Michael Ashcroft's reputation as a pollster has taken a battering today, after he admitted that the results of three key English constituency polls published towards the end of last year were inaccurate due to a basic error of arithmetic. At the time, it was reported that Ed Miliband was under severe pressure from UKIP in Doncaster North, that Nigel Farage was trailing the Tories by a significant margin in Thanet South, and that Nick Clegg was very slightly ahead of Labour in Sheffield Hallam. It turns out that all three narratives were totally misleading - Miliband was in fact light-years ahead of UKIP, Farage was more or less level-pegging with the Tories, and Clegg was slightly behind Labour. And to make it even more embarrassing, this is the second time the Doncaster North result has been revised.
When I heard about what had happened, my initial reaction was "how refreshing to see a pollster openly admit they've made a pig's ear of it". But in fact Ashcroft hasn't done that - he's instead tried to shunt the blame onto the (unnamed) firm he commissioned to conduct the fieldwork for the polls. That's rather unseemly. It would be fair enough for that firm to take the rap if they had been allowed to publish the poll under their own name. But if Ashcroft wants to set himself up as a pollster in his own right, then the buck stops with him - he chooses the weightings and adjustments to apply to the raw data, and it's ultimately up to him to ensure they are applied correctly. If he doesn't bother double-checking before hitting "publish", he has no-one but himself to blame.
The most important upshot of this is that we now have two constituency polls in Sheffield Hallam which agree that Nick Clegg is on course to lose his seat. That's not to say that he necessarily will lose - the decks are stacked in favour of any party leader in this situation because of all the free publicity they get, and because people seem to like the 'honour' of having their constituency represented by someone important. On the other hand, there are examples from other countries of even the most revered leaders being swept away when the electoral tide is strong enough - Helmut Kohl was defeated in his constituency seat in the 1998 German election. And Nick Clegg is scarcely a Helmut Kohl.
If he does lose, the Liberal Democrats will be plunged into chaos at exactly the moment they hope to be negotiating a new coalition with either Cameron or Miliband. They'll not only be without a leader, but also without a deputy leader, because Malcolm Bruce will have departed the scene by then. Presumably an interim leader will quickly "emerge", but it's difficult to imagine that person carrying much authority ahead of a leadership election that will decide whether the party returns to its radical traditions, or perseveres with the hellish Orange Book experiment. Would Labour or the Tories even know who they were negotiating with? Would a coalition deal be worth the paper it was written on?
In such circumstances, a deal involving the SNP might look like the only game in town.