Opinion Polling

Lord Bridges of Headley Excerpts

Thursday 18th June 2015 Grand Committee

[The Parliamentary Secretary, Cabinet Office (Lord Bridges of Headley) (Con)]

I am very grateful to the noble Baroness for that introduction. Sadly, no one has come to me asking for my services, although my son keeps asking me to partake in the National Lottery each week, as he is sure I can win. I also congratulate the noble Lord, Lord Lipsey, on securing this debate and am very grateful for the other contributions, which were very interesting. Indeed, this has been a very thoughtful debate, although a short one. I am delighted that we have risen above the sterile argument about regulation, good or bad, although I will touch on that.

The noble Lord, Lord Lipsey, has, as the noble Baroness said, extensive experience of polling and psephology, and I certainly cannot claim to rival that. However, he and I do share the honour—I think I am right on this—and probably the scars of having worked in the back rooms of Downing Street, only to be handed our P45s by the British people. My noble friend Lord McColl and I were there in 1997, so we all know what it is like—this could turn into a bit of a group therapy session—when you are desperately hoping that the polls are wrong but they turn out to be right; or, as in this case, not quite so right.

It strikes me that the noble Lord’s speech in this debate has focused on one particular aspect of opinion polling—the methodology—and in particular those polls that are deliberately designed to get the answers wanted by those who commission them. I do not want to get too much into the details around the ComRes issue; if your Lordships do not mind I would rather rise above that and just talk about broad issues.

Let me start by putting the noble Lord’s mind completely at rest by saying that this Government have indeed no plans to regulate opinion polls. I am delighted that there has been an outbreak of consensus on this point. Many of your Lordships would agree that statutory regulation is not the answer to the issue that we are concerned about: accurate opinion polling. There is widespread agreement that opinion polls lubricate political debate. They help to get that debate moving and to air views, and regulation of any form of opinion polling would put us on a slippery slope towards an unwanted intervention in free debate, benefiting only those with deep pockets who could afford their own polls, as my noble friend Lord McColl so rightly said.

Touching on a few of the points that I think we will discuss in tomorrow’s debate on the same issue, the power that a regulator would yield would be entirely disproportionate. It would end up sanctioning research which could then be portrayed as the official point of view. I have no idea how this would work during a general election. Would it be banned? Would the regulator be asked to adjudicate on which questions were permitted, the methodology and so on? Also, what is the scope of this regulator? While those of us within the bubble of Westminster are fixated on political polls, as I am sure your Lordships are aware the vast majority of pollsters’ business is with commercial entities who want to test what consumers think. Just think about this—you would have a cat food television advertisement that would read, “Nine out of 10 say their cats prefer it, as certified by the Consumer Research Authority, Cracom”, or words to that effect. That would be disastrous and a slippery slope. It would be unwanted regulation of business and bad for democracy. Is this necessary? I think not—but no doubt someone can produce an opinion poll to show whether it is.

As regards innovation, a number of your Lordships picked up on what I think is a key point. Regulation would threaten the debate and innovation on which polling depends. Polling is similar to that most dismal of sciences, economics. It was famously asked of the economics profession why it did not see the crash coming. Yet despite this collective failure, no one has yet called for statutory regulation of economists—not that I want to put ideas in your Lordships’ heads. This is because we understand that the technical problems inherent in economic forecasting cannot simply be regulated away. We know that improvement will only come through intensive research, open debate and rethinking of old assumptions. I would argue that it is just the same with the science of public opinion polling—a point that the right reverend Prelate the Bishop of Derby spoke eloquently about.

As the noble Lord, Lord Lipsey, well knows, this science is a far more complex business than simply phoning up random members of the public and asking what they think. Samples have to be weighted and there is no consensus about the best way to do this, obviously. Surveys have to take account of cognitive bias, and methodologies are constantly being tinkered with and adjusted. Indeed, there is a certain amount of competitive edge that companies have within that. I was particularly struck by the right reverend Prelate saying that asking someone how will they would vote days before a general election can have some bearing on how they actually behave when they enter the polling booth—picking up that little stubby pencil, their hand hovering over the box and then saying, “Actually, I am going to put my cross here”. To compare those two thoughts and those two reactions to the question is very difficult, and this is exactly what I hope the Sturgis inquiry into the last election is going to get to.

