A look back at the polls: November

by Stephen Tall on December 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 nine polls published in November:

Tories 43%, Labour 30%, Lib Dems 18% – ICM/Sunday Telegraph (9 Nov)
Tories 41%, Labour 35%, Lib Dems 16% – Populus/Times (10 Nov)
Tories 40%, Labour 37%, Lib Dems 12% – MORI, unpublished (14 Nov)
Tories 41%, Labour 36%, Lib Dems 14% – YouGov/Sunday Telegraph (16 Nov)
Tories 43%, Labour 32%, Lib Dems 12% – ComRes/Independent (16 Nov)
Tories 42%, Labour 31%, Lib Dems 19% – ICM/Sunday Mirror (21 Nov)
Tories 40%, Labour 36%, Lib Dems 14% – YouGov/Telegraph (26 Nov)
Tories 45%, Labour 30%, Lib Dems 18% – ICM/Guardian (29 Nov)
Tories 43%, Labour 32%, Lib Dems 15% – MORI/Observer (30 Nov)

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

Tories 42% (-1%), Labour 33% (+2%), Lib Dems 15% (-1%)

In one sense, the story of November’s polling is easy to state: Labour gains a bit (again), at the expense of both the Lib Dems and Tories. And that may well be the underlying reality. But, as we at Lib Dem Voice have been noting for many months, your view of reality really does depend on which polling company you most trust.

This is certainly the case if you’re a Lib Dem, with the different polling companies’ methods exerting a disproportionate effect on Lib Dem support. Below is the party’s average rating according to the last five polls conducted by the main five polling companies:

ICM – 19%
ComRes – 17%
Populus – 17%
Mori – 14%
YouGov – 14%

Glass half-full Lib Dems will, very understandably, prefer to believe ICM, which consistently tends to show the party doing better than other pollsters (and which has a proven track record of accuracy over many years). Glass half-empty Lib Dems will point to Mori and YouGov’s significantly lower ratings. Many of us will choose to split the difference, and guesstimate that the party is hovering around 17% as suggested by ComRes and Populus.

In the past month, it’s not only been Lib Dem ratings which have fluctuated according to pollster; so, too, have Labour’s. There does seem to be something of an inverse relationship currently between Labour and Lib Dem ratings: in November, Labour’s highest ratings were from YouGov (36%, twice) and Mori (37%) – the two pollsters which show the Lib Dems doing least well; while ICM is pegging Labour support at 30% (twice) and 31% – while showing Lib Dem support within touching distance of 20%.

For the moment, at least, Tory support seems to be reasonably consistent, ranging between a within-the-margin-of-error 40-45%. Interestingly, it’s currently ICM which is proving to be the friendliest to the Tories, while YouGov, sometimes regarded with suspicion owing to its links to top Tory Stephan Shakespeare, is showing them at the lower range.

Enjoy reading this? Please like and share:

No comments

UK Polling Report says that we are running at an average of 16%.

That seems a bit low to me, especially given the other parties’ woes right now.

Why do you think we are not doing better?

Has anyone found our economic narrative?



by Neil Stockley on December 1, 2008 at 4:56 pm. Reply #

[Apparently I posted this in the wrong thread earlier.]

A dramatically different poll by ComRes for the Independent:
CON 37% (-6)
LAB 36% (+4)
LDEM 17% (+5)

Just when it appeared the Tories had regained a double-digit lead after the pre-budget report, this would indicate (on a uniform swing assumption) Labour just 3 seats short of an overall majority.

by Clegg's Candid Fan on December 1, 2008 at 11:17 pm. Reply #

What I was going to say was how funny it was to see all the Tories on politicalbetting.com running round like headless chickens in response to this. It must be wrong, because … [any reason will do]

Anthony Wells on ukpollingreport.co.uk commented sensibly that “We won’t know for sure until we see some other polls …”

by Clegg's Candid Fan on December 1, 2008 at 11:20 pm. Reply #

Stephen is right to expres all of the ambiguity over polling and also the ICM/Non ICM divergence at the moment. However lets just look at the broad trend of the LD voice average over the last year. 16% last December including all of the coverage around the leadership election and new leader. Slowly climbing to 18% over a few months and then steadily downwards till 15% now.

If the glass is half full then the party seems to have a remarkably stable core vote at about the 16% mark plus what ever boost a GE campaign will give us ( which usually happens rather than always…)

If the glass is have empty then the party has gone absolutely no where and arguably a point backwards over the last year. has nothing happened over the last year that an opposition party could have exploited. If we can’t progress in the current climate then when will we ? It also begs the question as to whether there has been any benefit what so ever from the panicy executions of two federal leaders since the last elections.

My sumary would be “reason to panic? No. reason for soul searching ? Yes. “

by David Morton on December 2, 2008 at 12:04 pm. Reply #

Poll by Populus for the Times:
CON 39% (-2)
LAB 35% (-)
LDEM 17% (+1)

Changes are since the last Populus poll about a month ago.

Again on a uniform swing projection this would make Labour the largest party in the Commons. And again this indicates that the Lib Dems have improved by a few points since reaching a nadir a month or two ago.

by Clegg's Candid Fan on December 8, 2008 at 10:50 pm. Reply #

Ipsos-Mori poll for the Mirror:
CON 41% (-2)
LAB 36% (+4)
LDEM 11% (-4)

Changes are since the last Ipsos-Mori poll at the end of November.

