I. Introduction
A standard narrative about the constitution of the past 50 years is one of transformation. The core constitutional doctrine of parliamentary sovereignty is understood to have undergone – depending on one’s reading – either a quick death following R. v Secretary of State for Transport, ex parte Factortame Ltd. (No. 2) Footnote 1 or a slow one, starting sometime around then and continuing to this day. The basic idea is summed up by Lord Steyn in R. (on the Application of Jackson) v Attorney General:
We do not in the United Kingdom have an uncontrolled constitution […]. In the European context the second Factortame decision made that clear […]. The settlement contained in the Scotland Act 1998 also points to a divided sovereignty. Moreover, the European Convention on Human Rights as incorporated into our law by the Human Rights Act 1998, created a new legal order […]. The classic account given by Dicey of the doctrine of the supremacy of Parliament, pure and absolute as it was, can now be seen to be out of place in the modern United Kingdom.Footnote 2
Lord Steyn’s suggestion is that these various constitutional changes ushered in a transformation not just of the controls over Parliament but of the very notion of parliamentary sovereignty. The basic idea is that while Parliament of the past held absolute authority, subject only to the controlling force of political opinion, its power is now increasingly subject to legal qualification. Statutes could be disapplied if conflicting with EU law (and still can today, after Brexit, albeit in more limited circumstances).Footnote 3 In turn, the devolved institutions have their own lawmaking powers that cut across Parliament’s. Finally, the courts are required to interpret legislation “in so far as is possible” to make it compatible with human rights. These changes are understood to speak of a limited rather than unlimited sovereignty, a sovereignty that is conditioned by other constitutional doctrines. Hence, Lord Steyn concluded that were Parliament to act contrary to some further constitutional principle, it may be open to the courts, acting at their own behest, to refuse to give effect to such legislation: “In exceptional circumstances involving an attempt to abolish judicial review or the ordinary role of the courts, the […] Supreme Court may have to consider whether this is a constitutional fundamental which even a sovereign Parliament acting at the behest of a complaisant House of Commons cannot abolish.”Footnote 4
Many commentators have shied away from Lord Steyn’s radical conclusion.Footnote 5 Indeed, it is plausible to characterise the era in which he was writing as the high watermark for legal constitutionalism in the UK. Nonetheless, there is broad consensus that the model of parliamentary power that the constitution embraces has undergone transformation and that this transformation survives any climbdown in judicial attitude.Footnote 6
What I am going to do in this article is to argue that this narrative is mistaken. My core claim is that the constitutional innovations standardly understood to require a reconceptualisation of sovereignty are based on a misunderstanding of that doctrine, a misunderstanding attributable in the first place to A.V. Dicey himself. Famously, Dicey held that sovereignty involved two ideas. First, that Parliament has the ability to make or unmake any law. Second, that no other body has the right to set aside parliamentary legislation. What I will suggest is that the first of these principles qualifies the second in a way that went unaddressed by Dicey. While sovereignty implies that no other body may override legislation of its own volition, it may do so if empowered by Parliament. Recognising this, I suggest, reveals the hidden radicality of sovereignty as a tool for constitutional change. It also explains why Lord Steyn’s argument for an independent right on the part of the courts to override Parliament’s will fails.
Section II explains Dicey’s account of sovereignty. In Section III, I outline the argument standardly used to account for the transition from unqualified to qualified sovereignty: William Wade’s interpretation of Factortame (No. 2) as upending the rule of implied repeal. I go on to show how this argument has been buttressed by the doctrine of constitutional statutes. Section IV subjects Wade’s analysis to critical scrutiny, explaining how it misses its target. This places in doubt the doctrine of constitutional statutes, albeit for reasons different than those recently averred to by the Supreme Court in Re Allister’s Application for Judicial Review.Footnote 7 Section V explains the real tension between the Diceyian understanding of sovereignty and the holding in Factortame (No. 2). According to Dicey no other institution has the right to set aside legislation and yet this is precisely what the court did in Factortame (No. 2), disapplying the Merchant Shipping Act 1988 for the sake of EU law. I go on to advance a reading of sovereignty in accordance with which Parliament has always been capable of empowering other institutions to override its own legislation. Such an understanding is, I argue, implicit in the devolution settlements, the Parliament Acts of 1911 and 1949, and the wide variety of Henry VIII powers bestowed on the executive. It also, I suggest, licenses the result in Factortame (No. 2). Finally, Section VI explores some implications of this view of sovereignty, both for the practice and theory of the constitution.
A brief explanatory note: much of this article looks to arguments developed during the period of the UK’s membership of the EU and perhaps, more crucially, to the analysis of the mechanism giving effect to EU law within the domestic order. The UK is, of course, no longer a member of the EU and so one might wonder what is to be gained from revisiting these arguments. The first thing to note is that under the terms of withdrawal, Northern Ireland remains subject to directly effective EU law and many of the rights of EU nationals in the UK are a consequence of the same. To the extent that directly effective EU law challenges sovereignty, the challenge remains.Footnote 8 More importantly, however, contemporary constitutional thought has absorbed a range of “lessons” from the period of the UK’s EU membership; that sovereignty today means something different than it did before, that the doctrine of constitutional statutes is an established part of the architecture of the constitution and that the courts are now, more fundamentally than ever, guardians of the constitutional order.Footnote 9 I will argue that each of these ideas is mistaken, but to make these arguments I need to return to the source of such understandings and this begins with Dicey.
