Euro and county elections – polling and predictions
YouGov’s poll in the Sunday Times also included voting intention figures for the European elections. Topline figures (with changes from when YouGov last asked back in January) were CON 36%(+1), LAB 25%(-4), LDEM 20%(+5), UKIP 7%(nc), Green 4%(-1), BNP 4%(nc), SNP&PC 4%(nc). No other party reached 1%.
Despite their slump in Westminister polling, Labour are still above their actual vote share in 2004. The Conservatives and Lib Dems are also both substantially above their share of the vote in 2004, and I suspect the reason is that we still aren’t seeing the sort of level of “other” support in the polls that we will eventually do on polling day.
This may be because people haven’t considered the minor parties yet, and they will naturally rise during the campaign as their freepost addresses go out, alternatively it may be down to the problems of whether to prompt by minor parties in questions or not (a question I expect I’ll write about in a lot more detail as we get closer to the election), or perhaps support for minor parties has genuinely fallen away.
For the record though, on a uniform swing, these figures would result in the Conservatives winning 30 seats (up 3), Labour 20 (up 1), the Lib Dems 13 (up 4), UKIP 4 (down 8 ) and the SNP and PC one each (though this is solely because of the uniform swing calculation, in reality their increase would obviously be only in Scotland and Wales and would result in more seats). The Greens would lose both their seats, and the BNP would fail to gain one.
The Sunday Times also has Rallings and Thrashers predictions for the local elections, based on local government by-elections. They are predicting a national equivalent vote share of CON 40% (up 9 from 2005), LAB 25% (down 9), LDEM 27% (no change) – this would equate to the Conservatives gaining 300 seats, Labour losing 250 and the Lib Dems losing an unspecified number. Labour would loose control of all of their remaining county councils, though it would not be quite enough for the Conservatives to talk overall control of Derbyshire.
Good to see the BNP will get nothing, suprised that Labour and Liberal Democrats will gain. Not surprised that Greens and UKIP will do very badly.
Could it be what everybody has been talking about?! Lib Dems begining to overtake Labour in the polls?
Thanks for this Anthony.
Based on the Rallings and Thrashers prediction (which I believe has been pretty accurate in past years) clearly the Local Elections next month are going to be terrible for Labour. However I wonder if the Tories can take much comfort from not pushing past 40% (down from 44% last year)?
Using Electoral Calculus these figures show a Conservative majority of 46 on a national swing of 9% since 2005. As we know from past experience all past governments have a swing back to them in the final year before a general election (Blair recovered from 26% in 2004 and even Major had a small swing back to the Tories in 1997 of 1%-2%). Assuming these are the actual Local Elections results this June 4th and also assume a small swing back to Labour next year of just 2% (historical average) from the Tories that would show the Tories short of an overall majority by 6 seats (Con 38%, Lab 27%, LD 27%).
So a hung Parliament (or a small Con majority) next year becomes a real possibility compared to today’s talk of a Labour wipeout (similar to 1997) based on the latest polls.
Next month’s final results for the Local Elections will indeed be interesting.
No Michael, the Liberals nearly always do better at a local level than Labour, look at last years Local Elections. They may do better in the euros to thanks to proportional representation. But in a general election it’s another story. Every poll thats comes out says the will lose seats, mainly because of the tories.
Sorry my first post was directed at the first Micahel not Michael B.
I don’t think everyone who is thinking about voting BNP is going to admit to the pollsters that they are considering voting for them. This is probably still true even for automated polling systems.
I think this poll is certainly overestimating the level of support that Lab and LD will actually receive in the Euro election. On the other hand, the Con share of 36% looks like it could be right.
If Thrasher & Rawlings are predicting 40% for the Conservatives, on a national projection, in the County Council Elections, then I’m cautiously optimistic the Conservatives can reach about 45%, which would better last year.
They under-estimated the Conservative share decisively in both 2006 and 2007.
That said, local council by-elections have been a slight cause for concern in recent months, but if this explains the discrepancy from their projections in 2006 and 2007, then perhaps the Tories weren’t doing particularly well in by-elections then either, but still surpassed expectations in May.
But of course, the electorate will decide, and the last few weeks haven’t been that great for “politics” generally, which could boost “others”.
