h1

It can be argued that the flawed polls are those that don’t name candidates

March 11th, 2015

IMG_1179~2 (1)

Why the LDs are releasing some of their private polling

Yesterday I received the full media briefing on the controversial private LD polling which has attracted a lot of attention. I was able to ask about any seat and have a pretty good picture of how things are looking.

The reason for the part disclosure is to make a very simple point that naming the candidates can make a huge difference – as seen in the one poll that was published on Lynne Featherstone’s fight to beat off LAB in Wood Green & Hornsey.

She won the seat in 2005 and has built up a powerful personal support base that is directly linked to what she has done on the ground. Effectively she has become the brand not her party.

The Ashcroft polling of the seat was about parties even with the second voting question asking people to think of their own constituencies and the candidates who might stand. It showed her some way off. The LD poll in which candidates are named, had the gap at just 1%.

The fieldwork and tabulation was carried out by Survation but the poll designed by party. As has been pointed out the voting question was not, as is the norm, put first and that can have an impact. Also it was weighted back fully to 2010 without a misremembering adjustment.

What I came away with is that the polling is designed almost solely to help the party decide where to put its resources in the final stages of the campaign. Some current LD held seats are going to be given lower priority as s result of the data. Others are going to be given boosts.

The best guide to what the polling overall is showing is in the betting prices in each constituency. Clearly the data filters out to party workers many of whom are having a punt.

My conclusion is that where the LDs are tighter than 5/2 the polling has shown that they are in with a shout.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble