1. “But let's remember that we're all part of one American family. We are united in common values, and that includes belief in equality under the law; a basic respect for public order and the right to peaceful public protest; a reverence for the dignity of every single man, woman and child among us; and the need for accountability when it comes to our government.” Barack Obama, Statement by the President, Edgartown, MA (Aug. 14, 2014), at https://www.whitehouse.gov/the-press-office/2014/08/14/statement-president.

2. The original German reads: “Und nun gar das besondere Prinzip der sogenannten Gleichheit vor dem Gesetz! Es bedeutet nichts anderes, als daß die rechtsanwendenen Organe keine Unterschiede machen sollen, die das anzuwendende Recht nicht selbst macht. … Mit Gleichheit hat dieses Prinzip kaum noch etwas zu tun. Es besagt nur, daß das Recht so angewendet werden soll, wie es seinem Sinne nach anzuwenden ist. Es ist das Prinzip der Recht- oder Gesetzmäßigkeit, das jeder Rechtsordnung ihrem Wesen nach immanent ist, gleichgültig, ob diese Ordnung gerecht oder ungerecht ist.” Hans Kelsen, Was ist Gerechtigkeit? (2010), at 35–36; for related criticisms see Alf Ross, Om ret og retfærdighed (1953), at 371–372; Winston, Kenneth I., On Treating Like Cases Alike, 62 Cal. L. Rev. 1 (1974)CrossRefGoogle Scholar; Herbert Lionel Adolphus Hart, The Concept of Law (1997), at 160–161; Westen, Peter, The Empty Idea of Equality, 95 Harv. L. Rev. 537 (1982)CrossRefGoogle Scholar; Ingram, Peter, Procedural Equality, 21 Archiv für Rechts- und Sozialphilosophie 39 (1985)Google Scholar; Sadurski, Wojciech, Equality Before the Law: A Conceptual Analysis, 60 Austl. L.J. 131 (1986)Google Scholar; Peters, Christopher J., Foolish Consistency: On Equality, Integrity and Justice in Stare Decisis, 105 Yale L.J. 2031 (1996)CrossRefGoogle Scholar; Peters, Christopher J., Equality Revisited, 110 Harv. L. Rev. 1210 (1997)CrossRefGoogle Scholar; Fred Feldman, The Irrelevance of Equality Before the Law (2004) (unpublished manuscript), at http://people.umass.edu/~ffeldman/EBLforLund2.pdf.

3. See Hart, supra note 2, at 159–167; Sadurski, supra note 2, at 131; Lippert-Rasmussen, Kasper, Punishment and Discrimination, in Punishment and Ethics (Ryberg, Jesper & Corlett, J. Angelo eds., 2010), at 171–172Google Scholar.

4. This is not because equality in the law does not pose interesting problems of its own. See, e.g., Gardner, John, Liberals and Unlawful Discrimination, 9 Oxford J. Legal Stud. 1 (1989)CrossRefGoogle Scholar; Campbell, Tom D., Unlawful Discrimination, in Ethical Dimensions of Legal Theory (Sadurski, Wojciech ed., 1991)Google Scholar; Gardner, John, Discrimination as Injustice, 16 Oxford J. Legal Stud. 353 (1996)CrossRefGoogle Scholar; Doyle, Oran, Direct Discrimination, Indirect Discrimination and Autonomy, 27 Oxford J. Legal Stud. 537 (2007)CrossRefGoogle Scholar; Sepielli, Andrew, The Law's ‘Majestic Equality’, 32 Law & Phil. 673 (2013)CrossRefGoogle Scholar.