After a few polls showing (apparently) a slight recovery in the Lib Dem rating, this is another appalling one. It equals the worst rating recorded by Ipsos-Mori (or anyone else) since the last general election, shortly before Ming Campbell was dumped, in October 2007.

by Clegg's Candid Fan on December 15, 2008 at 12:44 am. Reply #

On a uniform swing projection, that would leave the Lib Dems with just 13 seats, unable to give either Labour or the Conservatives a majority.

by Clegg's Candid Fan on December 15, 2008 at 12:50 am. Reply #

There will be a Guardian ICM early this week almost certainly out tonight in advance of the Tuesday addition. I think ICM is the gold standard when it comes to Lib Dem polling. It tends to give the party higher but more accurate scores than other pollsters.

by David Morton on December 15, 2008 at 2:22 am. Reply #

nice to see you’re not obsessing or being selective!

by Orajepan on December 15, 2008 at 8:35 am. Reply #

“I think ICM is the gold standard when it comes to Lib Dem polling. It tends to give the party higher but more accurate scores than other pollsters.”

Higher, certainly, but I’m not convinced we can know which of the pollsters is evaluating Lib Dem support most accurately.

by Clegg's Candid Fan on December 15, 2008 at 8:54 am. Reply #

“nice to see you’re not obsessing or being selective!”

For your information – in case you’re interested in information, which I doubt – I generally post the results of any polls I see, unless someone else gets there first, as David Morton did on another thread for the last two.

And, incidentally, of those two the ComRes poll was also pretty awful. That company tends to produce favourable ratings for the party, and 14% would have been their lowest for nearly two years – if it hadn’t been for the 12% rating they produced last month.

Now, do you have any comment on the subject of the thread, apart from the usual “Nick Clegg has a secret plan and everything’s going to be all right” guff?

by Clegg's Candid Fan on December 15, 2008 at 9:01 am. Reply #

Mori are habitually the worst pollster for us LibDems because of the screwy weightings they apply to the pure numbers.

There is no need to take their results seriously when they always underestimate the level of our support, so accounting for this underestimation (by at least a couple of points) and adding it to the potential margin for error we can see our poll rating as holding steady.

I really don’t understand why you are so keen to be a merchant of doom.

Can I give you a pinch of salt for Christmas?

by Orajepan on December 15, 2008 at 10:03 am. Reply #


So you really _don’t_ have anything to say apart from “everything’s going to be all right”?

“Mori are habitually the worst pollster for us LibDems because of the screwy weightings they apply to the pure numbers.”

Of course all the pollsters differ, and it doesn’t surprise me that people tend to prefer the ones that are most favourable to the Lib Dems. But as I pointed out, this equals the worst Lib Dem rating by MORI since 2005. Just as the latest ComRes equals the worst Lib Dem rating ever recorded by that company – with the single exception of 12% last month.

According to other companies the Lib Dems aren’t doing quite as badly as that. But these polling figures are not at all good, and I can’t see any point in pretending otherwise.

by Clegg's Candid Fan on December 15, 2008 at 10:16 am. Reply #

I have plenty to say, but when you keep arguing that the glass is half-empty I feel obliged to point out that this is only one perspective.

by Oranjepan on December 15, 2008 at 10:23 am. Reply #

Just out of curiosity, I looked at the rersults of opinion polls in the fortnight before the 2005 general election. The average Lib Dem ratings for each pollster were as follows:

ComRes 20.5
ICM 21
Populus 21
BPIX 21.5
MORI 22.5
YouGov 23.6

Those figures hardly support the ideas that ICM is the most and MORI the least accurate in evaluating Lib Dem support.

by Clegg's Candid Fan on December 15, 2008 at 11:04 am. Reply #

ICM poll for the Guardian:
CON 38% (-7)
LAB 33% (+3)
LDEM 19% (+1)

Chenges are since the last IC poll 2-3 weeks ago.

by Clegg's Candid Fan on December 16, 2008 at 7:26 pm. Reply #

I think ICM is the most accurate in that for a number of reasons it seems to simulate some of the campaign boost that we get out of “Equal Air Time”. My instinct is to say its a better predictor of results where as some of the lower scores are more accurate in real time. I have already seen one blog post pointing out that if you add on our campaign boost of 3% to 19% then we are back at 2005 general election levels and so everything is alright. I think thats a bit foolish.

I note with interest that 19% was the extant ICM rating on the day nick took office. It seems to summarise the evidence of the whole year. No sign of any great collapse but absolutely no progress either.

by David Morton on December 17, 2008 at 1:39 am. Reply #

I think that ICM rating of 19% is just about at the median of all the ICM ratings for the Lib Dems over the last year. Obviously that’s a lot better than being at or near 2-3 year lows as MORI and ComRes would indicate, but it’s not a spectacular improvement either.

And obviously the difference between 11% (MORI) and 19% (ICM) is beyond the theoretical statistical error. I still think all the reweighting and normalisation that goes on is more likely to account for the disagreements between the pollsters than differences in the phrasing of the questions. Who knows which is closest to the truth?

I think all we can sensibly do is look at each result in relation to previous results from the same pollster, but even that gives mixed messages at the moment.

by Clegg's Candid Fan on December 17, 2008 at 10:22 am. Reply #

The polls are coming thick and fast now. Here’s another from MORI:
CON 39%(-2)
LAB 35%(-1)
LDEM 15%(+4)

Changes are since the last MORI poll, released 2 days ago.

by Clegg's Candid Fan on December 17, 2008 at 1:30 pm. Reply #

Tommorows Telegraph

You Gov
Con 42 up 1
Lab 35 n/c
LD 14 Down 1

by David Morton on December 19, 2008 at 9:36 pm. Reply #

Leave your comment


Required. Not published.

If you have one.