Carl Gardner, BPP University of Law – written evidence (FPA0006)
- I am a barrister, not currently practising. I worked as a government lawyer from 1995 to 2008 including postings at Cabinet Office European Legal Advisers, and the Attorney General’s Office. I now teach constitutional and EU law at BPP University Law School. I’ve written about law at my blog, “Head of Legal”, and in 2015 I self-published an e-book called What a Fix-Up! about the Fixed-term Parliaments Act 2011.
- I’m not a member of any political party, though I was in the Labour party for over 20 years until 2016. I voted Remain in the 2016 referendum, and have recently taken part in public demonstrations in favour of a “People’s Vote”.
- I’ll address the committee’s questions in turn.
To what extent has the Fixed-term Parliaments FTPA 2011 led to a meaningful transfer of power from the Prime Minister to the House of Commons, removing “the right of Prime Minister to seek the Dissolution of Parliament for pure political gain”?
- I don’t think the FTPA has really shifted power from the Prime Minister to the House of Commons. In fact, in some ways I think it’s shifted power to the PM.
- Those who support fixed terms in principle used to complain that, under the old system, the PM could unilaterally call an election at any time for pure political gain. But in my view this complaint was overdone.
- I can’t think of an actual election in my lifetime that was widely perceived as having been an abusively called. In reality, Prime Ministers’ requests for dissolution tended to be uncontroversial. Either a confident PM would ask for and be granted an election after about four years (e.g. 1983, 1987, 2001 and 2005), or an embattled PM would wait it out till much nearer the full five-year term (e.g. 1992, 1997 and 2010). Finally, on some occasions a PM with a very small majority or none at all would call an election to try to get a working majority (e.g. 1966 and October 1974).
- The risk of abuse was in my view mitigated by the background understanding that the Queen had power to refuse a dissolution, applying the “Lascelles principles” or something like them. In my view, this procedural safeguard (combined with public, political and media pressure) helped encourage reasonable behaviour from politicians. I think the role of the Queen under the old system was a good example of a “Chesterton fence”—an institution that reformers cleared away in 2011 before they understood its almost imperceptible function.
- I agree that the FTPA can be argued to have shifted power away from the PM in the sense that it removes the risk that an aggressive PM can, out of the blue, unilaterally secure a snap election in dubious circumstances against the wishes of the opposition by prevailing upon the Queen to grant one. I accept that this can now, in late 2019, be seen as a real benefit. The prorogation of Parliament about to take place unfortunately implies that the Queen’s power to block improper constitutional behaviour may no longer be the safeguard I’ve hoped it was.
- But even this modest benefit depends on assumptions that either are not or may not be justified.
- First, it wrongly assumes the requirement for two thirds of MPs to back a PM who wishes to “call” and election is a fairly strong procedural constraint on him or her. This assumption was undermined when Theresa May “called” the 2017 election, to the surprise even of some well-informed observers who were reasonably familiar with the FTPA. In many if not most circumstances the opposition will welcome an election, or feel they must appear to. The threshold of two thirds of MPs under section 2(2) is therefore only a weak constraint on the PM’s power to call an election. In my view, the 2017 election is the best example in modern history of a PM causing an election to be held for pure political gain.
- In addition it assumes the PM is unlikely to engineer a vote of no confidence in his or her own government, under s2(3) and (4), because of the political price the government could expect to pay at the ensuing election. The House of Commons Public Administration and Constitutional Affairs Committee has taken this view in its conclusion 5:
a governing party seeking to bring about a general election through the “no confidence” mechanism provided by Section 2(3) of the Fixed-term Parliaments Act 2011 in order to circumvent the requirement for a majority of at least two-thirds under the Section 2(1) risks a political penalty at the ballot box; and this is likely to be enough of a deterrent to prevent any Government from using section 2(3) to bring about a general election. It is our view that, while legally possible, it would be entirely inappropriate for a Government to use the simple majority route to a general election under section 2(3) to circumvent the requirement for a two-thirds majority in Section 2(1). However, it must be accepted that there is nothing in law to prevent such an abuse.
- I disagree with PACAC’s assessment, and note the last sentence quoted above. I think a PM would only want an election in the first place if the governing party was popular, and the opposition weakened. I think it’s therefore unlikely there would be a significant political deterrent. I note that this kind of device has been used by governments in Germany, and in my view it’s only a matter of time before a British government uses the FTPA in this way.
