Peter Kellner, YouGov and ASH
Wednesday, October 8, 2008 Prior to the smoking ban Forest commissioned a series of opinion polls to find out what the public thought about smoking in public places. Knowing that politicians and journalists would be sceptical of any poll carried out by Forest, we commissioned Populus, pollster to The Times, to do the work on our behalf.
We also took the decision to ask the same questions about public smoking that were being asked by the government-run Office for National Statistics. (Reassuringly, our results were very similar to theirs.)
Each time we commissioned a poll we were advised by Andrew Cooper, director of Populus. To this day I have no idea what Andrew's views are on smoking or government policy on tobacco. (He could be a rabid anti-smoker for all I know. Or a 40-a-day man, although I never saw him with a cigarette.) It was something we never discussed. It was neither relevant nor appropriate.
Compare this to Peter Kellner, president of YouGov, the country’s leading internet-based market research firm. On Monday I discovered that Kellner is chairman of the “editorial board of scientific and medical experts” that produced the latest ASH report, Beyond Smoking Kills. Yesterday (I quote from the press release issued by ASH) he declared:
“There is a terrible gap in life expectancy between the richest and poorest in British society. Smoking is by far the biggest single factor, accounting for half the difference. A child born today who never smokes will live, on average, ten years longer than a child who takes up smoking. The younger a smoker starts, the harder it is for them to quit.
“Our report sets out a comprehensive strategy for Government action to protect children from exposure to smoke and smoking, to support smokers to quit and to help smokers who are not yet ready to quit to reduce the harm they cause themselves. Smoking remains the greatest public health problem in our society – but we now have a great opportunity to stop the next generation from inheriting this lethal habit.”
And that's not all. Writing on the Guardian website, Kellner explained that:
“Earlier this year Action on Smoking and Health (ASH) commissioned YouGov to explore public attitudes in detail, as part of a report on the next steps for tobacco control, Beyond Smoking Kills, published today. Our first finding was that support for last year's ban is higher than ever. It had already grown from 51% in 2004, before the legislation was debated, to 72% before the ban took effect. It has now climbed further, to 77%. Even smokers now favour the ban, though by a narrower margin. This matters because a decline in support for existing measures would make it hard to introduce new ones. In fact, YouGov's research suggests that the ban has whetted the public's appetite – and, indeed, the appetite among many smokers – for further action.”
These figures may be true of ASH/YouGov polls, but it's worth noting that surveys conducted by Populus and the ONS prior to the ban produced very different results. For example, a poll by Populus for Forest, published on 7th February 2006 (a week before MPs voted for a comprehensive ban), showed clear public support for the government’s manifesto pledge to allow smoking in private members’ clubs and pubs that don’t serve food.
According to the survey, 59 percent believed that smoking should be allowed in pubs that don't serve food; with 63 percent supporting smoking in private members' clubs. The poll also found overwhelming support for the introduction of designated smoking rooms in pubs and bars that do not serve food (66 percent in favour, 30 percent opposed).
In 2007, a year after Scotland introduced a public smoking ban, another Forest/Populus poll (in Scotland) suggested that a clear majority continued to favour restrictions rather than a total ban. (See HERE.)
Meanwhile, annual surveys by the Office for National Statistics found that while the majority backed curbs on smoking, the majority did NOT support a comprehensive ban.
Why the difference between the various polls? Well, as I recall, ASH tended to offer people a stark choice: smoking or non-smoking (in "public" places). Forest/Populus and the Office for National Statistics gave people a choice of smoking, non-smoking, mostly smoking (with smoke-free areas), or mostly non-smoking (with separate smoking areas). (See THIS UK Polling Report from September 2005.) The most popular choice was non-smoking with separate smoking areas. Overall there was a clear majority in favour of restrictions but against prohibition.
The government knew this yet still went ahead with the ban. According to ASH, it was “literally a confidence trick”. Writing in the Guardian (19 July 2006), director Deborah Arnott and her sidekick Iain Wilmore boasted:
“It is essential that campaigners create the impression of inevitable success. Campaigning of this kind is literally a confidence trick: the appearance of confidence both creates confidence and demoralises the opposition. The week before the free vote we made sure the government got the message that we ‘knew’ we were going to win and it would be better for them to be on the winning side.”
Following the publication of Beyond Smoking Kills, some might think that ASH is repeating this "confidence trick". I couldn't possibly comment. What I am certain of is that there should be a clear distinction between a campaign group such as ASH and an independent polling organisation like YouGov. Thanks to Peter Kellner, that boundary is now blurred.
Simon Clark
Dolan Cummings of the Manifesto Club writes about the ASH/YouGov report HERE.
People
Reader Comments (9)
It is very easy to have a poll give the results your clients want. The questions can be loaded such as 'Do you think that babies should be exposed to their parents smoke in their cots?' as opposed to 'Do you think that home owners should be banned from smoking in their own private dwellings?' and 'Do you believe that there should be private members clubs specifically for smokers' versus 'Do you think that smokers should be entitled to blow smoke in your face when you go to a restaurant for your meal?'
