The constitution continually changes. The Victorian jurist A. V. Dicey still argued that the monarch could dissolve Parliament on his or her sole authority, without the advice of ministers: ‘A dissolution is allowable, or necessary, whenever the wishes of the legislature are, or may fairly be presumed to be, different from the wishes of the nation.’ But if the Fixed Term Parliaments Act 2011 removed this royal prerogative to dissolve, it explicitly provided that ‘This Act does not affect Her Majesty’s prerogative power to prorogue Parliament’ (6 (1)). That power is, indeed, regularly exercised to determine the dates of Parliament’s sitting. The Act did not subject that power to convention, or establish purposes that might make it subject to judicial review.
The situation Dicey envisaged clearly obtains today, and it could be remedied by prorogation rather than by dissolution. The prerogative to prorogue overrides the provisions of the Northern Ireland (Executive Formation) Act 2019 as amended, requiring ministers to report to Parliament on set dates: if Parliament were not sitting, such a requirement could not be performed, and falls away. Had the Northern Ireland Bill overridden prerogative, it would have required Royal Consent for debate on it in Parliament to proceed; no such Consent was given (this requirement is a matter of convention, not of law; but its absence is evidence of the nature of the Act). Indeed the Act contained no reference to its superseding or abrogating prerogative as restated in the Fixed Term Parliaments Act 2011, and the courts (if asked) are unlikely to guess that it did. If the 2019 Act had been intended to do so, it would have said so; but it might not, in that case, have commanded a majority.
The rationale for using prorogation to prevent attempts to block Brexit would be that the people (being sovereign) had voted to leave the EU in a referendum legislated for by Parliament, that Parliament, in compliance with this exercise of popular sovereignty, had voted to trigger Article 50, and had not challenged a leaving date extended by Statutory Instrument; but that some members of the House of Commons were now attempting to frustrate this democratic decision taken by the People and accepted by their Representatives. These are strong arguments.
If prorogation is not employed, however, further challenges are evidently being planned. Remainers are contemplating ways of frustrating the government’s policy of leaving on 31 October. What are the options?
First, unusual exercises of royal authority have been spoken of: that the Queen should be ‘sent to’ Brussels to request an extension of Article 50; that Jeremy Corbyn, as Leader of the Opposition, should be ‘sent to’ Buckingham Palace to demand royal recognition of his succeeding to the office of Prime Minister; that, more precisely, the Queen might independently dismiss her Prime Minister, Boris Johnson. This prerogative was last employed within the UK in 1834 (although against Gough Whitlam in Australia in 1975), but its discussion today is evidence that Remainers are willing to accept old prerogative as still usable in novel circumstances. Yet the monarch, seeking to avoid involvement in party politics, is unlikely to use it unless on the advice of her ministers (i.e. Mr Johnson himself) or unless she has clear evidence that another individual can command a majority in the Commons. Either currently seems improbable.
Second, Remainers in the House of Commons may seek to break convention by taking control of the order paper in order to legislate to block Brexit. The House may, by a motion, suspend or modify its Standing Orders in order to pass a Bill to instruct the Prime Minister to revoke Article 50. It may legislate to require a ministerial request to the EU for an extension of Article 50. It may amend the Fixed Term Parliament Act to regulate the date of an election.
How might the government respond? It might also break convention. If such a Bill were presented for the Royal Assent, the assembled peers and Commons might hear, instead of the familiar Norman French formula La Reine le veult, the formula La Reine s’avisera. This power was used by Queen Anne in 1708, importantly on the advice of her ministers, to veto a Bill that would have established a Scottish militia. Since France was about to invade Scotland in the Jacobite cause, and since the Scots were rightly suspected of widespread loyalty to the House of Stuart, this exercise of the royal prerogative on the advice of ministers was popular in England. But it was not the last such occasion when something similar was done.
Convention requires the prior Royal Consent (not to be confused with the Royal Assent) if Parliament is to debate any Bill that would affect the sovereign’s prerogative powers. In 1999 the Queen, on the advice of her ministers (notably her then Prime Minister, Tony Blair), explicitly withheld that Consent to the Commons’ debating the Military Action Against Iraq (Parliamentary Approval) Bill, the private member’s Bill of Tam Dalyell, which would have given Parliament control over the use of force in that instance. The convention was accepted by the Commons, and debate was duly blocked. We may question the wisdom of the ministers’ choice, but they succeeded, and the Iraq war went ahead. The remedy for those unhappy with that decision was the democratic removal of the government, not judicial review. Meanwhile, it is open to argument which of the Remainer tactics considered above would require the Royal Consent in order to be debated in Parliament, Consent which might be legally withheld.
Additionally, the Crown, on the advice of her ministers, has a reserve power to delay the presentation of a Bill for the Royal Assent. In the present case, the presentation of any of the Remainer Bills outlined above could be delayed until after 31 October, and so rendered of no effect.
These are uncommon remedies. Over time, the ministry’s increasing and normal control of business in the Commons had rendered unnecessary the veto of a Bill that had already passed the Lords and Commons. But an executive veto works in the United States as a recognised part of democratic constitutional practice; it is not ‘outdated’. If it were objected that the revival of such a prerogative is inconsistent with recent convention, we might recall the advice of Speaker Bercow that precedents may guide but they do not bind. Conventions are many and various, and sometimes inconsistent, not a collection of simple ‘right answers’.