- So in my view, it’s not clear that the FTPA has really reduced the PM’s power to initiate an election. In fact by removing the Queen’s reserve power to block an election, it has arguably increased the PM’s power.
- In my view the FTPA has also transferred power to a PM who is losing or has lost the confidence of the Commons and is threatened by, or has lost, an opposition-initiated FTPA no confidence motion under s2(3) and (4).
- Of course the PM does not control whether and when the opposition tables a no-confidence motion. But from the moment such a motion succeeds—once the House of Commons enters the ensuing 14 day period—the FTPA’s practical political effect is to shift power significantly towards the PM as compared with the pre-FTPA situation. This is for two reasons: first, confusion about how traditional conventions apply in this situation creates space for government to interpret and apply the FTPA in a self-serving way; and second, the fact that there is no longer any institutional “gatekeeper” with a reserve power to prevent this.
- Before the FTPA, a Prime Minister defeated on an issue of confidence had to either (a) ask for and be granted a dissolution by the Queen, or (b) resign in favour of a new government led by the MP most likely to command the confidence of the House.
- The FTPA clearly changes the first of these bullet points. A Prime Minister defeated on a motion of confidence can no longer ask the Queen for a dissolution, and a fresh election.
- The key question on which there is both confusion and disagreement is how the FTPA now interacts with convention, in particular whether a conventional duty to resign does or can still arise during the 14 days. Another way of looking at it is to say that there is confusion and disagreement about how strongly or weakly Parliament intended the FTPA to fix Parliaments.
- Since as I draft this written evidence in late August 2019 there has not yet been an FTPA no confidence motion, we have no evidence from political practice about which of several possible approaches will prevail. How much power the FTPA has transferred to the PM depends at least in part on which does prevail. I think it will help to convey my point if I take some time to set out what those possible approaches are.
- One approach is to say a Prime Minister defeated on a vote of no confidence under section 2(3)(a) is entitled to freely choose (a) to stay in power for 14 days and so cause an election, or (b) to resign in favour of a new government. I call this the free choice theory.
- Some proponents of this view explain it by saying the old convention on no confidence votes is “unchanged”, or “embodied” or “codified” in the FTPA; others explain it by saying the old convention has been abolished in its entirety, and replaced by a pure statutory scheme. This doesn’t so much imply that Parliaments are intended to be relatively weakly fixed, as that the term of each Parliament is intended to be fixed or not at the absolute discretion of a PM who’s lost a no-confidence vote.
- The key point about this theory is that it denies that the PM can be under any duty to resign during the 14 days. On this approach, a no-confidenced PM has an absolute right to an election if they want one.
- The alternative approach is to think that at some point within the 14 days, the government still has, or can have depending on circumstances, a duty to resign. According to this type of theory, the resignation part of the traditional convention is unaffected. It’s only the other part of the convention—the option of asking for a dissolution—that has been removed.
- There are a number of possible variants of this broad duty to resign approach to the FTPA, depending on when you think the duty to resign arises. Which variant you prefer depends partly on your understanding of the functioning of the traditional, pre-FTPA conventions (about which there is room for disagreement), and partly on how strongly or weakly you think the FTPA is intended to fix Parliaments. I’ll sketch these variants briefly.
- The immediate duty to resign theory says a Prime Minister defeated on any matter of confidence should immediately resign (in favour of a new government led by the MP most likely to command the confidence of the House). This implies strongly fixed Parliaments.
- More commonly, it’s thought a duty to resign arises at a conceptually later point. That point is either when the PM realises confidence in the government can’t be restored (which I call the less deferred duty to resign theory); or when it’s clear an alternative government is better placed than the incumbents to gain the House’s confidence, or that an alternative government does or would command a majority (I call these the more deferred duty to resign theories).
- Recently I’ve begun to think it’s important to distinguish between the last two theories, which we could call the somewhat more deferred duty to resign, and the much more deferred duty to resign.