These loaded questions have no place being used by a supposedly reputable pollster but I have my doubts now with YouGov particularly since they do not admit to ASH being a client. If this is a personal crusade by Peter Kellner or he is taking money to 'Produce' the right result for ASH then none of YouGov's poll can be deemed to be of any worth at all. This should be publicised as if the public do not trust a particular pollster then their results are worthless. Their clients would certainly be advised to take their business elsewhere to a less tainted company.
I was pleased to see the following story today:
In June, 2006, then Surgeon General Carmona released his report titled "The Health Consequences of Involuntary Exposure to Tobacco Smoke". Since that date, his report has drawn criticism from Scientists and Epidemiologists worldwide.
Four separate groups have filed complaints with the Office of Research Integrity, Health and Human Services against Ex-Surgeon General Carmona's 2006 Report
Michael, it would be wrong (and potentially libellous!) to suggest that Peter Kellner is "taking money to 'produce' the right result for ASH". I don't believe that at all. Ditto, I do not believe that YouGov is a "tainted" company. I have a lot of respect for YouGov and Peter Kellner and I don't doubt that PK is very sincere in his views. I am merely making the point that, in my view, a pollster should not only be independent of its client, but be seen to be independent - and that by allying himself so closely to ASH and its campaign for further tobacco controls, PK could cause some people to question his impartiality, regardless of the facts. This is a question of judgement - nothing else - and I wouldn't want to put it any stronger than that.
PS. Don't get me wrong, YouGov is far from secretive about the fact that ASH is a client. Peter Kellner is very open about it - both in his Guardian article and at a fringe meeting which he chaired at the Labour conference in Manchester. I was merely pointing out that ASH is not listed on the clients page on YouGov's website - and I wondered why this should be, given the amount of work that YouGov (very openly) does for ASH.
Simon. I do not particularly fancy being sued by Peter Kellner and to be honest I have always found him to be an excellent politcal commentator and pundit. But, if he was chairman of a Labour/ Lib-Dem/Tory think tank would anyone believe YouGov when they published political polls? If he was the chairman of Greenpeace could we trust any polls on global warming?
I agree he is showing a lack of judgement and he should see the conflict or perceived conflict. By ignoring this it calls into question his impartiality and thereby taints his company. I also noticed that his clients include the BMA, NHS, BUPA and the TUC all anti smoking organisations. I did not see any pub chains, tobacco companies or small grocers who have been most adversely affected by the bans and restrictions. Would he produce a poll result that contradicted his own views and also those of some of his other large clients? Maybe, but I for one will be taking YouGov polls with a very large pinch of salt from now on.
What a nasty, spiteful and vicious little person that Deborah Arnott is. If I had the misfortune to meet her I would slap her face. I would be worth the fine.
As for YouGov I would never trust anything they did as the very mention of Gov makes me wary.
Sylvia
Would that be 'armful? Did you know that the anagram for Deborah Arnott is Abhorrent Toad.
It is important to remember that the YouGov panel is not a random sample. As a paid YouGov panellist, you are profiled by them in two ways:
First, there is the profile you fill out on joining. This tells them your age, where you live, yours and if appropriate your spouse's employment, how much you both earn, your educational-level, what newspaper[s] you read and how often, home-ownership status, which ITV region you have access to, how often you watch BBC Newsnight, your ethnicity, your religion, who you voted for in 2005 and how you "usually think of yourself" politically.
Already quite a full bag from which to dip selectively from, you'd think. After all, panellists are never made privy to the eligibility-criteria for each survey and these criteria are unpublished.
But they also profile you through answers you give in response to subsequent survey-questions. As a panellist, you have no access to this profile. It is their hidden profile of you and helps determine your eligibility for subsequent surveys.
Perhaps my mistake was to be honest. In an early survey, I was asked to provide written reasons for my defection from Labour to Tory and I cited the smoking-ban as one reason. I have NEVER been asked my views on the smoking-ban since. Not once.
And it's taken four long years for me to amass the £50 in accumulated survey-fees they demand before the cheque gets sent. The only survey I am regularly asked to participate in is the interminable "Brand Index" effort where the most political it gets is "Top Shop" or M&S?"
Take YouGov results with all the salt in the Irish Sea.
Better late than never. Yes, a lack of judgement on PK's part... but take their findings with a pinch of salt? are you bonkers? time nd time again thay have proved that they are incredibly accurate when it comes to fining out what people think or do - e.g. London Mayoral, European elections, etc etc (I was actually searching for the Harriet harman vs YouGov spat, when I stumbled upon this blog.
Time and time again they have proved that selective polling can change public attitudes but that doesn't make YouGov accurate or right. It just makes it manipulative.