A small but very noisy section of the British news media seems to have come close to losing its collective sanity. The election campaign is maybe not quite what the people concerned would have wanted: the Tories are locked into an ever-deepening crisis, now crystallised by a gambling scandal, Labour is capably holding itself together, and the limited fireworks let off by Reform UK do not threaten the election’s seemingly inevitable result. So, for want of any other excitement, they have turned to another source of fun: opinion polls.
Has there ever been a campaign so dominated by them? For seven or eight years now, the most powerful polling companies have been developing so-called – and yes, I had to look this up – multilevel regression with poststratification (or MRP) surveys, which contact tens of thousands of voters, calculate results based on a range of granular demographic details, and result in findings that can be sifted constituency by constituency. The fact that YouGov used this method to unexpectedly predict 2017’s hung parliament has given it an air of quasi-scientific magic; now, the publication of one such poll after another is greeted in some quarters with a huge level of expectation.
The result is postmodern news that a certain kind of 20th-century social theorist would have loved. The Conservatives, the Telegraph screamed last week, are on track to “slump to just 53 seats”. The Labour party, it said, was predicted to win a mind-boggling 516. Here, it seemed, was full-blown Starmergeddon, and the advent of a one-party state. But no one had voted and nothing had actually happened. Nor, by definition, could anyone be certain that the predictions were in any way accurate. “‘We live in a world where there is more and more information, and less and less meaning,” said Jean Baudrillard in his 1981 masterpiece Simulacra and Simulation. In this election, that distinction hardly seems to matter.
Clearly, the whole ritual – new poll shock coming at 5pm! – suits the dopamine-driven world of social media. And entirely predictably, along with the bog-standard polling that echoes most MRP predictions of a Labour landslide, all this is now feeding into the contest itself. Predictions of a red “supermajority” offer the Tories one last hope of shoring up their vote, which explains why the rightwing papers have gone MRP bonkers. Meanwhile, there are two dangers for Labour. One is centred on people who want a change of government assuming it is written in the stars and so staying at home. The other lies in an election discourse that might view anything other than an imagined megavictory as an anticlimax – or, even more absurdly, a failure.
I have recently read predictions of a Commons majority for Keir Starmer’s party of 256, 162 and 382. Somewhere in these wildly diverging figures, there is probably a sign of what is going to happen in 10 days’ time. But constituency-level predictions cannot possibly predict the results of scores of local contests that look too close to call. The reduction of people’s thoughts and choices to such crude numbers, moreover, says precious little about their view of politics and politicians, how they decide who to back, and their expectations of a winning party. Indeed, in a political world brimming with volatility, ambivalence, cynicism, tactical voting and specific local factors, polling does not even scratch the surface.
So, what is actually going on? Predicting election results, in case anyone had forgotten, is not really the job of political journalists. But there are aspects of the current public mood that seem very clear, and they are still overlooked.
Over the past three weeks, I have spent time in 10 constituencies, from Surrey to North Staffordshire, via the West Midlands, outer London, and more. The fact that the Tories’ support is in freefall seems self-evident, and partly reflects the fact that younger people are more dismissive of Conservatism than ever before. In the once-Tory town of Woking, in Surrey, I interviewed a women’s football team made up of twentysomethings, whose views reflected what I had heard from plenty of other people and places. “There needs to be a vision,” one of its players told me. “We’ve gone through austerity and Brexit, but it’s like, what’s the actual future? What’s the UK going to stand for?”
Another said that all that mattered for now was getting rid of the Tories, “but that shouldn’t be the case. We should have something to believe in.” By contrast, a trainee nurse in Birmingham – who was queueing at a food bank – said he would vote Labour because “things were better when they were in power”. But even that level of certainty was hard to find: after so many crises and turnabouts, this still feels like a country full of unease and doubt, no matter which parties it chooses. There is, it seems to me, a distinct possibility of a very cognitively dissonant big Labour win, which will have extremely complicated consequences.
In that context, my mistrust of polling is only partly to do with its accuracy. The Guardian’s election reporting is based on the crucial idea that it should cover “not just the odds, but the stakes”. And for the moment at least, going out into the electoral fields and talking to voters and candidates remains a staple of journalism, in news outlets of every political stripe. There are caveats to that: particularly in the broadcast media, 10-second vox pops and quick sprints around marginals reduce public opinion to meaninglessness. But you can still view and read plentiful examples of on-the-ground reporting. It is part of how we understand not just how politics works, but each other.
The problem is that the deafening din generated by poll after poll threatens to drown that stuff out – and thereby debase the democratic process. Elections ought to represent potential moments of reckoning, when we all talk about the state of the country and its future, the struggles and travails people go through, and the different visions of the future they have. If the people meant to lead that conversation are hyperventilating about uniform swings, representative samples and which polling company is going to be proved right, those things get lost.
There is also an even bigger risk. As the uneasy public mood proves, this is an age of profound political alienation, when nasty populists can too easily accuse the mainstream media and Westminster politicians of being part of a remote elite whose members are all the same. Because it robs politics of meaningful substance, polling hysteria gives that kind of talk the ring of truth. So does the banal framing of the election as a wholly foregone conclusion.
Other countries ban polling during election campaigns, which is one way of dealing with the problem – though that seems a very clunky response to an issue that may have a more stereotypically British solution. We should treat all those polls with deep scepticism; the best thing, in fact, may be to marvel at their arcane machinations, occasionally recognise their prescience, and laugh. The election, just to remind everyone, happens on 4 July. The rest is noise.
John Harris is a Guardian columnist
Guardian Newsroom: Election results special. Join Gaby Hinsliff, John Crace, Hugh Muir, Jonathan Freedland and Zoe Williams on 5 July
