Brexit and Parliament: The Anatomy of a Perfect Storm


 The Westminster parliament lies at the heart of UK democracy. Yet its role and powers became increasingly controversial during the ‘Brexit’ process, following the 2016 referendum decision to leave the European Union. Fierce arguments were framed as ‘parliament versus people’, which fed a populist narrative and raised fundamental questions about where UK sovereignty should lie. This article charts the stages of parliament’s Brexit ‘perfect storm’, tracing its causes to four factors: the design of the referendum, a period of (unfamiliar) minority government, deeply divided political parties, and the weakness of parliamentary rules in facilitating a solution. In the end, the Brexit argument was primarily one inside the Conservative Party, but parliament got the blame.

The UK is famously the most parliamentary of democracies. Its unique status is based not just on being a 'parliamentary democracy' in the typical sense, of the executive depending on parliamentary confidence for its survival (Sartori, 1994), but beyond that on the central UK constitutional tradition of 'parliamentary sovereignty' (Goldsworthy, 2001). This makes parliament the highest authority, with no written constitution to bind its powers.
It was thus highly unusual, and potentially highly destabilising, that parliament's role came to be intensely contested during the UK's recent crisis over 'Brexit'. Following a narrow referendum vote in June 2016 that the UK should leave the European Union (EU) (secured by 52% of votes to 48%), parliament became the forum for increasingly bitter policy battles, leaving the executive and legislature divided in a way most uncharacteristic of parliamentary systems. Throughout this process, the correct site of constitutional sovereignty became widely disputed. Flashpoints included repeated defeats of the government's Brexit proposals in the House of Commons, and fierce procedural wrangles over control of the parliamentary agenda. Rarely used parliamentary tools, such as emergency debates, or those normally seen as 'toothless', such as private members' bills, were deployed against the government in anger-to significant effect. Amidst these bitter disputes, one Prime Minister (Theresa May) was forced to resign, and her successor (Boris Johnson) attempted to shut down parliament using the prerogative power of prorogation-only for his direction to the Queen to be overturned by the UK Supreme Court. When members of parliament (MPs) reconvened following the court's intervention, anti-parliamentary rhetoric reached new levels, with the Attorney General controversially suggesting that 'This Parliament is a dead Parliament . . . [that] has no moral right to sit'. 1 Shortly afterwards, the UK embarked on a general election (the first December election since 1923), marked by a rhetoric of 'parliament versus people'. Its result was a landslide win for the same Prime Minister that had unlawfully shut parliament down.
This article explores how such a strange and unsettling period developed, and what it teaches us about the place of parliament in the UK system of government-to date, and potentially in the future. It suggests that four key political and constitutional features, all of them unusual in a UK context, contributed to creating this 'perfect storm'. The four features-the referendum, a period of minority government, deep divisions within the governing party, and parliament's own mechanisms for handling a minority situation-are used to structure the main body of the article. The article ends by reflecting on where this leaves the UK-with the 'perfect storm' having fuelled an unprecedented level of populist rhetoric and public hostility to parliament. It suggests that the UK system, including its parliament, was ill-adapted to accommodate key recent changes. Overall, the behaviour of both prime ministers-in the context of a deeply divided Conservative Party-greatly fuelled the situation. In the process, the divisions over Brexit raised increasingly existential questions about sovereignty in the UK system, which remain raw and unresolved.

The referendum
The starting point for parliament's battles over Brexit was obviously the referendum, held on 23 June 2016-which itself had a substantially longer history. Referendums are relatively unfamiliar in UK politics, though their use has gradually grown (Independent Commission on Referendums, 2018). These 'direct democracy' mechanisms clearly sit somewhat awkwardly alongside the UK's tradition of representative democracy, introducing an element of popular sovereignty that could potentially conflict with parliamentary sovereignty. The Brexit referendum, as this section explores, had some unusual features that significantly exacerbated such conflicts.
The UK first joined the (then) European Community in 1973, via parliamentary decision rather than referendum. A public poll was later held in 1975 which strongly supported continued membership. In subsequent years, gradually greater EU integration was coupled with increasingly vocal 'Euroscepticism', and regular demands for referendums on the transfer of new powers to EU institutionsironically often framed as seeking to defend parliamentary sovereignty (Gifford and Wellings, 2017). From its emergence in the 1990s, the UK Independence Party, which favoured EU withdrawal, placed increasing pressure on the Conservative Party-exacerbating the latter's long-running internal splits over the EU. Eventually, under pressure from his Eurosceptic backbenchers, Prime Minister David Cameron committed to holding an 'in-out' referendum (Smith, 2017).
Although professing a degree of Euroscepticism, Cameron went into the referendum firmly favouring continued EU membership and was highly visible in the 'Remain' campaign. This was the referendum's first unusual feature, in both domestic and international terms. Recent prior referendums in UK politics, most notably on devolution in the 1990s, had offered the public a change strongly advocated by ministers. 2 In contrast, in bringing forward legislation to provide for the 2016 poll, the government's official position favoured continued EU membership. Ministers had merely conceded the principle that 'the decision . . . should be taken by the British people'. 3 Labour and the Liberal Democrats accepted this, despite also both being firmly committed to continued membership of the bloc. On 9 June 2015, the EU (Referendum) Bill thus passed its Commons second reading by 544 votes to 53.
