Professor Robert Hazell CBE, the Constitution Unit, Department of Political Science, University College London – written evidence (FPA0003)
This submission is based upon observation of the Fixed Term Parliaments Act in operation, and commentary upon it; but also upon comparative research into the operation of fixed term parliaments in other countries, summarised in a 40 page report on Fixed Term Parliaments published by the Constitution Unit in 2010. Because we have relatively little experience of the operation of our own Act, it is important to draw upon the much greater experience of other countries with fixed terms, in Europe and in other Westminster-style parliaments.
In addition to the issues and questions raised in the Committee’s call for evidence, I hope that the Committee might consider the following:
The power of prorogation
Section 6(1) of the FTPA says ‘This Act does not affect Her Majesty’s power to prorogue Parliament’. The power of prorogation remains a prerogative power, exercised on the advice of the Prime Minister. Its exercise has normally been uncontroversial, to end a parliamentary session. But given the risk of possible abuse, it would seem wise to ensure that Parliament cannot be prorogued against its will. One way to do this would be to make the prerogative power exercisable at the request of Parliament rather than on the advice of the Prime Minister. Another would be to put prorogation on the same footing as the power of adjournment, and to enable Parliament to be prorogued when the House of Commons passes a motion to that effect.
The caretaker convention
The caretaker convention applies when a government cannot command the confidence of Parliament: for example, after it loses a formal no confidence motion under s 2(4) of the FTPA. The convention is not sufficiently clearly explained or articulated in the Cabinet Manual (paras 2.27 to 2.34, on Restrictions on Government Activity). I hope the Committee might make recommendations on strengthening those paragraphs. If required, I could submit further evidence on this point.
Setting the date for polling day following a mid-term dissolution
Section 2(7) of the FTPA provides that following a mid-term dissolution, ‘the polling day for the election is to be the day appointed by Her Majesty by proclamation on the recommendation of the Prime Minister’. Given the recent controversy over the discretion this might confer on Boris Johnson as Prime Minister to choose a polling day after 31 October if he is defeated on a no confidence motion in early September, it might be preferable to provide that the polling day should be approved by the House of Commons, before the Prime Minister recommends a date to the Queen. Or, following the precedent in section 3 of the Scotland Act 1998, where the polling day is proposed by the Presiding Officer, at Westminster it could be proposed by the Speaker. This would have the same effect: the Speaker would want to propose a date that was supported by the House.
Power of Proclamation
The FTPA retains the system of issuing writs for the election, and a proclamation to summon the new parliament and appoint the date of its first meeting. Writs are issued by the Lord Chancellor and Secretary of State for Northern Ireland (s3(3), the proclamation is issued by the Queen, and the appointed day for the first meeting of the new Parliament is chosen on the advice of the Prime Minister. It does seem odd that election writs are still issued by government ministers, and the date of the first meeting of the new parliament is decided by the Prime Minister. The main justification for fixed term parliaments is to take power from the Prime Minister and give it to parliament (see below). Following the same logic, consideration should be given to the Electoral Commission issuing the writs for the election, and the Speaker deciding the date of the first meeting. In Tony Wright’s Fixed Term Parliaments Bill 2001 the power to decide the dates for issue of the writs, polling day, and the meeting of the new Parliament was conferred on the Electoral Commission. That seems right in terms of who should control issue of the writs and polling day. But parliament should decide the date of its first meeting, with the decision being made by the Speaker of the outgoing Parliament.
To what extent has the Fixed-term Parliaments Act 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”?
1 There is a belief among some Conservatives that the FTPA is simply the product of a short term political fix to bolster the 2010 coalition government. But fixed term parliaments had been advocated for the previous 20 years, by academics, think tanks and parliamentarians, with three private members’ bills being introduced, in 1994, 2001 and 2007. In the 2010 election manifestos, both the Liberal Democrats and Labour had pledges to introduce fixed term parliaments. The Conservatives did not, but had a more general pledge “to make the Royal Prerogative subject to greater democratic control so that Parliament is directly involved”. There is more detail about the genesis of the FTPA in chapter 6 of our 2010 report.
