Media and pollsters eat humble pie for getting election so wrong

The big story of the election, it turns out, is that the big story of the election has been utterly and completely wrong all along.

The media narrative of the campaign, like many others preceding it, was built on the edifice of opinion polls. The vital importance of these polls is not surprising. Election coverage tends to be focused on the horse race – the main emphasis is on predicting the outcome of the election and pointing to the winners and losers.

Polls provide fuel for the interpretation of campaign events – what are often very subtle swings in the fortunes of the key players. These swings, viewed through the lens of opinion polls, are then attributed to everything from campaign strategies, policy successes and failures, manifestos, minor gaffes and table manners.

Throughout the campaign – and up until the very last round of election polls – the dominant narrative of the election outcome predicted another coalition government, this one with Ed Miliband as the prime minister.

According to this narrative, Miliband ran a successful campaign simply by avoiding any major missteps beyond his tumble off the stage at the final debate. He did this in the face of a relentlessly negative – and miscalculated – Conservative campaign. The most conspicuous winner, however, was to be the SNP in Scotland, predicted to win across the board after a campaign fronted by Scottish first minister, Nicola Sturgeon.

The last of these predictions held true: the SNP scored a spectacular win with a historic swing in the party’s favour, taking 56 out of 59 seats in Scotland. The rest, however, proved to be entirely mistaken. Liberal Democrat veteran Paddy Ashdown swiftly came to regret his promise of eating his hat if the exit polls proved right.

Can I nominate someone else to eat my hat?Liberal Democrats/Flickr, CC BY-ND

Shamefaced pundits

Far from being punished for negative campaigning, the politics of austerity and the out-of-touch poshness of its leader, David Cameron and his incumbent Tory party were not only elected back in, but actually added to their seats. By winning an outright majority, the party will now be able to rule without support from a coalition partner.

By contrast, Ed Miliband went overnight from being teenage heart-throb to being out of a job as party leader. His shadow chancellor Ed Balls, long tipped as a potential future Labour leader, lost his seat altogether.

Whether the wrongness of the polls was caused by a systematic bias in terms of under-reporting of Conservative votes throughout the election, poor predictions of seat distribution or by a late swing in favour of the party, the fact remains that on the morning of May 8, politicians, journalists and pundits, with stunned looks on their tired faces, were forced to swiftly rewrite the campaign story from scratch.

While the analysts pondered the shock results, politicians from opposition parties, including Caroline Lucas and Peter Hain, blamed their parties’ losses on the Conservative “fear campaign”.

When the dust settles, there is one important lesson for political observers of all stripes: putting all our eggs in the basket of poll-driven, horse-race coverage is a fatal error. As Peter Kellner, president of YouGov, stated in his own hat-eating moment:

What seems to have gone wrong is that people have said one thing and they did something else in the ballot box.

The lesson, he conceded, was that politicians “should campaign on what they believe, they should not listen to people like me and the figures we produce”.

Whether Kellner and other pollsters up and down the country were consuming hats or humble pie, the statement points to an important insight: campaign coverage driven by issues and substance instead of strategy and outcome is more likely not just to better inform the voters, but also to get things right.

So what does this mean for the future of campaign coverage? Perhaps the spectacular failure of the media-pollster nexus will be viewed as a bizarre and laughable anomaly, dramatising what was always an unpredictable election. Or perhaps it could lead to careful soul-searching and a fundamental shift in polling paradigms, methods and reporting.

There is no doubt that however the poll fiasco of 2015 is appraised with the hindsight of history, it serves as a reminder that campaign narratives are no more than frail, if convenient fictions. Like all political creations, they stand or fall on the will of the electorate.