II. Dicey’s Constitution
What, then, are the details of Dicey’s understanding of sovereignty? In the first place, Parliament has the capacity “to make or unmake any law whatsoever”.Footnote 10 In contrast to the orthodoxies of modern constitutionalism, it is not bound by any entrenched bill of rights or similar set of overarching legal principles. As such, it is open to Parliament to enact law dealing with any subject matter and for it to be enforced. This implies, as Dicey famously asserted, that Parliament is constitutionally entitled to pass legislation mandating the killing of “all blue-eyed babies”.Footnote 11
So far so straightforward. In fact, however, there exists an ambiguity in the idea that Parliament has the power to make any law. Because one thing that such a body might do is to enact legislation that would impinge on its own future lawmaking powers. Does Parliament’s capacity to make any law imply that the courts are bound to give effect to such a statute or should they not do so precisely because the consequence of such legislation would be to deny any subsequent Parliament the ability to make any law that it chose. In parallel to the famous theological puzzle about an all-powerful god – is God so powerful that they can create a rock that even they cannot lift or are they so powerful that they are unable to create such a rock? – there is a need to resolve the paradox. The UK is widely understood to accept a continuing version of sovereignty – in accordance with which Parliament retains its lawmaking capacity across time – over a self-embracing conception of the doctrine – in accordance with which it would be able to impose limits on its own future lawmaking power.Footnote 12 As a necessary condition of its continuing authority Parliament may repeal any prior legislative enactment. Further, as an important aspect of the way in which continuing sovereignty is manifest in the UK, new legislation “impliedly repeals” any inconsistent prior legislation. As it will become important later, it is worth us spending some time on this idea.
The misleadingly named doctrine of implied repeal holds that in the case of conflict between two statutes the courts should give effect to that which was enacted most recently, even if that legislation does not expressly repeal or exempt itself from the effect of that which came before. The most cited understanding of the rule comes from Church Wardens and Overseers of West Ham v Fourth City Mutual Building Society: “The test of whether there has been a repeal by implication by subsequent legislation is this: Are the provisions of a later Act so inconsistent with, or repugnant to, the provisions of an earlier Act that the two cannot stand together?”Footnote 13
So, for example, if Parliament at one point enacted legislation providing that the national speed limit should be 70 miles per hour (mph) but later enacted a law providing for a 60-mph limit which made no mention of the earlier legislation, the courts should give effect to the 60-mph rule because the later law has “impliedly repealed” the earlier. I say the doctrine is misleadingly named because in the first place the rule has no special connection to anything implied by the later legislative Act. It is a common law rule that regulates the relationship between statutes.Footnote 14 Second, the later legislation does not in fact repeal the earlier law. Rather the courts give effect to it in preference to the prior law. As such, were Parliament to repeal the 60-mph limit, the legislation providing for a 70-mph speed limit would again take effect. The doctrine therefore provides a way of resolving conflict between laws, rather than a means by which one repeals the other.Footnote 15
Finally, what if Parliament were to attempt to prevent the express or implied repeal of its own legislation? According to the continuing model of sovereignty the courts should refuse to give effect to any such provision, as an attempt to do that which Parliament is incapable of doing: namely, binding its successors. Consider the famous case of Ellen Street Estates v Minister of Health.Footnote 16 This concerned the Acquisition of Land Act 1919 which created a regime for compulsory purchase. The Act also provided, by section 7(1), that any future legislation “shall have effect subject to this act, and as far as inconsistent shall not have effect”. The question before the court was about the bearing of this section on the operation of the Housing Act 1925, which contained conflicting provisions. The court both refused to give effect to Acquisition of Land Act 1919, section 7(1) – as an attempt to bind future Parliaments – and decided the case according to the 1925 Act, in line with the doctrine of implied repeal.Footnote 17
This, then, is the first aspect of Dicey’s understanding of sovereignty: that Parliament has the capacity to make any law, subject to the qualification that it cannot bind itself. It is buttressed by a second. According to Dicey, sovereignty implies “that no person or body is recognised by the law of England as having a right to override or set aside the legislation of Parliament”.Footnote 18 No court of law may refuse to apply a duly enacted statute. No member of the executive may elect to dispense with their statutory duties. It is easy to see how the two aspects of the theory are natural counterparts. Parliament’s authority to create law without restriction would be theoretical only, if that very same law could be overturned by the courts at their own choosing. Dicey therefore considered it a corollary of Parliament’s capacity to make any law that no other institution should have the right to “override or set aside” its legislation. In this way the defensive second limb of the theory is understood to buttress the offensive first.
Notice, however, that things cannot be quite so simple, for the very idea of continuing sovereignty gives rise to two circumstances in which the courts must refuse to give effect to parliamentary legislation. First, where later legislation conflicts with earlier the courts should give effect to the later in place of the earlier. Second, where legislation attempts to bind future Parliaments, the courts are required to disregard such provisions. Both ideas played key roles in Ellen Street Estates v Minister of Health. First, the court refused to give effect to the terms of compulsory purchase provided for by the Acquisition of Land Act 1919. Second, it disregarded Acquisition of Land Act 1919, section 7(1), which provided that future inconsistent legislation was to have no effect. In doing so it reaffirmed constitutional orthodoxy.
Two implications of the Diceyian picture deserve emphasis.Footnote 19 In one sense the model is hierarchical, because it places the legislation of Parliament above all other sources of law.Footnote 20 Parliament has unlimited capacity to make law, subject to the proviso that it cannot bind itself. Additionally, neither the courts nor the administrative state may curtail its authority. In another sense, however, the model is flat, because it confers on all Acts of Parliament equal worth: each is legally operative up and until repealed or overridden by subsequent Act of Parliament. On this understanding “neither the Act of Union with Scotland nor the Dentists Act 1878 has more claim than the other to be considered a supreme law”.Footnote 21 Both are ordinary legislation, capable of being overridden by any subsequent legislative Act.
