect, in Precedent in the United States Supreme Court 121 (Christopher J. Peters ed., 2013) [hereinafter Precedent in the United States Supreme Court]; Frederick Schauer, Has Precedent Ever Really Mattered in the Supreme Court?, 24 Ga. St. U. L. Rev. 381 (2007).

      [14].    Saul Brenner & Harold J. Spaeth, Stare Indecisis: The Alteration of Precedent on the Supreme Court, 1946–1992 (1995). See also Harold J. Spaeth & Jeffrey A. Segal, Majority Rule or Minority Will: Adherence to Precedent on the U.S. Supreme Court (1999); Jeffrey A. Segal & Harold J. Spaeth, The Influence of Stare Decisis on the Votes of Supreme Court Justices, 40 Am. J. Pol. Sci. 971 (1996). The empirical literature on the role of precedent in the Supreme Court is large and growing. An important overview and analysis is Thomas G. Hansford & James F. Spriggs, The Politics of Precedent in the U.S. Supreme Court (2006). See also the various contributions in Precedent in the United States Supreme Court, supra note 13. For more subtle analyses of the role played by precedent, especially in courts other than the Supreme Court of the United States, see Richard A. Posner, How Judges Think 43–51 (2008); Frank Cross, Appellate Court Adherence to Precedent, 2 J. Empirical Legal Stud. 369 (2005); Stefanie A. Lindquist & Frank B. Cross, Empirically Testing Dworkin’s Chain Novel Theory: Studying the Path of Precedent, 80 N.Y.U. L. Rev. 1156 (2005).

      [15].    A large part of the empirical problem in identifying the force of precedent is that the typical reference in a Supreme Court opinion to the importance of precedent or stare decisis is offered by a Justice who in fact agreed with the earlier decision. City of Akron v. Akron Ctr. for Reprod. Health, Inc., 462 U.S. 416 (1983), e.g., is offered by Justices who were not on the Court at the time of the original decision and thus may well have agreed with the earlier decision as a matter of first order substance, e.g., Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833 (1992) (joint opinion of O’Connor, Kennedy, & Souter, JJ.), or is cursorily acknowledged and dismissed in the course of a Justice’s refusing to follow an earlier decision. E.g., Leegin Creative Leather Products, Inc. v. PSKS, Inc., 551 U.S. 877 (2012).

      [16].    See references in note 14, supra. See also the unfortunately neglected Maurice Kelman, The Forked Path of Dissent, 1985 Sup. Ct. Rev. 227 (1985).

      [17].    410 U.S. 113, 170 (1973) (Stewart, J., concurring).

      [18].    381 U.S. 479, 530 (1965) (Stewart, J., dissenting).

      [19].    See Henry J. Bourguignon, The Second Mr. Justice Harlan: His Principles of Judicial Decision Making, 1979 Sup. Ct. Rev. 251, 279 (1979).

      [20].    Compare Edwards v. Arizona, 451 U.S. 477 (1981) (White, J., for the Court), with Miranda v. Arizona, 384 U.S. 436, 504 (1966) (Harlan, J., dissenting, joined by Stewart, J., & White, J.). On the genuine precedential force of Miranda, see also Dickerson v. United States, 530 U.S. 428 (2000). Although Chief Justice Rehnquist’s opinion for the Court sounds in precedential constraint, the issue is complicated by the Court’s obvious desire in Dickerson to reaffirm its own interpretive authority, and, indeed, supremacy, against congressional encroachment—against Congress’s claim that it has the authority to overrule the Court’s constitutional decisions.

      [21].    Compare Ring v. Arizona, 536 U.S. 584, 613 (2002) (Kennedy, J., concurring), with Apprendi v. New Jersey, 530 U.S. 466, 524 (2000) (O’Connor, J., dissenting, joined by Rehnquist, C.J., Kennedy, J., & Breyer, J.).

      [22].    As noted above, supra note 20, a somewhat more complex example is Dickerson, 530 U.S. at 428. It is true that at least some of the Justices who joined the Court’s opinion disallowing a congressional attempt to overrule Miranda, 384 U.S. at 436, would likely have been dissenters in Miranda had they been on the Court at the time, but the Court’s assertion of judicial interpretive supremacy in Dickerson likely provided an important justification for refusing to allow congressional overruling, the Court’s references to stare decisis notwithstanding. Dickerson thus involved three different desiderata—adhering to past decisions, making the right decision as a matter of first-order constitutional substance, and reaffirming the Court’s role as interpreter of the Constitution—that in this and other cases are likely in conflict.

      [23].    For a valuable description and application of the venerable distinction between sociological and moral (or political) legitimacy, see Richard H. Fallon, Jr., Law and Legitimacy in the Supreme Court 155–74 (2018); Richard H. Fallon, Jr., Legitimacy and the Constitution, 118 Harv. L. Rev. 1787 (2005). See also Deborah Hellman, The Importance of Appearing Principled, 37 Ariz. L. Rev. 1107 (1995).