5. See Carter, Ian, Respect and the Basis of Equality, 121 Ethics 538, 541 (2011)CrossRefGoogle Scholar; Heinrichs, Bert, What Is Discrimination and When Is It Morally Wrong?, 12 Jahrbuch für Wissenschaft und Ethik 97, 102 (2007)CrossRefGoogle Scholar; Westen, supra note 2, at 543; Feinberg, Joel, Noncomparative Justice, 83 Phil. Rev. 297, 310 (1974)CrossRefGoogle Scholar; Singer, Peter, Is Racial Discrimination Arbitrary?, 8 Philosophia 185, 186 (1978)CrossRefGoogle Scholar. Like Christopher Peters, however, I am skeptical that the view can actually be attributed to Aristotle. See Peters, Foolish Consistency, supra note 2, at 2058. The passage most commonly cited to support the attribution seems to me to argue for adjusting the allocation of goods according to desert and, perhaps, subsuming the distribution of goods under the issue of justice in the first place. See Nicomachean Ethics (Roger Crisp trans., 2000), at 86–87, 1131b–1132a. Feinberg, it is worth noting, is at places more careful in his reading; see Feinberg, supra, at 303, 319.

6. I shall speak throughout of “the court” as an agent that “treats equally,” “punishes,” etc. This is not because I assume that institutions can, properly speaking, be conceived as agents—perhaps they can, but I do not wish to rely on the assumption—but merely a form of shorthand for whomever we ought ultimately to consider the relevant agents, e.g., the judge, the jury, the lawyers, etc.

7. Some may prefer a subjective version of the principle, where all that is required is that the agent believes the cases and/or treatment to be alike. I believe the objective version is the more plausible and intuitive of the two, and so I focus on that here, but since nothing of substance in the following will turn on the distinction, I invite readers who disagree to assume that we are discussing the subjective version instead.

8. Does the fact that these treatments distribute burdens and therefore also affect telic considerations, say, of desert, beneficence, and outcome equality, make them unsuitable? I do not believe so—it is difficult to find good examples of deontic equality that do not—but if it aids the reader, I invite her to imagine that we are discussing instead how the court decides the admissibility of a piece of material evidence, estimates the reliability of witness testimony, or some other such action.

9. Hart, supra note 2, at 160.

10. Feinberg, supra note 5.

11. Note that the assumption is therefore deliberately neutral on the controversial issue of whether there are hard cases, i.e., cases where there is no unique action that the court ought to take. I leave open the possibility that sets of permissible actions contain more than one member partly so as to avoid having to take sides with Dworkin against this possibility. See Ronald Dworkin, Law's Empire (2006).

12. Westen, supra note 2, at 1220.

13. Note that we constrain the properties here, because, e.g., “being a case in which the court ought to φ” is a (moral) property, but including it and other moral properties will prevent us from cleanly separating the three possible ways of comparing cases. I assume that a property is normatively relevant iff it will give rise to reasons for treating the case one way or another.

14. Two such cases will, of course, thereby also share the sets of reasons and permissible actions, just as two cases that share the set of reasons, but not properties, will thereby also share the set of permissible actions.

15. Some might now raise the objection that I have not defended the idea that different sets of properties can give rise to the same reasons, and different sets of reasons to the same permissible action(s). If either or both of these claims are false, the objector might say, then we have no reason to prefer the third option. My response would be that, although I have not defended the claims, my argument here does not rely on them. All we require is that it is conceptually possible for the claims to be true, which I take it few will deny, since this allows us to imagine a situation in which differences in properties or reasons force us to classify cases as unalike, as per above. This possibility is what shows that the likeness of cases pertains to permissible actions, not properties or reasons.

16. Feinberg, supra note 5, at 313. Peters makes the same point: “Identity of situation is defined by reference to criteria for the treatment in question. If persons are identically entitled to the relevant treatment, they are ‘identically situated’ under our expression of equality.” Peters, Foolish Consistency, supra note 2, at 2059. See also Montague, Phillip, Comparative and Non-comparative Justice, 30 Phil. Q. 131, 135 (1980)CrossRefGoogle Scholar; Hoffman, Joshua, A New Theory of Comparative and Non-comparative Justice, 70 Phil. Stud. 165, 167 (1993)CrossRefGoogle Scholar.

17. The same observations hold true, mutatis mutandis, for One Size Fits All. Note also that a benefit of the definition is that it can account for our intuitions in cases where relevant circumstances of the cases change, e.g., because the legislature votes into effect an amendment to a body of law. Intuitively, if courts follow the rules and treat otherwise similar cases differently before and after the amendment, this ought not to constitute procedural inequality, and on the suggested definition it does not.