III. Wade’s Factortame
Diceyian orthodoxy is understood to have held, largely without interruption, until 1990, when, according to William Wade’s highly influential analysis, a “constitutional revolution” took place.Footnote 22 Fortunately, the revolution was bloodless. It was also recorded in the pages of the law reports. The source of Wade’s alarm was the second in the series of Factortame cases.Footnote 23 In this the House of Lords famously “disapplied” the Merchant Shipping Act 1988 which would otherwise have frustrated the exercise of rights recognised in EU law, specifically the rights of Spanish fishers to trawl in UK waters. As justification for its novel decision the court cited the European Communities Act 1972, itself an Act of Parliament. This Act was created to give domestic legal effect to the law of the EU and provided, by section 2(4), that “any enactment passed or to be passed […] shall be construed and have effect subject to [EU Law]”. Lord Bridge gave the lead judgment:
[W]hatever limitation of its sovereignty Parliament accepted when it enacted the European Communities Act 1972 was entirely voluntary. Under the terms of the Act of 1972 it has always been clear that it was the duty of a United Kingdom court, when delivering final judgment, to override any rule of national law found to be in conflict with any directly enforceable rule of Community law.Footnote 24
And this is what the court did, for the first time ever applying some external source of law in preference to an Act of Parliament. Here is Wade’s analysis of the case:
When […] the House of Lords […] forbid a minister from obeying an Act of Parliament, and the novel term “disapplied” had to be invented to describe the fate of the Act, it was natural to suppose that something drastic had happened to the traditional doctrine of Parliamentary sovereignty. The established rule about conflicting Acts of Parliament, namely that the later Act must prevail, was evidently violated, since the later Act in this case was the Merchant Shipping Act 1988, yet it was disapplied under the European Communities Act 1972 […]. When that Act was nevertheless held to prevail it seemed to be fair comment to characterise this, at least in a technical sense, as a constitutional revolution. The Parliament of 1972 had succeeded in binding the Parliament of 1988 and restricting its sovereignty, something that was supposed to be constitutionally impossible.Footnote 25
Thus, Wade suggested, Diceyian orthodoxy was fundamentally disrupted.Footnote 26 Two things about the analysis deserve note. First, it invokes without naming the doctrine of implied repeal, this being the “established rule about conflicting Acts of Parliament”. Second, while Lord Bridge lay responsibility for any constitutional upheaval at the door of Parliament, Wade’s argument had as its target the court, this being the body that “violated” “the established rule” and disapplied the relevant Act of Parliament.Footnote 27
How, though, could the House of Lords have been an agent of “constitutional revolution” when, as Lord Bridge made clear, it was merely giving effect to the terms of the European Communities Act 1972? Doesn’t sovereignty mean giving authority to Acts of Parliament? It is here that the particulars of continuing sovereignty become important. According to the “established rule” of implied repeal, later legislation must prevail over earlier law, yet here the European Communities Act 1972 was applied in preference to the later Merchant Shipping Act 1988. Wade’s claim, then, is that by deciding Factortame (No. 2) the way it did, the court disregarded this fundamental rule of the constitution, with profound implications for “the traditional doctrine of Parliamentary sovereignty”.Footnote 28
Judicial affirmation of core elements of this analysis, albeit on friendlier terms, came six years later in Thoburn v Sunderland City Council.Footnote 29 Here Lord Justice Laws suggested that the courts had “in recent years [created] exceptions to the doctrine of implied repeal”.Footnote 30 This was most apparent “in the field of European law itself, in the Factortame case”.Footnote 31 What justified such actions? The courts had created, Laws L.J. theorised, “a hierarchy of Acts of Parliament: as it were ‘ordinary’ statutes and ‘constitutional’ statutes”.Footnote 32 While “ordinary statutes [could] be impliedly repealed […] constitutional statutes [could] not”.Footnote 33 On what basis was a statute to be deemed constitutional? Such a law, Laws L.J. suggested, was one that either “conditions the legal relationship between citizen and State in some general, overarching manner, or […] enlarges or diminishes the scope of […] fundamental constitutional rights”.Footnote 34 Examples of constitutional statutes included the Magna Carta, the Bill of Rights 1689 and the Act of Union 1800. More recent examples included the devolution statutes, the Human Rights Act 1998 and, crucially, the European Communities Act 1972.Footnote 35 With this argument Laws L.J. gave theoretical foundation to what Wade had previously diagnosed as the special significance of Factortame (No. 2), as subverting the rule of implied repeal. What was for Wade a source of alarm became for Laws L.J. cause for celebration, with the courts taking up their role as arbiters of the constitutional order: “the traditional doctrine [of sovereignty] has […] been modified. It has been done by the common law, wholly consistently with constitutional principle.”Footnote 36
Laws’ reasoning in Thoburn v Sunderland City Council was, in turn, endorsed and amplified in R. (on the Application of HS2 Action Alliance Ltd.) v Secretary of State for Transport.Footnote 37 Here multiple members of the Supreme Court accepted obiter, not only that there existed constitutional statutes, but that in the case of dual loyalties between constitutional statutes and other constitutional doctrines that the priority of each would be judged based on the relative importance of the principles that they instantiated.Footnote 38 Hence, were EU law to require the courts to pass judgment on the merits of the legislative process in a manner inconsistent with the principle of parliamentary privilege, as enunciated in Article 9 of the Bill of Rights 1689, the courts would protect the latter in preference to the former.Footnote 39 This, Mark Elliott argued, suggested not only a distinction between ordinary and constitutional statutes, but between constitutional statutes and other constitutional principles at different levels of fundamentality.Footnote 40
It should be clear how far from Dicey’s model of the constitution this takes us. While Dicey held that all statutes had equal worth, subject to the principle that the earlier gives way to the later, Thoburn v Sunderland City Council suggests a hierarchy of legislative enactments. Going further, the Supreme Court in R. (HS2 Action Alliance) v Secretary of State for Transport implies a hierarchy within the category of constitutional statutes and between these and other “principles […] fundamental to the rule of law”.Footnote 41 Second, the picture complicates the hierarchy built into Dicey’s model. Where Dicey held that parliamentary sovereignty came before all other constitutional principles, these developments appear to suggest an understanding on which other principles stand alongside that doctrine, jockeying for space in a pluralistic landscape.Footnote 42 Whether this analysis is correct will be the subject of the next section.
Before we approach this issue, however, it is worth spending some time on the recent case of Re Allister, which has been taken by some, at least, to render the question of constitutional statutes “academic”.Footnote 43 The case concerned the relationship between the Act of Union 1800 and the Northern Ireland Protocol, an international agreement between the EU and the UK. The Protocol kept Northern Ireland within the EU single market in a manner that was inconsistent with Article VI of the Act of Union 1800, which provides that “all parts of the United Kingdom […] shall […] be under the same […] regulations of trade”. The Protocol was given domestic effect by the European Union (Withdrawal) Act 2018, which requires, by section 7A(3), that “every enactment […] is to be read and have effect subject to” the Protocol’s requirements. The argument before the court was that the constitutional status of the Act of Union 1800 rendered it immune from being overridden by the European Union (Withdrawal) Act 2018. It had not been expressly repealed and could not be impliedly repealed by that Act. As a result, the courts should give effect to the requirements of the Act of Union 1800 in preference to those of the Northern Ireland Protocol. The Supreme Court gave this argument short shrift. Lord Stephens held that “the suspension, subjugation, or modification of rights contained in an earlier statute may [always] be effected by express words in a later statute”.Footnote 44 Here Parliament had expressly provided for what should happen in the case of any conflict between statute and the Protocol. It was therefore unnecessary to resolve (hence, “academic”) whether the Act of Union was a constitutional statute.
