Added in the right hand margin is a new ConservativeHome poll of polls.
I'm grateful to those visitors who responded to my request of 24th February for advice on how to set up such a poll of polls. I've chosen a simple average of the last five polls of Britain's main pollsters - BPIX, ICM, Mori, Populus and YouGov. I don't pretend it is a perfect poll of polls. In many ways it is quite crude but it gives a rough idea of the state of the parties.
Opinion polls have tended to overstate Labour's position in recent years and it must be hoped that they still do. Nonetheless, the first poll of polls suggests that Labour has a very narrow 0.4% edge over the Conservative Party.
The poll of poll numbers will be input to the ElectoralCalculus website to produce the seats projection at the bottom of the graphic. On this occasion that produces a Labour majority of 50. This, of course, is based on a uniform nationwide swing. It is likely that results will be more complex than that with significant local and regional variations. The concentration of LibDem support would be very unlikely to reduce their number of seats to 32, for example, even with just 19.2% of the vote. Tactical voting - which has hurt the Tories at the last three elections - may even start to be used against a disliked Labour government.
The headline figures from the last twenty political polls are listed in the left hand margin - below the 'About' graphic.
"Tactical voting - which has hurt the Tories at the last three elections - may even start to be used against a disliked Labour government."
This is already the case. I was one of many who voted tactically for the Liberal Democrat candidate Jenny Willott in Cardiff Central in order to unseat the sitting Labour MP.
Posted by: Daniel Vince-Archer | March 19, 2006 at 19:43
The Lib Dems also seem to be particularly good at targeting 'winable' seats, so I doubt they would sink as low as 32
Posted by: comstock | March 19, 2006 at 20:04
Does Electoral Calculus take into account the new boundaries for seats in England and Wales? If they don't, then the editor should modify the seat numbers to take them into account.
Posted by: CDM | March 19, 2006 at 20:08
"Opinion polls have tended to overstate Labour's position in recent years and it must be hoped that they still do."
This was because many voters would not admit voting Tory, and caused some pollsters to adjust their results. Now that Cameron has made it "acceptable" to be a Conservative, there must be a danger that the polls overstate our vote.
Posted by: Nick | March 19, 2006 at 20:44
"The Lib Dems also seem to be particularly good at targeting 'winable' seats, so I doubt they would sink as low as 32."
They're also notoriously difficult to unseat when they do get in, although I would expect their total number of seats to fall back to the 50s at the next election.
Posted by: Daniel Vince-Archer | March 19, 2006 at 20:51
You ask a good question CDM.
ElectoralCalculus says "The first phase of the project to handle the new constituency boundaries is now complete. Labour's majority is calculated to reduce by 22 due to the boundary changes." I'm not sure if this reduction of 22 is incorporated in their projection for my poll numbers, or not.
I'll endeavour to find out.
Posted by: Editor | March 19, 2006 at 21:01
I don't think Electoral Calculus takes the boundary changes into account so you can knock 11 of the predicted Lab seat total.
From what I've seen of the YouGov polls though we've been a steady 37-39% since December whilst Labour has a range of 35 to 40%, with the movement being between the LDs and Lab. I think 'neck and neck seems to be the case whichever way you cut it.
Posted by: kingbongo | March 19, 2006 at 21:39
Surely it is impossible to know of the full implications of the boundary changes and what effect they will have on seats as a whole? Labour have previously been successful in hijacking Boundary Reviews for their own electoral purposes. There has been widespread malaportionment and gerrymandering. I only hope it hasnt happened again.
Boundary changes are the bane of my existance.
Posted by: Rob Largan | March 19, 2006 at 21:44
Electoral Calculus does not yet take the boundary changes into account. The best boundary changes adjustment calculation that I've yet seen is at Anthony Wells' at: http://www.ukpollingreport.co.uk/blog/newseatguide.pdf
- the benefits to our seat share depend upon the total swing.
