Ripening time? The Welsh Labour government between Brexit and parliamentary sovereignty

The Welsh Labour government occupies a unique position in UK territorial politics, favouring neither the status quo nor independence for Wales while advocating a new settlement for the whole state. This article provides a detailed examination of its policy, focusing on its position on the doctrine of parliamentary sovereignty. Drawing from a range of documentary sources, we analyse the Welsh government’s constitutional proposals and its decision-making in the wake of the 2016 referendum on European Union membership. We argue that Welsh policy is defined by ambiguity. While it advances an alternative constitutional vision, it refrains from rejecting Westminster’s sovereignty outright. In the aftermath of the referendum, it sought to accommodate that sovereignty with its own constitutional claims through enhanced intergovernmental collaboration. In light of the Johnson administration’s centralising reforms, the strategy appears to have failed. Caught in the fractious politics of the Union, Welsh constitutional policy now faces an uncertain future.

remainder of the article, we explore its decision-making in the aftermath of the 2016 referendum. Brexit processes have forced the Welsh government to confront parliamentary sovereignty through the legislative consent process (i.e. the 'Sewel Convention') and in the courts: confrontations which have revealed significant caveats to its position. Drawing from a range of documentary sources from the past decade, we analyse not only what the Welsh government thinks the United Kingdom ought to look like, but also its decisionmaking as major constitutional changes were in motion. We conclude by reflecting on the possible fate of Welsh government proposals amid the wider politics of the Union.
Welsh constitutional policy, we argue, is defined by ambiguity. While it questions the democratic merits of Westminster sovereignty in the light of devolution schemes sanctioned by referendum, grounding its position in a tacit claim of Welsh sovereignty, the Welsh government has stopped short of rejecting the doctrine outright. During the Brexit process, it sought to accommodate Westminster's power with its own claims through enhanced intergovernmental collaboration. In effect, it sought to sustain a condition of 'Schrodinger's devolution' in which different understandings of the state co-exist (Sandford and Gormley-Heenan, 2020). Its approach therefore contrasts with both the Scottish government and Boris Johnson's administration in London, which tend to focus on ultimate authority, and where it is located. Faced by the Johnson administration's centralising reforms -notably the UK Internal Market Act 2020 (see Wincott et al., 2021b) -the strategy appears to have failed. Consequently, the Welsh government's contestation of parliamentary sovereignty has intensified. With UK constitutional arrangements and territorial politics in flux, Welsh constitutional policy now faces an uncertain future.
Welsh Labour's proposals emerged from this permanent churn. On the one hand, soon after devolution the party began to emphasise its distinctive Welsh character. It declared itself 'Welsh Labour: the True Party of Wales' and pitched its social democratic credentials apart from New Labour with the rhetoric of 'Clear Red Water' (cf. Evans et al., 2021). Its Cardiff and Westminster wings, however, have been and remain divided over devolution. Wyn Jones and Scully (2012) recount how this contradiction shaped the initial devolution dispensations: weakness in their design reflected attempts to contain the divisions. The 'One Wales' coalition with Plaid Cymru (2007Cymru ( -2011) marked a shift: it paved the way for the referendum on primary law-making powers and 'dramatically blurred the ideological differences' between the two parties (Evans, 2018: 499). A pro-devolution, 'soft nationalism' has since been 'the hegemonic ideology within Welsh Labour' (Moon, 2016: 283).
On the other hand, the party remains committed to the UK state. It shares what Brown Swan and Cetrà (2020: 50) call the 'welfare unionism' of the wider party, which 'emphasizes the instrumental importance of the Union in providing for a welfare state'. For Carwyn Jones (2013), former leader and First Minister (2009, Welsh Labour supports a United Kingdom '. . . that guarantees the fundamental gains made by the welfare state in the 20th Century'. Thatcherism, diversifying health systems, post-2008 austerity, punitive benefit sanctions and the partial devolution of welfare to Scotland have not pierced this vision; it is deeply embedded in Labour traditions. The party thus seeks to blend support for Welsh autonomy with a commitment to UK-wide social solidarity, as it understands it. From 2012, the party began to think more deeply about the United Kingdom's constitutional trajectory. The Senedd had acquired full primary law-making powers, creating some parity between the United Kingdom's legislatures and a position from which to advocate a new settlement. Britain's political map was also changing. In 2010, Labour lost power to the Conservative and Liberal Democrat coalition at Westminster. The following year the SNP consolidated its position, gaining a Holyrood majority, in turn leading to the 2012 Edinburgh Agreement and 2014 referendum on Scottish independence. The UK coalition government also proposed a commission on Welsh devolution, which became the Silk Commission (2011)(2012)(2013)(2014). Welsh Labour was the last enclave of national power for the wider party and found itself confronted by major constitutional change.
