The office of Prime Minister is a creature entirely of constitutional convention. While legislation references the office itself, setting out pay for example, this is only statutory recognition of the existence of an office which arises purely by virtue of convention. It is by convention that the Monarch appoints as Prime Minister someone who is capable of commanding the confidence of the House of Commons and it is by convention that he exercises certain prerogative powers only on the advice of the Prime Minister (or another minister in his government). The mechanism for choosing a Prime Minister is, however, not entirely covered by convention. While they must command the confidence of a majority of MPs in the Commons, it is not necessary that they be chosen directly by the Commons. This is because of the party system and the presumption that the leader of a given party has the confidence of their MPs, even if they might not always have their full support. 

For all the talk we often hear about how the fusion of legislative and executive powers in Parliament means that the Prime Minister is effectively an elected dictator, it is clear (especially from the last few weeks) that cabinet policy can survive only to the extent that the back benches permit it. Bagehot’s efficient secret rests on the Prime Minister actually – in substance – commanding the confidence of the House of Commons because the majority of elected MPs support their policies. If that is the case, with the possible exception of the first-past-the-post electoral system, there is nothing democratically illegitimate about government proceeding to implement its agenda in a Commons that supports it. 

Of course, it’s a banal point to stress that if the Prime Minister or their cabinet lose that support, they cannot govern. Yet this should be stressed because some commentators and scholars continue to critique the fusion of powers on the presumption that the Prime Minister will have the support of their party and so will be able to act unimpeded. Anyone with any serious knowledge of how parliamentary politics works knows that this is ludicrous. A Prime Minister may have confidence in the sense that the majority of MPs do not want to remove them from office, but that should not be read to imply that they have support for their policies. The cabinet draws its legitimacy from the Commons. There is one source of national democratic authority under our system and that is the election of constituency MPs. The authority of the Prime Minister – despite claims by Boris Johnson and his supporters to the contrary – derives from the confidence of MPs. This is not a presidential system such as the U.S. where there are rival sources of democratic authority, creating tensions and competing claims to legitimacy between the executive and the legislature. Here the executive has authority only to the extent that it is supported by the legislature. When that support is weakened, the executive cannot govern effectively (whether it governs well is a separate matter), forced to U-turn on or adapt policies which do not have the support of backbench MPs. When support runs out, the Prime Minister will either resign or be removed. 

Liz Truss will be remembered as one of the weakest Prime Ministers in British history. Why she was so feeble is the result of several factors, including her own hubris and distinctive speaking style. In this post I want to suggest that her most important weakness, and the cause of her swift defenestration, was the manner of her selection as leader of the Conservative party. By outsourcing the final choice for leader to the membership, the party created the conditions for political failure and constitutional crisis. 

Not every political crisis is a constitutional crisis. The political crises in the 1970s – the oil crisis that led to the demise of Heath’s premiership and the breakdown of Callaghan’s premiership as a result of the winter of discontent – were examples of politicians being broken by events largely outside of their control. Other political crises are better seen as the result of a genuine decay of politics, where norms and traditions are abandoned by political actors seeking self-advancement or appealing to rival sources of political legitimacy. National politics in the U.K. has been in this later kind of crisis since the 2016 Brexit referendum, driven partially by a clash between competing sources of democratic legitimacy manifesting on opposite sides of the issue. It is unclear whether any of this is a constitutional crisis, however. If the general malaise that we have seen, peaking with the rise and rapid fall of Liz Truss, is causing a constitutional crisis, it must be a constitutional one. That is, it must threaten to undermine constitutional principles or conventions. Weak politicians or unprecedented political or economic events are not enough to establish a constitutional crisis. 

Having said this, the Truss premiership has revealed some serious constitutional issues at the heart of these swirling political crises. The constitutional convention that the Prime Minister must hold the confidence of the House of Commons is compatible with the party system only if political parties choose their leaders through processes that ensure that the leader has the confidence and the support of their MPs. It is not sufficient that the Prime Minister will survive a formal vote of no confidence should one be called. To govern effectively, they must be supported by the Commons, even if all of their policies are not. The current saga which led to the collapse of Liz Truss can be traced back to the 1980s and a change in party governance which made it possible for someone without the support of their parliamentary party to become Prime Minister. Liz Truss was removed as quickly as she was because she never should have been there in the first place. 

It’s a cliché at this point for Conservative MPs to blame the Labour Party for their failures but here there is actually some truth to this. Driven by a desire to democratise the governance of the Labour Party, Tony Benn championed a campaign to de-centre the parliamentary party from the choice of Labour leader, opening the choice up to the membership and ensuring that the views of the ordinary Labour voter were better reflected in their leader. The Campaign for Labour Party Democracy advocated for an electoral college but what was eventually introduced in 1993 was a one member one vote system, spearheaded by John Smith. The Conservatives followed a few years later when, in 1998, William Hague changed the internal party rules to allow members the final say on leadership. Crucially, these changes were introduced when both parties were in opposition and so the choice of leader did not automatically translate to a choice of Prime Minister. Since then there has been a fall in the membership of each party, although labour had a sizeable increase in membership following the 2015 general election. Membership of these parties has become the preserve of a minority of political activists whose views do not align with the average voter, nor the majority of MPs in their party. Jeremy Corbyn received the fewest parliamentary nominations of any candidate during his leadership contest but was overwhelmingly chosen by members. Similarly, Liz Truss became leader of the Conservatives with fewer than half of the MPs backing her. Huge membership spikes during the run up to the selection of Jeremy Corbyn and Boris Johnson, coupled with the decline in members actually campaigning for the party, suggest that membership is now used as the vehicle to change the policy direction of existing parties, contributing to an increase in polarisation and a disconnect between party members and MPs. 