A referendum was thereby held on a proposition for change anathema to both the Prime Minister and the majority of MPs. Had the question of EU membership been put directly to the House of Commons, it would almost certainly have had strong support-including among Conservatives (Heppell et al., 2017). Although collective Cabinet responsibility was lifted for the campaign, allowing individual ministers to support either side, the government's official position was 2 As indicated below, the 2011 referendum on the voting system was more complex, as the lead Minister Nick Clegg (Deputy Prime Minister and Leader of the Liberal Democrats) supported the change and campaigned for it, while the party's coalition partner, the Conservatives, opposed change. In addition, the 2014 referendum on Scottish independence presented a proposition that was opposed by the UK government, but which was strongly supported by the Scottish National Party government in Scotland.

Brexit and Parliament
to back 'Remain'. Rather than promoting change, the referendum's principal purpose was thus to clear the air and end a lingering political argument. After the result, the House of Commons Public Administration and Constitutional Affairs Committee (2017, p. 12), though chaired by a leading Brexit supporter, was highly critical of this staging of what it dubbed a 'bluff call' referendum, where government 'initiated the process . . . despite being against the suggested proposal, and with the aim of using a negative result to shut down the debate about the question'. The decisions of both the Prime Minister and parliament had clearly been influenced by consistent polling prior to the referendum which indicated that only a minority of the public supported leaving the EU (Curtice, 2016).
A second notable feature of the referendum was that it asked the public to decide on a broad principle, rather than a detailed manifesto for change. Of course, the terms of the UK's exit from the EU would by necessity be subject to negotiation with other Member States, so a fully agreed prospectus was impossible. However, the government's strong preference for the status quo meant that there was very little consideration of how the change option would work. As again identified by the Public Administration and Constitutional Affairs Committee (2017, p. 4), 'There was no proper planning for a Leave vote so the EU referendum opened up much new controversy'. The government offered no clear prospectus, and in the absence of this there was little strategic incentive for the Leave campaign to elaborate on its catchy slogan to 'take back control'. Issues which became central to the subsequent debates-most obviously how to manage the position of Northern Ireland, which shares a land border with the EU-were hence barely mentioned during the campaign.
These two characteristics were unusual not only in domestic, but also in international, terms. In previous UK referendums where ministers had supported change, its details were generally already clear. Prior to the previous UK-wide referendum in 2011, on changing the voting system for the House of Commons, there had been differences of view between the two government coalition partners. But the risk of arguments over implementation in the event of a 'Yes' vote was removed by parliament having legislated prior to the referendum to set out fully how the new system would work (ultimately the public voted 'No'). This approach was consistent with that typically used internationally for constitutional referendums. Such votes are often required as part of the constitutional reform process, but this is generally a multi-stage process, where the legislature, as well as the voters, must endorse the change (Independent Commission on Referendums, 2018). Additionally, of course, in most places, constitutional change comprises a clearly defined amendment to a constitutional document. Hence both its precise wording and the in-principle support of politicians are already in place before the public are presented with a change proposition. Instead, with Brexit, the 'new controversy' opened up by the referendum result was left for others to resolve. The decision had, in effect, been constructed as a two-part question: first, whether the UK should leave the EU, and then, if so, on what terms. The public had answered the first part of the question and it unexpectedly became the job of parliament and government (alongside EU partners) to answer the second.
Before this second part of the question could be addressed, post-referendum controversy emerged over sovereignty, as applied to the first part. Alongside providing no detail on implementation, the legislation had contained no explicit legal instruction on how ministers should act immediately following a Leave vote. The new Prime Minister Theresa May, who rapidly replaced David Cameron following his resignation on the day after the referendum result, took the outcome as politically, even if not legally, binding. She sought to act on public instruction, by using the executive's prerogative powers to trigger the Article 50 provisions in the Treaty of European Union (which allow for departure of a Member State), thereby adopting an implicit logic of popular sovereignty. There was relatively little resistance to this from within parliament, but a legal challenge against the government did ultimately end in the Supreme Court. 4 The court concluded that if ministers triggered Article 50 without statutory approval that would breach the core principle of parliamentary sovereignty, as it implied repeal of a prior statute (the European Communities Act 1972). A new statute was hence required, bringing parliament into the process.
This episode sparked one of the most notorious rhetorical responses of the period, when on 3 November 2016 the Daily Mail front page painted three High Court justices as 'enemies of the people' for deciding that Brexit should be handed back to parliament. The clear expectation of some on both sides of the Brexit divide was that MPs might now overturn the result. The ruling hence presented parliamentarians with a difficult dilemma, forcing them to decide very explicitly whether or not they had ceded their sovereignty to the public.
