Spoiler alert. Labour will win this week’s general election. However, important uncertainties remain. Will Conservatives be able to provide an effective opposition? Will Reform’s appeal surge or fade between now and Thursday? Will low turnout eat into Labour’s majority? Will the SNP still be Scotland’s top party?
Here are my predictions – and the assumptions that underpin them. I am certain about only one thing: at least one of my assumptions will be wrong. On Thursday night the biggest surprise would be if there are no surprises.
Labour: 37 per cent of the Britain-wide vote, 400 seats
Conservative: 24 per cent, 155 seats
Liberal Democrats: 13 per cent, 50 seats
Reform: 13 per cent, 2 seats
Green: 7 per cent, 2 seats
SNP: 33 per cent in Scotland, 18 seats
Other (including Northern Ireland): 23 seats
Labour majority: 150
On these figures, Labour is on course for a landslide in seats without coming anywhere near a landslide in votes. In parliamentary terms, Keir Starmer will have a gold-plated mandate. Whether it gleams so brightly to voters outside Westminster is another matter.
However, any prediction is subject to error. These are my assumptions, and why each could be wrong.
- The polls. Those showing Labour leads of 16-18 per cent are nearer the truth than those reporting leads of 23-25 per cent.
The companies showing lower leads adjust their raw data, in particular to take account of the two-plus million people who voted Conservative in 2019 but now say don’t know. This adjustments bring into play today’s equivalent of the “shy Tories” that gave John Major his surprise victory in 1992.
Why it might be wrong. Perhaps those ex-Tory don’t knows won’t “return home”. Instead they will vote Reform, or even Labour or Lib Dem, or simply register their protest by abstaining. In which case the larger leads will turn out to be right.
- The Farage factor. We have passed peak Reform. Some Tory defectors tempted by Farage will decide to vote Conservative after all.
Why it might be wrong. Reform’s support could stay solid and even rise further. That could help Farage have five or six colleagues in the new Parliament – and help Labour remove 20 or more Tory MPs who would otherwise hold their seats. Every extra Labour gain increases its majority by two. An extra 20 seats resulting from Reform’s vote holding up would increase Labour’s majority by 40.
- Turnout. This will be down. Labour will suffer more than the Conservatives.
One of the toughest challenges for pollsters is to decide which of their respondents will actually vote at all. In recent elections, the British Election Study has found that only half of people under 30 turn out, compared with around 80 per cent of those over 65. Labour’s lead is greatest among younger voters. The polls try to take all this into account; but if more under-30s stay at home than they expect, Labour will fall short of its predicted support.
This might be wrong in opposite ways. Anger with the Tories among younger voters might increase their turnout and help Labour increase its majority. On the other hand, Labour might suffer double trouble with many under 30s staying at home while others, confident that Labour will win anyway, feel strongly enough about Brexit, climate change and Gaza to vote Green or Lib Dem instead.
- Labour’s tax plans. Sunak will fail to persuade enough voters that he will tax them less than Starmer.
The polls have consistently found the most people think taxes will go up anyway, whoever wins the election. The dial didn’t moved when Sunak promised further national insurance cuts and predicted a four-year £2,000 hit to family incomes under Labour. According to both Ipsos and YouGov, more people still trust Labour than the Tories to handle taxes.
Why it might be wrong. In 1992, uncertain voters were wooed by a strong campaign against Labour’s “double whammy” on taxes. Sunak might yet regain some ground in the final days.
- How seats swing. For the first time since 1945, the Tories will lose most votes in normally safe Tory seats.
On the 13 point gap I am predicting, conventional models would show Labour struggling to win an overall majority, let alone a large one. This is because they assume the same swing in all kinds of seat – safe Tory, safe Labour or marginal.
This time, there is ample evidence that Conservative support is crumbling most in their strongest seats, making them far more vulnerable to Labour and the Lib Dems. Hence recent seat-by-seat MRP polling predictions of huge Labour majorities.
Why doubts remain. MRP surveys disagree wildly on how many normally safe Conservative seats will fall. If Tory MPs defending majorities of 10-20,000 are able to stem the haemorrhage – for example where they have built up strong personal reputations – then their party will not lose quite so many seats after all.
- Tactical voting. The Conservatives will suffer to roughly same extent as in 1997.
In that election they lost 30 extra seats because of the way Labour and Lib Dem supporters switched their votes to ensure the defeat of their local Conservative MP. Labour gained 20 of these, the Lib Dems the other ten.
Why this might be wrong. Tactical voting is certain to take place; what is in doubt is its scale. Some estimates suggest it could exceed what happened in 1997. This would add further to Labour’s majority and help the Lib Dems.
- Scotland. Labour will regain top place.
Based on recent polls, I predict that Labour MPs will outnumber SNP MPs by 28-18. depriving John Swinney, Scotland’s first minister, of his claim that SNP has a mandate for demanding a fresh referendum on independence.
Why it might be wrong. There are likely to be many seats won and lost by tiny majorities. A late swing back to the SNP and/or a small polling error could reverse those seat predictions, and revive Swinney’s mandate argument.
This analysis was first published by the Sunday Times
Peter Kellner. The Politics Counter 