UK Opinion Polls Thread

[Filuwaúrdjan]

When the Labour percentage goes up and down it's usually turnout (ie; within the various parts of society prone to Labour voting, rather than society as a whole) related. At least when we're in an environment in which the swingers have already swung firmly away anyway (which we are right now. Make no mistakes about that).

---

Slightly amusing that of all the polls published over the past few weeks, days even, the one to get the most attention is the one by the company with a very bad record in recent years and which is currently using a largely untested methodology (and much of the bits that aren't untested are a little dubious and tend to greatly magnify swings) Life is funny.

[Filuwaúrdjan]

Would you like to be the first to suggest that the poll isn't in the general areas of the public mood?

Not quite sure what you mean.

I'm broadly (and increasingly) sceptical of the value of opinion polls, though I've no doubt that they're correct as to who the winner of the next election will be. Besides, like I said...

You must be logged in to read this quote.

Don't mistake me for someone with me head in the sand loudly yelling "lalalalalaeverythingisfiiiiiiiiiiine". Srsly.

[Filuwaúrdjan]

Something that should be pointed out (as it is always, always forgotten) is that when there is a sizeable swing, it tends to be a lot higher in marginal seats, thus it's always easier to win a majority than a uniform swing suggests.

No, that's not strictly true. Swings in socially polarised marginals can often be relatively low (this contrasts with low turnout elections of course, where swings in socially polarised wards or constituencies are typically unusually high). It's more that swings tend to be highest where there are big concentrations of swing voters.

[Filuwaúrdjan]

Been half-expecting this for a while. Oh well. Things are clearly very volatile and unpredictable at the moment and little should be trusted much, but awful ratings for us are now the constant factor. If things are anything like this bad on Thursday, Brown probably goes, one way or another and would become the principle casualty of the expenses mess; which on one level would be monstrously unfair, but then politics isn't even slightly fair (which he knows and has benefited from in the past). The summer would then be filled up by a leadership election (they take a while and I don't think a coronation is all that likely, unless someone popular with activists, the PLP and the main Union bosses has been quietly plotting away for a while now) before the new leader has little choice to call a snap election, resulting in a Tory majority of uncertain size.

As for Labour not having the political will to do what might have to be done, well, when given the option of a possible return to the 1930's or a possible return to the 1980's, the less awful choice is likely to triumph. I don't put much stock in the rumours about a pact with the LibDems, but if that does happen it'd likely the the proverbial camel-back-breaking-straw. I reserve the right to be totally and utterly wrong.

[Filuwaúrdjan]

that aside however, it's the Labour Party that is damaged.

Regrettably, this is true. People do not expect Labour figures to have snouts-in-trough in such as way, and this is why we have been so badly hurt by this. The thing is, Wilson was right. The Labour Party is a moral crusade or it is nothing.

You must be logged in to read this quote.

...which would mean the continued existence of a big PLP.

You must be logged in to read this quote.

That ship sailed a long time ago...

[Filuwaúrdjan]

Kellner calls it: http://today.yougov.co.uk/commentaries/peter-kellner/honeymoon-over

[Filuwaúrdjan]

Who knows? Midterm poll ratings for government parties can often be laughably - and unrealistically - low. As we all saw in the last parliament wrt Labour. I would presume that the nature of the coalition would make removing Clegg under such a situation pretty difficult; if you were polling in single digits would you do something that would greatly increase the chance of a snap election that your party would be blamed for? But that's hypothetical upon hypothetical.

[Filuwaúrdjan]

Are they still doing daily polls? Weird.

[Filuwaúrdjan]

Didn't we fall below 20% at the height of the expenses mania? I don't have an especially good memory of that period, as I was... um... not in the best of health at the time.

By which I mean, opinion polls can show all sorts of strange things midterm. Really low ratings are often the result of depressed supporters claiming that they won't vote, rather than a genuine shift in partisan affiliation, fwiw.

Edit: back in the late 80s (the lowest point in Centre support in opinion polling since the 1950s - and maybe even including the 1950s - I think) the LibDems (SLD then) often polled around 5% or so.

[Filuwaúrdjan]

It's meaningless, really, but I smiled a little. Hah.

[Filuwaúrdjan]

Probably not much different, not that it really matters. But 39 = 9/10 points higher than in the General Election, or only slightly less support than in 2001. Most of the PLP would have killed for figures like those for most of the past five years.

The government isn't unpopular yet (neither is it popular anymore, of course. We're in that strange inbetween stage that can last for anything between weeks and years) so there's probably not much room for grossly inflated figures yet. You'll only see those once Tories start telling/lying the pollsters that they won't vote.

[Filuwaúrdjan]

ICM/Grauniad: Labour 37, Tories 35, LDems 18

First ICM Labour lead for three years, of course.

