Should Populus and ICM be showing Labour leads?
December 22nd, 2006-
Why is less importance attached to the view of Labour voters?
Detailed data from the December ICM and Populus polls shows that the reported shares followed the biggest scaling back of the views of those who said they voted Labour last time since the last election. If this process had not happened then, in spite of all the recent troubles, Blair-Brown’s party would have probably had poll leads.
For in each of the two polls that we have had in December more than half those who answered the “how did you vote last time” question said they had supported Labour at the last election. With both firms this vote recall proportion is the highest figure since May 2005. Normally they find an average of about 44%.
It will be recalled that the actual Labour share at the last General Election was just over 36%. So what the two leading phone pollsters do is adjust their samples, allowing for a level of misremembering, to bring it more in line with the General Election result and these are the figures that we see published. The December 2006 “adjustments” were the biggest I have ever seen.
-
If Populus and ICM had operated like Ipsos-Mori and Communicate Research, which don’t use such an approach, they would probably have reported significant Labour leads.
Understanding what is going on is central for those who like predicting and betting on election outcomes. I recently found this article from ICM, written after the 2001 General Election, which sets out the firm’s rationale for its methodology.
Pollsters hope that by getting the demographic profile of their samples to match the whole population, polls will give an accurate picture of voting intentions. Trouble is the days when all the toffs voted Tory and the flat capped working classes supported Labour have long gone. A demographically representative poll is no longer necessarily politically balanced. Added to which, response rates are low and falling. Some groups within the population are difficult to interview and some don’t want to reveal their voting intentions. But at present the pollsters simply replace refusers with others who share the same demographic profile, ignoring the possibility that, in doing so, it may be easier to find Labour voters…
You would have thought that most people could remember how they voted in the last election. Yet according to one poll conducted just three weeks after the 2001 election only 26% remembered having voted Conservative (7% too low) while 48% said they had voted Labour (6% too high). On the face of it, such polls simply contain too many Labour voters and too few Tories.
So why not use past voting behaviour to ensure the polls are demographically and politically representative? Some say you can’t trust past votes because some people forget how they voted and others align past votes to present intentions, but it’s the only candidate in town. Of course pollsters have to make some allowance for faulty recall, but the indications are that if the pollsters were to target recall votes closer to the outcome last time they will also get more accurate predictions.
I believe that there is a strong case for the ICM-Populus approach which is why I rate their polls ahead of Ipsos-Mori and Communicate Research. If, as happened this month, half your respondents are saying they voted Labour last time then clearly the sample is not balanced.
Latest prices on which party will win most seats at the next General Election are: CON 0.86/1: LAB 1.18/1.
Mike Smithson
MessageSpace Advertising
Malcolm Chisholm, Minister for Communities in the Lib-Lab coalition government, has been forced to resign after supporting a Scottish National Party motion in Parliament. Three other Labour MSPs also backed the motion, which fell by 72 to 45 votes. The Scottish Parliament has now failed to pass any motion regarding Trident (the Vanguard submarine fleet and nuclear missile stockpile is based at Faslane on the Clyde), with 3 amendments by Labour, the Tories and Lib Dems also failing.
Incidentally, while still an MP Chisholm was the first-ever minister to resign from Tony Blair’s new government in 1997, when he was a junior local government and transport minister at the old Scottish Office. He was the last member of Jack McConnell’s cabinet to have been a member of the Westminster parliament.
The other Labour rebels were Bill Butler (Glasgow Anniesland), Marlyn Glen (North East list) and Elaine Smith (Coatbridge and Chryston).
http://www.theherald.co.uk/politics/77338.html
http://news.scotsman.com/politics.cfm?id=1900072006
http://news.bbc.co.uk/1/hi/scotland/6201737.stm
1. Correction.
The Herald says that Chisholm was “the only member of the Holyrood cabinet to have Westminster experience”. Strictly speaking this is not true: he was actually the only member of the Scottish cabinet to have Westminster ministerial experience. Nicol Stephen,
Deputy First Minister and leader of the Liberal Democrats, served a very short stint of 5 months as an MP in 1991/92, after winning the Kincardine and Deeside by-election, which he gained from the Tories. That brief stint makes Stephen the equal-6th shortest-serving by-election victor of all time, after Bobby Sands, Michael Carr, Oswald O’Brien, Margo MacDonald and Charles Beattie.
Chisholm was the recipient of the Scottish Politician of the Year award in 2002, while serving as Health Minister.
An interesting article Mike. Another danger in the ICM/Populus approach is if there is a change in the paradigm of “Labour voters over stating and Conservatives under stating”.
That is quite possible given that the Conservative brand image is improving whilst Labour’s is in decline. We have to expect voters will be more willing to state that they voted for a positive image than a negative one. “I voted Conservative it is the fault of the others”.
Does anyone know how these adjustments relate to the data from ICM which indicates that 2/3 of Labour’s current voters could change to another party which indicates that Labour has not reached its core vote and could fall further. Is this 2/3 before adjustments or after adjustments?
YouGov in Telegraph little different from last month - Cons 37%, Labour 33% (+1), LDs 17% (+1). “if you had to chose” question gives the Cons under Cameron (45%) v Labour under Brown (32%).
