Will Keir Starmer be the UK Prime Minister at 23:59 BST on 30 June 2026?

TL;DR

My forecast is 79% that Keir Starmer is still UK Prime Minister at 23:59 BST on 30 June 2026.

Context

Starmer is the formal officeholder now: the official GOV.UK Prime Minister page lists him as the current role holder and says he became Prime Minister on 5 July 2024, and the GOV.UK ministers page still lists him as Prime Minister, First Lord of the Treasury, Minister for the Civil Service, and Minister for the Union.

The crisis is real, but the question is narrow. It asks whether someone else has been appointed Prime Minister by the King by one exact minute on 30 June. A resignation timetable, a leadership contest, or a public statement that Starmer will go later still resolves YES if he remains the constitutional PM at the deadline.

Evidence

The historical backbone favours survival through this deadline. The full history of completed Labour premierships since Labour first entered government is small, but it is the right reference class because Labour’s rules and norms differ from the Conservatives. The official GOV.UK list gives the Labour premiership sequence; no completed Labour premiership in that sequence ended through a successful formal party leadership challenge against a sitting Labour PM.

The rules make a fast forced removal hard. A challenger against an incumbent Labour leader needs written nominations from at least 20% of Labour MPs; at the current Commons count of 402 Labour MPs, that is about 81 MPs. The incumbent is automatically on the ballot if a challenge is triggered. If the Labour leader resigns while Labour is in government, the Cabinet, in consultation with the NEC, can appoint one of its own members as interim party leader until a ballot, but resigning as party leader does not itself mean immediate resignation as Prime Minister. The Institute for Government says outgoing party leaders normally remain PM until a successor is elected, and gives the 2007 Labour example: Brown launched his leadership bid on 11 May, became Labour leader on 24 June, and became PM on 27 June.

That timing point dominates the forecast. From 31 May to the deadline there are only 30 days. A normal Labour member ballot started now would be tight. A Burnham-led challenge after 18 June would be much tighter. Even a successful late-June challenge or resignation announcement can still resolve YES if Starmer remains PM while Labour chooses the next leader.

The current rebellion is serious but still short of a completed mechanism. ITV reported on 13 May that it had tracked 95 Labour MPs prepared to oust Starmer or demand a timetable, after contacting all 403 Labour MPs, but the same report said a contest would only be triggered if MPs united behind a named candidate with the necessary support. (itv.com) AP reported on 18 May that Starmer was widely expected to face a challenge, but no one had formally launched one yet. By 31 May, the live public posture was still preparation for a possible contest rather than a contest already under way: The Independent described Streeting as setting out policies for a potential leadership contest expected in the coming months and said he had resigned to prepare for a bid after Makerfield.

The political pressure is still high. ITV’s 9 May election report, with some councils still to declare, had Reform up 1,244 councillors and Labour down 1,022 councillors, and described the results as putting Starmer’s leadership under fresh scrutiny. YouGov’s 18 May Labour-member survey of 706 members found Burnham first choice on 47% versus Starmer on 31%, and found Burnham beating Starmer 59% to 37% in member rankings; the same survey also found 66% of members thought Starmer had done a good job as PM and 61% wanted him to remain leader for now. Ipsos fieldwork from 8–12 May among 2,191 British adults found 66% thought Starmer should not lead Labour into the next general election, split between 35% saying resign immediately and 31% saying go later.

Makerfield is the main live path. Wigan Council’s official notice says the Makerfield by-election will be held on Thursday 18 June 2026 from 7:00am to 10:00pm and lists Andy Burnham as the Labour and Co-operative candidate and Rob Kenyon as the Reform UK candidate. The House of Commons Library says Burnham can stand while Greater Manchester mayor, but if he wins he becomes an MP immediately on declaration and is immediately disqualified from the mayoralty. Survation’s constituency poll, fielded by telephone among 504 Makerfield adults from 18–22 May and published 26 May, had Burnham on 43% and Kenyon on 40%, with a headline margin of error of about ±5.4 percentage points; it also found Reform leading Labour by 11 points on a generic Westminster ballot without candidate names. I read that as a small Burnham edge, not a lock.

The other June trigger is the Mandelson document tranche. In the Commons on 19 May, Darren Jones said the second tranche would be published after Whitsun recess, called it one of the largest Government publications ever laid before the House, and said some vetting-related material could not be published raw. (hansard.parliament.uk) That creates a real shock path. It does not by itself solve the successor-selection problem.

My model treats NO as the union of four routes. Let $e$ be an early direct handover before Makerfield, $b$ the chance Burnham wins Makerfield, $w$ the chance Starmer is actually replaced by 30 June conditional on a Burnham win and no earlier handover, $l$ the same chance conditional on a Burnham loss, and $t$ a small tail for death, incapacitation, no-confidence mechanics, or another idiosyncratic route. I used $e=8.5\%$, $b=58\%$, $w=18\%$, $l=5.5\%$, and $t=0.4\%$.

$$P(\text{NO})=e+(1-e)\,[b w+(1-b)l]+t$$

This gives $P(\text{NO})=20.6\%$, so $P(\text{YES})=79.4\%$. The most sensitive input is $w$. If a Burnham win has only a 10% chance of producing a new PM by 30 June, YES rises into the mid-80s. If it has a 30% chance, YES falls to the low 70s.

What's non-obvious

Most coverage blurs “Starmer is finished” with “someone else is PM by 30 June.” Those are different claims. Labour can be preparing a succession, Starmer can be publicly doomed, and a contest can even be under way, while the constitutional officeholder at the deadline is still Starmer.

Burnham both raises and delays the risk. He is the strongest alternative in the member polling, but his route opens only if he wins Makerfield on 18 June. If MPs wait for him, they leave about twelve days for nomination, any NEC process, a political settlement, and a palace handover. If they do not wait, the obvious Commons-based challenger is Streeting, who polls badly with Labour members. That coordination problem is why the post-election revolt has not already become a completed change of PM.

Limitations

The largest uncertainty is private PLP and Cabinet coordination. Public counts of MPs who want Starmer to go are not the same as signed nominations for one challenger, and they are not the same as a Cabinet decision to force a PM handover. A coordinated Cabinet delegation could compress the timetable.

The Makerfield estimate is thin. The best constituency poll I found has a 3-point Burnham lead, a ±5.4-point headline margin of error, and early fieldwork before the campaign fully settled. By-elections are noisy, and this one is endogenous because voters know it may affect the premiership.

I did not use prediction markets in the model. This client question is narrower than many “Starmer out” framings: if he announces that he will resign after a Labour contest but remains the commissioned Prime Minister at 23:59 BST on 30 June, this question resolves YES.