Written evidence submitted by Dr Craig Prescott (FTP0016)

I am a Lecturer in Law at Bangor University. Previously, I was the Director, Centre for Parliament and Public Law at The University of Winchester. With Dr John Stanton (City University London), I am the author of Public Law, published by Oxford University Press. I regularly contribute to the national media discussing political and constitutional issues.

As my research focuses on Parliament and the monarchy, the Fixed-term Parliaments Act 2011 (“2011 Act”) is a natural focus. I have previously submitted evidence on the 2011 Act to the House of Lords Constitution Committee and House of Commons Public Administration and Constitutional Affairs Select Committee.[1]


-      The pre-2011 Act gave the Prime Minister considerable latitude over when to hold a general election, although in practice the Prime Minister did not have an entirely free choice. A practice emerged of elections taking place in the spring, usually in the 4th or 5th year of parliament.

-      The biggest advantage of the 2011 Act is that introduced certainty as to when the next general election would be held. This allows both the government and Parliament to plan, yielding the greatest benefits towards the end of the Parliament, as there is no “will they or won’t they” speculation over when the Prime Minister will request a dissolution.

-      The real cause for the unpopularity of the 2011 Act are the flaws in the mechanisms to hold an early election. The relationship between the Acts provisions and existing conventions regarding confidence and the duty on the Prime Minister to resign is at best unclear, and the Act is silent on what could, or should happen during the 14-day period it introduces. Triggering an election with the votes of 66% of MPs can simply be negated by another Act of Parliament, disapplying that requirement.

-      A far more certain way of dealing with those flaws is to provide for a maximum term of a parliament, with provision for an earlier election if a simple majority of MPs agree. This avoids any technical questions regarding the nature of a revived prerogative power, and any need for an ouster clause. This also avoids restoring any potential uncertainty over the ability of the Monarch to refuse a request for a dissolution.

-      It is unnecessary for the maximum legal length of a parliament to return to five years from the day it first meets, as opposed to the provisions of the 2011 Act, which provide a default election date of the first Thursday in May in the fifth calendar year following the date of the previous election. Parliament is dissolved by operation of law twenty-five working days before that date.

-      If the Monarch is required to grant the Prime Minister’s request for a dissolution, then this raises the possibility that future Prime Ministers could use the revived prerogative power differently to their predecessors from the 2011 Act era. As shown with the purported prorogation in 2019, the repercussions fall upon the Prime Minister, and not the Monarch. However, the role of the Monarch in more extreme or unusual circumstances remains ambiguous.

-      The ambiguity could be resolved by the Dissolution Principles document published alongside the Draft Fixed-term Parliaments Act 2011 (Repeal) Bill. At best, it is a partial success. The rules that it outlines are a mixture of constitutional convention and practice, and to fully make sense need to be read alongside broader principles of government formation. The most effective way to do this would be to update the Cabinet Manual, taking inspiration from the relevant paragraphs of the New Zealand Cabinet Manual, which deals with this issue in admirably clear and concise three paragraphs.


Given that the aim of the Draft Fixed Term Parliaments Act (Repeal) Bill (“Draft Bill”) is to restore the law, and broader constitutional position, that existed before the enactment of the Fixed-term Parliaments Act 2011 (“2011 Act”), it is necessary to consider what the position was before the enactment of the 2011 Act, what the 2011 Act did, the cause for its unpopularity, and a discussion of the provisions of the Draft Bill.

Given that the Draft Bill touches on a fundamental part of the constitution and does so in a manner that raises several issues of constitutional importance. Unfortunately, this means that any reasonably full treatment of those issues ends up being quite lengthy.


  1. Before the 2011 Act, legal regulation over the timing of general elections was scant. The only relevant provision was Septennial Act 1715 (as amended by the Parliament Act 1911) which set the maximum term of a parliament as being five years calculated from the from the date that Parliament first met. In practice, no Parliament (at least since the five-year limit was introduced in 1911) has run for its entire term as provided for by the Septennial Act.[2] Instead, Parliament was dissolved by the monarch exercising the prerogative on a date requested by the Prime Minister. This applies to parliaments that are often described as running for a “full term”. For example, the 1987 Parliament first met on 17th June 1987, but was dissolved by the royal prerogative on 16th March 1992 ahead of a general election on 9th April 1992.
  2. Although at its core, the decision of when to request a dissolution was a political judgement for the Prime Minister in practice, they did not have an entirely free choice. Looking at the precedents from the 20th century to 2010, it is clear that there are a range of factors that inform when Prime Ministers request a dissolution. These factors include the size of their majority, the state of the opinion polls, the length of time since the last election, and, if they’ve taken office since the previous election, the amount of time they have spent at Downing Street. With these factors in mind, it is possible to identify the following illustrative categories of dissolution.

