I. Introduction
In various publications Professor Rivka Weill has examined the role of “the people”, since 1832, in endorsing fundamental changes to the constitution of the United Kingdom. She proclaims a “revolutionary” thesis: that Britain underwent a “subtle, unacknowledged transition” “from parliamentary sovereignty to popular sovereignty in the nineteenth century [that] created a large gap between the British practice of popular sovereignty and its narrative of parliamentary sovereignty, which persists to this day”.Footnote 1
Popular sovereignty came to be upheld by new constitutional conventions, enforceable by unelected branches of government: in the nineteenth century, by the House of Lords and the monarch, and more recently, by the courts.Footnote 2 “[I]f … political branches do not step in to fulfil their duty to enforce constitutional conventions … the courts should serve as the last line of defense and possess both the authority and the responsibility to enforce them to protect popular sovereignty.”Footnote 3 Due to these “sophisticated enforcement mechanisms”, she rejects the orthodox understanding that because conventions are part of the political rather than the legal constitution, they are consistent with the legal sovereignty of Parliament.Footnote 4 The modern role of the courts shows that “[t]he political constitution converged with the legal constitution”.Footnote 5
I will criticise Weill's thesis, and defend the orthodox understanding, on several grounds. First, in the UK parliamentary sovereignty and popular sovereignty are not competitors; the former is a legal principle and the latter a political one, and they are perfectly compatible. Therefore, secondly, even if it is true that in the nineteenth century there was a constitutional convention requiring popular approval of fundamental constitutional changes, this would have been consistent with parliamentary sovereignty. The supposed requirement would have been inconsistent with parliamentary sovereignty only if it were legally binding rather than a convention. Thirdly, there is little evidence that this convention, if it did exist (which seems questionable), survived after 1911 or has been revived since. Fourthly, there is little or no evidence that courts today have – or even claim to have – the authority Weill attributes to them, to enforce either constitutional conventions, or a new rule of recognition, upholding popular sovereignty.
My third ground challenges Weill's claim that the UK today – like the US – exemplifies a “dualist” rather than “monist” model of lawmaking, which requires that constitutional changes must be approved by the people but ordinary lawmaking need not.Footnote 6 That claim would be vindicated if there were currently a constitutional convention requiring popular approval of constitutional change, which I deny. But even if there were such a convention, my first and second grounds show that this would still be consistent with parliamentary sovereignty.
I will refer to “the people” and “popular sovereignty” without attempting to analyse those concepts or justify their use. That is a task for those such as Weill who advance theories which depend on them. It is not a task I need to undertake, since my purpose is merely to clarify the compatibility of such theories with the legal doctrine of parliamentary sovereignty.
II. Weill's History of Popular Sovereignty in Britain
Weill is convinced that “[t]he U.K. has been operating under a popular sovereignty model that is remarkably similar to the U.S. model for the past two-hundred years. Each iteration of the U.K. model has exhibited a commitment to enacting constitutional change only with the people's endorsement”.Footnote 7 She identifies four main iterations: 1832–1911; 1911–49; 1949 to the mid-1970s; and since the mid-1970s.
She argues that from 1832 until 1911 it became generally accepted that fundamental constitutional changes must be approved by the people before being enacted.Footnote 8 The will of the people was ascertained at elections called specifically for that purpose, often due to prior obstruction of such changes in the House of Lords. If a majority of the electors approved of the change, but the Lords continued to obstruct it, the monarch would threaten to appoint enough new peers to ensure its passage.Footnote 9 The unelected House of Lords therefore possessed a power to review constitutional changes that resembled the US Supreme Court's power of judicial review: both powers could be used to stymie attempts at constitutional change that had not received the constitutionally required endorsement of the people.Footnote 10 A system of parliamentary sovereignty was thereby transformed into one of popular sovereignty.Footnote 11
Weill's analysis of the period between 1911 and 1949 has shifted over time. She originally wrote that the Parliament Act 1911 “made Parliament, and more particularly the Commons, sovereign in practice and not just in law”.Footnote 12 But she has since qualified that assertion; the Act “embodied a transformation from a strong-form model of popular sovereignty to a weakened commitment to popular sovereignty and a parallel strengthening of commitment to parliamentary sovereignty”.Footnote 13 “Although the House of Lords (HL) lost its ability to exercise an absolute veto, it nonetheless maintained a [two-year] suspensory veto that enabled it to protect the constitutional status quo and refer issues to the People's decision.”Footnote 14
This weakened model of popular sovereignty survived until 1949, when the second Parliament Act (assuming its legal validityFootnote 15), together with the Salisbury Convention, inaugurated full parliamentary sovereignty that lasted until the modern use of referendums commenced in the mid-1970s.Footnote 16 By affirming the validity of the 1949 Act (questionably, in her opinion), the court in Jackson in effect held that it had established full parliamentary sovereignty.Footnote 17
