Thursday’s elections showed how multi-party competition is transforming British politics. Two of the results illustrate the character of this change. In Hackney, Zoe Garbutt, the Green candidate, was elected mayor on a 19 per cent swing from Labour. A few miles to the east, Reform UK captured Havering Council, gaining 39 seats, most of them from the Conservatives. Two boroughs, two two-party contests, but each completely different.
If these two results have one thing in common, it is that many voters from left and right have been deserting the traditional main parties for the insurgents towards the edges of the political spectrum. The last time Thursday’s seats were contested, almost 80 per cent of them elected a Labour or Conservative councillor. This week, the combined tally of the Labour and the Tories is below 40 per cent. They are outnumbered by Green and Reform councillors, with Reform enjoying the largest share.
All that said, when we compare this year’s local election results with last year’s, we can see that left and right have diverged. Reform’s shares of seats and votes are down, while the Tories have recovered some ground. To the left of centre, Labour is down, the Greens up. For Sky News, Michael Thrasher has projected the Britain-wide share of the vote, compared with last year’s equivalent projection. Reform is down 5 per cent from 32 to 27 per cent. Labour is down four per cent, from 19 to 15 per cent. The Tories are up 2 per cent, from 18 to 20 per cent. The Lib Dems are down 2 per cent, from 16 to 14. And the biggest gain is for the Greens, who are up 7 percentage points, from 7 to 14.
The fact remains that Reform is still out in front. If our only comparison was with the 2024 general election result, Nigel Farage could justly claim that this week’s results represent a revolution in British electoral history. But we do have figures from polling and last year’s local elections. These all indicate that Reform has slipped back from its peak.
There is another feature of multi-party politics that the results have shown, and which Hackney and Havering also illustrate. We have five-party politics across England (and six-party politics in Wales and Scotland) but in the great majority of localities the battle remains essentially between two parties, largely due to first past the post. This explains the huge variety of results.
What we are witnessing is a local government version of tactical voting…This has potentially huge implications for the next general election
Normally, the movements in votes in one locality are broadly similar to those next door. No longer. In London’s 32 boroughs they were all over the place. Kingston-upon-Thames is a one-party borough: 44 of its 48 councillors are Liberal Democrats. Next door, Labour consolidated its control of Merton, with the Tories leading the opposition. Cross into Lambeth, and the Greens have thrashed Labour. In the great majority of London boroughs, either one party or two supplies at least 80 per cent of local councillors; it’s just that the identity of the two varies from place to place.
What we are witnessing is a local government version of tactical voting. Enough people know which parties are locally strong, and which aren’t, to shape the outcome of each contest. This has potentially huge implications for the next general election. In the 1990s, the Lib Dems established local defensive bridgeheads by winning council seats in different parts of the United Kingdom. These bridgeheads gave them credibility when the next general election came along. They helped the party make its big breakthrough in 1997, when their tally of MPs at Westminster jumped from 20 to 46, despite the fact that their national vote share had actually dropped.
Something similar could boost Reform and the Greens at the next general election. Two years ago, they won only nine seats between them, despite winning a combined total of 22 per cent of the Britain-wide vote. The next general election is likely to see both a big rise in their total vote, and a greater concentration of that increased vote in Reform and the Greens’ target seats. This week’s results have gone a long way to showing voters, and not just party strategists, where those seats are. The current parliament, in which Labour and the Tories won 532 out of 632 mainland seats, remains a bastion of two-party politics. It is now in danger of crumbling as never before.
Given the size of Labour’s landslide two years ago, Keir Starmer has an additional reason to worry. It’s not just the party’s terrible national support: 15 per cent on the basis of Thursday’s results, 18 per cent according to the polls. It’s also that Labour is threatened by different parties in different constituencies. In the past, the Tories were the challengers in the vast majority of Labour’s vulnerable seats. Next time, Labour will need to fend off significant numbers of Green, Reform, SNP and Plaid Cymru candidates, not just Tories.
