I. Introduction
This article examines the principle of legality (PoL). The PoL is a principle of statutory interpretation which is typically taken to mean that if Parliament wishes to infringe basic common-law norms it must do so through express language or by necessary implication.
This principle is of increasing prominence in the public law jurisprudence of the UK Supreme Court. It has been applied to determine recent significant cases such as Unison,Footnote 1 and its intellectual influence is evident more broadly in emergent methodologies adopted by the Supreme Court to scrutinise governmental power, as in the landmark prorogation case of Cherry.Footnote 2 Yet the principle has not featured prominently in public law scholarship. Indeed, what scholarship exists is dwarfed by that on other prominent topics in public law, such as substantive review. This is despite the real possibility that the PoL may come to eclipse substantive review as the Supreme Court's principal common-law tool for scrutinising the substance of executive action.
The aim of this article is to look behind this well-known and apparently straightforward principle of interpretation, in order to unpack its key elements, including many which are not apparent at the surface level of doctrine, and in doing so to tease out key controversies. The concern is thus to understand the PoL and its core features, and canvas central tensions in the case law which have not so far been identified or examined.
The article addresses four key issues. First, it considers the “triggers” which enliven the PoL and delineate its scope of application. Second, the article argues that, while the PoL has often been treated as a unitary principle by both courts and commentators, deeper doctrinal analysis reveals that several different variants exist. Third, having identified different variants of the PoL, the article examines whether the judicially-stated rationale for the PoL can sustain each variant. Fourth, having identified the emergence of a significant proportionality dimension within the legality case law, the article considers the reasons for emergence of this proportionality dimension and the interrelationship between the legality principle and substantive review.
One preliminary point requires mention. The PoL is a general principle of the legal order. It can arise in any legal context where statutory provisions prima facie interfere with important common-law rights or values, including criminal law, private law and administrative law. However, this article focuses on administrative law cases, specifically cases where statute (or the common law) purports to bestow a power on a decision-maker which could, prima facie, be exercised to infringe basic common-law norms. Administrative law is the principal context in which the case law on the PoL has evolved over the last 40 years. To have a full understanding of the case law, one must understand the broader doctrinal context in which it has evolved, and which has invariably shaped it. Focusing on a specific context also serves to demonstrate, and allows one to interrogate, how the PoL may give rise to specific issues within given contexts, which are not replicated in others, such as, in the administrative law context, the intersection of the PoL with substantive review.
II. When Does the Principle of Legality Apply?
Before examining the nature of the PoL, a logically prior question arises: what are the “triggers” for the principle? The PoL is only engaged where statute touches a common-law norm recognised as a trigger for the legality principle.Footnote 3 Over time the judiciary has expanded the range and type of triggers. This expansion gives rise to several emerging issues.
A. Triggers
Traditionally, the idea that only clear statutory words could oust or limit common-law norms was applied to protect vested rights, specifically private law rights in land, liberty and physical integrity.Footnote 4
That the PoL was generally limited to private law rights, clearly established in positive law, provided certainty in its application. Further, the principle's application to protect such rights was uncontroversial. Their status within the normative order of the common law is well-established: fundamental rights in person and property are a signal feature of the English constitutional tradition. These rights have long been afforded strong protection at common law through dedicated actions and powerful remedies. As such there could be no plausible argument that Parliament was ignorant of these rights when passing legislation. Reinforcing this point, their content is well-established and straightforward: do not enter another's land, touch another or confine another.
The interpretative requirement that clear words are required to authorise rights-infringements has synergies with the analytical structure of private law actions; so the PoL had natural application. If an official enters one's land, the official must demonstrate lawful authority. Because lawful justification is a defence it is construed narrowly, requiring clear authority to invade rights.
Thus, vested rights provided a relatively certain, stable and uncontroversial set of triggers. Yet the set is narrow. Torts protect a slim set of generally negative rights in property and person. It was inevitable that pressure would mount to expand the set of interests protected against inadvertent statutory infringement. But in England, through the 1980s and 1990s, there emerged a direct source of pressure: the European Convention on Human Rights (ECHR or Convention). Specifically in the prisons context, and absent a domestic rights-charter, the PoL emerged as a principal means by which courts could protect prisoner rights, against the background of multiple European Court of Human Rights (ECtHR) findings against the UK.Footnote 5 The PoL was a natural corollary of the common-law presumption that prisoners retain their civil rights, which emerged in the 1970s.Footnote 6 Bringing the two ideas together, the courts held that, if the presumption is to be defeated, clear statutory words are required.Footnote 7 But, significantly, the PoL could only mitigate the risk of non-compliance with the ECHR if the range of triggers was expanded, given many violations found by the ECtHR related to Article 6,Footnote 8 which protects access to court – an interest not protected traditionally by private rights in English law.
The courts could not directly apply Convention rights, given a commitment to dualism.Footnote 9 Instead courts expanded the range of triggers by recourse to the home-grown concept of common-law constitutional or fundamental rights. Over time various interests protected by the Convention were recognised as common-law constitutional rights and triggers for the PoL,Footnote 10 including, most prominently, free expressionFootnote 11 and access to court.Footnote 12 Importantly, while courts have leaned heavily on the language of “rights”, access to court and freedom of expression are not “rights” as that concept is traditionally understood in English law. There is, for example, no individual legal entitlement, enforceable against another, to access court or express oneself. Rather, within the common-law tradition these are civil liberties: one has freedom to go to court or express oneself to the extent permitted by law.Footnote 13 Importantly, that free expression and access to court have been refashioned as “constitutional rights” for the purposes of the PoL does not appear to have transformed them from liberties to rights-proper. Free expression and access to court do not cast correlative duties, and nor are they the subject of independent legal protection via dedicated legal actions (at common law), in the way the right to liberty is protected directly by false imprisonment. As Laws J. said in Witham, “the executive cannot in law abrogate the right of access to justice, unless it is specifically so permitted by Parliament; and this is the meaning of the constitutional right”.Footnote 14 When in Watkins the Law Lords squarely addressed whether to recognise these constitutional rights as independently actionable, they declined to so develop the law, preferring to protect these interests through interpretative methods.Footnote 15 Thus, whereas the PoL is “parasitic” on the existence of free-standing private rights, in contrast common-law constitutional rights do not exist independently of their role as trigger norms for the PoL.
Lastly, more recently triggers have been extended beyond “rights”. Specifically, constitutional “values” or “principles” have been recognised as triggers, such as the rule-of-law principle of non-aggravation of penalties,Footnote 16 bindingness of court decisions,Footnote 17 and preservation of the High Court's supervisory jurisdictionFootnote 18 (albeit this last principle has a longer history in ouster clause cases). Traditionally, such values may have been appealed to in judicial reasoning, provided justification for legal norms, or been immanent in judicial practice. Thus, in Cherry the Supreme Court explained that the constitutional principle recognised therein – executive accountability to Parliament – had traditionally been an “explanation” or “justification” for certain doctrines.Footnote 19 The shift involved in recognising values as trigger norms is that such values are elevated from the substrata that underpins legal norms to the surface level of the law, themselves now having the status of legal norms and, where engaged, having direct legal consequences.
B. Three Emerging Issues
The significant expansion of triggers beyond long-standing vested rights, and increasing rate of expansion, in turn give rise to several issues.
1. Selecting triggers
There are issues relating to selection of triggers.
It has often been unclear why certain norms have been selected for protection via the PoL but not others. Consider privacy. While certain “rights” such as free expression are recognised triggers, when cases concerning privacy have reached the Supreme Court the Court has routinely relied on the Human Rights Act 1998 (HRA).Footnote 20 Previously, in Gillan the Law Lords thought the liberty to go about one's life free from intrusion came close to being a constitutional principle, but did not quite warrant that status.Footnote 21 However, it was not explained why.Footnote 22 This begs the question why privacy is any less important than free expression. Under the Convention the interests are presumptively equal. This differential treatment is brought squarely into focus by the Supreme Court's repeated statements that the “ordinary approach” to human rights adjudication is to go to the common law first, before considering HRA/Convention protections,Footnote 23 and the Law Lords’ repeated claims of confluence between the protection afforded to basic rights by the common law and HRA.Footnote 24 Further, there is a privacy right at common law: the right against misuse of private information.Footnote 25 But perhaps invoking this right could bring into doubt repeated judicial statements that the common law provides protection equivalent to that of the Convention, given courts have not developed the common-law right to protect the full gamut of interests protected by Article 8.