But this proved to be merely a “short lapse to parliamentary sovereignty”.Footnote 18 “Since the 1970s, the U.K. returned to a full-fledged popular sovereignty model accompanied by formal written constitutional norms and the exercise of various forms of judicial review over primary legislation.”Footnote 19 The will of the people regarding proposed constitutional changes has been ascertained through referendums.Footnote 20 This provided a “substitute model to express the People's will”, replacing the House of Lords’ suspensory veto that had been so weakened by the Parliament Act 1949.Footnote 21 “The commitment to popular sovereignty, thus, remained strong. Only the means of expressing that will changed.”Footnote 22 Furthermore, judicial decisions in Jackson and Miller II demonstrated that ultimately, the courts may enforce constitutional conventions when other enforcement mechanisms fail to do so.Footnote 23
She concludes that “Britain indeed operates by a popular sovereignty model … [whereby] constitutional change requires the People's consent, expressed through elections, super-majorities or referenda”,Footnote 24 which “contradict[s] the very definition of a parliamentary sovereignty model”.Footnote 25 The orthodox view that the UK constitution has exemplified parliamentary sovereignty from well before 1832 until today is mistaken. Scholars “misunderstood the British model by treating it as a model of parliamentary sovereignty when it is, in fact, a model of popular sovereignty”.Footnote 26
III. Was and Is there a Constitutional Convention requiring Popular Endorsement of Constitutional Change?
A. 1830–1911
In earlier publications, Weill did not use the term “constitutional convention” in the British sense, but appeared to have it in mind. To establish the supposed requirement of popular approval of fundamental constitutional change, she relied on political practice, political rhetoric, and scholarly writings, especially that of A.V. Dicey.Footnote 27 In her most recent article, she explicitly argues that the requirement was embodied in constitutional conventions.Footnote 28
Her first article described Dicey's changing views,Footnote 29 which do seem to support her thesis. He described constitutional conventions as “understandings, customs, or conventions”Footnote 30 which are not laws but constitute “constitutional or political ethics”,Footnote 31 or “the constitutional morality of the day”.Footnote 32 They regulated how the three members of the sovereign body – the King in Parliament – should exercise their discretionary powers (prerogatives in the case of the Crown, privileges in the case of the two Houses).Footnote 33 The ultimate object of constitutional conventions was to ensure that these entities acted consistently with the will of the political sovereign, namely, the electors.Footnote 34 One major convention, concerning disagreements between the two Houses, provided that the Lords should “at some point, not definitely fixed” give way, and if they did not, authorised the Crown to create or threaten to create enough new peers to override their opposition.Footnote 35 “[T]he point at which the Lords must yield or the Crown intervene is properly determined by anything which conclusively shows that the House of Commons represents on the matter in dispute the deliberate decision of the nation.”Footnote 36 “The nature of the proof differs under different circumstances”Footnote 37 and “the will of the nation is often not clearly expressed”.Footnote 38 Elsewhere, he implied that in practice, though not in law, this had led to “a distinction between laws which affect the Constitution and laws which deal with matters of everyday life”.Footnote 39
Weill reports that in a journal article published in 1910, Dicey proposed that Parliament subject itself to a statutory requirement to hold a referendum before enacting any fundamental constitutional change, a requirement that would to some extent be judicially enforceable.Footnote 40 That proposal is intriguing. Although the device of self-entrenching such a requirement was first used in New South Wales in 1930, Dicey anticipated it. He advocated the enactment of a Referendum Act that would list important constitutional statutes in a Schedule, and forbid their amendment or repeal except with the approval of a majority of voters obtained in a referendum. He recommended that the Referendum Act itself be included in the Schedule, which would constitute self-entrenchment.Footnote 41 The Act would also require courts throughout the British Empire to declare invalid any Bill or “so-called Act” that purported to amend or repeal a protected statute without the approval of a majority of voters.Footnote 42 He confidently asserted that this “would clearly, as regards any enactment whatever included in its Schedule … make an alteration impossible without an appeal to the people”.Footnote 43
But Dicey immediately went on to consider the objection that his Referendum Act “would, after all, as long as the sovereignty of the Imperial Parliament is acknowledged, be futile, for Parliament could clearly evade … [it] by adding to any Bill which affected any scheduled Act … words exempting the Bill from the operation of the Referendum Act”.Footnote 44 He said that this objection was “verbally sound”, but “in reality … without force” because “[t]he electors may be trusted to resent an attempt to deprive them of legal power ensured to them by the Referendum Act” and “[n]o party leader will risk this resentment. The Referendum Act will be less subject to change, except by way of extension, than any enactment in the statute book”.Footnote 45 He qualified this by adding that the “latent sovereignty of Parliament” would, nevertheless, enable the Referendum Act to be overridden in a national emergency threatening the safety of the nation.Footnote 46 So Dicey's reference to the “impossibility” of the Referendum Act being evaded is a reference to political, not legal, impossibility – and even in that sense, not complete or therefore genuine impossibility.