In a series of speeches between 2012 and 2015, Carwyn Jones proposed wide reforms of the territorial constitution. In his 'New Union' pitch delivered in the aftermath of the Scottish referendum, Jones (2014) called for a new settlement after more than a decade of disjointed, ad hoc devolution. Following the 2016 referendum, the Welsh government channelled this New Union thinking into demands for the withdrawal process and post-Brexit governance, which it articulated in Securing Wales ' Future and Brexit and Devolution (Welsh Government, 2017;Welsh Government and Plaid Cymru, 2017). Two years later, by then under Mark Drakeford's leadership, Reforming Our Union (Welsh Government, 2019b, 2021c finessed the proposals further. This document remains the Welsh Labour government's most detailed constitutional policy statement, recently republished following the party's May 2021 Senedd election success. The proposals have grown increasingly sophisticated, yet the basic pitch has remained: 'entrenched devolution settlements within a strong United Kingdom' (Welsh Government, 2019b: 2).
Welsh constitutional policy, however, has an 'elite, official, character' (Rawlings, 2019: 299). So far, it has been produced by a small group of ministers and officials, not by democratic consultation, civic dialogue or debate among Welsh Labour's rank-and-file membership. Carwyn Jones made a key contribution: a barrister and former Counsel General, Jones' interventions established the basic ingredients of Welsh Labour's subsequent thinking. Mark Drakeford, who succeeded Jones as Welsh Labour leader and First Minister on 13 December 2018, carried it on. A professor of social policy and former special advisor to Rhodri Morgan, the former First Minister (2000)(2001)(2002)(2003)(2004)(2005)(2006)(2007)(2008)(2009), Drakeford also served as Cabinet Secretary for Finance ( ) and Brexit minister (2017 under Carwyn Jones. Mick Antoniw, another lawyer, has also been influential: twice Counsel General (2016-2017, 2021-present) and latterly Minister for the Constitution (2021-present). Jeremy Miles, also a lawyer, is another key figure. Counsel General (2017-2021) for much of the withdrawal process, he also held responsibilities related to Brexit in Drakeford's cabinet (2018Drakeford's cabinet ( -2021. Officials have also been critical; in particular, Hugh Rawlings. Another lawyer, Rawlings served as the Welsh government's Director of Constitutional Affairs under Jones and Drakeford. This elite authorship made far-reaching constitutional claims, advanced in the name of Wales, and for the United Kingdom as a whole.

Welsh sovereignty within a reformed United Kingdom
For Carwyn Jones (2014), the referendums establishing devolution 'created' a 'New Union': 'a presumption of popular sovereignty in the different parts of the UK . . . [which] fundamentally challenged assumptions about a centralised British state'. The United Kingdom should no longer be viewed in the 'Old Union' way, as a unitary state predicated on Westminster sovereignty, but as 'a voluntary association of nations'.
The Welsh government subsequently attributed two constitutional implications to this premise. First, '. . . it must be open to any of its parts democratically to choose to withdraw from the Union' (Welsh Government, 2019b: 4). Second, since devolution has grounded sovereignty in the constituent parts of the United Kingdom, parliamentary sovereignty is '. . . outmoded and inappropriate' and '. . . no longer provides a firm foundation for the constitution of the UK' (Welsh Government, 2017Government, : 19, 2019b. Electoral diversification between the constituent parts, it argues, has accentuated this shift: 'no one political institution or political party can legitimately claim alone to speak for the whole of the UK' (Welsh Government, 2017: 7). Wide-ranging reforms were necessary to reflect the multiple sovereignties 'shared' in 'a joint project . . . based on a recognition of our mutual inter-dependence' (Welsh Government, 2019b: 2).
Faced by multiple sites of democratic legitimacy and legislative authority, Welsh policy thus calls into question the normative basis of parliamentary sovereignty. At the same time, it invokes an 'autochthonous' claim of Welsh popular sovereignty (Jones, 2012b: 153). As recently as 2013 the Welsh government had been reluctant to invoke such a claim explicitly: it was 'a bit too strong meat', in the words of Hugh Rawlings (Osmond, 2013). The near-miss of Scottish independence seems to have emboldened the Welsh government; its claims about the nature of the United Kingdom became much more farreaching thereafter.
For a devolved government determined to enhance its intra-UK influence, perhaps the appeal to popular sovereignty was inevitable. Daly (2019: 629) notes that the appeal to constituent power, or 'popular will', has long been used as 'a technique of legitimation . . . a way, discursively and rhetorically, of rationalising a given kind of political order'. Welsh political elites, however, are unable to invoke the kind of historic (if contested) foundations of sovereignty provided to Scotland by the Declaration of Arbroath 1320, the Union of 1707 and the Claim of Right 1989. Neither can they point to a principle of democratic consent regarding the constitutional status of Wales, as exists in respect of Northern Ireland under the Belfast/Good Friday Agreement and Northern Ireland Act 1998. The devolution referendums, therefore, seem to provide the strongest basis for a Welsh sovereignty claim (Hadfield, 2011;Jones, 2012b).
As we have seen, however, the Welsh government invokes this claim not only as a basis for Welsh autonomy. It fuses autonomy with an appeal for 'shared' sovereignty and the reconstruction of the UK state. At one level, this chimes with long-standing critiques of 'exclusive' notions of sovereignty in the light of growing inter-dependence between places (e.g. Mitchell, 2020). More specifically, it reflects a 'relational' understanding of sovereignty (Loughlin and Tierney, 2018): changing understandings of constituent power, the political foundation of sovereignty, entail changes to constituted power, here the doctrine of parliamentary sovereignty. According to Loughlin andTierney (2018: 1009), various changes since the 1970s -including legislative devolution and the 'growing disaggregation of the cultural-political notion of the 'British people'' -have divested the UK parliament of the political foundations of its power. Consequently, it can 'no longer claim to be the sole repository of sovereign authority' (Loughlin andTierney, 2018: 1009). The same relational understanding is reflected in the Welsh government's position: devolution reflects a fundamental shift in constituent power within the United Kingdom; as a result, constituted power -parliamentary sovereignty -must also change.

'Entrenched devolution'
These foundational claims form the basis of numerous Welsh government proposals. To initiate a process of renewal, it has called repeatedly for a UK constitutional convention. It demands the formalisation of intergovernmental structures, fiscal reform and repurposing the House of Lords as an elected chamber with equal territorial representation. In place of ad hoc devolution, it wants subsidiarity, with powers 'allocated to the most local level at which they can be performed efficiently and effectively' (Welsh Government, 2019b: 6). In this spirit, it advocates the devolution of justice and policing, among other policy areas (Welsh .
More broadly, as we have seen, the Welsh government supports 'entrenched devolution'. It has made various suggestions, however, for how entrenchment could be achieved. On the one hand, it has -infrequently -indicated an openness to a codified constitution. Carwyn Jones (2012aJones ( , 2014 initially advocated 'a written constitution . . . which commits and binds' but later spoke more loosely of the need for 'federal thinking'. On the other hand, Reforming our Union comments favourably on a codified constitution but leaves substantive proposals to a constitutional convention. The Welsh government has always held a qualified position on parliamentary sovereignty. The problem, it suggests, is the 'exclusive' version of the doctrine, 'as traditionally understood' (Welsh Government, 2017Government, , 2019b. It 'needs to be adjusted . . . just as it was adjusted to take account of the UK's membership of the European Union' (Welsh Government, 2019b: 5). Reforming our Union sets outs different options for this 'adjustment'. These relate primarily to the Sewel Convention, the practice by which Westminster does 'not normally' legislate in relation to devolved competences or policy areas without the consent of the legislatures affected. One option, it suggests, is a political agreement to clarify the circumstances in which the UK parliament could legislate in the absence of consent. This could be combined with procedural adjustments to Westminster's legislative process to facilitate better consideration of a decision to withhold consent. In addition, it suggests, the criteria underpinning 'not normally' could be set out in statute and subject to judicial oversight.
Alongside these proposals, the Welsh government pitches a 'simpler, albeit more radical, approach' (Welsh Government, 2019b: 8). A new constitutional settlement, it argues, could stipulate a legal principle that the Westminster parliament cannot legislate on devolved matters without the consent of the relevant legislature. This would eliminate 'uncertainty, misunderstanding and distrust' (Welsh Government, 2019b: 8). In a speech delivered on the day that Reforming Our Union was first published, Mark Drakeford (2019) suggested that the two sets of reforms should be linked: the first provided shortterm solutions, the latter an endpoint.
These changes could dramatically alter the workings of the UK territorial constitution. Notably, however, even the 'more radical approach' leaves the door open to the retention of parliamentary sovereignty. It has long been argued that the UK parliament can make 'manner and form' changes to the exercise of its power. In other words, it can implement 'statutory change to the law-making process' (Gordon, 2019: 136). Thus, without divesting its formal legislative competence, a future law could stipulate that the Westminster parliament would require the consent of devolved institutions before it could legislate for their territories. This seems to be the arrangement envisaged: Drakeford (2019) argued that the retention of Westminster's ability to act on a UK-wide basis was desirable, since 'there will be occasions and purposes where those legislatures will want to act in that way'. As Rawlings (2015: 477-478) previously observed, the Welsh government's proposals do not insist on formally divided sovereignty: instead they embrace 'federal-type ideas of shared rule and self-rule in a flexible system of multi-level governance', organised around a 'strong common core'.
While favouring stronger judicial protection for devolution, therefore, Welsh government proposals do not reject Westminster's legislative supremacy outright. Instead, they reflect an amenability to various modifications to the exercise of its power, that range from mutually agreed, non-justiciable principles to more far-reaching 'manner and form' changes. Its proposals reveal a critical but qualified position on parliamentary sovereignty. Clearly, the Welsh government has been willing to compromise on questions of constitutional principle. The next section attempts to make sense of this position.

The ambiguity of the Welsh Labour way
Given its instrumental view of the Union, the near-constant political and economic upheaval which has faced the state for the past decade may have compelled the Welsh government to advocate constitutional renewal. The 'New Union' speeches were not only driven by flaws in Welsh devolution but also by the prospect -and narrow avoidance -of Scottish independence. Brexit gave renewed impetus to that advocacy. It threatened the economic stability of the Union while revealing limits to the Welsh government's power to respond. With the Conservatives in tentative command of the Westminster parliament as the Brexit negotiations began and subject to only informal intergovernmental structures (McEwen et al., 2020), the devolved governments faced a situation over which they had little control.
The 2016 referendum also exposed a vulnerability in Welsh Labour's electoral dominance. Against the advice of the Welsh government, a majority (52.5%) in Wales voted for Leave, marking a 'clear disconnect' between government and governed (Hunt et al., 2016: 828). Support for leaving was highest in some of Wales' most deprived areas, particularly among those on low income (Henderson et al., 2021). Yet the potential economic ramifications of Brexit were substantial: loss of funding from the Common Agricultural Policy and European Structural and Investment Funds, along with significant threats to Welsh agriculture, manufacturing, ports, and an economic strategy focused on foreign direct investment. In that context, a 'soft nationalist' demand for greater empowerment of the devolved institutions was perhaps unavoidable.
By contrast, the flexibility evinced by the Welsh government's proposals provides further evidence of the 'good unionist' strategy identified by Hunt and Minto (2017).
Compounding the absence of formal intergovernmental structures, historically Wales had no threat of secession comparable to Scotland's. Hunt and Minto (2017: 659) argue that the Welsh government has often adopted a 'cooperative approach', 'presenting itself as a ''like-minded' interlocutor for the UK Government' and preferring evidence-based contributions over political contestation. By positioning itself as the reasonable partner, it seeks to maximise its influence with the UK government.
The Welsh government's stance on parliamentary sovereignty can therefore be attributed to a tension between constitutional aspiration and intergovernmental strategy. More importantly, it reveals an attempt to manage those influences through ambiguity. Wincott et al. (2021a: 8) argue that ambiguity can provide decision-makers a way of managing disunity: Actors may choose to live with, rather than resolve, contradictions they perceive . . .
[A]mbiguities may allow a state to carry on -to endure at times when clarity or a decisive choice between constitutional interpretations would make it unsustainable.
Sub-state parties in various contexts have embraced ambiguous territorial goals in order to navigate constitutional constraints and competing political and electoral demands (Elias and Mees, 2017). Welsh Labour, it seems, is no different. Indeed, ambiguity has obvious strategic appeal for the party. The Welsh sub-state is made up of an increasingly complex myriad of national identities and constitutional preferences (Wyn Jones, 2019). Furthermore, a significant section of Welsh Labour's supporters -potentially up to half -now favour Welsh independence (Larner et al., 2021). For the UK government and much of the Labour Party, however, the idea of parliamentary sovereignty remains sacrosanct. An ambiguous position on the constitution -pitching alternative sovereignty claims but without rejecting outright those of Westminster -may therefore command greater electoral support and facilitate party unity while also leaving space for cooperation and compromise with a future UK government.
Constitutional ambiguity can also be viewed as the 'normal' mode of the UK state. Sandford and Gormley-Heenan (2020) observe how, since 1999, permanent and increasingly powerful devolved institutions have co-existed with a supposedly omnipotent Westminster parliament. In this condition of 'Schrodinger's devolution', the state appears as both changed and unchanged. The Welsh government's proposals reflect an attempt to reinvigorate this ambiguity. Like other recent initiatives seeking to preserve the Union, it tries to reconcile Westminster's legislative power with sub-state autonomy and the competing sovereignty claims which underpin them (Davies and Wincott, 2021).
As we shall see, the Welsh government also sought to maintain 'Schrodinger's devolution' during the Brexit process. It demanded that Westminster's sovereignty was exercised compatibly with its devolved interpretation of the constitution. Constitutional ambiguity, however, cannot be sustained by a devolved government alone; it requires self-restraint from Westminster. The success of the Welsh government's strategy thus hinged largely on the latter's willingness to restrain itself.

Prolonging ambiguity: The 2018 Intergovernmental Agreement
Official Welsh government policy provides important but limited insights into its constitutional perspective; its decisions during and since the Brexit process are equally revealing.
