Is Reform taking Labour votes?

Some seats between the two parties may be much tighter than many assume.

Reform is confident of winning millions of votes at the general election on Thursday. Insiders are certain they can beat Ukip’s total in 2015, when the party received 3.8 million. Some think the party could aspire to as many as five or six million – if these are taken directly from the Conservatives it would put Reform second to Labour in the popular vote. It is not likely, but it is not impossible either. But any noise of a split in the right-wing vote is music to the ears of Labour strategists and activists. It puts many more seats within Labour’s grasp, and makes its route to a landslide victory much easier.

But not everything is so simple. Ashford in Kent, for example, has never had a Labour MP before. At the moment, forecasts suggest it will go Labour, but only by the slimmest of margins – at 27.1 per cent, compared to the currently projected 24.7 per cent for the Conservatives and 23.8 per cent for Reform. The error margins leave this seat relatively open. And it is all predicated on two things: first, Reform taking an overwhelming number of otherwise-Conservative votes; and second: Labour retaining the popularity it has hypothetically won over the past two years. According to the data, this is what’s happening. But insiders are cautious: there are signs on the ground that votes once safely assumed as Labour are now wavering. There are even whispers that Labour – not just the Tories – might cede votes to Reform.

This runs contrary to the narrative at the start of this campaign, when Nigel Farage made his return to frontline politics. The dominant view was that Reform would pick up apathetic former Conservative voters, and fail to court those who voted Labour in 2019. This remains likely. But it is Labour‘s new base – the votes it assumes it is picking up from the Conservatives – that is vulnerable to a Reform sweep. For a party apparently cruising to an easy victory, this is enough to make some Labour organisers jittery.

If Labour’s share is overstated by 1 percentage point, the forecast number of Labour seats falls from 434 to 417. If it’s overstated by 2 points, it falls to 402. If by 3, it falls to 382.

But Labour winning 382 seats is at the far end of what to expect. The truth is that seepage to Reform is unlikely to reduce the party’s share by 3 points, the maximum point of an error margin in polling.

Here’s the maths. Labour should expect to win between 1.4 million and 2.1 million 2019 Conservative votes. That’s about 10-15 per cent of the Conservative base. If a portion of these are peeling off to Reform, such as, say, 5 per cent of the forecast Conservative-Labour defectors, that would represent a loss for Labour of 70,000-100,000 votes. It would swing a dozen or so seats back to the Conservatives. But it wouldn’t represent more than a percentage point fall in the Labour share.

In the grand scheme of things, it’s a very minor matter. Which makes it tricky to unearth in the data. But word on the ground is that a few seats will be much closer-run than assumed mere weeks ago.