This may be where the Brown bounce will come from.
If these “false Labour recallers” are habitual Labour supporters who for one reason or another — Iraq, civil liberties or whatever — stayed at home last time, then to the extent these issues are held against Blair personally rather than the party generally, these voters can be expected to return under a new leader.
The key question is not who is blamed by Tory or Lib Dem supporters but whether Labour supporters blame the whole government or just the Prime Minister.
Can somebody explain why ministers in the Scotland are sacrificing their careers for something over which the Scottish Parliament has no legislative role anyway? If they’re that bothered about Trident, what are they doing in the Scottish Parliament?
A majority of voters now see Labour as “sleazy and disreputable” after Tony Blair became the first serving Prime Minister to be questioned by the police as part of a criminal investigation, according to a YouGov poll for The Daily Telegraph today.
http://www.telegraph.co.uk/news/main.jhtml?xml=/news/2006/12/22/nloans122.xml
“When Mr Blair came to power in 1997 promising to clean up politics, only 19 per cent thought Labour gave the impression of being disreputable and sleazy, while 73 per cent said it did not. Now 58 per cent say they think Labour is sleazy and disreputable, compared to 26 per cent who say they do not have that impression.”
Labour’s image has clearly suffered with only 26% viewing them as not sleazy, down from 73% in 1997. At these levels their voters will have to take polly’s advice and “hold their noses” when voting next May.
Telegraph has a “just ahead” line on the Brown Vs Cameron in the article I quoted above.
“But Mr Cameron is just ahead of Gordon Brown, with 45 per cent saying they would prefer a Cameron-led government to 32 per cent opting for a Brown one.”
How on earth is being 13 percentage points ahead only “just ahead”. Cameron has in mathematical terms 40% more voter preferences than Brown.
I see Tony Blair is ahead of David Cameron by 27% to 26% as “Who would you prefer as Prime Minister”.
Poor David Cameron.
re 9
I hardly see the relevance of this,IF i was asked this question I would be replying its irrelevant, as is Blair to any poll.
The real question is Cameron v Brown. If spitting image were to be reborn, I’m sure Brown would be as “grey” as Major was!
10 MTf
The Cameron/Brown question is also pretty irrelevant until GB becomes leader and people can see what he is like in that role.
PA report of poll
Fewer than one in four people would vote Labour in a general election but the Conservatives are failing to capitalise, according to a poll released today.
According to a poll by the internet site YouGov in today’s Daily Telegraph, Labour has trailed the Conservatives for eight consecutive months but support for the Tories has remained at between 37 and 39 per cent since David Cameron took over from Michael Howard as leader a year ago.
In the poll of 1,874 electors, 37 per cent said they would vote Tory if a general election was held tomorrow, 33 per cent Labour, and 17 per cent Liberal Democrat.
Labour’s support has risen by one point since November and has fallen by three points since the general election.
But the poll suggests the party should not count on a “Brown bounce” in the event of Gordon Brown succeeding Tony Blair as Prime Minister.
It asked voters who would make the best Prime Minister, first with Mr Blair as the Labour candidate and then with Mr Brown.
The results showed that both men gained 27 per cent of the vote.
Following the police investigation into the ‘cash for honours’ affair, almost two thirds of people believed “the Government and the Labour Party probably did indicate to wealthy individuals that if they donated or lent money to the party they might expect to receive in return an honour such as a knighthood or peerage”.
The Conservative Party is also under investigation over the affair, but there are no figures for whether voters believe the same statement applies to it.
To an almost unprecedented extent, voters are evincing their dissatisfaction not merely with Labour and the Tories but with all three major parties.
According to YouGov, one respondent in four - 27 per cent - would either abstain at an early election (11 per cent) or at the moment lacks any voting preference (16 per cent).
These people were excluded from the figures for voting intentions, meaning the 33 per cent who would vote for Labour account for less than a quarter of the survey’s respondents.
Opinion is more evenly divided on whether or not the Government was right or wrong to halt the Serious Fraud Squad investigation into the alleged bribery of Saudi officials.
Nevertheless, in keeping with YouGov’s other findings, the largest single group of respondents, 38 per cent, believe ministers were wrong to halt the inquiry and should have “upheld the rule of law“.
Considerably fewer, 28 per cent, believe ministers were right to act as they did. A significant proportion, 34 per cent, were unable to judge.
end
Good article Mike. These very high shares for Labour voting recall suggest the original samples taken by the pollsters may be becoming increasingly unrepresentative. This is a little surprising given the Tories’ ‘makeover’ since Cameron took over - something one might have thought would reduce false recall and perhaps non-response rates by ’shy’ Tories. Perhaps this means that technology/privacy considerations are dominant in creating this sample bias. If so, then the problem isn’t going to go away unless polling techniques change substantially.