Early on, the Welsh government actively sought to place limits around the exercise of parliamentary sovereignty. In January 2017, for example, despite misgivings, it consented to the Wales Bill. By placing the Sewel Convention on a legislative basis, the Welsh government envisaged that the resulting Act would offer stronger constitutional protection for Welsh devolution. Around the same time, it intervened in the first Miller case to argue that the UK Supreme Court should interpret the common law in accordance with the Sewel Convention's requirements. It advanced its vision of a 'voluntary association of nations' in which devolution was 'fundamental and effectively irreversible' (Welsh Government, 2016: 3) -an 'expansive claim to constitutional power-sharing' (McHarg, 2018: 178). Had the submissions been accepted, and Sewel given legal effect, the Welsh government's position would have been strengthened significantly. In the event, the UK Supreme Court (2017) ruled that the convention was non-justiciable.
The tilt thereafter towards ambiguity became evident in the intergovernmental dispute over the European Union (Withdrawal) Bill (EU(W)B), introduced in July 2017. The UK government proposed to transpose all EU rules and regulations at the point of exit into domestic law (retained EU law). Until 'common frameworks' to replace EU regulation had been jointly agreed, the devolved institutions would be prevented from modifying this body of law, including in areas intersecting with devolved competences, such as agriculture, food standards and the environment. By contrast, UK ministers would have broad powers to amend retained EU law, including within devolved areas. For the Welsh and Scottish governments, this was 'an attack on the founding principles of devolution' (Sturgeon and Jones, 2017). They proposed joint amendments to the bill and enacted their own legislative schemes to prepare the devolved systems for EU withdrawal. Each stipulated the removal of the restrictions on devolved competence and a legal requirement of devolved institutional consent for any UK-level regulations applying to devolved areas.
Following months of negotiations between the three governments, in April 2018 the Welsh and UK governments struck a deal. Comprising various legislative and non-legislative commitments, the 'Intergovernmental Agreement' (UK Government, 2018) is a model of ambiguity. The EU(W)B would be amended so that former EU competences intersecting with devolved policy areas would fall to the devolved institutions. The UK government would have powers to 'freeze' those competences, subject to UK parliamentary approval and only after requesting the consent of devolved ministers and providing their reasons to the UK parliament if consent was refused. The 'freezing' powers and their effects would also be time-limited by sunset clauses, meaning that any frozen powers not subject to UK legislative common frameworks would be restored to the devolved institutions upon expiry. The agreement also emphasised the importance of constitutional conventions and the need to proceed by consensus. Building on principles agreed in October 2017, an additional memorandum of understanding set out further terms for the development of common frameworks.
The Welsh government achieved several of its objectives. The principle of the reserved powers model -powers devolved unless already reserved -was secure. There was an enhanced role for devolved consent: a 'consent convention plus' (Rawlings, 2019: 311). Furthermore, according to the then Brexit minister, Mark Drakeford, the agreement offered 'a first significant step towards an equitable approach to inter-governmental working' of the kind advocated in Brexit and Devolution (Senedd Cymru, 2018). Its emphasis on collaboration, consensus and intergovernmental parity meant the agreement also aligned with the Welsh government's aspirations.
The agreement sustained a degree of constitutional ambiguity: the Westminster parliament was sovereign, but the exercise of its sovereignty would be moderated in view of the Welsh government's claims. There was a price, however: the demand for a legal requirement on UK ministers to obtain devolved consent before suspending devolved powers was abandoned. Sandford and Gormley-Heenan (2020: 11) observe that the agreement thus 'sits firmly within the tradition of parliamentary sovereignty' and 'speaks to traditional UK territorial management: informal negotiations, with central sovereignty as a backstop'. This was acknowledged by Drakeford when he presented the agreement to the Senedd: . . . [T]here are those who argue that it is unacceptable, even in extreme circumstances, that Parliament could act to impose constraints on devolved competence. But until a new constitutional settlement for the whole United Kingdom is negotiated -for which this Government has long argued -it is the constitutional reality that Parliament retains that role. (Senedd Cymru, 2018) This was a notable concession on the insistence that 'no one political institution or political party can legitimately claim alone to speak for the whole of the UK'. Amid the pressures of Brexit, earlier bold sovereignty claims seemed to have been tempered. Rawlings (2019) identifies a number of factors which may have informed this stance: a perceived lack of leverage created by the Leave vote in Wales; concerns over the regulation-making capacity of Welsh devolved institutions; and a determination to avoid a precedent of Westminster enacting Brexit legislation without consent. Opting to play the 'good unionist' may have also bought the Welsh government credibility and influence for the common framework negotiations ahead.
