Abstract
This article investigates the impact of Brexit on the British political system. By critically engaging with the conceptualisation of the Westminster model proposed by Arend Lijphart, it analyses the strains of Brexit on three dimensions developed from from Lijphart’s framework: elections and the party system, executive– legislative dynamics and the relationship between central and devolved administrations. Supplementing quantitative indicators with an in-depth qualitative analysis, the article shows that the process of Brexit has ultimately reaffirmed, with some important caveats, key features of the Westminster model: the resilience of the two-party system, executive dominance over Parliament and the unitary character of the political system. Inheriting a context marked by the progressive weakening of key majoritarian features of the political system, the Brexit process has brought back some of the traditional executive power-hoarding dynamics. Yet, this prevailing trend has created strains and resistances that keep the political process open to different developments.
Introduction: The strains of Brexit
This article aims to ascertain how the politics of the UK has been reshaped by Brexit and, more tentatively, in which direction it is moving towards. While the Brexit process has shaken the political system and unleashed contradictory dynamics, we argue that a move back towards the ‘Westminster model’ has ultimately emerged. In so doing, this research places itself in continuity with studies that have provided analytical assessments on the ongoing transformation of the UK before the Brexit process started to unfold (e.g. Flinders, 2010; Matthews and Flinders, 2017). In line with these studies, analytical categories developed in the classic comparative literature, such as Arendt Lijphart’s Westminster model, are adopted as benchmarks in view of detecting longitudinal change. While other scholarship explores the Brexit–constitution nexus (see Bogdanor, 2019), the analytical focus here is on the impact of the Brexit process on the UK political system, in the period ranging from the Brexit referendum (June 2016) to exit day (January 2020).
The unsettling impact of Brexit has manifested itself in a series of rather unusual political occurrences: the alternation of three prime ministers (PMs) in 3 years; a highly volatile electoral market; and harsh clashes between government and parliament as well as between the central and devolved institutions, often leading to strong court involvement (including the Supreme Court) in the political process. As Brexit promises to leave a significant imprint on British politics, this article takes a close look at the trends that have surfaced at various points in time, looking especially at the party system and at the various strains between different institutions.
Three dimensions of the UK political system are singled out as particularly exposed to the potential effects of Brexit: the party system, executive–legislative dynamics, and the relationship between the central and devolved administrations. How important was Brexit in reshaping the party system? Has Parliament constrained government during the Brexit process? Has Brexit shifted the constitutional balance between central and devolved administrations? Our analysis shows that key features of the Westminster model – such as single-party majority government, executive dominance over Parliament and the unitary character of the state – have been reinforced.
Yet, this conclusion needs to be qualified, as it is based on the analysis of a limited period. First, important constitutional features of the UK political system that do not fit with the Westminster model – for example, bicameralism, devolution and the Supreme Court – are still there. Secondly, while the nexus between the two-party system, single-party government and cabinet dominance can still count on its institutional pillar (i.e. the majoritarian electoral system), the behavioural pillar (i.e. concentrated electoral preferences) has only partially recovered from the erosion of the previous half century. In addition, the persistence of high volatility casts doubts on the future solidity of this second pillar. Without indulging in difficult and sterile speculation on future developments, it is important to acknowledge that these consolidated transformations represent important constraints on the push towards the Westminster model that has recently re-emerged with Brexit.
Analytical framework: Lijphart and beyond
In his seminal work on democracy, Lijphart classified countries on two dimensions: executive–parties and federal–unitary (Lijphart, 1984; 1999). The first concerns the horizontal distribution of power, and consists of five interconnected variables: party system (V1; two-party vs. multi-party systems); type of cabinet (V2; single-party vs. multi-party coalition); executive–legislative relations (V3; executive dominance vs. balanced relationship); type of electoral system (V4; majoritarian vs. proportional); and type of interest group interaction (V5; pluralism vs. corporatism). The second dimension refers to the vertical distribution of power and comprises five more variables: relationship between central and territorial jurisdictions (V6; unitary vs. federal state); type of parliament (V7; unicameral vs. bicameral); type of constitution (V8; flexible vs. rigid); judicial review (V9; Parliament vs. Supreme Court as final arbiter on laws); and central bank (V10; dependent vs. independent). In this classification, the conceptual categories are derived from a basic dichotomy: majoritarian versus consensus democracies. Lijphart described the UK as the ideal-typical case of a majoritarian democracy, to the point that Westminster model and ‘majoritarian model’ have become synonyms in the literature (Russell and Serban, 2020). Indeed, the traditionally unitary character of the state (V6) and the flexible constitution (V8) have strengthened executives. The latter, as a result of the V4–V1–V2 nexus (i.e. majoritarian electoral system, two-party system and single-party government) and ‘the cohesiveness of the majority party’ (Lijphart, 1999: 12), has clearly dominated parliament. Despite the constitutional reforms adopted since the late 1990s, the last edition of Patterns of Democracy was still presenting the UK as ‘a prime example of majoritarian democracy’ (Lijphart, 2012: 20).
