The next election could be in as little as 18 months’ time and the possibility that it will produce a hung parliament is a very real one—compared to what would be needed for a majority Labour government it would require only a relatively modest, and entirely possible, shift in public opinion. Indeed, as of the time of writing the opinion polls are suggesting that an election held today would produce a hung parliament.
There are very good reasons why the Labour leadership and Labour MPs do not want to discuss that possibility in public: they are all hoping and working for a majority Labour government. But contingency planning is a good idea even when you hope that the contingency will never happen: and it is better done now, well before the pre-election period when other priorities will take over. Moreover, silence about what would happen if there is a hung parliament may not be sustainable. If the polls are still pointing towards such an outcome at the next election, Labour will need to be prepared to deal with the inevitable “coalition of chaos” and “Starmer in Sturgeon’s pocket” attack lines from the Conservatives. It will be easier for Labour to come up with an effective response to that attack if it has thought through in a robust way what its strategy would be.
There are some reasonably safe political assumptions:
- In a hung parliament the SNP is likely to be a sizeable bloc, given its concentration of support and the assistance the party is given by first-past-the-post. There will be—again thanks to first-past-the-post—hardly any Green MPs. As for the Liberal Democrats, they may or may not have significantly more MPs than they do now but, unless they do really well in taking seats from the Conservatives in rural and suburban England, first-past-the-post is likely to hand them a number of MPs well behind the SNP, even if they get a share of the vote in the mid-to-late teens.
- Even if Labour is significantly behind the Tories in terms of seats, Johnson will not find any parties willing to support a Conservative government in a confidence vote. The only conceivable exception is the DUP, but (to put it mildly) a lot of water has flowed under the bridge since the DUP supported May’s minority government, and its support cannot be taken for granted (and it may well lose seats at the next election anyway).
So let us take, as a central working assumption, a situation in which the Conservatives are short of a majority (even with the DUP) but a combination of Labour with the SNP, or with the SNP and the Lib Dems, would have a majority.
The rules of the game
Before playing any wargame, you need to master the rules. So what are the rules? As usual in what passes for the UK constitution, the rules are a patchwork of laws and conventions of varying degrees of clarity and certainty found in various places. However, they are conveniently summarised in the Cabinet Manual, an account that is generally regarded as authoritative and is likely to be followed by both the civil service and Buckingham Palace (although bits of it will need updating when the Fixed-term Parliaments Act 2011 (FTPA) is repealed).
After an election, the ball is in the hands of the incumbent prime minister. This is explained in Chapter 2 of the Cabinet Manual:
2.12 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 his or her resignation and the Government’s resignation to the Sovereign. An incumbent government is entitled to wait until the new Parliament has met to see if it can command the confidence of the House of Commons, but 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.
That needs to be read with another paragraph, 2.10:
it remains a matter for the Prime Minister, as the Sovereign’s principal adviser, to judge the appropriate time at which to resign, either from their individual position as Prime Minister or on behalf of the government. Recent examples suggest that previous Prime Ministers have not offered their resignations until there was a situation in which clear advice could be given to the Sovereign on who should be asked to form a government. It remains to be seen whether or not these examples will be regarded in future as having established a constitutional convention.
What all that means in our central scenario is that, after the election, Johnson would stay in office until he resigned. But he would not have to resign—and indeed should not resign—unless and until it became clear that an alternative government would be able to win a vote of confidence. It is “expected” that he should resign as soon as it becomes clear that he would lose a vote of confidence and that an alternative government can be formed which would command confidence (the last sentence of 2.12): but even if he defies that expectation, he would on any view have to resign at once as soon as he lost a vote of no confidence in the new House of Commons, if it is clear that an alternative government would win a confidence vote. In either case, the Queen would then ask Keir Starmer to form a government.
So, in our central scenario, if either the SNP or the SNP plus the Lib Dems (as the case may be) were able to say that they would support a Labour government on a vote of confidence, Johnson would have to resign either at once (probably) or after losing a vote of confidence (certainly), and would have to advise the Queen to ask Starmer to form a government (though his advice at that stage is essentially irrelevant, as that is what the Queen would do anyway).
Note, however, the point made in paragraph 2.10 and at the end of 2.12. Unless and until the SNP, or the SNP plus the Lib Dems, were prepared publicly to say that they would support a Labour government in a confidence vote, Johnson would not be required to resign—indeed, he would, probably, be required not to resign. The Queen’s government must be carried on, as it is sometimes said.
Three important further points about the rules.