The government responded by publishing the EU (Notification of Withdrawal) Bill-a one-clause measure authorising the Prime Minister formally to request the triggering of Article 50. Debate at the bill's Commons second reading was agonised, extending over two days with 189 speakers. 5 Many MPs expressed their own misgivings about the referendum result, but indicated that they felt it must be honoured. Some, representing constituencies where the balance of votes had favoured remaining in the EU, argued that they should be guided by constituency rather than national opinion, and voted against. Brexit and Parliament handful did the same notwithstanding both constituency and national opinion, preferring to follow their own judgement. The debate was marked, unusually, by numerous references to Edmund Burke-whose notions of 'delegate' and 'trustee' neatly reflected MPs' representational dilemmas (Judge, 2005). But both main parties whipped in favour of the bill, and it comfortably passed its second reading by 498 votes to 114. 6 No amendments were subsequently agreed in the Commons, and changes proposed by the House of Lords were easily overturned by MPs (after which peers immediately backed down). Nonetheless, in response to pressures from both chambers, ministers gave assurances that parliament would get a vote on the deal finally agreed with the EU.
Overall, therefore, the clear reservations of many parliamentarians about the Brexit decision were overshadowed by the referendum result, with most MPs on both the government and opposition side accepting that it must be respected. Parliament hence actively, if reluctantly, ceded its sovereignty to the public on the principle of Brexit. But the detail of how to implement the decision had yet to be agreed. While the executive had initially sought to act directly on public instruction, the principle of parliamentary sovereignty was somewhat uneasily restored-including through parliament's right to input on this next crucial part of the question.

Minority government
The tensions created by the unusual referendum were significantly exacerbated by the advent of minority government post-2017. While referendums are historically uncommon in UK politics, this occurrence was even less familiar and highly consequential.
The Conservatives had won the 2015 general election only narrowly, which clearly presented a potential challenge for delivering the unexpected Brexit result-although, as indicated above, Theresa May faced little immediate parliamentary resistance. Nonetheless, having triggered Article 50, she moved straight to proposing an early general election, aimed at boosting her parliamentary majority. This aspiration seemed, at the time, well justified by the polls (Cowley and Kavanagh, 2018). May had previously repeatedly denied that she would take such a route, so was concerned about how to justify her change of position (Seldon and Newell, 2019). She chose to cite difficulties in parliament. Speaking outside 10 Downing Street, May suggested that 'At this moment of enormous national significance there should be unity here in Westminster, but instead there is division' (BBC, 2017). She argued that 'In recent weeks Labour has threatened to 6 Those voting against the second reading included 50 Scottish National Party members, 47 Labour, 7 Liberal Democrats and just 1 Conservative. vote against the final agreement we reach with the European Union' and that 'unelected members of the House of Lords have vowed to fight us every step of the way.' The last of these statements had no clear basis in fact-the Lords having just completely capitulated over her bill. Targeting the opposition, meanwhile, was obvious election talk-but masked her primary problem. Opposition parties fairly routinely vote against government proposals; this remains unimportant, so long as the government has solid support from its own MPs (King, 1976). The unspoken challenge was thus whether May could hold together her own deeply divided parliamentary party to support a Brexit deal.
As is now well known, the Prime Minister's 2017 general election gamble backfired. Labour performed better than expected, and the result was a hung parliament and minority Conservative government. May agreed a 'confidence and supply' arrangement with the Northern Ireland Democratic Unionist Party (DUP), with whose support she would enjoy a notional Commons working majority of 13. In theory, this agreement extended to Brexit matters.
There had previously been just three brief post-war periods of minority government in the UK, only one of which directly followed a general election, in February 1974-and led to another election eight months later. While minority governments are relatively common internationally, their culture and the necessary strategies for governing were hence unfamiliar in both Whitehall and Westminster.
In anticipation of the hung parliament result in 2010, a report from the Constitution Unit and Institute for Government had concluded that a 'key lesson' for the UK from historical and international practice was 'that when minority prime ministers seek to govern as if they had a majority the result is instability, partisanship . . . and likely failure. By contrast, minority administrations which adopt a more consensual approach, negotiating and making concessions with opponents . . . are more likely to remain in office' (Hazell et al., 2009, p. 7). This is in line with the comparative literature, which emphasises the importance of cross-party bargaining and concessions (Strøm, 1990). Crucially, minority governments will normally steer clear of controversial policies likely to inflame their opponents, and avoid legislating on matters where they lack parliamentary majority support. But the referendum decision denied such options to Theresa May. Instead the two-year period specified in Article 50 now gave her a clear timetable to complete the Brexit process by 29 March 2019.
Westminster is typically seen as a classic 'majoritarian' institution (Lijphart, 2012). This can feed a myth that British politics is an 'elective dictatorship', where the government is guaranteed its way in parliament. But such views have always been simplistic; even under majority government, far more negotiating goes on behind the scenes within the governing party over policy than is often appreciated (Russell and Gover, 2017). Nonetheless key executive actors may fall for the myth, and slip into assuming that ministers can force their policy through. To do so in times of majority government is risky, but under minority government, it becomes positively reckless-as compromise with parliament becomes essential.