The most recent scholarship has challenged the aforementioned classic description of the British political system. Providing an overall assessment of constitutional change in the late 2000s, Robert Hazell wrote that ‘the Westminster model, formerly held up as the ideal type of unfettered majoritarian government, has seen the introduction of a series of new checks and balances’ which ‘have transformed the Westminster constitution’ (2008: 1–2). Even studies explicitly adopting Lijphart’s framework concluded that the British system has been moving away from some of its majoritarian elements (Flinders, 2010; Matthews and Flinders, 2017).
Based on this recent scholarship, we acknowledge that the majoritarian model, at least in its ideal-typical configuration, no longer accurately describes the UK political system. In this article, we analyse the Brexit years (23 June 2016–31 January 2020) to ascertain whether the British political system has been shifting further away or, rather, back towards the majoritarian pole of Lijphart’s typology. In order to conduct a focused analysis, we reformulated Lijphart’s analytical framework, which was configured for cross-country and long-term longitudinal comparative analyses. In contrast, our analysis concerns a single case – the UK – with a specific focus on a limited period.
First, it is important to consider the logical and empirical links between Lijphart’s variables (Taagepera, 2003). In this respect, we excluded the two variables that stood out for being logically disconnected (as well as weakly connected in empirical terms) from the other variables of the same dimension: interest representation (V5) and central bank independence (V10) (Lijphart, 1999: 306; Taagepera, 2003: 7, 12). In addition, while in a cross-country comparative analysis it is logical to expect a link between federalism (V6) and bicameralism (V7), unitary states can also have bicameral parliaments, whereby higher chambers are not designed to represent territorial units (Taagepera, 2003: 11). When such a case is studied in isolation, as in this article, it makes more sense to analyse bicameralism (V7) in relation to the horizontal distribution of powers, and in particular in connection to the type of cabinet (V2) and executive–legislative relations (V3), rather than in relation to the vertical distribution of powers. 1 Hence, we regrouped Lijphart’s variables under three dimensions that are particularly exposed to the Brexit process: elections and parties (V4 and V1); executive–legislative relations (V2, V3 and V7) and federal–unitary system (V6). The two remaining variables belonging to Lijphart’s second dimension – constitution (V8) and judicial review (V9) – are also included in our analysis: not as distinct variables but respectively as a contextual condition (constitution) and as an intervening factor (judicial review) that cut across our dimensions (especially the last two on the horizontal and vertical distribution of power). Finally, we integrated the quantitative indicators deployed in Lijphart (2012) and Matthews and Flinders (2017) with others, using additionally a wide array of qualitative data.
Elections and parties: A two-party system with a volatile electorate
A key feature of the Westminster model has always been the strict relation between the first-past-the-post (FPTP) electoral system and the two-party system, whereby the former helps alternation of single-party majorities between the two main parties – since 1945 the Conservatives and Labour. As Lijphart (2012: 13) remarks ‘the hegemony of these two parties was especially pronounced between 1950 and 1970’. As well as by the electoral system, this hegemony was helped by the predominance of the class cleavage. As famously remarked in one of the most frequently quoted analyses of British politics, ‘class [is] the basis of British party politics; all else is embellishment and detail’ (Pulzer, 1967). Here we show that Brexit has been an important issue during the three most recent UK general elections: triggering the referendum in 2015; being somehow less relevant than the calendar might suggest at first sight in the inconclusive vote of 2017; and the dominant issue in 2019. Before looking at how Brexit impacted on the party system, however, a caveat is in order. While in the next two dimensions citizens are only indirectly involved (mainly through their representatives and the executives, at national level or in the devolved administrations), in this dimension they have a key role. Hence, we complemented Lijphart’s institutional indicators (on party system fragmentation,V1, measured by the effective number of parties (ENPs) in its two variants of share of parliamentary party and of electoral parties, to which we added the two-party vote index) and disproportionality (V4, Gallagher Index) with some behavioural indicators that helped to contextualise the significance of the ‘Brexit effect’ in a long-time perspective.
Until 1970, in each general election the two major parties obtained around 45% of the votes: a small bonus in seats to the most-voted party was generally enough to ‘fabricate a majority’ in Westminster. Correspondingly, the disproportionality of the electoral system was low, and electoral alignments were very strong. Aggregate electoral volatility – measured here through the Pedersen Index – was very low, as voters tended to identify with (and to vote for) the same party for many years. The first column of Table 1 identifies these defining elements of the ‘classic’ Westminster model (Lijphart, 1984).
Selected indicators on party system and elections (1945–2019).
see Fieldhouse et al. (2020) and list below.
Sources: own calculation; Fieldhouse et al., (2020: 53).