18. Note that an action's “having reason to φ in case A and φ in case B” here denotes the result of weighing and comparing the applicable reasons, so that a court may be said to have a certain level of reasons to φ even when these reasons are weaker than the reasons supporting another action, or when the reasons make the action all things considered impermissible. As an example, a person can, on this terminology, be said to have more reason to cause unjustified harm of a certain severity than to cause even more severe unjustified harm, because there are stronger reasons against performing the latter.

19. Is there space for supererogatory actions in the context of the court's treatment of cases? Those who want to answer in the affirmative, and who may for that reason have bristled at the way in which I have hitherto identified actions that the court ought to perform with actions that are permissible, will presumably find that this definition has the added advantage of accommodating their view. On the present definition, a court that has a range of permissible actions, some of which it has stronger reason to perform than others, will fail to treat two cases equally if it performs two actions that differ in this respect.

20. See Peters, Equality Revisited, supra note 2, at 1227 (“[T]he alleged prescriptive strength of equality need not be absolute. … The egalitarian may believe that although the incorrect treatment of X is a reason favoring incorrect treatment of Y, other reasons exist that disfavor such treatment - reasons that outweigh the reason provided by equality.”). Greenawalt, Kent, “Prescriptive Equality”: Two Steps Forward, 110 Harv. L. Rev. 1265, 1270 (1997)CrossRefGoogle Scholar (“If the principle of prescriptive equality has normative force, it ‘reinforces’ and ‘pulls against’ independent reasons…”). Schauer makes the related point for a doctrine of precedent: “To say that precedent provides a reason for deciding in a particular way is not to say that following precedent is what we should always do, all things considered.” Schauer, Frederick, Precedent, 39 Stan. L. Rev. 571, 592 (1987)CrossRefGoogle Scholar. Note also that this appears to be the stumbling block on which Peter Westen's analysis ultimately trips. In his reply to Greenawalt, he comes eerily close to formulating the principle as I here set it out only to reject it as “untrue” because he believes it “tells us we should do something we know we should not do, that is, to take people who are entitled to certain treatment and either give them that treatment or not give them that treatment.” Westen, Peter, To Lure the Tarantula from Its Hole: A Response, 83 Colum. L. Rev. 1186, 1193 (1983)CrossRefGoogle Scholar. Although he is right that this implication would make the principle implausible, it follows only if the principle provides the court with a decisive reason that renders all other reasons irrelevant. In places Peters appears to make a similar mistake, when claiming that egalitarian justice must be “incoherent.” See Peters, Equality Revisited, supra note 2, at 1249–1250; see also note 45 below.

21. Peters, Foolish Consistency, supra note 2, at 2068.

22. What about the notion familiar from doctrines of precedent that the most recent way of treating like cases should carry added or even exclusive weight on how the present case ought to be treated? Whatever its merits it seems to me to involve a very different principle than the one at stake, one that must be both defined and defended independently, and I shall not discuss it here. It is worth mentioning, however, that I do think some of the challenges to the principle of equality before the law discussed below would apply to it too.

23. Note that we need not hold that the reason to treat a like case the way a minority of like cases have been treated disappears. We can say rather that it is outweighed by the reason to treat it the way the majority of like cases have been treated, but that the ratio between majority and minority (or minorities, if there is more than one alternate way of treating like cases) influences the strength of the reason to prefer treating the present case like the majority. This allows that only in situations where no majority exists would the reasons cancel out, which seems intuitively right.

24. See Peters, Foolish Consistency, supra note 2, at 2062: (“[T]he ‘true’ norm of equality … holds that the bare fact that a person has been treated a certain way is a reason in itself for treating another identically situated person the same way.”) (emphasis in original). Also Schauer, supra note 20, at 571 (“The previous treatment of occurrence X in manner Y constitutes, solely because of its historical pedigree, a reason for treating X in manner Y if and when X again occurs.”) (emphasis in original). Deborah Hellman also cites Gerald Postema in support of this point. Hellman, Deborah, An Epistemic Defense of Precedent, in Precedent in the United States Supreme Court (Peters, Christopher J. ed., 2013)Google Scholar, at 64.

