A Quick Look At The Polls

by Jackie_South on September 1, 2011

It’s been a while since we last looked at the polls here at AllThatsLeft. 

Probably the best place to start is UK Polling Report’s polling average, taken across all the polls over the last fortnight.  As a way of trying to smooth out blips in individual polls, it is as good a method as any, although it does have some problems: firstly as it is over a fortnight it tends to lag behind recent shifts in opinion. 

Secondly it is biased towards the more frequent polls: of the 8 polls used, 4 are YouGov polls due to their greater frequency.  YouGov has tended to show the lowest poll numbers for the Lib Dems and the highest for Labour, whereas ICM’s polls for The Guardianhave had the opposite tendency.

But anyway, here’s that poll average:

One thing that is worth noting is that this average does not seem to bear out the assertion made in The Guardianthis week that the Lib Dems have mounted a recovery in their share of the vote.  This was shown in two polls, an Ipsos Mori poll on 22 August which put them on 15% and an ICM poll (see my comment above) on 21 August, which put them on 17%.  These polls haven’t been supported by later polls.  A cynic might conclude that The Guardian is still working through its rash infatuation from last year with the party.

Projecting this against the results in each seat from the last General Election, this translates to the following:

So, it would give Labour an outright but relatively slender majority of 18. 

Interestingly, although these numbers are a bit better for the Lib Dems than the 9% they were getting back in March, it would mean them losing more seats – the March polls had them winning 18 seats.  The reason for this is that they are more vulnerable to the Tories and the Tory recovery since then hurts them more than the slight improvement in their position against Labour.

In total, these polls would have the following shifts:

  • Labour: +76 (55 from C, 18 from LD, 1 each from Plaid Cymru, SNP and Green Party)
  • Conservative: -32 (55 lost to Labour, compensated by 23 won from the LDs)
  • Lib Dems: -41 (18 to Labour, 23 to Conservatives)

A huge caveat on all this is the impact of the boundary changes, of which we are due to get our first sight next month, which gets rid of 50 seats.  The London School of Economics projected that these could cost Labour 18 seats, the Tories 15 and the Lib Dems 14, although presumably some of the latter would be the same that they are projected to lose to other parties.

The seats that I project would change hands are as follows:

Labour wins from Conservatives:

  • Amber Valley
  • Bedford
  • Blackpool North & Cleveleys
  • Brentford & Isleworth
  • Brighton Kemptown
  • Broxtowe
  • Bury North
  • Cannock Chase
  • Cardiff North
  • Carlisle
  • Carmarthen West & Pembroke South
  • City of Chester
  • Corby
  • Croydon Central
  • Dewsbury
  • Dumfriesshire, Clydesdale & Tweeddale
  • Ealing Central & Acton
  • Elmet & Rothwell
  • Enfield North
  • Erewash
  • Gloucester
  • Halesowen & Rowley Regis
  • Harrow East
  • Hastings & Rye
  • Hendon
  • Hove
  • Ipswich
  • Keighley
  • Kingswood
  • Lancaster & Fleetwood
  • Lincoln
  • Loughborough
  • Morecambe & Lunesdale
  • Northampton North
  • Norwich North
  • Nuneaton
  • Pendle
  • Plymouth Sutton & Devonport
  • Pudsey
  • Sherwood
  • Stevenage
  • Stockton South
  • Stroud
  • Swindon South
  • Thurrock
  • Vale of Glamorgan
  • Warrington South
  • Warwick & Leamington
  • Warwickshire North
  • Watford
  • Waveney
  • Weaver Vale
  • Wirral West
  • Wolverhampton South West
  • Worcester

Labour wins from Lib Dems:

  • Argyll & Bute
  • Bermondsey & Old Southwark
  • Birmingham Yardley
  • Bradford East
  • Brent Central
  • Bristol West
  • Burnley
  • Caithness, Sutherland & Ross
  • Cambridge
  • Cardiff Central
  • Dunbartonshire East
  • Edinburgh West
  • Gordon
  • Hornsey & Wood Green
  • Inverness, Nairn, Badenoch & Lochaber
  • Manchester Withington
  • Norwich South
  • Redcar

Labour wins from other parties:

  • Arfon (from Plaid Cymru)
  • Brighton Pavilion (from Green Party)
  • Dundee East (from SNP)

Conservative wins from Lib Dems:

  • Aberdeenshire West & Kincardine
  • Berwick-upon-Tweed
  • Berwickshire, Roxburgh & Selkirk
  • Brecon & Radnorshire
  • Carshalton & Wallington
  • Cheadle
  • Cheltenham
  • Chippenham
  • Cornwall North
  • Devon North
  • Dorset Mid & Poole North
  • Eastbourne
  • Eastleigh
  • Kingston & Surbiton
  • Portsmouth South
  • St Austell & Newquay
  • St Ives
  • Solihull
  • Somerton & Frome
  • Sutton & Cheam
  • Taunton Deane
  • Torbay
  • Wells

This would leave the Lib Dems with only 3 seats in their South West homeland (Bath, Thornbury & Yate and Yeovil) and one each in London (Twickenham), Wales (Ceredigion) and the South East (Lewes).

{ 1 comment… read it below or add one }

Mike Killingworth September 2, 2011 at 11:28 am

Mike Smithson argues, on the back of the referendum and Scottish Parliamentary results, that ICM is a far far better pollster than YouGov. And as he puts his money where his mouth is, I’m inclined to accept his view.

Lists of seats that might or might not change hands on the back of a poll or even a poll of polls are amusing merely – not only do we have the Boundaries Review, and not only is that Review fraught with potential legal and parliamentary thornbushes, but there are two or three other reasons:

- the incumbency factor, where the sitting MP is either retiring, very popular or a complete plonker (remember Lembit opik?);
- regional variation;
- IIRC while the polls did a good job last time of predicting the parties’ overall vote share, the variations within that from seat to seat were greater than ever before. And with the next election requiring the voters to deliver a verdict on a peacetime coalition government – which will be a new experience for all but centenarians – I predict that the results next time will be more even more variegated.


Leave a Comment

Previous post:

Next post: