Latest from UKIP Daily

The Trouble with Opinion Polls

Pollsters operate nationally, and aim to produce headlines for newspapers, which generally are based on the mood of the whole nation. When they say things like “UKIP only have 12% saying they will vote for them in a General Election, but Liberal Democrats are down on 8%” that is reflecting what they have measured across the whole country. Of course, we in UKIP know those results are skewed, by the way most of the polls do not prompt for UKIP, and the way they allocate the Don’t Knows in the proportions of the last election.

Ignoring that anomaly, how is it that elsewhere in those newspapers they will boldly predict the Liberal Democrats will still get 25-30 seats in the next election, whereas UKIP will get none? By what contortion of mathematics does a 12% party get no seats, and an 8% party 25 seats? The answer is geographic spread, and within that spread there will be “hot spots” like Eastleigh standing alone in otherwise mostly Conservative (for now!) Hampshire.

Of course, they will then dismiss UKIP, saying our party has an equal spread of support across the country.  Nothing could be further from the truth now. Support for UKIP is lower in London and much lower in Scotland – that’s 22% of the country, pumping up the UKIP vote in the rest of the country. Then within that, there are hotspots – the Euro and Council elections this year have highlighted that, with the West Midlands and the East Coast from the Humber to the Thames being the strongest UKIP citadels of support.

So, recently, YouGov published a poll for The Sun where they show the percentage of voters who will stick with the party they voted for in 2010, and if not, who they will vote for. Here’s the matrix I extracted from their numbers

2010 Vote Who they will vote for in next General Election
C Green Lab LD UKIP Don’t Knows
C 65.3% 0.9% 4.4% 1.7% 12.2% 13%
Lab 4.4% 1.7% 72.2% 0.9% 7.0% 13%
LD 10.6% 5.3% 27.4% 23.6% 7.6% 24%

As can be seen, a big shift away from the Liberal Democrats, some movement between the other two, and a very moderate shift to UKIP.  I would contest the accuracy of the poll given other indicators, but for more of that, read on.  When I ran these party loyalty shifts against the 2010 General Election vote counts, the results were surprising:

  • Labour with 330 Seats, Tories with 288
  • Both UKIP and Liberal Democrats with no seats.
  • Strongest UKIP numbers and 2nd places were mostly those in the Tory Shires.

This is the “Vote UKIP, Get Labour” nonsense that has been trumpeted by the Tories and the MSM. However, I have to say that by looking at the numbers nationally, the “election result” is complete garbage. Furthermore, we know from other polls and from actual votes that UKIP’s performance is stronger than that, especially in our hot-spots. Indeed, the Liberal Democrats will retain a few seats in their “hot-spots” too.

And contrast this with the analysis I did against the European Parliament Election results. Real results showed UKIP to be stronger in those areas that Robert Ford and Matthew Goodwin (in their book “Revolt on the Right”) have shown to be prime UKIP territory – those with the most number of “left behind” voters who do not feel LibLabCon has looked after their interests of late. The seats topping that analysis, in the West Midlands, South Essex, Kent and other parts of the country, barely figure in the analysis based on the recent YouGov poll.

Why? Because the poll doesn’t differentiate the shift from Labour to UKIP in the poorer parts of the country! The Labour-voting “state client” metropolitan elite are quite comfortable, thank you very much, in London, Manchester, Birmingham and other provincial centres like Bradford, whether that comfort is based on good public sector jobs or good benefits. But former Labour voters in the areas where UKIP did well in the May 2014 elections have seen that Labour does not represent their interests, and have shifted their loyalties to UKIP.

On top of that, there is UKIP’s “secret weapon”. A lot of people who didn’t vote in 2010, or who haven’t voted for 20 years or so, are now voting UKIP. They have waken up to the fact that their complacency has allowed LibLabCon to rule without check and balance, and not in the best interests of this country and its population.

However, we must neither be over-confident from analyses that predict success in winning a good number of Westminster seats, or depressed by those that predict no seats. We must continue to refine and spread our message of hope: nationally, regionally and locally. – there is a good deal of work to do before the next election!

Print Friendly, PDF & Email
About Brian Otridge (140 Articles)
Brian Otridge was Editor-in-Chief of UKIP Daily for 2 years till March 2016. He has been an aircraft engineer, currently works in the IT Industry and also enjoys writing on a wide variety of subjects. He was the UKIP Parliamentary Candidate for Southend West in Essex in 2015, but now lives in Lancashire

5 Comments on The Trouble with Opinion Polls

  1. There is one aspect of polling data which grabbed my attention just before the EU elections, and then during the Newark campaign.
    It’s not so much the data themselves, nor their bias, but the way these polls are used by hacks in the MSM.
    They are used as secondary propaganda material, in the way of ‘my poll is bigger than yours, so you can’t/won’t win’. Polling data are used not as snapshot of voter intentions, but as reaffirmation of a certain party political stance.
    I believe polling data are important to be used by political parties, e.g. as seeing where more effort might be spent for greater effect. They ought however not be used in the fashion I described above: as propaganda. I am sure this sort of usage has cost us a few fence sitter votes. And as we know – every vote will count next May!

    Isn’t there some sort of convention or even law that forbids the publication of polling data three days before the election? That ought to stop this misuse of polling data.

    • Brian Otridge // June 11, 2014 at 9:29 pm //

      THe only ban on polls is not to publish exit polls until the polling stations have closed.

      For the European elections, that ban existed from 22 May until 11PM on 25 May when the Italian polls closed.

  2. Bernard from Bucks // June 11, 2014 at 5:35 pm //

    There was a very interesting bit on ‘Political Betting’ this morning.
    It turns out that YouGov is fairly neutral in its reporting on UKIP. Survation is providing us with ‘good news’ whilst ICM, ComRes(phone poll) and Populus have it in for us. Very interesting and worth a look. I shall use it when reading future polls and adjust accordingly.

  3. Have a look at this. It shows what UKIP might be up against.
    Yes, the Tories might not be able to reproduce in a General Election the scale of what they managed in Newark but the crucial work is done before an official election campaign starts not during the three or four weeks of the campaign

  4. Great analysis. The latest straw I saw being clutched at was that the lowest share of the vote required to win a seat in 2010 was 29%, while UKIP’s best performance in the by-elections since has been 28%. The conclusion presumably being that this means UKIP will not win any seats next year.

    Of course this ignores the fact that the political landscape has changed beyond recognition since 2010. All bets are off. All previous assumptions should be unassumed. Seats will conceivably be won on smaller percentages. And of course none of the by-elections, to my recollection, have been in really strong UKIP areas, rather they have been against strong Labour/LibDem/Tory majorities.

    That these polls consistently underplay UKIP’s support is not necessarily a bad thing, I prefer the status of underdog and it will help guard against complacency. There will be a few shocked pollsters and pundits this time next year!

Comments are closed.