The Croydon conundrum: has Ashcroft got it wrong again?
Poll shows sudden rise in Tory fortunes quite at odds with general swing to Labour in the capital
Lord Ashcroft has come out with his final batch of constituency polls – and one of the results has prompted questions about whether he has got it wrong, as he did in Doncaster North in November.
Then a poll in Ed Miliband’s seat suggested the Labour leader might be defeated. In the event, it turned out Ashcroft’s pollsters had made an error in weighting. Miliband was safe – and Ashcroft was quick to apologise.
The Tory peer said at the time: “One of the most important principles behind my polling is transparency.”
The question mark now is over a poll in Croydon Central which Ashcroft says is “the most striking” of his latest constituency surveys, released on Friday.
It found the Tories four points ahead of Labour – an extraordinary turnaround since October when Labour enjoyed a six-point lead in the constituency. The explanation offered by Ashcroft is that the Ukip share in Croydon Central has nearly halved, from 19 to ten per cent, since the October poll.
But is there more to it than that? Mike Smithson of Political Betting has spotted an “apparent oddity” with the Croydon poll. He says the proportion of 2010 Lib Dem voters switching to the Conservatives “is totally out of line with just about anything we have seen in his [Ashcroft’s] constituency polling”.
In the new poll, the Tories appear to be getting more Lib Dem switchers than Labour – by 41 to 34 per cent. Yet back in October Labour were getting 45 per cent to the Tories’ 11 per cent.
Another reason for scepticism about the Croydon poll is that it shows virtually no swing from the Tories to Labour since 2010. That is starkly at odds with the latest London-wide polling by YouGov for the Evening Standard. This gave Labour a 12 per cent lead in the capital – representing a seven per cent swing from the Tories since 2010.
Another Ashcroft poll that raised concerns for the prospects of a party leader was last month’s constituency survey of Sheffield Hallam which found the Lib Dems' Nick Clegg trailing Labour’s Oliver Coppard.
Now an ICM poll for The Guardian puts Clegg a clear seven points ahead of the young Labour candidate, by 42 to 35 per cent, with the Tories’ Ian Walker trailing on 12 per cent. The Lib Dems will not, it seems, be looking for a new leader on 8 May (well, not for this reason, anyway).
The ICM finding suggests that Clegg is benefiting from massive tactical voting by Tory supporters: just under half of those who would normally vote Conservative say they will vote for Clegg.
Does this mean Ashcroft is wrong and ICM right? No. Ashcroft’s sample was 1,000 – twice the size of ICM’s - and was carried out ten days earlier, plenty of time for the tactical voting to gather force.
A key difference is that ICM named the candidates when asking the voting intention question. Ashcroft, on the other hand, asks respondents to “think about their own constituency”. Ashcroft argues, not unreasonably, that if a candidate has a personal following voters shouldn’t need prompting with their name.
So it’s just the Croydon Central finding which, in my view, appears suspect.
Whatever the doubts, however, nothing can diminish the importance of Ashcroft’s innovative programme of constituency surveys. They have added enormously to the information available to politicians and strategists - not to mention members of the public interested in voting tactically.
Over the past year Ashcroft has commissioned nearly a quarter of a million telephone interviews covering around over 150 constituencies. All his figures are in the public domain for checking and analysis by interested parties.
Not least, his polling in Tory-held marginals in England makes nonsense of David Cameron’s claim that the Tories just need to gain 23 seats to win a Commons majority.
Cameron’s calculations take no account of the fact that Labour are enjoying comfortable poll leads in 50 Tory-held marginals and that a tight race is predicted in another 20. We have Lord Ashcroft to thank for that information.