The crucial factor to remember with the county elections is that the last time these seats were up, there was a general election on the same day. This depressed the Lib Dem and Independent vote whilst boosting the Labour vote. The Tories would have seen their vote boosted in contests with the Lib Dems and depressed in contests with Labour.
We should therefore see some significant Tory gains from Labour in many councils and from Lib Dems in places like Devon and Somerset. But we should also see some significant LOSSES to Lib Dems and Independents in the stronger Tory areas.
I have analysed local election results in key Tory target counties held over 2007-08. If the county council elections this year go the same way as these district council elections did, we can expect the fgollowing results:
Derbyshire- NOC GAIN from LAB
LAB 26 (-12) HUNG- 12
CON 24 (+9)
LD 12 (+2)
BNP 1 (+1)
Ind 1 (=)
Devon- CON GAIN from LD
CON 36 (+13) MAJ= 10
LD 19 (-14)
IND 4 (+2)
LAB 3 (-1)
Lancashire- CON GAIN from LAB
CON 46 (+12) MAJ= 8
Lab 21 (-22)
LD 10 (+5)
Ind 5 (+4)
Green 2 (+1)
Nottinghamshire- NOC GAIN from LAB
CON 30 (+4) HUNG- 7
LAB 16 (-20)
LD 11 (+6)
Ind 10 (+10)
Somerset- CON GAIN from LD
CON 33 (+9) MAJ= 8
LD 22 (-8)
LAB 2 (-2)
IND 1 (+1)
Staffordshire- CON GAIN from LAB
CON 45 (+17) MAJ= 28
LAB 10 (-22)
LD 5 (+3)
UKIP 1 (+1)
IND 1 (+1)
For these results to happen, the parties don’t need to achieve anything MORE than they have done in the last 2 years. This is roughly how the county councils would have gone if they had been up for election at the same time as the local elections in 2007-08.
Should be a fascinating night.
Has likelihood to vote been factored in? The Labour vote may be inclined to have a not bothering to vote at this point as it’s not for the Westminster parliamnet?
The poll in The Times has one figure that seems rather strange. On their headline voting intention they breakdown the vote into regions and have the Conservatives on 29% tied with the SNP and 1% ahead of Labour. Can this really be accurate?
I am going on the basis that the results of elections in these SAME wards as are up this year are more accurate than other methods such as looking at local by-election results.
The turnout this year in the county divisions should not be markedly different from the distrct ward results held over 2007-08, so there should be little real impact.
The thing that could have an impact though compared to the 2007-08 local elections is individual county councillor’s ‘incumbancy effect’. This may particularly be important in divisions that cover wards where independents have done well in 2007-08 (such as in Mansfield and Ashfield). Whose to say that these victorious independents didn’t get electedon a personal vote that may not be carried accross to county elections for other (different independents). That will probably only be a small minority of the seats up for grabs though.
The BNP didn’t even register on the polls in 2004, yet they got 5% of the vote. Polls are not a good way to judge BNP support.
I think UKIP will get 5%, with the BNP on 12-13%. The BNP message is more in tune with the mood of the country at the moment, and for that reason I think they should do very well.
In defence of Thrasher & Rawlings, while I am unaware of their 2006/07 predictions, their current prediction is in line with recent local council by-election results in which the Tories have been under-performing compared to the national polls.
http://www.24dash.com/news/Local_Government/2009-05-01-Lib-Dems-gain-Harlow-seat-in-latest-council-by-elections
“Analysis of 14 comparable by-election results in April 2009 suggests an 11.3% nationwide Tory lead over Labour.
A calculation based on 13 wards fought both times by all three major parties gives a line-up of C 37.9%, Lib Dem 29.0%, Lab 26.2%.”
Actually those figures in a General Election, based on Electoral Calculus, would show the Tories short by 8 of an overall majority. However I think that Thrasher & Rawlings will be closer to the actual result on June 4th.
Michael – they would be in line – that’s what R&T base their predictions on (in fact, while it doesn’t cite them, there’s a fair chance the figures in that news story were probably based on calculations by Rallings and Thrasher).
R&T have had some success predicting local elections using local government by-elections. They’ve underestimated the Tories the last couple of times, but in at least one of those cases it was because Labour had some negative news stories a fortnight before the local elections and slumped in the polls, so a prediction from by-elections in the last six months obviously became out of date.
They have not had similar success at predicting general elections though, simply because the Lib Dems do far better at local elections than national ones. I did a long post a couple of months ago trying every which way I could to come up with a relationship between local by-elections and Westminster elections. I couldn’t.
For the record, I also did a long post on the commonly held belief that government’s always recover.
Oliver & Andy – My view is the same as Andy’s, the BNP are almost certainly under-represented in phone and face-to-face polls because people are reluctant to admit to another person that they are supporting them. That will be less of a factor in an online poll, since there’s a computer screen rather than a person, but I’d be surprised if the social acceptability bias vanished completely.
However, the BNP have not been that under-represented. In YouGov’s final poll of the 2004 European campaign the BNP were at 4%, so not grossly underestimated.
James – no, it probably can’t. Regional splits have very small sample sizes to start with, so give wacky results. Added to that nationwide polls are weighted to be representative overall, not to be representative within regions. My advice would always be to ignore the regional splits in voting intention polls.
If you take the historically correct view that the Conservatives will do better in the actual elections than these polls suggest then the on Euro election they could *JUST* scrape to a 40% vote share and on the county council elections a 40%+ share of the vote looks very much on the cards with Labour facing an absolute blood bath.
[...] surveys have the BNP polling at about the 4% mark. According to the YouGov poll in the Sunday Times the BNP are polling at 4% (nc) and according to Predict 09 it is polling at [...]
fascinating poll – The key factor will be differential turnout.
The last time the County seats were fought was general election day – this boosted the Tory turn-out vs the Lib Dems. I suspect the Euro will
has a similar but smaller effect. In county seats the Lib Dems won even when there was a genrel election on, they ought to be fairly safe – bar particular local factors.
I think the BNP 4% is spot on. The BNP have targeted some local govt by-elections but as we know, that is a poor guide to overall results.
GIN The Conservatives polled 40% of the votes in the CC elections held on GE day in 2005 They will certainly get more than that this year .
You may be meaning the National Projected Vote Share . The calculation that the BBC and R and T do in non election years is not possible to calculate in the same way this year as there are no elections in most of the wards they use every year .
This poll, taken before electors have started to think about the European Elections, continues to suggest that the levels of support for the major parties are remainlng fairly steady at present.
Concerning the European Elections, there are prediction issues not only in relation to the likely campaigns by minor parties, but also in relation to differential turnout, which, as has been discussed recently, is a very difficult issue in relation to extrapolation from opinion poll data to the prediction of actual elections.
There is an issue arising in relation to Westminster elections which may also make it harder to use poll results for prediction purpose in the coming months.
According to the “Sunday Times”, this poll was carried out as the Expenses issue broke. The papers said that there was no practical difference in the data collected on the first and the second of the two sampling days. Of course, this does raise questions as to what “lag” takes place to delay news being converted into changed voting intentions.
However, it seesm to me that there is another psephological issue about the Expenses scandal. This issue affects MPs of all parties, but it affects different MPs to different degrees.
In these circumstances, it seems to me that the Expenses issue may not affect expressions of political support. What it may do is affect the “incumbency” effect”. Whereas in the past sitting MPs have achieved comparatively good results at elections, next time round some MPs may do badly because of their expenses. Possibly, the effect will be significant only for a handful of scapegoated MPs, or ones who have been particularly stupid/venal. And it is difficult for polls to pick up such outlier results.
In recent months, the dominance of economic concerns on UK political life made it appear that factors influencing voting intentions in particular consituencies were going to be pretty irrelevent at the next General Election. There would be a fairly uniform swing simply depending on the state of the economy. But it looks as though, in some seats, electors may determine a voting intention on national issues, but then modify it nearer the election, perhaps even in the ballot box (or, sadly, on the day they receive their postal vote), when they find out if their particular candidate has a bad record on expenses.
The only real answer to this problem is to do, as PopulusHome did last year, polls with samples enabling trends to be identified in particular consitutencies. As indeed PopulusHome did last year. But of course such polls have resource problems.
Fred
Do you really think the vast majority of voters care about EU versus local elections? I would say no.