Since 1974, the rise of regional parties, especially the Scottish National Party (SNP), and of the Liberal Democrats (Lib Dems), increased party system fragmentation (see the increasing difference between the two indexes of ENPs until 2015, Table 1). Yet, the FPTP helped the two main parties to remain in control. 2 Party cohesiveness was then eroded by the European issue and by the slow emergence of anti-European parties. While in general elections the latter were contained by tactical voting (in favour of parties standing a chance to win the seat), elections for the European Parliament have provided – especially since 1999, when a proportional electoral system was introduced – a key opportunity for the rise of the UK Independence Party (UKIP). In 2019, this happened again with the Brexit Party. 3
Our argument is that Brexit has been a key source of turbulence for the party system. However, behind the electoral prevalence of the Conservatives (the plurality party for four elections in a row: 2010, 2015, 2017, 2019), several important party system indicators of the Brexit years appear to have re-approached the classic Westminster model. In a context marked by long-time weakening partisan identifications (i.e. of rising partisan dealignment: see Table 1), Brexit has added fuel to other recent ‘electoral shocks’. Fieldhouse and colleagues (2020) have identified five of them: the EU immigration shock after the 2004 enlargement (and the ensuing rise of UKIP), the 2007–2008 financial crisis, the coalition government (2010–2015), the 2014 Scottish referendum and the 2016 Brexit referendum. While the effects of these shocks are often interrelated, they are strongly associated with vote switching: around 60% of British voters have switched their vote at least once in the 2010–2017 period. Most importantly, these shocks ‘have the potential to change how parties are perceived and therefore to reshape the party system’ (Fieldhouse et al., 2020: 189).
Before looking more closely to the last two elections, let us recall a key lesson. Giovanni Sartori (1986) argued that for FPTP to produce a two-party system there are two joint necessary conditions: party systemic structuring and cross-constituency vote dispersion. The former can be assessed through electoral volatility (Table 1), the latter refers to electoral geography. In 2015, we see important changes in both these domains. Volatility peaked at 17.8 (we still have 15.5 in 2017 and 10.3 in 2019). The 2015 peak was due to the combination of UKIP’s breakthrough (third highest vote share, quadrupling its vote from 3% in 2010 to 12.6%) and the collapse of the Europhile Lib Dems (down to 7.9% from 23%, mainly because of their difficulties as the junior partner in the coalition government; Fieldhouse et al., 2020: Ch. 7). This election set the scene for the referendum, as David Cameron had to comply with his promise of holding one in the event of a Conservative election victory. UKIP only obtained one seat but finished second in 120 constituencies: the high FPTP threshold was no longer unsurmountable. In this election, the rise of immigration as the main electoral issue testified to the relevance of Brexit (Dennison and Goodwin, 2015). Similarly, the new electoral geography of 2015 saw the combined record-low level of Conservative–Labour marginal seats and the SNP’s predominance in Scotland (Johnston et al., 2017). Hence, the vote was already geographically polarised before Brexit: a ‘new electorally divided nation from 2015 reflects polarised support for three parties – the Conservatives, Labour, and the SNP’ (Johnston et al., 2017: 534).
One year later, the 2016 referendum mobilised British voters like no other electoral event since the 1992 general election (72.2% turnout). Although turnout does not always affect seat distribution during general elections, given the aforementioned patterns of de-structuring, it can become decisive in obtaining a majority. 4
In 2017 Brexit mattered less than one might have expected when Theresa May called a snap election to foster her slim majority. As both main parties had just voted in Parliament in favour of triggering the 2-year exit procedure (Article 50 (art. 50) of the Lisbon Treaty), Brexit left the scene to the successful anti-austerity campaign of Labour’s leader, Jeremy Corbyn, which, combined with May’s poor campaign, delivered a result with no clear winner (Allen and Bartle, 2019).
In contrast, Brexit was the dominant issue in 2019, when the third Conservative PM in 3 years – Brexit-supporter Boris Johnson – called yet another general election to win a majority and finally ‘get Brexit done’ (his main campaign slogan). Experts are divided on the importance of tactical voting in 2019. Norris underlines that the selective entry of the Brexit Party was decisive. By withdrawing from the 317 seats with a Conservative incumbent, the Brexit Party – despite its poor overall result: 2% – subtracted Leave voters to the Labour Party, especially in the famous former ‘red wall’ area of the North of England, further mutating the electoral geography of the country: ‘the share of the Brexit vote was large enough to allow the Conservative to . . . make up to 20 seat gains in former Labour areas, thereby doubling Johnson’s eventual parliamentary majority’ (Norris, 2019). Mellon (2020: 8) argues that ‘On balance, the evidence is that the Brexit Party pact was probably unnecessary given the high rate at which their voters defected to the Conservatives anyway’. Be that as it may, Johnson got the best of both worlds: by refusing pacts with Nigel Farage, he kept his hands clean, while the latter lured ‘Labour Leavers’: indeed, ‘Labour’s support declined more heavily in pro-Leave seats when there was a Brexit Party candidate than when there was not’ (Cutts et al., 2020: 9).