Privacy is one example. Many others could be given: why have courts not recognised religious freedom or freedom from discriminationFootnote 26 as triggers, or socio-economic rights, for example?
The courts have articulated more detailed explanations for why those rights that are protected, such as access to courtFootnote 27 or free expression,Footnote 28 are important. But many interests are important; the pertinent issue requiring explanation is why certain phenomena warrant the classification “constitutional”, and protection specifically via the PoL, but not others. What is missing is a transparent analytical framework to govern on a consistent basis selection of trigger norms.Footnote 29
The recent Elgizouli case provides a rare example where the Supreme Court directly addressed why norms should not be recognised as constitutional.Footnote 30 The court had to confront the matter as the pleaded norms were not established in Convention jurisprudence; so there was no option of relying on the HRA. But unfortunately, the reasons given reinforce concerns over lack of a consistently applied framework. The majority rejected a common-law norm that the Government shall not facilitate prosecutions abroad that could result in the death penalty, said to derive from a common-law constitutional right to life. The reasoning is difficult to reconcile with prior cases.
The majority placed heavy emphasis on the specific norm pleaded not being recognised in international instruments.Footnote 31 But this has not been a touchstone in prior cases. In Unison for example (see below) the court expanded the right to access court beyond the content of Article 6, never mentioning Article 6.Footnote 32 The whole point of the “ordinary approach” is that the common law is an autonomous source of norms.Footnote 33 In any case the majority doubted whether even canonical ECHR principles such as that in Chahal would find recognition at common law,Footnote 34 while the pleaded norms hardly cut across international standards – which are uniformly against the death penalty.Footnote 35
Emphasis was placed on Parliament having addressed the subject matter of the claimed norms in various statutes.Footnote 36 But access to court and free expression are both addressed in the HRA, yet this has not stopped parallel common-law developments.Footnote 37 Further, the pleaded norms are consonant with legislative policies, which plainly favour life and are against the death penalty.
The majority held there was insufficient precedent to recognise a constitutional right to life.Footnote 38 Yet it was explicitly accepted that the right to life is a recognised common-law “value”, immanent in tort, contract and public law.Footnote 39 Values immanent in the law have been accepted as constitutional norms, such as the value of executive accountability to Parliament recognised in Cherry (see below). Similarly, free expression was immanent in various doctrines, such as defences to breach of confidence,Footnote 40 prior to its recognition as a constitutional right.
It was observed that the claimed norms were not established private law rights.Footnote 41 But neither are access to court or free expression.Footnote 42 Moreover, it seems difficult to reconcile access to court and free expression being recognised as constitutional rights, but not the right to life, judicially recognised as the most fundamental of all human rights.Footnote 43
My concern is not to argue the pleaded norms in Elgizouli should have been recognised as constitutional rights (I express no view). It is to highlight how issues of inconsistency and incoherence, or the perception of such, can arise absent a worked-out analytical framework. In turn, this can raise other issues, including a perception that unstated normative concerns lie behind apparently incongruent lines of reasoning.
Turning from cases of “rights” to those of constitutional “values” or “principles” it is striking that recognised triggers have generally been limited to those relating to the courts’ institutional position.Footnote 44 This suggests a narrow view of the constitution. However, in Cherry the Supreme Court signalled a new willingness to recognise constitutional principles unrelated to the judicial branch, such as those regulating the relationship between Parliament and government.Footnote 45 However, this expansionary move, while addressing the criticism that courts have adopted an overly narrow view of constitutional values, precipitates further issues. Should every constitutional value be recognised by the common law, and rendered a trigger for the PoL or only some, and if some, how are the courts to decide? Consider devolution. Despite now being a fundamental feature of the UK constitution, devolution finds no recognition within the common-law corpus of constitutional principles. Value for money is a leitmotif of modern public life, and arguably a public expectation as to management of public resources: should Parliament have to speak clearly to authorise inefficient executive action?
Within the UK constitution, many constitutional values or principles have existed as principles of the political system. Courts may have taken cognisance of and reasoned by reference to these values, but the values were not legal norms as such. However, Cherry may mark a turning point. The principle that the executive is accountable to Parliament, traditionally a norm of the political order, was recognised as a legal norm, capable of directly triggering normative consequences.Footnote 46 The court's willingness to recognise a principle of the political constitution as a legal norm conditioning prerogative power raises difficult questions regarding which political norms may be recognised as legal norms, and on what basis. For example, the Supreme Court's readiness in Cherry to transmute a political norm into a legal one seems at odds with the generally cautious judicial approach to conventions, illustrated by the Supreme Court's cursory treatment of arguments based on the Sewel Convention in Miller, despite that case having high stakes for devolved nations.Footnote 47 It is oft-said in legality cases that the legislature does not legislate in a vacuum but against the backdrop of pre-existing fundamental norms and traditions, which are presumed to continue to apply unless there are express terms suggesting otherwise.Footnote 48 Why might conventions, or the values underpinning them, not be considered background fundamental norms, especially as the Supreme Court recognises conventions can play “a fundamental role in the operation of our constitution”?Footnote 49
These “why not?” questions are easy to pose. But they are hard to answer because courts have not articulated a framework to govern clearly and consistently the selection of trigger norms.
2. New and expanding triggers
Courts have recognised new triggers and expanded existing triggers. For example in Evans the bindingness of court judgments was recognised as a constitutional principle, apparently for the first time,Footnote 50 and in Pierson two Law Lords recognised non-aggravation of criminal penalties as a general principle, apparently for the first time.Footnote 51 In Unison the Supreme Court endorsed three major extensions of the right to access court. First, it was authoritatively expanded to include tribunals.Footnote 52 Second, the court authoritatively affirmed that the right protects not only against direct interferences with one's ability to access court, such as confiscation of a prisoner's correspondence with a court, but also against measures which create disincentives to accessing court, specifically economic disincentives.Footnote 53 Third, in addition to protecting identified individuals who are actually prevented from accessing court, the right protects against administrative measures which create a real risk that persons at large may be prevented from accessing court.Footnote 54
When courts have so developed the law, how would it have been possible for Parliament to have squarely confronted infringement of the trigger norm, given it was not previously recognised as a trigger, or the scope of the norm was previously narrower? One might argue that Parliament can respond after the event, but the judiciary's stated rationale for the PoL is ex ante: Parliament should, in passing this legislation, have squarely confronted interference with constitutional norms. This issue does not arise to the same extent where the trigger has a legal existence distinct from the PoL. For example, legal development of private law rights occurs outside legality cases whereas, for common-law constitutional rights, the only opportunity for legal development is in the very cases where the rights are being applied as triggers for the PoL.
Given the premise underlying the PoL is that Parliament should have known to speak clearly to abrogate established norms, legitimacy concerns arise where a court reads a statutory provision as subject to a norm that was not previously clearly established. Consider Pierson. Those Law Lords that held the principle of non-aggravation of penalties did not trigger the PoL based that view on there being a lack of authority for such a principle; it was not clearly established by precedent.Footnote 55 Without an anchor in precedent, there was a concern that to cut down a statutory power based on appeal to this putative principle would equate to reading into the statutory provision a limit based on the judges’ own conceptions of fairness.Footnote 56 One might view this emphasis on prior authority as a response to Lord Steyn's contrasting approach in the same case, which involved distilling the non-aggravation principle from a philosophical inquiry into the rule of law, and applying it to cut down the statutory power.Footnote 57
3. Institutional and constitutional concerns
It is seriously open to question whether it is for courts to recognise a host of new norms, said to be constitutive of the polity, especially as Parliament has prescribed which rights it considers fundamental in the HRA and other legislation, such as anti-discrimination statutes, providing dedicated models of protection, including interpretative principles such as HRA, s. 3. Questions over which norms are (or are not) basic to a society do not permit of easy answers, and their resolution implicates bare issues of politics and/or morality, in relation to which citizens will reasonably disagree and legitimately expect a voice. For these reasons such matters might be thought to be properly for the people or their representatives.
Even assuming consensus as to which norms are “constitutional”, there are myriad further issues which implicate contentious value judgements, in respect of which citizens may legitimately disagree.
We might agree norms are important, but disagree over their relative importance and how to protect them. As discussed below, courts view some triggers as more important than others, and thus apply more aggressive variants of the PoL. Some freedoms, such as the freedom to live one's life as one wishes, are protected by anxious scrutiny Wednesbury review,Footnote 58 but not recognised as triggers for the PoL.Footnote 59 Similarly, the right to life is an important “value” protected by anxious scrutiny,Footnote 60 but not considered a constitutional right. More broadly, there will be disagreement over legal versus political modes of protection.
