06 May 2024

Unconstitutionality à l’Anglaise

Is the UK having its Marbury v Madison moment?

After long and tortuous proceedings in Parliament, the Safety of Rwanda (Asylum and Immigration) Act 2024 finally received Royal Assent on Thursday 25 April. The bill had already grabbed attention before it was officially introduced, and it continued to excite constitutional lawyers at every stage of its enactment. There are so many problems with the Act and they are so fundamental that there has even been speculation that the courts might refuse to apply some of the Act’s provisions. This has never happened in the UK, though judges have dropped hints that it might if Parliament ever does the ‘unthinkable’. It looks like it just did.

Still, it is difficult to tell if the UK is really having its Marbury v Madison moment. It is clear enough that even if the courts are asked to disapply the Act’s provisions or strike them down, they are likely to be very slow to take such radical steps. The political context, if anything, encourages caution: the upcoming general election looms large, as does the likely winners’ commitment to backing out of the Rwanda affair. But it is just as clear that a judicial intervention is no longer seen as completely impossible. By British standards, this is already an extreme situation.

I will say no more about whether the courts will intervene and focus instead on how the courts might intervene. Specifically, I will argue that quite apart from the ‘hard-line’ approach of striking down or disapplying the statute in whole or in part, the courts also have a ‘soft-line’ option of declaring its unconstitutionality without denying its status as binding law. This may sound strange to many readers. In the first part of this post, I will explain how such an intervention would fit into the constitutional tradition of the UK. In the latter part, I will say a few words about what may make it attractive.

Unconstitutional but Legally Valid

In most constitutional systems, the notion that a piece of legislation may be ‘unconstitutional’ but remain legally valid is paradoxical at best. Not so in the British (or, if you prefer, English) tradition, defined as it is by the Diceyan notion of parliamentary sovereignty. Dicey wrote that ‘the expression “unconstitutional” … as applied to an English Act of Parliament’ simply ‘cannot mean that the Act is either a breach of law or is void’. Instead, he insisted, it merely indicates that ‘in the opinion of the speaker’, the legislation is ‘opposed to the spirit of the English constitution’. In short, it is a political epithet. It is not a legal term to be adjudicated on by a court of law.

This view seems fairly established as far as British courts go. In Madzimbamuto v Lardner-Burke (itself a Privy Council case from Southern Rhodesia), Lord Reid famously commented:

It is often said that it would be unconstitutional for the United Kingdom Parliament to do certain things, meaning that the moral, political and other reasons against doing them are so strong that most people would regard it as highly improper if Parliament did these things. But that does not mean that it is beyond the power of Parliament to do such things. If Parliament chose to do any of them the courts could not hold the Act of Parliament invalid.

More recently, in the first Miller case, the UK Supreme Court reaffirmed that it is not the role of the courts to articulate any limits on Parliament’s legislative powers as might be grounded in political conventions, even when breaching the conventions in question would attract the label of unconstitutionality. In so doing, it refused to follow in the footsteps of the Canadian Supreme Court, which—in the 1981 Patriation Reference—made the unprecedented step of determining certain actions of the Canadian federal government as lawful, but at the same time ‘unconstitutional in the conventional sense’.

My argument is that UK courts should be able to make a broadly analogous declaration in relation to the Safety of Rwanda Act 2024: that, although legally valid and thus binding (because Parliament is sovereign), it is unconstitutional. It may at this stage appear to fly in the face of the authority, but it does not. I will mention four reasons why.

Declarations of Unconstitutionality

First, the problems with the 2024 Act can be distinguished from the challenges raised in Madzimbamuto or Miller. In both Madzimbamuto and Miller, the argument concerned constitutional conventions prescribing that a regional (Miller) or colonial (Madzimbamuto) legislature consent to a proposed Act of Westminster Parliament before it is enacted. The constitutional standards at play were therefore conventional in nature and regulated relationships between political institutions. Now, the Safety of Rwanda Act 2024 is unconstitutional first and foremost because it enacts a legal fiction that Rwanda is a safe country, clearly aimed at displacing the judicial finding that it is not, and because it makes this fiction impossible to challenge in court, even against the criteria it accepts as definitive of what a safe country is. The Act does not infringe ‘political conventions’, therefore, but rather fundamental constitutional rights and principles which define the position of the judiciary as well as that of individual asylum seekers vis-à-vis Parliament and the government that enjoys its confidence. They have a different source and nature than the conventions in Madzimbamuto and Miller: they are grounded, rather than in past political practice, in normative considerations central to the principle of the rule of law. And the constitutional relationships they regulate are different, too. None of this means that they are necessarily more important, but it does suggest that they are more apt for some form of judicial enforcement.

