1. Two irreconcilable positions?

The conventional understanding, succinctly summarised by the Scottish Law Commission, is that there are “two competing theories” of the common law judicial method that “clash.”Footnote ¹¹ There is the “declaratory theory,” which reflects the idea that “the judges do not ‘change’ the law but merely ‘discover’ and ‘declare’ it.”Footnote ¹² And there is the “change theory,” which reflects “the ‘common sense’ truth that certain judicial decisions, for all practical purposes, do indeed change the law.”Footnote ¹³ The conceptions go by different labels: the “classical common law theory” versus the “positivist common law theory;”Footnote ¹⁴ the idea that judges find law versus the idea that they make it.Footnote ¹⁵

These two conceptions of judicial decision-making are considered to be fundamentally at odds. They underlie one of the great debates in modern jurisprudence—long a popular theme of dinner speeches and scholarly inquiry.Footnote ¹⁶

The idea that judges necessarily do not make law, but instead reveal and expound ever-existing law, is derived from the writings of Matthew Hale and William Blackstone. Hale wrote that judicial decisions, although binding “as a Law between the Parties thereto” in respect of their dispute, “do not make a Law properly so called (for that only the King and Parliament can do).”Footnote ¹⁷ Judges were “upon their Oaths to judge according to the Laws of the Kingdom.”Footnote ¹⁸ Hale regarded judicial decisions as “less than a Law” albeit valuable as precedent, “Expounding, Declaring, and Publishing what the Law of this Kingdom is,” at least until such judgments “are by the same Law revers’d or avoided.”Footnote ¹⁹

Blackstone characterised the law revealed in judicial precedent as evidence of a “permanent rule,” which a subsequent judge by virtue of their oath of office cannot “alter or vary from, according to his private sentiments.”Footnote ²⁰ The judge’s duty was to determine the law in each case “according to the known laws and customs of the land,” which meant that the judge’s role was not “to pronounce a new law, but to maintain and expound the old one.”Footnote ²¹ Blackstone certainly recognised that given precedents were not immutable and were sometimes overruled. He described this phenomenon as a matter of correcting the evidential record of law, rather than making law anew:

This conception of the judge as declarator of the law was derided by Jeremy Bentham and John Austin, who considered the idea of judicial law-making irrefutable. Bentham accused Blackstone of treating the common law as “an abstruse and invisible quiddity.”Footnote ²³ Austin dismissed the Blackstonian account as a “childish fiction,” which he took to entail judges pretending:

For Austin, ‘judiciary law’ entailed the exercise of legislative power through another means: “the sovereign, acting in the capacity of a judge, may make [laws] in the indirect mode of judicial decision.”Footnote ²⁵ Justice Oliver Wendell Holmes Jr. memorably expressed this perspective of the common law as being “not a brooding omnipresence in the sky but the articulate voice of some sovereign or quasi-sovereign that can be identified.”Footnote ²⁶ From this perspective, judges necessarily make law when they rule upon novel issues for which there is no binding precedent or statute. They make law when overruling precedents delivered by their predecessors. They certainly do so when quashing statutory and regulatory provisions promulgated by legislators. The common law has not remained frozen since the Saxon era. Judges have brought about its change.Footnote ²⁷ The common law is judge-made law. So the declaratory theory must be, as Lord Reid later charged, a “fairy tale[].”Footnote ²⁸

The rivalry between the conception of the judge as law-finder versus law-maker has hardly dwindled since. The debate might seem less intransigent today, in the sense that jurists tend to think of judges as pivoting between the two conceptions by ‘finding’ law in run-of-the-mill (easy) cases and ‘making’ it in novel (hard) cases. But it is still framed as a binary choice between the declaratory theory and judicial law-making. For example, the Supreme Court of Canada in Canada v Hislop, influenced by Lord Nicholls’ speech in Re Spectrum Plus, contrasted situations in which judges “are applying the existing law” with “situations in which judges are fashioning new legal rules or principles.”Footnote ²⁹ The declaratory theory was said to operate in the former situations, whereas “[w]hen the law changes through judicial intervention, courts operate outside of the Blackstonian paradigm,” and so “the declaratory approach is inapt.”Footnote ³⁰ Thus, even while considering that a judge might shift between the two conceptions from case to case, the Court framed the conceptions as rival: In any given situation, a judge either declares the existing law or makes new law. There is no in-between.

But is the declaratory conception of the common law necessarily rivalrous with the idea that it is made? Several scholars have come to the declaratory theory’s defence, arguing that its classical conception has always been misrepresented: that “Bentham and Austin were attacking a straw man.”Footnote ³¹ The conception of the common law as immutable and eternal seems to have been a “caricature” of Hale and Blackstone’s jurisprudence.Footnote ³² It cannot be what Blackstone actually believed, as throughout the Commentaries Blackstone addressed and even celebrated “the ingenious inventions of common law’s creative practitioners.”Footnote ³³ In the end, it may be that “[t]he claim that Blackstone regarded law as fixed for all time, unchangeable and merely awaiting discovery, is a calumny.”Footnote ³⁴ These accounts propose that the declaratory theory is not properly understood as a theory of immutability and stasis. It actually is a theory of how the common law changes.Footnote ³⁵ It represents a conception of judicial fidelity to law, according to which judges by oath subordinate their private views to the constraints of a system of law, determining in each case the law prevailing, if not for eternity, at least at the time the parties’ dispute arose, as construed through adjudication of the dispute rather than imposed as a matter of judicial legislation.

These responses suggest there is reason to reconsider the declaratory theory’s assumed incompatibility with the idea of the judge as law-maker. Indeed, Allan Beever has contended that “properly understood” the declaratory theory is reconcilable with the positions adopted by some of its “most vituperative detractors.”Footnote ³⁶ That is the enquiry I take up in this article. To guide my path, I turn to four twentieth-century inheritors of this debate. On the one side, the respective jurisprudence of H.L.A. Hart and Karl Llewellyn conceptualised judges as exercising law-making power and certainly did not think that the common law never changes. On the other side stand their interlocutors: Ronald Dworkin and Lon Fuller. Dworkin’s jurisprudence is generally taken to have reflected a “modern,”Footnote ³⁷ “sophisticated,”Footnote ³⁸ and “prominent”Footnote ³⁹ account of the declaratory theory. Fuller, while rejecting the idea that judges never make law, ultimately “proved to be a staunch defender of the declaratory theory of the common law, seeing in it an account of law and its normativity consistent with his own.”Footnote ⁴⁰

My suggestion is that by seeking out the common ground across these jurists’ perspectives on judicial decision-making, we might find a path to reconciling the idea of judicial law-making power with the essence of the declaratory theory. Along the way, I endeavour to expose some common misconceptions of these jurists’ respective positions. The destination I reach is a renewed regard for the view of adjudication advanced by Lord Goff of Chieveley, which sought to reconcile the two ‘competing’ conceptions of judicial method through what could be considered a declaratory theory of judicial law-making. The judicial role is to declare what the law fairly was at the time a dispute arose by determining how it ought to have been understood by the parties all things considered. It is not to rule based on what the judge considers the law should have been or should be from now on.

