In her Mansion House speech in March, in which she outlined her plans for leaving the EU, Theresa May stated that she ‘would not allow anything that would damage the integrity of our precious Union’. We have heard much about Brexit and ‘the will of the people’. Yet the people who voted in the 2016 referendum did not speak with one voice. They were divided by age and geography and, not least, by which of the four nations of the UK they belonged to: 62 per cent of those voting in Scotland, and 56 per cent in Northern Ireland, voted to remain in the EU, while 52.5 per cent of voters in Wales backed Leave, as did 53 per cent in England.
The basic problem for Scotland and Wales has arisen in the following way (I am of necessity simplifying a complex state of affairs). The Scottish and Welsh governments asserted that the European Union (Withdrawal) Bill (EUWB) currently making its way through the Westminster Parliament – an arcane and convoluted piece of legislation – is incompatible with the devolution settlement and that as a result they would refuse to give legislative consent to it. The Sewel Convention holds that the devolved legislatures must normally give their consent to any Westminster legislation touching on their powers before it can be adopted. The EUWB is supposed to pave the way for Brexit by transferring most EU laws into UK law; without it the UK would not have a functioning legal system on Brexit day.
The withdrawal bill will enable powers currently controlled by Brussels to be repatriated to Britain. However, under Clause 11 of the bill, even powers in devolved areas will return to Whitehall first, and UK ministers will then decide if they should return to devolved legislatures or become part of UK common frameworks. The devolved governments see this as a power grab. Until very recently, the UK government was resisting amending the clause (indeed it reneged on an undertaking to do so earlier this year). As a result, the Scottish and Welsh governments both published continuity bills intended to preserve their powers from appropriation by Westminster, and to prepare them for Brexit. Unless changes to the EUWB are agreed, in other words, devolved consent will be withheld, these continuity bills will be adopted, and UK law will not be ready for Brexit day – something approaching a constitutional crisis.
Late in April, Theresa May’s government did two things. First, it launched legal proceedings challenging the legality of these continuity bills. Second, it signalled its readiness to amend the withdrawal bill in a way that might be acceptable to the devolved authorities, by providing that most devolved powers currently exercised by the EU would pass to devolved institutions. However, critically, the UK government’s changed position also enables UK ministers to make regulations, freezing the devolved governments’ ability to legislate in specified areas. In these areas, the UK overall would retain control until agreement is reached about what to put in place of EU law. The Welsh government agreed proposals (‘capitulation’, according to Plaid Cymru) with its UK counterpart. The Scottish government rejected the UK government’s plan because it still allowed the Scottish Parliament’s powers to be restricted without its consent: the new amendment requires only that UK ministers ‘consult’ with devolved governments – which means the UK government could impose restrictions on them unilaterally. So the positions of the previously closely aligned devolved governments have diverged (a case of divide and rule?) and, for Scotland, the battle continues. Meanwhile, the UK seems no closer to settling the fiendishly difficult problems Brexit raises in Northern Ireland (where devolution is currently suspended, excluding the possibility of a continuity bill there).
Constitutions are meant to set out basic rules for a society: how it is to be governed; whether government will take a presidential or parliamentary form; whether the state is to be federal or centralised; whether or not human rights will be protected by judicial review, and so on. A key role of a written constitution is to prevent politicians and the public from acting on impulse by entrenching especially important provisions. Entrenchment isn’t possible in the UK, since it is one of the very few countries not to have codified its constitution (historically, it has prided itself on this arrangement, which it sees as having the merit of flexibility). But, as the late Lord Bingham said, the absence of codification means that, ‘constitutionally speaking, we now find ourselves in a trackless desert without map or compass.’
Determining what our constitution requires, what fundamental rules it sets out, is as difficult and divisive as Brexit itself. It may be misleading to claim, as some have, that the British constitution gives rise to a ‘Rashomon effect’ (after Kurosawa’s film, in which a murder is described in mutually contradictory ways by those who witnessed it), but still I would argue that there is no single way of understanding it. May’s government (in common with past UK governments, including Thatcher’s) sees the UK as a unitary, centralised state with an omnicompetent Parliament (though it also tried to subjugate Parliament to the executive in the Article 50 process – a matter that had to be sorted out by the Supreme Court in the Miller case). Since the EU referendum vote, although formally non-binding, was carried out on a UK-wide basis, with no protection for different results in the devolved nations, Scotland and Northern Ireland’s vote to remain in the EU is treated as being of no constitutional relevance. Although the devolution statutes (such as the Scotland Act of 1998) established which powers were to be devolved, Westminster, thanks to the doctrine of parliamentary sovereignty, retains its right to legislate on any matter, including those it gave over to Holyrood. Mechanisms such as the Sewel Convention or the Joint Ministerial Committee are intended to control intragovernmental discord, but they aren’t legally enforceable or binding. Foreign affairs, including EU membership, are reserved to the UK government and devolved authorities may at most be consulted as a matter of courtesy. Neither EU law nor UK law guarantees the devolved nations a role in the EU withdrawal negotiations. The UK government approaches treaty-making as matter of the prerogative. (Whether it accords the Westminster Parliament a formal and meaningful vote on the final exit deal – something more than take it or leave it – remains to be seen.)
