« The Governator can't make it | Main | Davis tops shadow cabinet league table again with Warsi at bottom »


Why can 't anybody read the polls right. The Sun says that Labour REMAIN on 41% with the Tories up 1% at 36%and the Labour lead staying the same. LibDems -1%. at 16%. Others 7%

Given those figures and - unless the blog is comparing them to a different previous survey than the Sun - the Others cannot have fallen.

Have I missed something?

The polls are everywhere- I don't think we can trust any until after the conference season (if we get one!).

By then the public will have had a proper chance to compare each leader with the other. With some clever press management the EU treaty could overshadow the Labour conference.

It is positive however to see upward movement on our side and shows we're still in the game.

christina, presumably The Sun was referring to the much talked about YouGov poll rather than the last MORI poll.

Interesting. At least we're on the way up, even if voters haven't had chance to see through Brown's sham yet.

The only way to look at this poll and others, Labour are holding their lead despite everything.

These figures are correct as the changes are based on the mori poll for the Sun published on the 9th August. Which were Lab 38% Con 33% LD 15%.

The Polls are getting a bit difficult to read we have you gov showing an 8 point Labour lead ICM and Mori a smaller 5 point lead whilst a private poll from poplus has labour's lead down to a single percentage point. I know everyone loves to think that every individual poll makes an early general election more or less likely but making sense of this months figures is no easy task.

The big thing we have to fear is that Brown hasn't used his clunking fist during August. He'll be back much sooner than Arnie and that's our time of maximum vulnerability.

Sorry Mr Deputy Editor but you add to the confusion.

Where please did you get YOUR + and - figures from? These are not the same as the Sun's figures and presumably theirs - as the paper commissioning the poll - were direct from the horses mouth, Ipsos MORI. They say their comparisons were with "last month" which obviously is the comparable poll to use.

The 7% for others is a ludicrous figure since consistently the Nationalists alone are chipping in ALL polls at around 3-4% which doesn't vary a great deal

Are you saying that the Sun is misreporting the poll or what? Sorry to be a a hassle

If crime and immigration are at the top of voters wish lists then why does the poll not show a narrowing of the Brown lead?


This is the poll I think the Dep Ed has used. Published by the Sun two weeks ago

Mori's Last Poll

Though I remain to be convinced about a GE just now (and will doubtless be proved instantly wrong), those wanting to wait until 'after the conference season' to see how things pan out may just find that we have had a rather larger sampling in the form of a GE in the meantime.

The private poll published yesterday was the most interesting since I have always assumed that the parties themselves would prefer to hear an accurate truth rather than inaccurate spin (otherwise why not just spend the money on a big punt on the 2.30 at Newmarket)and that that may explain a more confident performance by Mr. David Cameron this week, if he and others think there is all still to play for.

John, Christina,

That link should be :

MORI's last opinion poll

and that that may explain a more confident performance by Mr. David Cameron this week, if he and others think there is all still to play for.
David Cameron isn't likely to give up, after all if Labour wins the General Election he probably will want to continue and fight another election later on and so the numbers voting Conservative as compared with the previous few General Elections is crucial, if he is unable to take the Conservative Party into government then he needs something to show to indicate some kind of progress or his options otherwise would be to face either having to resign after the General Election or getting booted out.

If his strategy is discredited then the party will choose someone to replace him who will follow a different strategy and he doesn't want that.

Why is the private polling showing different findings to the other polls? They cannot both be right. Either we are in for a competitive election or there will be another Labour landslide.

If crime and immigration are at the top of voters wish lists then why does the poll not show a narrowing of the Brown lead?

Posted by: Cleo | September 01, 2007 at 12:46

Maybe people are more inclined to trust Broon ("granite-like man of principle" - yeah bollocks I know) with those issues than the flippety gibbet "Dave" who wanders from issue to issue in search of the odd cheap vote?

Thanks NaE for the link. That particular poll had the "Others" on 14% but it also had the "dustbin" area of Would not vote (10%) Undecided (6%) and Refused to say (3%) or a total of 19% of the WHOLE SAMPLE. If AlL the figures were given the Tories fell back to 29%, the Labour went up to 42% and Libdems were on 16% with the "others" on 13%

But none of these gel with what The Sun says . The trouble is that the journalists are so hankering after a headline that they are either too in-numerate to read the figures or too lazy.

As for the party's private poll. The organisation that disclosed it - UK Polling Report - is affiliated to YouGov who wanted to examine the discrepancy between the Populus and the YouGov polls. They are supportive of both and say why - convincingly. I circulated it yesterday

The Sun appears to have comprehensively loused up a perfectly adequate report

Why is the private polling showing different findings to the other polls?

What was the question ? Cleo

Can we believe ANY of these polls, the "dustbin" area of those who will not vote of 10% bears no relation to the last GE where the figure was close on 40%. I do not think that nearly 30% of the electorate are suddenly going to go to the polls.
Or is it me!

Ipsos MORI only includes those who say they will vote at scale of 10. It doesn't weigh its sample by past votes. What the Ipsos Mori poll produces is a snapshot based on a random sample of electors with headline figures based on certain to vote. This approach tended to overplay Labour in the past which is why most other polling companies try to correct the random sample by an additional weighting to reflect peoples votes in last election. YouGov which bases its sampling on the declared party preferences and voting intentions has a different weighting method.

Thats why (as Graeme Archer keeps pointing out) you need to compare like with like, Mori with Mori, ICM with ICM, YouGov with YouGov, Populus with Populus.