2. Judges exercise law-making power

Bentham and Austin’s perspective of judicial law-making is a hallmark of legal positivism, which understands law to subsist in the sources that are recognised as authoritative in a society. The law is made—posited—by people in requisite positions of authority.Footnote ⁴¹ H.L.A. Hart advanced the idea that legal rules invariably have a core of certainty and a penumbra of doubt. Much of the law is clear and readily discernible, but because legal language has an inherent vagueness, law is invariably imperfect and incomplete. These characteristics of law account for legal disputes, of which we can differentiate two kinds: There are regulated disputes, which can be resolved according to some discernible common law or statutory rule that regulates the outcome; and there are unregulated disputes, which have no answer in positive law. Unregulated disputes arise from incompleteness (‘gaps’) in the law. Positivism recognises room for judicial discretion in the resolution of both classes of dispute.Footnote ⁴² A judge might identify grounds not to follow some relevant precedent to resolve a regulated dispute (recognising that law is imperfect), and instead distinguish or overrule it.Footnote ⁴³ Such grounds could include, for instance, that the precedent conflicts with some other legal authority, or that it is unclear how it should bear upon a particular case. In departing from precedent, the judge makes law. In the resolution of unregulated disputes, the law-making role of judges is obvious: The judge makes law by filling in the gaps. In these cases the judge’s decision cannot be determined according to any legal sources, and so it must follow from the extralegal values or principles that the judge considers most pertinent in the circumstances. The positivist theory of adjudication affords judges “strong discretion” to look beyond the law to resolve hard cases.Footnote ⁴⁴

The American legal realist Karl Lewellyn pushed the claim of judicial discretion further. Legal realism understands law to be determined not by recognised legal materials (as positivism contends), but by the social practice of law.Footnote ⁴⁵ The realist claim is that legal doctrine as presented in cases and statutes cannot determine the content of law because legal doctrine is ultimately indeterminate. For every rule or interpretation, there can be found a counter-rule or counter-interpretation. In all cases there is room for judicial choice in construing the law. In this sense the law exists in a state of “flux.”Footnote ⁴⁶ Judges act as law-makers when they exercise choice to decide legal disputes. In Llewellyn’s words:

Thus, the realist perspective is that judging necessarily is law-making.

In short, both of these positivist and the realist perspectives hold that judges unavoidably exercise law-making power. By way of contrast, it is sometimes said that natural law and interpretive theories of adjudication reject this conception of judicial decision-making because they hold to the declaratory theory. The question here is whether these jurisprudential perspectives are incompatible with the conception of the judge as law-maker—which is certainly how the classical natural law perspective has been received. According to that perspective, the common law is not determined by positive sources but found through moral reason. Departures from precedent are seen as judicial error-correction, rather than as the judges changing the law: “the common law … works itself pure by rules drawn from the fountain of justice.”Footnote ⁴⁸

The precepts of this classical perspective came to be revised in the twentieth century in response to the critiques levelled by Bentham and Austin. Lon Fuller developed an account of the internal morality of law that centred on the process of law-making, propounding basic criteria that must be met in order for a rule to be acceptable as law.Footnote ⁴⁹ In setting out his account, Fuller surprised his critics by embracing the notion that judges are indeed law-makers. He thought that when novel disputes arise, the judge must render a (novel) decision. In the absence of binding precedent, judges must choose among possible options, and in so doing they exercise discretion. Fuller compared the creativity of the judge to that of a poet, “who respects the inheritance of his native language in the very act of exploiting its resources for novel uses.”Footnote ⁵⁰ Although the judge’s discretion is constrained, a novel judgment is ‘new law’—and what’s more, when applied to parties to a dispute, it can be understood as retrospective law-making.Footnote ⁵¹ This perspective came to be shared by another prominent natural law theorist, John Finnis, who considered that when a judge applies a moral standard to resolve a new dispute the judge is “in a sense ‘making’ new law.”Footnote ⁵² The key insight here (to which I will return) is not that judges lack law-making power, but that judges do not make law in the manner of legislators. Judges make law in a particular context (adjudication) and a particular way (interpretation).Footnote ⁵³ In each case that comes before them, judges must discern an interpretation that, all things considered, best resolves the parties’ dispute in its time and context. A judicial interpretation will be novel whenever it responds to facts that are without analogy in precedent or when it invokes reasons that have not previously been judicially endorsed (but that nevertheless best resolve the dispute). Such novel judgments are fairly understood as judge-made law.

In debates with both Hart and Fuller, Ronald Dworkin defended an interpretive theory of adjudication which he differentiated from the positivist and the natural law accounts. Dworkin understood judicial decision-making to be constrained and shaped by the institutional practice of judging and by principled interpretation. Dworkin became associated early on with the idea that judges do not make law.Footnote ⁵⁴ He defended the claim that it is “the judge’s duty, even in hard cases, to discover what the rights of the parties are, not to invent new rights retrospectively.”Footnote ⁵⁵ He argued that judges must interpret the law as if it were coherent—a seamless web.Footnote ⁵⁶ On this view, an ideal judge can properly interpret the law to reach the answer that best fits, and is justified by, the law as a whole. This led to the inference that principled judging does not entail making law. But Dworkin’s later writings clarified his jurisprudence as a theory of how judges develop the law. Dworkin did accept that judges have a creative role—that the common law can be seen to change and that it is judges who control such change. Echoing Fuller’s artistic judge, Dworkin offered the analogy of judges developing the law like authors of a chain novel: They add to the law in a manner that respects what has come before.Footnote ⁵⁷ In hard cases, the judge must exercise discernment in reasoning from legal principles to legal conclusions. Dworkin recognised that in adjudication, “fresh and challenging issues” do arise that may require a judge to “choos[e] one interpretation over another of a much contested line of precedents.”Footnote ⁵⁸ The judge’s choice over how the law is to develop is guided and constrained by the principle of coherence—but it is still a choice. The resolution of such cases “will often, perhaps usually, be controversial, so that a fresh judgment is needed.”Footnote ⁵⁹ If a precedent must be overruled, it is because the judge views its alternative as better cohering to the body of law as a whole. Overruling, on this account, is a feature of interpretation. It is not ‘legislating’, but it can be understood as a ‘weak’ form of judicial discretion.Footnote ⁶⁰

Dworkin’s reticence to conceive of judicial discretion in terms of ‘law-making’ at first blush seems to undermine my claim that we can find shared foundational features of the common law method across these four jurisprudential perspectives. Certainly, Dworkinian interpretivism is the most closely aligned with the classical Blackstonian account. On the other hand, as will be seen in the subsequent sections, the reason Dworkin eschewed the language of judicial law-making was precisely because he endorsed the other five features I explore as foundational to principled judicial decision-making. Dworkin conceived of judicial discretion as constrained, exercised in response to legal instability, operating with hindsight, incomparable to legislating, and entailing elucidation of the law. In that regard, as will be shown, Dworkin’s conception of adjudication does find common ground with the accounts of Hart, Lewellyn, and Fuller.Footnote ⁶¹

A further reason to think that there is underacknowledged common ground between Dworkin and his detractors on the nature of the judicial method is that hostility to the notion of judicial law-making does not seem to be an essential hallmark of Dworkin’s legal interpretivism. For example, Peter Jaffey has demonstrated how the idea of (constrained) judicial law-making can be reconciled with a Dworkinian perspective. Jaffey builds upon Dworkin’s thesis to elaborate a theory of how the common law changes. For Jaffey, “[i]t is indisputable that in the common law the judges do make law.”Footnote ⁶² Jaffey suggests that because on the interpretivist account courts are continually in a process of refining rules to cohere with the whole body of law—through distinguishing, extending, subsuming, and overruling—the rules laid down by the courts are essentially provisional:

Each chapter in the chain novel of the common law provides guidance to future judicial authors, but how a written chapter is interpreted and developed upon is a matter for those future authors. To say that judges exercise ‘weak’ discretion is to suggest that judges have a modest power to make law. This account of Dworkinian interpretivism understands judges to “make law as they declare it” and to do justice by applying their best interpretation of the law in the context of adjudication.Footnote ⁶⁴ Understood in this way, it is possible to find common ground across these rival perspectives around the idea that judges exercise law-making power.

3. Judicial law-making power is constrained

The insight that judges exercise law-making power is of limited import without an appreciation of how this power is exercised. As Patrick Atiyah cautioned when reflecting on the “highly complex business” of judging, “[t]o recognise that the judge has a creative role does not imply that he can create exactly what he pleases.”Footnote ⁶⁵ There are constraints on judicial law-making that serve to ensure judge-made law retains “some degree of predictability and certainty.”Footnote ⁶⁶ The Hartian, Llewellian, Fullerian, and Dworkinian views of adjudication outline rival accounts of what constrains judicial law-making. They also disagree over the extent of judicial constraint. These remain contentious issues for jurists and judges alike. As one Supreme Court judge has commented, “judges recognise that there is a boundary to judicial law-making but there is no consensus as to where it is.”Footnote ⁶⁷ My aim here is not to takes sides in the jurisprudential disputes over the place of rules, principles, policies, and social norms in empowering or constraining judicial creativity. The claim here is less ambitious. It is more descriptive than normative. It concerns the basic characteristics of the common law judicial method. The insight is that the rival theories of adjudication share important common ground in recognising that judges are constrained in how they resolve disputes. Certainly, none of these jurisprudential accounts hold that judges have complete discretion in their decisions to say what the law is.

Although Hart’s positivism attributed to judges ‘strong’ discretion to resolve hard cases, he also recognised that the context of adjudication significantly constrains judicial discretion. The constraints are structural and substantive. The interstitial nature of making law through adjudication limits the opportunities for and scope of innovation from the bench. Judicial creativity is also constrained by the requirement to give reasons that justify the court’s holding. In formulating their reasons, judges do not start from a blank slate.Footnote ⁶⁸ Rather, even in unregulated cases judges reason by analogy “so as to ensure that the new law they make, though it is new law, is in accordance with principles or underpinning reasons recognized as already having a footing in the existing law.”Footnote ⁶⁹ As Joseph Raz echoed, there are no “wholly” unregulated cases; judicial law-making is “circumscribed and hedged by legal limitations.”Footnote ⁷⁰

In his ‘Postscript’ to The Concept of Law, Hart suggested that this account of the limits on judicial creativity was not so far removed from Dworkin’s interpretivist account, which grounds constraints on legal interpretation in the principle of coherence.Footnote ⁷¹ When judges exercise interpretive choice, they must choose the interpretation that best fits the law and is justified by law. Interpretive theory is principally a theory of constraint on judicial law-making.

The core disagreement between Hart and Dworkin on this issue is about the nature and extent of judicial discretion in interpreting the law in hard cases. The positivist conception of strong judicial discretion and the interpretivist conception of weak discretion are best understood as different points along a spectrum. Both points are, however, closer together than are the two (implausible) poles—namely, that in hard cases judges either have the discretion of a legislator (a caricature of legal realism), or they have no discretion at all (a caricature of natural law theory). As William Lucy observes, both the positivist and the interpretivist accounts “take seriously adjudicators’ claims that they decide hard cases relatively constrained by standards relatively determinate of the dispute before them.”Footnote ⁷² Neither theory accepts as legitimate deciding cases according to arbitrary criteria—for example, the flip of a coin. While there may be disagreement over the possibility of right answers to hard cases, there is no question that there can be wrong answers.