However, there is another way of understanding the constitution, according to which the UK is a union state rather than a unitary state, founded on treaties (the Treaty of Union of 1707, the Good Friday Agreement of 1998) and reliant on continuing consent as well as constitutional practice. It recognises that the UK has been transformed by external developments such as EU and ECHR membership, and recalibrated internally by devolution, as well as by the Human Rights Act and a desire for more principled constitutional development than parliamentary sovereignty allows, i.e. a situation where the constitution is not constantly changing on an ad hoc basis. The constitution has also become much more legalised and susceptible to judicial review, especially in the murky area of executive prerogative, as demonstrated in the Miller case. In addition, the Scotland Act 2016 and Wales Act 2017 (both pieces of Westminster legislation) proclaim the permanence of the Scottish Parliament and the Welsh Assembly, flying in the face of orthodox constitutional law’s assertion of an omnicompetent Westminster Parliament.
Of course, Brexit has again and again been justified in terms of UK parliamentary sovereignty – as Britain ‘taking back control’ and taking charge of its own laws. Yet the classic Diceyan vision of parliamentary sovereignty was under challenge in the UK well before Brexit. In the 2005 case of R (Jackson) v. Attorney General, which concerned a challenge to the hunting ban and was heard in the House of Lords, then the UK’s highest court, Lord Hope stated that ‘parliamentary sovereignty is no longer, if it ever was, absolute … Step by step, gradually but surely, the English principle of the absolute legislative sovereignty of Parliament … is being qualified.’ Diceyan orthodoxy never held as much weight north of the border. In the 1953 case of MacCormick v. Lord Advocate, decided by the Court of Session in Edinburgh, the senior Scottish judge, Lord Cooper (a former Unionist Party politician and an eminent legal historian), stated that the ‘principle of the unlimited sovereignty of Parliament is a distinctively English principle which has no counterpart in Scottish constitutional law.’ Many Scots claim a tradition of popular sovereignty dating back to the Declaration of Arbroath in 1320, which held that there was a contract between the king and the people. As it was the Union of Scotland and England in 1707 (a consensual union brought about by treaty, not conquest) that brought Great Britain and its Parliament into being, perhaps we should recognise this historical event as a foundation of the British constitution, rather than rely on the Diceyan totem of parliamentary sovereignty. The historian Conrad Russell held that ‘a residual Scottish sovereignty’ derived from the Treaty of Union. The British constitution should therefore be seen as distinct from the English constitution, and Scottish constitutional principles (as well as Northern Irish and Welsh ones) should be acknowledged as playing their part in this multiple constitutional order. In which case, is it not a sign of constitutional confusion if the perceived popular sovereignty of the 2016 advisory EU referendum is used to justify a Leave vote in England, where popular sovereignty is not part of the constitutional tradition, but ignored for Scotland, where it is?
The decentralised Northern Ireland brought into being by the Good Friday Agreement, which has complex provisions regarding cross-community consent, self-determination and a role for the Republic of Ireland and the EU, is far from Dicey’s model. In Northern Ireland a painstaking, written constitutional order was introduced as part of the peace settlement, bringing into existence a consociational, pluralist system designed to ensure that major change happens only with broad support in both parts of the community. Brexit will severely challenge that model.
Indeed, the Northern Ireland issue seems almost impossibly difficult. Although Northern Ireland is currently subject to direct rule by a UK government dependent on DUP support, the prospect of an external border between the EU (Ireland) and non-EU territory (Northern Ireland) creates headaches for all concerned. This problem is entirely of the UK government’s making: if, as May insists, Brexit means the UK leaving the customs union and single market, there will have to be a customs and regulatory border between the Republic and the North. (May is also ignoring the Scottish government’s arguments – which she pledged to consider – in favour of single market and customs union membership for the UK, or Scotland at least.) But the UK is maintaining that there will be no hard border, no physical infrastructure, checks or controls. This would be possible only if Northern Ireland remained in the customs union and single market. Yet the UK government has repeatedly promised the DUP that there will be no barrier between Northern Ireland and the rest of the UK. These outcomes are incompatible. Some have described this as a modern version of the Schleswig-Holstein question, which Lord Palmerston claimed was understood by just two people aside from himself, one of whom was dead, the other gone mad. The question of whether Schleswig-Holstein was a part of Denmark or Germany was finally settled by referendum; the Good Friday Agreement allows for a border poll, a referendum allowing the people of Northern Ireland to decide whether they wish to become part of a united Ireland.