Prime Ministers with Workable Majorities - 4th/5th Year Dissolutions

  1. If the Prime Minister is enjoying a workable and stable majority, then in principle, they would want to go deeper into a parliamentary term in order to implement their manifesto. Yet, once a parliamentary term ran into its third year, speculation would inevitably arise as to when the Prime Minister would request a dissolution of Parliament. In essence, the Prime Minister’s choice would be whether to make the request in the fourth or fifth year of the term. Prime Ministers, such as Margaret Thatcher and Tony Blair, in a strong political position, with relatively high opinion poll ratings, requested a dissolution in the fourth year,[3] whereas those in a weaker position, would hope (perhaps forlornly) that their political fortunes would improve during the fifth year of the term.[4]
  2. The range of possible dates has also been limited by practice. Since the 1966 General Election, there has been a strong preference to holding elections between late March or early June, with only the two elections in 1974 deviating from this patter.[5] However, since 1997, the more precise practice of requesting a dissolution in order to hold the election in the first Thursday in May has emerged.[6] The key point is that in setting the default election date to the first Thursday in May, the 2011 Act  enshrined a tradition of more modern origin than is perhaps widely thought, and it does not necessarily follow that future Prime Ministers will follow this pattern.[7]
  3. Another complicating factor is when a Prime Minister takes office during the term. There is a temptation for a Prime Minister in this position to request an early dissolution in order to obtain a personal mandate from the electorate for their policies and their premiership. However, the only clear example of this, is Sir Anthony Eden’s request for a dissolution in April 1955, made almost immediately on taking office. Other Prime Ministers: Harold Macmillan, John Major and Gordon Brown, have all waited until the fifth year of the term in which they took office before requesting a dissolution.[8]

“Snap” Elections

  1. From the precedents, a Prime Minister with a smaller majority may request a dissolution earlier in a term in the belief that this will bolster their political position.  The best of example of this remains Harold Wilson’s decision to go for a snap election in 1966, with the result of turning his majority of four into one of ninety-six.
  2. Conversely, a Prime Minister may decide that their small majority, or their lack of one, is making it increasingly difficult to govern effectively, meaning that an election is required or is an inevitability, most likely following a vote of no-confidence. Again, it was Harold Wilson, who following the February 1974 election requested a dissolution six months into the term. In such cases, the decision to hold an early general election is not necessarily a surprise, but the precise timing could catch the opposition parties off-guard.
  3. By contrast, John Major enjoyed a relatively small majority of 22 following the 1992 General Election, but following Black Wednesday, several scandals and a popular opposition, the opportunity to request an early dissolution and hold a snap election never arose.

Motion of No Confidence

  1. Alternatively, as discussed in more detail below, the Prime Minister would be required by convention to call an election if they lost a motion of no confidence, as occurred in 1979. Although this occurred in the fifth year of the term, Jim Callaghan was required to seek dissolution the following day, meaning that he did not have a free choice over the date of the election.

(B) The Fixed-term Parliaments Act 2011

  1. These precedents rehearsed above, largely became of historical interest only following the enactment of the 2011 Act, given that none of the above illustrative categories survived the 2011 Act in any meaningful manner. This is because the primary focus of the 2011 Act was the setting of the polling date of general elections. It does this in two ways, by providing for the polling dates of “parliamentary general elections”, but if necessary, an “early parliamentary general election” can be held if the provisions of the Act are used.

“Parliamentary General Elections”

  1. Primarily, the 2011 Act operates by setting a default election date, with dissolution occurring by operation of law twenty-five working days before that date, leaving no role for the Prime Minister or the Monarch. The default election date is the first Thursday in May in the fifth calendar year following the date of the previous general election.[9] Alternatively, if an early election has been held before the first Thursday in May of that year, then the following general election will take place on the first Thursday in May in the fourth calendar following that date.[10] Otherwise, it will take place on the first Thursday in May in the fifth calendar year. Consequently, under the 2011 Act, the date of the next general election is Thursday, 2nd May 2024. Had the 2019 General Election taken place on (say) 4th April 2019, the next general election would have been due on Thursday, 4th May 2023.[11] As discussed in paragraph 4 above, fixing the election date to the first Thursday in May is a relatively modern practice, and it is not guaranteed that this would survive any repeal of the 2011 Act.
  2. By setting polling dates in this way, prime ministerial discretion over the timing of a general election has been removed. This has the benefit of preventing any speculation over the date of the next election, which as stated above, would inevitably arise once a parliament entered the third year of its term. As found by the Political and Constitutional Reform Committee, in its inquiry Fixed-term Parliaments: The final year of a parliament, fixed-terms have the advantage of encouraging longer-term planning particularly within the government, given that it knows it has the full term within which deliver its objectives.[12]
  3. As regards Parliament, a fixed term also has advantages. In particular, it allows the five-year term to be split into (roughly) five year-long sessions. This makes it easier for the government to plan its legislative agenda for the whole parliament, as it knows how much parliamentary time is available to debate and enact legislation. As shown in 2015, this does reduce the possibility for bills to be lost in the “wash up”. In addition, a full term should facilitate pre-legislative scrutiny as a draft bill can be published in one session, before being formally introduced in the following session.[13]
  4. Similarly, select committees can plan how they wish to fulfil different aspects of their remit over the five-year term. If a committee has a greater idea of the government’s legislative programme over the parliament, then committees can conduct pre-emptive inquiries into those policy areas, with the potential to feed into the legislative process. In the final year of a term, committees could look at the issues that are likely to arise during the next Parliament, again giving the potential to play a proactive role in the policy making process.
  5. The stability over the timing of elections that the 2011 Act introduced is its single biggest advantage.

