Wednesday 25 February 2015

#GE2015 Forecast from UK ELECTION TREND

Some improvement for Labour's chances over last week, however that's mainly due to the ICM outlier ageing out of the recent average, and the model forgetting about older polling from last year. As we move closer to the election, the window the model considers when judging momentum reduces. The window now starts at the period where Labour's figure dipped then stabilised last September. Which means that the model has just gone from seeing Labour's momentum as gently downwards, to being somewhat static around the current average.

The Conservative momentum is almost as static, with a very very slight downwards trend. However, even the weak revert to mean being applied is still more substantial than this trend so the prediction is still a close VI result.

Looking at polling this week, it does look like this static position is very resistant to change, even after major news scandals. However, this includes an Opinium survey following methodology changes, that altered how they re-weight their figures. Opinium use a voter identification weighting, and base part of that on votes cast at the European Election. I doubt if that's entirely appropriate, as European Elections are not General Elections, and the voting patterns are not the same.

No comments:

Post a Comment

Comment is open to all feel free to link to this blog.