First, it does not really matter for these purposes whether Labour or the Tories have more seats. Being the largest party is irrelevant if you cannot win a vote of confidence, but the second largest party can. In fact—though you have to go back to before the Second World War—there are several examples of minority UK governments formed by the second largest party under precisely these circumstances: the interwar Labour governments (supported by the Liberal Party) and the 1892-95 Liberal government (supported by the Irish Home Rule Party) come to mind. There are also many foreign examples (Labour in New Zealand from 2017-2020 is a very recent one).
Second, what about the possibility that Johnson would ask the Queen for another general election? Under the current government’s plans to repeal the FTPA, the position on calling an election will in effect be the same as that which applied before 2011. The Queen has power to dissolve parliament under the Royal Prerogative and generally will do so whenever the incumbent prime minister advises her to. However, it is generally understood that that presumption is qualified by what are known as the “Lascelles principles” (in classic British fashion, these were formulated in a letter written to the Times by a former private secretary to George VI). Under the Lascelles Principles, the Queen would refuse a prime minister’s request for a dissolution of parliament if “(1) the existing Parliament was still vital, viable, and capable of doing its job; (2) a General Election would be detrimental to the national economy; (3) [she] could rely on finding another Prime Minister who could carry on [her] Government, for a reasonable period, with a working majority in the House of Commons.”
It may now be questioned whether those principles are now quite right (principle (2) looks like a rather political judgment for a non-political monarch to be making) but it can be said with some confidence that, in the immediate aftermath of one general election, and in a situation where Labour would clearly be able to win a vote of confidence in the new parliament, the Queen would refuse a request by Johnson for another election, either before or after he lost a vote of no confidence (unless no alternative government could be formed, in which case there would be no alternative to a second election, with Johnson remaining in power in the interim). On the other hand, once Starmer was in office as prime minister, he would likely be able to obtain a general election whenever he wanted one (at least after a few months), including if he lost a confidence vote (unless, which is implausible, it was in a situation where the Conservatives were then able to attract either the SNP or the Lib Dems to support them as an alternative government).
Third, though minority governments may negotiate “confidence-and-supply” agreements with other parties, in which the other parties formally agree to back them on measures needed to secure government funding and on confidence votes in return for policy commitments, there is no formal requirement that such an agreement be negotiated: and if it were clear that (if they held the balance of power) the SNP or the Lib Dems would vote confidence in (and supply to) a Labour government, the absence of any formal agreement would be irrelevant. And it would be even more irrelevant that there was no coalition along the lines of the 2010-15 UK coalition or the 1999-2007 Lab/Lib coalition in Scotland, or even along the rather looser lines of the current SNP/Scottish Green coalition.
How to play the game
So those are the rules. What should Labour’s strategy be?
The real issue is how to deal with the SNP, given that (a) the foundational policy of the SNP is one to which the Labour Party is resolutely opposed and that (b) the idea of a Labour government doing deals with the SNP is likely to be unattractive to most Labour Party members, especially in Scotland—and to many voters, especially in England (a vulnerability that the Conservatives will seek to run hard as a reason not to vote Labour).
However, there is no need for Labour to negotiate any formal (or informal) deal with the SNP in order to form a minority government, even if that government relies on SNP support in order to win votes of confidence. Moreover, the SNP’s room for manoeuvre in such circumstances is quite constrained. If it refuses to say that it will support a Starmer government in a confidence vote then the result is to keep Johnson in power until he loses one: and even when he loses that vote, refusing to support Labour would lead to continued Conservative government pending a second general election—a position that would be hard for the SNP to explain to its voters and which would risk returning a parliament with a Conservative majority.
So Labour would be in a strong position to take a “back us or sack us” line with the SNP and to resist any deals (though of course assurances could be given on issues where no concession of important principle was at stake: examples might be handing over to the Scottish government control of Scotland’s share of the UK Shared Prosperity Fund, or giving legal force to the Sewel convention that restrains the UK parliament from legislating on devolved matters without the agreement of the devolved parliaments). In particular, it would have a good chance of calling the Scottish nationalists’ bluff if the SNP tried to make its support conditional on agreement to an independence referendum.
Once a minority Labour government were installed on such a basis, its legislative programme would be vulnerable to the SNP trying to exert pressure by threatening to vote with the Conservatives or abstaining. There might well be particular issues with legislation affecting only England because, even though the special procedure that restricted key votes on such laws to English MPs has now been abolished, the SNP might refuse to vote on such legislation (and there might also be presentational issues in getting English legislation through on the back of SNP votes). Labour would need to minimise the risk of defeat by concentrating on legislation which the SNP would find hard to oppose with any plausibility and, especially in England, on policies that do not require primary legislation to implement.