Early Parliamentary General Elections and Loss of Confidence

  1. The most problematic provisions of the 2011 Act, are those that allow for an early election. There are two ways in which any early election, identified as an “early parliamentary general election”, can be held under the 2011 Act, these are as follows:

   s 2 (1)-(2): If 66% of MPs vote in favour of the motion "That there shall be an early parliamentary general election” ("the 66% method").

   s 2 (3)-(5): "That this House has lost confidence in Her Majestys Government ", and within the following fourteen days, the House of Commons does not pass the motion, "That this House has confidence in Her Majestys Government" (“the no confidence method”).

  1. The main problem with the 66% method, is that it does not prevent a simple majority of the House of Commons providing for an election in another Act of Parliament, which applies notwithstanding what is provided for in the 2011 Act. This was how the Early Parliamentary General Elections Act 2019, provided for the 2019 General Election.[14]
  2. As extensively discussed by both the Public Administration and Constitutional Affairs Committee and Constitution Committee, the no confidence method creates significant problems. As I have extensively explored this issue in the written evidence I submitted to both committees, the following is only a short summary. The problem is that the Act does not explain what should happen during the fourteen-day period allowed for by the Act, or address the effect of non-statutory confidence motions, as existed before the Act. The particular issue is when does the Prime Minister fall under a duty to resign.
  3. With statutory no confidence motions, I agree with the conclusion of the Public Administration and Constitutional Affairs Committee, that the Prime Minister does not fall under an immediate duty to resign following a loss of statutory no-confidence motion. Instead, the Prime Minister’s duty to resign arises once it is clear that an alternative government exists.[15]  This ensures the Monarch is not left without a government. Should no alternative government emerge, the fourteen-day period should expire. If it is clear that no alternative government is going to emerge during the remainder of the fourteen-day period, then there is no obstacle to dispensing with what is left, by triggering an immediate general election using the 66% method.
  4. As regards non-statutory confidence motions, given that there is no indication the Act intended to abolish them in any way, and that more broadly, the fundamental principles of government formation, that the Monarch must always have a government and that government must have the confidence of the House of Common, remain untouched. Consequently, it remains possible for the Commons to withdraw confidence from the government in a manner outside the 2011 Act. Viewed in this way, the dynamic is the same as statutory no confidence motions - a Prime Minister has lost the confidence of the House of Commons but is unable to request a dissolution of Parliament. Given the principle that Monarch cannot be left without a government, that as with statutory no confidence motions, the Prime Minister, should only resign if it is clear that an alternative government can be found. In practice, defeat on a no-confidence motion would the likely start of a process leading towards an early general election using either of the methods provided for by the 2011 Act.
  5. Taking this common approach for both non-statutory and statutory confidence motions has several advantages. It places the Prime Minister in the same position as following a general election that returns a hung parliament. As seen in 2010, and as outlined in the Cabinet Manual, the position is that the Prime Minister should resign once an alternative government has emerged.[16]
  6. Another advantage is that it removes the possibility for the structure of the 2011 Act to be gamed by the political parties. Imagine that a government loses a non-statutory confidence motion, and the Prime Minister was placed under an immediate obligation to resign. The Monarch would presumably appoint the Leader of the Opposition as their successor to form a second government of the parliamentary term. Depending on the results of the previous election, this second government could be significantly short of a majority, but now lacks the ability to call an election without the agreement of at least some of the Opposition parties. This includes the party that formed the first government. This party could take the opportunity of being Opposition, to re-group, and perhaps hold a leadership contest. During this period, the second government would find it difficult to govern or pass legislation. Being the largest party, the party that formed the first government would exercise considerable control over the timing of any general election, as in practice their support would be required for any statutory motion of confidence or for the 66% method to be used.
  7. This sketch highlights the potential difficulties that flow from requiring an immediate resignation following a loss of confidence. More realistically, a loss of confidence in the government, in whichever way it occurs, is likely to only be a step towards an inevitable general election.
  8. This may be thought to run counter to the previous practice, as shown in 1979, that a loss of confidence gives the Prime Minister two options: to either resign, or, to request a dissolution. The argument being that the 2011 Act, in only abolishing the Prime Minister’s ability to request a dissolution, left the other option untouched. Yet, the rationale for resignation arises because of the existence of the option of dissolution. It was clearly constitutionally unacceptable for a Prime Minister to fail to pursue either option following a loss of confidence. Similarly, should they choose to resign immediately, their successor (presumably the Leader of the Opposition) had the option of seeking a dissolution at a time of their choosing.
  9. At a more theoretical level, the core issue is the relationship between statute and convention. The 2011 Act introduced into an area of the constitution governed by convention specific statutory rules, while leaving unaffected the broader principles that operate in this area. The 2011 Act has precluded the operation of specific conventions because to follow those conventions is no longer allowed by the law. Consequently, the Prime Minister can no longer request a dissolution, as that is now governed by law. The question left unanswered by the 2011 Act is the extent to which its provisions impact on those pre-existing conventions not within the scope of the Act. The discussion above seeks to fill that gap.
  10. Inherent in the notion of constitutional conventions is that they are capable of adapting to the changing constitutional context in which they operate. The Jennings test for the existence of a constitutional convention includes considering the reasons for the rule.[17] It follows that if the reasons for the existence of convention change because of the underlying constitutional landscape has changed, then the operation of that convention may need to change in order to take account of the new landscape. The argument is that the convention regulating when the Prime Minister is under an obligation to resign has been altered to take account of the changes introduced by the 2011 Act.