We might disagree as to the formulation of norms. In Privacy International the majority recognised a constitutional principle of maintenance of the High Court's supervisory jurisdiction, whereas minority Justices would have recognised a constitutional principle of access to a judicial body with capacity to quash for legal error, which need not be the High Court.Footnote 61
We might disagree whether any “hindrance or interference”Footnote 62 with given norms is sufficient to trigger legal protection, or whether “significant”Footnote 63 interference is required. For example, in Cherry the court seemingly adopted a de minimus principle: a prorogation would only infringe executive accountability to Parliament if it reached a minimum duration.Footnote 64
We might disagree whether norms should be absolute or qualified. In Witham Laws J. asserted access to court was more important than free expression and is thus an absolute right, whereas free expression is qualified.Footnote 65 In Cherry one of the constitutional norms applied was parliamentary supremacy, traditionally an absolute rule. But the court, having given the norm an extended meaning and describing it now as a “principle”, appears to have treated it as qualified: an exercise of the prerogative that undermined parliamentary supremacy might be capable of objective justification.Footnote 66 In Privacy International the majority considered parliamentary supremacy absolute, whereas the minority considered it qualified by the rule of law.Footnote 67
Courts have invoked different conceptions of fundamental principles. Lord Sumption in Privacy International considered the rule of law encompassed parliamentary supremacy, whereas Lord Carnwath considered parliamentary supremacy and the rule of law to be, in an important way, at odds, and thus discrete concepts.Footnote 68 In Anufrijeva Lord Bingham considered the rule of law required giving effect to clear statutory words, but did not conceptualise the requirement to give notice of an administrative decision as a rule-of-law principle.Footnote 69 In contrast Lord Steyn framed the notice-giving requirement as a manifestation of the rule of law, and did not connect to the rule of law the proposition that clear statutory words should be abided by.Footnote 70
If judges disagree about these basic questions, which go to formulation of, and the basic conception of relevant norms – before one even gets to application – then ordinary citizens will disagree too. As such one might expect the fairest and most transparent way to settle which norms are constitutive of society is for society to decide through an open process.
Moreover, judges have not addressed, on the one hand, many of the foregoing questions in a direct way, within a coherent framework. But, on the other hand, developing a more transparent and systematic approach will bring to the surface the contentious, political and/or moral nature of many of these questions, which may in turn undermine judicial legitimacy. Where judges have provided greater reasoning, they have walked a fine line. Some judges have expressed their personal moral beliefs.Footnote 71 In Unison Lord Reed, justifying the importance of access to court, explicitly criticised market-based models of justice.Footnote 72 Deciding amongst models for organising and funding a justice system implicates contentious political and economic value judgements. More generally, creating new legal norms involves bare distributive questions over allocation of entitlements. Within private law this is a key reason why courts increasingly consider that the creation of new rights is principally for Parliament.
Interestingly, in the recent case of Elgizouli the Supreme Court, for the first time in its new constitutional canon, seriously acknowledged legitimacy concerns associated with creation of new legal norms. In refusing to recognise the pleaded norms the majority stressed the common law's traditional incrementalism and the legislature's primacy in effecting law reform.Footnote 73 However, this new-found restraint is difficult to reconcile with the bold decisions in Cherry – where a wholly novel substantive limit on the prorogation power was established – or Unison, where the Supreme Court endorsed significant expansion of the right to access court. Time will tell whether Elgizouli marks a new phase of restraint, following-on from a period of expansive legal development.Footnote 74
III. One Principle of Legality or Many?
Let us now consider the PoL itself. The classical formulation of the principle holds: if Parliament wishes to infringe basic common-law norms it must do so by express statutory words or necessary implication. However, doctrinal exegesis reveals the existence of meaningfully different versions of the principle. In turn, use of the umbrella term “principle of legality”, coupled with its association with the classic formulation, obscures these nuances.Footnote 75
It is important to unpack these variants. If the law is to be rationally ordered and consistently applied it is fundamental to recognise that different principles are being applied in different cases. Further, once brought to light, different variants of the PoL may not mesh with the principle's stated rationale, and/or may raise greater legitimacy concerns than the classical version.
At least three variants can be identified. First, the classic PoL that requires express words to sanction any interference with basic norms. Second, the augmented PoL which holds that express words authorising an interference with basic norms are insufficient to authorise disproportionate interferences; such interferences can only be sanctioned by statutory words specifically authorising disproportionate interferences. Third, the proactive PoL which requires courts proactively to construe statutory language to minimise any interference with basic norms as far as possible.
A. Classic Principle
The classic PoL holds that clear words are required to sanction interferences with basic norms. An authoritative formulation of the classic PoL, in the administrative law context, comes from Lord Browne-Wilkinson in Pierson:
A power conferred by Parliament in general terms is not to be taken to authorise the doing of acts by the donee of the power which adversely affect the legal rights of the citizen or the basic principles on which the law of the United Kingdom is based unless the statute conferring the power makes it clear that such was the intention of Parliament.Footnote 76
Raymond provides an illustration of this classic principle in operation.Footnote 77 The Minister had a broadly framed power to make “rules for the regulation and management of prisons” and “classification, treatment, employment, discipline and control of” prisoners.Footnote 78 The Minister promulgated a rule that purported to prohibit prisoners from communicating with any person in relation to legal or other business. This rule would, on its face, prevent a prisoner from communicating with a court concerning legal proceedings, which would constitute interference with access to court. It was held that a general grant of power to promulgate rules for prison administration was “manifestly insufficient” to authorise interference with such a basic right.Footnote 79
In such cases, where there is a broad power to make miscellaneous rules, the classical principle operates in a fairly uncontroversial and straightforward way. There is no specific textual indicator that can be cited as evidence of a clear parliamentary intention to enable interference with given basic norms. Put another way, it is hard to maintain convincingly, in the context of a broad managerial power, that what Parliament specifically had in mind was authorising officials to prevent prisoners from accessing court.
But on closer inspection even the apparently straightforward classical principle has layers of nuance. It would appear there is a hierarchy of rights. The more important the right, the more resilient to statutory displacement. Whereas basic norms can generally be displaced by express words or necessary implication, in certain contexts the “necessary implication” limb is inapplicable given the importance of what is at stake. Thus, it has been said that access to court, given its fundamental nature, can only be precluded by express words; necessary implication is insufficient.Footnote 80 Similarly, given the demands of the rule of law, judicial review could only be precluded “by the most clear and explicit language and not by implication”.Footnote 81
The criteria of “express” or “clear” words are themselves highly malleable, and judges may, consciously or not, vary their demandingness depending on the importance of the trigger. Consider Anufrijeva. Regulations provided that an asylum seeker's access to welfare benefits ends when he/she “ceases to be an asylum seeker”, which is when his/her claim to asylum “is recorded by the Secretary of State as having been determined … on the date on which it is so recorded”.Footnote 82 The question was whether the words were clear enough to negative the ordinary procedural fairness requirement that a determination, to have legal effect, must be communicated to the subject. Lord Bingham, dissenting, considered the regulation's meaning “clear and obvious”: a matter is determined once recorded by the Minister, on the date it is so recorded. As such the wording made clear that the concept of “determination” was not one dependent on notice.Footnote 83 In contrast, Lord Steyn, in the majority, considered the provision had “not in specific and unmistakable terms” displaced the ordinary notice requirement.Footnote 84
It does not seem coincidental that the strictness with which the majority and minority applied the “clarity” requirement corresponded with differing views of the normative importance of the trigger norms. For Lord Bingham the trigger was the “public law duty”Footnote 85 to notify of a decision. There was no appeal to constitutional principles – this was ordinary administrative law fairness. In contrast Lord Steyn characterised the triggers as “fundamental principles of our law”.Footnote 86 First, notice is required by the right of access to justice, a right held in the highest regard by the courts; without notice a person could not challenge the decision. The second trigger was the constitutional principle of the rule of law. The importance of these common-law principles was reinforced by appeal to European law. Lord Steyn clearly considered the triggers ranked near the apex of any hierarchy of common-law norms – and he no doubt also had in mind compliance with supranational requirements. In turn, it is difficult to see as unrelated his relatively more searching application of the “clarity” requirement, compared to Lord Bingham, who did not characterise the triggers as fundamental.
B. Augmented Principle
The classic formulation is well-known. But other variants of the PoL are identifiable, which are materially different from the classical conception – albeit these differences have not always been judicially acknowledged.