25. These principles borrow heavily, as I am sure is obvious, from the discussion of what Larry Temkin has dubbed “The Slogan.” Larry S. Temkin, Inequality (1993), at 248; Temkin, Larry S., Equality, Priority, and the Levelling Down Objection, in The Ideal of Equality (Clayton, Matthew & Williams, Andrew eds., 2002)Google Scholar; Parfit, Derek, Equality or Priority, in The Ideal of Equality (Clayton, Matthew & Williams, Andrew eds., 2002), at 98–99, 110–115Google Scholar. Recall however, that unlike in the debate between egalitarians and prioritarians, we are dealing with procedural equality, a subspecies of deontic egalitarianism, rather than the telic concern over equality in the distribution of goods between persons. See Parfit, supra, at 88–90.

26. Note that I do not assume, nor need to assume, that equality before the law is a legal principle. Only that it is a type of principle that, whether legal or moral, is at least potentially capable of giving reasons for the court to act. I say “potentially” to include the possibility that the principle does not in fact give such reasons, e.g., because it happens to be false.

27. Greenawalt, supra note 20, at 1271–1272.

28. See Parfit, supra note 25, at 111–114.

29. That is, I assume that there cannot be a decisive reason to φ if there is not a pro tanto reason to φ, and thus, that if the stated version fails, e.g., because it turns out to be implausible, then this means that the strict version fails too.

30. Hart, supra note 2, at 159; see also Feinberg, supra note 5, at 310; Greenawalt, Kent, How Empty Is the Idea of Equality?, 83 Colum. L. Rev. 1167, 1173–1175 (1983)CrossRefGoogle Scholar.

31. Why not simply require that the court treats each case as it ought? Because, again, we presumably want to maintain that equality before the law could obtain even under conditions of non-comparative injustice, so long as the cases are treated with comparative justice.

32. Note that the target of comparison has thereby shifted. The narrow principle compares with and prefers actions that have most frequently been performed. The wide principle compares with and prefers strengths of reasons for action that have most frequently obtained. However, strengths of reasons cannot differ for the same action when comparing like cases, so it is possible that the focus on comparing actions in the narrow principle works because frequency of actions piggybacks on frequency of strengths of reasons there, and that the two versions of the principle are thus ultimately concerned with the same type of comparison.

33. Frederick Schauer, Profiles, Probabilities and Stereotypes (2003), at 201–203.

34. Peters, Foolish Consistency, supra note 2; Peters, Equality Revisited, supra note 2. Peters himself modestly claims that his work “…is an effort to carve a new path through very well-trodden territory - always a difficult task, and one that should inspire a healthy dose of humility in those who attempt it. I thus have no illusions that the analysis this Article offers is comprehensive, unassailable, or entirely original. I hope only that it contributes in some positive way to the current thinking about its subject.” Peters, Equality Revisited, supra note 2, at 1214. These admirably honest observations undoubtedly apply doubly to the present article.

35. Hart, supra note 2, at 160–161.

36. Sadurski, supra note 2, at 132 (my emphasis). See also Kelsen, supra note 2, at 1; Ross, supra note 2.

37. See Greenawalt, supra note 20, at 1268–1269, on whether (what I label) the principle of legal procedural egalitarianism is best conceived as part of or separate from other conceptions of “formal” equality.

38. See Knight, Carl, Describing Equality, 28 Law & Phil. 327, 335–338 (2009)CrossRefGoogle Scholar; Peters, Equality Revisited, supra note 2, at 1223–1224. In fairness to Sadurski, Ross, Kelsen, and their compatriots, the argument can be read as a more restrictive conditional claim, that if one accepts their broader analysis of law, then the only conception of equality before the law that has not already been ruled out is the one that is then subjected to critique. Since this reading presupposes that there is a convincing argument against the conception I have developed, we will not be in a position to evaluate whether it would leave their overall argument better off until we have explored the second strategy.

39. Sadurski, supra note 2, at 132 (emphasis in original); see also Hart, supra note 2; Westen, supra note 2, at 550–551; Winston, supra note 2, at 10.