Weill suggests that in this 1910 article Dicey may have been trying to deflect objections to his proposal by denying that it contradicted parliamentary sovereignty.Footnote 47 But as early as 1894 he had acknowledged in a private letter that Parliament could bind itself only politically, and not legally:
It is true that an Act of Parliament might repeal or override the Referendum Act itself, but this though a plausible, is not a valid objection. The Referendum Act would practically be secured by the odium which any Ministry or party would incur by depriving the people of their right to be appealed to. I am quite certain that once established the Referendum would never be got rid of by anything short of a revolution.Footnote 48
As well as relying on political rather than legal obstacles, Dicey in these passages seems to deny the principle of implied repeal: Parliament could expressly repeal or override the Referendum Act, but would otherwise be legally bound by it. That interpretation would make sense of his proposal that the courts should be directed to declare invalid any breach of its requirements. This is admittedly contrary to Dicey's previous commitment to implied repeal, but there is no good reason to think that implied repeal is essential to parliamentary sovereignty: Parliament is fully sovereign if it can change the law however and whenever it chooses, even if it must sometimes do so by using express words (which is not difficult).Footnote 49 It is also far from certain that if his Referendum Act had been enacted, the courts would have exercised the power to invalidate a later, inconsistent statute.
Even more importantly, as Weill observes, Dicey's proposal for a Referendum Act was never adopted, and indeed was repudiated when the Parliament Act was enacted in 1911.Footnote 50 At most, then, Weill's account shows that there may have been a constitutional convention in the late nineteenth century requiring that, before a fundamental constitutional change was made, the approval of the people be obtained in an election devoted specifically to the issue. She acknowledges this when she says that “between 1832 and 1911 the British constitution operated under popular rather than parliamentary sovereignty. This British version of the popular sovereignty model was part of the political rather than legal constitution. The political actors and primarily Parliament, not the courts, enforced this model”.Footnote 51
In 1913 Dicey declared that if Parliament passed a bill for Home Rule in Ireland, without the approval of the people being obtained in an election focusing on that issue, the resulting statute would be unconstitutional.Footnote 52 Weill deems this “the ultimate contradiction of Dicey's conception of parliamentary sovereignty”.Footnote 53 It amounted to “recognising the People as the legal sovereign of Britain even without the adoption of the referendum and notwithstanding the passage of the Parliament Act” – the “basic norm of popular sovereignty was entrenched even against the People's will”.Footnote 54
Weill relies here on Richard Cosgrove, who wrote: “By his own definition the laws Parliament passed were constitutional; Dicey himself had done most to banish the idea of unconstitutionality from English law.”Footnote 55 But Dicey had done no such thing: he emphasised that in Britain, the word “unconstitutional” is often used to denote a breach of constitutional convention, or “constitutional ethics”, rather than of lawFootnote 56: “The expression, as applied to an English Act of Parliament, means simply that the Act in question, as, for instance, the Irish Church Act, 1869, is, in the opinion of the speaker, opposed to the spirit of the English constitution; it cannot mean that the Act is either a breach of law or is void.”Footnote 57
For this reason, Dicey used the words “constitutional” and “unconstitutional” to describe compliance with or violations of constitutional conventions, even though he emphasised that British courts could not declare a statute to be void on the ground that it was unconstitutional.Footnote 58 That usage was entirely orthodox, long before Dicey's era.Footnote 59 Therefore, his condemnation of the proposed statute for Home Rule as unconstitutional was not inconsistent with his conception of parliamentary sovereignty.
But it seems debatable whether a constitutional convention, requiring popular assent to constitutional change, existed at the time, notwithstanding Dicey's apparent belief that it was.Footnote 60 The accepted tests for the existence of a convention require, among other things, general agreement among relevant political actors that a certain course of action is obligatory.Footnote 61 But Weill reports that “[t]he House [of Lord]'s exercise of its veto power [to force an election] subjected it to constant and vehement attacks … [and was] portrayed as partisan and biased in favour of the Conservatives. It was also claimed that as an unelected chamber, it should not obstruct the will of the nation as expressed by the elected Commons”.Footnote 62 “Within the Liberal party, some denied the legitimacy of the specific mandate theory altogether.”Footnote 63 Liberal Prime Minister Asquith attacked the theory as a “new-fangled Caesarism which converts the House of Lords into a kind of plebiscitary organ”.Footnote 64 It also seems doubtful that those who passionately advocated extensions of the franchise would have accepted that such an extension had first to receive the approval of a majority of those who currently enjoyed the privilege of being enfranchised; these advocates, presumably, regarded the unreformed electorate as not truly representing the people.
