Keir Starmer proved the crystal ball wrong this year – what has 2025 in store?

Keir Starmer is the second prime minister to take a close interest in the proceedings of a mysterious and secretive cabal, the annual Peter Mandelson Memorial Dim Sum Supper.

This is a group that has nothing to do with Lord Mandelson himself: it is a group of my friends who were dining in Chinatown on the day of his resignation from Tony Blair’s cabinet in 1998. By coincidence, we were dining there again on the day of Mandelson’s second resignation two years later.

Both resignations happened around Christmas, so we made an annual event of it, and used the gathering as a chance to make predictions about the year ahead in politics. I have written an end-of-year column about our deliberations for many years now, although I make no great claims about the quality of our forecasts. Berkshire Hathaway we are not.

However, it turns out that two prime ministers have been among the subset of the population who noticed our prognostications.

When I spoke to Keir Starmer at his election count in July, before he was declared duly elected the member of parliament for Holborn and St Pancras, he said he had read my “dim sum predictions” and asked: “Weren’t you saying it would be a hung parliament?”

I had to admit that this was the majority view of the dim sum diners this time last year, and that those predicting a Labour majority were in the (albeit quite vocal) minority. I felt Starmer was entitled, at that moment – poised on the threshold of a huge landslide victory – to take some satisfaction in confounding the doubters.

After all, the other prime minister who reminded me about a dim sum forecast was David Cameron, in the run-up to the 2015 election. At the end of 2014 we had also predicted a hung parliament – we thought that the coalition government of 2010 would continue – and I think he was already confident that he would win enough Liberal Democrat seats to gain a majority. “We need to win a lot of seats off them,” he told me. He, too, produced an unexpected election win.

Fortunately, this year’s dim sum dinner, held at an undisclosed location last week, did not feel under too much pressure to predict the outcome of the next general election. There will be plenty of time to make many different predictions before then.

For the moment, though, we thought the election would be in the summer of 2028, although there was also one vote for the autumn, and that Labour would win with a majority of around 50. The consensus view was that, although the government had got off to a bad start, the Conservatives had so forfeited the trust of the voters, especially on immigration, that it would take longer than a parliament to recover – and that Nigel Farage would continue to divide the anti-Labour vote without Reform actually breaking through to replace the Tories as the main opposition.

Turning to more immediate elections over prawn toast and pak choi, the general view was that the 2025 local elections would give Starmer some breathing space, because those are seats that were last contested at the height of Boris Johnson’s vaccine-boosted popularity in 2021. Labour MPs in marginal seats might postpone their inevitable panicking for another year. We thought that the Scottish parliament in 2026 would revert to its original design, which was intended to produce hung parliaments in which Labour would govern in partnership with the Lib Dems.

As the Peking duck arrived, we tried to predict the fate of individual politicians over the next three and a half years. Again, some of our predictions were unanimous. We thought that Rachel Reeves would survive as chancellor, but that David Lammy and Ed Miliband would not last the parliament. And we agreed that Boris Johnson would not return to the Commons – my view prevailed that Tory party members have worked out who was responsible for net immigration of nearly one million in 2022.

We predicted that Starmer would lead Labour into the election, but there was one late waverer, who argued that, even though it is hard to eject a Labour leader (and Starmer’s rule changes in 2021 made it harder), the “men in blue suits” might prevail upon him. In the end, though, we all agreed that he would survive.

After that, the consensus broke down. The majority thought that Kemi Badenoch would lead the Tories at the election, against the view of the minority who thought that her service in the last government would drag her down. The minority noted that Labour was already putting out social media adverts of her at the despatch box welcoming a relaxation of visa rules.

We didn’t have much time to devote to international politics. We thought Emmanuel Macron would serve out his term until 2027: a majority thought this would be Marine Le Pen’s last chance, and that she would lose. We didn’t think Donald Trump would try to change the US constitution to stand again, and we thought JD Vance would be the Republican nominee in 2028.

But at this point the napkin on which I was making notes is smudged with green tea stains, so you will have to wait until next year’s dim sum supper for a fuller prediction.

One projection I think is a fairly safe one is that in 2028 or 2029, I shall be back in Camden Town Hall for Starmer’s election count, and he will ask, deadpan: “Weren’t you saying it would be a majority of 50?”