Is Boris Johnson losing the red wall seats that he gained in 2019?

The Conservatives are doing worse in the constituencies that they won for the first time at the last election according to an opinion poll, writes John Rentoul

Tainted love?: Boris Johnson in Middlesbrough during the 2019 election campaign (Getty)

A Deltapoll survey carried out in the 57 constituencies gained by the Conservatives at the last election suggests that Labour is doing better in these battleground seats than the national average.

The poll, carried out for the Mail on Sunday, suggests that in these seats there has been a swing from the Tories to Labour since the general election of 14 per cent, compared with an 8 per cent swing nationally. This may not be particularly surprising, in that we might expect those voters who have only recently detached themselves from their Labour loyalty to be less secure in their attachment to the Conservatives, but it is a warning sign for Boris Johnson.

The finding has to be treated with caution, because the sample was relatively small: just 612 voters interviewed online between 23 and 30 December. This means the margin of error is larger than that for a normal national poll, which would usually have a sample of between 1,000 and 2,000 people. However, 600 is acceptable, and enough to suggest that there is a genuine difference in Labour’s favour between new Tory seats and the national picture (for which Deltapoll had a sample size of 1,567).

The poll suggests that, even if Labour were to fall back in the national opinion polls, its ability to outperform the national average in the seats it needs to gain to deprive the Tories of a majority next time means that it has a head start. It may even be enough of an advantage to offset the losses Labour is likely to suffer as a result of boundary changes that will be brought in next year to equalise constituency electorates.

This good news for Keir Starmer is supported by some of the detailed findings. In the 57 new Tory seats, he is preferred to Boris Johnson as prime minister by 38 per cent to 33 per cent. What is more, the team of prime minister Starmer and chancellor Rachel Reeves is preferred to Johnson and Rishi Sunak by 40 per cent to 33 per cent.

The damage done to Johnson’s reputation by the stories about Christmas parties in Downing Street a year ago is suggested by the finding that only 16 per cent of voters in these battleground constituencies think the prime minister has obeyed the rules, while 72 per cent say he has not.

Pedantically, these constituencies are not all “red wall” seats in the original meaning of the term. That phrase was coined by James Kanagasooriam, a statistician who noticed, before the 2019 election, that there were a lot of constituencies that tended to vote Labour although demographically – by education, income, ethnic minority mix and home ownership – they ought to have been Tory. The most important cluster of these seats he identified as “a huge ‘red wall’ stretching from north Wales into Merseyside, Warrington, Wigan, Manchester, Oldham, Barnsley, Nottingham and Doncaster”.

Those seats, along with what Kanagasooriam regarded as a separate cluster in northeast England, were the ones that became identified with Johnson’s victory in 2019, although there were others in the southwest, and in south Wales. But the Deltapoll survey includes all Tory gains, whether they were part of Kanagasooriam’s red wall or not. Many of them are simply marginal seats, which would have been expected to change hands if the Tories increased their share of the vote nationally.

So the Deltapoll survey doesn’t tell us precisely what is happening to the true red wall seats – those seats in which Johnson overcame cultural resistance to voting Tory among people who in any other part of the country would have been natural Tory voters.

But it does suggest that Labour is doing better in precisely those seats that it needs to win, and if that holds up between now and the next election, that could be significant.