Other prominent theories of adjudication do not deny this proposition either. Fuller’s natural law account, as we have seen, did not deny judicial discretion. Fuller did not labour under the illusion of the “Noble Dream” that judges never make law.Footnote ⁷³ He accepted that judges have a creative role in developing the law, while also elaborating on how judicial creativity is meaningfully constrained by the adjudicative context.Footnote ⁷⁴ Legal realism, meanwhile, is often portrayed as “a claim about judges’ exercising unfettered discretion in reaching results based on their personal predilections.”Footnote ⁷⁵ Such an extreme jurisprudence—“that judges always make and never find the law they impose on litigants”—was Hart’s “Nightmare.”Footnote ⁷⁶ But as Hart himself recognised, legal realism in the tradition of Llewellyn resists this view of judges as unconstrained law-makers.Footnote ⁷⁷ Rather, Llewellyn saw the social practice of law as a significant constraint on judicial creativity. It need not follow from legal realism’s critique of doctrine as indeterminate that law generally is indeterminate because, as Hanoch Dagan explains, “legal realism does not equate law with black letter law.”Footnote ⁷⁸ Law’s stability and predictability derive instead from the legal community’s shared understanding of the content and applicability of laws. Part of this understanding is that judges do not create law from a blank slate; rather, as Llewellyn put it:

Judges make law to ‘fit’ the corpus of law as they encounter it. At any given time and place legal doctrine has a “felt predictability” arising from “the convergence of lawyers’ background understandings” of it, giving law a “significant measure of stability.”Footnote ⁸⁰ Contrary to the nightmarish caricature, the Llewellian realist conception of adjudication is one of fettered discretion.

4. Judicial law-making responds to legal instability

Our first two common propositions are that judges make law but do so in a constrained manner—with the substance of the constraints remaining contested. This descriptive claim is sometimes countered by the normative critique that judges should refrain from exercising law-making power. Bentham’s criticism—still frequently echoed today—was that judicial law-making is objectionable because it creates instability by upending people’s expectations about what the law is.Footnote ⁸¹ When judges overrule precedent, quash rules, and articulate novel doctrines in the course of deciding cases, they contaminate the law with uncertainty. How can people follow the law if judges can change it ex post facto?

This critique has been influential, but the problem it presents is not as straightforward as it appears. First, the critique seems to assume a dichotomy between judges applying pre-existing law and judges making new law. The complaint of unfairness lies in the ‘new’ law replacing the ‘old’ law after the fact. However, as I signalled in section 1, this binary view can and should be challenged. If judges are understood to make law interstitially, within the context and constraints of adjudication, then novel judgments are invariably informed by the past rather than representing a clean break with it. Judicial law-making can be understood as a matter of the judge using hindsight to elucidate today what the law fairly was at the time a dispute arose. This perspective elides the dichotomy between ‘old’ and ‘new’ upon which the ex post facto complaint seems premised. It embraces continuity as a feature of judicial development of the law.Footnote ⁸² I return to this idea in section 7.

Secondly, there is a sense in which the critique gets the problem backwards. The problem is that even formal laws can become unstable and incapable of clearly guiding behaviour.Footnote ⁸³ Stability in law can be understood as “a state of affairs in which the content of the law of one country is settled over a considerable amount of time.”Footnote ⁸⁴ It too is best understood not as a binary but a spectrum—“a matter of degree.”Footnote ⁸⁵ Whether a law (or its interpretation) can be considered stable depends not only on whether it has been authoritatively stated but on many other factors, not least the likelihood that the law as stated will survive the next case that encounters it. Instability can arise when a law comes into conflict with some other relevant legal rule or standard, is questioned in obiter dicta, or is criticised in scholarly literature. As conflicting authority over a law grows, its authoritative force weakens. In this way, black letter law can fall out of accord with peoples’ expectations of what the law actually requires.Footnote ⁸⁶ Judges do not pronounce on laws that nobody has challenged. It is when people cannot agree on what law governs (or what the law is) that they bring their dispute before the courts. The judge must give an answer to resolve the disagreement, and may be justified in departing from an impugned rule or understanding of law when there is already tension between the rule or understanding and other authority. The value of keeping the law the same is undercut by the divergence in the parties’ expectation and reliance interests—disagreement that precedes and precipitates the judge’s decision.Footnote ⁸⁷ In this context judicial law-making is not simply a cause of instability in the law; it is a response to instability.

Hart maintained that when two rules are in conflict (or when rules and principles are in conflict), there may be no clear answer. There will be, instead, a penumbra of doubt over what the law is. Doubts about a rule’s authoritative force make for a hard case. The judge must exercise discretion to determine the law.Footnote ⁸⁸ Judicial intervention rarely comes as a surprise, however. A rule’s demise is typically signalled in advance by obiter dicta or scholarly critique.Footnote ⁸⁹ In such circumstances, Raz considered, “[t]he law is still formally settled but people are put on notice not to rely on it. The ‘vice’ of retroactivity is here avoided. Instead, we have the disadvantages of uncertainty which apply to all unregulated disputes.”Footnote ⁹⁰ The ‘vice’ of retroactivity is avoided because no stable rule is undermined by a subsequent overruling. In this context the law can be considered already unstable; judicial law-making then serves to stabilise the law.

The early realists understood judicial law-making as serving a stabilising role. Llewellyn queried why judges would ever hesitate to give “warning at a suitable juncture that a particular rotting tree of law is tottering to its fall.”Footnote ⁹¹ Concealing the instability of a precedent is futile. Skilled lawyers can recognise when a precedent has been so eroded that it is likely to be overruled or abandoned, and so know not to place great reliance on it. Indeed, when a conventional understanding of a precedent appears vulnerable to challenge, skilled lawyers will corral arguments against it to precipitate its demise.Footnote ⁹² Ill-informed parties, however, are liable to be caught off-guard if judges conceal their misgivings over an unstable precedent, and they will place unfounded reliance upon it. If the precedent comes to be overruled or abandoned, these parties will complain of activist judges disrupting their reliance interests. The realist response would be that their reliance was unfounded. The precedent was disputed and uncertain. It was unstable. The situation could not be improved by judicial obfuscation of the precedent’s prospects; it could only be improved by judicial intervention aimed at stabilising the law.

Fuller prized legal stability as a foundational feature of the rule of law. A legal system cannot function when rules are contradictory or change frequently.Footnote ⁹³ Fuller also recognised that in a common law system, where often rules are not laid down in advance but are determined in the course of adjudicating disputes between parties, the development of legal doctrine can appear “very messy” and dependent upon “accidents of litigation.”Footnote ⁹⁴ Yet for Fuller, the common law system was not as unstable as this description might imply. That is because a continuous task of the common law judge is to decide the cases that come before them in a manner that maintains “some systematic order in the law.”Footnote ⁹⁵ In resolving new controversies, the judge strives to bring stability to the law. This explains the common law’s penchant for precedent. Fuller understood precedent neither as a straitjacket on judicial law-making nor as a façade that obscured it. Precedent operates more like a shared language that helps to structure the law, enabling each generation of lawyers to understand and appreciate what they have inherited without closing off the potential for innovation in the law when the circumstances warrant it.Footnote ⁹⁶ The innovative judge maintains stability in the law not by casting off the force of precedents but by building upon their systemic elements in response to new problems.

Dworkin grounded the rationale for departing from precedent in the principle of coherence. When parties act upon a view of the law that conflicts with other precedents and principles in the law, the judge may have a duty to decide the parties’ case on the basis of law that some of them might not have subjectively anticipated in advance. This may be so even if the parties’ view of the law was widely held at the time.Footnote ⁹⁷ A precedent that does not cohere with other decisions is unstable, and once its incoherence is exposed, judicial overruling may be justifiable or even inevitable.Footnote ⁹⁸ The rule-of-law argument against overruling is that parties must be able to rely on settled rules and not be taken by surprise as to what the law requires of them. But in hard cases, reliance interests are weak, since what the law requires is unclear. Opposing parties will corral competing arguments to support their rival interpretations. When a point at issue is doubtful, a court may not be able to avoid upsetting the expectations of one or another of the litigating parties. It must determine what the law is in order to decide the case. One interpretation will be stronger than the other, and “if the court decides that on balance the plaintiff’s argument is stronger, then it will also decide that the plaintiff was, on balance, more justified in his expectations.”Footnote ⁹⁹ On this view, courts must always reach the right answer even when that entails overruling a prior precedent. The precedent is overruled because it was unstable and reliance upon it could not be justified. It is therefore not counterintuitive to see overruling in such cases as fostering the law’s stability.Footnote ¹⁰⁰

5. Judicial law-making is retrospective

The idea that judges act as law-makers in order better to stabilise the law prompts the question of whether the manner of judicial law-making is justifiable. Lord Diplock worried that “whereas law should be knowable in advance, judge-made law is law made in arrear: it is known only after it is broken.”Footnote ¹⁰¹ That is to say, judges make law retrospectively. This is widely recognised by jurists. It is also widely considered that retrospective law-making is unjust. It violates generally accepted accounts of the rule of law—including the accounts advanced by Hart, Fuller, and Dworkin—that new laws should only operate prospectively.Footnote ¹⁰² No one can be guided by a law that did not exist at the time they acted. From these observations, some scholars conclude that judicial law-making by its nature sits in fundamental tension with the rule of law.Footnote ¹⁰³

Yet, such a conclusion does not naturally follow from the jurisprudence of our four preeminent theorists. It would be quite surprising if it did—that is, if the jurisprudence of Hart, Llewellyn, Fuller, and Dworkin impugned the propriety of judge-made law generally. The retrospectivity concern is rather more nuanced: The theories of adjudication surveyed in this article recognise retrospectivity as both justified and inherent in adjudication.Footnote ¹⁰⁴ That is because disputes over which law governed some past action can only be resolved by judges determining the law with hindsight. Retrospectivity is a feature of judicial reasoning that serves to ameliorate the problem of legal instability. In this sense, the overruling of rules that provide no adequate guide to conduct is compatible with the rule of law.Footnote ¹⁰⁵

Moreover, retrospectivity may be thought to engender restraint in the resolution of disputes. The reality that a judge’s decision bears upon the parties raises the possibility of disturbing the basis on which past transactions and arrangements have been made, which tends to encourage sober judicial reasoning and discourage wanton departures from precedent.Footnote ¹⁰⁶ Thus, retrospectivity as a feature of the common law method serves to moderate judicial law-making.Footnote ¹⁰⁷

In his Postscript, Hart clarified his position that when judges resolve hard cases with hindsight, they inevitably apply newly declared law retrospectively to the parties. He thought it cannot be avoided. Retrospectivity is nevertheless justified, because it is the only means to resolve the uncertainty endemic in hard cases. The judge’s intervention does not usurp any stable rights or expectations of the parties because the legal conditions in which the parties interacted were already unstable. (That is so even if the parties did not subjectively appreciate the precariousness of their situation at the time.) The judge looks back to settle pre-existing uncertainty over what the law is and was.Footnote ¹⁰⁸

Legal realism, rather than perceiving cases to be either easy or hard, regulated or unregulated, understands law to be in a state of dynamism. Through the adjudication of cases, judges are continually revisiting and modernising the law with a view to improving it.Footnote ¹⁰⁹ Their conception of law is both backward-looking and forward-looking. As Holmes surmised, “[i]n order to know what [the law] is, we must know what it has been, and what it tends to become.”Footnote ¹¹⁰ The process of assessing and restating the law simultaneously promotes certainty and stability while exposing a persistent degree of unpredictability and retrospectivity. Dagan, drawing upon Llewellyn, construes this balance optimistically. It encourages people to anticipate the law’s development. The law in the books is not all there is. The context in which judges can be seen to make law encourages “a healthy expectation that law should work itself pure.”Footnote ¹¹¹

Fuller observed that the retrospective effect of judicial law-making is almost always taken for granted, and queried why it is “regarded as being exempt from the taboo against retrospective laws.”Footnote ¹¹² He gave a two-fold answer. First, there is no feasible alternative: Judges must decide disputes and can only do so from the perspective of hindsight. Attempts to separate judicial law-making from its adjudicative context “would only compound the difficulties” by injecting hardship, cost, and confusion into the functions of the courts.Footnote ¹¹³ Secondly, adjudication is a reflective and collaborative process in which litigants meaningfully participate. Adversaries present their respective cases with the understanding that the judge’s decision will resolve their dispute. Retrospectivity satisfies that understanding. This adjudicative context justifies retrospectivity in a manner that the legislative context does not. Whereas judge-made law emerges interstitially from (past) cases, legislatures make general rules in anticipation of future issues. Statute law does not depend on past disputes and so it should not govern such disputes. Retrospective statute-making tends to offend the rule of law in a way judicial decisions retrospectively settling disputes between parties do not.Footnote ¹¹⁴ For Fuller, these two types of law are not at all equivalent. Within the institutional context in which courts develop the common law and interpret statutes, “in the run of cases a retrospective effect is and should be taken for granted.”Footnote ¹¹⁵