Post-election research underlines the importance of Brexit as a determinant of electoral choice (Curtice, 2020). Other factors certainly played a role, starting from the dissatisfaction with Corbyn’s leadership. While the Conservatives increased their attractiveness to 73% among Leave voters – (from about 60% (Lord Ashcroft 2019), or 65% (McDonnell and Curtis 2019) in 2017) – Labour lost votes among ‘Remainers’, falling short of 50%. As the lead of the Conservatives over Labour among Leavers exceeded 50 percentage points in both middle- and working-class voters, Evans and Mellon argue that by 2019 ‘being middle or working class no longer mattered, while Brexit did’ (2020: 2). Cutts et al. (2020: 21) capture this dynamic very aptly: the surge of Leave voters may be the final element which cracked the brickwork, but the decay, reflected in Labour’s weakening relationship with working class Britain, had been setting in over much of the last decade, if not earlier. The 2019 general election was not a critical election that signalled a radical departure from the past, but rather, marked the continuation of longer-term trends in the dealignment and realignment of political loyalties in British politics.
To conclude, Brexit is changing the electoral/party system dimension in very important ways. After being triggered by the rise of UKIP, Brexit reunited the Conservatives behind the Leave camp, thus enabling them to further attract many working-class voters and conquer a strong majority in Parliament. Until 1970, the party system of the classic Westminster model stood on a solid ground of high partisanship, turnout and class fidelity, which is very unlikely to return anytime soon. However, while the overall high volatility means that any conclusion must be provisional, important elements of the two-party game are nonetheless now back in town.
Executive–legislative relations: The constrained dominance of the executive
In the classic Westminster model, single-party governments command a cohesive majority in parliament, ensuring that legislation – which they introduce, controlling the parliamentary agenda – gets smoothly approved. For Anthony King, the only tools available to the parliamentary opposition are ‘good reasons, and time’ (cited in Russell and Cowley, 2018: 19). This stands in stark contrast with the more balanced executive–legislative relationship of consensus models, where cross-party agreement is the norm.
Lijphart (1999) uses average cabinet duration as a proxy for executive dominance. He also considers the share of minimal winning and one-party cabinets. Both measures, which belong to his executive–party dimension, are of limited value when studying a single case. To present an overall picture of executive–legislative relations in the UK, we thus rely – following Matthews and Flinders (2017) – on the index of opportunity for opposition influence (Powell, 2000: 103–109). The index considers two factors within the legislature enhancing the opposition’s clout: the political strength of the government and the nature of the committee system. Its lowest value is for majority government and a weak committee system (0.1), while the highest score is for minority government and a distribution of committee chairs among all large parliamentary parties (0.75). In the first scenario, describing the classic Westminster model, ‘opposition party representatives may do little more than use the legislature as a forum for mobilising public opinion’ (Powell, 2000: 103). Classically, this situation characterised countries like New Zealand and Britain (Powell, 2000: 106).
We complemented this index with other indicators of parliamentary influence, such as the government’s defeats in parliament and backbenchers’ agenda-setting capacity. In addition, we also focused on the House of Lords. Lijphart included his variable, bicameralism (V7), in the federal–unitary dimension, being strongly associated with decentralisation or federalism. Here, instead, we consider the House of Lords as an additional constraint on the government.
We placed our empirical focus on some critical cases, where the institutional strains between the executive and the legislature have prominently come to light. We selected key legislation and votes on Brexit files. Not only were they likely to provide the strongest evidence of any Brexit effect, but such files characterised the 2017–2019 parliamentary session, lasting (unusually) 2 years in order to get the Brexit legislation through, in what came to be known as the ‘Brexit Parliament’ (cf. Lynch et al., 2019: 54). While the majority of parliamentary activity was focused on activities other than Brexit, it has been shown that, beyond Brexit legislation, the government dealt with specific areas of policy rather than broader reforms and major bills (Institute for Government, 2020).
Executive–legislative relations have changed significantly in the last few decades or so and, pace Lijphart, there is robust empirical evidence to assert that parliament has consolidated its position vis-à-vis the executive. Parliament exercises ‘preventive influence’ and, albeit not formally controlling the agenda, does affect what the government can and cannot propose. Furthermore, while most of the amendments made to the bills introduced by the government fail, several substantive changes are agreed before the bill reaches the plenary (Russell and Gover, 2017). Governments face increasing rebellions from backbenchers, with the 2010–2015 parliament being the most rebellious ever (Cowley, 2015). Research has also shown that reforming the House of Lords has strengthened bicameralism (Russell, 2013) and that the system of select committees is visible and influential (Russell and Cowley, 2016: 130–132). Such recent changes are partly captured by Powell’s index of opportunity for opposition influence. As Figure 1 shows, the index remained close to its minimum value for decades. Barring some limited exceptions, post-WWII single-party majority government is the norm. The creation in 1979, and then the reform of the committee system under the coalition – notably the election of committee chairs – pushed the index up. Due to the latter reform, Matthews and Flinders (2017) assigned the score of 0.35 to the coalition. In the Brexit years, the index reached its highest value, due to the combination of minority governments (May II and Johnson I) and a robust committee system. 5
The influence of the parliamentary opposition (1945–2020).