40. The strongest version of the premise holds that all three requirements apply. Indeed, the identification requirement is by itself clearly incapable of making the premise plausible. I include it regardless because it does the argument no harm and some critics seem to appeal to it or something like it.

41. Joel Feinberg, Harm to Others (1984); Joel Feinberg, Offense to Others (1985).

42. See Petersen, Thomas Søbirk, No Offense! On the Offense Principle and Some New Challenges, 10 Crim. L. & Phil. 355 (2016)Google Scholar.

43. Peters, Foolish Consistency, supra note 2, at 2063–2064; see also Miguel, Alfonso Ruiz, Equality Before the Law and Precedent, 10 Ratio Juris 372, 378 (1997)CrossRefGoogle Scholar.

44. Would the arguments do better if we required strong procedural legal equality? Only marginally, if at all; there could still be cases where procedural equality is satisfied but non-comparative requirements are not so long as all cases are treated with equal degrees of non-comparative injustice, and the status of situations where a court begins to treat cases as it ought against a background of wrongful treatment is entirely unaffected by the switch from weak to strong equality.

45. See Peters, Equality Revisited, supra note 2, at 1226 (“The alleged prescriptive scope of nontautological equality extends primarily to cases in which one person, X, already has been treated wrongly according to some nonegalitarian treatment rule, and the question of how to treat another, identically situated person, Y, arises. In such a case, prescriptive equality claims to provide a reason to treat Y similarly wrongly, a reason the nonegalitarian treatment rule does not give us. As we have just seen, that reason is the fact that X already has been treated wrongly. Prescriptive equality, however, does not claim to carry much force in a case in which X already has been treated correctly according to the nonegalitarian treatment rule.”).

46. Peters, Foolish Consistency, supra note 2, at 2070.

47. Id. at 2071; see also Peters, Equality Revisited, supra note 2, at 1249–1251.

48. See Greenawalt, supra note 20, at 1281–1283, for a related critique.

49. Parfit, supra note 25, at 98–99, 110–115.

50. Miguel, supra note 43, at 382.

51. Note that to constitute a more traditional example of leveling down we should instead imagine that the court considers how to treat the case, and is for some reason aware both that all previous treatments have been unjust and of what would constitute non-comparatively just treatment of the case at hand. I avoid this version to set aside potential complications to do with the mental states of the agents involved, such as the intention to discriminate between cases, which some might hold to be morally objectionable in themselves.

52. That is, many will feel at least some pull toward it being better to create situations in which persons are made more equal, or in which morally better persons are better off than morally worse persons, but to properly assess the procedural principle we must set such concerns aside. We can suppose for example that the court substitutes a procedure that unjustly assigns random outcomes in some dimension with a procedure that justly weighs the relevant factors, but that the outcome of that weighing is the same as the random outcome the court would have delivered. Both Greenawalt and Peters note this as a general problem for arguments in favor of the principle, but it seems to me that it is possible to construct examples in which such factors ought to play no part, and in which it would therefore be possible for the proponent of the principle to demonstrate its validity. See Greenawalt, supra note 20, at 1270–1271; Peters, Foolish Consistency, supra note 2, at 2066–2067. The problem is that, as the above example is intended to show, such examples do not in fact support the principle.

53. Kent Greenawalt makes what I believe to be the mistake of supposing the opposite; Greenawalt, supra note 30, at 1173 (“…when treatment is being decided for equal humans sensitive to possible inequalities, the moral power of the formal principle exerts a significant pull against unequal benefits or burdens”) (my emphasis). Note that the problem with that notion is not, I believe, the objection to which Greenawalt responds, that the reasons might be telic rather than deontic (Greenawalt writes “consequentialist” and “deontological”); see Greenawalt, supra note 20, at 1285–1289. The problem is rather that irrespective of whether the negative consequences of perceived inequality ground telic or deontic reasons to avoid inequality, they are not egalitarian reasons.