What I term the “augmented” PoL is one such variant. It originated in a pre-HRA line of prisoner cases, against the background of a series of findings against the UK by the ECtHR, which placed pressure on domestic courts to bring English law into compliance with Convention requirements, absent a rights-charter.Footnote 87 The augmented variant introduces a proportionality dimension into the classic version, with the effect that it is more difficult to demonstrate legal authorisation for rights-interferences.
Whereas early on in the development of the case law aspects of proportionality analysis, such as “necessity” or “pressing need”, were invoked in an unstructured way, over time these concepts were integrated into a more formalised and structured methodology, which now very closely resembles full-fledged proportionality of the type applied under the HRA. As these developments have “bedded in” courts have begun to explicitly acknowledge that the PoL, as applied in these cases, incorporates structured proportionality.
1. Beginnings: Leech
The origins of the augmented principle lie in the Court of Appeal decision in Leech.Footnote 88 Pursuant to section 47(1) of the Prison Act 1952 (the same provision at issue in Raymond), the Minister promulgated a rule allowing all correspondence to and from a prison to be read. Because the rule allowed screening of a prisoner's correspondence with his or her lawyer, the court held the rule engaged the common-law right to legal professional privilege, a right conceptually separate from but which “buttressed” the right to access court. However, the court accepted that the statute by necessary implication allowed some screening of legal correspondence and thus the making of rules that interfered with the right to legal professional privilege.Footnote 89
On the classic principle that would be the end of the story – the statute authorised screening of legal correspondence, and thus interference with the right to legal professional privilege. However, the court considered this insufficient to authorise the rule. The court effectively added a second stage to the legality inquiry. It asked whether there was a “pressing”, “objective” or “demonstrable need” for a rule that allowed screening of all legal correspondence, and whether the rule went beyond the “minimum extent necessary” to ensure correspondence was genuine legal correspondence.Footnote 90 Because the court held the rule was “extravagantly wide”,Footnote 91 infringing the right more than was strictly necessary or justified by legitimate objectives, more was required by way of authorisation.Footnote 92 What was required was statutory language which expressly or by necessary implication authorised an intrusion of the extent imposed by the rule – namely authorisation for a blanket rule that infringed rights more than strictly necessary. The statute contained no such language; so the rule was ultra vires.
Notably, the ground of intervention was not substantive review but vires.Footnote 93 Yet we see in the court's analysis, under the rubric of the PoL, application of concepts synonymous with limbs of structured proportionality, such as “least intrusive means”. The court explicitly drew on Canadian human rights jurisprudence for these concepts,Footnote 94 while the case was decided against the backdrop of Campbell, where the ECtHR found blanket screening of prisoner correspondence contrary to the ECHR – on proportionality grounds.Footnote 95 The court recorded the happy coincidence that the same result had been reached via domestic law,Footnote 96 while later cases explain Leech as having “followed” Campbell.Footnote 97
This augmented principle enhances the protection afforded by the PoL. Where a measure unnecessarily restricts rights, something more will be required than an explicit authorisation to interfere with rights. There must be authorisation specifically for an interference of such extent. Where courts determine a given intrusion is disproportionate it is unlikely the statute will be found to have provided such detailed authorisation. And the more intrusive the measure, the more courts will demand by way of detailed authorisation.Footnote 98
2. Coming of age: Simms and Daly
The augmented principle from Leech was subsequently endorsed and applied by the Law Lords in two important prisoner cases, decided as the HRA was entering into force.
In Simms the question was whether the Minister could adopt a rule under the Prison Act imposing a blanket ban on journalists visiting prisoners.Footnote 99 Lord Steyn, who had given the judgment in Leech, said freedom of expression was the analytical starting point, but as a qualified right may yield to other interests.Footnote 100 The importance of free expression was bolstered by appeal to access to justice: Simms wanted to engage a journalist because he believed he had suffered a miscarriage of justice.Footnote 101
Lord Steyn held that the Minister could demonstrate no “pressing need” to justify the ban.Footnote 102 The “pressing need” test is a well-established limb of structured proportionality, and echoes the approach in Leech. Lord Steyn's methodological approach had another feature in common with proportionality: it was steeped in evidence, and specifically evidence pertaining to “non-adjudicative” facts. He relied on evidence showing prisoner interviews had not adversely affected prison discipline and that without interviews it was virtually impossible for a journalist to take up a prisoner's cause, which served an important check on the justice system.Footnote 103 It followed that a blanket ban would be “exorbitant in width in so far as [it] would undermine … fundamental rights” and was “therefore ultra vires”.Footnote 104 Thus, in common with Leech the measure went further than strictly necessary, and therefore fell outwith the empowering clause.Footnote 105 Notably, the speeches, given on the eve of the HRA entering into force, are replete with claims that there is no real difference between the common law and Convention in the protection afforded to basic rights.Footnote 106
In Daly, decided soon after the HRA entered into force, Lord Bingham explicitly adopted the two-part analysis from Leech: “rights may be curtailed only by clear and express words, and then only to the extent reasonably necessary to meet the ends which justify the curtailment.”Footnote 107 The case concerned whether “the general terms of” the empowering provision in the Prison Act, “authorise … expressly or impliedly” a blanket policy of searching cells without the prisoner being present.Footnote 108 Lord Bingham's speech arguably marked the maturation of the augmented principle, as his analysis involved the most faithful application of the sequenced questions that characterise structured proportionality. His reasoning proceeded as follows:
(1) “It is necessary, first, to ask whether the policy infringes in a significant way Mr Daly's common law right that the confidentiality of privileged legal correspondence be maintained.”
(2) “The next question is whether there can be any ground for infringing in any way a prisoner's right to maintain the confidentiality of his privileged legal correspondence.”
(3) “It is then necessary to ask whether, to the extent that it infringes a prisoner's common law right to privilege, the policy can be justified as a necessary and proper response to the acknowledged need to maintain security, order and discipline in prisons and to prevent crime.”Footnote 109
His Lordship found the prisoner's right to privileged communications was engaged because there will necessarily be occasions when officers do more than merely examine a prisoner's legal documents, and apprehension that officers may read documents could inhibit prisoners’ willingness to freely communicate with their lawyer. There may be good reasons for examining privileged correspondence, for example to check it is not a hiding place for illicit materials, and reasons to keep prisoners out of cells, for example to protect staff. But the policy's blanket, indiscriminate nature, being unresponsive to variables such as category of prisoner, past disciplinary record etc., meant it went further than necessary to serve legitimate goals.
This may look like proportionality applied as a head of substantive review. But the analysis was all part of a legality inquiry. Tying the proportionality analysis back to the statute, Lord Bingham said: “Section 47(1) … does not authorise such excessive intrusion, and the Home Secretary accordingly had no power to lay down or implement the policy in its present form.”Footnote 110 He observed that while his decision was reached at common law, the same result could be achieved by reliance on Convention norms.Footnote 111 Thus, while Daly is often cited for Lord Steyn and Lord Cooke's ruminations on substantive review, in hindsight the case is arguably more important for Lord Bingham's contribution to the post-Leech line of legality cases.
3. Second coming: Unison
After Daly the augmented PoL entered a long sleep. It was not applied by the Law Lords for another 16 years,Footnote 112 until its spectacular return in Unison, the first Supreme Court decision to apply the augmented principle, and the first in the Leech line of authorities to apply it outside the prisoner context.Footnote 113 Why it returned after this hiatus is considered below.
The Tribunals, Courts and Enforcement Act 2007 provides that the Lord Chancellor may by Order prescribe fees for employment tribunals, and an elaborate system of fees was established.Footnote 114 Unison involved a challenge to this system, specifically to the level of fees and limited provision for fees remission.
The relevant trigger, broadly stated, was access to justice. Lord Reed articulated two ways in which the Order could be ultra vires, both involving the PoL:
(1) There is no power to promulgate a measure that completely bars access to justice, unless empowering legislation specifically addresses and explicitly authorises a total bar.Footnote 115
(2) Legislation may in general terms authorise measures which make it more difficult or harder for a person to access justice, but that will not – without more in the relevant provision – authorise measures which disproportionately interfere with a person's ability to access justice.Footnote 116
The best explanation for these two different paths to ultra vires is that different rights are implicated, one absolute and one qualified. On this explanation (1) concerns the right not to be completely prevented from accessing court, which is absolute; and (2) concerns the right to unimpeded access to court, which is qualified.