What to say about this argument? In the first place the question of timing is a red herring, because as the court itself acknowledged, European Union (Withdrawal) Act 2018, section 7A(3) applies as much to legislation created after the enactment of the Act as before it.Footnote 45 In turn, whether the Act of Union 1800 is to be treated as a constitutional statute depends on the veracity and scope of that doctrine. If it is necessary to explain the result in Factortame (No. 2), then it is certainly implicated by the facts of Re Allister, the crucial difference being that while the Merchant Shipping Act 1988 could not, on Laws L.J.’s understanding, be construed as a constitutional statute, the Act of Union 1800 certainly could. To determine how to approach this issue we should go back to Factortame (No. 2).
IV. Are there any Constitutional Statutes?
Recall that on Wade’s reading Factortame (No. 2) broke new ground by denying “the established rule about conflicting Acts of Parliament, namely that the later Act must prevail”. But was this rule implicated in the case?Footnote 46 No doubt there existed conflict between the terms of the Merchant Shipping Act 1988 and rights available under EU law. The Merchant Shipping Act 1988 required that companies fishing in British waters be 75 per cent. UK owned, whereas EU law confers rights not to be discriminated against on grounds of nationality when engaging in commercial activity. But these latter rights are not creatures of any statute. They are the creation of an international treaty.Footnote 47 The question is whether there was any incompatibility between the terms of the European Communities Act 1972 and the Merchant Shipping Act 1988.
We can say that two statutes conflict when one creates rights and obligations inconsistent with the other, when one renders true a proposition of law that the other denies. Hence, the statute imposing the 60-mph speed limit is inconsistent with the prior 70-mph rule. So too in Ellen Street Estates v Minister of Health, Housing Act 1925, section 46(1), which provided that compensation payable on compulsorily purchased land “shall be the value at the time the valuation is made of the land as a site cleared of buildings”, was inconsistent with Acquisition of Land (Assessment of Compensation) Act 1919, section 2(2), which provided that “the value of land shall […] be taken to be the amount which the land if sold in the open market by a willing seller might be expected to realise”.
What proposition did the Merchant Shipping Act 1988 stand for? That UK nationals be granted a special dispensation to fish in UK waters. Is there anything in the terms of the European Communities Act 1972 inconsistent with this? The Act says nothing about the freedoms of persons to fish or engage in any other commercial activity. Instead, it provides, by section 2(1), that “all such rights, powers, liabilities, obligations and restrictions […] created or arising […] under the Treaties […] shall be […] available in law”. Further, section 2(4) requires that “any enactment passed or to be passed […] shall be construed and have effect subject to the foregoing provisions of this section”. The first of these provisions requires that EU law be enforceable within the national order. The second ensures its priority over conflicting provisions of national law. Neither is inconsistent with the requirement that fishing vessels be 75 per cent. British owned because neither specifies a different answer to that question. In turn, there is nothing in the Merchant Shipping Act 1988 that speaks to the constitutional issue that animates the European Communities Act 1972. It regulates fishing, not the relationship between norms of discrete legal orders. The situation would have been different had the Merchant Shipping Act 1988 contained a provision either proclaiming its priority over conflicting EU law or excluding the application of EU law to it, in which case the court in Factortame (No. 2) would have been faced with a conflict between two statutes both of which spoke to the constitutional status of EU law.Footnote 48 As Adam Tomkins first recognised, in the absence of any such provision there is no conflict between these two statutes, only a conflict between the norms of EU law and the requirements of the Merchant Shipping Act 1988.Footnote 49
Another way to see this point is to ask what the situation would have been had EU law conferred no right against discrimination on grounds of nationality. In such circumstances there would be no live conflict of norms for the court to resolve. This would be so, moreover, despite there being no change to the terms of either the European Communities Act 1972 or Merchant Shipping Act 1988. This means that there are no provisions of the later Act “inconsistent with, or repugnant to, the provisions of [the] earlier Act [such] that the two [could not] stand together”.Footnote 50 The conflict in this case did not lie between these two statutes, but between the Merchant Shipping Act 1988 and the requirements of EU law. Hence the 1988 Act would not, according to ordinary principles, have impliedly repealed the 1972 Act.Footnote 51
To this it might be said that while these two statutes were not strictly speaking contradictory, the coexistence of both within the legal order gave rise to a conflict that the House of Lords did need to resolve: namely, between EU law, as given domestic purchase by European Communities Act 1972, section 2(1) and the registration regime under the Merchant Shipping Act 1988. And this is indeed true. But this was a conflict between norms of international law and a domestic statute, not between two domestic statutes. It is a conflict that thus falls outside of the scope of the rule of implied repeal.Footnote 52 By the terms of European Communities Act 1972, section 2(4), Parliament provided for what to do in such a situation.
If the above analysis is right, Factortame (No. 2) never disturbed the rule of implied repeal because that rule was not at issue in the case. This has the consequence that the case provides no support for the doctrine of constitutional statutes. Recall that on Laws L.J.’s analysis constitutional statutes are just those statutes that the law deems immune from implied repeal. Factortame (No. 2) was the one authority he cited for this.Footnote 53 But if the European Communities Act 1972 did not need to be protected from implied repeal to achieve the result the court reached, the case provides no evidence for the existence of the doctrine. What of R. (HS2 Action Alliance) v Secretary of State for Transport? The discussion of constitutional statutes was entirely obiter and did not, at any rate, implicate the doctrine of implied repeal. The question that the Supreme Court considered was about a potential conflict between provisions of EU law and Article 9 of the Bill of Rights 1689, not between two statutes created by the UK Parliament.
