Why we can’t be sure of the size of the Labour swing

Good morning. Three new MRP projections in the past two days have forecast the biggest Labour victory, and the largest Conservative defeat, in British political history. (A fourth merely projects a loss that might only be slightly worse than the one the Conservatives experienced in 1997.)

We have very good evidence that the opinion polls are about right: namely, the results in local elections and parliamentary by-elections, and the behaviour of the party leaders. On Tuesday, Rishi Sunak was campaigning in Torridge and Tavistock (a new seat with a notional majority of 41.9 points), and yesterday he was in Suffolk Coastal (35.2 points).

Thanks to my own travels around the country and the huge amount of election literature many of you have kindly taken photos of and sent my way, it is pretty clear to me that the Conservatives have essentially abandoned their most vulnerable seats — and sensibly so, given the political backdrop.

Although one reason why the Tories might not do quite as badly as forecast is that the Liberal Democrats do not have a great deal of spare resources to expand their target list of seats. George Parker and Anna Gross have an excellent piece on ‘Project A30’, the Lib Dems’ plan to extend the scope of their campaign. But it requires money that they may not be able to raise.

One reason why, beyond the big picture of “major Tory rout”, the polls and projections do not agree is that it is hard to exactly model the impact of local campaigns. That makes it tough to predict exactly how well the Lib Dems, Greens and Reform will do when it comes to picking up seats.

Some thoughts on some of the other things we can’t be certain of as they stand below.

It don’t mean a thing, if you ain’t got that swing

Generally speaking, what tends to happen in British elections is something close to uniform national swing, particularly in marginal constituencies. Simply put, if the Labour vote rises by five points at the Conservatives’ expense in Harlow, a perennial marginal, then it will also rise by a similar amount in Hackney North, an ultra-safe Labour seat — and in South Holland and the Deepings, a very safe Conservative constituency.

But what we’ve seen in local elections this parliament is something quite different; that is, incredibly efficient changes in the composition of the Labour vote. It is treading water and in some places, including its very safe inner city constituencies, actively going backwards — but it is gaining votes in the marginal constituencies and formerly safe Tory seats, where the election is ultimately decided.

Labour’s narrow win in the West Midlands mayoral race is a good example of that: Richard Parker got fewer votes in Birmingham than Siôn Simon or Liam Byrne did in their defeats in 2017 and 2021, respectively. But that didn’t matter, because he did a lot better in Walsall and Dudley than they did.

For the general election, that matters a lot — because there are a lot more constituencies that look and vote like Dudley than there are that look and vote like Birmingham. (And in any case, Labour holds those constituencies already.)

What we don’t know is how much the electorate in this general election is going to mirror the behaviour of the electorate in recent local elections. We often talk about what has been happening in this parliament as “tactical voting”. (This is essentially when Jane, a Lib Dem voter, votes tactically for a party she likes less because they are better placed to defeat a party she hates.)

Given our electoral system, that usually means voting Labour. But what has happened in the local elections is better understood as “negative polarisation”, when Jane is frankly indifferent to whether she is pulling the red Labour lever, the yellow Lib Dem one, or the green one for the, uh, Greens — because what she really cares about is defeating the Conservatives.

Voters like that are why Labour and the Lib Dems have enjoyed very good sets of local elections in terms of seats won but not so much in terms of actual votes. One reason to think that won’t hold in an election is that general election voters are just less engaged than the people who bother to vote in off-years.

But one reason to think it will is stories like this one:

When Terry Gibbs returned with his family to find his home near Esher in outer London being burgled last year, he called the police — only to be told “nobody could come”. “[They said] ‘You’ll just have to enter with caution’,” he recalled. Although police came the next day to take fingerprints, the force — Surrey Constabulary — has the second-worst record for solving burglaries in England and Wales.

The government’s record since 2019 is not good and Sunak’s administration can’t be said to be doing the basics right. Put that together, and the strikingly efficient defeats suffered by the Conservative party in local elections begin to look like the most likely outcome in the general election too.

Now try this

Continuing my election time trend of re-reading books I know I love, I am re-reading Ned Beauman’s excellent The Teleportation Accident. Beauman is a terrific novelist whose cleverness with language and whose humour pairs well with some engagingly dark set pieces and moments.

Top stories today

Below is the Financial Times’ live-updating UK poll-of-polls, which combines voting intention surveys published by major British pollsters. Visit the FT poll-tracker page to discover our methodology and explore polling data by demographic including age, gender, region and more.