General election result: Why did the opinion polls get it so badly wrong?
Former US political pollster Emily Hunt examines why even the lauded Nate Silver couldn't predict today's general election outcome.
I declared pretty early on in the run up to the general election that I would not be predicting the outcome. I cited all sorts of sensible reasons, the primary one for me being that I don't trust the sort of public political polling that we do here in the UK.
We just can't have our own version of Nate Silver because we don't have the endless piles of tracking and trending data on the constituency level that the US has. If Nate Silver himself – who famously got all 50 states right in the US presidential election –got yesterday's results entirely wrong, perhaps it is time to blame the data and not the interpretation. This isn't even an online in lieu of phone sampling issue, this is a the UK is just different problem. And I don't mean 'too polite' or 'too shy'.
Here's the thing, we have 650 parliamentary constituencies in the UK versus the 438 congressional districts in the US. America's electoral college may be convoluted but the data on the district level has been reliably lately. We have split out our much smaller population across nearly 200 more constituencies, all of which we need to understand in order to build proper predictive models. There just aren't enough people in all of them to do the kind of research needed.
But, predict I did. And I like everyone else got it completely wrong. The data didn't give us we needed to make the right picks, but I do also think that the pollster's plague of the not-quite-decided voter had a massive role to play in the results.
Media coverage had, rightly, in the run up to the election fixated on what would happen with a hung parliament. We were treated to all sorts of theories, none of them particularly savoury. Fundamentally, Ed Miliband declaring that he wouldn't make a deal might very well have been the last nail in Labour's coffin. The visions of upset and instability of not knowing what could happen with the government may very well have been a large contributing factor in voters deciding against Labour.
I suspect that there also may have been an almost protest vote nature to those not-quite-decided voters picking the Conservative candidate once they got in the booth. Miliband was not an inspiring choice for PM – few people saw him as at all in line with their ideal of a generic leader. The Lib Dems had lost voters' trust years ago. The Greens failed to provide a compelling alternative. And so the protest against the mainstream choices was, perhaps, to vote for the devil we know.
Either way, I would love to be doing focus groups today. I want to know how we got here, what the thinking was behind the decisions, in the voices of the real people who made them. And I hope I am not the only one. I hope that the parties are right now delving deeply into what has happened drawing on the best possible source: the people who made their decisions to vote.
Emily Hunt is a London-based strategy consultant. You can find her on Twitter @emilyinpublic.