Caerphilly hasn’t voted anything other than Labour in more than a century. A by-election there is meant to be a quiet contest in the incumbent’s back-yard. It is turning out to be anything but. Labour started off defending a relatively safe majority of 5,000 votes. Now the governing party is in a dead heath with Plaid Cymru and Reform.
Both challenger parties like their chances. Reform is confident, and Nigel Farage has visited the constituency. Plaid recently accepted the defection of the local Labour leader. And each challenger sees the other as their opposition – find me a leaflet of theirs that doesn’t portray this race as one between Plaid and Reform. Labour is a notable absence from perceived contention.
Political strongholds break gradually, then suddenly. Welsh Labour was showing cracks long before Reform’s insurgency. Ukip scooped up half a dozen Senedd seats in 2016. Labour held Caerphilly, but a dent was made. Ukip had picked up 22 per cent from nowhere, and Labour fell 14 points. When Ukip failed to stand a candidate in 2021, Labour’s vote soared 11pts. You don’t need a psephologist to tell you there’s a link.
If Ukip could get 22 per cent here in Caerphilly when the radical right’s national polling was half what it is now, then this week’s contest looks ripe for Reform.
Britain Predicts’s model has form for getting races right, or at least within a few percentage points. It was the most accurate model of the last general election, right about Runcorn, and only a few numbers off in Hamilton, Scotland. Its projection for Caerphilly is one to watch.
And here’s what it says. In 2021, Labour won Caerphilly with 46 per cent of the vote, with Plaid coming in second place at 28 per cent. The Conservatives polled 17 per cent.
This time around, Britain Predicts expects Labour to come third with 20 per cent of the vote – a fall of more than half. Plaid is predicted to come in second with 33 per cent, up five points on last time. And Reform, ahead of the pack, is expected to win with 36 per cent.
A health warning – this model has a tactical voting filter applied: top two polling parties becoming the preferred two. Were that not the case, then the seat would look thus: Reform 27 per cent, Plaid 27 per cent, and Labour 24 per cent.
Whichever way you chop it, whichever way you model it, this is a dead heat. Two percentage points in a race like this is five hundred votes, if that.
But Caerphilly proves an increasingly universal truth in British politics: no seat is a safe seat. Britain’s established parties have no idea what’s coming.
[Further reading: Does Reform already have more members than Labour?]
Listen to the New Statesman podcast
Content from our partners
Related
