How to change the government without causing a general election

David Howarth explains how the Fixed Term Parliaments Act 2011 has altered the options for no confidence motions in Parliament, and how an opposition party might form a government without there being a general election.

Picture: Luke Stackpoole on Unsplash

The Fixed Term Parliaments Act 2011 removed the ability of Prime Ministers to call early general elections and handed that power to Parliament. Parliament has two options: a motion supported by two-thirds of MPs (434 votes) that an early election should take place; or a motion of no confidence in Her Majesty’s Government, in the precise wording set out in s. 2(4) of the Act, carried by a simple majority, which is followed by a 14-day period in which the Commons can install a new government (or re-install the previous government) by a precisely worded motion of confidence.

The early election mechanism was used in 2017, on a motion put down by the government and supported by the opposition. The no confidence route has not yet been used but it is already having an important if paradoxical effect: some, including the former government Chief Whip, Mark Harper, argue that, despite what the government said at the time, the Act now constitutes the only way in which the House of Commons can declare its lack of confidence in the government, so that all the previous conventions about governments having to resign if defeated in a vote of confidence worded in a variety of ways or just declared by the participants to be a motion of confidence no longer apply. The argument is that the FTPA has replaced not only the previous method for calling a general election but also the previous method for replacing a government without calling a general election. The argument flows from the view that oppositions should not be permitted to move motions of confidence without risking a general election. Under the old system that risk existed because the government could choose between resignation and calling an election. Now, according to this view, because s. 3(2) of the Act abolishes any other way of dissolving Parliament, the old conventions would lead to governments having to resign with no prospect of an election, which would result in the possibility of risk-free motions of confidence. To avoid this result, supporters of Mr Harper’s view contend that the only possible method of declaring no confidence should be the one in s. 2(4) of the Act, a method that still carries the risk of an early election after the 14-day period.

Serious difficulties arise, however, if one adopts the view that the government cannot be changed except through the no confidence and 14-day process. First, as the Cabinet Manual illustrates, the doctrine still exists that after a general election an incoming Prime Minister must demonstrate that she commands the confidence of the House – usually by winning approval for a Queen’s Speech. What if, after an inconclusive election, the incoming Prime Minister loses on the Queen’s Speech? Will that now make no difference unless the Commons additionally passes a vote of no confidence? If so, what, apart from the inconveniences that flow from the Commons’ Standing Order 51 on Ways and Means motions, would be left of the requirement to pass a Queen’s Speech? No dissolution can follow, because of s. 3(2), but Mr Harper’s view implies that the Prime Minister would not need to resign either. The Queen’s Speech would have become irrelevant. That seems an unlikely consequence of the Act.

Perhaps more obviously, does the Fixed Term Parliaments Act mean that governments may no longer resign? If a government lost on supply, losing the Finance Bill or the Supply and Appropriation Bills, for example, so that at the end of the financial year the administration of the state would largely grind to a halt, would it have to carry on in office unless defeated on a no confidence motion? That seems very unlikely. And it seems just as unlikely that a government can no longer announce that it is treating a certain vote as a matter of confidence and will resign if defeated. Those involved with the debate on the Fixed Term Parliaments Bill certainly envisaged that the government retained the power to resign (see e.g. P. Norton, ‘The Fixed-term Parliaments Act and Votes of Conﬁdence’ (2016) 69 Parliamentary Affairs 3, 14).

Most fundamentally of all, does the Fixed Term Parliaments Act mean that the monarch no longer retains a reserve power to dismiss the ministry? As the Cabinet Manual points out, the last time a monarch exercised that power was in 1834, in circumstances that turned out to be something of a disaster. But the power is generally still acknowledged to exist, and it would be extraordinary for it to have been abolished purely by implication. Indeed, one can argue that without it the Fixed Term Parliaments Act itself could not work. What if, following a vote of no confidence, the outgoing Prime Minister sought to frustrate the formation of a new government within the 14-day period by the obvious method of refusing to leave office? The Act could only operate if the monarch exercised the reserve power to dismiss.

As a result, it seems wrong to treat the Fixed Term Parliaments Act as a comprehensive code not only for calling general elections but also for changing governments without an election. Governments can still be ousted by being defeated on the Queen’s Speech, can still threaten to resign if they lose on votes they specify as matters of confidence, and can still, ultimately, be dismissed by the monarch.

But that conclusion still leaves open a more practical question. How, outside the occasion of a Queen’s Speech, can an opposition put itself into government without risking a general election? What if the government says that it has adopted the view, regardless of whether it is correct, that no votes of confidence now exist except those mentioned in the FTPA and so will not allow any other sort of no confidence motion to be debated in government time? And what if it says that it will ignore any motions passed on opposition or back-bench days that do not use the magic words set out in the FTPA, for example a motion declaring that the House had ‘lost confidence’ in Her Majesty’s Government? The problem would be that the monarch would take some convincing that she should use her reserve power where doing so could lead to an impasse, with no one else being able to win the confidence of the House but no general election being possible.

But perhaps there is a solution in the newly rediscovered mechanism of moving humble addresses. A motion could be moved on an opposition or back-bench day in the form ‘That an humble Address be presented to Her Majesty, That she will be graciously pleased to dismiss her current ministers and to appoint X as Prime Minister’. If such a motion passed, the monarch could be confident not only that the current government had lost the confidence of the House but also that a specific person (presumably the Leader of the Opposition, but another person could be proposed) already commanded that confidence. In effect, we would find ourselves with the sensible rule in Article 67 of the German Basic Law that the Bundestag can only declare no confidence in one Chancellor by declaring confidence in another Chancellor.

No doubt this proposal would cause some trepidation at the Palace. No problem would arise as long as the sitting Prime Minister is an honourable person who would implement the House’s wishes without embarrassing the monarch. But what if the Prime Minister were not honourable? The basic principle the monarchy follows is that it should always take the least politically controversial course of action. The question would be, which would be less controversial – to defy the Commons or to accede to its request?

This article gives the views of the author, and not the position of Democratic Audit. It was first published on the UK Constitutional Law Association blog, and is republished with permission.

About the author

David Howarth is Professor of Law and Public Policy, University of Cambridge.