In Reciprocal Freedom, Ernest Weinrib provides a sustained analysis of the relationship between private law and public law, and in the book’s final chapter, he turns to the rule of law’s significance for private law.Footnote ¹ His is a rich and nuanced conception of the rule of law, whose dimension and consequences far outstrip its import for private law. But that import is my focus here, and given those aims, I begin with only the barest outline of Weinrib’s overall conception of the rule of law.

On Weinrib’s view, the rule of law demands first that the state operate through the law, which he takes to involve issuing laws that are general, able to guide conduct, and non-futile. This is a claim familiar from Lon Fuller’s famous “principles of legality,” which Weinrib incorporates, though he gives the claim a distinctively Kantian cast.Footnote ² Thus, “the state disrespects the freedom of its subjects if it requires them to comply with laws that were secret or retroactive or contradictory or incomprehensible or impossible to fulfill or constantly changing.”Footnote ³ Besides operating through the law, the state must abide by its own laws, acting in accordance with them. In this way, “state power is inherently self-limiting,” which is why the rule of law is the antidote to arbitrary rule.Footnote ⁴ The way that the state constrains itself and avoids despotism is to be organized specifically as a republic with its emblematic separation of powers. So organizing the state helps ensure persons’ equal freedom, because “only in a republic is the state institutionally organized to operate not only through law but in accordance with law.”Footnote ⁵ Legislative, executive, and adjudicative functions must be separated, more specifically, in order to preserve “the integrity of legislation’s claim to obedience.”Footnote ⁶ For only if there is such separation are persons subject (only) to the law as envisioned by the legislature—not some administrator or judge—thus retaining their freedom to do whatever the law does not enjoin. Echoing Joseph Raz’s well-known methodological point that the rule of law must provide a “distinctive basis for assessment,”Footnote ⁷ Weinrib maintains that “viewing the rule of law as dealing with the constitutive aspects of a republican polity preserves for it a distinctive and coherent role.”Footnote ⁸

This brings Weinrib to private law. Given the provenance of private law in the courts, private law submits to the rule of law in and only in adjudication. Moreover, insofar as the rule of law revolves around the constitutive aspects of a republic, private law’s narrow focus on adjudication translates the rule of law’s bearing on private law into a focus on the constitutive aspects of adjudication. So, according to Weinrib, the rule of law demands that private law proceedings be held in a public institutional forum, where claims are impartially heard and then evaluated by exclusive reference to the merits of the two parties’ arguments.

A rule-of-law puzzle seems to arise in its application to private law, though, on account of an apparent conflict between prospectivity, a central feature of the rule of law, and one of the constitutive aspects of adjudication, namely, that it resolves an antecedent dispute and thus cannot but be retroactive. The worry, in short, is that “parties to litigation are held to a norm of which they had no specific notice when the impugned conduct occurred.”Footnote ⁹ Weinrib offers a deflationary response to this worry.

He begins by arguing that the norms that are articulated in private law adjudication are fashioned out of preexisting legal materials: previous decisions, but also the reasoning behind those decisions. This stands in contrast to legislation, which can create a new rule based on novel logic out of thin air. Weinrib next argues that the principle of prospectivity needs to be applied differently to legislation and private law adjudication: “The rule of law should be understood in a way that reflects the distinct kinds of law … to which it applies.”Footnote ¹⁰ Because the norms created through adjudication are fashioned from pre-existing legal material, including norms of legal argument and reasoning, all of which is public, any court’s “decision makes definitive what ought to and could have been done earlier on the basis of the law as it existed earlier.”Footnote ¹¹ This is to say that there is no actual problem of retroactivity after all. It is, in Weinrib’s striking words, “a non-existent problem.”Footnote ¹² As creative as this deflation of the seeming conflict between the rule of law and adjudication is, I do not think it is successful. The Gordian knot cannot just be cut.

Weinrib makes two claims here, one about what could have been done earlier and one about what ought to have been done earlier, based on pre-existing legal materials. Let me take these two claims in turn, starting with the modal claim. Weinrib argues that insofar as all of the legal materials and patterns of legal reasoning to which courts must appeal in resolving a private law dispute before them are accessible to the parties ex ante, it is false to say that compliance with the norm that the court announces ex post is not possible. Retroactivity is avoided, on this picture, because the various elements of whatever rule the court will announce to resolve their dispute were in front of the parties all along. This, however, is too thin a conception of prospectivity and of possible compliance.

Were I either party in the 1916 New York Court of Appeals case, MacPherson v. Buick Motor Company, my attorneys would have made me aware that the privity rule—holding that one’s duty of care is limited by contract—was both the law of the land but also honored in the breach, as it were, depending on the circumstances.Footnote ¹³ New York courts had selectively applied the rule throughout most of the nineteenth and early twentieth centuries. When Benjamin Cardozo rejected the privity rule once and for all in MacPherson, it is true that neither party would have been shocked. After all, previous courts in the same jurisdiction had found for plaintiffs in similar enough cases. Neither the winning plaintiff nor the losing defendant, then, would have found the decision unintelligible. True, the reason MacPherson is reprinted in American tort casebooks is that it was an innovation and the definitive statement of the privity rule’s rejection. But that historically significant fact would not have mattered to the parties. That Cardozo resolutely rejected the privity rule as such, root and branch, had no greater impact on the parties in MacPherson than a narrower ruling would have had, holding more modestly that the privity rule, while still good law in New York, simply did not apply to the facts in the case at bar. The ruling was important for future litigants (and commentators and law professors) because it was innovative.

Still, for the parties themselves, it is hard to get past the thought that the case could have gone either way insofar as both plaintiff and defendant had at least some law on their side. That Cardozo’s opinion was entirely intelligible to the parties’ attorneys—they surely could recognize the existing legal materials, legal logic and all, that Cardozo drew upon to reach his conclusion—does not change the fact that Cardozo took an unsettled question and made the choice to settle it in a particular way. And that when he did, a party who might not have lost the case did in fact lose. This, it seems to me, is something close to the essence of retroactivity, and it is problematic for adjudication’s consonance with the rule of law.

When Weinrib maintains that, in resolving a private law dispute, the court’s “decision makes definitive what … could have been done earlier on the basis of the law as it existed earlier,” then, the “could” is bearing a great deal of weight. It is surely right that a court could have rejected the privity rule earlier than MacPherson did, based on extant legal materials. It is also true that parties could have conducted themselves as if the privity rule were not authoritative. But at a minimum, neither would be a case of norm compliance. For there was no norm. That there was law suggesting that the privity rule did not always apply, or that there are compelling moral reasons to acknowledge duties beyond those voluntarily undertaken, does not establish the existence of an anti-privity rule legal norm.

The weaknesses in Weinrib’s claim about what could be done are amplified when paired with his claim about what ought to have been done. So here I want to emphasize not the modal point but the normative one that he makes simultaneously. It is his position, once again and streamlined, that adjudication “makes definitive what ought to … have been done earlier on the basis of the law as it existed earlier.” This is an ambitious claim.

Even bracketing the problems surrounding the claim of what could have been done earlier, it seems to me exceedingly difficult to connect the fact that something could have been done earlier to a fact about what ought to have been done earlier. Possible norms are not actual or authoritative norms. We typically do not think that anyone—a court or a litigating party—is bound by a merely possible norm, but only by one that has been given authoritative articulation. The problem of retroactivity itself, in fact, depends upon this fact: The problem has bite because, and only once, a court actually does articulate the norm. Until that point, no one is subject to any remedial duty—the secondary obligation is merely possible. There is no problem of retroactivity until a court actually tries to articulate and then enforce a new norm. But once a court tries to enforce a norm against a party, which until its announcement by the court had only been a possible norm, it seems to me that there is a very real problem of retroactivity.