  1. Clearly, the 2011 Act is seen as a contributory factor to the parliamentary machinations during 2019 when the House of Commons struggled to come to decision over which form of Brexit to pursue, or whether to continue to pursue Brexit at all, with a range of options all failing to obtain a majority. These were wholly unusual circumstances, with tensions heightened by the tight timetable provided for by Article 50 TEU. Clearly, the dynamics are likely to have played out differently had the 2011 Act not been passed, and the old prerogative rules applied, but speculating as to what would have happened is a fool’s game as it would have been just one of several factors for consideration by MPs. However, it is not too great a leap of the imagination to suggest that a general election could have been held earlier in the year than it ended up being. Theresa May could have decided to have made one of the votes on her “deal” a matter of confidence; Boris Johnson, as a new Prime Minister taking office during a Parliament, could have sought a personal mandate for his renegotiations in order to bolster his position with the EU, or the opposition parties could have moved a motion of confidence triggering an election.
  2. Indeed, given the fluidity of the situation, and that the ultimate sanction of a no confidence vote would have rested with MPs, the government could have taken a more emollient stance towards the Commons, heading off as much opposition as possible before it gathered momentum leading to a confidence motion.[18] The pre-2011 Act rules may also have allowed some MPs to vote against the government on a no confidence motion, safe in the knowledge that this would trigger a general election, rather than the appointment of a new government that might not have been to their liking.[19] In summary, the Brexit process in the Commons would have been a very different game of chess, with many of the pieces occupying different places.


  1. It against the backdrop of these problems that the Draft Bill should be considered. The aim of the Draft Bill is to return to the legal position as it stood prior to the enactment of the 2011 Act. It seeks to achieve this by repealing the 2011 Act, and calculates the maximum life of a parliament in the manner that used to be provided by the Septennial Act, as five years from the date that parliament first met following the general election. Parliament could be dissolved earlier, through the revived power of dissolution. Yet, in doing all of this, the Draft Bill leads to more questions than it seeks to answer.

Question One: The Dates of Future Elections

  1. The first question is what is the date of the next general election? Under the pre-2011 Act rules, despite the Prime Minister enjoying considerable latitude over when to hold an election, as discussed above, the practice emerged of holding elections on the first Thursday in May. The 2011 Act enshrined this practice in law. Is the intention of the government to return to this practice? If it is the government’s intention, then it is strange to propose legislation that creates the very possibility of not doing that. The Draft Bill creates the possibility that, instead of this Parliament being dissolved on 26th March 2024 for the scheduled polling day of 2nd May 2024, it could be dissolved on 17th December 2024, leading to an election in January 2025.
  2. There are decent arguments in favour of holding elections in May (or the Spring). Yet, given the precedent of a successfully held election in December 2019, future Prime Ministers may feel that they have a freer hand than their pre-2011 Act predecessors when choosing an election date.[20]  There is also now the precedent that the 2010 Parliament was dissolved by operation of law, as provided for by the 2011 Act itself.
  3. Another concern is the appropriateness of Parliament extending its maximum term in this manner. The 2019 Parliament was elected under the provisions of the 2011 Act. Although repeal of the 2011 Act is a manifesto commitment, of both the Labour and Conservative Parties, it does not follow that the maximum term of this Parliament should be extended in this way. In modern terms, the term of a Parliament has only been extended during wartime.[21] By contrast, the 2011 Act trimmed the maximum term of the 2010 Parliament slightly, to ensure that the election was held on 7th May 2015.
  4. These concerns could be addressed by continuing with the approach of the 2011 Act, where the maximum term of a parliament is calculated so that it is dissolved twenty-five working days before the first Thursday in May, in the fifth calendar year following the previous election.[22] This would provide legal certainty as to the latest possible date of the next general election, while the Prime Minister retains the option of requesting an earlier dissolution for an election on earlier date.
  5. One further complicating factor is devolution. The Welsh and Scottish Parliaments are each dissolved 21 and 28 working days before the polling day for their respective elections,[23] with the UK Parliament dissolved 25 working days before polling day. Should a Prime Minister choose to hold a UK general election on the day Welsh and Scottish Parliamentary elections are scheduled, (which is the first Thursday in May) those devolved elections would be delayed as a consequence.[24] Using the pre-2011 Act practice as a guide, Prime Ministers would announce that they have requested a dissolution at short notice, which could potentially be only a matter of days prior to the expected dissolutions of the Welsh or Scottish Parliaments, or theoretically just after the Scottish Parliament has already been dissolved.

Question 2: How Does it Address the problems of the 2011 Act?