Third is the possibility that the government loses on a motion of no confidence in the Commons. In recent history, this would, by convention, have led at once to the resignation of the Prime Minister, an invitation to the leader of the party that had just secured a majority to form a government, and a general election. So it was when James Callaghan lost by just one vote in 1979. But the Fixed Term Parliament Act 2011 has explicitly superseded that convention. It provides for a fourteen day period for attempts to find a government that could command a majority. But it does not require that the incumbent Prime Minister resign while this is done. The Act is not unclear on that point; it is silent. Its silence entails that the incumbent may continue in office.
Indeed any number of motions of no confidence could be passed, but they would not of themselves, in law, unseat the Prime Minister. The qualification for holding office as Prime Minister is conventional: not surviving votes of no confidence, but that the sovereign invites that person to hold the office and continues to support him or her in it. So it was with William Pitt the younger in 1783-4, installed by George III and surviving successive defeats until opinion in the Commons rallied behind him and he was encouraged to face the electorate, which duly provided him with a majority. If the lack of a majority continues for long enough, a PM’s position would in practice become untenable if he were unable to pass legislation (especially a budget). But this practical consideration would hardly apply over a period of a few days or weeks, which is all that is at issue here.
Say a general election were triggered under the 2011 Act if Mr Johnson were to fail to assemble a working majority in the fourteen day period provided. The Fixed Term Parliament Act, 2 (7), provides that in such cases of ‘Early parliamentary general elections’ ‘Her Majesty by Proclamation on the recommendation of the Prime Minister’ appoints the polling day, which could therefore be after 31 October. Had Parliament wished itself to regulate the date of elections, it would have done so in the Act of 2011. Similarly, that Act, 3 (4) (a), provides that the Queen appoints the day for the meeting of the new Parliament, which could be after 31 October even if polling day were before that date.
But the 2011 Act does not guarantee a fourteen-day period during which another person might assemble a majority. Section 2 (2) provides for a motion at any time ‘That there shall be an early parliamentary general election’. If that motion were passed, the incumbent Prime Minister might at once appoint a polling day after 31 October.
Say no such motion is passed, and the fourteen-day period proceeds. Say another member emerges, during that period, commanding a majority in the Commons. Such a PM, if supported by a majority, may legislate to any effect; but the improbability of the current House rallying behind any of the candidates so far named has been much discussed.
The Cabinet Manual (2011) advises (2.27) that during a general election campaign ‘governments are expected by convention to observe discretion in initiating any new action of a continuing or long-term character’. Convention, however, does not specify what is ‘new action’: Parliament has already legislated to trigger Article 50 and leave the EU. It has rejected a deal with the EU three times, and has not disputed a date for leaving. No new action would therefore be taken by ministers were existing legislation merely to take its course. Similarly, a new tax approved in the last budget would continue to be collected, however unpopular among some. Even if this were not the case, since the Labour Party is torn two ways on Brexit it would be hard to invoke judicial review on the grounds that the policy of the most likely incoming ministry (to Leave or to Remain) could be known in advance. It is not clear that case law bears on this question, and the courts will not intervene on the basis of convention where Parliament has already legislated. The remedy lies with the electorate.
A more telling option has not so far been widely commented on, but it would be decisive. The European Union (Withdrawal) Act 2018 provides (20 (4)) that ‘A Minister of the Crown may by regulations (a) amend the definition of ‘exit day’ in subsection (1) to ensure that the day and time specified in the definition are the day and time that the Treaties are to cease to apply to the United Kingdom’. The House of Commons Library briefing paper of December 18, 2017 explained the draft text: ‘The regulations made under this power will not be subject to any parliamentary procedure.’ The exit date was duly changed to 31 October 2019 by Statutory Instrument 2019 No. 859, made on 11 April 2019, not by legislation.
Mr Johnson has not formally changed that date. But if he were to lose control of parliamentary business, it would be legal for him at once to name an exit date before 31 October. The UK may wake up one morning to find itself out of the EU. Whether the sun will continue to rise, or the heavens fall, will then remain to be seen. But UK statute law will have been complied with. It might be objected that EU law is still supreme in the UK, and that EU law sets the departure date as 31 October. But this would merely emphasise that the will of the electorate, not a disembodied ‘law’ or ‘convention’, will determine the outcome.
Too much, then, is made of the power of convention in the UK’s supposedly ‘unwritten’ constitution by those who wish to frustrate the UK’s departure from the EU. As the distinguished judge Lord Sumption reminded us, ‘The Supreme Court held in the Gina Miller case in 2017 that a constitutional convention is not justiciable in the courts.’
Too much is also made of the power of the law against the sovereign power of the People. Conventions now operate in a wholly new situation: Parliament’s legislation which provided for the arrival of direct democracy (via referendums) to challenge, modify, or override parliamentary democracy. The bland and hitherto uncontroversial claims of the Cabinet Manual (2011) that sovereignty resides with Parliament are now out of date. Both law and convention, then, tend to lag behind evolving realities.
In reality, most of the constitution is indeed written, though not in a single document. When disputes arise, it is normally because neither statutes nor conventions can provide in advance for every circumstance in the infinite variety of human life. Lawyers are used to arguing to apply old law and old conventions to new situations; in court, they have to claim that only one conclusion is tenable (their own). But the public understand that more than one outcome is possible. The outcome is settled (or partly settled) by the choice of the electorate. That is how it has always been; that is how it will be now, in the new context created by the electorate’s decision on 23 June 2016.
Of one thing we can be sure: the constitution continually changes, and often in ways that no-one expects. Today, all the evidence suggests that representative democracy is on the decline, and direct democracy is on the rise. Conflicts between two plausible ideals are the very stuff of politics.
Jonathan Clark is a leading political and constitutional historian of early modern Britain.
 Since this was not done in 2011, it is again evidence that the Fixed Term Parliament Act was not intended to limit the prerogative other than the prerogative to dissolve.