“The numbers have been exaggerated […] the public perception in an opinion poll last year was that one third of all Labour party members were somehow or other under suspicion of antisemitism. The reality is, it was 0.3 per cent of party members had a case against them which had to be put through the process.” 

That was Jeremy Corbyn’s claim after the Equality and Human Rights Commission published its long-awaited report into how Labour dealt with antisemitism during his time as leader.

Responding to the report, which found “specific examples of harassment, discrimination and political interference” and “a lack of leadership”, Mr Corbyn said in a statement: “One antisemite is one too many, but the scale of the problem was also dramatically overstated for political reasons by our opponents inside and outside the party, as well as by much of the media.”

Mr Corbyn has since been suspended from the party and had the whip removed, meaning he is not currently a Labour MP.

But is there a factual basis for his claim?

Did 0.3 per cent of Labour members have a case against them?

It’s hard to understand how Mr Corbyn can claim to know how many Labour members were investigated by the party for antisemitism throughout his leadership.

This is because people close to him have always argued that Labour did not have proper systems in place to track antisemitism cases until they took over the management of the party.

Although Mr Corbyn became Labour leader in 2015, it wasn’t until 2018 that Jennie Formby, a close ally, became general secretary.

Ms Formby said consistently that she could not provide comprehensive data on antisemitism cases before she took over the post from Iain McNicol – who is not considered a Corbyn loyalist.

In 2019, Ms Formby told Labour MPs “there was no consistent and comprehensive system for recording and processing cases of antisemitism” before she became general secretary.

She made a similar statement in January 2020 when the party published statistics on antisemitism cases on its website: “The Party does not have the same detailed breakdowns of figures for the period before 2018 as a comprehensive central complaints system was not in use.”

Labour obviously gave the same account to the EHRC, which says the party “was not able to confirm the number of complaints made, or action taken on antisemitism complaints, until 2018 when it started to record this data”.

But in March 2018, Mr Corbyn gave an interview with Jewish News in which he cited what appeared to be very precise figures on the number of cases of antisemitism that had been investigated by the party since he became leader in 2015.

FactCheck approached Labour last year to ask how Mr Corbyn was able to recite such statistics then if, according to Ms Formby’s later statement, there was no “consistent and comprehensive system” in place for recording antisemitism complaints until she took over a month later.

Those representing Labour at the time denied that there was any inconsistency between Mr Corbyn and Ms Formby’s statements, though they were unable to provide a full explanation of how the two might be reconciled.

Update: Professor Greg Philo has contacted FactCheck to offer a likely source for Mr Corbyn’s claim: his 2019 book Bad News for Labour. In the book, Professor Philo takes the figures for antisemitism cases provided by Jennie Formby in 2019 and extrapolates back to estimate numbers across the whole of Mr Corbyn’s leadership. He says this estimate of 0.3 per cent of Labour members being involved in antisemitism cases is “well within what the existing evidence would indicate”. This estimate is not referenced in the EHRC report.

Did the public believe a third of Labour members were suspected of antisemitism?

It looks like Mr Corbyn is referring to a poll conducted by Survation on behalf of Greg Philo, professor of Communications and Social Change at the University of Glasgow, in 2019. The results formed part of his book on Labour’s antisemitism row.

It’s clear from the context of Mr Corbyn’s remarks that he thinks the poll is evidence of the public having a distorted view of the scale of Labour’s antisemitism problem, perhaps because of media misreporting.

The survey asked two questions, the first one being “Have you seen or heard anything about accusations of antisemitism (hostility to or prejudice against Jewish people) made against members of the Labour Party?”

A full 30 per cent of people polled said they had not heard anything, while 9 per cent answered: “Don’t know”.

Only those who had heard about antisemitism in Labour were asked a second question: “From what you have seen or heard, what percentage of Labour Party members do you think have had complaints of antisemitism made against them?”

It’s true that the mean average answer to this question was “34 per cent” – which is presumably where Mr Corbyn’s claim comes from.

But the most popular answer to question 2, picked by 29 per cent of respondents, was that they “don’t know” what proportion of members had been accused.

And of those who did put a figure on it, the most popular answer was “0-9%” – the correct answer, according to Mr Corbyn.

Is the poll otherwise reliable? The sample size was respectable at just over 1,000 people and the results were “weighted” to take account of demographic factors. Academic statisticians we spoke to told us that standard practice was followed in the way people were asked to select numbers from a range of possible answers.

But experts often caution against putting too much faith in the results of one question in a single poll.

Nick Allum, professor of Research Methodology at the University of Essex, said: “I would not want to say that the 34 per cent is ‘wrong’. But it is clear that other answers would probably come if the question were framed in a different way.”

“People may not have a clear idea of who are Labour members or what ‘reporting’ means. Are they thinking of Labour MPs or elected representatives, or rank and file members or officials?” he said.

“Single survey questions are always in principle quite unstable as measures of an underlying stable concept or belief, which is why ideally one would ask several questions or formulations of the same question.”

Dr Martin Lages, senior lecturer in Psychology at the University of Glasgow, said “ideally, the total number of party members should be provided as the reference when asking about a percentage.” Without this, he said, “responses about percentages tend to be subjective and unreliable across participants”.

Survation confirmed to FactCheck that those who took part in the survey were not given information about the total number of members in the party.

Update: Professor Philo told FactCheck that his qualitative research alongside the poll supported the idea that the public had overestimated the extent of antisemitism in the Labour Party and added: “The point about interviewees not being given data on Labour membership figures is not relevant. To do so would make it a deliberative poll in which subjects are given information with which they can qualify their views. We were investigating the impacts of media reports saying that Labour was ‘riddled’ with antisemitism and that the Party was ‘wholly infected’.”

FactCheck verdict

Jeremy Corbyn said: “the public perception in an opinion poll last year was that a third of all Labour party members were somehow or other under suspicion of antisemitism. The reality is, it was 0.3 per cent of party members had a case against them.”

It’s not clear that Mr Corbyn can reliably claim to know what proportion of Labour members have been the subject of an antisemitism case – at least, not for the duration of his time as leader.

Labour told the EHRC that “it was not able to confirm the number of complaints made, or action taken on antisemitism complaints, until 2018 when it started to record this data”. The party has previously said that there was no “consistent or comprehensive system” in place for recording antisemitism cases before then, which means we can’t say what happened in the first two-and-a-half years of his tenure.

The poll Mr Corbyn refers to does have some strengths (a large sample that is demographically representative).

However, it also revealed that a third of the public hadn’t heard anything about antisemitism in the Labour party, and that a third of those who had couldn’t say how many Labour members had been accused (“don’t know” was the single most popular answer).

Some academic statisticians FactCheck spoke to raised concerns that the way the question was asked might make its results unreliable. As with any single poll, we shouldn’t place too much weight on its findings.