Opinion Polling

Baroness Hayter of Kentish Town Excerpts

Thursday 18th June 2015 Grand Committee

[Baroness Hayter of Kentish Town (Lab)]

My Lords, like the right reverend Prelate, I will keep more to the May election. ComRes has been mentioned quite a lot and I read just after the election that it said that it had had been a,

“difficult few days for pollsters”.

They should have tried being in the Labour Party. The nub of the problem is that the pollsters got it wrong, as in 1992 when I experienced the same 10 o’clock shock. I was sitting alongside the then deputy leader of the Labour Party, Roy Hattersley, whose minder I was at the time. I had anticipated him being Home Secretary within a few hours. So I am a bit bitten by this.

While 1992 and 2015 may have been bad for business for the pollsters, it raises bigger issues for the country and for those seeking to run it if the publication of misleading polls alters voting behaviour. Of course, in moments of loss such as I and my noble friend went through that night, we activists feel that blow, and we are reminded of course of Bertolt Brecht’s “Die Lösung”:

“Nach dem Aufstand des 17 Juni”.

Yesterday, therefore, was the anniversary of when the country,

“Had forfeited the confidence of the government”,

making the solution,

“for the government to dissolve the people and elect another”.

The temptation to do that on 7 May was great.

However, the more serious question posed by my noble friend Lord Lipsey is serious, albeit that he has concentrated on a different variety of polling: namely, that commissioned by or for a particular campaign, often with loaded questions. He is, of course, one of the most experienced in the field, having studied, used, commissioned, interpreted and reported on polls since I first worked alongside him in 1970.

The problem we discuss today is an old one: whether our reliance on soothsayers and fortune tellers, or indeed bookies, can affect our actions or policies. Pollsters are not soothsayers, but because of the role that they play in how we as politicians frame our campaigning and even our policies, and in how voters choose to vote, there is a special responsibility on them to raise their game, as there is on the media that report them.

My noble friend brilliantly covered the traps and shortcomings of some polling. While we acknowledge that the polls have often been accurate, today’s debate is about where the sampling, the methodology or the questions failed the industry and the body politic. I would worry about any pre-approving of sampling or other methods, as this could stifle innovation and lead to even more clustering or huddling. I also cannot see that there can be acceptable or, perhaps more importantly, unacceptable questions. However, there is some urgency to improving the industry, especially before we face the first recall ballot for an MP, where, in a single constituency, a vote to trigger a by-election could be heavily influenced by some local—and possibly shoddy or loaded—polling.

However, I wonder whether the industry has the appetite to do more itself. Has enough yet happened in the way of peer reviewing its academic approach in order to raise standards or to guard against the drive for cheap, headline-grabbing polling, undertaken for commercial rather than democratic gain? As my noble friend suggested, I do not perhaps share his faith in the ASA model, but I share with him the desire for improvement and for the industry to take a long, hard look at how it produced its figures.

However, that is only part of the story. As the right reverend Prelate suggested, we also have to look at how polls are reported, not only by newspapers but by radio and TV, which sadly too often take their agenda from the papers and can make the poll a lead story rather than background intelligence. We should also look at how this translates on the doorstep. Part of my own shock at the 10 pm exit poll came from the fact that I had mostly been campaigning in London, where my own experience pretty closely reflected that of the published polls and therefore gave me too much confidence that they were right elsewhere. I would be interested to hear from experienced campaigners outside London whether their feel was different from the published polls and whether voters’ responses appeared influenced by their expectation of the outcome.

The opinion poll inquiry has, I think, been called comprehensive. Sadly, I do not find that. It is very UK-focused, as if we have nothing to learn from elsewhere, and also misses the input of candidates and campaigners; and indeed of journalists, who may have tales to tell of how they were given the data—exactly when, how close to when they had to use them and with what spin. There is also the question of the degree to which news reporting and the polling were so intertwined that there was no independent review between one activity and the other. As the noble Lord, Lord McColl, noted, the inquiry will be holding a public meeting tomorrow afternoon, I think at the Royal Statistical Society. I hope that some of those wider questions can be posed there.

The issues raised today are important—perhaps too important to be left to pollsters. I congratulate my noble friend Lord Lipsey on initiating the debate and on sharing his considerable expertise. I look forward to the thoughts of the Minister, who—I told the House this last week but some noble Lords may not have been there to hear it—placed a bet 12 months ago on a Conservative majority of 12. Perhaps we should just replace the pollsters with the Minister.