Bad News for Labour: A response to Channel 4’s ‘FactCheck’

JVL Introduction

Last year Pluto Press published Greg Philo et al’s pathbreaking book Bad News for Labour: Antisemitism, the Party and Public Belief.

It showed quite how badly mainstream media were reporting and distorting the realities of Labour’s “antisemitism crisis”.

It was, predictably, virtually ignored by the same mainstream media…

Last week Jeremy Corbyn drew on information in the book about exaggeration of claims of antisemitism in his statement responding to the release of the EHRC Report.

Channel 4 has ‘fact checked’ Jeremy Corbyn’s statement and raised some doubts – wrongly and unnecessarily as authors Greg Phil & Mike Berry show in this blogpost.

This article was originally published by PlutoBooks blog on Tue 3 Nov 2020. Read the original here.

Bad News for Labour: A response to Channel 4's ‘FactCheck’

Buy Bad News for Labour

Channel 4 has ‘fact checked’ a statement by Jeremy Corbyn in response to the report by the EHRC on how the Labour Party dealt with antisemitism when he was leader. His statement was based on research in the book Bad News for Labour: Antisemitism, the Party and Public Belief, written by five academics including Greg Philo, Mike Berry, Justin Schlosberg, Antony Lerman and David Miller.

The Channel 4 FactCheck is critical of two points he makes:

  •  That ‘0.3% of party members had a case against them which had to be put through the process’

  •  Public perception was that ‘one third of all Labour party members were somehow or other under suspicion of antisemitism’. This was based on the results of an opinion poll commissioned for the book

The first criticism that they make of the 0.3% figure is that no comprehensive data was available on antisemitism cases before Jennie Formby took over as General Secretary in 2018. After taking office, she was then able to provide figures for the period from April 2018 to January 2019.

The figure of 0.3% appears on page 52 of our book. FactCheck did not contact us to discuss this and do not seem to have read all of the book. Had they done so, they would have seen that the figure is an extrapolation from the detailed figures from 2018 – 19 which take account of the absence of data from the earlier years. As we wrote, ‘if we assumed a constant level of cases over three years, the number would still come to just 0.3 per cent of the membership’. We also suggested that the number would possibly be less.

In January 2020, Jennie Formby published new figures which supported this. They related to the years 2017 – 2019. She gave figures on 1,201 complaints about members. Of those processed with decisions, 388 were dismissed as being without foundation and 321 were sent a reminder of conduct.

The total number who had left the Party was reported as 220. This included people who were automatically excluded as they were found to have supported another party, some whose membership lapsed, as well as others who decided to resign. This could be because they did not want to face the evidence against them or alternatively, they believed the process was wrong and that they were innocent. The actual number within the 220 who were expelled was just 56. A further 71 on top of the 220 were given formal warnings. There were still some unresolved cases but if we add those who have left to those receiving formal warnings, the figure would represent about 0.05 per cent of the membership at the time. In July 2019, Jennie Formby had written in a published reply to the Party Deputy Leader that ‘antisemitism-related cases that have been taken through the stages of our disciplinary procedures since September 2015 relate to roughly 0.06% of the Party’s average membership during this time’. It seems very likely that the 0.3 per cent figure cited by Corbyn is well within what the figures from the Labour Party would indicate.

The second issue that FactCheck discusses is the opinion poll from the book which Jeremy Corbyn quotes. FactCheck repeats the findings given in the book that a section of the public had not heard of the issue and others chose the ‘did not know’ option. They also note that, ‘The sample size was respectable at just over 1,000 people and the results were ‘weighted’ to take account of demographic factors.’

The respondents in the poll who knew of the issue then answered the 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?’ On this, FactCheck states: ‘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.’

They then seek to downgrade this result by focusing on the ‘most popular’ figure picked. This is a very strange way of reporting public opinion. The option which happens to have the most choices is unlikely to be very significant if there are another ten possible options. But they pursue this and give the ‘popular answer’ as 0-9 per cent. What they do not say is that only 14 per cent of those who expressed an opinion gave this figure while 86 per cent were above it.

They also note that the total number of Labour Party members could have been provided when asking the question. But this would then be a deliberative poll where respondents are offered information which they can consider when developing a more informed answer. We were investigating the impact of media reports saying that Labour was ‘riddled’ with antisemitism and that the Party was ‘wholly infected’. There is no point us giving information which the interviewees may not otherwise know, especially if that produces a more critical understanding of beliefs derived from the media.

They quote academics making the obvious point that if questions are asked in different ways then results may vary and that it is important to know how the question is understood by respondents. That is why academics like us conduct qualitative research using focus groups and interviews alongside the quantitative studies to check exactly on these issues. An extraordinary absence in the FactCheck account is that they make no mention of the qualitative research which we undertook for the study. This clearly showed the link between beliefs and media reports. Crucially, the qualitative work reproduced the trend which we found in the quantitative research adding to its validity.

