Things always look better after a bit of a polish

When you pay the cherrypicker and get to pick out the cherries you like the look of, you tend to get what you want.

This is true of this week’s Alan Bown-funded poll for UKIP, carried out by Survation in the Thanet South constituency.  Political Betting makes the polling its main story with a headline that UKIP is just 5 points off winning Thanet South, pushing figures that have UKIP polling at 30%, ahead of the Tories on 28% and behind Labour on 35% as shown in their graph below, if a General Election was being held tomorrow.

The problem is this selective snapshot is only made possible by excluding from the model those people who replied ‘Don’t know’ or refused to express a party preference.  That is 35% of the respondants, and makes the cherrypicked data virtually meaningless.

If you look at the poll numbers themselves rather than this top level take that Guido and others have seized upon without looking at the actual data, and include those who are yet to make up their mind and those who may well vote but refuse to tell the pollster which party they will support, we find somewhat more realistic figures (rounded from page 3 of the data).

Labour 23%
UKIP 19%
Conservative 18%
Lib Dem 3%
Other party 2%
Undecided 24%
Refused to say 11%

Make no mistake, 19% for UKIP, given their diminishing focus on the core issue of sovereignty, isn’t that bad.  As this is Kent, we are talking about a county that has suffered more than most from illegal migration, via France, of people from around the world.  In these circumstances, UKIP should be out of sight – not trailing behind the very Labour party that more than any other political entity threw open our doors to migrants regardless of the legality of their status or value to our society and skills base.

Thanet South is reportedly one of the constituencies Nigel Farage is considering parachuting himself into, in the hope of achieving his burning ambition of becoming an MP.  If he chooses to drop himself into Thanet based on the figures being played up, he should prepare himself to be disappointed.

Given previous polls about which parties voters would never vote for, the likelihood is that a majority of the undecideds will break towards Labour or the Conservatives, reducing UKIP’s overall share in the constituency.  UKIP’s core base across all polls is around 13-14%, and there is much between now and 2015 that could see floating voters drift elsewhere.

There is also a possible ’embarrassment factor’ to consider, which came to prominence in polling models after 1992 when Tory supporters who did not reveal their voting intentions because of the party’s unpopularity skewed polls to make Kinnock’s Labour appear more popular than it was.  The 11% who refused to declare their preference could easily break for the Tories, Lib Dems or even Labour.

When you take all this into account the picture looks rather different.

3 Responses to “Things always look better after a bit of a polish”


  1. 1 Bucko 27/11/2013 at 1:25 pm

    Is it me or do the ‘more realistic’ figures look pretty similar to the graph?

    The undecideds and refused to says make up 35% of the poll. That’s probably a fair estimate of the number of people who will not vote at all, rather than those who will pick their side eventually.

  2. 2 cosmic 27/11/2013 at 10:09 pm

    I’ve always thought that opinion polls were there to sway opinion as much as report it.

    We’ll see what happens on the day.

    For all the fact they are something of a mess, I wish UKIP well.


  1. 1 UKIP’s marginal seat polling – fun with hyperbole and badly spun numbers | Autonomous Mind Trackback on 04/12/2013 at 12:32 am
Comments are currently closed.



Enter your email address below

The Harrogate Agenda Explained

Email AM

Bloggers for an Independent UK

AM on Twitter

Error: Twitter did not respond. Please wait a few minutes and refresh this page.

STOR Scandal

Autonomous Mind Archive