Here’s what we’re all missing about the polls

It feels as if the lessons of 2017 - polls can change - and 2015 - polls can be wrong - have both been forgotten. 

Sign Up

Get the New Statesman\'s Morning Call email.

What’s behind Labour’s failure to pull into a decisive lead? I’ve been speaking to various people about the topic on an off-the-record basis and will discuss their thoughts in my column in next week’s New Statesman. But first, I wanted to talk through some thoughts of my own on the issue here.

There are, to my eyes, a couple of odd elements to the debate surrounding “Labour’s failure to establish a decisive lead/the Conservatives retaking the lead”. The first is that while everyone at Westminster is getting very excited about the polls, the movement is well within the “normal” margin of error – that is to say, plus or minus three per cent – and comes a little more than half a year after Labour all but eliminated a 20-point Tory lead to achieve a virtual draw.

Did we really all forget so quickly that the polls can change, well, so quickly? In addition, we appear to have not only forgotten the lesson of 2017 – polls can change – but also that of 2015, which is that polls can be wrong. We also know that it is particularly difficult for pollsters to capture public opinion four years from the next election, when most ordinary people have switched off – particularly as far as the crucial “how are you going to vote at the next election?” question goes.

This is not to say I think the polls now are entirely useless: they provide information about broad values and what people care about it. Were I Labour, I’d be deeply worried about the fact that between a quarter and a third of their voters, depending on the pollster, are some degree of lukewarm on Jeremy Corbyn, who is all-but-certain to lead the party into the next election. And were I a Conservative, I would be worried that the National Health Service is rocketing up the list of things that voters care and worry about.

But I’ll make a prediction, not because I expect it to be right – I think long-range punditry is even more valueless and prone to error than long-range polling – but simply because it’s always good to get your assumptions on paper in order to check against them later. It’s this: whatever the outcome of the next election, we will look at either the proportion of Labour voters who were uneasy about their Prime Minister designate or the growing number of all voters who were getting more worried about the condition of the NHS as more significant than the state of the voting intention scores.

Stephen Bush is political editor of the New Statesman. His daily briefing, Morning Call, provides a quick and essential guide to domestic and global politics.

Free trial CSS