When must a government leave office?

Image adapted from a photograph by Defence Images on Flickr under licence CC BY SA 2.0, and available for reuse under the same.

This is a follow-up, of sorts, to my article outlining the nine ways in which Theresa May might leave office. It looks in more detail at one option in particular: losing the confidence of the House of Commons. And in fact I want to argue that this now looks pretty unlikely as her exit route, largely thanks to a widespread misunderstanding of some recent constitutional change — namely the Fixed Term Parliaments Act 2011.

Let’s not leave the subtext unspoken: the prompt for this is Brexit. If Mrs May brings a withdrawal agreement back to the House of Commons for it to vote on (which now seems unlikely to me, but that’s another story), what happens if she loses the vote and her deal is defeated?

My answer to the question of what should happen is simple: she should resign and advise the Queen to invite the Leader of the Opposition to form a government. But I don’t believe that’s what actually will happen if we get there. So let’s look firstly a why it should happen, and secondly at why it won’t.

As I previously outlined, the rules around who can form a government are quite simple: you have to command the confidence of the House of Commons. If you’re in office but lose its confidence, you must resign. Your only other option is to secure a new House of Commons — ie call a general election — and hope that the new one has more confidence in you than the old one. But if that option isn’t feasible for whatever reason, you have to resign, and your entire government with you. But there’s a second rule: you may not resign as Prime Minister unless and until you are able to advise the Queen who she should send for to replace you — that is, who else can command the confidence of the Commons.

So, let’s map that onto a Brexit vote defeat. Brexit is the flagship issue for Mrs May’s government. It is the defining issue of the day, central to the future of our nation and by far the most important thing the government will ask this parliament to vote on. It is impossible to argue that a Prime Minister who loses a vote of principle on a matter such as this commands the confidence of the House of Commons. It is a confidence issue. Just as Major would have gone if he’d lost over Maastricht, and just as Jack Straw cleared his desk ahead of the crucial vote on Iraq, Mrs May cannot remain as Prime Minister if she is defeated on this issue.

Not only that, but in those circumstances it will be very clear that the Conservative Party is sufficiently badly split that it cannot govern — simply installing a new Tory leader would not cut it, even if it was procedurally possible to do so within a matter of days, which it isn’t. Worse still, the country will immediately enter a constitutional, political and possibly economic crisis, with an urgent need for the next steps on Brexit to be outlined (whether to reopen negotiations, ask for an extension to Article 50, or whatever). Mrs May certainly won’t be able to take such a decision and get parliamentary agreement to it. So there will be only one option: Mrs May has to resign, and advise the Queen to send for the only person with even a chance of commanding the confidence of the House of Commons: the default fall-back in these situations, the Leader of the Opposition — right now, that’s Jeremy Corbyn.

(Whether Mr Corbyn actually can command the confidence of the Commons will be another matter; but the bottom line is that the Queen will have to appoint him Prime Minister so that he can at least have a go. And that would add an extra dynamic to the vote: Conservative and Labour MPs voting against the deal would also know that they are voting for Mr Corbyn to become PM, which you might think would put a different complexion on the vote for at least some.)

Thanks to the two simple rules about who can become Prime Minister and when they must leave, the position could hardly be more clear-cut. Yet none of that will happen. Why not? Because of the Fixed Term Parliaments Act (FTPA).

Or more precisely, because the FTPA is so widely misunderstood. Take a look, for instance, at this article from the normally excellent Patrick Maguire, who says: “Under the terms of the Fixed Term Parliaments Act, budgets are no longer considered confidence votes.” This is nonsense. The terms of the Fixed Terms Parliament Act say no such thing. The Act is completely silent on the matter of budgets, or any form of confidence vote other than one specific type. It does not in any way debar a government from identifying a particular matter as an issue of confidence and staking its existence on it. Yet Patrick’s error is widely repeated: far more commentators, and also it seems politicians, think what he says is right than realise that it’s wrong.

Let’s remind ourselves what the FTPA actually says (it’s not a long or difficult piece of legislation, so I recommend reading it). It sets out that Parliaments will last for five years (with an additional power for the PM to make regulations to extend that period by up to two months, but no more — presumably intended to cover off scenarios like 2001’s outbreak of foot and mouth disease). It provides two mechanisms for varying the length of the Parliament away from the five year limit. The first is that two thirds of the Commons votes for an early election, in which case the PM then sets the date. The second mechanism is the bit that confuses people.

The second mechanism works in two parts. It stipulates firstly that if a motion of no confidence, using wording specified in the Act and no wording other than that, is passed by the Commons, then a 14 day countdown is commenced. If, within that period, a motion of positive confidence in the government is passed by the Commons — again, using wording specified in the Act — then the countdown is halted and nothing else happens: there is no election. But if no such second motion is passed, Parliament is dissolved and an election is called. And that’s all the Act says on the subject.