Secondly, there is nothing in Madzimbamuto or Miller that would preclude a court from declaring an Act of Parliament unconstitutional, so long as such a declaration is understood to have no effect on the Act’s validity. Indeed, it would be merely to recognise that what—in Lord Reid’s words—is already ‘often said’ in relation to an Act of Parliament can also be declared by a court of law. While it is true that in Miller the Supreme Court declined jurisdiction to ‘give legal rulings on [the] operation or scope’ of a ‘political convention’, I have already argued that a declaration of the 2024 Act’s unconstitutionality would not amount to such a ruling. Instead, it would amount to a non-binding declaration of the incompatibility of a piece of parliamentary legislation with fundamental constitutional rights and principles. And, as my third and fourth reasons show, British judges are familiar with making declarations of this sort as well as with this subject matter.

The third reason is then that British courts are already accustomed to issuing non-binding declarations of the incompatibility between an Act of Parliament and fundamental constitutional standards. They have been doing so for more than two decades under section 4 of the Human Rights Act 1998, which gives senior UK courts the power to declare an Act of Parliament incompatible with Convention rights. The sort of jurisdiction I am arguing for here would of course be different, since it would have to be grounded in the common law, not statute. Nonetheless, while it would be novel to the common law, it would not be so novel to British courts. They have exercised a similar statutory jurisdiction for some time already and without causing particular constitutional trouble.

Finally, UK courts already adjudicate on the scope of fundamental common law rights and principles and on their interaction with parliamentary legislation. Under the principle of legality, courts apply a presumption that Acts of Parliament are to be interpreted and applied in accordance with basic constitutional rights and principles. In effect, Parliament may only abrogate such basic rights and principles by using clear and unambiguous words. If the jurisdiction I am arguing for here corresponds to section 4 of the Human Rights Act, the principle of legality has its analogue in section 3—though the precise extent of the analogy is a matter of some controversy. The practical point is that an interpretive presumption of the sort envisaged under the principle of legality or section 3 is more intrusive than a declaration of incompatibility because it actually affects the way in which statutory words are applied in concrete cases. If common law principles justify the interpretive presumption involved in the principle of legality, they can also sustain a jurisdiction to declare, in a non-binding manner, the unconstitutionality of an Act of Parliament that violates basic constitutional rights and principles.

These four arguments would need to be developed at more length. Nonetheless, they suggest that a judicial declaration of the Act’s unconstitutionality would not be as odd a development as it might first appear.

A Middle Way

Issuing a non-binding declaration of unconstitutionality could also present some pragmatic advantages in a situation such as the present.

Let us imagine that the Supreme Court is seriously faced with a choice whether to disapply the unconstitutional provisions of the 2024 Act or not. Politically speaking, this is an unenviable choice. If the Court caves in, it will have to acquiesce in a fairly explicit attack on the judiciary’s capacity to provide an effective and independent check on the executive. Conversely, if the Court decides to disapply the legislation, it risks enormous political backlash—amplified, if anything, in the tense atmosphere of pre-election politics—and the possibility that the government will defy the Court’s determination anyway. The very passing of the Safety of Rwanda Act to reverse the AAA ruling was already a stark gesture of defiance. If the government was ready to take that step, then it might imaginably choose to ignore the court’s decision to disapply the legislation, too. It can then easily hold itself out as defending traditional parliamentary sovereignty—or even the sovereignty of the UK as such—from judicial and international interference. For the Conservative Party of 2024, such a move would likely present only negligible political costs, and perhaps even a net political gain.

Against this background, a declaration of unconstitutionality presents itself as a convenient middle way. It allows the Court to lay bare the constitutional problems with the legislation and bring them to the attention of both Parliament and the public. At the same time, because it is premised on a recognition of Parliament’s sovereignty and of the limited institutional competences of the judiciary, it largely defuses the risk of a constitutional crisis and is much harder for a defiant government to dismiss as a blunt judicial power grab. It is perhaps as safe an option as caving in, and certainly a safer option than the ‘hard-line’ approach of striking down or disapplying the statute while being equally (in)effective. A government acting on a good faith commitment to institutional comity would no doubt do its best to act within the bounds of what the courts regarded as its constitutional obligations. This is what happened in the Canadian case, and this is what usually happens in response to section 4 declarations. A government lacking such a commitment, by contrast, may be just as ready to disregard a ‘binding’ judicial disapplication as a ‘non-binding’ judicial declaration of unconstitutionality.

Conclusion

It remains unclear if the courts will even face these choices, let alone which option they will eventually pick. But if this really is a Marbury v Madison moment for the UK, then it will come as no surprise if it is met with a very British response. It remains to be seen what that response will be and what it will mean for UK constitutional law in the long run. For better or worse, the story is far from over.