      [24].    See Cross, supra note 7, at 38–102; Edmund M. Morgan, Introduction to the Study of Law 155 (1948); Edwin W. Patterson, Jurisprudence: Men and Ideas of the Law 300–10 (1953); Eugene Wambaugh, The Study of Cases (2d ed., 1894); Glanville Williams, Learning the Law 62–68 (10th ed., 1978).

      [25].    See Jerome Frank, Law and the Modern Mind (1930).

      [26].    Herman Oliphant, A Return to Stare Decisis, 14 A.B.A. J. 71, 159 (1928). A particularly valuable historical and jurisprudential analysis of both precedent traditionalists such as Wambaugh and precedent Realists such as Oliphant is Charles W. Collier, Precedent and Legal Authority: A Critical History, 1988 Wis. L. Rev. 771 (1988).

      [27].    See Karl N. Llewellyn, The Theory of Rules 124–25 (Frederick Schauer ed., Univ. Chicago Press 2011) (1938–1939). Somewhat more guarded is Karl Llewellyn, The Case Law System in America 68–69 (Michael Ansaldi trans., Paul Gewirtz ed., Univ. Chicago Press 1989) (1928–1929) [hereinafter, Llewellyn, Case Law System]. See also Karl N. Llewellyn, The Rule of Law in Our Case-Law of Contracts, 47 Yale L.J. 1243, 1244–46 (1938). See generally William Twining, Karl Llewellyn and the Realist Movement 203–69 (2d ed., Cambridge Univ. Press 2012) (1973).

      [28].    See Felix S. Cohen, Ethical Systems and Legal Ideals 33–40 (1933); Brian Leiter, Naturalizing Jurisprudence: Essays on American Legal Realism and Naturalism in Legal Philosophy 107–09 (2007).

      [29].    Edward H. Levi, An Introduction to Legal Reasoning (Univ. Chicago Press 2013) (1949).

      [30].    It is worth noting that the Realist challenge to traditional views about the constraining force of precedent has two distinct prongs. One is the claim that individual precedents are written in vague and/or non-canonical language and are thus malleable in subsequent cases. See Llewellyn, Case Law System, supra note 27, at 3–4. The other is that common law systems, and especially in federal systems such as the United States, offer an enormous menu of potential precedents, thus giving judges the power to select which ones to use, a selection typically made, according to the Realists, on the basis of the judges’ outcome preferences.

      [31].    See Schauer, Thinking, supra note 11, at 36–60.

      [32].    Leiter, supra note 28, at 54–57; Schauer, Thinking, supra note 11, at 138–42; John Henry Schlegel, American Legal Realism and Empirical Social Science (1995); Walter W. Cook, Scientific Method and the Law, 13 A.B.A. J. 303 (1927); Frederick Schauer, Foreword, in Twining, supra note 27, at ix, xii; Frederick Schauer, Legal Realism Untamed, 91 Texas L. Rev. 749 (2013); John Henry Schlegel, American Legal Realism and Empirical Social Science: From the Yale Experience, 28 Buff. L. Rev. 459 (1979); John Henry Schlegel, American Legal Realism and Empirical Social Science: The Singular Case of Underhill Moore, 29 Buff. L. Rev. 195 (1980).

      [33].    And perhaps especially members of the United States Senate in judicial confirmation proceedings.

      [34].    For the jurisprudentially inclined, the statement in the text might be understood as a reference to the sources of what H. L. A. Hart called the “ultimate rule of recognition.” Hart, supra note 4, at 100–10. Just as the ultimate rule of recognition determines in a legal culture what counts as law and what does not, so too does the same diffuse amorphous process (see Frederick Schauer, Is the Rule of Recognition a Rule?, 3 Transnat’l Legal Theory 1 (2012); A.W.B. Simpson, The Common Law and Legal Theory, in Oxford Essays in Jurisprudence (Second Series) 77 (A. W. B. Simpson ed., 1973)) determine, at bedrock, which forms of legal argument, legal reasoning, and legal decision-making will be considered valid or legitimate in a legal culture and which will not.

      [35].    It is possible, however, that Kozel’s focus on the Supreme Court gets things backwards. Given the increasingly political makeup of the Court, the increasingly political nature of the selection and confirmation processes, and the ideologically charged character of much of the Court’s docket, expecting the Court ever to take precedent seriously on a regular basis may be little more than wishful thinking. In appellate courts with non-discretionary and less pervasively ideologically salient dockets, however, including the United States Courts of Appeals, some state supreme courts, and all of the intermediate appellate courts in states whose supreme courts do have discretionary dockets, a strong commitment to precedent may more often make more of a difference.

      [36].    Arthur L. Goodhart, Determining the Ratio Decidendi of a Case, 40 Yale L.J. 161 (1930).

