Archive for April, 2015

h1

Ipsos-MORI Scotland phone poll has SNP a staggering 34% ahead

Wednesday, April 29th, 2015

This could be down to “shy unionists”

With the ongoing debate about phone polls versus online ones there’s a new Ipsos Scotland survey for STV which has extraordinary figures with the SNP on its biggest level yet.

The firm is the only one that regularly carries out phone polls in Scotland and its findings this morning are remarkable.

There’s been a lot of talk about “shy responders” to phone polling with supporters of less popular positions said to be reluctant to volunteer that information in a live conversation with a pollster. Could that be happening here?

The answer is that we don’t know but this poll will become a reference point on the shy factor

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble




h1

My podcast with Keiran Pedley of Polling Matters + latest YouGov has LAB back in the lead

Wednesday, April 29th, 2015

Collage-DC-EM-NC-NF

Yesterday I recorded a podcast with the excellent Keiran Pedley (@keiranpedley), the professional pollster who runs the Polling Matters website. We covered a lot of issues that are very familiar to PB regulars including comparing betting markets to opinion polls, Lord Ashcroft and regional polling, Scotland and the future of the Liberal Democrats after the election. Also Keiran asked about how PB was set up in the first place more than eleven years ago

Meanwhile overnight we have had the daily YouGov which has reverted to a small LAB lead.

Coming up today I’m expecting a new round of ComRes battleground polling for ITV. My guess, based on the fact that my wife was called by ComRes on Sunday, is that this latest survey will focus on tight LAB-CON battlegrounds.

Tomorrow we are promised a new phone poll from Ipsos-MORI.

Mike Smithson

For 11 years viewing politics from OUTSIDE the Westminster bubble




h1

Marf for tonight on Dave “being pumped up”

Tuesday, April 28th, 2015

Cameronpumpedup



h1

If you’ve been seat betting based on the Ashcroft polls then Nuneaton could be the first test

Tuesday, April 28th, 2015

A look forward to election night

The point I’ll be waiting for on the night will be the first LAB-CON marginals to declare. These are the seats that will determine which party comes top and all that could mean in the post-election bantering.

Looking at the Press Association expected declaration timings we are not going to get much on these seats before 1am when the Nuneaton result is scheduled.

This Midlands seat is LAB target number 32 from CON in England and Wales which would change hands on a 2.32% swing. A successful CON defence would be a good pointer to them winning most seats. A loss a bad one.

Nuneaton looks set also to be the first LAB-CON marginal in England and Wales where Lord Ashcroft has carried constituency polling. His last there, carried out in March, (see above) had LAB taking it on a 5% swing. What will be interesting for me is not just the result here but what it says about the many other 100 or so seats that Lord Ashcroft has polled which have had a big impact on constituency betting markets.

If on the night the Ashcroft polling prove to be good predictors then betting prices linked to his seat polls will harden. If not then the other way.

Now I know that Lord A describes his polls as snapshots but they have had a big impact on the betting. What is important is to look at the fieldwork dates. Polling done as far back as July might be less relevant.

Mike Smithson

2004-2014: The view from OUTSIDE the Westminster bubble




h1

What we need is for individual pollsters to produce results that are completely out of character

Tuesday, April 28th, 2015

IMG_20150310_134447700_HDR~2

A Populus CON lead perhaps would speak volumes

Mark Pack makes an excellent point that if you look at movements pollster by pollster then there has actually been very little volatility.

This is one of the basic rules of polling analysis – you shouldn’t compare one firm’s poll with another and then deduce that there has been a trend.

In the run up to this election Monday and Friday morning’s, as we saw again yesterday, have generally brought reassuring news for LAB with the twice-weekly Populus/FT poll. Throughout 2015 it has not shown a CON lead and has had LAB ahead by 1-3% in all but one survey. Monday afternoons, again like yesterday, have been good for the Tories with the ICM and Ashcroft phone polls generally showing solid Tory leads.

The monthly Ipsos-MORI phone poll has during the year tended be good for LAB with the party always ahead. ComRes phone, now in the Mail, has tended to show solid Tory leads.

YouGov can move about though in recent weeks there have been few Tory leads such as we have seen overnight.

    What I’d love to see in this final period is for pollsters to produce numbers that are out of character. Populus showing CON leads in its remaining surveys would send out a stronger message about what’s happened than ICM having similar results.

I am expecting a level of convergence as we get closer and that the final polls will have most of them in the same territory.

  • Meanwhile the decline of UKIP in the latest Ashcroft marginals polling was good news for LAB and CON. The Tories are delighted that Castle Point and Great Yarmouth look safe though they’ll be less pleased by the 6.5% swing to LAB in Cannock Chase that points to a LAB gain. Great Grimsby, regarded as the best LAB target for UKIP had the purples 17% behind.
  • Mike Smithson

    2004-2014: The view from OUTSIDE the Westminster bubble




    h1

    The big unknown – the large number of polling respondents saying they don’t know

    Monday, April 27th, 2015



    h1

    CON 6% lead in latest Ashcroft national poll while in new seat polls UKIP continues to struggle

    Monday, April 27th, 2015



    h1

    ICM have 3% CON lead while Populus have a 3% LAB one. In Scotland TNS has SNP 32% ahead

    Monday, April 27th, 2015

    It’s almost no change with Populus

    TNS has SNP with 32% lead in Scotland