With (2), Lord Reed restated the two-stage analysis associated with the augmented principle. An express power to put in place impediments, which make it harder to access court, will be effective in authorising some interference.Footnote 117 But, unless statutory terms indicate otherwise, such a provision will not authorise interferences which lack objective justification.Footnote 118 As authority Lord Reed invoked cases where courts had applied forms of proportionality analysis including Leech, Simms and Daly.Footnote 119 Thus, “[e]ven where a statutory power authorises an intrusion upon the right of access to the courts, it is interpreted as authorising only such a degree of intrusion as is reasonably necessary to fulfil the objective of the provision in question”.Footnote 120 Significantly, Lord Reed expressly drew an analogy with the ECHR proportionality method,Footnote 121 while previously in Pham he had explained the interpretative principles from Leech and Daly as imposing “in substance, a requirement of proportionality”.Footnote 122
In contrast, with (1) there is no scope to objectively justify completely preventing access to justice.Footnote 123 As Laws J. said in Witham, the right against being completely barred from court is “absolute”;Footnote 124 unlike the qualified right to unimpeded access to court, there is no circumstance in which a person can justifiably be barred from accessing court. Therefore, the only way such outcome could be lawful is if it were specifically prescribed by statute. There would have to be something in the statutory words which explicitly alerts the reader to the possibility of completely barring persons from court: the statute “would provide in terms that in defined circumstances the citizen may not enter the court door”.Footnote 125
Thus (1) could be explained as involving application of the augmented PoL to an absolute right. But another possibility is that (1) involves strict application of the classic PoL. The classic principle requires express words authorising infringement of a right. The relevant right here is very specific, concerning one outcome: being excluded completely from court. It follows that a specific statutory formula is required to defeat a right of such specificity: there must be express words stating persons can be barred completely from accessing court. Thus, perhaps in this context the augmented and classic principles are indistinguishable in operation.
The Order in Unison was scrutinised on grounds (1) and (2). Applying (1), Lord Reed held that the Order created a real risk that people would be effectively prevented from accessing the tribunals, by making it impossible or futile for certain groups to bring claims, due to the level of fees charged and limited fees remission.Footnote 126 The Order was ultra vires as the empowering provision did not specifically authorise a fees regime with such effect. As noted above, the analysis is striking because, unlike the prisoner cases, the challenge was not brought by an individual alleging actual interference with their “rights”. The court had no evidence of an individual case in which fees had actually made it impossible for someone to access the tribunals.Footnote 127 Rather this was a “systemic” claim, brought on the basis that the fees system, as a system, created a “real risk” that persons unknown will be effectively prevented from accessing justice.Footnote 128
Lord Reed's conclusion on (1) was sufficient to dispose of the appeal. But His Lordship also applied (2) for completeness. The empowering provision clearly authorised some interference with the right to unimpeded access to court, by explicitly enabling imposition of an impediment: fees. According to (1), fees could not lawfully be set so high as completely to prevent access. But, according to (2), even fees that would not completely prevent access, but simply hinder access, are subject to a proportionality requirement: fees could be no higher than demanded by legitimate public goals.Footnote 129 The objectives pursued by the Order, such as rationing scarce resources and deterring vexatious claims, were legitimate.Footnote 130 But the Order failed the “least intrusive means” requirement because the Government could not prove a less intrusive fees regime would not be just as effective in achieving legitimate goals.Footnote 131 This analysis clarifies that, consonant with proportionality under the HRA, the defendant bears the burden of justifying rights-infringing measures. Ultimately, the Order was unlawful under (2) because it was disproportionate and there was no clear statutory authorisation for such unjustifiably intrusive measures.
Overall, Unison is a very significant decision. It brought back the augmented principle after a long hiatus, involves a strong endorsement of that principle in the Supreme Court era by the court's now President, and openly acknowledges that the PoL involves proportionality analysis.
4. Expanding influence: Cherry
With its high-profile endorsement in Unison the augmented principle's sphere of intellectual influence has grown.
Cherry involved a challenge to the Prime Minister's advice to HM The Queen to prorogue Parliament for several weeks in the lead up to, what was at the time, the date the UK was in law scheduled to leave the EU.Footnote 132 The case raises many issues. My concern here is the influence of the methodology associated with the augmented PoL on the court's approach to legality review of a prerogative power.
Lady Hale and Lord Reed said the PoL does not apply to prerogative powers: as a principle of statutory interpretation the PoL requires a text to be interpreted, whereas prerogative powers are sourced in the common law.Footnote 133
However, the Justices, in formulating their approach to legality review of the prorogation, explicitly drew on Unison for inspiration.Footnote 134 They observed it was of assistance to consider how courts had approached review of statutory powers where they affected constitutional principles in considering how courts should approach situations where prerogative power touched on constitutional principles.Footnote 135
The influence of the augmented PoL is reflected in obvious methodological similarities between that principle and the approach in Cherry. The court held that because the prorogation infringed constitutional principles, it required objective justification.Footnote 136 In common with the augmented PoL this analysis in terms of constitutional principles went to delineating the power's scope, and thus whether the Prime Minister had the power to give the advice that he did in the first place.Footnote 137 Questions of scope, or alternatively vires or legality, are to be contrasted with application of substantive review (which was not applied in Cherry) to scrutinise the qualities of a specific exercise of power that is otherwise within scope.Footnote 138 A further commonality with the augmented PoL is that once an interference with basic values is proven, there is at least an evidential onus on the defendant to justify the interference.Footnote 139 This contrasts with the ordinary rule on review that a claimant bears the onus of proving unlawfulness.
Thus there are strong similarities between the approach in Cherry and the augmented PoL. But there are also important differences. The constitutional norms articulated in Cherry form invariable limits on the scope of the prerogative power. Any purported exercise of power that contravenes constitutional norms without justification, will ipso facto be unlawful. In contrast, with the PoL there is always the possibility that the empowering statute could explicitly authorise interferences which lack objective justification.
Furthermore, while there are clear methodological similarities, there are differences at the level of detail. Notably, when the court in Cherry scrutinised possible justifications for the prorogation, it applied an approach more lenient and less structured than the form of proportionality analysis associated with the augmented PoL. The court called for “reasonable justification”, not explicitly invoking the language of proportionality and its associated concepts, such as necessity; the standard of justification was simply that the Prime Minister have a “good reason” for the prorogation.Footnote 140 Indeed, the assessment of justifications carried echoes of the two limbs of Wednesbury: the court would only intervene exceptionally,Footnote 141 while the mode of scrutiny was “proceduralist”, focusing on whether the Prime Minister took into account relevant considerations.Footnote 142 Thus, rather than the court itself applying the “least restrictive means” test, as it would pursuant to proportionality, it asked whether the Prime Minister had considered less drastic measures.Footnote 143 This approach was underpinned by a recognition that prorogation falls within the Prime Minister's area of responsibility and implicates an array of considerations including those requiring political judgementFootnote 144: “the Government must be accorded a great deal of latitude in making decisions of this nature.”Footnote 145
One might explain this as application of a “sliding scale” approach to proportionality, with a deferential variant applied here. One difficulty with this explanation however, as with the concept of sliding scale proportionality generally, is that at the most deferential end of the scale the approach bears no meaningful resemblance to proportionality, and is more reminiscent of Wednesbury. As such it is questionable whether proportionality as such was applied in Cherry, even if other aspects of the court's approach resemble the methodology of the augmented PoL.
Why we see a form of “deference” applied in Cherry but no appeal or even allusion to deference in the augmented legality cases is a matter to which we will return below.
Overall, the court's approach in Cherry reveals the increasing intellectual influence of the augmented PoL, but it would be inaccurate to conflate the approaches.
C. Proactive Principle
Section 3 of the HRA provides that courts should interpret legislation compatibly with enumerated rights “[s]o far as it is possible to do so”. This is a positive injunction to proactively interpret legislation compatibly with rights as far as possible.
This principle, which I term the “proactive” PoL, is conceptually distinct from the classic principle in that clear words may not be sufficient to authorise rights-infringements. Well-known HRA cases such as Ghaidan show that even where the plain meaning of words is clear courts may proactively seek to remove or minimise any rights-interference by adopting a strained meaning.Footnote 146 In light of such cases it has been acknowledged judicially that section 3 of the HRA is distinct from and goes further than the classic PoL, in that the classic PoL does not permit a court to “disregard an unambiguous expression of Parliament's intention”.Footnote 147
The proactive principle is also conceptually distinct from the augmented principle. Whereas the augmented principle renders disproportionate interferences outside power, the proactive principle involves courts seeking to minimise any interference. The goal is to maximally protect rights. In principle, this may involve ruling even proportionate interferences outside power. As such there is no balancing involved in applying the proactive principle; the controlling question is how far words can plausibly be stretched to afford basic norms maximal insulation from interference. Further, whereas under the augmented principle clear words will suffice to authorise disproportionate interferences, under the proactive principle even clear words may be “read down”.
