December Populus Poll

Populus’s final monthly poll of 2006 in Tuesday’s Times shows a drop in the level of Conservative support. The topline voting intention figures are CON 34% (-2), LAB 33%(nc), LDEM 19%(-1). Populus tend to show the smallest Labour lead of the main pollsters, and their Tory lead was as low as this as recently as October, but the drop in the Conservative share of the vote is more notable – they have been steady on 36% for the last four Populus polls, and this is their lowest level of support since prior to the local elections. The beneficiaries seem to be the smaller parties, with the “other” parties up to 14%, the highest level recorded by Populus since the election, including 4% for the Green party and 2% for UKIP.

The polls at the moment seem to be somewhat contradictory – the last YouGov poll also seemed to show a weakening of the Tory position, but the last poll from ICM showed them widening their lead to 8 points. With the Christmas holidays approaching I expect we will get the ICM and YouGov monthly polls earlier rather than later, which may shed some light on the situation.

Populus also asked their, now standard, question of how people would vote if Gordon Brown were Labour leader. This had appeared to show a narrowing of the Conservative lead last month, but this month seems to be back to normal – when David Cameron and Gordon Brown’s names are included in the question the Conservative lead grows to 7 points, 39% to 32%.

Populus also asked about how well people thought that Brown would perform as Prime Minister. While 57% of people said that they thought that Gordon Brown had been a good Chancellor of the Exchequer, it doesn’t necessarily translate into thinking he’ll be good at the top job. The 57% was made of up 34% of people who thought that he had been a good Chancellor, and would also be a good Prime Minister…and 23% who thought that while he made a good Chancellor, he would not make a good Prime Minister (there were 6% of people who thought that he hadn’t been a good Chancellor, but would be good as Prime Minister, the rest presumably didn’t rate him in either job).

Latest Scottish Polls

There was a new TNS System Three poll in the Sunday Herald yesterday. The constituency vote, with changes from TNS’s last monthly poll, breaks down as CON 11%(-1), LAB 35%(-3), LDEM 14%(nc), SNP 32%(+2), GRN 3%, SSP 4% (though the Greens do not put up candidates at the constituency level, so come the actual election these voters will obviously have to go elsewhere or not vote). Regional support stands at CON 11%(+2), LAB 32%(+2), LDEM 15%(-2), SNP 30%(-3), GRN 5%(nc), SSP 4%(nc).

There doesn’t seem to be any consistent trend since the TNS poll last month – at the constituency level the figures are better for the SNP, at the regional level the figures are better for Labour. In both votes Labour and the SNP are very close to one another.

System Three continue to paint a somewhat better picture for Labour than to the other pollsters – at the constituency level System Three tend to put Labour around the mid-30s and the SNP nearer 30%; ICM and YouGov tend to do the opposite. In the regional vote System Three have put Labour at 30 or above in their last two polls, while ICM and YouGov have put them in the mid-to-high 20s. The difference between the polling companies is very likely down to ICM and YouGov using political weighting in their samples.

On a separate note, a fortnight ago The Herald reported a leaked internal Labour party poll by Populus that put the Labour party 8 points behind the SNP on both the constituency and the regional vote. No other details were given about when it was done, what adjustments were done to take account of turnout and so on. Populus are, of course, members of the BPC and, as such, I said a week or two ago that I would try and get them to disclose the full tables. The Labour party have stated that the figures were genuinely leaked without their approval and, since only the lead was disclosed and no actual figures were, Populus had decided that it probably doesn’t fall under the aegis of the disclosure rules.


Since the election have been nineteen polls asking how people would vote with Gordon Brown as Labour party leader. Populus seem to have pretty much made it one of their regular tracker questions. Every month when one is published I try to add a caveat of some sort to them saying not to worry too much about them, but how useful are they as a guide to how Gordon Brown will perform?

It is easy to see why the papers commission this question. British politics is in a strange interregnum. We all know Tony Blair is going very soon, we know that whatever policies are announced now are liable to be changed once Gordon Brown enters Number 10, whether people would vote for Tony Blair or not is pretty much irrelevant. Naturally the papers are keen to look past this to proper politics, once the new Prime Minister is in situ so are trying to do so by asking how people would vote with Brown in place.

The results have been very consistent. Labour would do badly. Since David Cameron became Conservative leader in December no hypothetical poll of how people would vote with Brown as Labour leader has shown a Labour lead. In the vast majority of cases, it has shown the Conservatives performing better against Brown than they are at present. But does this actually mean anything?