People vote on the day. They vote on the issues of the day. So, currently it will be a plague on all your houses (look out for a low turnout) and even worse for Labour. Regrettably I think it will be interesting for those who can turn out their core vote; once a a BNP nutter always a BNP nutter. I hope Greens beat BNP by a lot… (Anyone object to that?)
I think that it will take some considerable time for the expenses issue to unwind fully; especially as July will see the FoI requested info published; i.e. all of it. Should make some interesting headlines if the media hasn’t exhausted itself by then.
Then we may see a clearer picture. The “run in” would have started as well.
Last day to declare a general election tomorrow!
25/1 at bookies!
Can I sleep tonight as a candidate at the CC s?
right based on R&T the tories will gain.
CON GAINS
devon from lib dem
nottinghamshire from labour
somerset from lib dems
staffordshire from labour
warwickshire from no overall control
LABOUR LOSE
lancashire to no control con biggest
cumbria to no control con biggest
derbyshire to no control labour biggest
bristol to no control lib dem bigget
LD LOSE
somerset
devon
and with the new councils the figures point to a conservative win in wiltshire, bedford, central bedfordshire, shropshire and a lib dem win in cornwell
On the subject of the rise of the BNP et al., I’m not too worried.
I think that this sort of thing serves as a warning for mainstream parties about “consensus politics” and trying to shut out “wacky” ideas from reaching influential ears. They have risen, but they will fall again, having served their purpose as a “shock tactic”, like other fringe parties. UK voters seem to think that EU elections really don’t matter, but are national, so are ideal for a protest vote.
Perhaps you could use Adam Smith’s comment about when leaders of an industry get together, a conspiracy against the public forms, and suggest that when leaders of the major parties start to get too cosy about an area of politics, they may forget/disregard their voters’ views.
@Jack. yes I object to that.
I think the polls are interesting, and would concur with others that BNP support is almost certainly understated.
As you are bringing personal opinions into it, I think many people would far rather support a party that wants to leave the EU, clamp down on dole scroungers and bring back the death penalty for certain crimes, than a bunch of vegetarian hippies who want us to go back to rubbing sticks together to get warm.
However, I am likely not to vote for either of them.
Jack “Do you really think the vast majority of voters care about EU versus local elections? I would say no.”
That just isn’t true. UKIP do much better in EU elections than in a GE. So do the Greens. The LibDems do better in Local elections. So that shows, all things being equal, that people vote differently in different types of election.
I think by having the two on the same day, that might mitigate the effect – it will be nteresting to see the difference.
Pete B,
“a bunch of vegetarian hippies who want us to go back to rubbing sticks together to get warm”
I’ll admit that I’d never vote for them but I have to say that a forest covered, technology free little island with non of the rat race that modernity has brought sounds quite nice really!
Add a touch of hanging and flogging of the worst criminals and a lower population through zero immigration of course and you have the best of both worlds
Perhaps the BNP and Greens should merge?!
Jack,
“…interesting for those who can turn out their core vote; once a a BNP nutter always a BNP nutter”
Not sure that’s entirely true in the case of the BNP. Of all the partys their support tends to fluctuate the most. A very weak left of centre mainstream gives them a boost historically.
Not all are ‘nutters’, just a small section of the apathetic working class/ poor I would suggest.
But Stuart, as I have already demonstrated, the Tories have a far bigger chance of gaining control of Lancashire than they have of gaining Nottinghamshire.
Also, I think Bristol is already no overall control with the Lib Dems as the biggest party isn’t it?
What do people think about how the UKIP vote might effect the Conservatives locally? In my Surrey town its a two horse race between us and a local residents’ assocation (the incumbents). Judging on last year’s local elections the Conservatives would easily gain the seat from the assocation, but I’m thinking that UKIP may take votes from the Conservatives. UKIP seem to have stood candidates in a few Cons marginals in my area
Jack,
“I hope Greens beat BNP by a lot… (Anyone object to that?)”
No objection, but you should not let hope rule your expectations as you are then more likely to be disappointed.
Two predictions for Euros:
(1) BNP will gain more votes nationally than Greens.