The post-2017 situation would have presented a challenge to any Prime Minister. But classically both 'structure and agency'-that is, political constraints and individual personality-matter to leadership success (Byrne and Theakston, 2019). Theresa May's attitude to parliament as Home Secretary had been that 'she wouldn't make any concessions on anything'-strongly suggesting that she had bought the 'elective dictatorship' myth. 7 The first book-length analysis of her premiership concluded that she was 'Renowned for her intransigence' and 'Renowned for her tribalism' (Seldon and Newell, 2019, p. 601). Yet, as Hazell et al. (2009, p. 70) concluded, prime ministerial 'personality and leadership style' are crucial to the success of minority government, and 'An obdurate [Prime Minister] with a confrontational style is likely to alienate the opposition and have greater difficulty in passing the government's legislative agenda'. This could almost have been written with Theresa May in mind. Her political style was fundamentally unsuited to minority government and, precisely as these authors warned against, she sought to govern as if she had an assured majority.
An added source of inflexibility came from the 'red lines' which May had set out prior to the 2017 election. During the Conservative leadership contest to succeed Cameron, she coined the seemingly meaningless phrase 'Brexit means Brexit'-designed to reassure 'hard' Eurosceptics on her backbenches, given her own Remain position during the referendum. Her 'Lancaster House' speech of January 2017 then went further, setting out negotiating objectives that included extracting the UK from the EU Single Market and Customs Union (Smith, 2017). These positions would make a cross-party deal all but impossible. Labour's manifesto had promised to respect the referendum result, but supported a 'close new relationship with the EU', proposing 'fresh negotiating priorities that have a strong emphasis on retaining the benefits of the Single Market and the Customs Union' (Labour Party, 2017, p. 24). Had the Prime Minister pursued a bipartisan approach, this might potentially have offered both a comfortable Brexit majority in the Commons, and the quashing of opposition criticism. But it would have required changes to her negotiating objectives, and made splits in her own parliamentary party a certainty. In the end, by failing to appeal to other parties beyond the pro-Brexit DUP, May remained captive to her own Eurosceptic wing-adopting an even more rigid and confrontational style that would be typical under majority government.
Only in spring 2019, in the dying days of her period as Prime Minister, after being defeated on her Brexit deal in the House of Commons three times, did May open formal negotiations with Labour. Simultaneously, she softened her rhetoric to admit 'that delivering Brexit "involves difficult choices for all of us"' (quoted in Brusenbauch Meislova, 2019, p. 3)-a message which she consistently failed to deliver earlier and which might have helped build a more constructive atmosphere both inside and outside parliament. By this point, her position inside her own Cabinet was fragile. Labour, having been locked out of discussions and facing its own internal divisions (see below), had shifted to adopt an increasingly Brexit-sceptical position focused on a further referendum. Hence battle lines had become too firmly drawn to be dismantled. The cross-party talks made little progress, and May's Conservative colleagues forced her eventual resignation.
Returning to sovereignty, minority government significantly exacerbated the tensions that had occurred at the previous stage. May continued to try and harness executive power in order ostensibly to respect popular sovereignty. But with no clear mandate for the form that Brexit should take, and no solid Commons majority, she failed to acknowledge the constraints of parliamentary sovereignty-creating a growing schism between government and the majority in parliament. As discussed further below, her tribalism, and consistent failure to articulate the problems caused by her own backbenchers, instead led her to lay blame on an undifferentiated collective 'parliament'. These tendencies became more pronounced as the crisis deepened, particularly in the endgame under Boris Johnson.

Divided parties
The legacy of the referendum, coupled with minority government, already guaranteed a difficult parliamentary situation. But circumstances were made far worse by the divided nature of the parliamentary parties-particularly, as already emphasised, on the government side. Had Conservative MPs voted cohesively, and the DUP supported May's Brexit deal, it would have narrowly passed the House of Commons. Instead, it was overwhelmingly defeated on two occasions: by 432 votes to 202 on 15 January 2019 and by 391 to 242 on 12 March. A further attempt, on the original 'Brexit day' of 29 March 2019, narrowed the margin of defeat-but still resulted in a government loss by 344 votes to 286. May's strategy of relying on her own backbenchers failed badly. In fact, the extent to which this was a root of the difficulties was significantly masked by expectations of instability under minority government.
Unlike the previous two factors, there is nothing very unusual about party dissent in the House of Commons-but the Conservative Party's splits over Europe were particularly long lasting and deep (Lynch and Whitaker, 2013). In the first parliamentary vote, 118 Conservative MPs defied the whip, as did 75 in the second vote. On the third occasion, the number of rebels had dropped to 34. Of these, 28, rather than being opponents of Brexit, were hard-line Eurosceptics.