The first question is whether this is actually anything to do with Gordon Brown. Normal voting intention questions don’t mention party leaders, just the party names. Obviously if a question mentioned just Gordon Brown it would be skewed to Labour, so the hypothetical questions also mention David Cameron and Sir Menzies Campbell. The difference in the results could, therefore, not be a result of Gordon Brown at all – it could be a positive result of mentioning David Cameron’s name (or a negative one of mentioning Menzies Campbell’s).

Back in July Populus sought to solve this question. Using a split sample they asked three questions – one was a normal voting intention question. The second was a voting intention question with the names of the present party leaders. The third was a hypothetical voting intention question with Gordon Brown as Labour leader. The results were that mentioning the names of David Cameron and Tony Blair also increased the Conservative lead, from 2 points in the unprompted question to 7 points in the prompted one. Changing the Labour leader from Blair to Brown further increased the Conservative lead – up to 9 points.

What this suggests is a large part of the apparant change is just the effect of mentioning David Cameron’s name in the question (or a negative result of mentioning Tony Blair’s name) but some of it is also people being less willing to vote for Brown’s Labour party than Blair’s labour party. Certainly it doesn’t suggest that Brown would increase Labour’s support relative to the Conservatives.

The second question is whether it means anything. People are not very good at predicting their future behaviour, and obviously they cannot take into account “events” that haven’t happened yet, future policies and publicity. At the moment they are judging Gordon Brown on his record as Chancellor. As yet people have no way of knowing what policies Brown will announce when he becomes premier, of how they will react to him as Prime Minister. At least, they don’t in some ways.

My personal prediction is that, when Gordon Brown actually becomes Prime Minister Labour will experience a strong boost in the polls. Obviously it depends where they are starting from, but it wouldn’t surprise me if they regained a healthy lead. There will obvious be a flurry of press coverage, TV profiles, interviews and so on. Brown will undoubtedly start his premiership by rolling out new policies and new agendas designed to boost his support. However people think they’ll react, in reality I suspect many people will want to give Gordon Brown a chance. Whatever the polls say now, I am confident that reality will be different and Brown will take Labour up in the polls, not down….in the short term.

In the longer term, once the bells and whistles of Gordon Brown’s accession have faded away, I’m afraid that the present polls probably do give a good indication of the sort of effect he will have on Labour’s fortunes. Here is why.

The polls are pretty consistent on questions of why people seem to dislike Gordon Brown. They consistently show that people rate Brown as having run the economy well, of being strong, reliable, experienced, competent and efficient. On every rational measure people rate Gordon Brown very highly, yet if you ask whether people would rather have a Brown or Cameron led government, Cameron leads. Brown’s weak point is that people think he is not charistmatic, caring or likeable. However highly they rate his performance as a politician, they don’t like him as a man.

That is why I think the present polls do say something about Brown’s future success. If they showed people didn’t like Brown because they doubted his competence or experience, or disagreed with his policies or principles, then it would be perfectly possible that Brown in office would change peoples’ opinions, impress them with his competence or change their minds with new policies and ideas. Brown’s negatives though are far more nebulous – they just don’t like him. This will be very hard to change.

Gordon Brown has been on the public stage for well over a decade. People have a very firm image of him in their minds of a dour, brooding Scotsman, and it would be incredibly difficult to shift. Perceptions of Tony Blair have changed in the last ten years, but the core impression that he is at heart a decent, fairly normal, sort of family chap is unmoved. Polls show that people think he is an untrustworthy liar, he is firmly associated with a now deeply unpopular war and seen as a poodle to an even more unpopular President. Yet polls show that people still consider him likeable. Impressions of Blair as a person haven’t shifted. The public formed an impression of William Hague as a bit of an oddball very early in his leadership and he never shifted it. Michael Howard was never able to escape the public image that Ann Widdecombe had summed up as Howard “having something of the night” about him. I have great doubts that any rebranding exercise could do anything to change perceptions of Brown’s character.

If you wish Gordon Brown success, as I am sure the majority of Labour supporters reading this do, then I am sure you are saying to yourself that this doesn’t matter. People are not so shallow as to vote on such trivialities as who is the nicest chap. Alas, while I’m sure democracy would be far healthier if everyone cast their vote based on a lengthy consideration of the rival manifestos, they do not. Most people vote on broad perceptions of the parties, including the leaders, and the fact is that people do not make purely rational decisions. This applies to all of us, however smart or immersed in politics we are. You cannot turn off your subconscious.

