The Politics of Polling (Liaison Committee Report)

Lord Hayward Excerpts

Thursday 19th May 2022 Lords Chamber

[Lord Hayward (Con)]

My Lords, before I start my comments in relation to both the committee and the Liaison Committee’s review, I pay credit to the noble Lord, Lord Lipsey, for his chairmanship. He alluded to his ill health. I think that everybody who was on that committee, and Members of this House who were not, are extremely pleased to see him looking so well in his place in this Chamber. I think we all welcome that.

In relation to polls, we automatically assume that everything is fine as long as the polls got the figures and the direction right in the last general election. The moment the pollsters fail to get the correct party in government, they are blamed for everything, and that is not fair. As the noble Lord, Lord Lipsey, said, there are large numbers of polling companies in operation now. I believe they work incredibly hard. It is in their own interests to produce the right results, because that is what they market to the public at large, but I do not share the confidence of the noble Lord, Lord Lipsey, in relation to the possibility that they might make errors in future. He indicated that they could, but I am not convinced that they are any more or less likely to make an error than they were previously. I say that because of the reason he identified: polling companies have to identify major demographic shifts, such as university education as against age or class, and the changes that have taken place over the last decade.

It is very interesting that the noble Lord referred to Peter Kellner, who in some of his research recently identified that some 50% of the nation’s voters have voted for different parties in the last decade because of the issues of Brexit and the changes in the red wall. Polling companies always tend to be somewhat backward-looking, because they are using how people voted last time by class, age, education or part of the country. Unfortunately, that is rather like fighting a war using the weapons they won with last time around. There is the probability that they will make the same errors again in future, when there is that same demographic shift.

The other aspect that still concerns me—I will come on to a further, small element at the end—is the way polls are reported. Again, I disagree with the noble Lord, Lord Lipsey, that journalists report things well. I am afraid that the pressures to which he referred—to sensationalise—result in a high level of hyperbole being applied to figures in any poll.

Take Ipsos MORI, which this afternoon published a poll saying that this year 31% of people have reduced their expenditure on their holiday plans. When people are asked questions, there is a temptation to give the answers they think they ought to give. Polls try to deal with that issue, and in fact the wording in Ipsos MORI’s question is quite clear—it says “noticeable reductions”. But if one compares that 31%, which produces a dramatic headline if you choose to use it—and all too often the media do use those sorts of figures—does that actually match up with what is happening at the moment?

The other day I listened to Michael O’Leary from Ryanair talking about revenge tourism. I checked just now, having seen this figure from Ipsos MORI. A year ago, his load factors were 67%; he had 1.04 million passengers in April last year. In April 2022 he had 14.24 million people flying on Ryanair, with a load factor of 91%. Does that really match up with the 31%? No—one has to take things in the round.

What worries me is that, when the media present figures from polls, they fail to take things in the round. The tendency is to give two or three questions and then we are on to the next subject. I recognise the pressure that journalists are under from producers and editors, but I wish that the media would look at things in much broader detail, rather than dashing on from one subject to the next, and that they would try to help educate.

In conclusion, I shall pick up one point from the response to the Liaison Committee by the BPC. This is associated in part with my previous comments on the use of figures and concerns where organisations pay for polls which are clearly intended to produce a certain result so that the result can be portrayed to the media, who can then portray it in hyperbolic form. This is not good for anybody. The Library briefing notes that

“The BPC also rejected the suggestion that its members should be obliged to disclose who funded each poll”.


I understand matters of commercial confidentiality; I have been involved in negotiating such agreements, though not in the field of polling. I understand the issue, but it really ought to be clear in every poll—on the can—who has funded it, so that the listener, reader or viewer can clearly understand what the intended message was in the first place. I do not think it does polling, or anybody, any good to be less than open on such matters.