  1. Retaining the approach of the 2011 Act for the maximum length of a term, would still allow the core problems with the 2011 Act to be addressed, which are the mechanisms for holding an early general election, especially during a hung parliament. As discussed above, there is some uncertainty regarding what happens during the fourteen days following a successful no confidence motion.
  2. As regards the 66% method, the experience of 2019 reveals that the 66% threshold can be both unobtainable and insufficient. In 2019, three motions achieved a simple majority of votes, but due to abstentions, the threshold was not met, as Labour made their support for an early election conditional on obtaining concessions from the government.[25] By contrast, if a vote for an early election merely required a simple majority of votes, then different dynamics are at play, and the support of the main opposition party is much less critical, as they enjoy less bargaining power.
  3. This dynamic was seen with the Early Parliamentary General Election Act 2019. The Labour Party abstained on three motions tabled by the government to trigger an early general election under the 2011 Act, knowing that their lack of positive support meant that no election would be held.[26] Yet, Labour ultimately moved to support an election held under the 2019 Act, knowing that it had the support of the smaller opposition parties, meaning that the legislation was all but certain to pass, regardless of their stance.
  4. This all leads to the conclusion that many of the problems the 2011 Act could be said to have caused in 2019, and may cause in future, are avoided by requiring that a motion for an early election needs a simple majority of votes cast.[27] This would be a relatively simple amendment to make to the 2011 Act. This would also negate the need for the Act to become embroiled in the issue of confidence, and that method could be repealed. Given that a no confidence motion would not lead to an early election, future Leaders of the opposition would be more likely to table a motion for an early election, which would in practice operate as confidence motions did before the 2011 Act. While a Prime Minister with a majority would, in practice, still be able to hold an early election if and when they desire.

Question Three - Is the Ouster Clause Successful?

  1. It is also the case that amending the 2011 Act in this way would negate the need for the ouster clause contained at Clause 3 of the Draft Bill. For the purposes of the Bill of Rights, Article IX, such a vote, would clearly be a proceeding in Parliament, and so could not be questioned by the courts. This is because it would be Parliament itself deciding when it would be dissolved. As Erskine May outlines, that “the primary meaning of proceedings”, for the purposes of Article IX, is “usually a decision, taken by the House in its collective capacity”.[28]
  2. The courts recognise that Article IX is a fundamental aspect of the constitution, speculating that it could even take precedence over other constitutional fundamentals.[29] Furthermore, the basis of the decision in R (Miller) v Prime Minister is that parliamentary accountability is a constitutional principle, which alongside parliamentary sovereignty, is predicated on the basis that Parliament is the elected institution within a constitution based representative democracy.[30] Indeed, the very justification for Article IX, and related aspects of parliamentary privilege, is that they buttress representative democracy, by ensuring that the political process remains a matter of politics and not law. Should elected representatives decide that a general election is to take place, it would an extraordinary move for the courts to intervene and prevent that election. It is difficult to envisage the circumstances that could lead to even an arguable case.
  3. By contrast, the Draft Bill seeks to use statute to prevent the exercise of the revived power of the executive from being reviewed by the courts. The concern is that by restoring dissolution to an executive action, exercised in relation to Parliament, the courts could deploy a similar line of analysis to that deployed in R (Miller) v Prime Minister, where prorogation was described as “something which is imposed upon [Parliament] from outside”.[31] Because prorogation is an executive act, exercised in relation to Parliament, the court, in the name of the separation of powers, was able to review the scope of the power, and found that the purported prorogation lay outside the scope of the power enjoyed by the Prime Minister. The risk is that an exercise of the dissolution prerogative, could be construed in a similar manner. This is what the ouster clause seeks to prevent.
  4. One of the difficulties is that the success of the ouster clause itself is a matter for the courts to determine. Previous clauses of this type have yielded litigation that has reached the highest courts, with the ouster clause proving unsuccessful. This ouster clause attempts to catch “purported exercises” of the revived prerogative, in an attempt to deal with the reasoning of Anisminic v Foreign Compensation Commission, where despite the relevant Act stating that a ‘determination … shall not be questioned’ by the courts, the court was still able to review the underlying legality of the decision in question.[32]
  5. Even on a relatively superficial reading of the Draft Bill, it seems possible to raise an arguable case that could avoid the effects of clause 3. That clause is focused on exercises of the power under clause 2, the revived prerogative power. Yet, clause 4 states, ‘if it has not been dissolved earlier, a Parliament dissolved at the beginning of the day that is the fifth anniversary of the day on which it first met’. This raises the possibility of the courts being asked to make a declaration for the purposes of clause 4 as to whether a Parliament has been lawfully dissolved earlier or not. Potentially, a court could purposively interpret such a request in a manner that frees it from the strictures of clause 2, by attempting to deal with the question in isolation, separating it out from whether the power revived by clause 2 has been validly exercised or not.
  6. Overall, given the uncertainties with any attempted ouster clause, if the aim is to avoid any possibility of judicial involvement, the safer course of action would be to make the question of dissolution one for the House of Commons, and not for the government to decide.

