This is an interesting post, but I’m wary of using historical trends as a guide given that we are in a coalition scenario, and that Labour’s small lead has come almost entirely through switchers from the junior coalition partner. Breakdowns suggest there have so far been few if any direct Tory-Labour switchers.

Because most of the switchers have been LD to Labour, they represent a different ‘swing electorate’ than we have seen in previous mid-term polling. While LDs have dipped in previous parliaments, their fall has not been so conspicuously to the benefit of one party as it is now. So the behaviour of this ‘swing electorate’ is difficult to forecast by past precedent.

I’d be interested to know your thoughts on this and how it affects your model.

r-squared 0.4?

Not the strongest result.

Still if that encourages you to stick with Milibean then great!

Oh, spare us. I’m not disputing the methodology but the idea you can predict the next election on a line of best fit based on how far “Weath” was ahead of “Hilson” in polls through the 60s and 70s is nonsense, not least because we’re now in coalition and Labour has probably already drained its potential store of Lib Dem tactical voters by now.

I can quite easily point at a similar bunch of polls at the analogous stage of the 1979 Parliament (which would be January 1981) that “prove” Michael Foot is going to sweep into power. In fact on UKPR I can see one (NOP Feb 13 1981) that shows Labour on 51% with a 16% lead! Even into late summer 1981 Labour were routinely getting 6-8% leads. Proof Labour were going to win in 1983, then.

The situation is exactly analogous – a weak, awkward, left-wing leader of a party talking to itself. LibCon can post as many articles as it likes saying all is hunky-dory on the Good Ship Milipede, it makes no difference to the fact that the British people are not in the mood for his (or your) brand of leftie student-union hackery. Can’t afford it.

As the link at the bottom of the post isn’t working, the full article and detailed methodology are available here: http://bit.ly/xtufCw

Ahem. “Past performance is no indicator of future success”.

I’ve long said that Labour should have a larger lead, but its clear that the next 12-18 months is crucial for Labour, and of course Ed.

Today Miliband has stated this is New labour, no not that New labour but a Newer type of New labour, he says we must accept austerity, we understand we will not have money any more, but we will cut.

If your a hard working member of society then labour is the party for you.

So basically if your a teacher or nurse or public service working , then labours your party.

Had your legs blown off in one of Labours wars, remember this is not that labour party, Miliband was not involved in that, oh yes you were mate.

problem is when you see MIliband talking , walking or standing speaking it’s simple your not leadership material…. good bye

The fact is Miliband is not a leader, well he maybe but he does not look it, and sadly Cameron does, and to be honest voting labour after todays report from Miliband about hard working people, sorry he’s not got any understanding at all.

@1 The problem with assuming that Tory-Labour switchers is the key factor is that it ignores differential turnout.

If Labour needs to build a double digit lead in the next year the real question is whether that is likely. Based on current polling trends with a Tory rise I’d suggest it isn’t.

The Polls for this Parliament are different this time because of the simple fact that the Tories did not win, infact with 36% of the vote they were nearer to the share of the vote obtained by Hague and Howard than Major in 92, so at 36% they are close to their core vote

The only real movement since then has been the anti-Tory voters who have left the Lib Dems due to the coalition. The Tory core vote appear happy with the Government.

Normally you would expect the Government to lose some of the swing voters they attracted during the campaign and these voters would either return as happened to the Tories in the 80s and 92 or not as happened in 97.

Unless something spectacular happens we will go into the next election with 36% Tory and The rest divided between Labour and Lib Dems (and Natioanlists). The unknowns will be
• Will the anti-Tory voters believe the Lib Dems and return in any numbers, and will there be any tactical Lib Dem/ Tory voting (In Scotland ill the Lib Dems be replaced by the SNP?)
• If the Lib Dems remain at 10% how will that break down in seats – will it boost the Tories more than Labour
• Will the Tories regain the 2-3% of UKIP voters
• What difference will the new boundaries make (if they are passed)

The Polls for this Parliament are different this time because of the simple fact that the Tories did not win, infact with 36% of the vote they were nearer to the share of the vote obtained by Hague and Howard than Major in 92, so at 36% they are close to their core vote

The only real movement since then has been the anti-Tory voters who have left the Lib Dems due to the coalition. The Tory core vote appear happy with the Government.

Normally you would expect the Government to lose some of the swing voters they attracted during the campaign and these voters would either return as happened to the Tories in the 80s and 92 or not as happened in 97. Which is why we assume the opposition lead needs to be much higher than it is at the minute.

Unless something spectacular happens we will go into the next election with 36% Tory and The rest divided between Labour and Lib Dems (and Natioanlists). The unknowns will be
• Will the anti-Tory voters believe the Lib Dems and return in any numbers, and will there be any tactical Lib Dem/ Tory voting (In Scotland ill the Lib Dems be replaced by the SNP?)
• If the Lib Dems remain at 10% how will that break down in seats – will it boost the Tories more than Labour
• Will the Tories regain the 2-3% of UKIP voters
• What difference will the new boundaries make (if they are passed)

My guess is Labour will win next time, but they may need a coalition with the SNP and other nationalists

As fun and yet pointless as bungee jumping

weeeeeeeee

How do you deal with the fact that polling in the 1980s and much of the 1990s was frankly crap? Shy Tory effect and all the rest.

Surely that means that some of your data points are likely to be wrong, since strong Labour polling results in the 1980s were never backed up by election results. This would obviously effect the lead you ‘need’ to win an election from this point.

ARe you using straight Averages. If you are then its not possible to compare Average polls now with those in the past because of the way current polling is dominated by Yougov. That wouldnt matter so much if Yougov werent at one end of the scale, consistently giving Labour the highest scores.

The regression curves have a whopping amount of variance to it. As a result, the figures on both charts are not sufficient to draw any reliable conclusion.

Much as I am attracted to the idea that “Labour is doing better than commentators think”, I don’t think we can conclude this.

In the interests of fairness, a chart showing how commentators think the opposition is doing vs how they then perform (I conjecture) is also likely to have a wide range of variance too:

What the commentators say is itself not worth drawing conclusions from.

8 9 well said