[Filuwaúrdjan]

If the past five years have taught us anything it's that there is never a fuller or clear picture

[Filuwaúrdjan]

I can't believe the latter, however amusing it would be.

[Filuwaúrdjan]

Fwiw, GE results...

Sheffield Hallam: LDem 53.4, Con 23.5, Labour 16.1, UKIP 2.3, Green 1.8, Others 2.8
Eastleigh: LDem 46.5, Con 39.3, Labour 9.6, UKIP 3.6, Others 0.9

The nastiest polling for the LibDems right now has them hovering around half of what they polled in the General Election. Given that, the Eastleigh figures aren't so strange for a poll that obviously works on worst-assumptions; it's also worth pointing out that a sizeable residual Labour vote that had previously avoided tactical voting more than halved in May (down from around 20% to 9%). But the Hallam figures look like trolling; even though boundary changes mean that it isn't quite the Labour dead-zone it was before and even though Cleggmania and the post-97 tactical vote are no more, such a high Labour vote in a constituency that includes Dore & Totley ward is very unlikely under any circumstances.

[Filuwaúrdjan]

As someone who lives there, I think the Hallam one is quite plausible, actually.

Really? Wow.

You must be logged in to read this quote.

A few are done at every election and they're usually pretty dreadful. This time round I'm only aware of some being done for the Brighton seats.

You must be logged in to read this quote.

Yes.

[Filuwaúrdjan]

Fresh elections are not in the interests of the LibDems (and probably won't be for years). Removing Clegg = fresh elections suddenly much more likely. Which means that Clegg is safe enough. There was far less trouble at their conference than many were suspecting/hoping for. We're more likely to see a split than to see the left of the party take over for real; besides they aren't in a great position as they mostly acquiesced to the coalition and the ones that didn't just stayed quiet.

[Filuwaúrdjan]

Poll internals are...

Anyway. I do note that teh nayshun is united on one thing: England's footballers are crap.

[Filuwaúrdjan]

Then why are Labour doing so much better than at the General Election and the LibDems so much worse? The CSR didn't come as a shock to the system because the build-up for it has been going on for months. Government popularity tends to erode rather than suddenly collapse and that's very much been the pattern we've seen so far. Government support only tends to fall suddenly when there are material factors at work; we saw this with price rises in 2008, for example.

[Filuwaúrdjan]

Marginally, maybe. Institutionally the map favors Labour, but this is because the map is drawn based on population, not voters, and voters in really Labour areas are almost universally less likely to turn out than voters in Conservative areas (really Conservative seats can get as high as close to 80% turnout while really Labour seats can go as low as 50% turnout).

No, it's based on the number of registered electors. But it is true that what 'bias' exists (which is more that it is easier for Labour to win a majority than the Tories, rather than Labour being grossly overrepresented in relation to the Tories - which they aren't at the moment) is down almost entirely to differential turnout.

[Filuwaúrdjan]

Basically, your core assumptions on this issue are all wrong.

Mostly wrong anyway. Working out what's rural and what isn't in a country with such a pronounced tendency towards industrial villages and with an unusually traumatic experience of industrialisation/urbanisation is problematic, and there's an unfortunate tendency to define it based on the pastoral idyll rather than anything halfway objective. So large parts of southern England that are clearly not rural in terms of social or economic structures are commonly regarded as such, while it is more or less a taboo to acknowledge the essentially rural nature of most of the coalfields and a lot of early industrial districts.

[Filuwaúrdjan]

Ah, the famous villages of Glasgow.

---

All questions... answered. Once again:

You must be logged in to read this quote.

I may have to adjust this at some point, due to the changes in the position of the Yellow Party, but as a guide to the last election it's fine.

[Filuwaúrdjan]

Amused that Angus Reid are still working here; presumably they've made certain major adjustments to their weighting. Anyway, why are YouGov still doing daily polls? During an election that's quite interesting, outside one? Obviously because The Scum is paying them to, but why is The Scum paying them to do that? Someone at YouGov must be very good at scamming newspapermen.

[Filuwaúrdjan]

The tables show that the LibDems are tied with the Communist Party in the North Wales region.
http://today.yougov.co.uk/sites/today.yougov.co.uk/files/YG-Archives-Pol-ITVWales-DecemberPolling-221210.pdf

Amusing, though poll internals are always dodgy, YouGov poll internals doubly so (a consequence of the panel system) and North Wales has been one of the weakest parts of the U.K for the LibDems since 1992. Still, lol.

[Filuwaúrdjan]

And we're only 7 months in folks...

Doesn't that work both ways though? If the situation improves, wouldn't the Lib Dems benefit?

The usual tendency is for the polling position of the government to erode throughout the Parliament, until the last few months before the election. O/c the Con/Lib split shown by YouGov is a little different to that shown by other polling firms; they tend not be quite so extreme.