As I noted in my blogpost this morning, a consensus is emerging among the commentariat that the Tories are coasting to victory, and that some opinion polls bear little or no resemblance to what is really happening on the ground. One popular theory is that polls are picking up too many Labour supporters because those people are more enthused, and thus keener to be interviewed. I certainly don't dismiss that consensus out of hand - it may well be entirely right. But when you see an enormous survey of 11,000 people, seemingly bang up-to-date, conducted by a firm that actually didn't perform at all badly in the 2015 election, and it points to the Tories losing their majority, you at least have to take a step back and consider the possibility that there may be two sides to this story.
GB-wide voting intentions (SurveyMonkey) :
Conservatives 42% (-2)
Labour 38% (n/c)
(Other parties' vote shares are not available yet.)
To put SurveyMonkey's limited track-record into perspective, in their final 2015 poll they put the Tories on 34% and Labour on 28%, which meant they slightly underestimated both parties but got the Tory lead almost exactly right. It's possible they may have just got freakishly lucky in that individual poll due to sampling variation, but it has to be said there were hardly any other polls that were so close to being accurate (with the famous unpublished Survation poll being an obvious exception).
* * *
You might be interested in Alasdair Soussi's article on the Al Jazeera website about the battle for Scottish seats at the general election - it features quotes from myself, Simon Pia, Ian Duncan and Professor James Mitchell. You can read it HERE.