Welsh poll - read the small print!

Before Labour supporters start drowning their sorrows and Conservative supporters start dancing in the street in reaction to today's YouGov poll - a word of caution. Apparently The figures have been weighted and are representative of all GB adults (aged 18+). This has been one of the reasons why most previous Welsh polls have proved disastrously inaccurate. Wales is not a mini GB. In Wales the C2s Ds and Es are a much more important factor in elections than the C1s who swing middle England for one party or an other.

Its good to see the beginning of regular Welsh polling, but until there have been a number of Welsh polls which can be used to show trends and until there have been enough polls conducted near real elections in order for the weighting to be adjusted to reflect actual voting patterns the results of Welsh polls must still be treated with caution and digested with a large pinch of salt.


  1. Good point, MOF. YouGov admitted this in their Cardiff Bay presentation today. The social groups you mention have least access to and use of the internet, on which their poll is founded.

    However, they want to improve their methodology. And thank goodness there's regular opinion-polling here at last!

  2. It is a very good point - and I agree with every word of it. It's still a pretty extraordinary poll, though.

    The Euro elections provided a significant warning to Labour that they are living in a fools paradise if they think they are entitled to the Welsh vote, even in ultra-traditional Labour-supporting Carmarthenshire.

    I'm probably wrong (I usually am) but I can see only one party capable of benefiting from what seems to be the decay in the connection between Labour and its core vote here, and that's Plaid.

    My aunt's voting for them, for instance, (she told me so) - and her father was head of a Labour town council for decades, for heaven's sake. I doubt Labour will actually lose a town like Llanelli - its Labour MP is pretty good, actually - but there's a definite and powerful groundswell of dissatisfaction that could translate into a protest "stay-at-home" factor that can only help Plaid in a first past the post situation, rather than the PR of the Euros.

    Key point for Labour, though: there's an interesting thing about traditions, especially in politics - they don't really exist!

  3. Nice catch.

    So which seats might this actually flip? The obvious ones are Aberconwy (to Plaid gain) and Brecon & Radnor (to LibDem hold). Are there any others that look suspect?

  4. Plaid can only gain from the demise of Labour and the Liberal disarray, particularly as the Tories will form the government and the great majority of Welsh voters are traditionally anti-Tory. It is a real opportunity for Plaid to seize the initiative.

  5. You might need to check this. Looking at the tables today the weightings appear to be adjusted to Wales. Happy to be proved wrong as I only scanned through the tables but suggest you look. I noticed comments on the Political Betting site about the same issue which i think concluded they were Welsh weighted.

  6. Okay I've gone and checked UK Polling Report that has the full tables. This was discussed on that site and Anthony Wells confirms they were Welsh weightings

  7. The comment that "The figures have been weighted and are representative of all GB adults (aged 18+)" was in the press release from YouGov that accompanied the poll. Do you believe the polling company's own press release or a comment on a blog?

  8. Hi Alwyn,

    Anthony Wells works for You Gov. He states categorically that the ABC1 in use for UK polls are 55% of the sample. The Welsh poll was 45% ABC1.

    Never trust a press release!!


  9. PLaid should get 5 seats...no more than 7 and that would be breakthrough to take Aberconwy and Llanelli.....buts its not impossible especially with election scandal.