The UK’s territorial constitution is, at present, under a great deal of pressure. Those familiar with one force unsettling the devolution framework — the attempts to override the Northern Ireland Protocol — will no doubt recall the legislation that first countenanced a similar approach: the UK Internal Market Act 2020 (UKIMA). This piece of legislation is, however, once again causing its own stir, this time in the form of a clash between Scottish and UK ministers over gene-editing regulations.
The UKIMA, in short, effectively deprives future provisions of devolved legislation—even within competence—of effect to the extent that they are incompatible with its “market access principles”. In other words, although such devolved legislation would remain law, its practical effect is limited by the application of the market access principles. The UKIMA, we suggest, contributes not only to a growing momentum towards a minimalist conception of devolution, but does so by fundamentally challenging the way competence has usually been understood in the UK’s territorial constitution. We contend that the UKIMA, complemented by recent Supreme Court jurisprudence, is arguably abandoning the dynamism that defined the orthodox approach to devolution in favour of a more restrictive model of competence, opening a gulf between how that concept is understood in the devolved and Westminster contexts.
Devolution: A Generous and Dynamic Settlement
Tasked with reviewing proposals for devolution within Great Britain in the 1970s (devolution in Northern Ireland being outside its scope), the Kilbrandon Commission, by a majority, called for devolved competences to be as generous as possible, with legislative or policy intervention from London appropriate only when all other options had been exhausted. It did this in light of Westminster’s earlier experiences of granting legislative autonomy which began over a century earlier, to territories such as Canada, Australia, Ireland and India.
Initially, the earlier attempts were restrictive, both in the text of the statutes conferring autonomy, and in judicial approaches to their interpretation. Over time, however, as disputes arose between colonial officials (and subjects) seeking greater autonomy and the interests of the metropole, grants of legislative autonomy moved into a more generous phase, culminating in significant moments such as the Statute of Westminster 1931, which considerably (but not totally) unshackled the Dominions from London’s influence. Although London’s control would remain tight in certain areas, the interests of workability led generally towards a reduction in its oversight, replacing it with a model wherein, within their fields of competence, these precursors to the devolved legislatures became far more autonomous.
This trajectory was possible despite provisions almost identical to section 28(7) of the Scotland Act 1998 (for example section 75 of the Government of Ireland Act 1920 and section 110(a) of the Government of India Act 1935). Such “sovereignty clauses” never animated the Privy Council or the House of Lords in the way they have the Supreme Court in the Continuity Bill and Treaty Incorporation References cases where, by contrast, the Court identified a further limit on devolved competence in the protection of Westminster’s “unqualified power”.
As a result of this generous approach to competence, where there was a conflict between the interests of London and its imperial possessions, it would usually be resolved politically – for example through the reservation and denial of royal assent to Bills passed by autonomous legislatures.
The development of the modern devolution settlements for Scotland, Wales and Northern Ireland, as well as the majority of the jurisprudence concerned with them, tended to reflect this maximalist approach to devolved competence, which was also endorsed by the Kilbrandon Commission. This is so both out of principle, intended perhaps as a signal of the constitutional significance of those institutions, and because experience from Empire demonstrated that more generous grants were often more pragmatic and workable, if not inevitable.
The resultant framework is one which is dynamic and powerful. Each of the devolution statutes is set out in generous terms, and each is primarily concerned with providing exhaustive negative lists of subject-matters and protected legislation over which the legislatures do not have jurisdiction; every other matter is devolved (though the Northern Ireland Act 1998 contains an additional – and for our purposes irrelevant – list of powers that can be devolved over time). Equally, devolved legislation is potent: within competence, the devolved legislatures may modify and even repeal Acts of the Westminster Parliament. The Northern Ireland Act 1998 expressly provides that “an Act of the Assembly may modify any provision made by or under an Act of Parliament in so far as it is part of the law of Northern Ireland” and the other devolution statutes implicitly acknowledge the same.
This statutory framework recognises and facilitates the evolution of diversity within the UK’s constitution, and devolved legislatures have, in various capacities, enthusiastically modified Acts of the Westminster Parliament which fall within their respective competences. Significant examples include modifications to the Welsh and Scottish local government franchises by, respectively, the Senedd Cymru and the Scottish Parliament as well as the recent (though unresolved) amendment to the functioning of the Northern Ireland Executive Committee by the Northern Ireland Assembly.
Devolution, then, is a system where both Westminster and a relevant devolved legislature can modify and repeal the enactments of the other. Acts of the devolved legislatures within competence have not, therefore, needed to be compatible with the body of Westminster’s legislation on the same issue; within the bounds of competence provided by subject matter or protected enactment restrictions, a devolved legislature can remake, or unmake, an entire scheme provided by Westminster. However, the UKIMA, and recent trends in the Supreme Court, represent a change of course.
The UKIMA, by contrast, presents a far less dynamic view of devolution. Rather than allowing the devolved legislatures to legislate freely within their competences, devolved enactments will only lead to effective law to the extent that they are compatible with the framework provided by Westminster in the UKIMA. The devolved institutions do have the opportunity to negotiate common frameworks which can qualify the impact of the UKIMA in this regard, however.
Devolved legislation has always been confined by other legal norms (for example, variously EU law, retained EU law and the ECHR), but these norms always provided a “hard” competence limitation: devolved legislation is not law so far as incompatible with them. The UKIMA, by contrast, deprives legislation incompatible with its market access principles of much of its effect, a more uncertain model which challenges the received understanding of competence. Concerns raised by this approach are not hard to find: for example, in giving evidence to the House of Lords Committee on the Constitution, Counsel General for Wales Mick Antoniw MS set out the subordinative implications of this legislation from the perspective of devolved law-making.
