The constant flow of GB-wide surveys is giving a distorted picture of what is going on
Just about every day at the moment I find myself having to Tweet or write on PB that general elections are not decided by national party vote shares but by first past the post elections in 650 separate constituencies.
This has never been the case more so than in what has for decades been regarded as Labourâ€™s most important bastion – Scotland where 59 of those constituencies are.
The demographic data suggests that Scotland represents just 8.68% of the total number of voters in a GB-wide sample. So if LAB, and this is being generous, has, lost to the SNP a third of those who voted for the party in May 2010 that accounts for just 1.2% of the GB-wide vote share.
Yet in terms of seats that fall of 1.2% in GB vote share could be catastrophic for the red team maybe reducing its current contingent of 41 Scottish seats to fewer than 10.
The surge in SNP support could be even more disastrous for the LDs who account for 11 of Scotland’s 59 MPs.
Yet in spite of these huge developments there have been just three Scotland only three Westminster voting polls since the week of the IndyRef in mid-September. What could be totally re-shaping the UK political map is based on the views of just three thousand Scottish voters.
We urgently need more Scotland-wide polls and we also need single constituency surveys to test whether some of the SNP surge is being tempered by the impact of incumbency and maybe anti-SNP tactical voting.
We are promised some single seat Scottish surveys from Lord Ashcroft but this urgently needs to be supplemented by frequent Scotland-wide polls.
In the meantime we get at least 8 GB polls a week. Cannot just part of that effort be diverted to Scotland?
That is my plea.