Are the Ashcroft polls wrong?

An ICM poll of Sheffield Hallam puts Nick Clegg seven points ahead of his Labour rival - and raises questions about the accuracy of the Ashcroft polling.

Sign Up

Get the New Statesman's Morning Call email.

IA new poll for ICM puts Nick Clegg seven points clear of his Labour opponent, Oliver Coppard, with 42 per cent of the vote against Coppard’s 35 per cent.

The usual caveats apply: it’s just one poll. It may be no more accurate than Lord Ashcroft’s previous polls, which showed Coppard ahead of Clegg. It could be that naming the candidates isn’t a better gauge of popular opinion in a constituency. But it appears to answer a question that is being said more and more – with confidence in Tory circles and panic in Labour ones – “what if the Ashcroft polls are wrong?”

The charge sheet is this: it was Lord Ashcroft who was operating as the Tories’ private pollster when that party wrote off their chances in Hampstead & Kilburn, which they lost by just 42 votes. And his new type of polling – constituency-by-constituency – is viewed dubiously in some circles. Greg Cook, Labour’s polling supremo, believes that you cannot accurately poll individual constituencies. Another Labour strategist is less absolute but still dubious. “If you look at how wrong [Gallup, one of the earliest pollsters] got it in 1945, the mess all the pollsters made in 1992 [when the polls forecast a Labour victory but the Conservatives won on the night], how many elections before Ashcroft polling is good enough to be useful? More than one, I’d say.”

Added to that, Ashcroft’s published polls contrast sharply at times with Labour’s canvass returns and private polling. He shows thumping victories for Labour in Hampstead & Kilburn and Hornsey & Wood Green over the Tories and the Liberal Democrats respectively – but in both cases, Labour's rivals are putting up a tough fight and Labour is having to put more resources into holding onto the seats.

The most recent wave of Ashcroft polls confirms the scepticism. The thumping Tory lead in Battersea – where both sides expect a tough, close fight – and the Labour lead in Norwich North all go against what the parties’ own data suggests. “The question I have to ask myself is: do I trust Patrick [Heneghan, Labour’s director of field operations] and the decisions he’s making?” one Labour insider pondered, “Or do I trust a new pollster? And I’m afraid that I trust Patrick.”

“Afraid” is the right word. Ashcroft’s polling suggests a rather better performance for Labour around the country than their own canvassing returns suggest. Campaign data isn’t polling, of course: but remember that Labour’s data was instrumental in deciding where to make a stand and fight against the tide in 2010. For the most part, those calls were validated by either narrow holds against the odds, or defeats by smaller margins than the national swing would have predicted.

But the problem for Labour is if their own data is right, it looks increasingly likely that not only will Clegg survive in Sheffield Hallam, but he will return to government as David Cameron’s deputy once again.

Stephen Bush is special correspondent at the New Statesman and the PSA's Journalist of the Year. His daily briefing, Morning Call, provides a quick and essential guide to domestic and global politics.