- I think it’s important to be explicit that which (if any) variant of a duty to resign theory prevails in constitutional practice will determine how likely it is that a duty to resign will arise at all during the 14 days. If you think the PM should resign immediately on losing a no confidence vote, then by definition the duty must arise on day one. If you think the PM should resign only if it’s clear within the 14 days that confidence can’t be restored, then it’s quite likely that the duty to resign will arise, or “crystallise”. If on the other hand you think there’s only such a duty if someone else is better placed to govern, or only when someone else has a majority, then it’s decreasingly likely that the duty will arise within the 14 days.
- The duty to resign theories therefore represent a spectrum, from more strongly fixed Parliaments (if the PM should resign immediately on losing a no confidence vote) through less strongly fixed Parliaments (if there’s a less deferred duty to resign, when the PM can’t restore confidence), and less weakly fixed Parliaments (if any duty to resign is somewhat more deferred, till someone else is better placed to govern) to more weakly fixed Parliaments (if the duty to resign is much more deferred, till someone else actually has a majority).
- PACAC, at conclusion 7, seems to reflect a “much more deferred duty to resign” approach, and therefore weakly fixed Parliaments,
It is clear that, during the 14-day period following a vote of no confidence under Section 2(3), the Prime Minister is under a duty not to resign unless and until it is clear another person commands the confidence of the House. It is also clear that in the event that it becomes apparent that another person could command the confidence of the House the Prime Minister would be expected to resign.
- But it’s fair to say that the committee’s use of the word “could” in the second quoted sentence may be thought to introduce a slight ambiguity. Seeing a duty to resign in favour of someone who only could command the House, rather than already clearly doing so, would imply only a somewhat more deferred duty to resign. I think there is a real distinction here that makes a difference, and that it’s important to avoid all possible ambiguity.
- The Cabinet Manual addresses what should happen during the 14 days at paragraph 2.19. It says
The Prime Minister is expected to resign where it is clear that he or she does not have the confidence of the House of Commons and that an alternative government does have the confidence.
- This also implies the much more deferred duty to resign theory, and therefore weakly fixed Parliaments.
- But it’s also possible to find a tiny hint of ambiguity in the Cabinet Manual. At paragraph 2.12, which deals not with the FTPA 14 days but with the arguably analogous situation of a hung Parliament immediately following an election, it says
An incumbent government … is expected to resign if it becomes clear that it is unlikely to be able to command that confidence and there is a clear alternative.
- A “clear alternative” is not necessarily the same thing as an alternative government that does have the confidence of the House. A practical example of this is given by Harold Wilson’s incoming minority government in 1974. It would help if future editions of the Cabinet Manual made it explicit whether the government thinks duties to resign arise at different points in the two analogous situations, and if so why.
- Let me return to the question of whether the FTPA has shifted power away from the PM.
- In my view the theories I have outlined imply differing relationships of power between the PM and government on the one hand, and the rest of the House on the other. Broadly, the more fixed you think Parliament intended Parliaments to be, the more you constrain the PM’s power.
- Of the duty to resign theories, in my view the ones that constrain the power of the PM most are those that see his or her duty to resign as arising conceptually earlier within the 14 days. These theories reduce the PM’s power as compared with the pre-FTPA position. This is the reason why I favour the less deferred duty to resign theory, that the PM should normally be expected to resign as soon as it’s clear confidence cannot be restored—regardless of whether anyone else has a majority or is better placed than the incumbent. This approach would put more power in the hands of other MPs, since it would enable them to give a new, potentially fragile government a chance to establish confidence, rather than have an election. I think it’s unfortunate that PACAC and the Cabinet Manual seem to have decided against this approach.
- The theory PACAC and the Cabinet Manual seem to prefer—the much more deferred duty to resign theory—does not in my view shift power away from the PM. It’s more accurate to think of it as shifting the safeguard power the Queen used to have to block an election in limited circumstances under the old system from Her Majesty to the House of Commons, without affecting the PM’s power. I accept that this could, if applied in practice, represent a gain in the legitimacy and (particularly given the current prorogation controversy) the effectiveness of the safeguard.
- The free choice theory that the PM has not duty to resign at all during the 14 days, but rather is simply free to choose whether to resign or to “call” an election by running the clock down would, if applied, in my view significantly shift power to the PM as compared with the pre-FTPA position. This is because it removes two constraints on the PM: the old duty to resign that could arise under convention, and any right of another political institution (whether Queen or Commons) to block an election sought for pure political gain.
