Labour 1% ahead in MORI's world

ConservativeHome has questioned the reliability of MORI polls before.  The news of a 1% Labour lead in a poll (see PoliticalBetting.com) that was completed on 6th September looks very unlikely.

Comments

[Andrew Woodman]

Should be ashamed to publish that. If I was commisioning a poll, I think Mori would be a very long bottom of my list.

[MH]

Seems unlikely, although core Conservatives are definitely becoming much more agitated and less likely stay with Dave.

[Jaz]

This must be a joke....

[CCHQ Spy]

Normally the CH poll grapghics have "Populus for The Times" or "YouGov for The Telegraph". "MORI (for no newspaper)" says it nicely. Who would buy a MORI poll?

[Mick]

MORI do as they are told!...

[Voice from the South West]

Even if this poll is not a rogue, it is time for MORI to change its methodology. You don't get the Lib Dems for example falling 5% in one poll. Even so for Labour support to increase by 4% after everything that’s happened strikes me as ludicrous.

[Editor]

VftSW: The poll was carried out from 3/8 to 6/9 so didn't capture all of Labour's difficulties.

[Voice from the South West]

Its still ludicrous Editor. Labour was still losing support during that time anyway!

[Afleitch]

As mentioned in the other thread, this poll uses older date than the BPIX (who are a bit dubious) and YouGov polls which gave us a 10% and 8% lead and coincided with the last Populus poll that gave us a 4 point lead before the crises really got underway. It's an old poll with old data based on old weighing methods and is quite probably an embarrasment for MORI.

[Ted]

The Observer last Sunday carried part of the poll - the sample that were questioned about Blair/Brown. They omitted the Cameron responses in poll and no-one seemed to want to run with the larger Voting Intentions part - unless I missed it somewhere in the Sunday Times who seem to have been the client?

Interesting that while most media/comentats thought the bigger poll not worth using the Obsever felt secure enough to base an article on one segment. Surely not using the poll to fit a particular agenda?

[Prodicus]

Check Politicalbetting

[Fred Baker]

Is it the result of Howard/Davis and Co, trying to fix another coup?

[UKout]

Oh dear. Suggests that the public are as happy with the real thing as with Blair 2.

[Jonathan M Scott]

MORI has got it wrong, as usual. It is only a matter of time until newspapers stop commissioning MORI polls because as they now never seem to correlate with the other polls (and it's not a simple margin-of-error issue - it's a simple not-knowing-what-they-are doing issue).

And let's hope the BBC stop asking Bob (any relation to Bertie?) Worcester/Wooster to come on their election programme to comment....

[kingbongo]

I don't think MORI polls should appear in the (already statistically challenged) poll of polls - same goes for BPIX.

As with all MORI polls this is an embarrassment to them as a company, and why anybody uses them for research I do not understand - Ming's not been such a bad leader for the Limpdums as to send voters flocking back to lucky Tony.

[TomTom]

Maybe Brown and Blair should stage a duel on Hampstead Heath with Sky Live coverage if this is what it takes to boost their appeal ?

No doubt news that Clare Short is to stand down boosted the MORI results !

[Martin]

Dumbest poll ever. This poll is so dodgy I wonder if something more sinister is going on.

[Justin Hinchcliffe]

"This poll is so dodgy I wonder if something more sinister is going on" - reminds me of a Lib Dem bar chart.

[James Maskell]

What is it with today...the joke that is the party logo and now this poll. That cannot possibly be accurate. It cant be.

Mark this all down to the silly summer season...

[Wallenstein]

So what's strange about it? Labour have been leading for years. Looks like things are just swinging back to normal.

Hardly surprising with Cameron in charge of our party. The man is a walking disaster area. If Davis had been elected we'd be 20 points ahead.

Get over it guys, and get used to it.

Pip! Pip!

[G-MaN Wild]

Ridiculas, and obviously very unaccurate. How can they even publish such a poll publicy? If I was Mori Man In Charge, id be saying, "hang on a minute, lets try this one agian".

[G-MaN Wild]

I like this comment on PoliticalBetting.com -

"I wouldn’t be surprised if it came up on ConsHome followed by a stream of “Look what Evil Dave is doing to us” comments.
by Julian H September 14th, 2006 at 4:17 pm"

HaHa True.

[verulamgal]

Shock horror... Boy Wonder has feet of clay after all!Why vote for the monkey when you can have the organgrinder?

[Editor]

G-MaN Wild: HaHa Untrue. The clear majority of comments are suspicious of MORI's findings.

[Wallenstein]

I detect something stronger than a whiff of panic in the air.

Very probably at this precise stage in the game this will prove to be a rogue poll, but the comments above say it all.

These jittery Cameron clones know as well as I do that our recent poll leads are based on very shaky foundations. They can yap as loud as they like but the fact won't alter.

In truth our recent "success" is totally determined by the ups and downs of the Labour Government. Had we elected David Davis we would now be streets ahead, and with real policies. Sadly we chose not to.

I see in the paper this morning that Gordon Brown is having a family-friendly makeover. Very possibly Labour has learnt its lesson from the recent debacle.

And that could be very bad news for Mr Cameron and his acolytes.

Pip!Pip!

[Peter]

We should not forget Bromley. There are tens of thousands of Conservatives like me who may well stay at home because the Cameron Tory party is to the left of Blair on nearly every major issue. If a Tory general election failure means getting rid of Cameron/Maude/May etc then it would be a blessing in disguise.

[Sean Fear]

I can't see any reason to rubbish this poll. I think that we are likely to be ahead of Labour, but it's within the accepted margin of error.