If neither Factortame (No. 2) nor R. (HS2 Action Alliance) v Secretary of State for Transport provides positive evidence for the existence of constitutional statutes, did Re Allister authoritatively reject the doctrine? It did not, for reasons that are structurally on all fours with Factortame (No. 2). For here any conflict that existed was between the terms of the Northern Irish Protocol and the Act of Union, not between that legislation and the European Union (Withdrawal) Act 2018. Only a conflict between the latter two would be sufficient to implicate the rule of implied repeal and so give grounds to authoritatively pronounce on the doctrine of constitutional statutes. Because no such conflict existed, no such issue presented itself. The question of whether the Act of Union 1800 was a constitutional statute was academic in Re Allister not because Parliament had, by European Union (Withdrawal) Act 2018, section 7A(3), provided for its resolution, but because it never arose in the first place.Footnote 54
If all of this is correct there is no positive evidence for the doctrine of constitutional statutes in UK law. Nonetheless it remains open whether the courts should embrace such a doctrine. If there exists conflict between two Acts of Parliament, the latter of which does not expressly repeal the former, should preference be given to the former when it has as its object “the legal relationship between citizen and State [or] fundamental constitutional rights”.Footnote 55 There are several issues. For one, such a doctrine would complicate the constitutional landscape. Rather than a single rule determining the relationship between conflicting legislative provisions we would operate with two: in the case of ordinary statutes the latter would supersede the former, but in the case of constitutional statutes the order of priority would be reversed. This would create pressing need to resolve whether legislation fell on one side of the divide or the other.Footnote 56 Given the inherent contestability and uncertainty of any such definition, subjects and officials would in many cases have to await the decisions of the courts about the status of particular legislative acts. Given such institutional costs we therefore need good normative reason to introduce such a doctrine. Two arguments have been made.
The first has to do with the importance of the rights and obligations contained in constitutional statutes. As Laws L.J. put the point: “The special status of constitutional statutes follows the special status of constitutional rights.”Footnote 57 The difficulty with this is that it fails to give reason to prioritise some earlier understanding of such rights over the later. For example, if Parliament were to now enact a new bill of rights without first repealing the Human Rights Act 1998, the doctrine of constitutional statutes would require us to enforce the 1998 Act. But if the Human Rights Act 1998 is especially important because of its concern with “constitutional rights”, then the very same would be true of the new legislation. This argument therefore gives us no reason to embrace the doctrine of constitutional statutes, which involves a preference for the former over the latter.
The second argument qualifies the first with a concern for consistency.Footnote 58 It is not just that constitutional norms are especially important, but that it is especially important that there is continuity in their application across time. Unlike Laws L.J.’s suggestion, this is an argument of the right form, for it would, if persuasive, give us reason to prefer the earlier legislation to the later. What to say of it? In the first place it is not obvious why constitutional continuity takes priority over continuity in other areas of the law. Why, in other words, should the courts be especially invested in the stability of human rights across time but not, say, the stability of employment rights? Second, concern for continuity cuts both ways, because, of course, one way to ensure continuity in constitutional issues would be to retain the doctrine of implied repeal, which has been part of the constitutional fabric of the UK for at least several centuries and plays an important role in the democratic renewal of law.Footnote 59 Finally, while it is true that there is often some value in ensuring that constitutional rules remain stable across time, it may be that such issues are more naturally handled as part of the political process,Footnote 60 rather than through judicial jerry-rigging of the procedure for their resolution.
V. Dicey, Again
The doctrine of constitutional statutes has both questionable normative and doctrinal foundations, the latter being so because Wade’s argument for the novelty of Factortame (No. 2) fails to hit its target. Whatever innovation that case involved it did not have to do with “the established rule about conflicting Acts of Parliament”.Footnote 61 This is not to deny, however, that something of constitutional importance happened in Factortame (No. 2). For, as Wade himself noted, the House of Lords both “forbid a minister from obeying” the Merchant Shipping Act 1988 and disapplied that statute on the instruction of an earlier Parliament. True, the courts have always had the capacity to disregard legislation that would otherwise have the effect of binding future Parliaments. Here, however, the court showed itself willing to disapply legislation created after the enactment of the European Communities Act 1972 and for the sake of some source of law other than parliamentary legislation, namely that of the EU.Footnote 62 Factortame (No. 2) certainly deserves its place on our constitutional law reading lists. The particulars of Wade’s analysis aside, we can see how the case fails to fit with Dicey’s picture of the constitution.
Recall that, according to Dicey, sovereignty has both a positive and negative limb. According to the positive limb Parliament has the capacity to make or unmake any law, subject to the proviso that it cannot bind its successors. According to the negative limb, no other body has the right to “override or set aside the legislation of Parliament”. But this, of course, is precisely what the court did in Factortame (No. 2). It set aside the provisions of the Merchant Shipping Act 1988 to the extent of their inconsistency with EU law. The result of this being that Spanish trawlers were permitted to fish in UK waters despite failing to satisfy the conditions provided for by an Act of Parliament. On the Diceyian understanding of sovereignty the court did something it was constitutionally debarred from doing, namely setting aside parliamentary legislation.
Does it matter for this argument that the court in Factortame (No. 2) was acting with the imprimatur of Parliament itself, as per the requirements of European Communities Act 1972, section 2(4)? On the Diceyian understanding it does not. If sovereignty implies that no other body has the right to set aside legislation, then it follows that Parliament cannot license them to do so. Just in the same way that Parliament cannot bind future iterations of itself, so it cannot give power to other bodies to override its own legislation. Hence, just as the court in Ellen Street Estates v Minister of Health refused to apply Acquisition of Land Act 1919, section 7(1) as an attempt to bind future Parliaments, then, by analogy, the court should have refused to apply European Communities Act 1972, section 2(4) as an injunction to disregard parliamentary legislation. The upshot of this analysis is that the court in Factortame (No. 2) set aside the wrong statute. Rather than disapplying the Merchant Shipping Act 1988, it ought to have refused to apply European Communities Act 1972, section 2(4) because it was inconsistent with the requirements of sovereignty. By invoking the Diceyian understanding of sovereignty we therefore arrive at Wade’s conclusion – that in Factortame (No. 2) “something drastic […] happened to [that] doctrine” – while avoiding the misattribution of this to the rule of implied repeal.