Equally, it seems that deeper constitutional questions could be left for another day. Shortly after the agreement was reached, and not long before he became First Minister, Drakeford (2018) gave an illuminating speech on 'Brexit and Devolution' in which he invoked the principle of 'unripe time' favoured by his predecessor, Rhodri Morgan. This, he explained, is a code of political self-restraint, predicated on the constructive potential of ambiguity and the scope for solutions to emerge over time. As he put it, 'which problems do you really need to solve, and which problems are better left?' Drakeford was speaking of the UK-EU future relationship negotiations, but his comments might be read across to the agreement which he had just negotiated. Clearly, for the Welsh government, the moment to challenge Westminster's legislative power was not yet ripe: better for now to maintain devolved competences and seek to develop post-Brexit governance consensually. The approach aligns with the 'more pragmatic territorial strategies' which Elias and Mees (2017: 13) observe in other sub-state contexts, '. . . whereby ambitions to radically overhaul the territorial organisation of the state have been put aside in order to focus on other more pressing policy challenges'.
The 2018 Agreement provided a template for negotiations on subsequent Brexit bills (Miles, 2019). Wherever the UK government proposed to grant itself powers with a reach into devolved areas, the Welsh government sought assurances that devolved consent would be sought and political commitments on intergovernmental collaboration. For the duration of Theresa May's tenure (2016)(2017)(2018)(2019), this approach seemed to be vindicated. Work on common frameworks progressed and the powers to suspend devolved competences were never used. The reign of constructive ambiguity and cooperation, however, proved short-lived. After Boris Johnson became Prime Minister on 24 July 2019, the Welsh government's attempts to accommodate Westminster sovereignty with its own constitutional vision were tested to destruction.

Relying on Westminster, defending its sovereignty
Despite the 2018 Agreement, constitutional ambiguity was dissipating before Johnson assumed office. The devolved governments had been excluded from decision-making over the form, timing and nature of EU withdrawal. The May administration had enacted the EU(W)A, despite the refusal of the Scottish parliament to consent to the legislation. The UK Supreme Court (2018) had also confirmed the right of the Westminster parliament to alter the powers of the devolved legislatures unilaterally. In place of a New Union, the territorial constitution appeared to be reconfiguring on a more hierarchical basis.
In important respects, however, the Welsh government's own approach to Brexit may have undercut its attempts to preserve constitutional ambiguity. For instance, frequently it has relied on Westminster legislating for Wales on devolved matters rather than seeking to pass legislation through the Senedd, citing pressures on the Senedd's legislative timetable. The Agriculture Act 2020, Fisheries Act 2020 and Environment Act 2021 are among stark examples of this turn to London. Furthermore, in place of legal requirements for devolved consent, the Welsh government has repeatedly accepted non-binding intergovernmental agreements with its UK counterpart to overcome disagreement and regulate the latter's new law-making powers under Brexit legislation within devolved areas (Davies, 2020). The Fifth Senedd's Legislation, Justice and Constitution Committee (formerly Constitutional and Legislative Affairs Committee) argued that these approaches are 'constitutionally unacceptable' and risk 'weakening the devolution settlement' (Constitutional and Legislative Affairs Committee, 2019: 31; Legislation, Justice and Constitution Committee, 2020: 18).
The Welsh government's tactical use of parliamentary sovereignty is even more striking. Intervening in the legal challenge to the UK government's prorogation of Westminster in August 2019, it submitted 15 pages of argument to the Supreme Court in defence of 'not a narrow, but a broad, conception of Parliamentary Sovereignty, and one which is delineated by the Courts as part of their constitutional function' (Welsh Government, 2019d), drawing on a range of legal authorities to demonstrate the principle's 'breadth and dynamism'. Of the 15 pages, just 2 dealt with devolution. The prorogation, it said, had denied the scope for inter-parliamentary 'dialogue' between Westminster and the Senedd, which was a 'fundamental feature of the United Kingdom's constitution'. Its submissions differed considerably from those of the Scottish Government (2019), which mostly avoided parliamentary sovereignty.
Justifying the intervention, Counsel General, Jeremy Miles, stated that his duty as a law officer was 'to uphold the rule of law and the constitution' (Welsh Government, 2019a). Even so, the submissions represent the high point of dissonance between the Welsh government's ambition to 'adjust' parliamentary sovereignty and its decision-making in the Brexit context. The UK Supreme Court (2019) held that the prorogation had indeed been unlawful, invoking a broader interpretation of parliamentary sovereignty. No straightforward victory for the Welsh government; however, arguably the judges' reassertion of parliamentary sovereignty was a setback for its aspiration to rebalance territorial power on the basis of alternative constitutional principles. Facing a potential 'no deal' rupture from the EU and a UK government decision which it regarded as unlawful and undemocratic, it seems the Welsh government saw the occasion as calling for a very different stance on Westminster's sovereignty.
Taken together, these instances show that the Welsh government's critique of parliamentary sovereignty remained secondary to other Brexit-related concerns, be they devolved institutional capacity, intergovernmental relations or the rule of law. Its position has been tailored to political circumstances, and as a result not always consistent with overarching constitutional aims.