Cutting against this grain, Dworkin’s interpretive account expressly denied that principled judicial decisions are in substance retrospective law-making. At first glance this seems like a serious obstacle to the thesis of this article. It is also a curious obstacle, given that the Dworkinian theory emphasises the adjudication of (past) cases as the material context in which judges interpret the law. How can this context avoid retrospectivity? For Dworkin, the basis of the judge’s decision is found in the principles already established in the body of law generally. Through interpretation, a judge can reach the best expression of a law. Hindsight will sometimes bring to light arbitrariness or unjustifiability of previously recognised rules or distinctions. Such rules or distinctions must be reinterpreted to cohere with the law generally. This means that when past precedents are found to be wrong, judges may depart from or overrule them. According to Dworkin, the reinterpreting judgment will seem novel. It may involve recognising a law that had not previously been recognised before. But because it is driven by pre-existing principles, the decision is not substantively retrospective.Footnote ¹¹⁶ Rather, it is corrective. In resolving hard cases, “judges do not violate the rule of law when their decisions are correct.”Footnote ¹¹⁷

Building upon Dworkin’s interpretive account of adjudication, Jaffey argues that overruling ‘wrong’ precedents can be reconciled with the idea of justified judicial retrospectivity. He suggests that Dworkin’s disagreement is less one of substance than of form and terminology—Dworkin eschews the term ‘retrospective’, seemingly in order to steer clear of the language of legislating. But Dworkin accepts that judges interpret the law with hindsight and that they have a creative function. Interpretivism need not disregard the novelty of novel precedent. On the Dworkinian account, the ‘novel’ judgment is not legislation because it is grounded in pre-existing principles that govern the case at hand. But the recognition of pre-existing principles that are discoverable at a later time reflects the essence of retrospectivity—discerning the law with hindsight. Jaffey elaborates on how judicial retrospectivity plays a central role in interpretive theory:

On the interpretivist account, courts must determine and apply the best rule to each case that comes before them. Sometimes that will mean applying a new rule to parties retrospectively—in the form of “a fresh judgment.”Footnote ¹¹⁹ The rule is ‘new’ in the sense that it had not previously been articulated, not in the sense that it could not previously have been anticipated or discerned by the parties from the broader legal context in which they were operating. It is because the new rule is the best interpretation of the law that governed the parties’ past dispute at the time it arose that justice requires the court to apply it in the case at hand.Footnote ¹²⁰

6. Judicial law-making is not legislative

Dworkin’s reticence to acknowledge judicial law-making and its retrospectivity stems from his conviction that courts are not “deputy legislators.”Footnote ¹²¹ That conviction is well-founded. It is also shared by other prominent theories of adjudication. Judges make law, but not as legislators: This distinction is too often ignored or denied. Then-Lord Justice Diplock took the distinction to task in a provocatively-titled address at the University of Birmingham, in which he advanced the claim that “Courts by the very nature of their functions are compelled to act as legislators.”Footnote ¹²² Legal scholarship is replete with assertions, variously invoking Hartian positivism or Llewellian realism, that it is “a plain fact that judges legislate, and should legislate;”Footnote ¹²³ that “courts are necessarily legislative bodies;”Footnote ¹²⁴ that “[l]egal change through overruling is no different in outcome from legal change through statutory amendment or repeal;”Footnote ¹²⁵ and that “the courts make law, just as legislatures and agencies do.”Footnote ¹²⁶

Such contentions are the legacy of Bentham and Austin. But they do not encapsulate the perspectives of Hart or Llewellyn—nor, as is perhaps more apparent, of Fuller or Dworkin. While Hart did sometimes describe the juridical techniques of distinguishing and overruling precedent as “forms of legislative activity,”Footnote ¹²⁷ he emphasised that the constrained and interstitial nature of judicial law-making was substantively “different” from legislative law-making power.Footnote ¹²⁸ He expressly rejected “the Nightmare view of the judicial process as mere crypto-legislation.”Footnote ¹²⁹ He thought the very notion was based on a misunderstanding of Llewellyn and others in the legal realist movement.Footnote ¹³⁰ Llewellyn placed considerable weight and attention on “the art and craft of the judge’s office,”Footnote ¹³¹ the nature of the judge, and their distinction from the legislative office and role.Footnote ¹³² He did not seek to expose myths surrounding judicial power by turning to myths conflating it with legislating.

Fuller admonished the claim that courts act like legislative assemblies and detailed no fewer than ten distinctions between statutory law and adjudicative law.Footnote ¹³³ And for Dworkin the distinction was basic and crucial.Footnote ¹³⁴ These theorists appreciated the profound differences in the judicial and legislative roles.Footnote ¹³⁵ They did not conflate the law-making of judges with that of legislators.

There are three foundational distinctions. First, as discussed in section 3 above, judicial law-makers are subject to constraints that that legislative law-makers are not. They are constrained by the context of adjudication: They do not directly control what cases are brought to them and their decision-making scope is narrowed by the justiciable issues and the weight of precedent and institutional role. From the judicial bench “[t]here are no pure law-creating cases.”Footnote ¹³⁶ Moreover, judicial precedent remains subject to interpretation and vulnerable to revision in subsequent cases. In this sense “judge-made law is less ‘binding’ than enacted law.”Footnote ¹³⁷ Legislators, by contrast, have wide scope for enacting new laws. They choose the timing and content of their law-making: They may introduce grand reforms and new codes, and their enactments bind and trump judge-made law.Footnote ¹³⁸

Secondly, as section 4 explored, judicial law-making occurs in the context of legal instability. It arises from disputes over the proper understanding of the law. The constraints of adjudication mean that courts will rarely depart from laws that are clear, certain, and widely accepted. They either will not have an opportunity to do so or will not have a reason to do so. Legislatures, by contrast, have freedom to reform the law to reflect their preferred policies. They routinely enact statutes to replace previously stable law with new stable law; they are not constrained to legislate only when existing law is unstable. Moreover, whereas the process of litigation and adjudication itself can tend to expose the instability of a precedent, the legislative process does not have this tendency. The process of proposing and debating legislative reform does not in itself make current law more unstable, since it is clear that a bill is not law.