May's primary adversaries were the increasingly well-organised European Research Group (ERG). It operated close to a party-within-a-party and claimed support of 60-80 Conservative MPs. Without cross-party cooperation, their backing was clearly essential. But the constraints of the negotiations with the EU27 made this challenging. The question of Northern Ireland became central, with a determination to avoid a 'hard border' on the island of Ireland. May's DUP allies fundamentally opposed any customs border down the Irish Sea, which resulted in her agreeing to the so-called 'backstop'. This would keep the whole UK in a customs union with the EU until alternative arrangements could be found (Wright, 2019) and was hence unacceptable to hard-line Eurosceptics. Several Brexit-supporting ministers departed her Cabinet in protest, starting with David Davis and Boris Johnson in July 2018. At this stage, it should already have been clear that Conservative backbench Eurosceptics would not be assuaged, and that without support from opposition MPs there would be no Commons majority. In the end, even the DUP opposed May's deal.
Divisions inside Labour also complicated matters. The party's leader, Jeremy Corbyn, had been elected unexpectedly in a membership ballot in 2015 despite lacking support from the parliamentary party (Heppell et al., 2018). He was more Eurosceptic than the great majority of his party-which in one respect could have made him a viable negotiating partner for May; but he was otherwise unattractive, as a staunch left-winger who lacked trust from many on his own side. Labour MPs generally felt bound by the Leave result-a handful representing Leave-voting areas even rebelling to support the government. But most preferred some kind of compromise, either through a relatively 'soft' Brexit or subjecting the deal to approval in a further referendum. Hence views in the Parliamentary Labour Party were fragmented, and the leader could not reliably speak on its behalf.
The standard logic at Westminster, as in other contemporary parliaments, is that governments build parliamentary majorities through agreements between stable party blocks. Had the Conservative block remained stable, and the DUP partnership held, Theresa May's deal could have been approved. When it became clear that this would fail, some other majority needed to be found. But May stuck rigidly to her tribal approach, while Corbyn was both a difficult and an unattractive partner. With splits in both parties, questions increasingly grew over whether other parliamentarians could build an alternative majority.

The failure of parliamentary rules
The preceding three factors (or even just two of them) were sufficient to generate a chaotic parliamentary situation. They were further exacerbated by the inability of the Commons' rules to facilitate a solution-either through allowing an alternative majority to be found, or perhaps (though this is more debatable) through allowing the government to force its MPs into line. Which of these routes the rules should have facilitated was contested, reflecting different interpretations of parliamentary sovereignty itself. But either way, the rules failed to alleviate the situation.
In some respects, it could be argued that parliament rose well to the Brexit challenge. In particular, it provided an essential forum of accountability for May and her ministers, through regular statements and questions about the negotiations and government preparedness, and through the Commons Exiting the EU Select Committee (which was created to shadow the government department of the same name). Innovative means were used to extract information from ministers, notably through revival of the 'humble address' procedure (Defty, 2018). But as time wore on, the degree to which a recalcitrant minority government controlled the agenda of the House of Commons, and hence the ability of MPs to reach a decision point, became a problem.
In comparative terms, the Commons agenda is seen as strongly government controlled (Döring, 2001). That control is manifested in various ways, most obviously through Standing Order no. 14 (SO14). This states that 'Save as provided in this order, government business shall have precedence at every sitting', and then lists various exceptions (whose timing in practice is also government controlled), including a limited number of days for opposition or backbench debates, and private members' bills. Pre-Brexit, a 2009 report from the Select Committee on the Reform of the House of Commons (colloquially the 'Wright committee') had recommended various key changes-including establishment of a broadly representative 'House Business Committee' to decide the Commons' weekly programme, which would be put to MPs for a vote (Russell, 2011). This arrangement would have brought the Commons into closer line with other parliaments more often subject to minority or coalition government. Had it been implemented, the Brexit battles might have been resolved in a more controlled and constructive way. 8 However, under the standard setup of majority government pressures for change were limited, as the parliamentary majority was coterminous with the governing party (or parties), and hence this proposal was shelved. But the dynamic changed radically under minority government. 8 The Wright committee's proposals included backbenchers in membership of a House Business Committee, and crucially required its proposals to be agreed by the chamber, in an explicit attempt to prevent agenda control simply being a 'stitch up' between party leaders. An alternative model, as for example, applies in Germany, is for parliamentary agenda time to be divided proportionally between party groups.
Concerns about the government's direction on the part of Remain or 'soft' Brexit supporting MPs began early on, following May's 'Brexit means Brexit' rhetoric. This led to defeat during passage of the EU (Withdrawal) Bill in 2017, on an amendment from former Conservative Attorney General Dominic Grieve, to enforce ministers' promise that MPs would get to vote on the final Brexit deal. Consequently, Section 13 of the bill as passed made ratification conditional on approval in a so-called 'meaningful vote' by MPs. In the end delays in the negotiations, due partly to continued Cabinet splits, meant no final agreement with the EU emerged for approval until late November 2018. To meet the Section 13 requirements, the government proposed five days of Commons debate in December, to be followed by a vote. But the vote was postponed by ministers after just three days' debate, as it was clear that defeat was imminent-which significantly angered MPs, and eventually led to the first heavy defeat on 15 January 2019. 9 By this point, the March deadline was looming, and the government apparently had no Plan B.
