Who’ll benefit from Robert Kilroy-Silk’s UKIP split?
October 28th, 2004
Given the impact that the rise of UKIP has had on UK politics in the past six months any move or split within the party has to be taken seriously. With both Labour and the Tories polling in the 30s small shifts in support could have a huge impact on the General Election.
-
Yesterday’s ICM poll recorded support for UKIP up 1% at 4%. With its most public figure calling his party “barmy” then any reduction in the national support could benefit the Tories and also the Lib Dems which suffered from the UKIP surge in the Euro elections in the West Country.
There’s also the issue of Kilroy-Silk’s plan to stand for a Westminster seat at the General Election. Will he now be able to do this under the UKIP banner? If he can’t it will reduce drastically the media coverage that the party will command.
At the 2001 General Election UKIP recorded 1.5%. It is hard to see that rising much and this will benefit the Tories who before the UKIP May surge had started to build a recovery and were even talking of taking 42% of the vote in the Euro Elections . It will also help the Lib Dems to retain and perhaps add to their seat tally in Devon and Cornwall.
From the betting perspective it might be worth a UKIP sell on the spread markets. IG have them at 1-1.8 seats and it’s hard to see that they could make even one gain without Kilroy-Silk. SportingIndex have suspended their price on UKIP seats. If they come back on at one or more then it will look like a good bet.
-
The IG spread of 202-210 seats on the Tories starts to look good value for a buy.
This is reinforced by a new ICM feature. They now ask “Which of the folllowing best describes you - “I always vote in General Elections”; “I almost always vote…”; “I sometimes vote” etc. The “always” figure from Tory voters was 64% against 55% for Labour. To the “likely to vote on scale of 1 to 10″ question the Tories had the same proportion with the top score - 64% but Labour rose to 60%. The difference might be one of the explanations why polls mostly overstate Labour.
Full round-up of political markets.
MessageSpace Advertising
This site is becoming progressively more biased. Is there any political development that is not seen to favour the Lib Dems. Surely all the evidence is that UKIP damage Conservatives almost to the exclusion of anything else (Hartlepool, Europe etc). Why then should their implosion benefit the Lib Dems? Surely it will unite the Conservative/Right vote and damage the Lib Dems. Surely Ukip being active would imperil the Tories and split their vote - to the aid of the LibDems
I think the comment above about bias is a little unfair - see Mike\’s recent call not to buy Lib Dems on the spreadbet any more as they are high enough.
However, Mike\’s argument that any UKIP slump could help the Lib Dems in the Westcountry needs a bit more explaining - I cannot really see it. Sure, some people who vote UKIP would otherwise vote Lib Dem, but on balance I would guess Lib Dem candidates in Lib Dem/Tory marginals would welcome a well funded, strong UKIP campaign in their area. I can just about see how the Lib Dems could conceivably benefit from a UKIP slump in the handful of Labour/Lib Dem marginals or how they could benefit if UKIP were generating negative headlines for being too right wing rather than merely a shambles (reinforcing the message that UKIP are a party for right wing Tories, not those towards the centre who dislike Brussels). But a generalised UKIP slump can only help the Tories and, on balance, harm the Lib Dems can\’t it? Or am I missing a crucial point?
I think we had this debate before. Devon and Cornwall LibDems pretty much all believe a strong UKIP challenge will help them retain and even gain at the next GE. Of course there are some of the more idealogical brethren hate UKIP so much and don\’t want to see them do well… but not many.
As I have said on this site before, the only clear comparison you can make is 2004 euros vs 1999 euros, where in the SW region UKIP were up substantially, with the Tories down substantially Labour small down and LibDems actually UP. (vote share of course). I can\’t see how you can argue anything other than an increasing UKIP GE vote share would not therefore hurt the Tories on the basis of this evidence.
I think the argument was that parts of the West Country are anti-establishment in general, and anti-EU in particular. In the past this has manifested itself by voting Liberal, then Lib Dem (although the LDs are pro-EU), but that if an anti-EU party emerged they would pull away anti-EU, anti-establishment protest votes from the LDs to UKIP. Hence this is not pro-LD bias but an assessment (up for debate) of how UKIP will affect voting intentions in a distinct region.
Mike - as far as I know it isn\’t a new ICM development, I think they\’ve been doing it for ages. The difference is that they have suddenly started including the question in the information they publish. It is, perhaps, also worth pointing out that ICM do use this second question when weighting their data. Responses are turnout weighted based upon both their answer to the 1-10 question and also their answer to the how regularly have you voted in the past question (I do not know exactly what formula is used to come up with the weighting figures).
One explanation for why polls have a tendency to overstate Labour is, as we were discussing the other day, \”Refusals\”. Demographically refusals tend to be disproportionately elderly women, and disproportionately people you would expect on demographic trends to be Tories. ICM and Populus include 50%(ish) of those who refuse to state a voting intention, but say how they voted in 2001. These people are likely another facet of the \”shy Tory\”/\”Bashful Blairite\” phenomenon. On the other hand people who won\’t say how they vote on principle don\’t give voting intentions for either election and are therefore totally excluded from all the polls. These \”double refusals\” made up about 6% of people in ICM\’s last poll.