While this argument is valid, in the sense that its conclusions follow from its premises, I am going to argue that it is false, and false because the second aspect of Dicey’s definition of sovereignty contains a mistake. Consider the relationship between the two parts of his conception. The positive and negative limbs of Dicey’s definition are standardly thought of as complementary, with Parliament’s unlimited capacity to make law being shored up through prohibition on any other institution setting aside its legislation. Indeed, in most circumstances the two parts of the theory form natural counterparts. If Parliament passes a law concerning the fate of blue-eyed babies, no court or administrative body has the capacity to set aside, invalidate or refuse to apply it. The law that Parliament enacts is the law that must be honoured by these institutions.Footnote 63 Nonetheless it is possible for tension to arise between these two principles, for one thing that an all-powerful Parliament might try to do is to empower another institution to set aside or disapply its legislation. Were it to do this it would be acting under the auspices of the first principle – exercising its capacity to make any law – but in a way that threatens the second – by giving power to another to set aside its legislation. It is easy to conjure up hypothetical instances of such a phenomenon, for example a statute mandating that the courts invalidate legislation contrary to human rights. But we can also consider actual examples, such as the European Communities Act 1972, section 2(4) or European Union (Withdrawal) Act 2018, section 7A(3), both of which require the courts to disapply legislation if it conflicts with EU law. Such clauses are consistent with the positive first limb of Dicey’s definition, but they mandate behaviour inconsistent with the second.
How, then, should we resolve this conflict? Dicey did not directly address this question, most likely because of his belief that these were really two different ways of looking at the same thing.Footnote 64 Nonetheless, the belief implicit in the constitutional self-conception of many thinkers is that the second principle qualifies the first. Hence, the disapplication of the Merchant Shipping Act 1988 in Factortame (No. 2) constituted “a constitutional revolution” – a decision with far-reaching implications for the doctrine of sovereignty. Implicit in this is the belief that Parliament cannot license the abrogation of statute. Consider also the Human Rights Act 1998. Notably the Act avoids imposing on the courts any obligation to invalidate legislation that is contrary to human rights, instead requiring the courts, by section 3, to interpret legislation consistently with rights “so far as is possible”.Footnote 65 By section 4, the courts are entitled to issue a declaration of incompatibility as between rights and rights-offensive legislation, but this does not affect “the validity and continuing operation” of the relevant statute. Such constitutional innovations were widely understood to be necessary to preserve the doctrine of sovereignty.Footnote 66 The implication is that sovereignty stands in the way of Parliament empowering any other institution to set aside its legislation.
The problem with this aspect of our constitutional self-conception is that it is inconsistent with the reality of our constitutional practices. Central to this realisation are Henry VIII clauses. These are powers, granted by statute, that enable institutions other than Parliament to pass laws that amend, repeal or take the place of primary legislation. The first record of such a clause was contained in the Statute of Sewers 1531.Footnote 67 This gave the commissioner of sewers the power to impose taxes and penalties in a way that would supersede legislation. The Statute of Sewers was followed, in 1539, by the much more notorious Statute of Proclamations.Footnote 68 This empowered Henry VIII to issue decrees with the force of Acts of Parliament. Today such clauses are used to confer powers on ministers, the devolved institutions and on the House of Commons as a unitary legislative body.Footnote 69
Consider first Henry VIII powers conferred on the executive branch. Such clauses are features of many statutes including the Human Rights Act 1998, which provides, by section 10(2), that a minister may amend rights-incompatible legislation following a section 4 declaration of incompatibility. So too both the European Communities Act 1972 and European Union (Withdrawal) Act 2018 contain Henry VIII clauses, the first to remove conflict between UK and EU law, the second to facilitate the compatibility with the domestic legal order of retained EU law.Footnote 70 Henry VIII powers are also features of legislation on topics as diverse as electronic communications and regulatory reform.Footnote 71 Especially if drafted in broad terms and without adequate opportunity for parliamentary oversight, such clauses can give rise to serious problems for the separation of powers. But few think that their enactment or exercise amounts to “a constitutional revolution” or upends the doctrine of parliamentary sovereignty.Footnote 72 Why? Because enacted by Parliament, and subject to repeal by the same, they are an output of its sovereign power, rather than its subversion.
Consider next the devolution statutes. These enable the Scottish Parliament, Northern Irish Assembly and Welsh Assembly to amend, repeal or supersede legislation created by the Westminster Parliament unless it falls outside their area of competence.Footnote 73 Such powers certainly allow institutions other than Parliament to “override or set aside” parliamentary legislation. Nonetheless, devolution is premised on the preservation of sovereignty. In contradistinction to federal models of power sharing, which depend upon the idea of constitutional supremacy, devolution maintains the central role of Parliament. Rather than a system that grants central and regional governments co-equal authority under the constitution, devolution involves a revocable grant of power by the centre, which maintains its authority over the national units.
Finally, consider the operation of the Parliament Acts of 1911 and 1949. These create a procedure for the House of Commons to pass legislation without the support of the House of Lords provided certain conditions are met: that the bill in question is a money bill that has been delayed by more than one month,Footnote 74 or else is legislation that has been held up by the Lords for at least one parliamentary session.Footnote 75 Neither legislation designed to extend the lifetime of Parliament beyond five years can be passed by this method nor can private members’ bills, nor bills sent to the Lords less than a month before the end of the relevant parliamentary session.Footnote 76 Although there is controversy about the precise nature of the law created under these Acts, it is clear that it is both the creature of a non-sovereign lawmaker – as we have just noted, there are substantive limits on the powers of the House of Commons under the Parliament Acts of 1911 and 1949 – and yet, a lawmaker whose legislation is in all other respects on a par with that created by the House of Commons and House of Lords acting together, including in its capacity to override or revoke prior legislation.Footnote 77
Why does all of this matter? It matters because Dicey told us that sovereignty implies that no other body has the right to “set aside or override” legislation and yet Henry VIII powers involve precisely this. On the Diceyian understanding we are therefore faced with a dilemma. Either we must reject Henry VIII clauses as unconstitutional interferences with sovereignty or, accepting the reality of such clauses, we must hold that Parliament is not truly sovereign. The first of these options requires us to treat accepted aspects of the constitutional order – including the devolution statutes and the power of the Commons to pass legislation – as aberrant. On the second, we can accept such powers, but at the cost of sovereignty. Indeed, because Henry VIII clauses have been part of the UK’s constitutional practices since before parliamentary sovereignty is standardly understood to have taken root – the Statute of Sewers preceding the Glorious Revolution of 1688 by more than a century – we would be forced to hold that Parliament never truly became sovereign.