“I hardly see the relevance of this,IF i was asked this question I would be replying its irrelevant”
……as are all questions three years from an election but as a snapshot of opinion on the leader of the opposition it’s important.That he isn’t rated more highly than Tony Blair after all the sleaze allegations and the disaster that’s Iraq and the general dissatisfaction is a stark and surprising indication of how little the public rate Cameron. Even I would have been a “don’t know”,
Often with share prices you will see a company’s price make a step change up then stick around that for sometime until for some reason investors decide to push another step change. Cameron’s election drove a step change from 31/33% to 37/39% with what looks like a step fall for Labour to 31/33% and LDs to the 17/20% range.
What’s the event or series of events that will drive the next change in sentiment? Brown’s election? the Scots & Welsh elections? Honours for Cash? EU constitution beiing revived in some form? I can’t see it being anything that the Conservatives or Lib Dems do but only a success or failure by Government and by Brown.
Cameron has had some leadership testing - his rebranding/repositioning give voters some idea of his leadership potential but really in 10 years Gordon hasn’t been. I sense a country waiting to see how events play out before judging on the next change in sentiment. Roger & others believe Gordon can do this, I have my doubts. I see the next step change as being in Cameron’s favour but it’ll be fun to find out - +10% by next Christmas? but for which leader?
14. Morning all. That Cameron is fractionally behind Blair on that question (and well within the margin of error), probably has more to do with Blair being in post and Cameron being an unknown quantity in as far as how he would do as PM. There’s always an inclination to ‘the devil you know’ in important matters (such as who should run the country). Do people really think Brown would be that much worse than Blair, if the figures are to be entirely believed? I think the same effect is probably present in those stats and it’s there that there’ll be a Brown bounce - not the overall party figures.
On topic, I think Fred puts his finger on it with his point about the increasing use of TPF, other privacy techniques and the multiplicipty of phone providers - including people who are mobile-only. My utterly unscientific gut instinct is that the use of these is likely to be higher among Tory and Lib Dem voters than among Labour and minor party voters (except perhaps Greens). If so, it will become ever harder to gain a representative sample.
16. Exactly true.
The 13 point difference between Brown and Cameron in the YouGov poll is explosive. How the Telegraph can describe it as “slightly ahead” is beyond me.
But what are the actual reasons behind such large number of people misremembering how they voted?
I can imagine/guess some: people want to identify with the winning side after the event; Conservative voters are more reluctant to participate in polls in the first place (they have less time? they value their privacy more?); Con voters are reluctant to admit voting Conservative for various reasons - because they don’t want to identify with losing side, because Tories are still seen as the nasty party (although this one must be becoming less relevant)
What other reasons could there be?
Does anyone have any more expert knowledge of this phenomenon?
RE 18, Tony A, It is also harder to get a Conservative voter on the other end of the phone. They just don’t do phone polls.
Before anyone gets too excited about the higher weighting than normla in the December polls, I don’t think it’s any sort of privacy trend or anything, more it’s a Christmas shopping trend. You get funny samples over bank holiday weekends because people go on holiday - in the run up to Christmas you start getting skewed samples because people are out shopping (it isn’t to the same degree as bank holidays of course).
Tony A - firstly, don’t get false recall mixed up with response bias. If Tory voters are less likely to take part in polls (and in phone polls it does seem to be the case) that is response bias, not false recall. It doesn’t involve anyone incorrectly saying how they voted in the past.
As for the reasons for false recall, no one knows for sure. The main reasons are probably (1) people not wanting to admit they didn’t vote at all, and therefore claiming to have voted for the winner or the party they would have voted for (2) people who voted tactically either forgetting that they voted tactically, or saying the party that was their first choice, rather than the party they actually put a cross next to (this is why recall of voting Lib Dem is low), (3) people saying how they would liked to have voted with the benefit of hindsight, as opposed to how they actually did.
Thanks for the expertise Anthony.
Must be almost impossible to proove why people misremember things. After all we are in the realm of human psychology here, the most imprecise of ’sciences’.
What you say about false recall and response bias certainly makes sense on reflection. After all if they didn’t take part in the survey, they wouldn’t be asked how they voted last time anyway.
I’m not convinced about this idea of ‘misremembering’. Obviously some people are confused or are sufficiently indifferent to politics for their vote not to register when asked later how they voted. But some people will deliberately say Labour as they will wish to be seen backing the winner last time.
I’m not sure how you can quantify these people, but we all know that they exist - in sport for example the thousands of ‘gloryhunting’ Man Utd (and now Chelsea) fans who have only a passing interest in football but want to be seen backing the winners.
Assuming that there exists even one or two percent of voters behaving in this way (I’d suspect it is more) then it has implications for pollsters. They shouldn’t be treated as real Labour voters and therefore adjusted to fit to make them ‘political representative’.
It’s just one of the problems you get when you move away from a ‘pure’ market research approach. At some point the pollsters assumptions become more important than the data - something that you cannot ever accuse MORI of, which is why I’ll go against the grain on this site and say I’d rather see slightly more volatile, but statisticaly valid polls rather than pollsters best guesses as to how representative the people they talk to are.
What i would like to know is is this 50% labour recall thing a feature of face-to-face polling, telephone, internet polling, or all of thew above?
That figure should raise alarm bells. Something ssems to be going wrong in the polling if that much correction is needed.