(2) BNO will have more MEPs on 8th June than Greens
(though personally I hope that both have none)
Sorry, for “BNO” read “BNP”
@Stuart and Shaun
Yes, Bristol are currently NOC with LibDems the largest party and a LibDem leader.
LIB 42
LAB 34
CON 13
GRE 1
Total 70
Anthony,
Why not have a sporting bet amongst us all on the level of turnout on 4th June.
No prizes – just an announcement of top three places and your praise!
All answers to have a decimal point.
Thanks
Larry
Apologies if this has been covered, but reading Michael B’s comment at the top of this discussion I am slightly confused.
Can we do as he has done and extrapolate from these local election polls to a general election result? Or is it best to extrapolate from the broader polls to the GE result?
Thank you
CharlieJ
As Anthony has documented, the evidence suggests that one cannot extrapolate from local election results to general election results except in the broadest terms (i.e. basically which parties are up or down) – in which case the national opinion polls are a more accurate guide.
Note also that while Anthony has commented that local council by-elections can give some guidance to local election results, it is even clearer that local council by-elections cannot predict general election results.
Michael B @ 12:03 yesterday said “all past governments have a swing back to them in the final year before a general election”
This is one of the things that everybody knows but isn’t actually true. Mike Smithson debunks it here: http://politicalbetting.com/index.php/the-swing-back-myth-icm-199720012005/
@Michael B you said, “Blair recovered from 26% in 2004″
But Labour polled:
April 2004 – 34,38,36,35 averaging 35.75%
May 2004 – 36,32,37,37,33 averaging 35%
Labour actually gained in May 2005 – 36.19%
Would anyone like to explain the by election result in Hartlepool UA last Thursday?
It seems to go against every trend.
I’ll eat my hat if the Greens poll better than, or come anywhere near the BNP.
I still think the BNP will comfortably beat UKIP, with both parties doing better than I first predicted due to the expences scandal that is ongoing.
My predictions for the EU elections:
BNP – 13% UKIP – 9% Green – 5%
Ivan,
As you guessed, my tongue was ever so slightly in my cheek with my previous comment, but what a lovely idea about the merger. They could become the BGP which has the advantage of standing for B____y Good Party!
Seriously, I agree with Oliver that we could well see a big surge of votes for non-mainstream parties in both local and Euro elections, whether the polls show the trend in advance or not. Some locals only need a few hundred votes for the winner.
It could get quite interesting, with hung councils being run by a coalition between Liberals and Greens, or Labour and Socialist Worker, or Tories and English Democrats. I can’t wait. I’m an election junkie, and I’ve already booked June 5th off work so i can stay up all night to watch the electiosn.
Sorry, last word was ‘elections’ of course.
Anthony,
As people keep making predictions on here why not have a predictor where people log in their pre poll predictions and we get a rolling average of them in the run up to polls.
It would be interesting to see how close or far we were from what the polsters come up with.
Peter.
Just an explanation of the Hartlepool by election blip referred to above.
This was a freak/unique result because it was one of the very few areas outside the SE where UKIP have any presence at all. At the last local elections, they threatened Labour but this time, the presence of a spoiling BNP candidate reduced their impact radically. I think it is the only instance in the past six months or so where UKIP have beaten the BNP in a local election – any corrections gladly received.
I suspect Stuart Wheeler’s money will shore up UKIP support in their brain-dead heartlands but in the North West and West Midlands, they will disappear without trace, thus letting in the BNP. I suspect that the BNP will get three, or possibly, four Euro seats: NW definitely, Yorkshire probably, West Midlands probably and East Mids possibly. UKIP will hang on in the SW and SE but be wiped out entirely elsewhere.
I hope the public don’t buy this shallow media driven response from Cameron and rightly punish a party that tried to portray themselves as whiter than white but turns out were actually far more extravagant than Labour.
Cameron would have known about this before it came out in the papers; but only now does he decide to be outraged and ask his MP’s to pay back! He clearly hoped the Tory press wouldn’t publish these and keep quiet, however it’s good to see they followed through and published these.
How Cameron can play the strong decisive leader when he failed to impose a banning of multiple jobs in his MP’s – funny how this got swept under the carpet but I personally think this is just a great an issue.
A great example of how media driven this guy is.
Anthony – “In YouGov’s final poll of the 2004 European campaign the BNP were at 4%, so not grossly underestimated.”
