Guardian Utd (Wednesday): Senior Tories prepare for defeat

From the Guardian Unlimited (UK)
Dated Wednesday May 4

Senior Tories prepare for defeat
By Nicholas Watt

Senior Tories are preparing for an election postmortem amid growing fears that the party is heading for another heavy defeat.

Jitters were compounded when news spread of today's Times Populus poll, which put the Tories on a historic low of 27%, and frontbenchers hit the phones to discuss tactics when polls close tomorrow night.

MPs on the modernising wing of the party will target the "dog whistle" strategy devised by the Tories' Australian campaign director, Lynton Crosby, which was meant to reach out to core voters without alienating those on the middle ground . . . .

Tories on the right will take the opposite view and declare that Michael Howard has failed to do enough to shore up the core vote by offering only modest tax cuts of £4bn. There is close interest in the movements of John Redwood, the rightwing shadow cabinet minister, who is said to be enraged that he has been silenced by the leadership after he described the £4bn tax cuts as a "downpayment".

Read more.

As quoted by the BBC

YouGov (taken for Sky News):

Labour: 36%
Conservatives: 32%
Liberal Democrats: 25%
Others: 7%

Populus (taken for the Times of London):

Labour: 42%
Conservatives: 29%
Liberal Democrats: 21%
Others: 8%

The Populus Poll has been showing a steady widening of Labour's lead for the last two weeks.

Labour: 38%
Conservative: 34%
Liberal Democrats: 22%
Others: 6%

Harris is an internet pollster like YouGov, so on one hand it is the internet pollsters (YouGov, Harris and BPIX) predicting a small-ish Labour lead, and the phone and face-to-face pollsters predicting a larger lead. Someone is going to have eggs on their faces...

The Populus poll is a tracker poll (the last 1/4 sample for it was yesterday) so is not entirely reliable. They are doing a standard full 1,000 sample poll for tommorow's Times of London, which is their eve-of-poll and their forecast for the result. Lots of other pollsters are doing the same.

I'm always more inclined to rely on face-to-face/telephone methods. The sample is more likely to be random

The MORI poll, which is face-to-face/telephone, intrigues me. When it was published yesterday, it showed a sudden upswing to Labour with the figures in line with Populus poll published today. Yet Populus simply shows a steady trend for two weeks toward Labour and away from the Tories.

because it's a 'tracker poll' (350 people polled each day, showing 4 days results at a time, ie 1400 people).

Labour 41% (no change since yesterday)
Con 27% (-2%)
Lib Dem 23% (+2%)
Other 9% (no change)

http://www.timesonline.co.uk/section/0,,21169,00.html

4 days ago, the Tories were at 31%.

Could that indicate a major re-alignment of British politics?

...the Lib/Dems as the official opposition and the Tories consigned to minority status.....

It could be very interesting...there will be a lot of strategic voting going on (for those that will bother to vote) and at this point the ONLY hope for the Tory party is that enough dis-affected Labour voters either stay home or vote Lib/Dem as a form of protest and the Tory candidates get in by virtue of dis-interest....

To a large extent those folks that do vote, pay attention to the issues and are well aware of the power of their individual votes in key constituencies...My father is one that is caught between a rock and a hard place, we will NOT vote Tory no matter how much Blair has pissed him off, he would like to vote Lib/Dem but understands that a vote that way could actually benefit the Tory candidate because he knows that there are alot of people that will not vote Labour because of Tony Blair....He is completely torn....

However, it is that the Liberal Democrats will pick up seats at the Tories' expense.

I've been playing a little with the BBC seat calculator. The tighter polls will give the the Tories a pick up of about 20 seats and the LibDems between 5 and 10; the ones that have Labour pulling away show a shift of a handfull of seats from the Tories to the LibDems. For example at 40/30/20/10 for Labour/Tories/LibDems/Other, the Tories lose 5 seats with LibDems picking up 4 of them and Labour the other.

"Weighting: Data is weighted to the profile of all adults aged 18+ (including non telephone owning households).

The data is further weighted on the basis of past vote.

Figures are adjusted for turnout on the basis of respondents' declared likelihood of voting.

In a further step Populus takes 60% of those who say they are going to vote at the next election, but who also refuse to answer the vote intention question, or say they don't know how they'll vote, and allocates them to the party they voted for in 2001."

http://news.bbc.co.uk/1/hi/uk_politics/vote_2005/basics/4275273.stm
Taken from the section on Populus.