Overall, FactCheck seems to be straining to find flaws where they do not exist.


Since the original publication by FactCheck, they have now updated it to note that the 0.3 per cent figure comes from an extrapolation. They have also noted the existence of our qualitative research and that to give data on Labour membership figures would have changed the research method to deliberative polling.

Comments (13)

  • Kuhnberg says:

    The establishment media is desperate to find Corbyn guilty of something, however trivial, so that they can both misrepresent it and blow it up into a major crime. For example, if Corbyn said that the public perception of the proportion of members of the Labour Party accused of antisemitism was 33% (not 34%) and that the proper figure was 0.3%, they would say, ‘Corbyn grossly understates the percentage of Labour members accused of being antisemitic!’ Naturally no mention would bd made of the 0.3% since that would mean confronting people with the truth of the matter.

    Sometimes I despair of the true situation ever being publicly acknowledged. I would like to comfort myself with the reflection that truth is truth, and that in time it is bound to persuade all but the most biassed — those who like Trump fans will always passionately believe the worst of the left. Alas, even Haaretz, the most liberal of Israeli papers, seems to believe that Corbyn is a dangerous antisemite.

    Oddly enough, today’s Guardian reprints Hilary Wainwright’s Red Pepper article — also reprinted on this site — on why Corbyn’s suspension should be lifted. Does this signal a change of heart for the paper that was once the voice of the liberal conscience? I fear not. Only yesterday morning they instantly deleted a comment I made below the line to the effect that Starmer had suspended Corbyn for voicing an inconvenient truth. And of course the tone of commentary on Corbyn will continue to be set by Jonathan Freedland, Nick Cohen, Andrew Rawnsley and John Crace.

  • Emma says:

    Thanks for keeping an eye on the detail this is what is needed when there is so much inaccuracy is so important to highlight the truth. You’d think a ‘Factcheck’ would have lived up to its name and not have to be corrected!

  • Harry Law says:

    So true Kuhnberg. I saw a picture of Corbyn going to a meeting he was wearing a mask, of course the comment was made that his mask was not fitted correctly, not entirely flush with one of his cheeks. Fortunately they did not comment [or spot] his tie was not pulled up to his collar. [scruffy anti-Semite /sarc

  • DJ says:

    No point fact checking the EHRC report. Hardly any hard facts to check!

  • Philip Ward says:

    I haven’t seen the details of the survey that found the “mean average” estimates of antisemitsm cases in the LP was 34%. Clearly, Channel 4 are wrong to use the mode – the range that is guessed at by the highest number of people. But I wonder if it actually correct to use the mean when the mode is the lowest decile, in other words in a highly skewed distribution? Surely the median would be a better measure – the value that has 50% of the respondents on either side?

  • John Bowley says:

    ‘Mainstream’ media is institutionally incompetent and morally corrupt.

  • Linda says:

    “Oddly enough, today’s Guardian reprints Hilary Wainwright’s Red Pepper article — also reprinted on this site — on why Corbyn’s suspension should be lifted. Does this signal a change of heart for the paper that was once the voice of the liberal conscience? I fear not …”.

    The “Guardian” takes this approach quite often – publishing just one article with contrary views to the dominant narrative.

    That said, maybe there’s some back-tracking. Enough CLPs, unions and MPs have protested at the perceived unlawfulness and unfairness of the behaviour towards Corbyn to surprise and perhaps worry his attackers.

  • George Wilmers says:

    Philip Ward is right to point out above that it is arguably more reasonable to use the median value than the mean value of people’s estimates in the above context, but it is clear from the limited data given that the resulting median value would, if it did not actually exceed the mean value of 34%, surely be closer to it than to 0.3%., so that changing “mean” to “median” would not alter Corbyn’s substantive argument. On the other hand the suggestion by Channel 4 to use instead the mode in this context is so bizarre as to suggest either mathematical illiteracy or bad faith in an attempt to discredit Corbyn.

    What also needs pointing out however is the sheer absurdity of the intellectual contortions in which Channel 4 is now indulging in order to challenge the statistical evaluations of data cited by Corbyn in defence of his assertions concerning political distortion of the extent of Labour antisemitism. This is the most astonishing chutzpah on the part of an organisation, which in common with the entire corporate media, has repeated wholesale for four years a mass of politically motivated scurrilous accusations on the subject of antisemitism, without bothering to do the most elementary fact checking.