Labour, then, is in deep trouble. No wonder its MPs are discussing whether to find a new leader. Starmer’s future was not any ballot paper on Thursday, but in a way it was on every ballot paper. The coming weeks might tell us if the prime minister will face a challenge. While the uncertainty persists, we can at least address the question objectively: does changing prime minister in mid-term help or hinder a government’s prospects of recovery?
The two examples most commonly cited are 1976, when James Callaghan succeeded Harold Wilson, and 2007, when Gordon Brown succeeded Tony Blair. Labour went on to lose the subsequent election after both handovers. Starmer’s allies point to this history to warn MPs not to repeat the same mistake. However, as any competent social scientist will tell us, two data points are insufficient to give us a general rule.
We can do better than that. There have been eight parliaments in the past 70 years when the prime minister at the end of the parliament was different from the one at the beginning. In crude terms, four of those end-of-parliament prime ministers stayed in office, while the other four were voted out. The successes were Harold Macmillan (1959), John Major (1992), Theresa May (2017) and Boris Johnson (2019). The four who were voted out were Alec Douglas-Home (1964), James Callaghan (1979), Gordon Brown (202) and Rishi Sunak (2024). May lost her overall majority in 2017, but stayed in office thanks to the Democratic Unionists. But the Conservatives, with 318 seats, remained well ahead of Labour, 262.
On those figures, the case for or against a change of PM can be argued either way. However, we should note that both Callaghan and Brown had opportunities to remain in office by calling earlier elections: Callaghan in 1978, before the winter of discontent, and Brown soon after he entered Downing Street in 2007. We can’t be sure that either would have kept their jobs, but it is far from certain that they would have been defeated. Brown enjoyed a three-month honeymoon, when he was ahead in the polls. He considered calling an early election but backed out. In retrospect it was a clear mistake.
In only two of the eight parliaments was the incoming prime minister facing the near certainty of defeat: Douglas-Home in 1964 and Sunak in 2024. Both took over when their party had been in office for 12 years and, unlike John Major in 1992, could not dispel the public mood that it was time for a change. Mind you, in Douglas-Home’s case, the surprising thing was how close he came to keeping Wilson out.
So there is no clear rule for deciding whether to change prime minister. It’s a risk either way. However, the record suggests some advice for Labour MPs pondering whether to move against Starmer.
First, a new leader must be the champion of change, not the status quo. Major scrapped Thatcher’s poll tax; Johnson sorted out Brexit. (The price we have all paid for his “achievement” came later.) What change could a new leader offer? An ambition to rejoin the European Union? Replacing first past the post with a fairer voting system?
Second, a new leader must avoid carrying the lingering odour of their predecessor. Eden had been broken by the 1956 Suez crisis. Although Macmillan had initially backed the doomed venture, he ended up insisting on withdrawing British troops. He escaped blame for launching a doomed war. Major did not just scrap the poll tax, he presented a more emollient and consensual style than Thatcher. If Labour’s new leader is a current member of the cabinet, they will need to work harder on this than Andy Burnham, who is not even an MP, let alone a minister or, say, Al Carns who, though a minister, is a fresh face and voice outside the cabinet. The more involved the new leader has been in running the country for the past two years, the more vital it will be to admit the mistakes made since 2024.
Third, a new leader must choose the right date for the next general election. Johnson got this right. Brown got it wrong; so, probably, did Callaghan. May would have got it right had her 2017 campaign not blown up over an unforced error on the politics of welfare reform. The Tories were stunningly successful in the local elections held shortly after that general election was announced; and they remained well ahead in the polls until the last fortnight of the 2017 campaign.
I know that this falls short of providing a simple answer to today’s exam question: should Starmer go? Either way, the crystal ball is murky. That’s politics for you. The one thing that is beyond doubt is that settling the leadership issue is just the start of the challenges that Labour MPs face.