Fortunately, for its ability to survive and implement large parts of its programme, a minority Labour government would have in its arsenal the huge powers that the current parliament is busily handing to the executive to legislate in very substantial ways without any, or any substantial, parliamentary scrutiny. Those powers include rewriting whole regulatory regimes and a plethora of “Henry VIII” clauses that allow the executive to make huge changes to previous Acts of Parliament. A minority Labour government would also be able to make use of the refusal by the current government to give the Commons any significant role in controlling the executive’s negotiation of trade agreements. It could, for example, renegotiate—without getting any prior approval from the Commons—the terms of the Trade and Cooperation Agreement with the European Union (a review of which is due in 2026): the new terms can then be presented to the other parties as a fait accompli which they can either vote against or vote for, but cannot amend.
A minority Labour government could also make use of its control of the Commons timetable—a key weapon in the hands of the executive—to reduce the number of votes that actually mattered on which it was liable to defeat (the Conservatives, in their period as a minority government in 2019, set a number of precedents for simply ignoring inconvenient votes, for example on Opposition motions). There would, inevitably, be at least some votes which Labour would have to hold and win—on the Budget and Queen’s Speech, most obviously. But ultimately, the question for the SNP would always be whether it was prepared to bring a Labour government down, leading to an election (since no alternative government could be constructed) and the risk of either a majority Labour government (if Labour was doing well) or another majority Tory government (if it wasn’t). Though having a Conservative UK government suits the SNP in many ways, the party cannot afford to look as if a Conservative government suits them or appear to have deliberately facilitated its return to power.
Further, Labour could promise constitutional reform of a kind that would (or certainly should) incentivise SNP support. Electoral reform would reduce the likelihood of a majority UK Conservative government in future, and be hard for the SNP to oppose or frustrate. Extension and entrenchment of devolution would be a further incentive.
As for the Lib Dems, they are less likely to hold the balance of power, as matters stand. But if they are in a key position, and assuming that a formal coalition is unlikely to be attractive to either party, there is obviously scope for a confidence-and-supply agreement—particularly if Labour has gone into the election supporting fair voting and reform of the House of Lords (so that electoral reform and a new second chamber can, without difficulty or political embarrassment, be included in the agreement, thus pretty well guaranteeing Lib Dem support for the government while those reforms are being implemented).
What would minority government be like?
Minority government is hard. Legislation—some of which cannot be avoided—becomes a major effort and a saga of late nights and lost votes. Keeping a sense of direction can become very difficult. It would be well worth (quietly) planning for the possibility well before the election: working out what the priorities are, what legislation is likely to be hard for the SNP or Lib Dems to oppose, abstain on, or disrupt with wrecking amendments, and what parts of the programme can be implemented without any primary legislation (either by secondary legislation or by use of other statutory or prerogative powers—remembering that the current government’s addiction to giving itself such powers would prove highly useful to a minority Labour government). It would also be well worth making use of the convention that allows the opposition to discuss the implementation of its programme with the civil service in the period leading up to an election.
Despite the difficulties, however, minority government can work and deliver electoral success in the subsequent election—see the SNP from 2007-2011 and Jacinda Ardern’s 2017-2020 Labour government. It is true that in the UK the living memory precedents for minority governments (Labour in 1976-79, the Conservatives from 1995-1997 and 2017-2019) are unhappy: but the first two were at the end of a period of government rather than at the start of it, and the May minority government had no choice but to devote itself to getting complex legislation through parliament on a very important issue where its own MPs were deeply divided—not a problem that a Labour minority government would be likely to face in 2023/4.
In a general election the Conservatives could well try to play on those memories in order to frighten voters into renewing their majority. But thinking through the strategy and how a minority government could actually work provides Labour with the foundation for a robust answer to that “hostage to the SNP” charge: the response would be along the lines that, if there is a hung parliament, there would be no deals on an independence referendum or on any other issue of principle: Labour would set out its stall, invite the SNP to support it, and defy the SNP to keep the Tories in power or to cause another election which would risk putting them back in power. Given that the SNP cannot credibly say that they might support the continuation of a Conservative UK government, that is an entirely solid position for Labour to take (before, of course, moving on to persuade voters to give it a majority).
The main danger is, perhaps, that once a Labour minority government were formed, its legitimacy would come under intense attack: particularly if (as is entirely possible) Labour has fewer seats than the Tories on a similar vote share, or fewer votes as well as seats. Although there are plenty of precedents for a minority government formed by the second largest party, the fact remains that such an attack is bound to be made. The best way of countering it will be to point to the (almost certain) fact that the number of people who voted for Labour, when added to those who voted for the Lib Dems, Greens and SNP—the large majority of whom will prefer a Labour government led by Starmer to a Conservative one led by Johnson—will greatly exceed votes cast for the Conservatives. That, of course, is an argument for a fair voting system that reliably gives effect to that anti-Conservative majority—and is a further reason why a fair voting system should be a firm commitment in Labour’s manifesto.