- An essential insight in my view is that the FTPA has undoubtedly shifted power to the PM in the sense that all the theories I’ve explored are at least superficially plausible, and the FTPA provides no mechanism whereby the House of Commons, or the Queen, or any other institution, can enforce and apply its preference among them. The FTPA has created the situation where the PM and government will decide which theory it seeks to apply, and it’s not easy to see how its preferred approach can be resisted except, improbably but just conceivably, in the courts.
- I note that a spokesman for the current government is reported as having made statements that imply it intends to apply this theory of the FTPA. In my view this approach is worrying, and if successfully applied in practice will mean the FTPA has shifted power from other institutions to to the PM in an important respect.
- This situation is all the more troubling given that section 2(7) of the FTPA in effect gives the PM power to time the ensuing early election, constrained not by any other political institution but only by principles of public law enforceable by the courts.
- Some reassurance might be found in the so-called “caretaker” or “purdah” convention, if it is felt that ministers can safely be relied on to respect it. But a government that had decided to apply the FTPA in the most self-serving way, ignoring conventions opponents think applicable, might not respect the caretaker convention either, or else interpret and apply it too in a self-serving way.
- These dangers are intensified where a government has already used prerogative power in a way thought by many to be improper, unconstitutional or even testing the boundaries of lawfulness.
- In conclusion, my view is that the FTPA has not led to a meaningful transfer of power from the Prime Minister to the House of Commons.
- It has transferred power to the PM if he has lost the confidence of the Commons by (a) enabling the PM discretion to decide, without any check from other political institutions, how the FTPA interacts with conventions during the 14 days following an FTPA no confidence vote; (b) admitting of a superficially plausible (thought flawed) reading that can be applied by the PM during the 14 days for political ends, and (c) removing any institutional check on the PM’s ability to force an election on the expiry of 14 days if he or she chooses, even in extreme circumstances.
- It has also transferred power to the PM if he or she still has a majority by allowing him or her to obtain an election at will by engineering a “self no confidence” motion, subject only to the Speaker’s acceptance of such a motion.
- Finally, the FTPA initially transferred power to the PM in that, since it seemed not to be possible for him or her to call an election, it removed the pressure PMs have sometimes faced to seek a mandate. For example, Theresa May was initially free of the expectations that Gordon Brown faced in 2007. This advantage has now diminished however, since the surprise of the 2017 election.
Is five years the appropriate length for fixed-terms between general elections?
- I have no view on this.
Does the certainty of knowing when the next election will be – notwithstanding the section 2 provisions for triggering an early general election – have an impact on good governance?
- I don’t believe there is any such certainty. I don’t think the FTPA has any positive or negative effect on governance generally.
Are the mechanisms in the FTPA to trigger an early general election appropriate?
- The two-thirds majority rule for government to initiate an election in section 2(1) and (2) is in my view acceptable looked at in isolation. It represents a form of institutional check on the government’s power to call an election.
- Good constitutional behaviour and governmental restraint might be more assured if there were a shared understanding—a convention, perhaps—that no early election should be triggered except as a result of a section 2(1) and (2) motion, or as a last resort. I will return to this point later.
- The provisions relating to votes of no confidence in sections 2(3) and (4) are not acceptable as they stand, for the reasons I have already set out.
What impact has the FTPA had on the notion of the House of Commons having “confidence” in a Government? Is it still possible for the Government to make a vote in the House of Commons on a matter of policy a “confidence” issue?
- In my view it has had a damaging effect.
- Most importantly, I have heard the view expressed that the FTPA abolishes no confidence motions and confidence issues except those in the statutory section 2 forms. Indeed, the committee’s question implies that that may be a plausible view.
- In my view it is not seriously tenable. There is no basis for thinking the FTPA is intended to be an exhaustive code in this area, rather than governing when an early election can take place.
- Nonetheless, the idea has some superficial plausibility to people unaccustomed to thinking in terms of constitutional conventions, who may be attracted to the idea of applying apparently simple, literal rules. It’s entirely understandable that very intelligent, well-informed observers can think in this way, particularly where Parliament has legislated in a field previously governed by convention. In other words, the FTPA has damaged public understanding of the constitution and loosened the political force of constitutional conventions.