Sadly (for me at least) that isn\’t a magic extra 6% of Tory voters who win us the election, but the chances are those people are disproportionately Conservative and those same people probably go a slight way to explaining why polls have tended to err more often in favour of Labour than against.
I think I was one of the first people to put forward the view that a strong UKIP vote in Devon & Cornwall was harmful to the Lib Dems, despite the fact that in most parts of the country it would be harmful to the Tories, for the same reasons that Steve puts forward.
The point is that the best UKIP performances in those counties (apart frm Plymouth)came in the districts that returned Lib Dem, rather than Conservative, MPs in 2001, and the Lib Dems saw a much bigger drop in vote share (compared with 2001) than either Labour or Conservatives.
David - your accusation of Lib Dem bias does not add up. In the London Mayoral election - we correctly said Norris was the challenger - not Hughes. In the three by-elections this year we only called for the LDs in Leicester South which they won. On the Commons spread markets we said buy Lib Dem at 58 seats - and then advised on selling and taking the profit when the spread reached 72-77 seats.
In this latest piece we are saying sell UKIP on the General Election spreads and consider buying the Conservatives.
We said repeatedly to back Labour to win most seats at the election until the price got below 1/4 when we felt there was not any value anymore.
I do have a bet - at an effective 132/1 - on the Lib Dems winning most seats at the General Election. This is my personal thing and I would not advise anybody to follow suit
Continuing from point 6, it\’s also worth noting that, overall, the Lib Dems saw a smaller drop in vote share across the country, comparing 2001 with the Euros, than either the Conservative or Labour parties, so Devon and Cornwall really stand out in that regard.
Isn\’t it important, though, to differentiate European and national elections. Yes, quite a lot of anti-European Lib Dems - particularly in the South West (fishing industry etc) - voted UKIP in June. But nobody except a few UKIP diehards really expects UKIP to pick up anywhere near as many votes as they did in June - at a General Election (after all, they did well in 1999 but poorly in 2001).
But it could make a big difference to seat calculations if they pick up 5-10% in some important marginals. The question is, where will the general election voters come from? In 1999, I believe the Lib Dems came FOURTH in some seats in the Westcountry which they proceeded to win in 2001, indicating that their UKIP defectors came back for the general election. The obvious appeal of UKIP at a European election is simply to say \”no\” to Brussels, but at a general election it is to make a more general right wing statement and tell the Tories to stop being wishy washy on key issues.
Anecdotally, Lib Dem activists/candidates seem to be more than relaxed about UKIP running strong campaigns in the Westcountry as elsewhere. I find it hard to believe that a UKIP slump will benefit anyone but the Tories even in the Westcountry at the general election.
Naturally, I hope that you are right and that I am wrong.
Ant - thanks for your very helpful comment and the contributions you make to the site. Given that ICM must know of the demographic and social-economic make-up of the 6% who won\’t say why don\’t they build in another adjusting factor? It doesn\’t make sense.
All pollsters need to draw a line between measuring public opinion and predicting public opinion. MORI have historically been very strict about it and have religiously pointed out that polls are only a snapshot and aren\’t supposed to predict elections. Hence they exclude don\’t knows on the grounds that we really don\’t know what these people think - what Populus and ICM do are just educated guesses based upon other factors that have, in the past shown themselves to be reasonably good predictors of how people vote.
The predictor used by Populus and ICM of how \”don\’t knows\” and \”refusals\” will vote is how they voted last time round, although even that is a long, long way from being perfect (about 50% of people vote the same way - the reason that ICM only reallocate that proportion). Without that (and obviously \”double refusals\” are without that) we really are moving away from polling people\’s opinions and moving towards educated guesswork. Most pollsters are only willing to move so far along that line.
In 1997 NOP did try and re-allocate don\’t knows and refusals using their opinions on which party was best for the economy, or which party\’s leader would make the best Prime Minister. The correlations between these two questions and how people actually voted in 1997 in the BES survey and the ICM post-election study are pretty much in line with people\’s voting intentions and both questions should, in theory, result in fewer double refusals (I cannot see people having the same principled objection to saying whether they prefer Blair or Howard). In the event though NOP\’s actual results weren\’t as good as ICM\’s.
What is boils down to is, no matter how much data we get on these people\’s other opinions, by definition we can never tell how well it correlates with an actual vote so it boils down to educated guesswork and the polling companies, rightly, don\’t want to step across the line between polling and punditry.
As a resident in Devon, I think Mike has made a valid point on the political environment of the South West.
i fink that bnp should win!!!!!!!!
I like what you have done with your site, and can’t wait to see more.
What about the previous post? I think that’s an important note as well.