Opinion: Attack YouGov if you want to – but at least say who you are

Last week’s Daily Telegraph article attacking YouGov’s polling raises some disturbing issues about the quality of political debate as we fast approach the general election.

Firstly, the article’s authors seem to have no understanding about how polls should be conducted. They complain that the raw data in one large aggregated survey “were…‘weighted’ using an undisclosed YouGov formula which reduced the lead to sex per cent [from 12 percentage points in the raw data].” But all reputable pollsters know that their sample will not usually be representative of the population, for example by having too few women or too many Guardian readers, and so adjust the weight given to each person’s responses so that the sample matches the population. And as for the formula being “undisclosed”, it is laid out openly at the end of each YouGov poll (on page 3 of this pdf, for example). If the political staff of a leading newspaper to have no understanding of the polls they used to commission (YouGov worked for the Telegraph until a few weeks ago) it makes you wonder whether they are up to the job.

Secondly, from the e-mails between Peter Kellner and Robert Winnett, one of the article’s authors, it is clear that the newspaper was fed questions, perhaps by the anonymous source posing as a polling expert in the article. Unfortunately, this person either knows as little about polling as Mr Winnett or has an agenda. The source implies that just because the sample is large it doesn’t need to be weighted. Anthony Wells, who works at YouGov, rightly points out that a large unrepresentative sample is a lot worse than a small but representative sample in an excellent post that demolishes the insinuations in the article.

Unless Mr Winnett is a complete nitwit there is only one conclusion to draw: someone high up in the Conservative party (perhaps the anonymous shadow cabinet member quoted in the article?) is orchestrating a campaign to discredit one of Britain’s most accurate pollsters through the newspaper that used to commission those polls, simply because they don’t like the results of the polls. (Though this has not stopped the Tories from employing YouGov for private polling.)

For far too long British newspapers have tolerated anonymous politicians using their pages to wage vicious campaigns to discredit the opponents, and sometimes even their colleagues – remember the “forces of Hell” unleashed on Alastair Darling. Unless more journalists are as gullible as Mr Winnett this particular storm will probably blow over. But, on principle, I feel that politicians who want to make such attacks should at least have the decency to name themselves when doing it. The way the media can encourage a more responsible political debate is by refusing to report hatchet-jobs unless the people giving the briefings are willing to stand up and be counted. Until then, the kinds of smears so devastatingly documented by Nick Cohen in an article last year will continue to sully our newspapers and drive decent people out of politics.

Niklas Smith is Junior Treasurer of Cambridge Student Liberal Democrats. The Voice has also covered the Telegraph story in this post.

Read more by or more about , , , or .
This entry was posted in Op-eds and Polls.


  • the reason the telegraph went for yougov is because they walked away from the newspaper and went to the sun. which offered them a more lucrative daily polling deal for the election. it’s what passes for journalism at the telegraph these days – giving over pages to personal and corporate vendettas…

  • Tony Dobbie 3rd May '10 - 5:55pm

    Well, you obviously need to learn how to read and comprehend Mr. Smith.

    The Telegraph is fully aware about weighting. If you think a formula consists of 1+1 = 11 as per the ‘disclosure’ at the end of the You Gov poll you cite then you might try going back to school. The point is how they decide their weighting, not what the weighting is you fool.

    One with an analytical mind might ask, for example, how they determine social class for their polling?

    Seriously, if your analytical skills are the best the LibDems can come up with Niklas then thank god we do not have proportional representation.

Post a Comment

Lib Dem Voice welcomes comments from everyone but we ask you to be polite, to be on topic and to be who you say you are. You can read our comments policy in full here. Please respect it and all readers of the site.

To have your photo next to your comment please signup your email address with Gravatar.

Your email is never published. Required fields are marked *

Please complete the name of this site, Liberal Democrat ...?


Recent Comments

  • Rif Winfield
    Sadly I think that some of your contributors don't appreciate the scale of the assault on the least well off in society. Look for example at pensioners - those ...
  • Alex Macfie
    Roland: Nitpicking? Not really, just pointing out the unenforceability of a distance-based restriction for free train travel, The refund-based model is interes...
  • Peter Davies
    I would challenge you to name the wife of a premiership footballer who is not declaring enough income in their own right to use up their tax allowance. This is ...
  • Joe Bourke
    Peter Davies, A £100 per week MIG and tax reducer does not end poverty but it is £1200 (£23/week) more per year than what you get from the UC standard all...
  • Joe Bourke
    Michael BG, I have to keep correcting your statements because this a minimum income guarantee. It is not (as you keep stating) a Universal basic income for t...