MPs hence increasingly sought innovative procedural solutions to avoid a 'no deal' exit, whereby the UK would depart the EU with no agreement in place. This included moderate backbench Conservatives and opposition MPs repeatedly seeking to 'seize the agenda', which in turn sparked anger among 'hard' Brexit supporters, some of whom were content to leave without a deal. It brought great controversy to the position of Speaker-with the incumbent (John Bercow) frequently required to decide which initiatives were permissible. Controversy over proper use of parliamentary procedure hence became increasingly heated, and politicised along Brexit lines.
Numerous backbench initiatives were attempted in this period. Votes against 'no deal' took place on 29 January and 13 March, effectively forcing the Prime Minister to seek an extension to Article 50. Once this had been agreed, and the deal had been rejected twice, the first successful attempt was made-in late March-to set aside SO14. On an amendment from senior Conservative backbencher Sir Oliver Letwin, MPs agreed to stage 'indicative votes' on a series of Brexit options on 27 March. But these became a fairly chaotic free-for-all, with all eight alternative propositions-including for a 'softer' Brexit or a new referendum-defeated. 10 This demonstrated starkly how difficult it is to organise a parliamentary majority without the usual apparatus of party groups (leadership, staffing and whips) to facilitate decisions, and in the absence of the usual stable 9 The Speaker made a short statement following the deferral of the vote, in which he referred to it as 'deeply discourteous' given that 164 members had already spoken in the debate (House of Commons Hansard, 10 December 2018, column 27). 10 Letwin repeated the exercise, using the same voting system and with very similar outcomes, on 1 April. voting blocs. The standard Westminster voting process (based on yes/no propositions) also contributed, as members could not express ordered preferences between the numerous options. 11 Had this been possible, a compromise around a 'soft' Brexit or a referendum might have been reached (Renwick, 2019). A government-sponsored attempt at achieving compromise using some of these methods could perhaps have succeeded at this point-but no such attempt was made.
Another tactic deployed by the increasingly emboldened anti-'no deal' alliance was to use private members' bills to force legally binding outcomes on the government. Such bills can normally readily be blocked by government whips, but rebel business motions were used to guarantee them time. The first successful such bill (the EU (Withdrawal) (No. 5) Bill) was sponsored by Labour's Yvette Cooper in April, to require a further Article 50 extension. This helped force Theresa May into agreeing a (second) new deadline of 31 October. The most dramatic such intervention came in early September, when May's successor Boris Johnson was made subject to a bill (the EU (Withdrawal) (No. 6) Bill) sponsored by Hilary Benn, Labour chair of the Exiting the EU select committee. This required a further extension request to 31 January if parliament had not agreed a deal by 19 October. Even more than most other initiatives the Benn Bill's passage stretched previous precedents for the use of Commons rules, being based on a rebel business motion brought forward and agreed during an emergency ('SO24') debate.
Many Brexit supporters were highly critical of these moves. Opponents argued that it was not parliament's role to take the place of government in setting such significant policy direction (e.g. Ekins and Laws, 2019). But government's persistent failure to broker a compromise, and the ticking clock, left many MPs feeling compelled to act. Ironically, key Eurosceptic opponents of these moves in parliament had themselves voted to block the government's policy, preferring a 'no deal' option-which the Prime Minister, as well as the parliamentary majority, was working to avoid.
These episodes raised fundamental questions about whether SO14's default of government agenda control was appropriate under minority government, or whether the more appropriate principle should be parliamentary majority control. This touched, once again, on key disputes over sovereignty-this time in terms of the interpretation of 'parliamentary' sovereignty itself. Under a 'liberal' view, parliamentary sovereignty means just that-that the ultimate authority should lie with parliament. However, an alternative 'Whitehall' view asserts that the true authority is the crown-in-parliament-meaning, in practice, ministers (Birch, 1964). Under the latter view, parliament should primarily be an electoral college for choosing and maintaining a government that sets policy, rather than setting policy itself. Crucially, under single-party majority government these interpretations can normally remain blurred; but minority government, coupled with large-scale backbench dissent, brought them into open conflict. Again, some claimed that government control was necessary in order to uphold popular sovereignty, and thereby that parliament should be sidelined. But the key argument was between forms of Brexit, on which the public mandate was unclear-while some purporting to support executive control had voted repeatedly against the government.