Our assessment of the impact of the Brexit process on the executive–legislative relationship is made against this background. In the words of a former cabinet member, Brexit was designed to allow ‘parliament to take back control of UK laws and policies’ (Davis, 2017). However, from its very beginning, Brexit brought to the fore the difficult relations between government and parliament, as the executive did not consider the involvement of parliament necessary to start the process of withdrawal. This uncertainty about institutional competences had to be resolved by the courts, with the Supreme Court ruling that relying on the royal prerogative to start the withdrawal process was unlawful. However, the involvement of parliament did not stop the government from fast-tracking a 133-word bill giving the PM the power to notify the UK’s intention to withdraw, which parliament rapidly approved. This was an early confirmation of the hypothesis that the Brexit process put the executive in the driving seat, unfolding under tight deadlines, set by the executive itself or jointly with the EU, with a more reactive role for parliament (cf. Baldini et al., 2018).
The event that fundamentally changed the relations between government and parliament, weakening the former and strengthening the latter, was May’s surprising decision – encouraged by favourable opinion polls – to hold a snap election in June 2017. Her gamble backfired: with only 48.8% of the seats controlled by the Tories, the PM needed the ‘confidence and supply’ of the 10 Members of the Democratic Unionist Party (DUP) to form a minority government. The agreement with the DUP ensured a majority of six – and a working majority of 13 – which became progressively smaller as some Conservative Members of Parliament (MPs) quit the party.
A minority government in itself is a significant deviation from the Westminster model. It suffices to note that this was only the fourth minority government in post-WWII history: Harold Wilson led a minority government after the 1974 elections, while in two other cases – James Callaghan in 1977 and John Major in 1996 – the government majority was lost during the term. Yet, May was not only leading a minority government. Her party was also fundamentally divided over Brexit. Therefore, the two conditions that Lijphart identified as necessary for ‘strong cabinet leadership’ (1999: 12) were missing. As Table 2 shows, comparing different legislatures starting in 2005, a high proportion of significant rebellions – with 10 or more MPs voting against the Conservative whip – characterised the 2017–2019 parliament, boosting the number of defeats for the government (28 for May alone, all but one in her second year of minority government).
Backbenchers’ rebellions and government’s defeats in the Commons (2005–2019).
Sources: Slapin et al., (2018); Institute for Government (2020); The Public Whip (www.publicwhip.org.uk).
The first defeat of the May government in the House of Commons happened on 13 December 2017, when 12 Tory backbenchers supported an amendment, tabled by a Conservative MP, to give parliament a ‘meaningful vote’ on the withdrawal agreement. Yet, it was only after the UK and the EU had agreed on the text of the withdrawal agreement (28 November 2018), that the government lost its capacity to command a majority. On 4 December 2018, the government lost three times, including on an amendment to give parliament more control over Brexit. The vote featured 26 Tory rebels – the largest rebellion up to that moment – and cross-party support. After several other defeats, the government lost the ‘meaningful vote’ by an unprecedented margin (230 votes), the largest ever defeat for a British government in the democratic era. The next two votes on the deal (the latter on the withdrawal agreement only) were also lost, with a reduced majority of 149 and 58, respectively. As a result, Brexit was postponed well beyond the originally scheduled date of 29 March 2019, bringing May’s premiership to an end.
Without a majority in parliament, the government lost control of the agenda. When it became clear that there was no support in parliament for the withdrawal agreement, the House of Commons held a series of indicative votes on the shape of the future relationship with the EU. The amendment to hold indicative votes came (again) from a Conservative backbencher, had cross-party support and was backed by 29 Tory rebels. The indicative votes were cast on 27 March and 1 April 2019, but there was no majority for any of the options. While this attempt shows that the parliament was incapable of finding cross-party support for any alternative ‘Brexit’, the clear opposition to ‘no-deal’ made it possible for a Private Member Bill (PMB) – the EU Withdrawal (No. 5) Bill, sponsored by a Labour MP – to be approved in less than a week. 6
While the number of successful PMBs was not exceptional – 15 of them were approved in the 2017–2019 session (3.8%), while 8 (4.6%) in the shorter 2016–2017 session (Hansard Society, 2019) – it was noteworthy that backbenchers voted twice to take control of the House of Commons business and twice – the second time was in September 2019 – they forced the government to extend the deadline set by art. 50 (cf. Institute for Government, 2020: 43–44).
An additional constraint for the government came from the House of Lords. The lack of a Conservative majority in the upper house, coupled with the pro-EU attitude of most peers (including several Conservative peers), made it even more difficult for the government to pass legislation (Smith, 2019). A revealing case is that of the EU (Withdrawal) Bill. The bill received extensive scrutiny by the Lords – about 5 months passed between the first and the third reading – and the government lost 15 times. Eventually, the government accepted one of the Lords’ amendments and made concessions on eight of the remaining 14 (Institute for Government, 2018).