Question Four - Do the Pre-2011 Act Conventions Return? The problems with the Dissolution Principles document

  1. One reason for setting all of this out in such tedious detail is that the aim of the Draft Bill of returning to the pre-2011 Act situation raises the question as to whether this also revives the constitutional conventions that existed before the 2011 Act. If the analysis above at paragraphs 25 and 26 is followed, then the constitutional reasons for the old conventions to exist would also be revived. As it was before 2011, the Draft Bill would make it constitutionally outrageous for a Prime Minister to neither seek a dissolution nor resign following a loss of confidence. The symbiotic relationship between those two options is restored. The final point is that the previous ways in which a government could lose confidence regain their political value, as they lead to the Prime Minister making that choice between resignation or requesting a dissolution. As discussed in paragraph 20, the 2011 Act did not abolish the ways in which confidence could be lost, but changed the consequences of a loss of confidence, similarly repeal would not affect the ways in which confidence could be lost.
  2. This is the basic approach followed by the Dissolution Principles document published by the Government alongside the Draft Bill. However, it’s statement that ‘… Prime Minister, having a lost a designated or explicit vote of confidence, can either resign or seek a dissolution’, excludes a vote on the Queen’s Speech, which has traditionally been considered to be a vote of confidence.[33]
  3. Although, the discussion in Section C shows that the Dissolution Principles document is broadly correct, however it is not without problems. The document appears to confuse constitutional conventions with constitutional principles, without elaborating quite what is meant by “principle”. The rule that the Prime Minister holds office, because of their ability to command the confidence of the House of Commons, is described as a principle, but this is generally recognised as constitutional convention rather than a principle.[34] This appears to be distinguished from the rule that the Monarch must always act on the advice of the Prime Minister, which is described as a “convention”. It is not clear why these two rules are distinguished in this way.
  4. This reflects the core problem with the document, which is that these dissolution “principles” cannot be seen in isolation, but only against the broader context of government formation. This includes the basic principle that the Monarch must not be left without a government, which is not mentioned in the Dissolution Principles document.[35] The most natural way to do this would be to incorporate these “principles” in a new edition of the Cabinet Manual, which in any case requires significant updating due to Brexit.

Question Five - What is the Role of the Monarch?

  1. The mere existence of the ouster clause is a perhaps admission that the revived prerogative could be exercised in such a manner that its exercise could be challenged before the courts, and that the one possible restraint on its use, the Monarch, simply no longer operates in that way.
  2. The crux of the problem is that it has become a strongly established expectation that that the Monarch personally exercises no discretion, and by convention is expected to act on the advice of the Prime Minister. This means that if she refused that advice, the Monarch would be embarking on the more hazardous path, even if the advice was of questionable propriety. As we saw with the purported prorogation, the repercussions of any such advice would primarily fall upon the Prime Minister. It was he who was taken to court, and it was his advice that was declared as unlawful.[36]
  3. If, in substance, the Monarch out of the picture, future Prime Ministers may feel less restrained over the exercise of the prerogative, knowing that the consequences will fall upon them, rather than the Monarch. In this sense, the question of when to exercise the prerogative becomes another political calculation amongst many others, and can used as a tool to achieve other, “higher” political objectives. This was the dynamic of the purported 2019 prorogation, where the Prime Minister, after a vote to hold a general election failed to meet the 66% threshold, prorogued Parliament for an unusual length, in the run up to then Brexit deadline of 31st October.[37] In the particular circumstances of 2019, prorogation was the lever available for the Prime Minister to pull, only to be undone by the Supreme Court.
  4. Reviving the prerogative of dissolution risks creating another lever for the Prime Minister to pull as part of their broader political calculations, with Prime Ministers using the power of dissolution in different ways to their predecessors from the pre-2011 Act era. This is likely to be less objectionable than an unusually long prorogation, as the ultimate sanction rests with the electorate to vote against a Prime Minister that has brought them to the polls unnecessarily. However, good governance would be likely to suffer if Prime Ministers started to repeatedly request dissolutions in, say, the second or third year of a parliament for no other reason than maintaining their party advantage,[38] or to shore up their own position in the face of opposition amongst their own MPs.[39]
  5. All of this assumes that the Monarch’s role is purely a formality. If that is the case, then it is unclear why the Draft Bill amends the Regency Acts 1937 to 1953 so that Counsellors of State can only dissolve Parliament on the ‘express instructions’ of the Monarch. It is understandable that this returns the Regency Acts to their pre-2011 Act position, but this also suggests that the role of the Monarch is more than merely ceremonial and requires their personal input. By contrast, with royal assent, it is accepted that the Monarch plays no active role, and Counsellors of State are given the powers to assent to any Act of Parliament.[40] Consequently, the Draft Bill restores the legal distinction between the role of the Monarch as regards the revived dissolution prerogative, and the great majority of other the Monarch’s other legal functions.
  6. The core problem is that we are in the dark as to this aspect of the constitution. In his autobiography, Ken Clarke shines some light, stating that in October 1992, when the government’s difficulties with ratifying the Maastricht Treaty were at their most acute, John Major spoke to Robert (now Lord) Fellowes, ‘who hinted that if the Queen was asked by her prime minister for a dissolution, she would refuse’, as she did not want a repeat of 1974 when Edward Health called an election over the miner’s strike.[41] This hints at how the relationship between the Monarch and Prime Minister may play out, where the Prime Minister floats the possibility of making an early request for a dissolution, but is dissuaded from pursuing that idea any further.
  7. The difficulty is that any such discussions between Monarch and Prime Minister lack any criteria or guidelines that could structure how they each undertake their respective roles. There is the Senex Letter, written by Sir Alan Lascelles, Private Secretary to George VI and Elizabeth II, which outlined three circumstances where the Monarch could refuse a request, which were;
    1. that the existing Parliament remains viable,
    2. an election would damage the economy, or
    3. if someone else other than the Prime Minister making the request could command the confidence of the Commons.