Moreover, if the employment of a veto by the House of Lords had proven a satisfactory remedy, Dicey would surely not have advocated its replacement by a popular veto in a referendum. In 1894, he implied that the House of Lords had lost its power.Footnote 65 Weill does provide evidence of some bipartisanship: Liberals as well as Conservatives sometimes “opposed their respective party's constitutional initiatives” and “demanded the People's decision on the fundamental constitutional issues of the day”.Footnote 66 But this implies that other members of those parties, or at least of the Liberal Party, ignored or resisted these demands, which would be inconsistent with the general acceptance required for a constitutional convention to exist. I do not assert that Weill is wrong; to do so would require more detailed historical investigation than I can undertake. I merely suggest that she herself (commendably) provides some grounds to question her claim.
B. 1911 to Mid-1970s
To support her contention that even after the enactment of the Parliament Act 1911 “Britain was still operating under a popular sovereignty model, though weaker than before”,Footnote 67 Weill relies on the fate of the two Bills whose passage the Act was intended to facilitate: one for Irish Home Rule and the other for disestablishment of the Welsh Church.Footnote 68 The House of Lords was still able to obstruct both Bills, but only delayed their enactment by two years.
As evidence of the continued operation of a popular sovereignty model, this seems slim. Both Bills were enacted in 1914, despite the obstruction of the House of Lords, although their coming into force was delayed until the end of the War. The Welsh Church Act 1914 came into effect in 1920, while Home Rule (the subject of a further Act passed in 1920) became redundant when the Irish Free State was established in 1922. Obstruction by the House of Lords did not lead to public opinion being consulted in an election or referendum devoted to either issue, although as Weill notes, it did enable armed opposition to Home Rule to be instigated in Ulster, which was hardly an expression of the will of the people as a whole.Footnote 69
Weill presents no evidence of a constitutional convention in existence between 1911 and 1949 requiring that the approval of the people be sought for major constitutional changes. Sir W. Ivor Jennings confidently asserted that the convention concerning disagreements between the two Houses, discussed by Dicey (who said that resolution depended on which House better represented the will of the nation), disappeared when supplanted by the Parliament Act 1911.Footnote 70 Weill acknowledges that between 1949 and the mid-1970s, full parliamentary sovereignty was generally accepted.Footnote 71
C. After the Mid-1970s
Whether or not new constitutional conventions regarding referendums have developed since the mid-1970s has been discussed by Stuart White.Footnote 72 He concludes that while there may now be a convention that the outcome of a referendum must be complied with, it is doubtful that there is any convention regarding the holding of referendums. He draws on two recent parliamentary enquiries into the use of referendums, by the House of Lords Select Committee on the Constitution in 2010, and by the House of Commons Public Administration and Constitutional Affairs Select Committee in 2017. He notes that “it is striking that neither the Lords report nor the 2017 Commons Report actually confirm” the existence of a convention in relation to the holding of referendums.Footnote 73
This is corroborated by Michael Gordon's careful account of recent practice regarding referendums, which draws on the House of Lords report as well as the 2018 Report of the Independent Commission on Referendums (established by the Constitution Unit at University College London, comprising senior civil servants, politicians, journalists, academics and other experienced people).Footnote 74 Prominent features of that practice include the lack of both a “specific constitutional framework” and clear prescriptive criteria determining when referendums should be held,Footnote 75 and inconsistency in decisions to hold them, especially in terms of subject matter,Footnote 76 due to “ad hoc, circumstantial decision-making”Footnote 77 driven by “political pragmatism, not constitutional principle”.Footnote 78 It seems clear that political actors have not yet developed “a principled constitutional practice”, and whether they ought to do so is contested.Footnote 79 Both the House of Lords Committee and the Independent Commission doubted the possibility or desirability of establishing a comprehensive list of issues that should require a referendum,Footnote 80 and expressed scepticism about a more formal incorporation of referendums into the constitution.Footnote 81
It seems that at most, it can be argued that Parliament is morally obligated to hold a referendum before making a fundamental change to the constitution, but as yet there is no generally accepted and consistent practice of doing so. To be fair, Weill at one point seems to acknowledge this: “the extent of British commitment to hold referenda is still developing. While some commentators have suggested that its use is partisan and intended to serve governmental needs alone, others are suggesting that it become a prerequisite for conducting any further major constitutional change. If the latter view prevails, the referenda may become a tool of a popular sovereignty model”.Footnote 82
IV. Would such a Constitutional Convention Contradict Parliamentary Sovereignty?
Weill assumes that popular sovereignty and parliamentary sovereignty are rivals: that there is a feature of government called “sovereignty”, which can be possessed by the people, or by Parliament, but not by both (although it can apparently be divided between them).Footnote 83 But that assumption is wrong.