- Again, the biggest danger arises from the fact that the FTPA enables the government to adopt exactly this view of the FTPA and to apply it, without any institutional check. The House of Commons may believe that the government should resign if, for instance, it is defeated on the Queen’s speech. So may the Speaker. But there is no mechanism for either to enforce their view on the government.
- Nor is it obvious how the courts could do so, since the courts do not enforce constitutional conventions.
- The FTPA has therefore created a serious risk that an aggressively self-serving government could decide in its own political interests to refuse to resign even if defeated on a Queen’s speech, for instance having lost its majority at an election. We don’t know, at a time when the electorate is deeply divided, whether the government would pay any political price for this.
- There is a certain flawed intellectual consistency in adopting this very pro-executive view of non-FTPA confidence matters if one is already committed to a “free choice” theory of FTPA no confidence votes. Each theory is connected to the idea that the FTPA is an exhaustive constitutional code in the fields it touches, bleaching out all traditional conventions.
- I am afraid that the current government might adopt this approach, and refuse to take any action after losing a non-statutory no confidence vote. At worst, this approach could lead to a government losing an FTPA no confidence motion and insisting on running down the clock to an election for pure political gain even where a united opposition has a clear majority and wants to govern, and where the House of Commons has in addition, within the 14 days, formally resolved that the government must resign in favour of the opposition.
- I note that PACAC, at conclusion 11, concludes that there is still a conventional duty to resign following such a vote. I strongly agree with PACAC on this.
What challenges arise for the political parties, the House of Commons and the civil service in the 14-day period following the passing of a motion of no confidence in the Government? Is there a risk of the monarch being drawn into the political debate during this period and, if so, how should this be mitigated?
- I don’t have a view on challenges to political parties. I will simply say in passing that I believe political parties as institutions, and their members in particular, have recently started to play a disproportionately powerful and damaging role in our constitution.
- The challenge for the House of Commons is how to enforce its will under the FTPA—even the will of a clear majority—against the PM.
- The challenge for the civil service is how to advise on what’s a matter of constitutional duty and propriety rather than law. Ministers attracted to the most self-serving theories of the FTPA may simply reject the idea that civil servants can legitimately advise them on matters that in their view are purely political.
- I think there is a particular challenge for the Law Officers and their staff. They are traditionally expected to advise on the propriety of government action as well as its lawfulness. There must be a danger either that Law Officers or their officials might feel obliged to resign over disagreements about propriety involving the FTPA; or that Law Officers might lose public respect; or in any event that the role of the Law Officers may again come under scrutiny, as it did after the invasion of Iraq.
- I think there is a risk that the Queen might be drawn into politics, for instance if the apparent inability of the Commons to impose its will on government (and perhaps the unwillingness of the courts to come to its aid) creates public pressure on the Queen to sack the government. Another risk is that an aggressively self-serving government tries to use the Queen to prorogue Parliament to avoid non-statutory votes of censure during the 14 day FTPA period.
- I see three ways of mitigating these risks. The first is for political actors to develop a shared understanding that the FTPA will not be applied in the aggressively pro-executive way I fear; but of course this requires the cooperation of ministers. Alternatively, the FTPA could be repealed. Finally, the FTPA could be amended to correct its flaws.
- If Parliament formally, or MPs and peers politically, fail collectively to mitigate these risks , then I think it’s inevitable that the courts will face pressure from claimants to make rulings that could significantly change the character of our constitution from one that’s mainly politically shaped to one that has much greater legal content. As a constitutional conservative or centrist who has long had faith in the British model, I would regret that. But it would be far preferable to seeing the country slide into a sort of authoritarianism in which the executive was able to increasingly assert its dominance over Parliament.
If the FTPA was repealed, what provisions for the lengths of Parliaments and the timing of general elections would need to be made in its place? Would the prerogative power for the Prime Minister to dissolve Parliament and call a general election be revived in the event of repeal?
- It’s often argued that it’s legally impossible to return to the pre-FTPA position. In my view this is simply wrong.
- Repeal would not create a new prerogative power, or extend one. All it would do is remove the statutory “rug” to uncover the common law floor underneath it, to use Francis Bennion’s image. So long as the repealing legislation clearly disclosed Parliament’s intention to disapply section 16(1)(a) of the Interpretation Act 1978, the result of repeal would be to “revive” the Queen’s prerogative power to dissolve Parliament.