For those arguing the Whitehall view, the problem was perhaps not SO14 but instead the Fixed-term Parliaments Act 2011, which had unbalanced the executive-parliament relationship (for discussion see Norton, 2019). The Act removed the Prime Minister's unilateral power to call a general election-thereby boosting 'parliamentary' sovereignty in the purest sense. But it also removed the Prime Minister's ability to threaten recalcitrant backbenchers with an early election when they failed to fall into line. Whether the existence of this weapon could have forced support for Theresa May's Brexit deal is distinctly questionable, as Labour offered only a limited electoral threat; but it might perhaps have changed backbenchers' calculations. 12 Undoubtedly, under any conception of British politics, parliament's ultimate sanction is the ability to remove the government via a no-confidence vote. It is hence a final peculiarity of this period that, with a weak government failing on its central policy, no such vote occurred. The explanation rests on factors discussed above, including the Brexit timetable, moderate MPs' fear of 'no deal', and the absence of stable building blocks from which to create an alternative majority. Crucially, rejection of Jeremy Corbyn as a viable prime minister-by Conservative rebels, Liberal Democrats, and even some on his own side-was a clear obstacle when cross-party talks did finally take place among anti-'no deal' MPs about possibly installing an alternative 'unity' government (BBC, 2019). This was not until summer 2019, and any such prospect was soon-as discussed next-overtaken by other events.

'Parliament versus people'-the descent into populism
The combination of these four features-the constraints on MPs resulting from the (highly unusual) referendum, the unexpected minority government, the grave divisions on the Conservative benches and the shortcomings of parliamentary 12 Again, some of the harshest critics of the Fixed-term Parliaments Act were those who had voted against the government. This appears paradoxical, as it implies that with a tougher enforcement of the whip they themselves would have fallen into line. rules-fuelled an increasingly adversarial and ill-tempered environment both inside and outside parliament. This laid the ground for the start of Boris Johnson's premiership, beginning in July 2019 after May had been forced out, which was characterised by an increasingly overt populist rhetoric, setting 'parliament' against 'the people'.
Johnson had been a key player in the 2016 Leave campaign and departed May's Cabinet in protest at her Brexit proposals. He then voted against her deal twice, though backed it on the third occasion. Upon becoming Prime Minister, he appointed a strongly pro-Brexit Cabinet, including several ERG sympathisers. His administration was clearly even less likely than its predecessor to conform to the usual consensus-building expectations of minority government or to find common cause with Labour.
The timing of the Conservative leadership contest saw Johnson subject to just one day's parliamentary scrutiny before the summer recess-providing no opportunity for MPs to question his detailed Brexit plans. He openly entertained a 'no deal' Brexit during the contest, which helped motivate the cross-party summer discussions among rebel MPs. With parliament set to convene on 3 September, it was clear that this group intended either to pursue what became the 'Benn Act' (see above) or a no-confidence vote and alternative 'unity' government.
But on 28 August, it was announced that Johnson had sought a five-week parliamentary prorogation from the Queen, to run from mid-September to mid-October. Leadership contender Dominic Raab had floated use of this mechanism during the contest, as a means to prevent MPs from blocking a 'no deal'. Other contestants had condemned the idea, while Johnson had indicated that he was 'not attracted' to it (Honeycombe-Foster, 2019). By choosing to prorogue, Johnson severely curtailed opportunities for rebel activity; nonetheless, the 'Benn Act' to require a further Brexit extension was rushed through before parliament closed on 9 September. 13 A group of 21 moderate Conservative MPs who voted in support of the bill was summarily stripped of the whip.
There followed a second Supreme Court case, which judged the prorogation to have been unlawful. The Court's unanimous judgment rejected the 'Whitehall view' of parliamentary sovereignty, emphasising that 'The Government exists because it has the confidence of the House of Commons. It has no democratic legitimacy other than that'. 14 Hence, as the senior partner in the relationship with government, parliament could not simply be dispensed with by ministers. 13 Fearing a prorogation, rebels had inserted amendments in the Northern Ireland (Executive Formation, etc.) Bill in July to require some autumn sittings-without which the prorogation might have come into force before 3 September.
As seen above, anti-parliamentary rhetoric had already featured in the Brexit debates under Theresa May-including in her speech calling for the 2017 election. It was also evident in another, much-criticised, Downing Street statement following the second Commons defeat on her deal. As ever, not wanting to single out her own backbenchers for criticism, the Prime Minister stated that 'MPs have been unable to agree on a way to implement the UK's withdrawal', suggesting that 'you the public have had enough' and, distancing herself from the parliament on whose support her government depended, claiming that 'I am on your side' (May, 2019). This had clear undertones of populism, which presents society as 'ultimately separated into two homogeneous and antagonistic groups, "the pure people" versus "the corrupt elite"' (Mudde, 2004, p. 543). May sought to portray herself as the people's defender against intransigent MPs-one of whom at the time, of course, was Boris Johnson.
Although Johnson's argument had been with May, his own administration, particularly after the Supreme Court judgment, adopted even more persistently and manifestly populist anti-parliamentary language. The Attorney General's comments (above)-made during the Commons' first debate on its return-illustrate this. Johnson used his Conservative Party conference speech the following week to suggest that 'If parliament were a school, Ofsted [the educational standards body] would be shutting it down', and to proclaim the views of a united people, suggesting that 'what people want, what leavers want, what remainers want, what the whole world wants' was to 'get Brexit done' (PoliticsHome, 2019). By now, the very question of whether parliament should sit was becoming politicised along Brexit lines; polling showed that 76% of Remain supporters approved of the Supreme Court's ruling overturning the prorogation, but 62% of Leave voters opposed it (Curtice, 2020).