Thus we find the proactive principle in the HRA. But significantly we also find evidence of this approach at common law. Unlike the augmented principle this variant has not been acknowledged explicitly by courts. It has typically been applied under the umbrella of the classic PoL.
The proactive principle explains the judicial approach to interpretation of ouster clauses. Such clauses are routinely “read down” to preserve the High Court's supervisory jurisdiction as far as possible, despite statutory terms plainly evincing an intention to oust that jurisdiction. In Privacy International Lord Carnwath explained the House of Lords’ famous decision in Anisminic Footnote 148 as an application of the PoL, reciting the classical formulation requiring clear words.Footnote 149 But in ouster clause cases courts do not apply the classic principle as we know it. Lord Carnwath explained that in Anisminic Lord Reid did not dispute that the plain meaning of the clause was apt to exclude judicial challenge, “but this ordinary meaning had to yield to the principle that such a clause will not protect a ‘nullity’”.Footnote 150 As such a provision's plain meaning, which manifests a clear intention to oust the High Court's jurisdiction, is substituted with another meaning which maximally preserves that jurisdiction. Thus in Anisminic a clause, providing that an administrative body's “determination” shall not be questioned in court, was read as applying only to validly made determinations, which in turn opened up the possibility of judicial scrutiny.
Of further note in Privacy International is Lord Carnwath's statement that in applying the PoL in ouster clause cases, courts are not engaged in the ordinary interpretative task of discerning Parliament's intention.Footnote 151 This statement helps us to fully understand the nature of the proactive principle: in applying the principle courts are not seeking to identify and give effect to Parliament's intent, but seeking to maximally protect basic norms – even to the extent of overriding the plain meaning of clear statutory words. As has been argued in the ongoing debatesFootnote 152 over whether legislative intent ought to remain the touchstone of interpretation, a paradigm of parliamentary intention reminds judges of the constitutional boundary between interpretation and legislation.Footnote 153 Whereas if parliamentary intention is abandoned as the object of the interpretative enterprise, “the way appears clear for the courts to import normative content of which they approve, even if it is not plausible to think that the legislating Parliament would have accepted it”.Footnote 154
In addition to the well-known cases on ouster clauses, the proactive principle is also evident in cases concerning discretionary powers. Evans is the paradigm example.
Evans concerned attempts by journalists to obtain under the Freedom of Information Act 2000 Prince Charles's correspondence with the UK Government. The Act creates a right to information subject to public interest exceptions.Footnote 155 In Evans the Government refused to release the information on public interest grounds. When the matter reached the Upper Tribunal, it ordered release. However, the Act provides a broadly framed power by which the Attorney General can veto such tribunal decisions, by producing a certificate stating that on reasonable grounds he does not consider the refusal breached the ActFootnote 156; he did so, giving detailed reasons.
Lord Neuberger, with whom Lords Kerr and Reed agreed, considered the PoL was triggered as the veto power cut across two rule-of-law principles: (1) a court's decision is binding, and (2) executive decisions are reviewable by a court.Footnote 157 The empowering provision “flouts” (1).Footnote 158 And if interpreted as allowing the Attorney General to set aside a court decision because he disagrees with it, the power would “stand” (2) “on its head”.Footnote 159
Lord Neuberger articulated the classical PoL: unless there is the clearest provision to the contrary Parliament is presumed not to legislate inconsistently with the rule of law.Footnote 160 But on the classic principle it is difficult to see how the provision is unclear in its intent to abrogate the trigger principles. In regard to (1) the section provides the Attorney General can by certificate override a tribunal decision, which shall “cease to have effect”; these words explicitly spell out that a certificate has the legal effect of nullifying the bindingness of the tribunal's decision. And the provision does stand (2) on its head, as the Attorney General plainly is empowered to set aside the tribunal decision. Indeed, Lord Neuberger acknowledged “section 53, expressly enables the executive to overrule a judicial decision”.Footnote 161
On the classic principle the case would be open and shut. However, while Lord Neuberger invoked the classical formulation, the interpretative approach in fact applied was rather different.
The interpretation adopted was that the veto could only be exercised “on few occasions and on limited grounds”.Footnote 162 Two such grounds were where there had been a material change in circumstances or the tribunal decision was manifestly flawed.Footnote 163 This interpretation ruled out the Attorney General overriding a tribunal decision simply because he took a different view on reasonable grounds.
Lord Neuberger's approach involves a deliberate reading down of the power, so as to minimise as far as possible any intrusion upon constitutional principles. The plain words of the provision provide the Minister may overrule a tribunal decision on reasonable grounds. There is no indication of a further limitation beyond reasonableness, and no indication of a restriction on the subject matter of permitted grounds. Lord Neuberger's prescription that the grounds are “limited” and his two examples of permissible grounds were his own creation and find no reflection in the statutory text, while the articulated grounds are unlikely to ever arise. Indeed, Lord Neuberger's interpretative approach is rendered explicit at the tail-end of his judgment: “the common law ensures that [the] grounds are limited so as not to undermine the fundamental principle, or at least to minimise any encroachment onto it.”Footnote 164 This is the proactive PoL in action: provisions which touch basic norms are read down to remove or minimise any incursion.
In contrast, Lord Mance and Lady Hale, while they would have impugned the Attorney General's decision based on substantive review, considered the parent clause clearly had wider application than attributed to it by Lord Neuberger.Footnote 165 Lord Hughes thought the provision's plain meaning was clear,Footnote 166 while Lord Wilson considered Lord Neuberger's approach involved re-writing the provision.Footnote 167
The concerns raised by the proactive principle can be stated briefly. There are transparency concerns: the classic PoL, a well-known public law headline, is invoked but the actual approach is closer to HRA, s. 3. There are concerns that courts are effectively imposing substantive limits on Parliament under the guise of interpretation (while courts applying the proactive principle at common law cannot claim a democratic mandate as they can in applying HRA, s. 3). If courts read down even clear words so as to minimise interference with common-law norms, one may question whether legislation could ever be framed clearly enough to override given norms. In this regard a minority in Privacy International would seemingly have preferred to give up the pretence of interpretation altogether, and openly acknowledge legal limits on Parliament's capacity to oust the High Court's jurisdiction.Footnote 168
IV. Rationale
Lord Hoffmann's statementFootnote 169 in Simms of the rationale for the PoL has emerged as the locus classicus, being repeated as incantation by the House of Lords and Supreme Court.Footnote 170 His twin justifications both relate to Parliament's democratic role. First, if the court does not insist on clear terms there is a risk that the full implications of general or ambiguous provisions may go unnoticed through democratic process. Second, the PoL requires Parliament to confront squarely what it is doing and accept the political cost.
The rationales are not without difficulty. For example, if we wish to know if MPs turned their minds to the matter of rights-infringement why would we only look at statutory text rather than Hansard, committee reports or media statements? Despite a more general opening up of the sources courts consider in interpreting statutes, in applying the PoL courts have rarely used any aid beyond the text.Footnote 171
But my concern here is not with critique. Rather, it is to examine the extent to which Lord Hoffmann's democratic-process rationales, taken on their own terms, provide justification for the new variants of the PoL, specifically the augmented and proactive principles. It has not been uncommon for senior judges to run together these different variants, and explain them all on the basis of Lord Hoffmann's rationale.Footnote 172 But analysis suggests that, as the jurisprudence has evolved, the new variants of the PoL have “outstripped” their stated rationale.
The augmented principle can, prima facie, be sustained by the democratic-process rationales, as the principle requires Parliament to make patent the extent of the envisioned rights-interference it considers justifiable. In a sense this principle is more faithful to Lord Hoffmann's rationale than the classic principle. On the classic principle rights-infringements are authorised where Parliament clearly sanctions some curtailment of rights. But one might consider this falls short of textual evidence that Parliament was aware of the “full implications” of the relevant provisions, given there is no inquiry into whether Parliament recognised the extent of the interference the statute might authorise.