- This is supported by R v Secretary of State for Foreign and Commonwealth Affairs, ex parte CCSU  IRLR 309. In that case, claimants successfully argued that a prerogative power (in that case, ministers’ prerogative to manage the civil service) revived when a 1927 statute abridging it was repealed.
- I don’t know of any authority to the contrary; if any such precedent is identified, I’d need some time to think about it before giving my view of it.
- In my view, repeal of the FTPA and restoration of the previous position is probably Parliament’s best option. It would remove the confusion caused by the FTPA, and in principle restore a reasonable balance between government and opposition, without the risks involved in an attempt at better, more detailed codification. I have little trust that this Parliament or its immediate successors could agree a clear, wise statutory scheme that attracted consensus.
- I do think however that the FTPA has done lasting damage to the constitution. It will be difficult for repeal to put the genie back in the bottle, because so much confusion has now been sown about the traditional constitutional conventions in this area, and because the current prorogation crisis has helped encourage an already growing, dangerous belief that the Queen must always comply with the government’s wishes even in extreme circumstances.
- So repeal would work best if accompanied by some fresh statement or understanding of the principles on which the Queen would exercise her reserve power to refuse a dissolution. This could resemble the Lascelles principles or be modelled on the approach taken in other democracies.
- If Parliament chooses instead to amend the FTPA, then it should consider the following improvements. First, it should make explicit provision for what should happen within the 14 days. Second, it should remove or constrain the PM’s power to unilaterally set the election date.
- It’s important that the House of Commons should be able to remove the government and install a new one within the 14 days, should it wish to, thus avoiding an early election. As we have seen, one of the FTPA’s flaws is that it provides no enforceable mechanism for this to be achieved. This will need careful thought, and I have suggestions for consideration, rather than fully worked-out solutions.
- One option would be to make clear that the government must by law resign with immediate effect, if the House of Commons so resolves. An important question in relation to this is whether such a resolution should be required to nominate the MP who should be summoned by the Queen to form a new government. Provisions such as these would provide an enforcement mechanism for the Commons, and help prevent pressure on the Queen to sack the PM.
- Another option would be to set down in statute a mandatory structured timetable of events within the 14 days. For example, an amended FTPA might require a no confidenced government (if it had not resigned) to attempt a clock-stopping vote of confidence within 5 or 7 days, and resign if it lost. Or it might require that government to choose to try a clock-stopping vote of confidence or else an early election motion under section 2(1) and (2), and resign if it lost. Or it could require them to try both, resigning if they lost both. These could be practical ways of putting on a statutory footing something like the less deferred duty to resign theory.
- It might be useful to make provision for a period (the middle five days, for instance, or days 7 to 10) during which the Commons must consider and vote on potential alternative governments. Something like this was suggested by Robert Craig in a recent blogpost.
- It would and should always be possible if the FTPA is retained for the Commons to vote for an early election under sections 2(1) and (2) of the FTPA. Provision might be made to compel the Commons to vote on such a motion at some point within the 14 days.
- It might be beneficial if provision were to be made (or a convention were to develop without statutory provision) that all these other options should be tried before the expiry of the 14 days, so that an automatic early election could only happen then as a truly last resort.
- Of course MPs and peers should consider how analogous provisions work in other democracies, and in our own devolved settlements, and consider adopting provisions that would appear workable in the Westminster context.
- Finally, Parliament should consider amending section 2(7) of the FTPA to remove the power of the PM to unilaterally fix the date of an early election. One possibility would be to create a new institution to fix this date impartially, or to give the function to the Electoral Commission. Another would be to require the House of Commons, or both Houses, to approve the date proposed by the PM.
5 September 2019
House of Commons 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 Fourteenth Report of Session 2017–19, 11 December 2018. https://publications.parliament.uk/pa/cm201719/cmselect/cmpubadm/1813/181302.htm
 Tweet by Robert Peston, August 28 2019 https://twitter.com/Peston/status/1166644250915815424
 Bennion, ed. Jones Statutory Interpretation. 6th edition, LexisNexis London 2013, section 32 page 168