politicalbetting.com

Remember what happened here in the run up to the 1979 election?

With suggestions going the rounds that that Geoff Hoon might replace Peter Mandelson as Britain’s EU commissioner in Brussels should we be preparing ourselves for a by election in his constituency at Ashfield in Nottinghamshire some time next year?

The 2005 result is reproduced above and certainly with current polling and by election performances the Tories would fancy their chances. In fact a failure to win here would not bode well for the general election.

There would also be a uncomfortable historical parallel for Labour. Back in 1977, just two years before Mrs Thatcher came to power in the 1979 general election, there was a by election in the seat that was caused by the sitting Labour MP being given a Brussels job. The Tories took it on a 20% swing.

This time Cameron’s party would require just a 12% swing to scoop up the seat which is well below the 17% that was achieved in Crewe & Nantwich three months ago. However you look at it the omens don’t look good for Labour.

Hoon - who went to the same £10,000 a year public school as Ed Balls - is no stranger to Brussels and was an MEP for ten years.

Maybe letting Peter Mandelson have another five years would be a better option but given what we know of Brown’s relationship with Mandy that probably won’t happen.

Mike Smithson

Why Labour should always assume the worst from the polls

This is Deborah Mattinson - Labour’s polling advisor who was part of conference session at the weekend on the polls trying to make the case that Cameron has “..yet to seal the deal”.

I don’t know about that but the lesson from polling history is that Labour’s default position should be that the most accurate poll is the one with the party in the least favourable position - what I term “The Golden Rule”.

Remember Glasgow East in July. Two polls showing Labour leads of 14% and 17%. Labour lost.

Remember Crewe & Nantwich in May. Two pollsters did surveys showing that Labour would lose - but in each case the margin predicted was lower than the outcome.

Remember the London Mayoral Election. Four pollsters carried out surveys and three suggested that it was neck and neck between Labour’s Ken and Boris. The fourth had Boris ahead by what turned out to be the precise margin of victory.

Remember the Blaenau Gwent by-election in June 2006. The only poll had Labour regaining the seat with a margin of 12%. The independent, Dai Davies, won by a margin of 9%

Remember the 2005 general election. All the pollsters bar one had Labour with a bigger vote lead than was actually achieved. The one exception, NOP for the Independent, got it precisely right and then got dropped by the paper.

Remember the Hartlepool by election in September 2004. The one poll during the campaign had Labour a whopping 33% ahead. The eventual winning margin was 6.5%.

Remember the 2004 London Mayoral race. Two pollsters did surveys - the one with Labour in the least favourable position got it almost precisely right.

Remember the 2004 Euro elections. Two firms did firms both of them overstating Labour’s eventual position.

Remember the 2001 general election. Labour won with a 9.3% lead on votes. None of the pollsters had this in single figures and one campaign poll had the party 30% ahead.

I could go on but the message is the same. Anybody advising Labour on polling has to make the default position that the survey showing the party doing the worst has proved to be the most accurate.

The next occasion where this will be put to the test again is the Glenrothes by election. A weekend poll had Labour and the SNP on 43% each. My money will be against Labour.

Mike Smithson

But will the polls be as wrong as they were in Glasgow East?

A night of four polls all doing very different things has been finalised with an ICM survey in Glenrothes, scene of the upcoming by-election, which suggests that Labour is in for a hard fight.

In the seat that adjoins Brown’s and where the party had a 10,600 majority at the general election ICM found that both Labour and the SNP were on 43% each. That’s fine except as the paper reminds us a similar poll by ICM ahead of the Glasgow East by election had Labour 14% ahead. The only other poll in that election was reporting a Labour victory by 17%

But the numbers that are highlighted most by the Mail on Sunday on its by-election poll is the response to this: “If the Glenrothes by-election was a vote of confidence on Gordon Brown how would you vote”. The response: SNP 44%: LAB 41%: LD 8%: CON 5%

In another poll a YouGov study of 1,200 Labour members, suggests that a majority (53%) view the PM as “indecisive and dithering” while 34% thought he has “an exciting vision for the future”.

About 40% of the members want immediate leadership contest, against 52% who are opposed. Among potential successors, David Miliband is clear favourite - backed by almost a quarter, ahead of Health Secretary Alan Johnson and Justice Secretary Jack Straw on 13%.

Politicians always look for the good news and there’s little doubt that the ComRes poll showing Labour’s deficit is down to 12% will give real cheer to delegates at the party conference in Manchester. They will seek to dismiss the mega PH-YouGov marginal seats survey with a 34,000 sample because the field-work took place in July. That suggested and reported seat-by-seat a Tory majority of 146.

As I noted yesterday polling at conference time should be treated with care. The broadcasters are under an obligation to give extended coverage and this usually produces a boost like the one that the Lib Dems have enjoyed. We really do need to wait until the end of October.

Whatever - polling which had got dead boring because the findings were almost all the same has got interesting again. Watch out next week for Brown’s conference bounce which will probably be eclipsed the week after by Cameron’s.

Mike Smithson

Does Brown really have a chance of saving the day?

There’s an interesting piece by Kevin Maguire of the Mirror quoting an as yet to be published article by Peter Kellner of YouGov suggesting that Labour might still have a chance.

The arguments are ones we have seen before - governments, it is said recover from mid-term lows and we all know that the Tories need a vote margin of 6% just to be equal on seats and a 10% gap for a majority.

Where I take issue with the two of them is on the lesson from the Tory victory in Crewe and Nantwich last May. Kellner points to the massive loss of Tory votes at Dudley West in 1994 which was on a far bigger scale than Labour’s loss in Crewe.

But where they are all making a huge mistake is judging by elections by percentage movements which can be distorted by very low turnouts. That happened in the 1994 case and Labour won its great victory even though it lost actual votes on its general election total two years earlier.

What was dramatic about Crewe was that turnout was nearly at general election levels and that the Tory vote in actual numbers rose from 14,162 in 2005 to 20,539. There has been no real precedent for this.

At Glasgow East in July the same thing happened to the SNP vote. It rose from 5,268 at the general election to 11,277 in the by election. Like Crewe turnout was not too far off the general election level.

These two results very much support the huge swing in the polling that we have seen since the March budget.

Yes Labour can recover and if they can get to 30% or more then it could be really interesting. But we are, I believe, heading for a general election where turnout is considerably up and based on the by elections the “new” voters will be coming out to oppose Labour.

Mike Smithson