It is true, on the one hand, that parliamentary sovereignty and popular sovereignty cannot co-exist if a constitution gives direct legal effect to the latter principle, by mandating that certain laws may be enacted only by or with the consent of the people, or by empowering the people through initiative and referendum procedures to bypass or override the legislature. On the other hand, parliamentary sovereignty and popular sovereignty can be quite different things, operating in different domains, and are in principle perfectly compatible.Footnote 84 That is the case in the UK, as I will now explain.
Parliamentary sovereignty concerns legal authority to legislate: it consists of legally unlimited authority to enact whatever legislation Parliament chooses, or in other words, authority that is not subject to any judicially enforceable limits.Footnote 85 Popular sovereignty concerns the ultimate source of governmental legitimacy – of moral authority to govern, including to legislate – which it identifies as “the people”.Footnote 86 With respect to legislation, parliamentary sovereignty concerns its legal validity, while popular sovereignty concerns its moral authority. They operate in different domains because they have different sources. Parliamentary sovereignty is part of a fundamental norm of the domestic legal system – the rule of recognition – whereas popular sovereignty is a fundamental norm of political morality.Footnote 87
They also differ in another respect. Parliamentary sovereignty, on the one hand, is only minimally limited by its source. The authority that the law confers on Parliament is not limited by law, except that it may not be used to impose limits on itself (that would require a change in the rule of recognition). Popular sovereignty, on the other hand, is necessarily limited by its source: no person or body, not even the people, can possess authority that is not limited by morality. Morality does not confer authority to violate its own most fundamental norms. It follows that while parliamentary sovereignty consists in being almost entirely unlimited by the norms within its (legal) domain, as well as not being subject to constraint or correction by any other person or body within that domain, popular sovereignty is necessarily limited by the norms within its (moral) domain, and therefore can only consist in not being subject to control or correction by any other body within that domain.
The moral authority of the people, to make a new constitution or to authorise or approve of government decisions, is necessarily subject to moral limits. On the one hand, the moral legitimacy of acts of government including the making of laws necessarily depends, not solely on the will of the people, but also on compliance with fundamental moral principles. The will of the people is an important but not exclusive or decisive consideration. On the other hand, in a system of full parliamentary sovereignty the legal validity of legislation depends solely on its having been enacted by Parliament.
These two kinds of sovereignty, being so different in terms of sources, domains and extent within their respective domains, are in principle perfectly compatible. It is possible for Parliament to possess unlimited legal authority to legislate, and to derive its moral authority to do so from the people. Conversely, it is possible for the people to be the source of moral authority to govern, including to legislate, without themselves possessing any legal authority to legislate.
Dicey was very clear that in the UK, this was not only possible but a reality:
Parliament is, from a merely legal point of view, the absolute sovereign of the British Empire, since every Act of Parliament is binding on every Court throughout the British dominions, and no rule, whether of morality or of law, which contravenes an Act of Parliament, binds any Court throughout the realm. But if Parliament be in the eye of the law a supreme legislature, the essence of representative government is, that the legislature should represent or give effect to the will of the political sovereign, i.e. of the electoral body, or of the nation. The electorate … is a body which does not, and from its nature hardly can, itself legislate, and which, owing chiefly to historical causes, has left in existence a theoretically supreme legislature. The result of this state of things would naturally be that the conduct of the legislature, which (ex hypothesi) cannot be governed by laws, should be regulated by understandings of which the object is to secure the conformity of Parliament to the will of the nation. And this is what has actually occurred.Footnote 88
No doubt, Dicey would have agreed that these understandings amounted to constitutional principles, but only insofar as they were constitutional conventions; he never suggested that they were justiciable. But he did go so far as to assert that “the fundamental dogma of modern constitutionalism … [is that] the legal sovereignty of Parliament is subordinate to the political sovereignty of the nation”.Footnote 89 In that sense, Weill could rightly say that his model was ultimately one of popular rather than parliamentary sovereignty. But it was not one in which the two kinds of sovereignty were incompatible. Parliament was for Dicey, and remains today, the legal sovereign.
It is clear that when, in a typical referendum, the people are asked their opinion, they do not exercise legislative sovereignty, or any kind of legislative power.Footnote 90 They do not initiate the passage of legislation, nor do they participate directly in its enactment. At most, they are taken to approve of Parliament enacting legislation, thereby giving it their moral imprimatur. The legislation must still be enacted by Parliament. The people act as political sovereign and Parliament as legal sovereign.
If Parliament, before exercising its authority on particular occasions, chooses to seek the approval of the people, this does not diminish its legally unlimited power to legislate. Even if Members of Parliament believe themselves (indeed, even if they really are) bound by a moral obligation, or a constitutional convention, to do so, its legislative sovereignty remains intact. Neither moral obligations nor constitutional conventions amount to legal limitations. As I have recently shown, Parliament's sovereignty would be diminished if it were to succeed in enacting a legally binding, and self-entrenched, referendum requirement. That would make it legally impossible for Parliament to enact particular legislation without first obtaining popular approval in a referendum, which would be inconsistent with its continuing to possess legally unlimited legislative authority.Footnote 91 But the UK Parliament has not yet done this, or even attempted it.
Historically, parliamentary sovereignty has been given many different justifications, at different times, aimed at demonstrating Parliament's moral authority to legislate. Ten such justifications are listed in my book The Sovereignty of Parliament: History and Philosophy.Footnote 92 One of them has appeared in various forms ever since the thirteenth century: the claim that Parliament represents the entire community, which is deemed to consent to its Acts.Footnote 93 This justification is compatible with, if not tantamount to, the proposition that the people are sovereign. It has provided the principal justification for parliamentary sovereignty since at least the nineteenth century. When Parliament seeks the approval of the people in a referendum for particular legislation, it seeks a more direct and specific expression of the community's consent.
I note in passing a difficult issue for proponents of popular sovereignty: who exactly are the people, in a state such as the UK, whose opinion should prevail on questions such as devolution or secession? Is it the people of the state as a whole, the people of the nation or region whose self-government is in question, or both? Weill maintains that to legitimise secession, the consent of both the “seceding People” and (separately) of the “remaining People” is required, with negotiation between them required if they disagree (she does not discuss what should happen if agreement cannot be reached).Footnote 94 When it came to Home Rule for Ireland, opponents such as Dicey insisted that the people of the UK as a whole rather than just those of Ireland had to approve, expecting that this would be unachievable.Footnote 95 Ulster Unionists insisted that the approval of the people of their region was essential.Footnote 96 It is necessary for Parliament to choose between competing demands of this kind, as well as to determine other matters such as precisely what question should be put to voters. That makes the practical exercise of popular sovereignty dependent on the sovereignty of Parliament.
When Parliament defers to the will of the people it does not diminish its legislative sovereignty, which should always be exercised in accordance with its Members’ considered judgments about what it morally ought to do. They might judge that that a certain law is so important and controversial that they morally ought to ascertain the will of the people. Contrary to Weill's suggestion, it is therefore not “unnatural” or “incomprehensible” for the UK to be committed to parliamentary sovereignty but also to test proposed constitutional changes in referendums.Footnote 97 She says that Brexit should make no sense to anyone committed to parliamentary sovereignty: “How could a nation committed to parliamentary rule find itself bound by a consultative referendum's result”, especially “after the majority in Parliament had repeatedly opposed Brexit!”Footnote 98 But this ignores the ambiguity of the word “bound”: Parliament might well find itself morally bound by the result of a referendum, even though it is not legally bound by it.
Weill may be right to say that the British constitution is “based on popular sovereignty”.Footnote 99 But that is in the domain of moral legitimacy. In the domain of law, it is based on parliamentary sovereignty. The strongest claim she can plausibly make is that the constitution is based on both kinds of sovereignty.
V. Law, the Rule of Recognition and the Role of the Judiciary
Weill sometimes suggests that the practice she purports to describe, of requiring the people's assent to fundamental constitutional changes, and the justifications given for it by politicians and constitutional scholars such as Dicey, “are paramount for understanding Britain's ultimate rule of recognition … as one of popular sovereignty”.Footnote 100 “Where sovereignty (the highest authority in the legal system) lies – with the People or with parliament – is defined externally, by the political facts and is the product of assertion of power. It later receives the courts’ recognition that it forms the basis of legal validity in the system.”Footnote 101
Here, we again observe Weill's assumption that parliamentary sovereignty and popular sovereignty are competitors for the position of “highest authority in the legal system”. But it is possible for Parliament to be the highest legal authority in the system, while the people are simultaneously the highest moral or political authority. A rule of recognition concerns legal authority: it governs the recognition of norms by legal officials, particularly judges, as valid laws within a legal system.Footnote 102 Weill adduces no evidence that British courts, in any of the periods she discusses, would have held invalid an Act of Parliament that had not been endorsed prior to its enactment by a majority of the people, in an election or referendum devoted specifically to the proposal to enact it. It is almost inconceivable that any court would have done so. As Weill acknowledges, before 1911 the practice she describes was “enforced” not by the courts, but by the House of Lords,Footnote 103 which of course was an element of Parliament itself. The sovereignty of Parliament can hardly be said to be diminished due to one of its own Houses declining to pass proposed legislation.
Weill maintains that “because the 1911 [Parliament] Act was enacted by Parliament along with the people … [it] amounts to a constitutional document. The decision to amend the Act is thus left to the People”.Footnote 104 Moreover, she takes the same view of every “constitutional statute” enacted only after receiving the approval of the people in a special election or referendum:
we treat all constitutional law enacted by the People as enjoying the same constitutional status and entitled to similar judicial protection. Once the constitutional enactment reached the level of a “We the People” enactment (in terms of the breadth, depth, and decisiveness criteria), the strength of the courts’ commitment to upholding it should not vary according to the strength of its underlying claim to a popular imprimatur.Footnote 105
She says it is arguable that in Jackson, the Appellate Committee of the House of Lords should have declared the Parliament Act 1949 to be invalid, because it amended the 1911 Act without the explicit approval of the people.Footnote 106 By not declaring the 1949 Act to be invalid, she claims, the court contributed to a change in the rule of recognition.Footnote 107 The 1949 Act already amounted to a shift towards parliamentary sovereignty, but this was fortified by the decision in Jackson which marked “the ultimate triumph of parliamentary and, more specifically, the Lower House's sovereignty”.Footnote 108
But the fact that the 1911 Act and other statutes can reasonably be labelled “constitutional documents” does not entail that they are legally protected from amendment or repeal by Parliament using its ordinary lawmaking procedures. This remains true even if we define a “constitutional statute” narrowly, as “any statute dealing with fundamental constitutional matters, which was enacted only after and because it received the approval of the people in an election or referendum devoted specifically to the proposal to enact it”. To assert that such a statute may, as a matter of law, be repealed only through the same or an equivalent process, is to make a radically novel claim. It also involves a non sequitur. If the original requirement – that the prior approval of the people had to be obtained before such a statute's enactment – was one of constitutional convention rather than law, then surely any proposed requirement that such a statute can be amended or repealed only by the same process must equally be a matter of constitutional convention rather than law (unless, perhaps, a self-entrenched manner and form requirement is included in the statute). By suggesting that such statutes are entitled to judicial protection against repeal, Weill obliterates the distinction between constitutional convention and law, and ignores the pre-1911 dependence of the supposed convention on enforcement by the House of Lords – a House of Parliament itself – rather than by the courts.
Weill might reply that I have fallen into the same error as others: “Constitutional scholars of the twentieth century have mistakenly focused on a legal analysis of sovereignty rather than a historical and factual analysis of the workings of the British constitution. Substance and practice in constitutional law are no less important than legalistic appearances.”Footnote 109
This is unfair to those scholars, who paid enormous attention to the historical and factual basis for constitutional conventions, and thereby went far beyond “legal analysis”.Footnote 110 But parliamentary sovereignty is about legality: it asserts that Parliament's legislative authority is not subject to legal, judicially enforceable limits. That is a matter of legal substance and practice, not mere “legalistic appearance”.Footnote 111
Weill attempts to bolster her thesis by relying on recent developments in the role of the judiciary, which have enhanced their ability to protect constitutional norms. But it seems doubtful that these developments play a significant role in enhancing popular sovereignty, in her sense of requiring public approval of constitutional change. They have more to do with enhancing judicial protection of certain constitutional norms against legislative incursion.
One of these developments is the relatively recent judicial recognition of “constitutional statutes”, and reputed requirement that they may only be repealed expressly. Weill insists that this “is inconsistent with the norms of parliamentary sovereignty that allow for implicit repeal of legislation”.Footnote 112 But in fact, implied repeal of such statutes remains possible, provided that the implication is “necessary”.Footnote 113 As John McGarry and Samantha Spence correctly summarise the current position, “it is fair to say that … the implied repeal rule has been amended [note: not excluded] with regard to constitutional statutes – so that such statutes may only be amended by express words in, or as a necessary implication of, a later Act of Parliament”.Footnote 114 But even if implied repeal of these statutes were now completely impermissible, implied repeal is not a necessary feature of parliamentary sovereignty, as previously noted.Footnote 115 Dicey did state that Parliament could amend or repeal any law, even a constitutional one, in the same manner and as easily as any other law, and he described this as a “trait” (apparently meaning an “essential feature”) of parliamentary sovereignty.Footnote 116 But that ability was neither part of nor entailed by his definition of parliamentary sovereignty,Footnote 117 and the better view is that he was wrong to suggest that it is essential to it.Footnote 118