Yet, the UK Government’s response, identifiable in its approach to the legal challenge from the Welsh Government, is straightforward: depriving devolved legislation of effect does not interfere with devolved competence per se. In the view of the UK Government—though not yet of any court—and from the text of the UKIMA, a devolved competence is still a competence, even when it is merely a power to make laws that are deprived of much of their practical effect by the UKIMA (to the extent that the market access principles provide).
The reason, however, that this is such a surprising contention is because—if it is right—this claim represents a step backwards for devolution, it being clear from the Supreme Court’s recent jurisprudence that the deprivation of the legal effect of Westminster’s legislation is repugnant to its competence. In the Continuity Bill case, the Supreme Court said this:
An enactment of the Scottish Parliament which prevented such subordinate legislation from having legal effect, unless the Scottish Ministers gave their consent, would render the effect of laws made by the UK Parliament conditional on the consent of the Scottish Ministers. It would therefore limit the power of the UK Parliament to make laws for Scotland, since Parliament cannot meaningfully be said to “make laws” if the laws which it makes are of no effect.
If the Government’s contention in the context of the UKIMA is right, it would bifurcate the meaning of competence: one rule for Westminster and quite another for the devolved legislatures. Competence would be conceived differently depending on which institution is concerned, widening the normative gap between the devolved and central legislatures.
The distinction between how competence is understood at Westminster and how it is understood in the devolved territories might be because, unlike Westminster, the devolved institutions do not possess the “unqualified legislative power” identified by the Supreme Court. If this is so, it demonstrates how the Court’s recent protectionism of Westminster’s legislative freedom, and its failure to accord that same protection to the devolved legislatures, buttresses the retreat from a dynamic model towards a more pre-emptive, subordinative model of competence.
This retreat from dynamism is also identifiable elsewhere. In its judgment in the Treaty Incorporation References case, the Supreme Court notably downplayed the delicate and complex interplay between devolved and central legislation. The Supreme Court said that if the Westminster Parliament were, by virtue of a Scottish Bill, “confined to enacting legislation for Scotland with a limited life expectancy… That would plainly be a modification of Parliament’s ‘unqualified legislative power’” and outwith the Scottish Parliament’s competence. In other words, any Bill that “would qualify Parliament’s power to allow existing legislation to remain in force unamended” is outwith competence. Yet, on an orthodox account, this is precisely what devolution provides: within devolved competence, Westminster legislation, unless a protected enactment, is always vulnerable to amendment or repeal by the relevant devolved legislature. If Westminster intends to insulate a particular provision or statute against this, the relevant mechanism for it to do so is to protect that enactment.
The Problems with this Approach
A key part of the retreat towards a minimal view of devolution is, in our view, this reconfiguration of competence and the new distinction between its meaning from one legislature to another. We suggest that this retreat is both problematic and unnecessary. The imperial experience at least suggests it is impracticable, and the Welsh Government’s legal challenge demonstrates it is arguably unprincipled. But there are deeper pragmatic and conceptual problems, too.
The UKIMA, and the less dynamic vision of devolution that accompanies it, means that the devolved legislatures can only legislate—even within competence—so far as permitted by Westminster’s legislative frameworks. This creates practical problems for the policy laboratory often considered a key benefit of devolution and encourages a race to the bottom on market standards. It also creates an uncertain framework, antithetical to devolution’s oft-observed pursuit of “workability”, by creating exactly the kind of disconnect between the text of devolved legislation and the reality of the law that the Supreme Court recently found so intolerable.
On the conceptual side, this approach highlights the problems with the under-theorisation of competence in the UK’s constitution. That the Welsh Government’s legal challenge was found to be premature, despite the issues we have identified, is indicative of how ill-equipped the UK’s institutions might be for the questions that arise in the regulation of its own internal market. Doctrines like field-occupation, pre-emption, and supremacy, which have all been at best footnotes in contemporary domestic territorial thinking, are making their way into constitutional reality by the back-door of the UKIMA and the Supreme Court’s jurisprudence. Yet, whether obscured by sovereignty, exceptionalism or something else, we seem to largely lack the language – or perhaps even the conceptual apparatus – to express these ideas clearly in the context of devolution, where primary legislative parity and sovereignty have been thought to suffice. The under-theorisation of competence in the UK reveals the problems with viewing devolution solely through the lens of enacted text in a constitution famous (or, perhaps infamous) for being uncodified. The earlier periods of legislative autonomy demonstrate that this problem is not a new one, but it is one that might need resolution more urgently than ever.
Having left the EU, Westminster may have “taken back control”, but the devolved legislatures have not made similar gains and the UK’s territorial constitution is under considerable strain. It is important that the nature of this challenge is not concealed behind a smoke-screen provided by the under-theorisation of competence, or a failure to draw lessons from constitutional history.
Many thanks are owed to Mark Elliott, Alison Young, Mike Gordon, Louise Mallinder and Conor McCormick for their helpful comments on an earlier draft. Any errors or omissions are the authors’ own.
The authors will be presenting a more detailed analysis of the issues explored in this blogpost at a forthcoming conference at Swansea University on Monday 11th July 2022. Details of how to attend are available at this link: “Undoing Devolution by the Back Door? The Implications of the United Kingdom Internal Market Act 2020”.
Anurag Deb is a PhD researcher at Queen’s University Belfast.
Nicholas Kilford, PhD Candidate at the University of Cambridge
(Suggested citation: A. Deb and N. Kilford, ‘The UK Internal Market Act: Devolution Minimalism and the Competence Smoke-Screen’, U.K. Const. L. Blog (4th July 2022) (available at https://ukconstitutionallaw.org/))