But if one digs a little deeper the augmented principle begins to strike difficulties, considered against the democracy-based rationales. Whether an interference of a given extent is justifiable can typically only be answered on specific facts. Yet is it realistic to expect primary legislation to provide for the extent of permissible interference in the full gamut of factual scenarios? And would it be desirable? Often provision for discretion or rule-making by expert officials is necessary because of a need for decisional flexibility, especially where it is difficult to foresee every scenario that could arise, conditions are changeable and/or a comprehensive set of statutory prescriptions would produce unfairness. In such circumstances is it not disingenuous for a court to hold that an interference of a given extent is not authorised because an interference of that extent has not been explicitly provided for? Where Parliament confers discretion or rule-making power via a provision that clearly authorises rights-interferences, is there not a patent intention to empower the repository to determine the appropriate extent of interference and mediate competing interests case-by-case? As Gageler and Keane JJ. have observed, if legislation makes clear the end sought to be achieved, the legality principle should have limited application; it is of little assistance in working out the meaning of a provision, to invoke a presumption against the very thing the legislation seeks to achieve.Footnote 173
Turning to the proactive principle, Evans and the ouster clause cases show application of the proactive principle involves substituting a provision's plain meaning for a different meaning. Lord Hoffmann's democratic rationale has been invoked in these cases.Footnote 174 But while that rationale can buttress the classical PoL, the proactive principle has nothing to do with enhancing democratic process. No one could suggest that in the legislation in Evans, Anisminic and Privacy International Parliament had shied away from confronting the political cost of its policy choices; its intent was patent in statutory terms. The proactive principle is rather concerned simply to afford maximal protection to basic norms, notwithstanding whether Parliament has squarely confronted the abrogation of those norms;Footnote 175 indeed, judges in applying the proactive principle have said they are not concerned with discerning what Parliament intended.Footnote 176 Thus, it is unsurprising to find the proactive principle included in rights-charters, which have the same principal goal of maximally protecting basic norms. But interestingly, at common law the proactive principle has not been applied in rights cases. Rather, courts have been most willing to take this approach and stretch the stated rationale for the PoL (and the courts’ own legitimacy) where courts are protecting their own institutional role. This suggests that within any hierarchy of common-law values the institutional integrity of the courts has primacy, given this value has attracted the most potent variant of the PoL.
Lastly, turning to Cherry, the democratic-process rationales cannot apply as prerogative powers are not the product of legislative process, but sourced in the common law. Thus, subjecting prerogative powers to basic principles seems simply to be based in a concern to maintain those principles given their importance. This is not to say the approach in Cherry may not have democracy-enhancing effects. It might prompt Parliament to define in legislation the scope of prerogative powers, if it disagrees with how courts have defined those powers, thus placing ancient powers on a democratic footing. But these are not the effects contemplated by Lord Hoffmann.
V. The Principle of Legality and Substantive Review
This last section considers the interrelationship between the PoL and substantive review, given the development of the PoL to include features closely resembling substantive review – specifically, proportionality. It charts an emerging preference on the part of the Supreme Court for developing and applying the augmented PoL ahead of substantive review, and considers possible reasons for this trend and concerns raised by it.
A. An Emerging Preference for Legality
It is important first to distinguish two lines of case law which both involve adoption of proportionality. These have often been conflated by courts and commentators,Footnote 177 but are distinct.Footnote 178 Over time the courts have developed common-law substantive review, specifically in fundamental-rights cases, from the highly deferential Wednesbury standard, first fashioning a new “anxious scrutiny” variant of Wednesbury Footnote 179 and, in Kennedy and Pham,Footnote 180 applying a structured proportionality method. Concurrently, the courts added a proportionality dimension to the PoL in the Leech line of cases, culminating in Unison.
Lord Reed's judgment in Unison rightly treats these as distinct lines of authority. Lord Reed only cites, as support for his approach in Unison, the augmented legality cases, not citing any substantive review case. This is consistent with his prior judgment in Pham which distinguishedFootnote 181 “anxious scrutiny” cases, such as Brind Footnote 182 and Smith,Footnote 183 from PoL cases, such as Leech and Daly.Footnote 184
Notably, in Pham while Lord Reed endorsed proportionality as an aspect of the PoL, he favoured a less structured and more deferential approach to substantive review, observing that there are material differences between common-law substantive review and proportionality as applied in EU and ECHR law.Footnote 185 His view contrasted with the majority, which applied structured proportionality as a form of substantive review at common law. In Evans too Lord Reed preferred to join Lord Neuberger's judgment, which impugned the Attorney General's decision as ultra vires, in contrast to other Justices who impugned the decision based on an aggressive form of substantive review. In AXA Lord Reed similarly held it would be inappropriate to apply substantive review to legislation of the Scottish Parliament, but nonetheless held the PoL applicable.Footnote 186
Thus, Lord Reed's reinvigoration of the augmented PoL in Unison – and recognition of a like principle in Cherry – is perhaps unsurprising given his consistent preference for that principle, where it applies, ahead of developing or applying other substantive bases for intervention.
This reinvigoration of the PoL coincides with a “cooling-off” in development of strong-form substantive review since Pham in 2015. This is evident in the Supreme Court's circumspect approach to developing substantive review in the 2016 Keyu decision,Footnote 187 and its Elgizouli decision in 2020, which reasserted anxious scrutiny Wednesbury as a head of substantive review in fundamental-rights cases (with no mention of proportionality).Footnote 188
As already noted, before Unison the augmented PoL had not been applied by the Law Lords since Daly, decided in 2001. Why has the Supreme Court now reinvigorated the augmented PoL? One might postulate,Footnote 189 given the timing, that the common law was being readied to fill any gap left by potential loss of EU rights through Brexit, and continuing Government threats to repeal the HRA. Or perhaps the new “constitutional” jurisprudence is the sign of a newly established Supreme Court establishing its place in the constitutional order.
These points are part of the story. But crucially the court had already developed substantive review to include a proportionality limit, in cases such as Kennedy and Pham, arguably in response to the drivers in the foregoing paragraph.Footnote 190 Given these advances in substantive review, why should the court now reinvigorate another principle which provides for proportionality, namely the augmented PoL?
B. Legality's Advantages
There may be several possible underlying reasons for resurgence of the augmented PoL, and the Supreme Court's and its new President's preference for the principle, all of which have to do with legality's advantages over substantive review.
First, in terms of standing, if a question is framed as one of legality then public interest standing is easily established, as it was in Unison, where the claim was brought not by an affected individual, but by a union.Footnote 191 Whereas if the normative basis of a claim is breach of an individual's right, it becomes harder to accord standing to anyone other than the right-holder. Thus, under the HRA only a victim has standingFootnote 192 – a point made explicitly by Lord Reed in Unison.Footnote 193
Second, in terms of remedies, the courts have held that if there is no power to do an act in the first place the purported exercise of power is void,Footnote 194 rather than voidable as it might be if the ground of intervention were substantive review. This principle was asserted explicitly by Lord Reed in Unison.Footnote 195 In Cherry, the fact the court intervened formally on the basis of “scope” resulted in the Prime Minister's advice and the subsequent prorogation being void ab initio by operation of law; Unison was given as authority.Footnote 196 In both cases this outcome relieved the court of having to exercise remedial discretion in contentious circumstances, with significant administrative and political ramifications. On ordinary principles those ramifications would have had to be accounted for in any exercise of discretion.
Third, integrating proportionality into the PoL enables courts to apply a strict form of scrutiny in contexts where substantive review has traditionally been applied deferentially. Nearly every case where the augmented principle has been applied has two fundamental features: (1) the empowering provision was extremely broad and (2) it empowered the making of legislation. Thus, the prisoner cases all involved the same parent clause, which provided a broad power to make secondary legislation.
Why do breadth and rule-making matter? First, if an empowering provision is extremely broad that favours a more deferential approach to substantive review, as there are no objective parameters provided by the text against which a court can judge the qualities of a decision. As Lord Reed observed in AXA, the greater the scope for the decision-maker to determine the basis on which its power may be exercised, “the range of decisions which are reasonably open to it are correspondingly widened”.Footnote 197 In contrast, with the PoL, the more general the clause, the more difficult it will be to argue rights-violating conduct is authorised.Footnote 198 Thus, in Witham Laws J. seemed to acknowledge that, given the breadth of the empowering provision, a Wednesbury challenge might fail, yet the court could nonetheless intervene based on the PoL.Footnote 199 In Leech the trial judge rejected a Wednesbury challenge to rules made under the Prison Act's broad empowering provision, whereas the Court of Appeal did not apply Wednesbury, observing “the matter should be approached rather differently”: the key question was “simply one of vires”.Footnote 200 Having asked a different question, the court reached a different outcome.