Why do companies spend so much time and money on the packaging of their goods? Because it makes people buy them. It isn’t just a case of good packaging making them look more appealing on the shelf though, good packaging changes opinions of the goods inside them. In taste tests of identical products in different packaging people will pick one over the other and believe it is on grounds of taste, there are companies devoted to it. Put margarine in foil, people think it tastes better. Put more yellow on a can of 7UP, people report that it tastes more lemony, people think that ice-cream tastes better if they’ve bought in a round box. If you ask people why they prefer that particular margarine, soft drink, ice-cream, etc they won’t say – it was in foil or a yellow can or a nice round box. They are under the impression that it is actually better.

At this point many of you reading this – possibly nearly all of you – will be thinking something along of lines of “Politics is different. You can’t just apply the logic of flogging ice lollies to selecting the head of government. People take it more seriously.”

The reason you think people look at politics differently is because you do, right now you are reading a blog about political opinion polling. It’s a long post so you’ve probably spent quite a while doing it. To you politics is far more important than fizzy drinks and margarine. You aren’t typical. Most people aren’t particularly interested in politics. Most people cannot identify politicans beyond the party leaders (Brown himself is one of the few exceptions). To most people politics isn’t that important. More importantly, we are talking about subconscious reactions. You cannot turn them on and off depending upon how important the issue is. People don’t think “I am only buying carrots so I shall allow my subconscious thoughts free rein” and later, “Now am I voting, so must become a dessicated calculating machine, banishing all but rational thought”. You couldn’t if you wanted to anyway.

You will probably have done the Implicit Association Test at Harvard University’s website in the past. It flashes rapid pictures and words on the screen, black faces and white faces, negative words and positive words, and sees if people are microscopically slower at associating positive words with black faces than they are with white faces. It measures your unconscious prejudice over age, gender, race and so on. If you haven’t yet tried it and you’ve got a spare 15 minutes or so do so, the chances are that you’ll find that you have some minor degree of subconscious prejudice on age, race, etc. The thing is, even if you do your level best and really concentrate on not being biased at all in the test, it is very hard to fool it. The point is, you can’t turn off your prejudices at whim. If you are prejudiced towards white people, or young people… or English people, or pleasant, likeable people (or indeed, against grouchy, insular, Scottish people) you aren’t going to turn it off.

Choosing someone to run a major international company is presumably a vitally important issue. The CEO of major companies like BP, General Motors or Microsoft arguably has more power than the heads of government of some countries. When appointing CEOs company boards presumably make decisions on things like competence and ability and not trivialities like, say, how tall they are. A third of the CEOs of the top 500 American companies are over 6’2″ tall. Under 4% of American men as a whole are over 6’2″. Presumably no company boards chose their CEO on the basis of height, but clearly there was some subconscious bias at work here.

I’ve referred to Malcolm Gladwell’s Blink here before (in fact, if you’ve read it you may have spotted that some of the packaging examples and the comment about the height of CEOs are drawn from it). In his book Gladwell talks about the “Warren Harding Error”. Warren Harding is normally rated as one of the very worst American Presidents. He had an undistinguished record in Ohio politics. As a Senator he didn’t turn up for the majority of votes. He was an ill-educated, was having a long term affair with a friend’s wife and illegally drank alcohol. Despite his manifest shortcomings, Gladwell argues, he got elected because he looked like a President, a tall, handsome, patrician figure with a strong, rumbling speaking voice. Subconsciously people assumed that because Harding looked like a great President, he would be a great President. They were wrong.

People’s opinions of other people are influenced by unconscious prejudices. So are people’s votes. People consistently voice their approval of Brown’s performance in his job and say how strong, effective and competent he is. These things are clearly not the problem. The reason he polls worse than Blair, must be because he is seen as having an unattractive personality, people just don’t like him, and that will be difficult to change.

Of course, that doesn’t mean he won’t win. James Callaghan was always regarded as far more likeable than Margaret Thatcher yet she won in 1979 and twice thereafter. I do think it will be a real obstacle though. My prediction is that there will indeed be a Brown boost when he becomes leader, but that it won’t last. The public already have firm perceptions of Brown, and they won’t be easy to shift, especially when he is up against an opponent in the form of David Cameron who seems to have the exact opposite effect upon the public.