If we are disinclined to embrace either form of constitutional revisionism, what options are available to us? Is there any way to square Henry VIII powers with the sovereignty of Parliament? The answer, I hope, should be obvious. While sovereignty implies that no other institution has an independent right to set aside legislation, it may do so if empowered by Parliament itself. Hence, while it is not open to the executive to use its prerogative powers in a way that contradicts the requirement of any statute,Footnote 78 it may use Henry VIII powers to override the same. So too while the monarch may not assume any power inconsistent with legislation,Footnote 79 they may issue proclamations with the force of statute if granted such power by legislation.Footnote 80 In turn, the House of Commons and devolved institutions have no power to amend or override statute except as granted to them through legislation. For the House of Commons in 1911 to have declared a unilateral power to issue law with the force of statute would have been for it to have engaged in something approaching a revolutionary act. For it to exercise functionally equivalent powers under the Parliament Acts of 1911 and 1949 is for it to play its role as part of a continuous constitutional order.
Notice that by modifying the second of Dicey’s principles in this way we render its two parts consistent with one another. According to the first, Parliament has the capacity to make any law, including, by implication, law empowering other bodies to affect the operation of its own legislation. According to the second, no other body has the right of its own volition to set aside legislation. This modification of Dicey’s account not only allows us to square sovereignty with Henry VIII powers but it also renders consistent with sovereignty the operation of provisions such as European Communities Act 1972, section 2(4) and European Union (Withdrawal) Act 2018, section 7A(3), both of which require the courts to set aside legislation that is inconsistent with provisions of EU law. The implication is that Factortame (No. 2), while in some sense a constitutional novelty, is entirely consistent with the sovereignty of Parliament. Moreover, it is so for the very reason cited by Lord Bridge as justification for the decision, namely the courts’ duty “under the terms of the Act of 1972 […] to override any rule of national law found to be in conflict with any directly enforceable rule of Community law”.Footnote 81
Is there any reason for doubting that the evidence described above licenses a principle broad enough to implicate the powers of the courts in this way? For one, it could be argued, Henry VIII powers are legislative powers of amendment conferred on institutions other than courts, such that the argument fails to extend to the role of judges under the constitution. While it is true that Henry VIII powers are legislative in nature, and hence not the kinds of powers conferred on courts, the question is not about the nature of the relevant powers, but about the constitutional principles licensed by their existence and operation. The idea that sovereignty enables Parliament to confer power to affect the operation of legislation on all institutions except the courts looks ad hoc. There is, moreover, no evidence for such a restricted version of the principle and plenty for the unrestricted version, including in Factortame (No. 2) itself. So too while a power to amend, repeal or enact legislation is different from the courts’ capacity to deny effect to legislation or to set it aside, the question of constitutional principle is not about the precise means by virtue of which other institutions can override Parliament’s will, but about whether they are able to override that will consistent with sovereignty. My argument is that they are, provided Parliament has itself licensed the possibility.
Notice that this claim differs from the manner and form or “new” view of parliamentary sovereignty.Footnote 82 According to this school of thought while Parliament has no capacity to bind itself on matters of substance, it does have the ability to redefine the procedures or formal requirements for the creation of valid legislation, for example by stipulating that a supermajority must be met before an Act can be repealed or that a particular form of words must be used for the creation of valid law.Footnote 83 The claim being defended here is not that Parliament has the ability to redefine its own power to legislate, but that it has the capacity to empower other institutions to override its legislation.Footnote 84 This is the idea behind Henry VIII powers, the European Communities Act 1972 and its constitutional successor in the form of the European Union (Withdrawal) Act 2018. It is this outward-looking dimension of Parliament’s sovereignty that is missed by extant theories of the constitution.Footnote 85
Before moving to consider the implications of this understanding of sovereignty, I want to consider two objections to it. The first objection is that it cannot be right because the UK is widely understood to adopt a Diceyian conception of sovereignty and on Dicey’s view no other body has “a right to override or set aside the legislation of Parliament”. The central force of my argument is that, in one respect at least, the doctrine of sovereignty diverges from Dicey’s understanding of it. The constitution may be broadly Diceyian, but this does not mean it precisely matches the strictures of his theory. More generally, it is worth us remembering that Dicey was but one interpreter of the constitution among many.Footnote 86 Without meaning to deny the significance or influence of his work, it is important to remind ourselves that what we are trying to understand is not the thoughts of this man, but the constitution.Footnote 87
A second objection to the account is that it is inconsistent with an aspect of the theory of sovereignty which we do have good reason to accept. If, as I have suggested, Parliament has the capacity to confer on institutions other than itself the power to override legislation, then surely this would allow Parliament to bind its successors in a manner inconsistent with its continuing sovereignty. Certainly, this was Wade’s view of the impact of the European Communities Act 1972 in the aftermath of Factortame (No. 2). “The Parliament of 1972 had,” he suggested, “succeeded in binding the Parliament of 1988.”Footnote 88 In a similar way, one might consider Henry VIII clauses with prospective reach to be problematic fetters on the powers of future Parliaments.Footnote 89 Whatever constitutional issues these devices do raise, however, they do not involve the capacity of Parliament to bind its successors. It is always open to Parliament to expressly repeal or modify any such clause, as it did, for example, by repealing the European Communities Act 1972 following Brexit. Beyond this, Parliament can stipulate that legislation take effect “notwithstanding” such a clause, thereby protecting new legislation from the effect of any overriding or invalidating power.Footnote 90 Moreover, if my above argument concerning the doctrine of constitutional statutes is correct, such powers remain subject to implied repeal, even if they form part of “constitutional” as opposed to “ordinary” legislation.Footnote 91 Finally, were Parliament to attempt to protect such clauses from repeal, the courts should, following Ellen Street Estates v Minister of Health, ignore any such injunction, as inconsistent with the continuing sovereignty of Parliament.