Note that the Act is all about general elections and the duration of parliaments. It says nothing at all on the subject of how governments are formed, and when they resign. It is completely silent on that. And that is why it is such a bad piece of legislation, that may well prove to have very significant unintended consequences for our politics and constitution.

As we saw above, the issue of ‘confidence’ — that is, the confidence of the House of Commons — is really about how governments are formed. Not about how elections are called, but rather about who gets to be Prime Minister, and therefore form a government. If you can command the confidence of the House of Commons — that is, you’d win a vote in it on any issue of substance — you get to be PM. This means you need to have a majority of MPs willing to vote in favour of whatever you propose, whether entirely from your party or from your party plus others.

By using the concept of ‘confidence’ as a mechanism for instead defining when elections are held, the FTPA has confused the constitution, and more or less everyone involved in operating it. The confusion has arisen, of course, because if you’re a PM who has suddenly lost the confidence of the Commons, resigning isn’t your only option: you can get a new House of Commons (by calling a general election) and hope you’ll command the confidence of that one instead — as James Callaghan tried in 1979.

An Act that was intended to take away the Prime Minister’s prerogative power to call an election has therefore also massively interfered with the key concepts of who gets to be Prime Minister, and when. Politicians and commentators now seem to assume that the concept of ‘confidence’ is only material when we’re talking about the timing of elections, and also that confidence can only be expressed in, or inferred from, a formal vote of confidence under the terms of the FTPA. All of this is nonsense.

Take, for example, a ‘normal’ general election that produces a change of government — like in 1997, say, when John Major lost and Tony Blair won. Major immediately went to the Queen and resigned, because he had lost the confidence of the Commons, and he didn’t need to hold a formal vote to find that out — Labour easily had enough MPs to vote his government down. Take 2010 as an example as well: Gordon Brown resigned only when the Lib Dems confirmed that they would not support Labour in the Commons — hence he knew he could not command the confidence of the Commons, and resigned without putting it to the test. Similarly in 1974, the first election was held on February 28th, but Heath only resigned on March 4th, when it was clear the Liberals would not back him and he therefore could not command the confidence of the Commons, but again no vote was held to check that. The confidence of the Commons was the key test in who formed a government.

It’s not clear that this is still the case. Informed comment — that is, informed about developments in contemporary politics, not necessarily about how the constitution works — suggests that in the event of defeat on a Brexit vote, Mrs May intends to continue in office. There seems to be no serious suggestion that Labour will try to force her resignation. It will probably seek an election under FTPA terms, but dissident Conservative MPs will no doubt fall back into line, and probably the DUP as well if they’re happy with the outcome on Brexit, and defeat any no confidence motion. Even though, as we’ve seen, by any meaningful assessment Mrs May will have lost the confidence of the House.

In other words, the widespread belief that ‘confidence’ is now only ever voted on under the FTPA’s arrangements means that big, serious issues that would once have brought governments down now no longer can. The consequence of this is that unhappy backbenchers can now have a ‘free hit’ on the government, even on the biggest issues of the day. They no longer need to consider whether the government will have to resign as a result — because it won’t. They can merrily vote against it on Brexit, then straight away vote with the whip on a formal confidence vote, no harm done.

Except I’d argue quite a lot of harm has been done. The changed considerations around major votes of this sort is surely a recipe for greater political instability: reckless decisions can be taken by MPs much more lightly. Taking Brexit as an example: an MP voting on the withdrawal agreement in the knowledge that not only will there be no deal if it falls, but also Jeremy Corbyn will become Prime Minister and another general election might follow soon, would surely think twice — for plenty of Labour and Conservative MPs, the threat of such a change of government and period of instability would make backing Mrs May’s deal seem palatable, in a way that it might not be if they were voting on it in isolation. Who knows what other big decisions might get voted on by the Commons differently in future if MPs know the government can afford to lose the confidence of the House and still remain in office?

This is significant constitutionally as well as politically. In an uncodified constitution such as ours, the elements of the constitution work as the actors involved believe them to work, not as they are written down anywhere (with the exception of codified elements such as the FTPA, which exists in statute). So this change to the constitution could well prove to be real, even though it has arisen from a dunderheaded misunderstanding. And it is a major constitutional change, completely unintended by anyone at the time of the passage of the FTPA.

Frankly, it is now not clear how we form governments, and the circumstances in which a PM must resign. For instance, if Mrs May loses an FTPA-style confidence vote, is it acceptable for her to ‘run down the clock’ of the 14 days trying to persuade the Commons to pass a new confidence motion? If not, how long is she allowed, and in what circumstances must the incumbent Prime Minister make way for the Leader of the Opposition? What if the government lets it be known politically that it considers a vote to be a matter of confidence, and then loses it — can they change their minds and not resign anyway? Is it even acceptable for a Prime Minister to conclude they have lost the confidence of the House and resign without a formal vote? The Act offers no guidance at all on those or other issues.