2 It is right to remind people of the justification for fixed terms. Essentially there are five main arguments in favour:
3 Allowing the incumbent government to call the election when it chooses can give it tremendous tactical advantage; it has been famously compared to an athlete being allowed to fire the starting pistol. Comparative research by Belu and Schleiter has shown the advantage to be significant: “strategically timed opportunistic elections have allowed governing parties to realise an average vote-share bonus of just under 6 per cent and seat-share bonuses of 12 per cent, doubling the probability that the Prime Minister survives in office”.
Reduction of prime ministerial power
4 The power to determine the date of the election is a source of additional power for the Prime Minister over his colleagues. If they threaten to rebel he can in turn threaten them with an early election, as John Major did with the Maastricht rebels. With fixed term parliaments it is harder for a Prime Minister to threaten a snap election in this way.
Better electoral administration
5 The Electoral Commission has long had an interest in fixed term parliaments, which enable electoral administrators to be better prepared because the election date is potentially known long in advance.
Better government planning
6 Fixed term parliaments create an expectation that the parliament will run for the whole term, allowing governments reasonable time to develop and implement their legislative agenda or programme, and reducing short termism. It also removes the debilitating speculation which characterised previous parliaments from the mid-term onwards, wondering when the Prime Minister might call the next election.
Protection of the Crown
7 By minimising the discretionary use of prerogative powers, constitutional crises can be avoided. The King-Byng affair in Canada in 1926 and the Australian constitutional crisis of 1975 demonstrate the difficulties for the Crown when faced with requests for early dissolution.
8 There are two arguments against fixed terms:
Loss of flexibility
9 Fixed terms might limit the capacity of a new Prime Minister appointed mid term to seek a fresh mandate (as Eden did in 1955). And rigidity could lead to lame duck governments, lacking the capacity to govern but not capable of being brought down, as we have seen with Theresa May’s government in 2017-19.
10 Experience in other jurisdictions shows that governments have been able to circumvent the fixed term requirement and call elections at their convenience: as happened in Canada in 2008, and in Germany in 1982 and 2005.
11 The limited British experience suggests the FTPA may also be ineffective: when Theresa May decided to call for an early election in April 2017, she had no difficulty in getting the necessary two-thirds support in the Commons, which voted for early dissolution by 533 to 13. But the Prime Minister can no longer threaten MPs with an early election: it is now at the discretion of the House of Commons to deny one. Whether they have the backbone to do so is another matter; but the power is undeniably in their hands, not the Prime Minister’s.
Is five years the appropriate length for fixed-terms between general elections?
12 Five years is long by comparison with most other parliamentary systems. In the Westminster world, Australia and New Zealand have three-year maximum terms. The legislatures of Canada and many of its provinces have four year fixed terms, as do most Australian states. The devolved legislatures in Scotland, Wales and Northern Ireland were all introduced with four year fixed terms. Ireland’s lower house has a five year maximum, as in the UK.
13 In continental Europe most countries have four year fixed terms, and only three (France, Italy, Luxembourg) have five years. The length of parliamentary terms in other Westminster parliaments and in Europe is shown in Figure 4.1 of our 2010 report.
14 Five years is also long by comparison with Westminster’s recent experience. The average term since 1945 has been 4.0 years. This has remained stable, with an average term length of 4.2 years since October 1974. Analysis of those parliaments which ran for a full term records seven parliaments which lasted around four years (1951, 1966, 1970, 1979, 1983, 1997, 2001); three which lasted four and a half years (1945, 1955, 1974); and five parliaments which ran for five (1959, 1987, 1992, 2005, 2010).
15 The balance between four and five years is thus more even than folk memory might suggest. But, save for the 2010 parliament, those parliaments which lasted for five years did so because the government had become unpopular and did not want to hold an earlier election. The Prime Minister stayed on hoping that his or her party’s luck might change. It did not, save for the case of John Major, who scraped through with a narrow majority in 1992.
16 Politicians and academics who have addressed the issue have all supported parliaments running for four years rather than five. During the debates on the Parliament Act 1911, which reduced the maximum term from seven to five years, Prime Minister Asquith said that this “will probably amount in practice to an actual legislative working term of four years”.
17 Professor Robert Blackburn has also argued for a four year term:
In the UK, there can be little doubt that the period between general elections should be four years. The proposal for fixed term Parliament as a whole should fit as closely as possible into existing constitutional expectations, and the idea that four years is about the right length of time between elections is very prevalent.