The illusion fades
Despite these compromises, the Welsh government did try to sustain constitutional ambiguity during the Johnson premiership. The European Union (Withdrawal Agreement) (EU(WA)) Bill provides an illuminating example. Introduced by the UK government following the general election in December 2019, the legislation made provision for the implementation of the UK-EU Withdrawal Agreement. Rather pointedly, Clause 38 asserted that 'the Parliament of the United Kingdom is sovereign'. Despite its suspicions, the Welsh government did not demand the removal of this clause. Instead, it requested an amendment to acknowledge the Sewel Convention alongside it. Echoing this willingness to accommodate, Jeremy Miles described cl 38 as 'a form of parliamentary sovereignty that . . . fundamentally misunderstands the changed constitution' (Senedd Cymru, 2020b).
Rather than objecting to the doctrine outright, the Welsh government took exception to a particular form of parliamentary sovereignty -consistent with its position in Reforming Our Union. It wanted Westminster's legislative supremacy to be exercised with greater sensitivity to devolution. These efforts, however, were in vain. The bill was enacted unchanged, notwithstanding the Senedd, the Northern Ireland Assembly and the Scottish Parliament refusing their consent to it.
Ambiguity faded further with the Internal Market Act 2020. The legislation introduced a 'Market Access Commitment', underpinned by the principles of mutual recognition and non-discrimination, ostensibly intended to ensure that goods and services produced in one part of the United Kingdom are not subject to different regulatory requirements or competitive disadvantages in another part. While the devolved institutions retain an ability to regulate goods and services, in future their regulations will only apply to goods and services originating from within their respective territories. UK ministers, meanwhile, are granted powers to amend the application of the market access principles, subject to a requirement to request the consent of the devolved governments. The Act also reserves state aid and equips UK ministers with new spending powers in a broad range of devolved areas.
More generally, the legislation accentuates the disparity between the UK and devolved legislatures, already entailed by parliamentary sovereignty. It now features in the devolution statutes as a 'protected enactment', immune from amendment by the devolved legislatures. By contrast, the Westminster parliament's unlimited legislative competence means it is subject to no such constraint, strengthening the hierarchical position of UK-level over devolved institutions. The UK government also invoked parliamentary sovereignty to justify the legislation. The United Kingdom, it argued, is 'a unitary state with powerful devolved legislatures' (Department for Business, Energy and Industrial Strategy, 2020), and the 'fundamental principle' of parliamentary sovereignty, as reasserted under the EU(WA) Act 2020 s.38, meant that even the most contentious provisions were perfectly constitutional (UK Government, 2020).
The Welsh government's opposition was led by Jeremy Miles. He had wanted the market access principles to operate as a measure of last resort: limited to those areas where agreement could not be found on common frameworks and given the same flexibility granted under EU rules. Reflecting a shift from the 2018 Agreement, he proposed that UK ministers be required to obtain the consent of devolved ministers before they could amend the market access principles, along with the removal of UK ministerial spending powers in devolved areas and the statutory reservation of state aid. Miles' proposals were not accepted. The legislation was enacted, despite the Senedd and the Scottish Parliament again withholding consent. The attempt in the 2018 Agreement to mediate parliamentary sovereignty through 'consent convention plus' mechanisms and enhanced intergovernmental collaboration was undermined. The Act also provided flat contradiction of the Welsh government's vision of the United Kingdom as a 'voluntary association' with permanent, consensual devolution schemes.
The Internal Market Act was followed by a further blow. On 24 December 2020, the United Kingdom and the EU struck an agreement regarding their future relationship. The implementing legislation, the European Union (Future Relationship) Act 2020, was passed in a single day. The significant ramifications for devolved governance were not considered or debated. The Senedd was recalled and a motion was passed expressing regret that it was 'not in a position to determine legislative consent' (Senedd Cymru, 2020a).
As the Johnson administration reasserted Westminster primacy, the Welsh government hardened its position. In January 2021, it formally launched judicial review proceedings over the Internal Market Act. It sought a declaration that neither the 'protected' status of the Act nor the broad Henry VIII powers granted to UK ministers could limit the scope of the Senedd's competence (Welsh Government, 2021a). The challenge was dismissed on grounds of prematurity (no Senedd legislation or UK government decisions were in issue) (Divisional Court, 2021). Permission to appeal to the UK Supreme Court has been requested at the time of writing, but the fact of the challenge alone is significant. It is the first time that a devolved government has initiated proceedings against the UK government to challenge UK parliamentary legislation. Furthermore, while the Welsh government's submissions did not confront parliamentary sovereignty squarely, they invoked the principle of legality, and in particular the judicially recognised 'constitutional' status of the devolution statutes, to limit the effects of the Internal Market Act. Later, in June, the Welsh government also appeared as a respondent in the UK government's referral of two Scottish bills to the Supreme Court, where it argued that the devolved parliaments are entitled to amend UK parliamentary legislation under certain conditions (Welsh Government, 2021b). In both instances, the attempt to circumscribe Westminster's legislative power was clear.