More generally, in future parliaments with tight arithmetic, we can expect to see more opposition parties angling cynically for a crisis that might prompt an early election through a vote of no confidence, such as Labour are now, and presumably we will also therefore see more general elections as a result, where in the past we might have seen an opposition leader appointed to a minority government and lead it for a while without going to the country. This wouldn’t be possible without the apparent trend towards small majorities or hung parliaments in Westminster elections, but even so the FTPA is a major element in determining how the parties behave in a tight House of Commons. All told, this set of consequences is deeply ironic for legislation intended to make the duration of parliaments both longer and more stable, and therefore general elections less frequent and less dependent on political contingency.

The moral of the story is of course: don’t mess with the constitution. There’s plenty wrong with it, and it needs a lot of change, but that needs to be done as a ten-year programme involving a great deal of care, thought and effort, not piecemeal tinkering to put odds and ends of stuff on a statutory footing as seems politically expedient. Just changing one thing inevitably has knock-on effects elsewhere, just as changing the voting system or lowering the voting age would if done in isolation. Sadly, looking at British politics today, one has to wonder when we will ever have the luxury, politically, of being able to commit to a properly considered, ten-year programme of constitutional reform. Or, for that matter, when our political class will ever have the intellectual wherewithal to do it successfully.

UPDATED, OCTOBER 10 2018: One permutation I didn’t consider in the above was the Budget getting voted down, which both the DUP and, reportedly, the European Research Group of Conservative MPs have threatened to do if progress on Brexit negotiations is not to their liking (the Budget occurring on October 29th, before any deal is expected to be finalised). This has emerged as a case study in the effects I outlined above.

Firstly, let’s consider what actual votes are held on ‘the Budget’. There are three lots. The first is a motion on the provisional collection of taxes, which is put to the House on Budget day and allows any changes in the rates of duties to apply from 6pm. The second is a set of motions called Budget Resolutions, to give immediate effect to some more of the measures announced in the Budget, after a customary four days of debate. The third is the voting on the Finance Bill itself, which puts everything on a full statutory footing and proceeds essentially like any other legislation.

As Philip Cowley has pointed out helpfully on Twitter, the oft-repeated belief that the Finance Bill can’t be amended without the government resigning is incorrect: it has happened 15 times in the last century (and I have updated the above to reflect that, though it argued that amending the Finance Bill would not be such a big deal anyway). So the Finance Bill would only traditionally be considered the subject of a confidence vote, if at all, on its second reading, like any other major legislation. Presumably therefore threats to ‘the Budget’ as a whole apply to the first two sets of votes, on the provisional collection of taxes (just after the Chancellor’s speech), and the Budget Resolutions (four days of debate later). The House of Commons explainer doesn’t specify, but I’m presuming these votes are held as normal: initially voted on by acclamation in the chamber, with a formal division if the results of that aren’t clear.

A government defeat on either of these items would be very serious, but not merely because it would be a loss of confidence (though it would). Voting money — by taxation — to the government is Parliament’s most basic function. Indeed, when the institution first emerged in the middle ages, that was its central purpose, and kings only called parliaments when they wanted to levy a tax. So in principle losing a vote on one of those motions most certainly constitutes a loss of confidence of the House, but in practice the government might try to ride it out on the pretext of the FTPA as discussed above. But it would be more serious, because it would not be clear that the government could actually levy the revenue it needed in order to carry out its work. So even leaving the confidence aspect aside for a moment, surely any government that lost one of the big symbolic votes on the Budget as a whole would have to go, as in principle it would not be able to sustain its finances?

There’s a further specific implication to the DUP’s threat to deny the government supply by defeating the Budget. Its deal with the government, in return for its notorious £1 billion bung, was to provide confidence and supply. Defeating the Budget means denying the government both. By rights, therefore, Mrs May should resign, and whether it’s on the supply issue or the confidence one shouldn’t matter: the Queen invited her to form a government on the basis that she would have a deal with the DUP to provide her government with confidence and supply. With that deal gone, the basis on which Mrs May was invited to form a government will have lapsed, and she should be advising Her Majesty to send for the Leader of the Opposition.

However, all of this appears to have been thrown out of the window by the FTPA. Downing Street has briefed today that a defeat on the Budget would not be a resigning matter, explicitly because of the FTPA. And there we have it: the FTPA has indeed torn up the rules on how governments are formed, when they may remain in office, and when they must resign. Of course, it has yet to be tested in practice: maybe, just maybe, a government lacking both confidence and supply will realise the hopelessness of its situation and quit, and re-assert the old conventions, weirdly, by setting a new precedent. But it’s far from clear that this will happen: instead, there must be a significant danger that any major political crisis will quickly turn into a constitutional crisis as well, because the rules for resolving problems as they arise will be disputed. What a good job there’s no sign of any sort of major political crisis on the horizon, then.

Working in public policy and writing here about politics, infrequently, in a purely personal capacity.

Get the Medium app

A button that says 'Download on the App Store', and if clicked it will lead you to the iOS App store
A button that says 'Get it on, Google Play', and if clicked it will lead you to the Google Play store