A look back at the polls: September

by Stephen Tall on October 1, 2008

We tend not to be too poll-obsessed here at LDV – of course we look at them, as do all other politico-geeks, but viewed in isolation no one poll will tell you very much beyond what you want to read into it. Looked at over a reasonable time-span and, if there are enough polls, you can see some trends.
Here, in chronological order, are the results of the eight polls* published in September:

Tories 44%, Labour 25%, Lib Dems 17% – ComRes/Independent (6th Sept)
Tories 46%, Labour 27%, Lib Dems 16% – YouGov/Sunday Times (14th Sept)
Tories 52%, Labour 24%, Lib Dems 12% – MORI, unpublished (18th Sept)
Tories 39%, Labour 27%, Lib Dems 21% – ComRes/Independent on Sunday (21st Sept)
Tories 44%, Labour 24%, Lib Dems 20% – YouGov/Telegraph (20th Sept)
Tories 41%, Labour 32%, Lib Dems 18% – YouGov/Sun (27th Sept)
Tories 44%, Labour 29%, Lib Dems 19% – ICM/Guardian (19th Aug)
Tories 41%, Labour 29%, Lib Dems 18% – ComRes/Independent (30th Sept)

Which gives us an average rating for the parties in September as follows, compared with August’s averages:

Tories 44% (-2%), Labour 27% (+1%), Lib Dems 17% (n/c)

As regular LDV poll-watch readers will know, I’ve been using the above standard introductory paragraph for months now: rarely has it seemed to apply more than in a month where all three party conferences are taking place, and when the financial markets are currently in crisis.

Just 10 days ago, most of the press was splashing with the MORI poll showing the Tories topping 50% of the vote, with a massive 28% lead over Labour, and the Lib Dems down at a pretty dismal 12%. Cue much breathless excitement among not only the Tory blogosphere, but among much of the media, whose then Labour-are-collapsing narrative the poll suited.

But then the Lib Dem conference happened, and (somewhat to most folks’ surprise, including my own) the party got a boost, with both YouGov and ComRes showing us polling at 20% or above. And this was followed by the Labour party conference, which also boosted Gordon Brown’s party, with two polls showing Labour above 30% for the first time since the local elections in May.

Both conferences coincided with the economic turmoil, which has itself now produced a new media narrative backed up by some selective polling: that Mr Brown is the most trusted leader to weather the financial storms ahead.

Though our LDV ‘poll of polls’ shows little movement between August and September, the averages conceal a little of what’s happened. Between 2nd May and mid-September, in the 36 polls conducted the Tories averaged 45%, Labour 26% and the Lib Dems 17%. In the five polls conducted since mid-September, the average is Tories on 41% (-4%), Labour on 29% (+3%) and the Lib Dems at 19% (+2%).

Whether this Tory downturn is a blip which will be reversed by the additional media coverage their party conference will attract; or whether it is a correction to the unduly high Tory summer ratings; or whether this does demonstrate that David Cameron and George Osborne are not yet trusted to manage the British economy in the tough times – well, only time will tell us that. And certainly not the next individual opinion poll, however much it cheers up the party activists whose party is shown to be on the up.

* LDV excludes all BPIX polls, as the company has failed to register itself with the British Polling Council, and refuses to publish any of the questions or data on which its headline polling figures are based.

Enjoy reading this? Please like and share:

No comments

I agree strongly with your last sentence.

And we won’t know the impact (if any) of the conferences until the round of surveys immediately after the Conservative Conference.

I expect that Labour will have the best conference bounce — but then, as uk polling report has shown, they quite often do.

by Neil Stockley on October 1, 2008 at 10:11 am. Reply #

Why does LDV always claim not be poll-obsessed? It doesn’t really ring true. All of us watch every last opinion poll with great interest, surely.

by Mark Littlewood on October 1, 2008 at 1:13 pm. Reply #

In my experience as a Lib Dem of 25 years, Lib Dems always get a boost from conference, as the coverage simply reminds some people we are here. Labour and Tories get far more coverage anyway, so their conference bounce is smaller than ours. However, we usually get a bounce from a General Election campaign, when the media has to give us more attention because of the requirement for balance. So the post conference 20-21% is probably closer to our true level if a General Election was called now.

by Terry Gilbert on October 1, 2008 at 6:14 pm. Reply #

Why does LDV always claim not be poll-obsessed? It doesn’t really ring true. All of us watch every last opinion poll with great interest, surely.

Erm, Mark – did you read the first sentence? “of course we look at [the polls], as do all other politico-geeks, but viewed in isolation no one poll will tell you very much beyond what you want to read into it.”

The PoliticalBetting-style obsession with what each and every poll might signify is plain silly.

by Stephen Tall on October 1, 2008 at 6:32 pm. Reply #

Interesting to see that the first poll conducted after Cameron’s speech – ICM for the Guardian, shows a negligible boost for the Tories since last week:

CON 42%(+1)
LAB 30%(-2)
LDEM 17%(-1)

by Clegg's Candid Friend on October 3, 2008 at 3:41 pm. Reply #

It’s interesting that Cameron’s post-conference bounce (+1) is much smaller than for either us (+4) or Labour (+7) and produces a significant net cut in the Conservative lead.

Is this down to how the conferences went, how coverage mixed with outside events, or how successful the parties were in responding to ongoing events?

by Oranjepan on October 3, 2008 at 4:13 pm. Reply #

Well, it is only one poll. I presume there will be one or two more soon (a YouGov one was rumoured to be on its way last night, but hasn’t materialised yet).

by Clegg's Candid Friend on October 3, 2008 at 4:21 pm. Reply #

Um, ICM is usually very positive for us this actually looks like one of the lowest they’ve given us for a long while.