The shift in the Welsh government's stance was also evident in an appearance by Drakeford before the House of Commons Welsh Affairs Committee on 4 March 2021. At one level, his evidence was layered with the usual ambiguities. Sovereignty, he said, '. . . has established itself in the different component parts of the United Kingdom'. It has not 'disappeared from the UK Parliament . . . but it is not exclusively there'. This aligned with the relational account of sovereignty articulated in Welsh policy documents. Notably, however, Drakeford also spoke of the need for '. . . devolution that cannot be unilaterally rolled back by any one party', and a constitutional settlement in which 'the powers that we have . . . would be guaranteed' (Welsh Affairs Committee, 2021).
The notion of irrevocable devolved powers may point to a lasting fallout from the Internal Market Act. Furthermore, whereas previously the Welsh government had insisted on waiting for a UK constitutional convention, following the 2021 Senedd election it initiated a 'national conversation' of its own with the establishment of the Independent Commission on the Constitutional Future of Wales. Drakeford also appointed Mick Antoniw as Counsel General and Minister for the Constitution. The latter has led recent calls within Labour for 'radical federalism'. This proclaims 'Ein Hawl' 'the right of people in Wales to determine the form of Government best suited to their needs' (Radical Federalism, 2021: 3) -and advocates Wales and UK-wide processes of civic dialogue on constitutional change. While the more ambiguous statements of Reforming our Union remain official policy, these developments suggest a potential shift in the Welsh government's goals. They point to a preference for the legal-constitutional entrenchment of substate autonomy, along with a new willingness to instrumentalise the views of the Welsh population to achieve that end.

Ripening time
'Soft nationalist', 'good unionist', 'ambivalent unionist': analysts have devised a variety of labels to capture the nuance that runs through the constitutional policy pursued by successive Welsh Labour governments. By focusing on parliamentary sovereignty, our analysis has drawn out the ambiguities that underpin this policy. For almost a decade, Welsh Labour governments have made known their dissatisfaction with the doctrine, evincing a relational understanding of sovereignty far removed from Anglo-British constitutional orthodoxy. And yet, they have shown considerable willingness to accommodate. Constitutional proposals and decision-making since the 2016 referendum have conveyed their openness to the Westminster parliament retaining its formal legislative competence in exchange for enhanced intergovernmental processes and further decentralisation. Under the multifaceted pressures of EU withdrawal, Welsh government decision-making often suggested that the more fundamental questions about parliamentary sovereignty had to be put on hold: they were a matter of 'unripe time'.
As reflected by multiple refusals to consent to Westminster bills and a legal challenge to the Internal Market Act, however, the Welsh government's willingness to contest parliamentary sovereignty has intensified. Ambiguity may be giving way to a demand for constitutionally entrenched, sub-state autonomy. Nevertheless, y Ffordd Lafur Gymreig/the Welsh Labour Way continues to be defined by a commitment to the United Kingdom which, though reformist, is unequivocal. Welsh ministers have indicated a preference for a federal arrangement even in the event of Scotland's independence . Notwithstanding increasing dissatisfaction with parliamentary sovereignty, therefore, there are evident limits to Welsh Labour's 'ambivalence' about the Union.
It remains to be seen what will become of Welsh constitutional policy. Clearly, UK-wide constitutional reform is not in the gift of the Welsh government. Constitutional arrangements and territorial politics, however, remain in flux. Live questions around 'levelling up' and the relationship between central and local government in England carry significant implications for the central governments of Scotland and Wales. Meanwhile, even after the contentious politics of the Internal Market Act, there are signs of more effective intergovernmental co-operation in the negotiation of common frameworks (Cabinet Office, 2021). Interest in territorial politics also seems to be growing in England, mostly focused on the role of metro-mayors, and there is some evidence of proposals for 'radical federalism' gaining wider traction in the Labour party (Rodgers, 2021). Viewed in this wider context, the Welsh government's proposals may prove to be an important contribution to a broader debate about federalism at a formative stage.
Equally, Welsh constitutional policy could have a short shelf life. An inclusive national conversation about Wales' future could yield new thinking and priorities. The Welsh Labour government is also caught between potentially powerful forces: the prospects of Scotland's independence and Ireland's reunification, significant support for Welsh independence among its own supporters, and a UK Conservative government with no interest in federalism. If time was ever ripe for Welsh Labour's vision of a reformed United Kingdom, it may have already passed. The state seems to be entering an uncertain stage of 'catastrophic equilibrium' or even 'failure', marked by a sustained inability to resolve its contradictions (Hay, 1999;Wincott et al., 2021a). As 'New Union' dreams fade in an 'Old Union' reality, time may ripen for a new Welsh way.