18 Previous reform proposals by the political parties have all supported a four year term. This can be found in the Labour party’s 1992 report Meet the Challenge, Make the Change; in the 1993 report of the Labour party’s Plant Commission; in the 1992 and 1997 Liberal Democrat manifestos; and in the Liberal Democrat policy papers Real Democracy for Britain (2006), and For the People, By the People (2007). The three Private Member’s Bills introduced in the last ten years, by the Labour MPs Jeff Rooker and Tony Wright, and the Lib Dem David Howarth MP all supported four year terms (see chapter 6). Indeed, none of the proponents of fixed term parliaments have advocated a term of five years.
19 If Westminster fixed terms were reduced to four years, thought would need to be given to how that would fit with the electoral cycles for the devolved assemblies (and the European Parliament elections if the UK remained a member of the EU). There would need to be consultation with the devolved assemblies about whether they would also revert to a four year term. If Westminster wished to avoid any clash between UK general and other elections, the simplest solution might be to move the date of general elections to October, with the transitional parliament serving a four and a half year term.
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?
20 The Committee may want to hear from Whitehall officials about the impact on good governance. When he was Cabinet Secretary Sir Gus O’Donnell certainly anticipated that the FTPA would have a beneficial effect, allowing the Treasury and other departments to plan their budgets and legislative programmes in reasonable confidence that parliaments would last a full five year term. But in practice we have had only one such parliament, from 2010 to 2015. Since 2016 the disruption and uncertainty caused by Brexit has made any sensible planning impossible.
21 Two further comments can be offered. First, there are some areas of government policy, such as pensions, or long term care, or major projects such as HS2, which require a much longer time horizon than five years, and require a degree of consensus which is difficult to achieve in the UK’s adversarial political system. Second, one set of public officials who would welcome fixed terms are electoral administrators: the Electoral Commission had long commented on the difficulty of preparing for unpredictable general elections, compared with all other elections in the UK which are on fixed terms.
Are the mechanisms in the Act to trigger an early general election appropriate?
22 Section 2 of the FTPA envisages two routes to early dissolution:
23 A motion for voluntary dissolution requires a two-thirds majority of all MPs; while a no confidence motion can be passed by a simple majority. No confidence leads to dissolution only if confidence cannot be restored within 14 days, by the incumbent or an alternative government.
24 It seems right to set a high barrier for the Prime Minister to trigger an early election under section 2(2); but such a high threshold is rare in other parliaments. Figure 7.1 in our 2010 report set out the threshold requirements for confidence motions in 16 other countries in western Europe. In all cases the threshold for a no confidence motion is a simple or absolute majority. In those countries where dissolution can also be triggered by a parliamentary vote, equivalent to FTPA s 2(2), the threshold is the same as for confidence motions.
25 The main incentives against early dissolution are political, as explained below. But there are other ways in which the use of dissolution motions may formally be restricted:
26 It is also quite common to limit or prohibit dissolution towards the beginning or the end of a parliamentary term. The French Constitution prohibits dissolution in the twelve months following a general election (article 12), as does the Spanish Constitution (s115(3)). Other legislatures only allow dissolution in the final year of their four year term, while the South African lower house cannot move to dissolve itself in the first three years of its five year term (art 51(1)(b). At such times, there can of course still be the possibility of confidence motions.
27 A strong disincentive to a government inclined to call an early election is whether a parliament elected mid-term merely serves out the remainder of that term, or whether dissolution restarts the clock and it serves a full new term. Here there is a big difference between the fixed term regime for Westminster and for the devolved legislatures. In Scotland, Wales and Northern Ireland only the remainder of the term is served, unless the dissolution takes place within six months of the next scheduled election. (In that event, the next scheduled election is not held; but the one after that is, to get back on schedule). If Theresa May had been faced with the prospect of the next general election after 2017 still being held in 2020 rather than being shifted back to 2022, she might have made a different calculation.
28 But political incentives are as important in guiding behaviour as legal rules. The main question for political parties in any mid-term crisis will not be whether dissolution is too easy or too difficult, but whether they stand to gain or lose by forcing a dissolution. And much of the time, the political incentives in themselves may prove a further force for stability. Political parties do not like excessively frequent elections: they are expensive for the parties, and exhausting for the participants. And whatever the polls say, the outcome is always slightly uncertain. A party which forces an unnecessary election risks being punished by the electorate, who also dislike frequent elections.