It is possible, however, that this mess can be avoided. In December, the UK accepted, as a ‘backstop’ position, full alignment with those EU internal market and customs union rules that support North-South co-operation and the Good Friday Agreement. This would mean that Northern Ireland in effect stays in the EU. That will sit comfortably with the government’s assurances to the DUP only if the rest of the UK stays in the customs union and single market too (or if the people of Northern Ireland exercised their referendum right and opted for a united Ireland). It’s possible that the UK government will move towards customs union membership (the cabinet is split on what to do about the border), but the bigger question is whether Brexiters’ ideological (and deluded) ambitions for absolute sovereignty in trade justify threatening the peace in Northern Ireland.
Can Brexit be achieved in a manner that satisfies all four nations of the Union? Although it is not always acknowledged, EU membership provided an external support system for devolution. It insulated the devolved nations from a unitary UK-wide approach by, for example, distributing EU funds – agricultural subsidies, money for new piers in the Hebrides – more to some parts of the UK than others, something unlikely to continue after Brexit, when a unified approach is likely to dominate however unfavourable to devolved interests. A unitary approach will permit ministerial cherry-picking of EU law even as it applies to devolved matters without any consultation, which is the reason the Scottish government continues to insist on its own continuity bill. It is not parliamentary sovereignty that seems in evidence here, but executive control underpinned by a vigorous concept of government prerogatives (illustrated by May’s recent authorisation of a missile attack in Syria without consulting the Commons). So much for Parliament ‘taking back control’.
The UK Supreme Court may be called on to determine the legality of Scotland’s continuity bill if no agreement is reached on the withdrawal bill. The court may agree with Holyrood’s presiding officer, Ken Macintosh, who (taking the opposite line to his Welsh counterpart) declared that the Scottish Parliament lacked the power to enact the continuity bill. His reasoning rested not on any perceived conflict between the bill and the powers reserved to Westminster, but on the belief that Holyrood could not legislate in this area until EU treaties no longer applied in Scotland. The Scottish government does not accept Macintosh’s opinion, and the Lord Advocate, Scotland’s chief law officer, argued that if the Scottish bill was breaking EU law in legislating for powers that its Parliament would obtain after Brexit, so too was the UK government in its withdrawal bill – an argument that seems persuasive to me. There are few precedents to consult: the Supreme Court has heard only a handful of cases in which it pronounced on the legality of devolved legislation prior to its becoming law. However, in at least a couple of Welsh cases it has upheld the legality of devolved legislation. The Supreme Court’s last (and, so far, only) Brexit case, Miller v. Secretary of State for Exiting the EU, was decidedly not a victory for the UK government. The court concluded that parliamentary consent was necessary for the triggering of Article 50: executive action alone would not do.
It should be noted that even if a continuity bill were upheld in court, the UK government has the option of enacting the withdrawal bill for the entire UK, even in the absence of devolved consent, because Westminster, as a sovereign Parliament, may override legislation passed by devolved legislatures. The UK is not a federal state but a union state that employs devolution, and devolved authorities lack the protection federalism would provide. Such is the price of maintaining the sovereignty of the Westminster Parliament. (Small wonder that Hugh MacDiarmid considered Scotland’s experience within the UK to be colonial in nature.) Of course, this would not be a happy situation. The UK Parliament would be overriding Scotland’s competing continuity bill and, by enacting its withdrawal bill, ignoring the Sewel Convention (now also set out on a statutory basis in the Scotland Act 2016 and Wales Act 2017). To be sure, in Miller, the Supreme Court unanimously held that constitutional conventions, such as Sewel, are not enforceable by the courts. But it is still a grave matter politically to ignore such a convention.
In any case, it seems unlikely that the framework of the English constitutional settlement of 1689, which gave rise to parliamentary sovereignty, can resolve the revolution unleashed by Brexit, not least because the devolved nations have already experienced constitutional transformation. James Bryce, an Oxford law professor and Dicey’s rival, wrote of the centripetal and centrifugal forces that operate on constitutions. Centrifugal currents carry a risk of dissolution, allowing small sections of populations to determine outcomes. Scottish independence could have been accomplished by 50.01 per cent of those voting in the 2014 referendum in an already small nation, and it is possible that Brexit may give rise to another independence bid. Brexit is seen as requiring the UK to leave the EU even though it was an ‘advisory’ referendum in which only 37 per cent of eligible voters voted Leave. Indeed, Brexit, although argued by many to be centripetal (taking back sovereignty for the UK), may ultimately prove centrifugal, tearing the Union apart.
The informal, asymmetrical nature of devolution in the UK constitution doesn’t sit well with Brexit. But it is unlikely that Brexit will provide the ‘constitutional moment’ necessary to inspire a written constitution or a federal solution. It seems instead that Brexit has deepened existing divisions. Those who support Scottish independence, or an all-Irish state, have little interest in a settlement that would lock them into the UK, even if it were to guarantee equality of rights between different national communities. Similarly, those most acutely concerned about EU encroachment on national sovereignty have little interest in a more federal Britain. Ultimately, there seem to be only two possible outcomes: either Brexit will shatter the devolution settlement, or taking back control from one Union will threaten another.