Thursday, 16 May 2013

UKIP's 'national breakthrough' fantasy turns to ashes as police rescue Farage from Scottish protesters

Over the last couple of weeks I've maintained that UKIP's "national breakthrough" is a media-led fantasy that collapses when the numbers are crunched.

When the BBC stated after the May 2nd elections that UKIP were on 23%, I blogged that even in that election, they'd do well to reach 20%.

 In the days following the May 2nd election I blogged for the Fabians and came up with a best-case scenario figure of 14% for UKIP.

 Then when the Guardian released ICM figures that showed UKIP on 18% I countered again on Left Foot Forward and argued that UKIP's likely polling numbers would have to work very hard to reach that number.

YouGov then went with 14% and last night numbers released by Ipsos MORI have UKIP back down to 13% - a full 10% below the BBC's bogus 23% number from only two weeks ago - a staggering collapse.

Then today, we have Farage visiting Scotland. It didn't go well.

ITV reported that
UKIP leader Nigel Farage has been locked in an Edinburgh pub for "his own safety" as police escorted him through angry protests which had sprung up as he tried to launch his party's Scottish campaign.
So much for the "national breakthrough", eh?


  1. This is nonsense, I'm afraid. I'm not a UKIP voter but you are comparing different things. A projected national share of the vote in local elections is not the same thing as a YouGov poll, which is not the same thing as a Mori poll. It's like saying Manchester United are getting worse because they scored three goals against Villa but only two against Swansea.

    1. Hi Anonymous.

      Think you need to go and read my actual numbers and how I arrived at them. The BBC's 23% was a PNVS - when I crunched the numbers it came down to 14%. A week later YouGov posted exactly the same number and Ipsos have taken it down to 13%. The point is that the BBC took a number that was a good fit for their narrative rather than something that was in anyway accurate. It's my view that 13%ish is about right now but this would likely soften to, at the very best, 8%ish in a general election. Let's wait and see but there is no way UKIP will secure 23% as a PNVS in a general election.

    2. You're still wildly confused (my name's John for what it's worth, anonymous is the only option without logging in or putting in a URL).

      Ipsos can't "take it down" they are a polling company, their polls are not equivalent to other polling companies, and they are certainly not equivalent to local elections - for starters they are polling for a hypothetical general election.

      Most importantly - you have failed to account for people who would have voted UKIP but didn't have a candidate. So your 18/19% is their total percentage vote, but they only stood in 75% of seats, so it is (let's say 18.5%) divided by three quarters, or just under 25%.

      Knock off a bit on the basis that the areas voting are a bit stronger for UKIP, and they are likely to have had candidates in places where they would get more votes (though the evidence for this latter is quite weak, they're not effective targeters) and you arrive at the 23%.

      It was still a PNVS *for local elections* whereas the other polls are forecasts for General Elections. Different issues, different voting motivations, different levels of turnout.

    3. John,

      It seems you still haven't read the links to my other blog posts. Extrapolating the numbers achieved by UKIP in the places they stood in May 2nd's election over the whole country is one method to create a PNVS, sure. Is it sound? No. Why? Because the places UKIP stood on May 2nd were the places they felt they could secure a decent vote - mainly in the shires in the south and east of England (there was no vote in the cities or Scotland, Wales, N Ireland). If UKIP don't have enough candidates to stand nationally or enough support to make it worth their while to stand a candidate then that too could be factored into a PNVS. I arrived at the numbers Ipsos and YouGov polled for UKIP a week before they published - how? By making some rough calculations that looked at pre-existing voting patterns. I think UKIP will still do ok in the south and east of England in a general election but that's it.

    4. Fine but to get to your 14% you have to assume they would have scored around 6% in the parts of the country not facing election.

      I don't think that holds. Tonight UKIP have won a Labour ward in Rotherham where they had never previously even stood a candidate. In South Shields they got 25% of the vote.

      Where are these places where you think they would get hardly any of the vote - Scotland and Central London aren't actually a huge proportion of the population. Even in Wales they got 13% in 2009, and in outer London they are winning by-elections.

    5. John,

      Sorry, but it's kind of tiresome if you don't actually read how I arrived at that 14% figure. When you do so you'll see I went for some pretty high numbers for UKIP.