This brings me to the question of conflicting polls—one poll suggesting the public support something and another poll suggesting they oppose it. My response is, as a number of your Lordships have been saying, let us interrogate the methodology and debate the issue further, and then let the public decide. This, I would argue, is what freedom of speech and expression is all about. I strongly believe that the public—aided by a free press and vigorous debate in Parliament and elsewhere—can smell a dodgy poll. As the noble Lord, Lord Lipsey, said himself, the poll he is concerned about seemed to receive very scant coverage or mention in your Lordships’ House during the debate.

If people discover that a poll is dodgy, there are means of making complaints, as the noble Lord, Lord Lipsey, is now following. However, I would further argue that just as frightening from the pollsters’ perspective is the route that ends by being placed in the stocks of public opinion, and the shame of one’s work being lampooned and castigated by the public. Having read the weighty analysis of the poll on the parent embryo survey and what has been said about it, my strong sense is that this remains perhaps the best route to address the noble Lord’s concerns, despite what he says. It is not for me to say whether the self-regulatory bodies should do more, but if I were in their shoes I am sure that, in light of this debate and others about polling, I would want to take note of what the noble Lord is saying.

This brings me to self-regulation. I declare that in the private sector I did not just place bets, as the noble Baroness predicted, on what I thought a company might or might not do; I actually commissioned a number of opinion polls from reputable companies on issues that were of relevance to private companies. In my experience, great care was taken by pollsters to ensure that no question was seen to be leading or partial. Any suggestion from me or anyone else in the organisation I was representing that a question was, and it would be rejected and changed.

I endorse the comment of the noble Lord that the vast majority of opinion-polling companies abide by the rules and standards of the Market Research Society and the British Polling Council. One has to ask why these companies do so. It is clear that there is a simple reason: it is in their interests to ensure that their research observes the letter and, crucially, the spirit of the code of practice, and that they are seen to be asking balanced questions and presenting answers in an impartial way. Only then does their research command the respect of politicians, the media and, in turn, the public.

Furthermore, the industry fully understands that transparency and trust go hand in hand. Members of the British Polling Council must already publish their results in full, with the questions exactly as asked, a description of the sampling methodology, the raw unweighted data and, crucially, the name of the client commissioning the survey.

On the specific point the noble Lord, Lord Lipsey, has been addressing about the assembling of the questionnaires themselves, if he or anyone else who is interested cannot sleep at night they might to turn to the MRS’s guidelines for questionnaire design. It is a nice, weighty document of about 28 pages, which states:

“Members must take reasonable steps to ensure … that participants”—

that is, those who opinion pollsters are polling—

“are not led towards a particular point of view”.

This applies to the objectives of the research, and to structuring and writing questionnaires. I am sure these guidelines will be taken into account in considering the issue the noble Lord raised; as he himself said, let us see what happens.

Moving on, all this shows that like any other business or service, the polling industry’s prosperity is built on trust. If opinion polls are to be taken seriously, people—be they the public, journalists or your Lordships—must trust them. If opinion polls become a laughing stock, pollsters go out of business. Why would anyone commission research if they feel they cannot trust the results? This is why, as my noble friend Lord McColl and other noble Lords have mentioned, the polling industry is undertaking such a thorough investigation of what happened at the general election. It is in not just our interests but its own that it does this. It is just as concerned as everyone else to get to the nub of what went wrong. As has been mentioned, the inquiry’s first evidence session is tomorrow.

To my mind this is the right response to a poll failure—a transparent review of what went wrong, followed by innovation and experimentation. The methodology has to be, and be seen to be, robust so we all await the result of this autopsy with interest. I trust that the noble Lord will make his voice heard in this inquiry, and that it will be heeded. Furthermore, the noble Baroness made a number of interesting points about the inquiry taking on board experience from other countries.

Government regulation certainly would not solve many of the issues relating to methodology that the noble Lord mentioned. Regulating the industry would simply centralise the debate and decision-making process, with no guarantee that the challenges surrounding sample size, questions and so on would be overcome. A statutory regulator would be too slow and unwieldy to respond to the innovation and change brought about by big data, cognitive psychology and the digital revolution. Indeed, it would be an analogue solution in a digital age. Crucially, such regulation could—and in my view definitely would—stifle the very debate that opinion polls seek to inform. That is why government regulation is the wrong answer to the right question—a question about conduct and methodology. It is a question that the noble Lord has every right to highlight, and the existing self-regulatory bodies have every reason to heed. The Government do not plan to regulate the opinion polling industry. As Walter Bagehot wrote, and I am sure the noble Lord will say:

“The place of nearly everybody depends on the opinion of everyone else”.

Whether the decision not to regulate will have an impact on the place of government, I am not sure; I suspect we would need an opinion poll to find that out.