I wonder if the differential weightings are being manipulated?

by Oranjepan on October 3, 2008 at 4:36 pm. Reply #

“Um, ICM is usually very positive for us this actually looks like one of the lowest they’ve given us for a long while.”

Yes – apart from from one at the beginning of August, it’s the lowest ICM rating for the Lib Dems since November 2007.

by Clegg's Candid Friend on October 3, 2008 at 4:51 pm. Reply #

The YouGov poll for the Telegraph has now materialised:

CON 45%(+1)
LAB 31%(+7)
LDEM 15%(-5)

(Changes are since the last YouGov poll for the Telegraph a fortnight ago.)

by Clegg's Candid Friend on October 3, 2008 at 9:18 pm. Reply #

The ICM, particularly the monthly Guardian one is the Gold Standard of Lib Dem polling. It persistantly scores us highly and accurately.

While that was holding up at 19%/20% then I was wasn’t too bothered by our recent anemic numbers from other pollsters.

However if even that is erroding then we are in trouble.

The rolling Poll of Polls over on COn Home has us at 16.0%. 0.4% lower than when Nick took over. We have been hovering over polling negative equity for a while but always seem to rally.

I don’t take polling seriously from the begining of Septemeber till mid October because of the static of conference coverage.

Interesting times.

by Clegg's Ardent Admirer on October 4, 2008 at 1:07 pm. Reply #

CCF & CAA, the sharp recorded fall in ICM’s ‘others’ should make you suspicious that all is not as appears in this ICM poll. The politics is subtle, but it’s there, so don’t ever be fooled by any claims of impartiality.

by Oranjepan on October 4, 2008 at 1:34 pm. Reply #

Oranjepan:
“CCF & CAA, the sharp recorded fall in ICM’s ‘others’ should make you suspicious that all is not as appears in this ICM poll.”

I’m not quite with you there. As reported by the Guardian, the ICM poll has Others at 11% (up 1).

by Clegg's Candid Friend on October 4, 2008 at 1:41 pm. Reply #

Sorry, easy mistake to make when there are two polls being discussed. I meant the YouGov poll saw a large drop in ‘others’ this time round. Jeez, CCF, don’t you allow any margin for error?

Still, the point holds that differential weightings given to party ID’s needs to be highlighted when attempting to accurately compare and contrast any two polls (even from the same pollster).

The statistical implications of YouGov’s practice of retaining voter ID’s (and creating a growing database on their panellists) has not yet been quantified as far as I’ve seen, nor has any discussion been raised about the subsequent purposes to which that growing database of voluntarily surrendered information can be mined.

I suspect that the YouGov internet methodology (which depends on traceability and tracking to provide the extra benefit of reliability) is not sustainable beyond the medium-term without creating additional safeguards covering the collation and use of data, and that the related issue of weightings may already be starting to distort their sums on the periphery as a result.

by Oranjepan on October 4, 2008 at 2:15 pm. Reply #

Oranjepan:
“Jeez, CCF, don’t you allow any margin for error?”

Shame on me if I did, engaged as we are in the sacred task of proffering advice to the Dear Leader.

Incidentally, another poll by ICM for the News of the World has just been published. As it was conducted in Labour/Conservative marginals, it’s of limited interest to Lib Dems, though it was interesting to see the verdict on the Dear Leader’s new tax cutting/spending cutting policy – 51% for, 39% against:
http://www.newsoftheworld.co.uk/news/article38758.ece

The other thing that struck me is how different – how very different – the links on the right-hand side offered by the News of the World are from the ones on this site. I mean, given a choice between “Tory Chris Grayling misleads the public” and “Sarah, 24, suffers from Permanent Sexual Arousal Syndrome”, how many people can honestly say they would find the former more interesting?

by Clegg's Candid Friend on October 5, 2008 at 12:34 am. Reply #

Must be THAT Sarah!

by David Allen on October 5, 2008 at 12:36 am. Reply #

Populus poll for tomorrow’s Times:
CON 45%(+2)
LAB 30%(+3)
LDEM 15%(-3)

The changes are since last month, before the party conferences.

This is the lowest rating for the Lib Dems from Populus since October 2007.

by Clegg's Candid Friend on October 7, 2008 at 11:51 pm. Reply #

Stephen makes a fair point about “not being obsessed” about polls. Although I tend to take this intro with about as much seriousness as a politician saying “I will be brief”.

Obviously, sites such as politicalbetting.com are basically definitionally obsessed with polls.

But, I suspect most readers of this site are too.

Also, my guess is that if the latest Populus poll had put us at, say, 24%, there would be a LibDem Voice post saying “While we are not obsessed with polls, we thouhgt it was worth highlighting that the latest one puts us at 24%”. Whereas if the poll puts at 15%, there is no such article.

This isn’t wrong or immoral, but it does underline a curious editorial dilemma that makes it difficult to judge the extent to which LDV is independent and the extent to which its part of the party’s spin operation. Both are totally laudible, by the way, but its useful to know which is which.

by Mark Littlewood on October 8, 2008 at 4:12 am. Reply #

A YouGov poll, for tomorrow’s Sunday Times, shows a further rise for Labour, and further falls for the Lib Dems and the Conservatives:
CON 43%(-2)
LAB 33%(+2)
LDEM 14%(-1)

Changes are since the last YouGov poll, last week.

by Clegg's Candid Friend on October 11, 2008 at 10:40 pm. Reply #

Leave your comment

Required.

Required. Not published.

If you have one.