Thirdly, the temporal reach of judicial and of legislative law-making are incomparable. That was the lesson of section 5. Judicial law-making is inherently retrospective and justifiably so. It operates to resolve a past dispute. Legislative law-making is presumptively prospective-only because that is what the rule of law demands. Legislators enact general laws by which we must organise our conduct, and this can only be achieved if enactment precedes the conduct to be governed. Legislation is not by its nature retrospective, whereas adjudication innately is.

These distinctions are perhaps trite, but they are significant because they help to expose a possible fallacy in complaints that the “retroactive effect of judicial legislation” is invariably “harsh and unfair” and so must be avoided.Footnote ¹³⁹ The idea that judicial law-making is legislative provokes conflict with the notion that it is retrospective. It leads to the criticisms referenced in section 5 that retrospective judicial law-making violates the rule of law. Such criticisms compel jurists to search for workarounds to realise judicial legislation without retrospective effect. The most dramatic such workaround at common law is the juridical technique of prospective overruling.Footnote ¹⁴⁰ This is not the place to deconstruct that “constitutionally controversial topic.”Footnote ¹⁴¹ The point, rather, is to intimate why such workarounds may be solutions in search of a problem.Footnote ¹⁴² If judicial law-making is not legislative, then it is a mistake to equate judicial retrospectivity with retrospective legislation. Instead, we must consider whether retrospective judicial law-making is justifiable in its own right. Our reading of Hart, Lewellyn, Fuller, and Dworkin reveals reasons, grounded in the context and constraints of adjudication, for viewing retrospectivity as justifiable and not contrary to the rule of law. This understanding then allows us to move beyond a conceptional binary of judges as either discoverers of the law or makers of it, and to appreciate these functions as intertwined in the judicial role in a manner that is intrinsically not legislative.

7. Judicial law-making entails recognising law

Common law courts should not be considered legislatures because the way judges make law is simply incomparable to legislating. Judges make law through adjudication. Adjudication, as described by Finnis, is “the effort to identify the rights of the contending parties now by identifying what were, in law, the rights and wrongs, or validity or invalidity, of their actions and transactions when entered upon and done.”Footnote ¹⁴³ Judges look back to what came before when adjudicating disputes, and they use hindsight to determine what the law is that governs the dispute before them. In this light, to debate whether judges either find law or make law in hard cases misses the point. As our preeminent theorists reveal, the concepts are inextricably intertwined.

Fuller observed that the nature of adjudication elides the distinction between whether a court’s judgment “creates new law or simply derives its standards of decision from implications already contained in existing law.”Footnote ¹⁴⁴ Dworkin “reject[ed] as unhelpful the ancient question whether judges find or invent law” because he saw in legal reasoning a “sense in which they do both and neither.”Footnote ¹⁴⁵ He considered that “when judges construct rules of liability not recognized before,” they “make their common-law decisions on grounds of principle” that subsisted at the time the parties interacted, not on the basis of the policies of today.Footnote ¹⁴⁶ On this, Hart did not disagree. For Hart, in hard cases “the judge both makes new law and applies the established law which both confers and constrains his law-making powers.”Footnote ¹⁴⁷ Llewellyn, more bluntly, considered debates over whether judges find the law or make it to be “meaningless: because the judges in fact do both at once.”Footnote ¹⁴⁸ Judges make new law in accordance with preexisting principles, given materials, and institutional and social constraints.Footnote ¹⁴⁹ They advance stability in the law through resolving disputes over what the law, properly understood, required of parties. It is in this light that even landmark cases, such as Donoghue v Stevenson—perceived as the House of Lords settling genuinely contested legal understandings of the scope of manufacturers’ liability to consumers—have been said to involve “no radical departure from what came before.”Footnote ¹⁵⁰

We can thus see jurisprudential overlap in conceptions of the judicial method that move beyond the dichotomy of the judge as law-finder or law-maker. There is even a convergence in phrasing. Fuller described the judicial method not “as enacting, but rather as recognizing, law.”Footnote ¹⁵¹ Dworkin sought to understand which principle “is decisive over what a judge recognizes as law.”Footnote ¹⁵² Hart considered that judge-made law, “though it is new law,” is grounded in “principles or underpinning reasons recognized” in the existing law.Footnote ¹⁵³ Llewellyn described judicial decisions as being “quite literally found and recognized, as well as made.”Footnote ¹⁵⁴

The notion that judicial decision-making entails judicially recognising the law ultimately reflects the core insight of the declaratory theory.Footnote ¹⁵⁵ It is that the law-making function of judges is inextricable from the adjudicative function. It is not that a judge in deciding a case discerns the law for and from “eternity,” as Austin portrayed the Blackstonian position. On the contrary, the judge’s task is to determine the law for a particular time period. What the declaratory theory insists is that the critical time period is not the date when the prevailing position in law is declared—i.e., not the date of judgment. Rather, the enquiry is into the law that properly governed at the time of the parties’ dispute. The law as recognised today is tethered to the past by the context of the parties’ dispute. The judge’s role is to discern what the law fairly was when the dispute arose by determining how it ought to have been understood by the parties, all things considered. The judge then declares the law with the benefit and the constraints of hindsight. The room for juridical novelty is counterbalanced by the need to do justice in the case at hand.