And it seems to me that it may introduce errors rather than eliminate them. "The data is further weighted on the basis of past vote."

Surely this is going to favour Lab too much, and the Tories to a lesser extent.

In statistics, one might give a chi test to parts of a sample to determine if weighting would be justified.

For example, if one examines any election polling data in the United States, there is a correlation between being Afro-American and voting Democratic. Consequently, a survey might weight the Afro-American response in the sample to fit the percentage of Afro-Americans in the overall population. If 30% of the respondents in a voter preference survey are Afro-American, but Afro-Americans were only 12% of the voting population, then one would expect a sample like that for the data to be biased toward Democrats. The statisticain would weight the sample accordingly so as not to skew the data.

but I am pointing out that an awful lot of weighting is being done, and we, at least I, don't know how much weight is being given to certain factors.

The fact that there is such a large discrepancy between polls supposedly taken at the the same time by different organisations surely means that some of these weights are either over, or under done. (And that's before we get into the standard +- 3% debate.)

To take your Afro-American example, yes it makes sense unless a particular factor mitigates against it eg Dem Presidential candidate is a racist (I know it wouldn't happen, just an illustration).

If these polls are taking people who say they won't vote, or who won't declare a preference and marking them down as the party they voted for last, then this favours Labour (as they did extremely well last time out).

But those votes will probably not materialise in real life and if they do they may go elsewhere, because if they are lab votes, they are very soft.

I think it's more to do with Labour policies making their supporters stay at home...

http://ellissharp.blogspot.com/2005/05/35-reasons-for-not-voting-labour.html

They're hoping everyone else stays home.

The Guardian helping to keep Labour voters home?

That's as likely as FoxNews reporting "facts" desinged to discourage wing morons from voting.

Labour party is sending.

If it was a different message, OK.

But the opposite message?

1. The Guardian, while its editorial policy over the years has been to support Labour, is still a newspaper and an excellent one at that. They wouldn't be saying that the Tory leadership is extremely concerned about another lackluster showing unless the Tory leadership were concerned.

2. Labour is a political party. Labour's leaders want to get their voters to turn out. So their message to their supporters is not to get overconfident. We heard the same thing from President Reagan's campaign just before the 1984 election, even though there wasn't a snowball's chance in Death Valley that Mondale would pull an upset.

The British press is a far less homogenous group than its American mainstream counterpart. I really don't know where you are getting the idea that the whole British press is engaged in some sort conspiracy to drive Blair's vote down.

but then they did in 1992 when John Major confounded the pundits by winning. Historically, the Conservative party have tended to perform slightly better at the ballot box than the opinion polls suggest. Nonetheless, I suspect that both the Tories and Liberal Democrats realise that the best they can achieve is to dent Blair's majority.

You have to love the British system of 'democracy' where a party with 37% of the popular vote gets twice as many seats as a party polling just 4% less.

One prediction I will confidently make is that whichever party wins the current General Election in the UK will go on to lose the next one badly .

Things look bad for the Tories, but then they did in 1992 when John Major confounded the pundits by winning. Historically, the Conservative party have tended to perform slightly better at the ballot box than the opinion polls suggest. Nonetheless, I suspect that both the Tories and Liberal Democrats realise that the best they can achieve is to dent Blair's majority.

The Tories were declining before 1992 after many years of electoral success under Margaret Thatcher. Labour was at least within striking distance of the Tories by that time.

The Tories would have to pick up almost as many seats as they currently hold in order to win a solid majority in Commons. To do that, they would have to reverse the figures against Labour in what are the more pessimistic (for them) polls quoted in post 14. That's not going to happen.

You have to love the British system of 'democracy' where a party with 37% of the popular vote gets twice as many seats as a party polling just 4% less.

That is good argument for proportional representation and preferential voting.

One prediction I will confidently make is that whichever party wins the current General Election in the UK will go on to lose the next one badly.

Labour will be led in the next election by Prime Minister Gordon Brown. The only reason they're having a bad time now is Blair: he isn't trusted, and for good reason.

The change of personalities will almost certainly help Labour next time (2009 or 2010); what other circumstances their might be will be a bigger factor.

having nothing to do with politics and everything to do with my belief that the world economy is heading for a nasty and long recession some time between 2006 and 2008. Whoever is in power is likely to get the blame. It does not really matter whether they are Labour, Conservative or Liberal Democrat. Of course if the Peak Oil pessimists are right we might not have any elections at all.