  • Unlike my usual, tolerant (?) self I have to admit to extreme irritation on the constant and largely unnecessary examination of, and debate on “antisemitism” What was said, who said it, when was it said and so on!
    My possibly, simplistic mentality, demands a simple answer to all this!
    Lets cut the crap and tell it like it is! (Sorry, two days of U.S. election) But everything I read from those of us who refuse to accept that either Corbyn or the Labour Party are institutionally antisemitic takes a defensive tone!
    I say, without hesitation, that this MUST stop! It is NOT on us to prove our innocence of these accusations! The burden of proof is squarely on the accusers as established in British law!
    Instead of debating individual points we must simply refute all and every utterance that alleges what we KNOW to be an abuse of factual reality.
    What John Bowley says of the media (morally corrupt) applies equally to those who use the media to to justify and disseminate their abuses of factual reality.
    Allegations of A.S. in labour or Corbyn are based on a nonsensical and flimsy construct. Tear it down and chuck it! Why not?
    Taking them seriously and trying to debate got Jeremy nowhere, what`s changed? Call them what they are, blatant abusers of factual reality, and WHAT is the point of trying to have dialogue with those who are known to have not the slightest interest in anything else but their own lurid inventions.

  • DJ says:

    On the question of anti semitism in the Labour Party the MSM is an evidence free zone. The EHRC report like the MSM assumes the Labour Party is infested with anti semites. It doesn’t actually try to establish the scale of the problem it was asked to deal with by the JLM and the CAA. It’s purpose was to provide something that the establishment and Israeli lobby could utilise to “shame” Jeremy Corbyn. The odd bit of so called harassment and illegality was all the MSM needed to shock and horrify the public with this”damning”report.

  • Harry Law says:

    The overall number of members removed from the party as a result of our disciplinary processes for antisemitism, broken down by quarter and by type of removal.
    Expelled 2018 10
    Expelled 2019 45

    Table 1 displays the overall number of members removed from the party as a result of our disciplinary processes for antisemitism, broken down by quarter and by type of removal
    116 members chose to resign in that same period, yet the above paragraph claims wrongly that they were removed.
    The 55 figure represents .01% of the approx 550,000 membership.

  • rc says:

    [Your JVL web editor has failed to follow the argument in the post below.

    In the belief it is maybe making an interesting and/or important point it is being approved so others can assess it. But please – we don’t want an extensive correspondence on this one!]

    I am struggling with the table on p 5 off ‘Bad News for Labour’. I reproduce below Column 1 of Table 1.1 ( estimates of the percentage of LP members accused of racism (AS), by estimated % and number of respondents) converted from vertical to horizontal form: (for ease of handling I have lettered each estimate band: i and l are not used because of their resemblance to the figure ‘one’:
    a. b. c. d. e. f g. h. j k m. n
    %:0-9 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-99 100 DK
    14 11. 13 8. 5. 6. 4. 4. 3. 2. 0. 29.

    The mode is actually Don’t Know (n) , not, as misreported by Channel 4 0-9 (a).
    But that is dependent on including DK (n), which is surely irrelevant. Excluding DK again, the median is in the 20-29 band (c). Because a+b = 25 and d+e+f+g+h+j+k+m = 32. To achieve the equality required by the concept of median, we must divide c’s total of 13 into 10, to be added to a+b, making 35, and 3 to be added to 32, making 35. We know nothing of the distribution of the estimates WITHIN band c, which might be every one at 20, or everyone at 29, or surely something in between. The minimum possible of all the actual estimates (ie omitting band n), is therefore 20%; the maximum is 29%. So on this basis a minimum of one fifth of the British public believed that a minimum of one fifth of the LP membership had been accused of racism (AS). And on the same basis over a quarter of the British public believed that over a quarter of LP members had been accused of racism (AS).
    Criticisms of the concept of mean as applicable to these representations suggest that we need not now go into the calculation of the mean. suffice it for the time being that 70 (the sum total of bands a through m) divided by 11 (the number of relevant bands) is 6.3686 recurring. Including band n (DK) at 29, we divide 99 by 12 and find a mean of 8.25.
    I offer the above rough workings to assist other comrades, who have, unlike me, formally studied maths since 1970, to polish them up. (for technical reasons I cannot save the workings and and reluctant to lose them to the inadequacies of my edit-copy-paste programme.

  • RC says:

    I have gone through my workings and found an elementary mistake. The average (mean) of 34% of LP members estimated by the public to have been accused of AS racism, given by Philo et al is absolutely correct. You may rely implicitly on their p 5 table and on their interpretations thereof.
    Whatever he may say now, Corbyn was absolutely right to comment that the extent of LP members accused (let alone convicted) of antisemitism had been over-dramatized for political reasons.
    My apologies for any confusion that may have arisen from my prior incompetence.

Comments are now closed.