VI. Implications
The above argument implies a radical divide in the architecture of the constitution, as between constitutionally legitimate interferences with statute licensed by Parliament itself and constitutionally illegitimate interferences by other institutions acting at their own behest. The analysis suggests that the constitution is, in one sense at least, political constitutionalist in shape, with Parliament as both ultimate legal authority and arbiter of controls on its own lawmaking acts.Footnote 92 However, in another sense the theory suggests that the constitution is more open to legal controls on parliamentary power than it is standardly understood to be. For example, while the Human Rights Act 1998 is premised on the idea that a power to strike down legislation would be inconsistent with sovereignty, the above argument suggests that it need not be. Provided Parliament itself confers such a power on the courts, it would be compatible with the existing constitutional order.Footnote 93 This is not to say, of course, that Parliament should grant such authority to judges. It may be that the present institutional arrangement – a duty to interpret legislation compatibly with rights and a power to issue a declaration of incompatibility – provides for the best balance of power as between Parliament and the courts in the protection of human rights.Footnote 94 What it is to say, however, is that the relative worth of different regimes for the protection of rights is to be adjudicated according to their costs and benefits, not forestalled by an overly restrictive understanding of sovereignty.
Consider next the devolution statutes. These permit the courts to invalidate legislation created by devolved institutions if it trespasses on matters reserved for the Westminster Parliament, but they impose no legal controls on Westminster itself. It retains the ability to legislate on any issue, including devolved matters, with or without the consent of the relevant devolved institutions. However, it is possible to imagine a more radical form of devolution, one which grants greater independence to the devolved institutions. At the most extreme end, it would be open to Parliament to impose on the courts a duty to strike down its legislation if trespassing on the powers of devolved institutions. Again, none of this is to say that such arrangements ought to be enacted. What it is to say, however, is that systems for the distribution of power considerably closer to a federal model than the current scheme of devolution are possible within the bounds of a constitution that preserves the sovereignty of Parliament.
These two examples show how the constitutional order is capable of accommodating much that legal constitutionalists prize – robust protection of rights, a legally policed distribution of powers between central and national or sub-national authorities – within a fundamentally political constitutionalist order. Such a model has the attraction of making legal controls on political power subject to democratic legitimation, through requirement that they be passed into law by Parliament. In contradistinction to traditional models of legal constitutionalism that are premised on the supremacy of law over politics we can think of this as a model of parliamentary constitutionalism. Constitutionalism because it means that law may be used to limit the effect of political decision-making. Parliamentary because this happens at the behest of political institutions themselves. In a well-functioning political system, it would prevent the dead hand of the past from becoming a stranglehold on the constitutional design of the present.
The other side of the coin is that such controls are subject to political whim and, legally speaking, as open to being revisited as any other aspect of the law. Indeed, the Human Rights Act 1998 came close to finding its fate in 2022 at the hands of the then Government’s deftly named Bill of Rights Bill. This vulnerability is one cost of opening up constitutional structures to ordinary political processes. Conversely, however, politics is also capable of shoring up that which is, legally speaking, no more permanent than any other statute. It is, for example, very hard to see how the powers of the Scottish Parliament could be repealed or unilaterally diminished by Westminster, without presaging the breakup of the UK. If legal constitutional controls are enacted through the medium of the political constitution this means that their relative durability depends on the headwinds of politics, which blow at different speeds and in different directions, in different places and at different times.
A final aspect of the picture that deserves emphasis is the way in which it circumscribes the role of the courts as guardians of the constitutional order. While it is doubtlessly true that many basic rules of the constitution, including the norms that structure the doctrine of sovereignty, are such because grounded in the practices of the courts,Footnote 95 these rules assign central responsibility to Parliament as author of the controls on its own lawmaking powers. This exposes the fatal flaw at the heart of Lord Steyn’s inference in Jackson v Attorney General, from the constraints imposed by the European Communities Act 1972 and Human Rights Act 1998 to the curtailment of Parliament’s power at the hands of the courts.
To this it will no doubt be said that the above involves an idealisation that paints large swathes of the constitution out of existence. No mention of the principle of legality whereby the courts require Parliament to speak in explicit terms if it is to be taken to have licensed the executive to violate fundamental rights.Footnote 96 No mention of the courts’ approach to the interpretation of ouster clauses, reading them down to enable effective judicial review of administrative action.Footnote 97 Do these doctrines not suggest a picture of the constitution in which the courts have some independent role in constraining the authority of Parliament?Footnote 98 There are several things to say about this.Footnote 99 For one, such practices have never involved the courts explicitly according themselves power to override statute, premised as they are, albeit implausibly, on the retrieval of Parliament’s true intention.Footnote 100 Second, the point of the courts’ actions in such cases is really to control the decision-making of bodies other than Parliament: to prevent the executive from acting contrary to common law rights or to hold administrative bodies subject to judicial review. The strained interpretation of legislation is merely the means through which this end is achieved. Third, there is very little consistency in the courts approach to ouster clauses. For every decision in which they read such a clause out of existence we can find another in which they give effect to its terms.Footnote 101 Finally, the interpretation of statutes by way of presumptions that modify what would otherwise be the effect of target legislation are a mainstay of all municipal legal orders, including orders that do and do not subscribe to the doctrine of sovereignty.Footnote 102 In many cases they are also a legitimate part of the judicial power, provided they protect important interests and are capable of being overridden by legislative institutions when their intention is both clear and targeted at the issue subject to dispute.Footnote 103 If judicial constraint on democratic political power is to move beyond these bounds it must find some external source of legitimation. I take this to be a true proposition of political morality. It is also, as things stand, a truth about the constitutional order of the UK.Footnote 104
Open access