The element of choice in adjudication is not given its due in Weinrib’s discussion. Cardozo’s holding in MacPherson may read as obviously right to contemporary eyes, but it was not inexorable in the way that Weinrib seems to think it was. Were it inexorable, then Weinrib’s claims about what could and should have been done earlier based on extant legal materials would be easier to establish—even obviously true. For we would be dealing with entailment or logical consequence. But while I am no Holmesian, I nevertheless agree with him that the life of the law is not logic.Footnote ¹⁴ In many cases, and especially in hard cases, courts make choices and have discretion about how to resolve a dispute. The reason they make choices, moreover, is that they have choices to make—the outcome is not foreordained by existing legal materials, even if it is constrained by them.

That courts have choices to make especially in hard cases is not new. It is a position most closely identified with contemporary legal positivism. Courts are of course required to settle the disputes before them. They have no choice over whether to do that. In the vast majority of cases, moreover, they also have no choice over how to settle them. The law is mostly determinate. In a significant range of cases, however—and it seems to me that MacPherson is as good an exemplar as any—judges are indeed confronted with a choice. The language of “discretion” is often used to characterize this choice,Footnote ¹⁵ for perhaps it better conveys the way in which judges remain constrained to decide cases both by reasoned judgment (as opposed to flipping a coin, say) and within the confines of the law (they cannot just come to any decision, one that is foreign or antagonistic to all existing law). But assuming those conditions are satisfied, courts are faced with a choice to make in hard cases. It is just this sense of choice to which Hart refers in a (recently surfaced) rejoinder to Dworkin:

It seems to me that choice—suitably constrained, as Hart explains—is unavoidable in hard cases, and I believe that Weinrib underplays its place and importance.

The reason the holding in MacPherson seems obviously correct today, I think, is its moral justification: The idea that the only obligations we have are voluntarily undertaken is itself morally repugnant. I gather that recognition of that fact accounts at least in part for Cardozo’s own conclusion in the case. Now, appeal to this moral fact was made possible in MacPherson precisely because the law was not settled. Had it been settled, there would have been no space ceteris paribus for straightforward moral reasoning to enter. But given the alternating treatments of the privity rule in the caselaw leading up to MacPherson, there was indeed an opening. Once Cardozo rejected the privity rule, he created a new legal norm—one that was based on a moral norm—but the legal norm was not itself the moral norm.

It seems to me that something like this story is necessary to make sense of the problem of retroactivity in the resolution of private law disputes. One could have recognized the moral deficiency of the privity rule long before Cardozo rejected it legally—the privity rule was never morally sound—but that fact is inapposite because it was, until MacPherson, legally recognized and enforced. That it was the law makes the innovation in MacPherson problematic from the rule of law’s point of view. Weinrib just disputes this—it is, recall, a “non-existent problem.” But here I worry that he is not taking seriously enough his own methodological commitments. In the Preface to Reciprocal Freedom, Weinrib contends that “the task of legal theory, as I see it, is to bring to the surface the most pervasive ideas latent in law as a normative practice.”Footnote ¹⁷ It seems to me that there is a very real tension between the adjudication of private law disputes and the rule of law, given the latter’s ostensible commitment to prospectivity, and part of what makes a successful legal theory on Weinrib’s own terms will be explaining this tension, not explaining it away.

Does private law adjudication ultimately comport with the rule of law? If ‘ought’ implies ‘can’, then it seems to me that it does. Private law adjudication cannot be impugned for being inherently incapable of complying with the principle of prospectivity. Adjudication can abide by the rule of law, even if not by its prima facie commitment to prospectivity, because the other elements of the rule of law—the ones by which adjudication can and does abide—are also central to the rule of law. We can therefore say the following three things together: that private law adjudication cannot comply with the rule of law to the extent that (prospective) legislation can; that adjudication nevertheless respects enough of the other elements of the rule of law, and indeed all of the elements that it can respect, to abide by the rule of law; and finally, that the elements to which adjudication can and does conform are important and indeed obligatory. Private law adjudication may be retroactive, but that does not undermine its claim to abide by the rule of law, nor does it give us reason to think the remaining features of the rule of law are not worth living up to. But it does give us reason to lament the fact that a central case of law-making cannot maximally abide by the rule of law, insofar as adjudication cannot but be retroactive. In my view, this conception of the relationship between private law adjudication and the rule of law has the key virtue of preserving the felt tension that prompts Weinrib to examine the relationship in the first place.