54. Feinberg uses the specific example of a justice system that metes out non-comparatively unjust punishments in a comparatively just fashion in his discussion. See Feinberg, supra note 5, at 312–316. For a critique of Feinberg's notion of comparative justice that to some extent mirrors my arguments in the present, see Montague supra note 16, at 133; Hoffman, supra note 16. Montague, however, also joins the critics discussed above: “But one who acts in accord with principles of non-comparative justice will deny no one his due, and will automatically meet the requirements of comparative justice. Thus there can be no conflicts between comparative and non-comparative principles relative to actions required by the latter.” Montague, supra, at 136.

55. An anonymous reviewer emphasized that this was their response.

56. Henry Sidgwick, The Methods of Ethics (Hackett 1981) (1907), at 341–342.

57. Nelson, Caleb, Stare Decisis and Demonstrably Erroneous Precedents, 87 Va. L. Rev. 1 (2001)CrossRefGoogle Scholar.

58. Peters, Foolish Consistency, supra note 2, at 2068–2069; Peters, Equality Revisited, supra note 2, at 1252–1253. Note that Peters introduces this as a general objection to the principle. I have reserved it as an objection for here both because I believe the non-permissibility-restricted version is defeated by the leveling-down objection, and because the objection strikes me as stronger in cases where the court merely chooses between permissible alternatives, and the order is therefore more obviously the result of mere chance.

59. Consider that there may well be other reasons that would constitute an answer. It seems perfectly legitimate to answer her complaint by saying either that her preference did not give the court reason to act, or that there was at least one other reason counting in favor of φ₂ that did not count in favor of φ₁. The point is not that the court cannot give an answer to the complaint; it is merely that the answer cannot plausibly be based on procedural egalitarianism.

60. Bernard Williams, The Idea of Equality, in Problems of the Self (2006), at 231.

61. I am not the first to think of this way of defending equality. Kenneth Winston argues in a somewhat parallel way that treating like cases alike may be beneficial in particular contexts; see Winston, supra note 2, at 36–39. Similarly, at least part of Kent Greenawalt's defense of equality is an argument along roughly these lines, and Peters affirms repeatedly that there are “consequentialist reasons” for adhering to the principle of procedural legal egalitarianism; see Greenawalt, supra note 20, at 1285–1289; Peters, Foolish Consistency, supra note 2.

62. It is worth emphasizing that, although the argument is concerned with the consequences of having and reinforcing a certain norm, it is not therefore consequentialist. Since it rests on telic reasons it relies on the positive consequentialist claim that consequences matter to the moral status of an action, but not the negative consequentialist claim that nothing but consequences matter to the moral status of an action.

63. The literature on the nature and function of norms is both complex and contentious. However, the central disagreements concern the mechanisms and origin of norms, that is, how they affect behavior and how they emerge and develop as social phenomena. C. Bicchieri, The Grammar of Society: The Nature and Dynamics of Social Norms (2005); M. Hechter & K.D. Opp, Social Norms (2001); E.A. Posner, Law and Social Norms (2009). Bicchieri, Christina & Muldoon, Ryan, Social Norms, in The Stanford Encyclopedia of Philosophy (Zalta, Edward N. ed., 2011)Google Scholar provides a good overview. On these issues I take no stand.

64. The legal literature contains a number of additional arguments sometimes cited in defense of equal treatment in the context of the doctrine of stare decisis, including benefits of stability and predictability. See Peters, Foolish Consistency, supra note 2, at 2039. Although considerations of space preclude my developing that argument, I believe the three benefits I discuss here to be the most plausible, but if stability, predictability, or other benefits also apply this simply means that the case for the norm is even stronger than I here suggest.

65. See Maitzen, Stephen, The Ethics of Statistical Discrimination, 17 Soc. Theory & Prac. 23, 25–31 (1991)Google Scholar.

66. Schauer, supra note 20, at 599, and Kornhauser, Lewis A., An Economic Perspective on Stare Decisis, 65 Chi.-Kent L. Rev. 63 (1989)Google Scholar both make the similar point concerning the benefits of stare decisis.