What impact has the Act 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?
29 As the Commons Public Administration and Constitutional Affairs Committee explained in their December 2018 report, the principle that the government must retain the confidence of the House of Commons remains fundamental to our constitution. The House remains free to express its confidence, or lack of it, in any manner it chooses. The government can make any issue a matter of confidence; as can the Opposition. But the only motions now which are formal confidence motions are those in the statutory wording required by section 2 of the FTPA. That happened in January 2019. On 15 January the government lost a vote seeking parliamentary support for Theresa May’s Brexit deal by 230 votes (432:202), the largest defeat for a sitting government in modern history. Jeremy Corbyn immediately tabled a no confidence motion, but on 16 January the Commons rejected it by 325 votes to 306.
30 The government became a zombie government, formally maintaining confidence, but unable to progress the key policy issue of the May administration. But that was more to do with the politics of Brexit, the breakdown of discipline in the governing party, and the weakness of the Opposition, than the requirements of the FTPA. Had the FTPA not been in force the result would have been the same. Theresa May, having survived a confidence motion, was entitled to remain in office; but equally, having failed to deliver a key policy, she was entitled to resign – as subsequently she did.
31 The Queen’s Speech and the Finance Bill are still to be regarded as matters of confidence; as would any matter which the government declares to be a confidence issue. What the FTPA does is create an intermediate step. Instead of defeat on the Queen’s Speech, or the Finance Bill, being seen as requiring the Prime Minister to offer their resignation, it would now lead to a formal no confidence motion tabled by the Opposition. Or – as can happen during the 14 days triggered by a successful no confidence motion under the FTPA – it could lead to the government changing course, in order to restore confidence.
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?
32 The FTPA provides that Parliament is dissolved, leading to an election, unless during the 14 days an affirmative confidence motion is passed, ‘that this House has confidence in HM Government’. That could be confidence in the incumbent government, if the Prime Minister is willing to change policy or personnel in order to restore confidence. Or it could be confidence in an alternative government. The difficulty is that the FTPA is silent on how that alternative government might be appointed before testing whether it can command confidence. If there was support for an alternative government, the House of Commons would have to demonstrate that by passing a resolution to that effect, or a motion on a humble address inviting the Queen to appoint an alternative Prime Minister. That would enable the incumbent Prime Minister to resign, and recommend that the Queen appoint the alternative nominated by Parliament. The Queen should not be expected to guess who might be able to command confidence; she would only act when given a clear signal by Parliament. But there is no clear procedure to enable the House of Commons to make a decision on whom to recommend to the Queen: no clear mechanism, and no guarantee of a debate being given time.
33 One way of formalising this, and filling the silence in the FTPA on how to nominate an alternative Prime Minister, would be to adopt the procedure for appointing a First Minister in Scotland. Under section 46 of the Scotland Act, the Scottish Parliament must nominate one of its members to be appointed as First Minister, and the Presiding Officer then recommends that person for appointment by the Queen. This must happen within 28 days of an election, or the resignation of the First Minister. The procedure for electing the First Minister is set out in Rule 11.10 of the Standing Orders of the Scottish Parliament, titled Selection of the First Minister. Similarly at Westminster, the silence in the FTPA on how to nominate an alternative Prime Minister could be filled by providing for a new procedure in the Standing Orders of the House of Commons.
If the Act 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?
34 The prerogative power of dissolution would not necessarily be automatically revived if the FTPA was repealed. If the monarch’s power to dissolve parliament were to be restored, it would have to be re-created in statute; and it might be difficult to define the circumstances in which a Prime Minister could properly request a dissolution. The prerogative power of dissolution is one of the most controversial prerogative powers, as shown in the 1975 dismissal of Gough Whitlam in Australia, and more recently the request in 2008 for an early dissolution by Stephen Harper in Canada. The statute might simply have to side step the issue, by saying that ‘the Prime Minister may request a dissolution of Parliament, which may be granted by the Sovereign’; but that would not avert a lot of discussion in Parliament about when a dissolution could or could not properly be granted.