Second, long-standing precedents on rationality review of secondary legislation, such as Kruse,Footnote 201 favour a very deferential approach. Thus, in Raymond the Law Lords were presented with submissions by Government that, based on authorities including Kruse, the Minister, who is directly answerable to Parliament and acts on expert advice, must be the prime arbiter of what constitutes a reasonable prison rule, and the reasonableness test sets a high threshold for intervention.Footnote 202 Unsurprisingly, given the strength of these arguments, the case was decided on the alternative basis of the PoL. In this connection it is notable that the Supreme Court has reinvigorated the PoL, a powerful tool for scrutinising secondary legislation, just as the legal system is, following Brexit, being-flooded with secondary legislation, made pursuant to very broadly-framed statutory powers.Footnote 203
In addition to the foregoing two features – breadth and rule-making – in most of the augmented legality cases there are further features which would ordinarily favour of a deferential approach to substantive review, but which have not affected judicial application of the PoL, such as: secondary legislation having being laid before Parliament, as with the fees order in Witham Footnote 204; legislation being subject to annulment by resolution of either House of Parliament, as with the Prison RulesFootnote 205; a decision-maker with direct democratic credentials, such as the Minister who promulgates the Prison Rules; and/or a decision-maker with significant expertise and/or experience, such as the four heads of division who approved the fees order challenged in Witham.Footnote 206
Thus the augmented PoL offers courts a way to apply strict scrutiny to the substance of administrative action in cases where substantive review would not permit such scrutiny.
Consider AXA. A majority agreed it would be impermissible to subject legislation of the Scottish Parliament to substantive review given that the Parliament possesses broad plenary powers and has direct democratic legitimacy, and substantive review could draw the courts into politics.Footnote 207 But Lord Reed added that the Parliament's powers are determined by applying principles of statutory interpretation – which include the PoL.Footnote 208 As such, absent express words, the Westminster Parliament cannot be taken to have established, through the Scotland Act, a devolved legislature with the freedom to abrogate basic norms. Therefore, legislative powers which cannot legitimately be subjected to substantive review at all are potentially subject to proportionality constraints via the PoL.
Similarly in Cherry, there were serious questions over the justiciability of the exercise of the prorogation power, so application of substantive review – or any ground other than legality – may have been impermissible.Footnote 209 However, requirements of objective justification were imposed under the guise of legality review, which were equivalent to limits one would associate with substantive review.
More generally, substantive review is dogged by legitimacy concerns. It raises the spectre of courts substituting their decisions for those of the repository specifically empowered to make the decision by Parliament, and straying beyond their proper constitutional role. Thus, it is often repeated that the primary decision is for the repository, with courts necessarily limited to a supervisory role.
In contrast, statutory interpretation is a matter quintessentially for courts, in respect of which courts exercise primary judgement and afford no deference. Thus in DSD, the court, applying the augmented PoL, said: “Even if some degree of infringement is impliedly authorised, it is incumbent on the executive to justify this by a pressing social need and as being the minimum necessary to achieve the objectives sought. These are matters for the court and not for the decision-maker.”Footnote 210 In Unison there was no consideration by Lord Reed of contextual factors which might counsel restraint in applying proportionality, despite that case involving scrutiny of a complex administrative system. In a recent speech Lady Hale, having considered a number of legality cases including Unison, observed: “the courts have been prepared to construe Acts of Parliament in the light of the principle of legality without a hint of deference or pragmatism, indeed some might say quite the reverse.”Footnote 211 Furthermore, were courts to afford deference in applying the PoL this would cut across the basic principle that statutory interpretation is for courts alone; unlike in North America there is no deference on questions of law. In Cherry the court did speak openly of affording the Prime Minister “a great deal of latitude”.Footnote 212 But tellingly that case did not involve statutory interpretation.
Where exercises of statutory power are impugned based on proportionality reasoning under the PoL, the formal basis of intervention is not that the court would have struck the balance differently, but rather that the decision-maker acted outside Parliament's grant of power. As such there is, formally at least, a democratic basis for judicial intervention; the courts can claim to be enforcing Parliament's will, not their own. In turn, the PoL offers courts a way to shore up the legitimacy of intervention on substantive grounds.
C. Substantive Review by Another Name
Thus the PoL has advantages over substantive review, and these likely go some way towards understanding the Supreme Court's reinvigoration of the augmented PoL.
However, the augmented principle raises concerns. There are questions of transparency. The augmented PoL is effectively substantive review by another name. The courts in applying the PoL may claim to be examining questions of scope rather than the qualities of an exercise of power, but, given determination of questions of scope involves considering the substantive justification for an individual exercise of power, including the balance struck between competing interests, this approach collapses any meaningful distinction between scope and exercise. Indeed, that distinction was revealed to be unsustainable a long time ago, as illustrated by collapse of the division between jurisdictional and non-jurisdictional errors.
Ultimately, all of the reasons to be sceptical of strong-form substantive review, such as judges supplanting the statutory decision-maker and substituting their own view of how interests should be balanced, are not washed away by a semantic re-framing of the basis of intervention as legality. Lord Browne-Wilkinson was alive to this in the legality case of Pierson:
Parliament having chosen to confer wide powers on the Secretary of State intends those powers to be exercised by him in accordance with his standards. If the courts seek to limit the ambit of such powers so as to accord with the individual judge's concepts of fairness they will be indirectly arrogating to the court the right to veto a decision conferred by Parliament on the Secretary.Footnote 213
Furthermore, conducting what is effectively substantive review under the guise of statutory interpretation raises a set of additional legitimacy issues. By doing so courts side-step the established framework of substantive review. This matters because that framework has been specifically developed to ensure the legitimacy of judicial scrutiny on substantive grounds, with factors such as the breadth of the power, extent of parliamentary oversight or decision-maker's institutional expertise affecting the degree of leeway afforded to the repository. By affording the repository an appropriate sphere of decision-making autonomy judges avoid exceeding their legitimate role within a supervisory jurisdiction.
Yet, as we have seen, where proportionality has been applied under the PoL, on the one hand, courts have not afforded deference, themselves striking the balance between competing interests, and doing so precisely in circumstances where the framework of substantive review would demand restraint.
On the other hand, should courts seek to respond to these legitimacy concerns by affording deference in application of the PoL this would conflict with the basic principle that courts do not afford deference on questions of statutory interpretation. Indeed, affording deference risks opening a can of worms. Government lawyers would argue that, given deference has been afforded in the interpretation of statutes affecting basic rights, it should most definitely be afforded in interpreting statutes governing socio-economic issues, such as housing or welfare.
Overall, applying the established framework of substantive review instead of the augmented PoL would resolve many of the foregoing concerns. Having said this, there are still real concerns even with courts applying proportionality as a common-law ground of substantive review.Footnote 214 But such an approach would at least resolve the super-added legitimacy concerns associated with applying proportionality specifically through the PoL.
VI. Conclusion
There is more to the PoL than meets the eye. Sitting behind the apparently straightforward principle, that express words are required to oust basic common-law norms, is a complex and sophisticated jurisprudence which gives rise to a host of contentious questions of principle, policy and legitimacy. This article has brought these complexities to light and drawn out and interrogated the normative controversies.
The article has raised questions of coherence and consistency in relation to the judiciary's selection of norms for protection via the PoL, and deeper questions over whether it is properly for courts to delineate those rights and values constitutive of society.
Doctrinal exegesis reveals there is not one PoL but different variants of the principle. The newer variants – the “augmented” and “proactive” principles – make significant inroads into executive discretion, and Parliament's capacity to reshape the common law, even where it manifests its intent by clear words. Courts have not always recognised or acknowledged these more invasive variants.
There are questions over whether the judicially-stated rationale for the PoL, that it operates to enhance democratic process, can sustain these new variants which seem less concerned with promoting healthy legislative practices and more with providing as much protection as possible to judicially-articulated constitutional norms.
Lastly, the addition of a proportionality dimension to the PoL has enabled the courts to apply the PoL effectively as a surrogate form of invasive substantive review, while freeing the courts of the ordinary constraints that apply to substantive review – and maintain its legitimacy – such as the concept of deference. These developments raise significant legitimacy concerns, not so far recognised or addressed by the Supreme Court.
Given the Supreme Court's growing preference for the PoL, one would expect the foregoing issues to be a central focus for the public law academy and higher courts in the years to come.
