Archive for the 'Pollsters/polling' Category


Latest wave of Ashcroft seat polling sees UKIP taking Thurrock but losing Rochester

Saturday, April 25th, 2015

Like in all Ashcroft seat polls the names of candidates were not included in the voting questions.

Mike Smithson

2004-2014: The view from OUTSIDE the Westminster bubble


LAB lead up a notch with Populus and not much change in London

Friday, April 24th, 2015

The morning’s polling news

Voters think there’s too much emphasis on Scotland


The three 4% CON lead polls this week cannot all be dismissed as outliers

Friday, April 24th, 2015

But only 1 CON lead from YouGov in a fortnight

The final poll to come out last night, YouGov’s 2% LAB lead, will have eased some nerves amongst the red team. But inevitably they should be worrying about the fact that we have now had 3 surveys in 6 days which have had the Conservatives 4 percent ahead.

At the same time the blue team must be concerned about the lack of progress with YouGov and some of the other online posters. Of the past 14 polls for YouGov there have been 10 LAB leads, three ties and only one showing the blues ahead and that by the narrowest of margins.

That is not easy to dismiss and remember that last time out in national elections, the May 2014 Euros, YouGov was the top pollster with ComRes and Survation trailing quite a way behind.

The two big question marks hanging over this election are whether we are seeing a real CON improvement and how do we work out what national vote shares mean in terms of seats.

Today the pollercoaster continues with the Friday Populus survey and, most likely, another batch of constituency polling from Lord Ashcroft.

My own view is that this election remains too close to call but that LAB still have the electoral system on their side. Even with the Scottish disaster LAB will chalk up more seats than CON with the same vote share.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble


The Thursday pollercoaster continues

Thursday, April 23rd, 2015

LAB increase lead with YouGov

New Survation poll boost for Farage in South Thanet

A second poll today has CON with 4% lead


A Marf cartoon at the start of what’ll be a busy polling night

Thursday, April 23rd, 2015


UPDATE Survation/Mirror poll has CON 4% ahead

First up this afternoon was Panelbase

New ComRes phone poll out at 10pm

Survation for the Mirror is due.


Marf for the evening and David Herdson with the half-time PB Poll Average score

Wednesday, April 22nd, 2015


But it’s still very much in knife-edge territory

The profusion of election-period polls means that it’s now possible to produce a PB poll average figure for shorter periods than before. Consequently, I’ve split the period from April 1 to May 6 into two parts and am using these instead, the first of which is for the scores for up to April 21:

Con 34.5 (n/c)
Lab 33.7 (-0.4)
UKIP 11.4 (-0.8)
LD 8.5 (+0.6)
Grn 5.8 (+0.1)
Oth 6.1 (+0.4)

Good news for the Blue team? Perhaps. Certainly the headline figures are relatively positive with the highest Tory lead since October 2010 and a decline in their two main rivals but CCHQ shouldn’t get too carried away for at least three reasons.

Firstly, a lot rests on that early April ICM poll which put the Tories six points clear. That wasn’t a result wholly out of line with their previous polling but was still some way top-side of what anyone else is reporting, or that they have previously and subsequently found. That doesn’t mean it was wrong but we ought to be sceptical. Having said that, even if we exclude it (which we shouldn’t), the overall Con lead would still be up on March.

Secondly, how much further will UKIP fall – if at all? Their share hasn’t been lower since February 2013 and means they’ve lost almost a third of their votes in half a year and it’s probably more than coincidence that their decline coincides with the recovery in the Con share. In one sense, there’s scope for further reduction: their current rating is still well over treble their 2010 total. In another, it means peeling off voters who’ve been with Farage and Co since midterm. Perhaps the SNP question will help Cameron there but it’s late in the day now.

And thirdly, a 0.8% lead is not much to write home about in the big scheme of things. Remember that the Tories were 7.2% ahead across GB in 2010, so the polls still suggest a net 3.2% swing to Labour overall, and substantially higher in England. Unless that’s inefficiently distributed – and previous Ashcroft polling suggests the opposite – Labour would probably just end up ahead even after losing 35 or so to seats the SNP, but Miliband would have far more supporters in the Commons.

Elsewhere, the Lib Dems return their best score this year but 8.5% is still only just over a third of their 2010 score. Indeed, in terms of vote share per constituency contested, it would be their worst ever at a general election. Incumbency may help to an extent but they do seem to be looking at a sizable cull given that there’s now precious little opportunity for the Yellows to make a mark on the campaign.

With just a fortnight of campaigning left, can anything deliver the country from a deeply hung parliament? Barring a dreadful error from one side or another (including the pollsters), it seems unlikely.

David Herdson


New ComRes battleground polling finds UKIP struggling in its key targets

Wednesday, April 22nd, 2015


This does not bode well for Farage in Thanet South

South Thanet
Boston and Skegness
Forest of Dean
Great Yarmouth
North Thanet
East Worthing and Shoreham
Sittingbourne and Sheppey
South Basildon and East Thurrock
Castle Point

A new ComRes/ITV news battleground seats poll finds the Tories holding on reasonably well in 10 seats which UKIP has made key targets. Like in the similar poll last week of LD defences in the SW the pollster has not named candidates which I think is wrong this close to the election.

The aggregated numbers speak for themselves and suggest that the Tories are set to withstand the UKIP threat in almost all of them. The list includes, of course, Thanet South.

    One key factor is that the LD collapse has bolstered both the CON and LAB votes, and is one of the key reasons for UKIP’s failure to make first or second place. 25% of 2010 LD voters say they will now vote LAB, and 21% say CON with just 8% opting for UKIP.

As you’d expect older voters are much more likely to vote UKIP than younger voters. Only 8% of 18-34s say they will vote UKIP, compared with 25% of voters aged 55+.

This was a phone poll and as we have seen these tend to show smaller shares for UKIP and bigger shares for CON than surveys carried out online.

I think Rob Ford, one of the leading academic who has studied UKIP, makes valid points in these Tweets.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble


New study points to UKIP’s support base being more middle class than was perceived

Wednesday, April 22nd, 2015

farage and carswell

British Electoral Study data sheds different light on the party

An analysis of data from the UK’s longest-running study of electoral behaviour has revealed how the bulk of UKIP’s support surprisingly comes from professional and managerial middle classes.

British Election Study Co-Director Professor Geoffrey Evans and BES Research Fellow Dr Jonathan Mellon, from Nuffield College Oxford, say contrary to the popular view advocated by some academic researchers, working class voters are only a little more likely to support Ukip.

And Ukip voters, they add, resemble European ‘radical right’ parties: an alliance between the working class, the self-employed and employers – and not the disenfranchised, ‘left behind’ voters, described commonly by the media and influential academic commentators.

Professor Evans said: “The idea that many Ukip voters are working class and that they therefore pose a threat to Labour’s support in the election has gained considerable currency ….But we find this is wrong; the working class basis of Ukip has been strongly overstated.The Party’s strongest supporters are often the self-employed and business owners.”

“Even within the working class, Ukippers tend to be low level supervisors, and not the disadvantaged semi and unskilled workers often thought to provide the core of the Party’s support.”

Dr Mellon said: “Ukip’s support is very similar in social composition to many other so called ‘radical right’ parties elsewhere in Europe – an alliance between the working class and the self-employed and employers – rather than a party of the ‘left behind’.

“And significantly in electoral terms, the differences in sizes between social classes means that numerically, the bulk of Ukip’s support comes from the larger professional and managerial middle classes.”

Quite what all this means on May 7th is hard to say. If this analysis is right then LAB would appear to be less vulnerable to UKIP than had been thought.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble