One of Britain’s leading polling companies has denied claims by Nigel Farage that it is downplaying Reform’s popularity among voters.
YouGov, whose voting intention surveys are published in The Times, was accused by the Reform UK leader of using “bizarre adjustments” to “suppress the true figures” about his party’s support.
In a letter to the company, Farage said that across YouGov’s past ten polls, Reform was on average five points lower than surveys conducted by other companies.
His attack follows YouGov’s latest poll, which shows Farage’s party just four points ahead of the Conservatives, with the support of 23 per cent of voters. Other polls have Reform’s share of the vote as high as 30 per cent.
However, YouGov says that its approach to voting intention polling is more accurate because it is designed to pick up tactical voting. This approach correctly predicted Reform’s share of the vote at the last election.
Overall, most pollsters have found that Reform’s support has declined since its high point last autumn.
YouGov changed its approach to its weekly voting intention polls during the last election campaign in 2024 in an attempt to pick up tactical voting that has become an increasing feature of UK elections in recent years.
Unlike other pollsters, YouGov now asks two questions of voters: how they would vote if a general election were held tomorrow and how they would vote in their own constituency.
It is the answer to the second question that drives the final polling number because the company believes that this represents a more accurate assessment of true voting intention, given the number of people who are known to vote tactically.
Data released by the company showing the results of five polls carried out between December and January reveal that Reform’s poll rating on YouGov is lower because the answer to these two questions is different.
When asked how they would vote if there was a general election tomorrow, 29 per cent said Reform, largely in line with the results of other polling companies. However, when asked how they would vote in their constituency, this figure dropped to 26 per cent.
The poll is also weighted for turnout and people who say they “don’t know” how they will vote. YouGov notes that this addresses the problem of more people in polls saying that they will vote than actually do.
A YouGov analysis of its performance at the last election found that its new voting intention model was significantly more accurate than its previous system that simply asked voters how they intended to vote without asking about their constituency.
Its “final call” of the election accurately predicted that Reform would win 15 per cent of the vote compared with its previous method that would have had Reform winning 20 per cent support.
YouGov’s final seat projection was also the most accurate of the constituency-based polls in terms of seats called, with 92 per cent correct.
In an article explaining YouGov’s polling method, Anthony Wells, global head of politics and elections, said that the company’s post-election review found its approach was “substantially more accurate than more traditional methods”.
He added that the company also published tabulated results on its website that showed the “results to the questions without the modelling, and with ‘don’t know’, ‘won’t say’ and ‘would not vote’ responses included”.
However, in his letter to the company, Farage claimed that YouGov was underestimating Reform’s support. “Political polling plays a significant role in shaping public perception, particularly on social media,” he wrote.
“I would welcome a thorough explanation of what you believe accounts for the consistent discrepancy between your results and the rest of the polling industry.”
In a recent article, Peter Kellner, the political analyst, said that it was “clear” that YouGov’s figures were “out of line” with those of other polling companies.
However, he said that the differences could be “explained by YouGov’s innovations in measuring voting intention”, adding that “competition in polling methods is a good thing”.
“As long as we understand what is being done, we can make sense of the figures and why pollsters differ, even as we acknowledge the element of uncertainty around individual percentages,” he said.
“All the pollsters I have contacted are continuing to test and improve their methods. They have a clear vested interest in being as right as possible [and] should be applauded for their candour.”