So we get the picture compared with previous polls
Latest 36:41:16 previous 9th August 33:38:15 so improved Tory, improved Labour, slightly higher LDs (others down)
Latest 36:37:16, previous 29th July 33:39:15 so improved Tory, lower Labour, slightly lower LDs
Latest 33:41:14 previous 10th August 32:42:14 improved Tory,slightly lower Labour, no change LD
Latest 34:39:18 previous 10th August 33:39:18 slightly improved Tory, Labour & LDs unchanged.

As Christina says a mixed picture. Brown bounce seems to be at or just over its peak, Tories recovering a bit, Lib Dems in the doldrums.

There is evidence in regional breakdowns that the Tories are doing very well in the South, not badly in Midlands & Wales and overall share hit by very poor share in Scotland and fall in support in the North. So in election we could expect to lose very few seats and pick up many in South/Midlands & a few in Wales. Labour seem to be doing very much better in their heartlands and racking up support where they already have seats so would despite their improved showing IMO suffer overall losses, retaining their safe seats with improved majorities, regaining share from LDs. LDs are the conundrum as they haven't had this many seats before which might make their usual tenacity in holding onto gains less than it was and as 2005 showed they can be beaten in the South - they are looking very weak in Scotland so may suffer there as Scotland is currently polling as an SNP/Labour two party country. In Wales they may also have been weakened as they are very much the fourth party now.

Picture looks to me that an October election would show Conservative gains at expense of LDs and southern Labour seats but maybe not sufficiently to take away the Labour majority, though a hung parliament is a strong possibility.

A lot of people think they should vote or will vote, but when it comes to the time will probably forget, or might not have arranged a postal vote and might decide it was impractical then to turn out and vote, or they might on the day just be too lazy to vote.

I think though that Turnout could well be around 70% with both many past Labour voters and Conservative voters who have stopped voting at various stages since 1992 turning out. Most of the Liberal Democrat percentage gain in vote since 1997 has been down to a fall in turnout. Their 2005 total vote was only about the same as that of 1992.

Interesting thoughts .I've learned a lot about polling reading this thread.

We're nearer resolving most queries except the reporting
of the Sun's poll today. Thanks esp: to Ted for laying out the reported headline figures so clearly .

What was, however, said in the Sun that Labour REMAIN on 41% with the Tories up 1% at 36% and the Labour lead staying the same. LibDems -1%. at 16%. Others 7%, is not compatible with the figures. The blog's account IS compatible so : IT'S THE SUN WOT BLEW IT" (to nick one their old headlines.)

The idea of 7% (a fall of -7% ) for others is ridiculous since the SNP and PC would account for almost all of the remainder and the SNP is still riding high leading the field in Scotland.

That's a very good summary, Ted.

Broadly speaking, because of boundary changes, the impact of first time incumbency, and shifts in regional voting patterns, I'd guess that Labour would need a lead of about 5 or 6%% to maintain their current majority (unless it really was 41% Labour to 36% Conservative, in which case Labour would probably pick up enough Lib Dem seats to push up their majority).

If they retained their 2005 lead of 3%, then I think their majority would be 30-40.

Anything under 3%, and their overall majority must be at risk.

Ted - your analysis is spot on - and it is exactly why Brown is not going to rush to an election now. But by the time he does the polls will be a lot worse for him.

Didn't the internal poll from last night have a 4% margin of error?

In that case Lab 37% Con 36% could well be Lab 41% Con 32%, but CCHQ 'let slip' on purpose with this one because it showed an okay result.

Just going by what UK Polling Report hinted.

Edison - CCHQ didn't 'let slip on purpose ' - they were compelled under the terms of British Polling Council's rules.

(As the blog said "UK Polling Report's Anthony Wells has forced the release of the leaked internal Populus poll, under the British Polling Council's disclosure rules)

Everyone is right to point to the utter unreliability of polls to forecast a result until an election has been called. Then people harden their decisions quite sharply.

Also this latest poll misses oner vital point which we shall not get until the detailed tables are available. The "dustbin area" of won't vote / refuse to say or undecided was for MORI 8% in july - rose to 19% in August and has done what now ??? Each variation reduces or increases the "universe" upon which the headline figures are based. To make any sense of these polls =one cannot just look at the headline figures. They are there solely to sell newspapers.

And as Edison says the margin of error is significant - it's somewhat less usually for YouGov who have a sample which is normally twice the size.

Edison, and in that case with the 4% Margin of Error, Lab 37% Con 36% could just as equally be Lab 33% Con 40%!

Christina, you're right: the underlying regional figures, and the changes from month to month, need to be looked at - not just the headline figures.

Out canvassing again today, not detecting vast differences from May,


If CCHQ released all their daily tracking figures it would be significant. Leaking just one daily poll indicates it probably isn't representative of the overall polling they are seeing.

Populus had to release the full details of the poll because their client, the Conservative Party, had given out the headline figures.

So Labour still well ahead regardless of Dave's sudden 'conversion' to the right - no surprise.

So Labour still well ahead regardless of Dave's sudden 'conversion' to the right - no surprise.

Well didn't the 'Roons warn us that this would be the inevitable result of a return to the 'failed core vote strategy'

Maybe they were right after all.

The comments to this entry are closed.



ConHome on Twitter

    follow me on Twitter

    Conservative blogs

    Today's public spending saving

    New on other blogs

    • Receive our daily email
      Enter your details below:

    • Tracker 2
    • Extreme Tracker