The most prominent concession was on the amendment prescribing that if either the House of Commons did not approve the deal with the EU, or the government did not meet the deadlines set by parliament, the executive would follow ‘any direction’ indicated by a parliamentary resolution. The executive did not accept being so constrained by parliament. A compromise was eventually found and, in the case of the deal being rejected, a minister would have made a statement in the House of Commons, with the latter voting on a motion ‘in neutral terms’ (emendable at the Speaker’s discretion). This case is worth mentioning because the Lords’ amendment had already been tabled in the House of Commons (Simson Caird, 2018), showing that the upper chamber cooperates closely with the lower chamber in what is, effectively and not only formally, a bicameral system (cf. Russell and Cowley, 2018).
The conflict between government and parliament reached its climax when Johnson, PM since July 2019, ‘prorogued’ parliament, which was not expected to meet from 9 September until 14 October. Johnson claimed that a fresh start was necessary, as the parliamentary session had lasted far too long. For the opposition, such a lengthy prorogation was deemed unnecessary, and was being instrumentally used by the PM to avoid scrutiny on his Brexit agenda. The issue escalated to the Supreme Court, which found prorogation unlawful as it effectively curtailed ‘without reasonable justification, the power of parliament to carry out its constitutional functions as a legislature and as the body responsible for the supervision of the executive’ (The Supreme Court, 2019). The Supreme Court, as it had done at the start of the Brexit process, protected the constitutional role of parliament.
From the 2017 general election up to November 2019, the Conservative minority government, weakened by continuous backbenchers’ rebellions and occasional cross-party alliances within and across chambers, failed to implement its Brexit agenda. This was the case notwithstanding the successful renegotiation of the withdrawal agreement with the EU in October, which included a different arrangement for Northern Ireland opposed by the DUP.
Yet, with the snap elections called by Johnson in December, the Tories took control of the 56.1% of the Commons’ seats, holding a comfortable majority of 80. In addition, Johnson controlled a relatively cohesive parliamentary party on the EU, having deselected most Europhile MPs who had previously rebelled against his government. With such a majority, the PM was back in the driving seat and the vote on the withdrawal agreement opened the path to Brexit, finally happening on 31 January 2020. As former Chancellor Philip Hammond – one of the most prominent Tory rebels losing the whip – put it: ‘all of the things that we’ve got very excited about . . . will be completely irrelevant, because a government with a majority can decide everything’ (Guerrera, 2019).
With a substantial majority and an overall cohesive party, it may be argued that the conditions for a power-hoarding executive are back in place, at least as long as Brexit remains an important issue in the agenda. However, parliament is now structurally stronger than it used to be – for example, through the system of standing committees – and, whatever the government majority, executive–legislative relations cannot be ‘back’ to the heyday of the Westminster model.
The federal–unitary dimension: Brexit as a clarifier
Matthews and Flinders (2017) argued that devolution reforms moved the UK away from the ‘centralist and unitary’ edge of Lijphart’s index, pushing it towards an intermediate position between the ‘unitary and decentralised’ and ‘semi-federal’ categories, which they called ‘quasi-federal’. Their evaluation was based on the considerable depth and impressive pace of decentralisation reforms that allowed three UK regions (Northern Ireland, Scotland and Wales) and, to a lesser extent, local authorities, to choose their own policies in certain areas. However, they did not give due consideration to the absence of formal federal arrangements, especially in terms of co-decision powers in UK legislation and veto powers on constitutional matters (McEwen, 2018). Our analysis shows that some ambiguities of the devolution settlement and the diverging interpretations of the UK territorial constitution that they permitted have been clarified by key events, including sentences of the Supreme Court, which have substantively restated the unitary (albeit increasingly decentralised) nature of the UK. Indeed, from the point of view of regional governments, in the face of the Brexit test ‘the structures of devolution have failed’ (Michael Russell, Scottish Minister for Brexit, reported by BBC News 7 ); as a consequence, the devolved administrations are now ‘significantly more vulnerable than before’ (Carwyn Jones, First Minister of Wales, reported in Paun, 2018). In our view, this justifies the ‘demotion’ of the UK to the ‘unitary and decentralised’ category of Lijphart’s federal–unitary scale.
Leaving aside the fact that the way in which the Brexit referendum was organised and the way in which the ‘will of the British people’ has been interpreted are typical of unitary states, the federal–unitary dimension is here evaluated through the analysis of three post-referendum key cases: (i) the judicial fight over the right of the UK executive to trigger art. 50 of the Treaty of the EU without consent from the regional legislatures; (ii) the clash over the assignment of repatriated powers (Clause 11 of the EU Withdrawal Act 2018 and the Scotland Continuity Bill); and (iii) the row over a second independence referendum in Scotland.
As a former Remainer called to lead a Brexit government, Theresa May was very eager to show resoluteness in following the ‘will of the British people’, as expressed in the referendum. This led to a judicial case on the necessity of a parliamentary act to trigger art. 50, which was resolved in favour of Parliament by a decision of the High Court in November 2016. When the UK government appealed against this decision in what came to be known as the Miller case, the devolved administrations jumped on the quarrel, claiming that, on the basis of the so-called Sewel Convention and the Belfast Agreement, 8 the UK government also needed consent from regional legislatures. The core of the argument brought forward by the legal representatives of the devolved administrations was that triggering art. 50 had direct consequences on many devolved policy areas and, therefore, a decision made by the central government without their consent would be unconstitutional (McHarg and Mitchell, 2017). However, on 24 January 2017, the Supreme Court pronounced that the UK executive only needed an act of Westminster. The court stated that the Sewel Convention remains a political (gentlemen’s-) agreement, not a constitutional provision. 9 As such, Westminster retains the power to evaluate whether the circumstances under which legislation is taking place are to be considered ‘normal’, in which case the Sewel Convention should be respected, or exceptional, thus justifying London intervention in devolved policy areas without consent. The inevitable corollary of this judicial sentence is that the regions have no legal powers to stop Westminster from stepping into their legislative competences.
The second case concerns the assignment of repatriated competences. The EU (Withdrawal) Act – announced by the government in late June 2017 and approved in Westminster a year later – was designed to import all EU legislation into UK domestic law under the heading ‘retained EU law’, with a view to providing legal continuity in the immediate aftermaths of exit. Clause 11 of the bill prescribed that 24 devolved competences that were to be repatriated from Brussels would be temporarily allocated (for up to 7 years) to Westminster. This was justified with the need to put in place temporary common frameworks that would protect the UK internal market from legislative and regulatory divergence across UK regions. However, the devolved administrations evaluated these provisions as totally disproportionate vis-à-vis the stated objectives. The first ministers of Scotland and Wales reacted by denouncing Clause 11 as a ‘naked power grab’ and ‘an attack on the founding principles of devolution’. 10 The intense talks amongst the UK, Scottish and Welsh executives, 11 which occurred within the Joint Ministerial Committee for EU Negotiations (JMC-EN), appeared to have come to a deadlock in late February 2018. The Scottish and Welsh legislatures reacted by passing their respective Continuity Bills on 21 March 2018, aiming to secure all repatriated powers falling under devolved competences. However, in May 2018, Scotland and Wales took different routes. Wales, amidst minor concessions, reached a deal with London and gave its consent to the EU Withdrawal Bill (with only the opposition of the Welsh nationalist party, Plaid Cymru). In contrast, Scotland, whose Continuity Bill had been challenged by the UK government before the Supreme Court, denied its consent. However, this did not stop Westminster from passing the EU Withdrawal Act on 20 June 2018. Once again, institutional clashes between London and Edinburgh were settled by the Supreme Court. The sentence was given in December 2018 and, perhaps unsurprisingly, it declared the pre-eminence of Westminster. The substance of the ruling was that the EU (Withdrawal) Act 2018 is protected against modification and overrides the contrasting provisions contained in the Scotland Continuity Bill (The Supreme Court, 2018).
The third case concerns the political backlashes created by the aforementioned constitutional clarifications, which appear to be particularly indigestible for the SNP. The Scottish government is also denouncing an unfairly different treatment between Scotland and Northern Ireland. In spite of the fact that Scotland had voted against Brexit more strongly and more uniformly than Northern Ireland, the Brexit Agreement between the UK and the EU prescribes that Scotland will have to leave the EU’s custom union and common market with the rest of Britain, while Northern Ireland will remain in the EU’s common market of goods and an entry point in the EU’s custom union. This has contributed to accelerating the resolve of the Scottish government to try and arrange a second independence referendum. To be sure, the Scottish First Minister (backed by the Scottish Parliament) had already issued a request to Theresa May in March 2017. However, at that time there were still hopes that Brexit could be stopped, leading to a generic request for a referendum to be held within the next 24 months. In contrast, after renegotiation of the Brexit Agreement in October 2019 and the victory of the Conservative Party in the December 2019 general election, the Scottish government came back with an extremely ambitious set of requests, including not only permission to hold another binding referendum in 2020 (under the constitutional status quo) but also the devolution of new powers to the Scottish Parliament, with a view to holding future independence referendums and co-legislating (in co-ordination with Westminster) to implement Scotland’s independence from the UK (Scottish Government, 2019). On the one hand, this move, whose consequences are difficult to predict, represents an open challenge to the constitutional status quo. In this case, the flexible/political nature of the UK constitution might work in favour of the SNP, who envisage exercising pressure on London by strengthening their already dominant electoral position in Scotland. In this respect, the results of the forthcoming 2021 Scottish election might prove crucial. On the other hand, for the time being, this episode represents a further confirmation of the unitary character of the UK, as the question was formally closed by the refusal of the new British PM to grant permission for a second referendum.
