The three things Labour got wrong in Caerphilly – and how to fix them

The Labour Party must take this devastating Caerphilly by-election defeat as an opportunity to assess how it turns around its rapid decline, writes John Rentoul – and I can think of three ways to do just that

The result of the Senedd by-election in Caerphilly was a triumph for Plaid Cymru, pointing towards the nationalists taking over the government of Wales for the first time at next year’s elections. And it was a disappointment for Nigel Farage. His party has a lot of leftover balloons that were ready for the celebration of a Reform victory – the party increased its vote hugely but still fell some way short.

The result looks bad for Labour, and indeed the party did even worse than the only opinion poll in the constituency suggested. The Survation poll put Labour on 12 per cent of the vote but, on the day, Richard Tunnicliffe, the Labour candidate, won 11 per cent.

Yet there may be some hope for Keir Starmer if his party can learn the right three lessons.

Firstly, the idea that there is a Labour core vote that will turn out for the party, come what may, needs to be downgraded. That should focus minds. The party might have thought that it could rely on a proportion of voters in formerly safe seats to remain loyal. But in this case, Labour’s vote fell off to such a degree that it does not provide much of a base for next year’s all-Wales elections for a new Senedd, let alone for the next UK-wide general election.

Labour cannot afford to dismiss the result as a protest vote (Andrew Matthews/PA)

It was a by-election, which as Peter Kellner, the former president of YouGov, says is a “time to moan” rather than a “time to decide”. But Labour cannot afford to dismiss the result as a protest vote, because the people of Caerphilly might be perfectly happy to vote for Plaid Cymru for the Senedd and for the House of Commons if Labour does not give them a reason to return. The party must incentivise voters to choose them, not simply rely on its core turning out to vote.

Secondly, he will note that Plaid probably won this by-election only because of tactical voting. This may be the key to heading off Reform. It seems that the voters recognised that the election was essentially a contest between Plaid and Reform – as the Survation opinion poll suggested – and enough former Labour voters switched to Plaid to deliver the result. The Green and Lib Dem were squeezed too: their candidates won only 1.5 per cent of the vote each.

It also appears that the threat of a Reform victory had a galvanising effect on anti-Reform voters. Turnout was up since the last Senedd election in 2021, at a respectable 50 per cent, which would be high for a parliamentary by-election.

This means that Labour can take a paradoxically hopeful message from this result. If anti-Reform sentiment is strong, that gives Labour the chance of defending the Labour-Reform marginals in England, which will probably be the main battleground at the next general election.

People will vote against Labour if there is an alternative – unless the alternative is Reform, in which case anti-Reform tactical voting will trump anti-Labour tactical voting. That is Starmer’s plan for the next general election, known as the “Macron strategy”, after Emmanuel Macron twice fought elections in France demanding that voters choose between him and Marine Le Pen, the anti-immigration candidate.

If it works in Britain, it would mean that Labour would lose seats to Plaid in Wales, to the SNP in Scotland, to the Lib Dems in their few remaining target seats, to the Greens in a few graduate-heavy urban seats, and to the Gaza-Corbyn-Sultana collective in a some other urban seats – but that it would hold onto its seats where the main challenger is Reform.

As there are many more seats where Reform is the challenger to Labour than where the other parties can win, this Macron strategy is currently Starmer’s best option.

Lastly, what Starmer must take to heart most of all is that a lurch to the left is not what Labour needs. There are many voices, especially from within the Labour Party, who say that the lesson of the Caerphilly by-election is that Labour should do just that and give people who have voted Labour in the past “something to vote for”.

Alun Davies, a minister in the Welsh government, said on the Today programme that “Labour is not really being Labour.” He said: “People didn’t like the way that we spoke and sometimes dehumanised refugees ... we shouldn’t be using the language of Reform.”

I think this is a misreading of the vote borne of wishful thinking. Of course Labour should be trying to give people something to vote for, but if that means putting up taxes to pay for further increases in public spending, and allowing the small boats to continue to arrive, then that is only going to make it harder to fight off Reform in those Labour-Reform marginal seats.

The left-turners say that Labour isn’t going to beat Reform by aping Nigel Farage, but it could equally be said that Labour isn’t going to hold onto its voters by aping Zack Polanski, the Corbynite leader of the Greens.

The lesson of Caerphilly is that Labour lost a large chunk of its vote to Reform, as well as losing a lot of tactical voters to Plaid. Labour needs to hold onto as many of those voters as possible in those seats across England where the next general election will be decided, as things currently stand, while at the same time benefiting from anti-Reform tactical voting.