Euros ‘Exit’ Poll

By crunching some numbers I’ve potentially found something very interesting indeed.

Lord Ashcroft also did a Euro ‘exit’ poll of people saying they voted, full results to be published tomorrow night, but in the meantime he published some ‘issues’ tabs.
What’s interesting is the table on page 2, because it shows breaks by party and also for all voters. ‘All voters’ is of course a weighted average of each party’s voters (and ‘others’) but by applying an optimisation algorithm, we can estimate the weights used. These aren’t totally precise estimates because the published numbers will have been rounded, but the weights in the solution (x100) are:
CON 23.8
LAB 23.3
LIB 5.8
UKIP 29.3
OTH 17.8
Not sure whether these are weights are indicative of the toplines that will be published tomorrow, but if they’re in any way indicative, I’d say they look good for UKIP, as expected for CON, bad for LAB, shocking for LIB and probably very good for GRN (Others ex-GRN have averaged about 6.3 in the last week by my reckoning, so if that played out, GRN weight would be around 11.5).
On this basis, a UNS would give these seats:
CON 18
LAB 19
LIB 0
UKIP 24
GRN 6

Make of this what you will, I haven’t had it peer reviewed! If you want to check the numbers it can easily be done in Excel.

 





About The Author

Related Posts