To conclude, the UK remains a decentralised system and, in time, the repatriation of devolved competences from Brussels will even strengthen the self-government powers of the regions. However, all the constitutional clashes created by the Brexit process have seen the Supreme Court upholding Westminster sovereignty. As such, Brexit has worked as a constitutional clarifier that has substantiated the unitary character of the UK and, above all, of Britain. Indeed, Brexit has also re-marked a qualitative difference between Northern Ireland, whose status is now accepted to be an international (UK–Ireland–EU) matter, and Scotland and Wales, which are only subject to the UK constitution and, therefore, to Westminster (see McEwen and Murphy, this issue). Predictably, this constitutional clarification has triggered political backlashes, starting from the will of the Scottish government to hold a second independence referendum. Ultimately, the fate of Scotland will be decided more by the struggle for winning the ‘hearts and minds’ of the Scottish people than by constitutional norms. However, for the time being, the Brexit process has reaffirmed the irrelevance of Scotland (and Wales) on constitutional matters.
Conclusions: Westminster transformed and reloaded
During the chaotic Brexit process, from the June 2016 referendum to exit day, the British political system experienced a move back towards the Westminster model across the three dimensions analysed in this article.
It is all too evident that the persistence of a FPTP voting system has reduced fragmentation in the party system: despite the high turbulence of the 2019 European Parliament election, the 2017 and 2019 general elections strengthened the two-party vote. In 2019 a solid single-party parliamentary majority reappeared in British politics after 9 years of, respectively, hung Parliament (2010–2015), slim majorities (2015–2017) and (supported) minority governments (2017–2019). In addition, except for Labour, all other parties have come to terms with the centrality and polarising effect of the Brexit issue, adopting neat positions (hard Brexit vs. no Brexit) and homogenising their parliamentary groups in support of those positions. Yet, the prolonged persistence of high volatility leaves the two-party system in a somehow vulnerable position.
As expected, these changes in the party system have had a direct effect on executive–legislative relations. Not surprisingly, executive–legislative tensions have risen considerably during minority governments, due to attempts by the executive to compress the constitutional powers of Parliament and, conversely, MPs exploiting the difficulties of the government and seeking to assert their control over the parliamentary agenda. Traditional executive dominance reached its nadir during the Johnson I government. Yet, the electoral success of a relatively cohesive Eurosceptic Conservative Party in December 2019 changed the picture. Meanwhile, Parliament is now structurally better equipped to challenge the government.
On centre–periphery relations, the Brexit process has worked as a clarifier of the territorial constitution, stating the unitary, albeit decentralised, character of the UK. The rulings of the Supreme Court have clearly confirmed that Parliament can legislate on devolved matters and, consequently, that devolved administrations do not have the legal power to stop constitutional laws that affect their own competences. The Brexit process has triggered a strong reaction in Scotland, where the SNP is set to use the 2021 regional election to legitimise its request for a second independence referendum. However, whether an electoral legitimation will suffice to break the veto of the UK government remains a moot point.
Moving beyond the above three dimensions, the Brexit process – and particularly the judicial row over Parliament’s prorogation – has triggered a determination by the governing Conservative Party to reconsider the role of the Supreme Court. Should this institution be substantively diminished or suppressed, then a powerful pushback to the institutional changes of the last two decades would be inflicted. However, so far, the move back towards the Westminster model is clearly constrained, as Brexit has not cancelled the important constitutional reforms adopted since 1997.
Finally, this work has confirmed the heuristic validity of Lijphart’s conceptual map of democracy, which allows one to bring the complexity of the analysis back to a meaningful synthesis, capturing the position and the direction of movement of a political system along the majoritarian–consensus continuum. However, we have relied on a wealth of additional evidence that goes well beyond the numerical indicators deployed in macro-comparative analyses à la Lijphart. Therefore, this type of research retains a higher degree of complexity. Indeed, besides identifying the prevailing trend at specific points in time, single-case analyses can also highlight persisting strains and detect contradictory dynamics, keeping the political process open to different developments. While the most recent developments in the context of COVID-19 seem to point to a further consolidation of the executive, more long-term, structural changes that have transformed the classic Westminster model in the UK over the last few decades indicate the need to interpret them cautiously.
Footnotes
Acknowledgements
We would like to thank Theresa Reidy and the two anonymous reviewers for their constructive comments. Nicola Chelotti, Marco Giuliani, Richard Hayton and the audiences at the 2019 EUSA Biennial conference in Denver (10 May 2019), at the UACES 49th Annual Conference in Lisbon (1 September 2019) and at the 33rd annual congress of the Italian Political Science Association in Lecce (13 September 2019) also provided helpful advice on earlier drafts of this article. We are particularly thankful to Andrea Pareschi for his research assistantship and contribution to this work.
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