Another recent development Weill cites is the enactment and operation of the Human Rights Act 1998 (UK), which obligates courts to interpret statutes, as far as possible, to be consistent with the rights it protects. “The more the courts interpret this obligation as requiring purposive interpretations of statutes to prevent noncomformity with the Convention, the more they deviate from traditional norms of parliamentary sovereignty.”Footnote 119 “In substance, British courts exercise judicial review under that HRA. The Act constitutes a superior law against which other legislation is measured. That prior acts – in this case the HRA – prevail and implicitly overrule later legislation contradicts the fundamental hallmark of parliamentary sovereignty.”Footnote 120
I have previously discussed and rejected claims of this kind.Footnote 121 Parliament enacted the HRA and retains the legal power to repeal it at any time. Weill emphasises that it may be immune to implied repeal,Footnote 122 but we have already dealt with that point.Footnote 123 Occasionally, courts in applying section 3(1) of the Act interpret statutes so creatively that they in effect rewrite them, whether or not they ought to.Footnote 124 Commentators disagree about the extent to which this has happened. Richard Ekins says that in “many cases” courts have adopted “unreasonable readings that depart from the will of Parliament”.Footnote 125 But a very recent survey of 593 cases, over a 13-year period, reached this conclusion:
Overall, we found relatively few cases in which s.3 was decisive to a case's outcome. However, when it was, its use, although important, was not radical with the courts being vigilant to not undermine Parliament's intention … rather we found that s.3 was often used to address unforeseen drafting issues or factual situations that clearly fell within the overall intention of the legislative scheme. This is not the same as interpreting statutes in a manner inconsistent with Parliament's intention and can in fact ensure that Parliament's overarching intention is realised.Footnote 126
Weill also places considerable emphasis on statements by three Law Lords in Jackson expressing doubts about the doctrine of parliamentary sovereignty, and suggesting that the courts possess a common law power to invalidate statutes.Footnote 127 She speculates that these statements were motivated by a perception that the judiciary had to step up and replace the House of Lords in upholding the constitution.Footnote 128 But relying on these notorious statements is problematic, because they were (1) obiter dicta, (2) made by a minority (three out of nine) and (3) based on known falsehoods.Footnote 129 In addition, they have been contradicted by the subsequent, unequivocal endorsement of parliamentary sovereignty in the two Miller cases.Footnote 130
Weill also relies on the decision in Miller II, in which the Supreme Court arguably enforced a constitutional convention, by declaring unlawful a prorogation of Parliament that had temporarily prevented it from exercising its legislative and supervisory functions in relation to Brexit.Footnote 131 Once again, she claims, the Court stepped up to replace the House of Lords as guardian of the constitutional status quo.Footnote 132 Judicial review thereby replaced the Lords’ legislative veto, and “the political constitution merged with the legal constitution”.Footnote 133
But there are two difficulties with this claim. First, the decision in Miller II was explicitly based partly on giving effect to the principle of parliamentary sovereignty, which makes it difficult to construe as undermining that principle.Footnote 134 The Supreme Court reviewed the legality of an exercise by the executive government of a prerogative power, not the validity of a statute. Secondly, although the decision was arguably based partly on giving effect to the constitutional convention of responsible government, elevated to a justiciable “constitutional principle”, that provides little support for a claim that the Court may enforce constitutional conventions – if any are extant – upholding popular sovereignty as Weill conceives of it.Footnote 135 In particular, it offers no support for the proposition that the Supreme Court has authority to prevent the enactment of, or to invalidate, a statute that brings about a major constitutional change without a referendum first being held. That would contradict the Court's explicit affirmation of parliamentary sovereignty.
VI. Conclusion
I have found Weill's historical evidence interesting, and her arguments provocative, but I am unable to accept her claim that the UK constitution, for the last 200 years, has exemplified popular rather than parliamentary sovereignty.
I have offered various grounds for defending the orthodox understanding that the constitution has exemplified parliamentary sovereignty, even if it has also exemplified popular sovereignty in some sense of that term. First, in the UK parliamentary sovereignty and popular sovereignty are not competitors; the former is a legal principle and the latter a political one, and they are perfectly compatible. Therefore, secondly, even if in the nineteenth century there were a constitutional convention requiring popular endorsement of fundamental constitutional changes (which seems questionable), this would have been consistent with parliamentary sovereignty. The supposed requirement would have been inconsistent with parliamentary sovereignty only if it were legally binding. Thirdly, this convention, if it did exist, did not survive after 1911, nor has it been revived more recently. Fourthly, there is little or no evidence that the judiciary today has – or even claims to have – the authority Weill attributes to it, to enforce either constitutional conventions or a new rule of recognition upholding popular sovereignty in her sense of requiring public approval of constitutional change.
I therefore reject Weill's thesis that “[t]he U.K. and U.S. have shared a common constitutional model of popular sovereignty for over two hundred years. This model demarcates constitutional from regular law, entrenching the former and entrusting unelected branches with the role of guardians of the constitutional status quo”.Footnote 136
