If the trend in the ComRes marginals’ survey is seen in Lord Ashcroft’s weekend mega-poll then LAB is in very serious trouble

May 21st, 2014

CON doing far better in the marginals than GB as a whole

The main reason why we have marginals polling at all is to find out whether what is happening in the key seats is different from the country as a whole.

At a general election voters in these constituencies have a different experience from those elsewhere with far less intensive campaigning and where there has been little activity in the years leading up to the election.

The chart above shows the key numbers from today’s ComRes Battlebus poll for the Indy. As can be seen the latest ComRes Indy phone poll pointed to a 6% CON to LAB swing. In the separate battleground poll which was carried out at about the same time there was a swing of just 1%.

These were the figures from the poll with changes on what happened in the 40 seats in the survey at GE2010: CON 33%-4, LAB 35%-2, LD 8%-10, UKIP 17%+14

In 25 of the seats polled the Tories were the incumbents while in the other 15 LAB was. On the figures in the poll no seats would change hand.

My main caveat is the overall sample size of 1,030. The Lord Ashcroft mega marginals poll to be published on Saturday is likely to be based on a sample ten to fifteen times that. Also the Ashcroft marginals asks a two stage voting question – the latter seeking to ask respondees to focus on their own specific seat.

We are promised more Battlebus polls in the run-up to GE2015.

Mike Smithson

2004-2014: The view from OUTSIDE the Westminster bubble