Opinion Poll graphic

If you stepped outside your house last week, you almost certainly collided with an opinion poll. There were no less than five on the subject of Europe alone: Lord Ashcroft’s, of course, as well as ones from ComRes, Ipsos MORI, ICM and Survation. And that wasn’t all. There were others with the usual figures on voting intentions and impressions of the party leaders. The polls had taken over.

It was much easier to avoid a document that one of the pollsters published the week before. This was YouGov’s investigation into why it, like so many others, had miscalled the result of this year’s general election. The conclusion? That, for the most part, they exaggerated Labour’s support because they over-represented the youth vote in their samples, and under-represented the over-70s. These errors are now being corrected.

The juxtaposition was almost brazen in its juxtapositioning. One week: a reminder of the pollsters’ fallibility. The next: a glut of new opinion polls, all enthusiastically reported in the press. This was meant to be a period of disdain for what the surveys say, at least judging by all the tua culpas written in May and directed at the polling companies. It has turned out to be nothing of the sort.

Plus ça change, as they sigh in the polling houses of Paris. Opinion polls always did provide easy copy for journalists, and that hasn’t changed with the events of seven months ago. It is still simpler – and, crucially for struggling newspapers, cheaper – to write about numbers delivered into your inbox than about what’s happening in the field.

Besides, opinion polls can provide good copy too. There’s a usefulness in knowing not just what the man on the street thinks, but also what every man and every woman on every street thinks. At their best, polls help us to ascertain this. Even when they’re not at their best – as during the last Parliament – they might still tell us something. Are the Conservatives going up? Then they’re probably going up. Are Labour going down? Then they’re probably going down. And so has it always been. Trends can contain truth, even if the actual numbers are false. They are worth reporting.

But something has also changed since May. Mostly, it’s in the types of polls that are being highlighted by the papers. Where it used to be daily scrutiny of voting intentions, it’s now near-daily scrutiny of people’s views about Europe and about Jeremy Corbyn’s leadership. There’s also a bit, which will surely become more, about who should succeed David Cameron as Tory leader.

This is topicality at play. Yet I wonder whether it’s also some of the scepticism we were promised after the election. Perhaps commentators are now warier of any numbers that put one party on top of another. Perhaps they are waiting for when the pollsters correct their methodologies.

In any case, the recent by-election in Oldham West felt like the first attempt at a new order. Had this occurred during the last Parliament, there would have been a dozen polls predicting the outcome. As it was, there was barely anything. Journalists were left straining for stories on the breeze. They spoke to sources who were, apparently, familiar with the situation in the constituency. They reported back on what they had heard. And – delicious irony – it was wrong.

Despite all the stories, UKIP wasn’t actually very close to Labour. There were some 10,722 votes between them in the end. Did any pollsters smile grimly at the news, and ask: do journalists have to launch inquiries when they make mistakes too?

There were very few lessons to emerge from Oldham, but one was that the truth is always mixed. Polls, like journalists, have their strengths and their weaknesses. The problems arrive when they are allowed to dominate what Westminster calls “the narrative,” as happened before the last election. So much energy and ink was expended – including by me – on the idea that the Conservatives could never ever hope to win a majority because that was what the polls suggested. So much energy and ink was wasted.

Could it happen again? Of course. There are already signs that opinion polling will shape, or even define, the narrative over whether we should Leave or Remain. But there is also the danger that it will shape, or even define, the counter-narrative too. It will suit some people, in this Parliament, to point towards the failures of May and dismiss the polls entirely. And so our politics could become an argument over whether the polls are right or wrong, which isn’t really politics at all. It’s bookkeeping.

Opinion polls are meant as one guide among many, not the Christmas Star by which we should set our course. By all means, follow them as they return to newspapers, but don’t do so too rigorously. Remember their limitations, but don’t ignore them completely. Everything in moderation.

22 comments for: The return of opinion polling

Leave a Reply

You must be logged in to post a comment.