Even if its authoritative nature in 1950 is accepted, it is difficult to judge its applicability after seventy years of constitutional development. Indeed, as Peter Hennessy has stated, it appears that the second circumstance had withered away at some point by the 1990s.[42] Arguably, the other two are likely to be matters where the Prime Minister, rather than the Monarch, is the better judge.

  1. The Cabinet Manual would be the natural place for any such guidelines to be developed for the modern era. A similar exercise has already been undertaken, with success as regards government formation in a hung parliament. It was extremely useful to be able to point to some definitive guidelines after both the 2010 and 2017 General Elections.[43] It should be possible to arrive at a formulation of words, allowing for the necessary flexibility while ensuring, as the Cabinet Manual does on government formation, that the Monarch is not faced with a difficult political decision. The New Zealand Cabinet Manual provides an example of how this can be done, in three clear, concise paragraphs.[44] The Dissolution Principles document falls far short of this example.


  1. All of this shows that the Draft Bill raises many questions which raise fundamental issues involving several aspects of the constitution. It’s no means certain that the Draft Bill successfully resolves all of those issues. It is a brave path to choose. A far safer, and less hazardous path would be for the House of Commons to decide if Parliament should be dissolved before its full term expires, and not the Prime Minister or the Monarch.


Craig Prescott

24th January 2021


[1] Evidence to the Constitution Committee - https://committees.parliament.uk/writtenevidence/458/html/;
Evidence to the Public Administration and Constitutional Affairs Committee - http://data.parliament.uk/WrittenEvidence/CommitteeEvidence.svc/EvidenceDocument/Public%20Administration%20and%20Constitutional%20Affairs%20Committee%20/Status%20of%20resolutions%20of%20the%20House%20of%20Commons/Written/91689.html

[2] The 1910 (December) and 1935 Parliaments are wartime exceptions to this, their terms were extended by annual legislation throughout the duration of World Wars I and II respectively.

[3] Such as Wilson in 1970, Margaret Thatcher in 1983 and 1987, and Tony Blair in 2001 and 2005. Edward Heath’s decision to call an election in February 1974, is the only post-war example of a Prime Minister who took office with a working majority following an election, making a request for a dissolution in the third year of the term. Of course, Heath’s primary concern was not necessarily opposition in Parliament, but from outside Parliament, from the trade unions.

[4] Both John Major and Gordon Brown are examples of this. They both let the 1992 and 2005 Parliaments respectively run into their fifth year and lost the election.

[5] Between 1966 and 1992, elections have been held on the following dates, 31st March 1966, 18th June 1970, 3rd May 1979, 9th June 1983, 11th June 1987 and 9th April 1992. The two 1974 Elections were held on 28th February and 10th October. All of these dates are also Thursdays, continuing the unbroken practice since 1935 that poll day falls on a Thursday.

[6] Between 1997 and 2010, elections have been held on the following dates, 1st May 1997, 5th May 2005, and 6th May 2010. The 2001 election was scheduled for 3rd May, but due to the foot-and-mouth outbreak, it was delayed to 7th June 2001.

[7] There are practical reasons in favour of holding elections in May. It allows for voting at general elections to be combined with scheduled local elections, the weather is likely to be better than at other times of the year, there are longer hours of daylight, and it does not fall within a school holiday period.

[8] Harold Macmillan took office in January 1957, but waited until October 1959 before requesting a dissolution, in the fifth year of the term. John Major took office in November 1990, three and half years into the term, waiting until March 1992 to request a dissolution, winning the election that April. However, Gordon Brown on taking office in June 2007, just over two years into the term, declined to hold an early election, losing office at the 2010 General Election that took place towards the end of the parliamentary term.

[9] Fixed-term Parliaments Act 2011, s 1(3)

[10] ibid, s 1(4).

[11] It is perhaps one of the paradoxes the 2011 Act that, considering its name, the maximum term of parliaments provided by the Act vary more than the Septennial Act 1715 which it replaced. In this sense, the 2011 Act would be more accurately named as the “Polling Dates Act” as that is its primary concern.

[12] HC 976, 2013-14.

[13] It is also the case that other rules are predicated on the notion of fairly regular sessions of Parliament. The Parliament Act procedure requires one year to have passed between the day of second reading of a bill in one session, and its third reading in the second session. In the House of Commons, Standing Order No 14 provides (inter alia) for twenty days in each session when the business of the opposition parties has precedence over government business, and thirty-five days for backbench business, with twenty-seven of those days taking place in the chamber. The longer a parliamentary session, the thinner these days are spread out, with any additional time in the gift of the government.

[14] As legislation, it also requires the approval of the House of Lords, but it is difficult to envisage circumstances when the unelected House would block the desire of a majority of the House of Commons to hold a general election.

[15] Public Administration and Constitutional Affairs Committee, The Role of Parliament in the UK Constitution Interim Report - The Status and Effect of Confidence Motions and the Fixed-term Parliaments Act 2011 (HC 2017-19, 1813) para 35.

[16] HM Government, Cabinet Manual (1st edn, 2011) para 2.12.

[17] The full test is, first, what are the precedents; secondly, did the actors in the precedents believe that they were bound by the rule; and thirdly, is there a good reason for the rule. Sir Ivor Jennings, The Law and the Constitution (5th edn, London University Press 1959) 136.