Shortly before the October Brexit deadline, Johnson secured a new deal with the EU27, which avoided customs union membership for the island of Great Britain but retained this for Northern Ireland (to the consternation of the DUP). On 22 October, MPs agreed a second reading of his Withdrawal Agreement Bill by 329 votes to 299, so that it might move to its committee stage. But they rejected the bill's programme motion, which had proposed just three days' Commons scrutiny. At this point, Johnson pressed, for the third time since early September, for an immediate general election. To circumvent the requirements of the Fixed-term Parliaments Act he did so via a short bill, allowing it to be achieved by a simple (rather than two-thirds) majority. This facilitated, as Johnson had almost certainly wanted from the outset, an election fought on a 'parliament versus people' theme. In his own pre-election statement in Downing Street, he suggested that 'our MPs are just refusing time and again to deliver Brexit and honour the mandate of the people', and (deploying what was now his key election slogan) what was needed was instead to 'get Brexit done' (Tapsfield, 2019). What he omitted to mention, of course, was that he and several members of his Cabinet had been central protagonists in preventing Theresa May from doing exactly that.

Analysis: what happened and where next?
This article has argued that four key factors contributed, in interrelated ways, to the 'perfect storm' that caused the UK executive to pit itself increasingly explicitly against parliament over Brexit. The 2016 referendum presented the public with a change option that government and parliament did not support, which was illspecified, and which left parliamentarians to navigate the unexpected result. Theresa May's desire to strengthen her hand in parliament then resulted in a minority government, for which the UK system was ill-adapted, and to which she was personally ill-suited. There were deep divisions inside her own parliamentary party, but as an instinctively tribal politician, May made little attempt to build cross-party support, while expressing increasing frustration at an undifferentiated 'parliament' rather than her own backbenchers. Had the House of Commons rules been better suited to a minority situation (e.g. as proposed by the Wright committee in 2009), others might have been better able to construct a compromise around a 'soft' Brexit deal, and/or a further referendum. But this proved impossible; while Jeremy Corbyn's unpopularity undermined attempts to create an alternative government. Repeated chaotic scenes, with MPs failing to back any concrete Brexit alternative, no doubt helped build support for Boris Johnson's eventual 'people versus parliament' rhetoric. That he and his allies had been part of the blocking majority at earlier stages remained something that his predecessor, and oddly even the Labour leader, never highlighted. In the end, the result of a 'bluff call' referendum driven by divisions in the Conservative Party failed repeatedly to be delivered, due to continued divisions in the Conservative Partyand the rhetoric of the party's leaders increasingly heaped the blame on parliament.
Throughout this process, fundamental tensions grew about where sovereignty should lie in the UK system, and these remain unresolved. Referendums are devices favoured by populists (Müller, 2017), and have the potential to challenge the UK's traditional central tenet of parliamentary sovereignty. The peculiar setup of the 2016 poll, and parliament's ambivalent attitude to the result, led the government to claim a direct mandate from 'the people': invoking a principle of popular sovereignty effectively to justify demands for executive sovereignty, which became more insistent as the tensions with parliament mounted. But there was no instruction from the people about the form that Brexit should take, while many pro-Brexit MPs who argued against parliament's interference had themselves voted to block the government's deal. In truth, divisions in parliament-between supporters of remaining in the EU, or of a 'soft' or 'hard' Brexit, or a further referendum-reflected fairly accurately the divisions in the country. Just as there was no undifferentiated 'parliament' to criticise, there was no undifferentiated 'will of the people'. Nonetheless a minority government, and its own backbench opponents, both claimed ever more vocally to speak on the people's behalf.
The general election of December 2019 was won by Boris Johnson's Conservative Party with a House of Commons majority of 80 seats. This promised to bring to an end many of the preceding difficulties. It ensured that Brexit was delivered and saw a return to the familiar territory of single-party majority government. The 'perfect storm', of a minority government seeking to pursue a highly controversial referendum result with a divided party, could in principle be left behind. But the scars are likely to remain.
Prior to the referendum trust in parliament was already at a historically low level; it since appears to have declined further, providing a fertile breeding ground for populism. 15 The conflicts also fuelled frustration among Brexit-supporting politicians about parliament's power, and demands for reform. The Conservative manifesto hence promised a 'Constitution, Democracy and Rights Commission', with a far-reaching remit to examine 'the relationship between the Government, Parliament and the courts' (Conservative Party, 2019, p. 48). One obvious conclusion from such an exercise could be that both Whitehall and Westminster need to adapt better to the possibility of future minority governments. But with the very meaning of parliamentary sovereignty itself now in contention, the Brexit 'perfect storm' has exposed fundamental questions about the core principles of the UK democratic system.