Could you put up the corresponding 2004 YouGov figures and the actual percentages achieved ?
What I’m interested in is where, if anywhere, 2004s hefty UKIP vote is going. Then they had something like 18% of the vote, Greens had over 5% and BNP just under 5. So getting on for a third of the vote went outside the three main parties – a ‘plague on your houses’ vote.
Given the events of the last five years I find it hard to believe that people are less cheesed off with the main parties than they were in 2004. Yet the total for those three parties is apparently down from 28% to 15%. Does that seem likely ? Can it really be that the UKIP vote is bleeding back to the Tories ? I’d have been thought it was more likely to split 50/50 BNP/Tories.
I
Laban,
YouGov archived their European polls here. There seems to be 3 before the 2004 elections, dated 25/05 (for UKIP; poll taken 12-14th); 25/05 (for the Telegraph, taken 20-21st); and 8/06 (taken 2-4th June). The combined UKIP, BNP and Green vote goes from 14, to 23, to 29% in those 3 polls, with BNP going from 1 to 3 to 4%.
Actual vote for the BNP was 4.9% (other parties there too).
On the basis of that, I’m quite liking my BNP-to-fail wagers. However much the BNP under-reporting bias is, I doubt it has changed much in 4 years, so the qualitiative conclusion ought to be robust that BNP support is going nowhere. Is there some reason why everyone’s presuming a) a UKIP collapse and b) that the UKIP support will therefore go to the BNP rather than its natural home in the Conservative party?
Dsquared,
Good luck with your wagers.
a) Likelihood of UKIP collapse:
Actually, the consensus prediction is that UKIP will fall back towards the level they achieved in 1999 rather than the spike they reached in 2004 courtesy of the high profile campaign led by Robert Kilroy-Silk.
Several factors support this view, not least that UKIP has had well documented organisational difficulties and personality splits. (Only 8 of the 12 UKIP MEPs elected in 2004 still sit for the party, and several of them are not standing again).
Another major factor is that, from an EU perspective, the Tory position appears more credible / palatable to the type of natural conservative who may have switched to UKIP in 2004.
b) UKIP switchers to BNP:
This is a contested view. I agree that the natural home of the large majority of UKIP voters in 2004 is in teh Conservative Party. This can easily be demonstrated by comparing the respective fall/rise of Con/UKIP votes in 2004 vs 1999 in Euro region by region.
Some, particulalrly on the left, make two lazy assumptions:
(i) that the BNP is a “right-wing” party; and
(ii) that the BNP is therefore viewed as an acceptable repository of Tory protest votes.
(i) is only true in terms of the BNP’s “nationalist” attitude. But this is quite distinct from “patriotism” or upholding traditional British values. In all otehr respects, the BNP platform has much in common with socialist / authoritarian parties, hence its support tends to come from traditional working class areas.
(ii) is founded on the mistaken assumption in (i). In reality, the majority of those who voted UKIP in 2004 will either:
– vote UKIP again
– vote Conservative
– stay at home
A few may well vote BNP, but the numbers will be insignificant in polling terms, and are likely to be concentrated in the E Midlands region.
Thanks for that, Barney. What’s notable is the rise of the combined UKIP/Green/BNP vote between May 12 and June 4 – from 13% to 29%, a percentage which was pretty much reflected in the results.
d2 – where can you bet on Euro results ?
In my case, the pub (plus a charity bet with a couple of CiF commenters).
[...] shows a rise of 8% in less than a week, compared to the Sunday Times YouGov poll last Sunday which put UKIP on 7%. Most observers are putting the surge in popularity down to [...]
Paul h-j
“(i) is only true in terms of the BNP’s “nationalist” attitude. But this is quite distinct from “patriotism” or upholding traditional British values. In all otehr respects, the BNP platform has much in common with socialist / authoritarian parties, hence its support tends to come from traditional working class areas.”
How exactly would you differentiate between “nationalism” and “patriotism” and “upholding British values”?
Which policies make you regard them as “authoritarian” exactly (compared with Labour for example)?
What makes you think the support tends to come from “traditional working class areas”, given the spread of members on those lists that were published recently?
I just don’t see where the evidence is for your assertions.