[Sir Robert Worcester]

We do not regard the findings as "sensational". Only those solely interested in the horse race would describe it as such, particularly those who do not understand differences in fieldwork, differences in timing, differences in weighting and the difference, above all, in statistical reliability.

The best guideline, I would suggest, should be "watch the share, not the gap".

Populus' poll was conducted over the same period and published in the Times on 6th September. It was conducted by telephone interviews with 1,504 respondents on 1-3 September. Ipsos MORI's figures, published on 10th September in the Sunday Times (Labour figure only), was among 1,886 respondents of whom the voting intention is based on the 57% who said they were absolutely certain to vote. Our poll was conducted face-to-face between 31 August and 6 September. Virtually all fieldwork was completed by 5th Spetember, i.e., before the MPs’ letter to Tony Blair was published and the latest round of bloodletting, and may well now have been overtaken by events.

The comparison between the two polls is as valid as can be expected given differences between telephone and face-to-fact methodology and a shorter fieldwork period (3 days) vs longer (7 days) one. The Conservatives and Liberal Democrat shares are each one point different, 'others' 12% in Populus poll and 10% in ours (well within statistical reliability tolerance levels), and Labour in Populus 32% and the Ipsos MORI figure as published in the Sunday Times 36%. The previous Populus poll had a two-point Tory lead. The previous Ipsos MORI had a four point Tory lead.

Neither of the current polls show significant change in the lead from July nor are they significantly different from each other.

Sensational? No. Volatile? Yes, compared to the July findings. But as a wise man once said, "A week is a long time in politics".

I seem to recall some personal abuse coming from the Conservative camp in the past, and would just remind those who made such comments before the last election about our methodology:

MORI 5 May for the Evening Standard:
Conservative 33%, Labour 38% LibDem 23%
Election Result
Conservative 33%, Labour 36%, LibDem 23%

And the MORI/NOP exit poll for the BBC:
Labour 66 seats
Election Result
Labour 66 seats

[Peter]

Sir Robert's points are well made.

I think Lord Tebbit's point needs to be taken on board that the Labour and Conservative parties are so keen to be like each other in the middle, that they have disenfranchised the left and the right. Millions didn't vote last time for either party, and in 1997, millions of Conservatives voted for the Referendum party or UKIP.
Why Cameron thinks Conservatives on the right( like me) will vote for him is beyond me, especially as he has already ratted on his leadership campaign promise to withdraw the Conservative party immediately from the European federalist group they are in now, and is now pursuing a high tax liberal/ green agenda. I would rather vote for UKIP or an Independent.

[Yet Another Anon]

Either you have a healthy scepticism about the usefulness of opinion polls on voting intention or you accept what they say based on their methodology.

The pollsters can't say that because a particular figure looks a long way out that they will not publish it, it might lead them to look at their methodology but empirical methodology involves accepting that research is there to use a method to produce results for analysis, the conclusion might be that the survey itself was flawed based on an analysis of the method against other data but to say that it is wrong because it shows Labour's vote higher than expected and the Conservatives slightly lower is to prejudge what is rather a slapdash activity in the first place with what people say subject to passing fads among the general population and perhaps even more so in social studies than experiments in the physical sciences the person conducting the survey affects the results of the research\experiment - not neccessarily deliberately but even the fact that they are there might result in someone saying something other than they actually intended to do, people felt uncomfortable in the 1980's and 1990's and even into the last parliament about being perceived as Conservative voters so frequently either they were non-committal or said they would vote for another party, this may well shift around to affect Labour.

[Yet Another Anon]

Equally it could be accurate and be suggestive of a return of many who had gone over to the Liberal Democrats from Labour starting to switch back perhaps seeing little difference between the Liberal Democrats and Conservatives and seeing a recovering Conservative Party resulting in a firming up of the vote and a more 2 party system, after all many of those who switched to the Liberal Democrats did so because actually from 1994 to 2005 the Liberal Democrats were of the 3 main parties the most strongly opposed to privatisation and favourable towards an increase in state involvement in the economy and this has changed in the past few years, it is possible that it is a reflection of this. None of the 3 main parties have been doing all that fantasically since 1997, in fact the Liberal Democrats successes since 1997 have more as a result of what a bad state the 2 main parties have been in rather than anything on their own part.

[Monday Clubber]

Sir Bob Worcester is the doyen of pollsters and a very charming and deeply impressive personality.

Frankly I think it's utterly pathetic that one or two desperate individuals seek to rubbish his highly-respected organisation for narrow party political motives.

[Ted]

It's good that Sir Robert has posted a defense of the poll and yes it can be defended as within sampling error margins. PoliticalBetting points out the weighting methods used in the two polls he finishes his posting with are different from those IpsosMori uses in the monthly political monitor. Mori produced in the month leading up to the election two startling polls - one on 3rd April showing a 5% Conservative lead and another 4 days before the election on 1st May a 10% Labour lead.
It may well be that this reflects volatility among the voters or alternatively that with more polls prior to election the chances of outliers rises. However if a different weighting was applied on last day, as claimed, why doesn't Mori adopt that as it seemed to produce a more representative result?

[comstock]

Should be ashamed to publish that. If I was commisioning a poll, I think Mori would be a very long bottom of my list.

Don't shoot the messenger, folks. Blair has finally annouced he's going, and a leadership change should give Labour a further boost. People being sick and tired of Blair does not equal 'sudden desire for the Tories back', as I predict we will see over the next couple of years.

(wanders off whilsting 'The only way is up')