ICM correction

The ICM poll published in the News of the World was actually their unadjusted figures. Once ICM’s spiral of silence adjustment is taken into account the correct figures are CON 39%, LAB 31%, LDEM 20%.

Now the full figures are up on the ICM website we can compare them to ICM/News of the World poll back in February, which asked many of the same questions about Gordon Brown and David Cameron. While there have been lots of surveys asking these sort of questions on which politician would be better at this or that, there haven’t been many with comparable questions like this.

Overall perceptions of David Cameron seem to have improved and perceptions of Gordon Brown have fallen, relative to one another. In February Brown recorded a 7 point lead as the more in touch of the two men, now Cameron leads by 10 points. In February Brown had a 11 point lead on trust, now Cameron has a 5 point lead. In February Cameron had a 5 point lead on the man most likely to stop and help, he now has an 11 point lead. In February Brown had a 1 point lead as the man most likely to have a good idea, now Cameron leads by 12 points.

The ironic thing is that, despite this, Brown led Cameron in the Best Prime Minister question.

UPDATE: Looking more closely at the two polls, it looks as though Cameron may not have had a big boost after all. These questions were conducted as part of a voting intention poll, and were therefore weighted by past vote. The figures from February were not, and it doesn’t look as if they were weighted by past vote. At least part of the difference between the figures is therefore probably just down to the different weighting.

A YouGov poll in the Telegraph suggests continuing unease over ID cards and the national database, and the potential for outright refusal to co-operate with any scheme from a minority of the public.

50% of people said they were in favour of ID cards, with 39% of people opposed. This is very similar to the last two times YouGov asked the question.

Asked how much they would be willing to pay, only 11% of people said they would be willing to pay extra, although it may well be that people misunderstood the question. The wording asked how much people would be willing to pay for a combined biometic passport and ID card, given that passports currently cost £66, and I suspect some people interpreted as how much extra they would be willing to pay.

Of the 39% opposed to ID card, 25% think they will do more harm than good and 30% think they will be too expensive, but 43% say they are opposed on principle. The poll suggests that many of those people would refuse to co-operate once cards were introduced. 21% of those opposed to cards said they would refuse to have a card even if it meant paying a small fine, 7% said they would refuse even if it meant a long fine. 15% said they would refuse even if it meant a prison sentence.

This works out at about 17% of the population who say they would refuse to co-operate with the ID card scheme. Of course, a large proportion – probably a large majority – of this will turn out to be empty bravado. It is far easier to claim in a survey that you would be willing to go to prison rather than have a card than to actually go to prison. However, if even a small proportion of that 17% of people actually do stand firm then there could be a severe problem with non-compliance and media focus on “ID card martyrs” (not having an ID card will not be an imprisonable offence, but the ultimate punishment for not paying fines for not having an ID card would be).

The survey also asked about the national database that will back up the ID card system. 43% of people thought that the information held upon it would be accurate and reliable, but 48% thought it would not. 66% said they did not trust the government to keep such data confidential and 82% thought there was some danger than civil servants working on the data would divulge it improperly to others. Taking these things into consideration, 52% said they were unhappy about a national database 52%.

The survey also asked about CCTV cameras, also regularly cited in articles about increasing state surveillance, and in contrast to ID cards and the national database found strong support for them. 37% of people said that “CCTV cameras and so forth” made them feel they were being spied upon, but when asked if they approved or disapproved of CCTV in various places, approval was overwhelming. 97% of people approved of CCTV in banks, 93% on public transport, 86% outside pubs, 85% on high streets. The lowest support was for CCTV in taxis, but even then 65% of people supported it.

On other facets of the “surveillance society”, 72% supported photographing airline passengers, 56% of people supported roadside fingerprinting of alleged suspects, 50% supported speed cameras and 45% supported fingerprinting airline passengers (39% were opposed). A majority of those expressing an opinion of people (48% of people overall) were opposed to maintaining DNA records of people who have not been charged or have been acquitted, with 37% of people supporting it. The idea of using the chips within ID cards to track people’s movement met with sharp opposition – 70% said they would disapprove with only 16% approval.

Overall people are supportive, but uneasy, about increasing state surveillance. CCTV cameras have wide support, as do security provisions on flights (presumably because of memories of past terrorist atrocities). However, there are doubts about whether the national database will be secure or accurate and public attitudes towards ID cards are ambivalent – around half of respondents still support them, but a minority seem to be staunchly opposed to the extent that they claim they will be willing to break the law rather than accept them.