In this sense, if it be found that a former precedent conflicts with a more fundamental rule or principle or understanding of the relevant law, to overrule that precedent is to declare that it was not law at the time of the parties’ past dispute. What the judge does not presume to do, under a declaratory conception of judicial law-making, is to legislate over what the law was with what the judge thinks the law should have been or should be from now on. That would be “to pronounce a new law,” rather than “to maintain and expound” the law prevailing for the parties who have come before the court.Footnote ¹⁵⁶

This reconciliation of supposedly competing conceptions finds support in Lord Goff’s noted endeavour to rehabilitate the declaratory theory for a modern era. Lord Goff—described by one commentator as a judge “firmly on the side of the positivist camp”Footnote ¹⁵⁷ —certainly did not think judges only find the common law in the Blackstonian sense. Both as a judge and as a jurist, Lord Goff rejected the notion of the common law as immutable.Footnote ¹⁵⁸ He perceived the law “not so much as Maitland’s seamless web, but as a mosaic, and a mosaic that is kaleidoscopic” in nature due to the constant influence of judges forming and reforming the “tiny pieces” of the law.Footnote ¹⁵⁹ In an Oxford lecture delivered in the wake of his appointment to the House of Lords, Lord Goff implored his audience “not [to] be surprised that judges have the power to create law.”Footnote ¹⁶⁰ In the face of so many examples in the law reports, he did “not see how it can possibly be said … that judges do not have the power to change the law.”Footnote ¹⁶¹ Echoing the Llewellian notion of law existing in a state of flux, he went so far as to suggest that “[t]he only truly constant feature of the law is that it is in a constant state of change.”Footnote ¹⁶² Yet, Lord Goff also embraced the declaratory theory and rejected the suggestion that judicial law-makers act as legislators.Footnote ¹⁶³ He reconciled these facially paradoxical positions by emphasising how the context of adjudication constrains judicial power. In every case, judicial law-making is constrained by the facts presented by the parties; by the necessity to resolve a past dispute; by the doctrine of precedent; and by the vulnerability of judicial reasons to subsequent reinterpretation—features Lord Goff saw as grounding a judicial commitment to “gradualism.”Footnote ¹⁶⁴

In Kleinwort Benson Ltd v Lincoln City Council, Lord Goff described the common law as “a living system of law” that is developed by judges.Footnote ¹⁶⁵ It is developed not by judges articulating rules distilled from a fictional “ideal system of the common law” that is supposed always to have existed, but rather by each judge determining the law as they understand it to be, informed by statute, precedent, scholarship, comparative sources, and legal argumentation.Footnote ¹⁶⁶ It is in the course of stating how the law applies to past events presented in new cases that the common law develops. In some cases that development will appear to be more radical than modest, but it can nevertheless be considered development of the law when it reflects the court’s considered understanding of what the law required in the case at hand. In each case the judge is ascertaining “the law applicable not only at the date of the decision but at the date of the events which are the subject of the case before him, and of the events of other cases in pari materia which may thereafter come before the courts.”Footnote ¹⁶⁷ The judges’ decision thus inevitably operates retrospectively (to resolve the dispute) as well as prospectively (as precedent). This is how the declaratory theory should properly be understood: “not … as an aberration of the common law,” but as “an inevitable attribute of judicial decision-making.”Footnote ¹⁶⁸

Lord Goff’s conception of the declaratory theory has spurred both scholarly critiques and refinements.Footnote ¹⁶⁹ Notably for this article’s thesis, the critiques have not suggested that Lord Goff’s reasoning was incompatible with a positivist conception of the judge as law-maker. What has been attacked is his claim that the declaratory theory is an inevitable feature of the judicial role. Richard Tur, long a trenchant critic of the declaratory theory,Footnote ¹⁷⁰ rejected the premise adopted by Lord Goff that judicial law-making is inextricable from the adjudicative context. For Tur, it was obvious that when a judicial decision overrules a longstanding legal doctrine, it reveals two rules: an “old” (previously-but-no-longer-valid) law that is being replaced by a “new” (now-valid) law.Footnote ¹⁷¹ The decision to afford that “new” law retrospective effect by applying it to the facts of the case at hand is just that: a choice. It is not inevitable. Tur argued that where justice and fairness demand it, “judges may and sometimes should change the law for the future only.”Footnote ¹⁷² They can choose to apply the “old” law to accrued and pending cases, while holding that the “new” law applies going forward. In this way the adjudicative and law-making functions of judges can be disentangled, and the declaratory theory can be discarded.

Tur’s critique expressly proceeds from an Austinian perspective of “the judicial role as creative and strongly legislative.”Footnote ¹⁷³ Indeed, he favourably quoted Austin’s lament that judges on the whole have tended to be too “timid, narrow, and piecemeal [in the] manner in which they have legislated.”Footnote ¹⁷⁴ Tur’s favoured solution—the juridical technique of prospective overruling—purports to free judges to make law for the future unshackled from concerns of disturbing pre-existing rights and interests. The doctrine is controversial because, as Lord Devlin remarked, “[i]t crosses the Rubicon that divides the judicial and the legislative powers.”Footnote ¹⁷⁵ For Tur, that was no real objection, because he considered that “any clear blue water between legislation and adjudication seems itself to be a theoretical artefact and not a neutral pre-theoretical datum by which theories may be tested for descriptive fit.”Footnote ¹⁷⁶ My (admittedly partial) response to Tur is that, to the extent the distinction is normative, it is a premise that influential jurists have taken seriously. The idea that judicial changes in the law are inextricable from the adjudication of past disputes cannot be readily dismissed.

Observing the development of the law over time from the perspective of a legal scholar, it is natural to identify landmark judgments as the points of change. But this is not the perspective of the judge delivering the judgment that is later taken to be a landmark. The judge is constrained to adjudicate a dispute, and the task is to determine the law governing the parties’ past situation that is now being litigated. This task necessarily employs hindsight. It may entail having to make sense of a longstanding but unstable doctrine: a doctrine over which the parties have advanced competing, compelling interpretations. Tur argued that departure from a longstanding doctrine may be necessary to “do justice.”Footnote ¹⁷⁷ But for whom is justice to be done? The declaratory theory maintains that it is for the parties whose dispute is being adjudicated.

When judges interpret the law governing a dispute with the benefit and the constraints of hindsight, they do not (necessarily) make a claim about the correctness of their interpretation for time immemorial (as Austin, perhaps unfairly, characterised the Blackstonian view). The judge need not claim that the judgment reflects the law as it always was. The claim is narrower than that. The judgment is simply the best understanding we have today of the law that governed the dispute at the time it arose. It is the considered understanding that, on (judicial) reflection, can be fairly recognised to govern the parties in the circumstances. From this internal vantage-point, the perception of “change” in law precedes the (novel) judgment in which it is articulated.Footnote ¹⁷⁸ That is what makes judicial retrospectivity inevitable and not unfair from the parties’ perspective. As Finnis puts it:

In this sense, the declaratory theory is “a statement of the judge’s vocation and responsibility.”Footnote ¹⁸⁰ It encapsulates how the common law changes. It is reconcilable with the notion of judicial law-making, as I have endeavoured to show from this traversal of four prominent and otherwise rival accounts of adjudication.