      [37].    See J. L. Montrose, Ratio Decidendi and the House of Lords, 20 Mod. L. Rev. 124 (1957); J. L. Montrose, The Ratio Decidendi of a Case, 20 Mod. L. Rev. 587 (1957); A.W.B. Simpson, Notes of Cases, The Ratio Decidendi of a Case, 20 Mod. L. Rev. 413 (1957); Julius Stone, The Ratio Decidendi of the Ratio Decidendi, 22 Mod. L. Rev. 597 (1959). Goodhart’s rejoinder is A. L. Goodhart, The Ratio Decidendi of a Case, 22 Mod. L. Rev. 117 (1959). See also Geoffrey Marshall, What is Binding in a Precedent?, in Interpreting Precedents: A Comparative Study 503–17 (D. Neil MacCormick & Robert S. Summers eds., 1997).

      [38].    Alexander, supra note 11; Larry Alexander & Emily Sherwin, Judges as Rule Makers, in Common Law Theory 27 (Douglas E. Edlin ed., 2007).

      [39].    Including myself. Schauer, Thinking, supra note 11, at 50–54.

      [40].    See Grant Lamond, Do Precedents Create Rules?, 11 Legal Theory 1 (2005).

      [41].    Although not directly relevant to Kozel’s book or to this article, it is worth noting that both vertical and horizontal precedent may be part of the decision-making norms on non-judicial decisional domains. See, e.g., Trevor W. Morrison, Stare Decisis in the Office of Legal Counsel, 110 Colum. L. Rev. 1448 (2010).

      [42].    410 U.S. 113 (1973).

      [43].    On the distinction between the same question and a merely similar question, and thus on the distinction between precedential constraint strictu sensu and reasoning by and from analogy, see Frederick Schauer, Why Precedent in Law (and Elsewhere) is Not Totally (or Even Substantially) About Analogy, 3 Persp. Psychol. Sci. 454 (2008).

      [44].    410 U.S. 113 (1973).

      [45].    Vast amounts have been written about analogical reasoning in law, some of it celebratory and some of it deflationary. An overview of the issues, as well as an addition to the array of conventional views, is Frederick Schauer & Barbara A. Spellman, Analogy, Expertise, and Experience, 84 U. Chi. L. Rev. 249 (2017).

      [46].    Once we understand that ascribing similarity for the purposes of precedential reasoning is largely about treating different acts, different events, different facts, and different questions as nevertheless similar, we can glimpse the relationship between the topic of precedent and the topic of legal fictions. See generally Legal Fictions in Theory and Practice (Maksymilian Del Mar & William Twining eds., 2015). Many of the classic fictions are not so much genuine fictions as fabrications as they are about consciously deeming things that are in some or many respects plainly different as nevertheless similar for some legal purpose, as with deeming corporations as natural persons for various legal doctrines, or deeming the child born of a married woman as the child of the marriage even if the husband was not the biological father, or deeming the island of Minorca as being in London for jurisdictional purposes, as in the famous case of Mostyn v. Fabrigas, 98 Eng. Rep. 1021 (1774) (see Lon L. Fuller, Legal Fictions 18–21 (1967)).

      [47].    See Lawrence v. Texas, 539 U.S. 558, 586 (2003) (Scalia, J., dissenting); Antonin Scalia, A Matter of Interpretation: Federal Courts and the Law 139 (Amy Gutmann ed., 1997) (“The whole function of [stare decisis] is to make us say that what is false under proper analysis must nevertheless be held to be true . . . .”). For discussion of Justice Scalia’s views about stare decisis, more of which appeared in extra-judicial writings and speeches than in judicial opinions, see Amy Coney Barrett, Originalism and Stare Decisis, 92 Notre Dame L. Rev. 1921 (2017); David A. Strauss, Tradition, Precedent, and Justice Scalia, 12 Cardozo L. Rev. 1699 (1991).

      [48].    See supra notes 17–18, and accompanying text.

      [49].    See Daniel A. Farber & Suzanna Sherry, Judgment Calls: Principle and Politics in Constitutional Law 63–74 (2009).

      [50].    For commentary on the position, often associated with Planned Parenthood of Southeastern Pennsylvania v. Casey, 505 U.S. 833 (1992), that the Supreme Court should promote through its opinions the view that constitutional law and the Supreme Court’s decisions are not dependent on the makeup of the Court, see Hellman, supra note 23; see also Jason Iuliano, The Supreme Court’s Noble Lie, 51 U.C. Davis L. Rev. 911 (2018). But if the Court’s decisions in fact do depend heavily on the Court’s staffing at any particular period of time, then the position reduces to the claim that it is good for the Court’s legitimacy that the public remain unaware of the extent to which the pre-legal social, political, and moral attitudes of the Justices play a predominant (but not exclusive) role in determining the votes of the Justices, and therefore in determining the Court’s outcomes.

      [51].    Ronald Dworkin, Law’s Empire 400–03 (1986); see also Ronald Dworkin, Freedom’s Law: The Moral Reading of the American Constitution 10–11 (1996); Ronald Dworkin, Justice for Hedgehogs 260–64 (2011); Ronald Dworkin, Justice in Robes 176–78 (2006); Stephen Guest, Ronald Dworkin 101–23 (3d ed., 2013).

      [52].    I am grateful to Walter Sinnott-Armstrong for suggesting the example.

      [53].    On Dworkinian integrity as community reinforcing and community creating, see Guest, supra note 51, at 101–23.