A consequence of the first-past-the-post system is the existence of large parties which are themselves coalitions, exceptionally broad churches with diverse political views. At the centre, it can be hard to distinguish the left of the Tory party from the right of the Labour party. At their extremes, each party contains members and MPs with views radically divergent from the average voter. Under a system of proportional representation, these more extreme wings would be separate parties. But under our current system, coupled with the change to leadership selection, these wings are not only a part of a larger coalition party, they have been able to choose the leader, notwithstanding the majority of MPs supporting a different candidate. An extreme libertarian or socialist party would likely not win enough seats in the Commons to command a majority, especially not under a system of proportional representation. But with the current rules of both of the major parties, all they need to do is pay a small fee and gather enough numbers to choose a Prime Minister. Still, if Truss has shown us anything, it is that making it to the top spot means very little if you cannot command the support – not just the confidence – of the majority of the Commons. Support of the MPs in your own party is important, of course. But it is the Commons which ultimately matters. The larger a party’s majority in the Commons, the more leeway a leader has to manage discontent within their party. But if discontent rises to the point where the working majority collapses, the Prime Minister may remain in place but not in power.  

Constitutional conventions can often be presented as strictly formal requirements, with compliance and normativity being necessary but only to the extent that they guide conduct in line with the convention. But this normativity must be informed by constitutional principle. As Jennings has argued, it matters that conventions are politically binding for good constitutional reasons because, if they are not, they will eventually decay, losing their normative force and ceasing to be conventions at all. The convention that the Prime Minister command the confidence of the House of Commons should not be perceived to be a mere formality, tested only by reference to votes of confidence. To govern effectively, the Prime Minister must have both confidence and support; confidence in form and in substance. 

We rely on the presumption – and it is a presumption – that the leader of the party with the majority of seats in the Commons has the support of the Commons and so is entitled to form a government. If we change the method of selecting party leaders, displacing the Commons, we lay the foundations for a genuine constitutional crisis wherein a Prime Minister formally has the support of their party’ MPs but in substance does not. This can be distinguished from instances where a Prime Minister generally does have the support of their parliamentary party but chooses to pursue an unpopular policy and must respond accordingly, by convincing MPs to support them, changing the policy, or U-turning. Often the government will take things off the table because it knows it would find it difficult to get it through the backbenches. That is simply politics and not a crisis on any definition. But the selection of a Prime Minister who does not even formally command the support of the majority of their MPs is neither legitimate nor sustainable. Liz Truss never had the support of her parliamentary party. She was the second choice behind Rishi Sunak and alienated him and many of his supporters in the Commons when selecting her cabinet. Here, just as with the final months of the May premiership, the line between support and confidence is being tested to breaking point, presuming that the leader has confidence (MPs do not wish to remove them) even when they don’t have support. The result is a leader in place but not in power and a serious uncertainty about whether they actually command the confidence of the Commons. Formally they may, but substance is needed if the efficient secret is to remain efficient. 

It may be acceptable for a party in opposition to choose its leader by member vote, but it cannot be justified for a governing party. The Prime Minister must command the support as well as the confidence of the Commons. If they command support, it is perfectly ordinary for some of their policies not to and for parliamentary politics to respond accordingly. Political crisis ensues when the Prime Minister lacks support. This is usually – with the notable and telling exception of Theresa May –resolved by swift and ruthless defenestration. Where this does not occur, we are approaching constitutional crisis. The final months of the May premiership and the short-lived weeks of the Truss premiership each represent a breakdown in the conditions needed for our constitution to function, one arguably more serious than the other, given the cause.  It is one thing for a Prime Minister to lose the support of the Commons. It’s another thing entirely for one to be elected when they never had support in the first place.  It is imperative that the choice of Prime Minister remain with the House of Commons. 

If the parties do not remove members from this decision-making process, it will call into question the strength of the presumption – and again, it is a presumption – that the leader of a party with a majority of seats in the Commons can command its confidence. The implications that a breakdown of this presumption has for the choice the Monarch must make in whom to invite to form a government are serious. It is unclear whether any change to the party constitutions will happen as a result of the catastrophic Truss premiership, but if this is not resolved, these problems will only get worse.  

Michael Foran is a Lecturer in Public Law at the University of Glasgow.

(Suggested citation: M. Foran, ‘Prime Ministers, Party Members, and the Efficient Secret’, U.K. Const. L. Blog (26th October 2022) (available at https://ukconstitutionallaw.org/))