67. Heiner, Ronald A., Imperfect Decisions and the Law: On the Evolution of Legal Precedent and Rules, 15 J. Legal Stud. 227 (1986)CrossRefGoogle Scholar develops and formalizes this point elegantly. Note too that, as Lewis Kornhauser observes in a slightly broader context, there are obvious parallels between this argument and Joseph Raz's argument for authority; see Kornhauser, Lewis A., The Economic Analysis of Law, in The Stanford Encyclopedia of Philosophy (Zalta, Edward ed., 2011)Google Scholar; Joseph Raz, Authority, Law, and Morality, in Ethics in the Public Domain (1994). Finally, note that the argument is importantly different from Richard Posner's familiar argument for an efficiency norm in common law; in fact it might be said to be its inverse, in that Posner seeks to justify a norm of efficiency, while I take promotion of efficiency to justify the establishment and reproduction of a norm. Nor is my argument here required to endorse the more controversial premises of Posner's argument, such as his claim for the ethical superiority of a principle of wealth maximization; see Posner, Richard A., Utilitarianism, Economics, and Legal Theory, 8 J. Legal Stud. 103 (1979)CrossRefGoogle Scholar; Posner, Richard A., The Ethical and Political Basis of the Efficiency Norm in Common Law Adjudication, 8 Hofstra L. Rev. 487 (1980)Google Scholar.

68. Goodin, Christian List & Robert E., Epistemic Democracy: Generalizing the Condorcet Jury Theorem, 9 J. Pol. Phil. 277 (2001)Google Scholar.

69. See Kornhauser, supra note 66, at 68–73.

70. Macey, Jonathan R., Internal and External Costs and Benefits of Stare Decisis, 106 Chi.-Kent L. Rev. 93 (1989)Google Scholar, Kornhauser, supra note 66, at 76, and Hellman, supra note 24, at 74–75, all make similar points regarding the benefits of stare decisis. For criticism of that view see Lee, Thomas R., Stare Decisis in Economic Perspective: An Economic Analysis of the Supreme Court's Doctrine of Precedent, 78 N.C. L. Rev. 643, 652–653 (1999)Google Scholar. Nelson, supra note 57, at 55–61, provides further arguments that pertain specifically to a weak doctrine of stare decisis, elements of which would also apply to a sufficiently weak norm of procedural legal egalitarianism.

71. See Hellman, supra note 24, at 65–66.

72. Schauer, supra note 20, at 575.

73. Hellman, supra note 24, at 64–69.

74. Hellman's primary concern is arguing this latter point, that (applying her terminology) epistemic reasons can lead to decisions that are “precedential” because such reasons satisfy the “independence thesis”: the reasons to treat a case C2 similar to a past case C1 persist regardless of whether C1 was decided correctly or wrongly.

75. This is a somewhat broader and less psychological definition than is common in the literature. I intend for it to be complementary rather than opposed to such definitions, but since the point I develop here does not rely on a particular account of the psychological mechanisms that produces the agent's disposition, e.g., whether it is attitudinal or doxastic, I shall rely on the broader definition. Brownstein, Michael, Implicit Bias, in The Stanford Encyclopedia of Philosophy (Zalta, Edward ed., 2015)Google Scholar, provides a good overview of the current debate.

76. Although it is of course consistent with the definition, e.g., that one could openly affirm a prejudice and simultaneously suffer from a bias pertaining to the same group.

77. Rachlinski, Jeffrey J. et al. , Does Unconscious Racial Bias Affect Trial Judges? (Cornell Law Faculty Publications Paper 786, 2009)Google Scholar; Kang, Jerry & Lane, Kristin, Seeing Through Colorblindness: Implicit Bias and the Law, 58 UCLA L. Rev. 465 (2010)Google Scholar; Kang, Jerry et al. , Implicit Bias in the Courtroom, 59 UCLA L. Rev. 1124 (2012)Google Scholar.

78. Hellman makes the related point in favor of according precedent some weight in judicial deliberations, that doing so “forces decisionmakers to engage with a contrary point of view and tak[ing] that view seriously improves decisions.” Hellman, supra note 24, at 74. More pessimistically, we might suspect that bias will affect the court's perception of the reasons at stake enough to distort its perception of which cases are alike, and hence ought to serve as suitable comparisons. If so, the norm may do little good, because the court will distinguish cases rather than challenge its initial assessment.