[18] The government could have sought parliamentary approval for a mandate for different stages of the negotiations, in order to prevent a big showdown vote on the deal”.

[19] Should one have emerged, as was speculated at various points during the autumn of 2019.

[20] Although perhaps not a typical election, the fact that it was held in December did not appear to significantly harm turnout, at 67.3% (1.5% less than June 2017), the second highest turnout since 1997. Increasing use of postal voting, also makes casting a vote easier for many regardless of the time of year.

[21] During World War II, this was achieved by passing annual Prolongation of Parliament Acts from 1940-1944.

[22] See s 1(3). In addition, if an early general election has been held before the first Thursday in May of that year, then the approach of 2011 Act at s 1(4) could be retained so that the default election date is the first Thursday in May in the fourth calendar year after the election, with Parliament dissolving as a matter of law twenty-five working days before that date.

[23] National Assembly for Wales (Representation of the People) Order 2007/236, Art 148; Scottish Parliament (Elections etc.) Order 2015/245 (Scottish SI), Art 84

[24] Government of Wales Act 2006, s 3(1A)(a), as inserted by Wales Act 2017, s 6; Scotland Act 1998, s 2(2A)(a) as inserted by Scotland Act 2016, s 5(3); although there is no equivalent provision for Northern Ireland.

[25] See for example, HC Deb, vol 664, col 140, 3rd September 2019.

[26] HC Debs, 4th September 2019, vol 664, cols 291-315; 9th September, vol 664, cols 616-639; and 28 October 2019, vol 667, cols 54-79.

[27] There is a genuine question as to whether this should be a majority of seats, and so require 326 votes in favour. Requiring a majority of votes would more closely reflect the government’s intention of restoring the maximum flexibility afforded by the prerogative.

[28] Erskine May, Parliamentary Practice (25th edn, 2019) para 13.12, cited by the Supreme Court in R (Miller) v Prime Minister [2019] UKSC 43 [67].

[29] R (HS2 Action Alliance Ltd v Secretary of State for Transport [2014] UKSC 3 [206]-[208].

[30] R (Miller) v Prime Minister [2019] UKSC 43 [46].

[31] [2019] UKSC 43 [68].

[32] See, Anisminic Ltd v Foreign Compensation Commission [1969] 2 AC 147 and R (Privacy International) v Investigatory Powers Tribunal [2019] UKSC 22.


[33] The Cabinet Manual is silent on this point, although a draft of the Cabinet Manual at para 44, states that a vote on the Queen’s Speech has traditionally been regarded as a [vote] of confidence. HM Government, Draft Manual December 2010 (2010)


[34] Indeed, in the section headed “Background”, the document uses the phrase “non-legislative constitutional principle”, but then only discusses “principles”, is there a difference?

[35] For that matter, the Cabinet Manual only hints at this, but does not make this explicit. At para 2.7, states that “Prime Ministers hold office unless and until they resign. If the Prime Minister resigns on behalf of the Government, the Sovereign will invite the person who appears most likely to be ablate command the confidence of the House to serve as Prime Minister and to from a government”. Similarly, at para 2.12 it states “Where an election does not result in an overall majority for a single party, the incumbent government remains in office unless and until the Prime Minister tenders [their] resignation”; HM Government, The Cabinet Manual (1st edn, 2011).

[36] In R (Miller) v Prime Minister makes clear at [30] that none of the appeals suggested that the Monarch should have done anything other than act on the advice of the Prime Minister, and the Supreme Court did not develop this point further.

[37] HC Deb, 4 September 2019, col 291–316.

[38] This could happen if in the future, one political party is in such a position of strength, that more frequent elections allow that party to considerably outspend the others, further consolidating their strength.

[39] This would be to prevent any leadership contest arising in the first place. If a Prime Minister sought to request a dissolution once they had lost the support of their own party, then this request could be refused on the basis that someone else is better placed to command the confidence of the Commons. This is made clear in the New Zealand Cabinet Manual, at paragraph 6.63, which states that “the Governor-General will act on the advice as long as the government appears to have the confidence of the House and the Prime Minister retains support as the leader of that government”.

[40] The only other restriction imposed on Counsellors of State by the Regency Acts themselves is that no powers granting, any rank, title, or dignity of the peerage can be delegated to them. The Letters Patent appointing Counsellors of State can impose restrictions on their powers. The text of the last Letters Patent to be issued in 2015, provided that Counsellors of state could “…approve and sign on Our behalf Letters Patent signifying the Royal Assent to Acts of Parliament except any Act touching any of the matters for which provision is made in the Act of Settlement or touching Our Royal Style and Titles”.

[41] Ken Clarke, Kind of Blue (Macmillan 2016) 312.

[42] Peter Hennessy, ‘The British constitution shimmers through: the ‘hung’ general election of May 2010’ in Andrew Duff (ed) Making the Difference: Essays in Honour of Shirley Williams (Biteback 2010).

[43] And for that matter in the lead up to the 2015 General Election, when another hung parliament was a possibility.

[44] New Zealand Cabinet Manual, paras 6.62-6.64, available at [https://dpmc.govt.nz/our-business-units/cabinet-office/supporting-work-cabinet/cabinet-manual/6-elections-transitions-6].