For tactical voting adjustments, the accumulated factor is currently equivalent to:
+5% (on the 1992 baseline) Labour to Lib (+4% in 1997, further +1% in 2001, no change in 2005) and
+6.5 (against the 1992 baseline) Lib to Lab (+5% in 1997, further =2% in 2001, unwind of 0.5% last year).
So a full unwind of the tactical vote would be -5% in the red box and -6.5% in the yellow box*
There's no regular measurement of tactical voting; for me, the best meter is the "forced vote" question ("if you had to choose between a Labour government headed by x and a Conservative government led by y, who would you choose?"). This forced vote should then be compared with the declared headline support to see how the non-supporters of the Conservatives and Labour split.
For example, at the election, YouGov measured the forced vote question at 53/35 for Labour, on top of the headline 37/32 in their poll. This meant (after adjusting for "would not votes" and "don't knows" from the dataset) that the non-supporters broke 71%/29% for Labour, a lead of 42%; not much changed on 2001 (thus there was little change in the tactical vote measured by Baxter).
In the last poll for which I have this data to hand (the YouGov poll of 15 Feb, where we trailed 39/37), the forced vote had been pegged back to 43/37 for Labour. This meant that the non-supporters broke 59%/41% in favour of Labour, cutting the lead here to 18%. This could arguably result in an unwind of half of the accumulated tactical vote.
This is completely hypothetical - the numerical calculation depends on a lot of estimates and could well be worthless - but the undeniable fact is that Cameron is catching up well on the "forced vote" question and this seems to reflect a significant reduction in anti-Tory feeling..
*Proviso - when the negatives of these numbers are put in to Electoral Calculus to test the unwind, Labour and Conservatives end up level on seats when level on vote share (when adjusted for boundary changes). As there should be a remaining bias to Labour from the Welsh seat excess allocation (by about 6 seats) and differential turnout (by about 8 seats), a slightly smaller number to reverse may be appropriate (say +4% Lab-Lib and +5.5% lib-Lab)
Posted by: Andy Cooke | March 19, 2006 at 22:09
We've just started doing a more sophisticated analysis at www.forecastuk.org.uk. As well as aggregating the last months polls, we're also taking into account local government returns and regionalised opinion polls.
A private project during the 2005 election got us within 2 seats of the actual Conservative figure for gains - we think we can do even better now, and our current prediction is a Labour minority government short by 10 or so seats - quite a bit different from your forecast!!! Check us out and see what you think.
Posted by: Peter O | March 19, 2006 at 22:58
Conclusion after 100 Days of Cameron Project:
-some upward movement in the polls but no sign of significant increase;
-but this likely would have happened with Cameron as young, attractive leader even if he hadn't dumped so many of our policies;
-the Leader has taken the party much further afield that anyone expected but without producing overwhelmingly positive poll results--this will lead to shallow support and a lot of face scrachting:many tories will wonder why exactly did we ditch all of our policies if the result is a project Lab majority of 50?
It's also very hard to see what new DC can bring the next three years to fundamentally change the situation.
Take into account that Brown is going to get his honeymoon too and the change of guards will be like getting a new government, thus satisfying voter demand for a change, and it all promises to be very dark and gloomy.
I predict that unless DC starts to deliver results, he will come under immense pressure. His position at the other side of an General Election is questionable.
Posted by: Rebel | March 19, 2006 at 23:21
I think tactical voting will only be effective if Lib Dem voters start voting Tory to unseat Labour MPs. This still seems unlikely to me.
Posted by: John Hustings | March 19, 2006 at 23:34
The current position of the polls is very interesting. Labour and the Tories are roughly neck and neck and the Lib Dem level of support is down on last time. There is though a considerable degree of anti-Labour feeling out there and a large portion of Lib Dem voters who are more centre-right then centre-left (i.e 1990s defectees from the Tories) will actually have little problem in voting Tory as a tactical vote, especially in the more marginal Labour seats which tend to be pre-1997 Tory seats.
That's why in our prediction we're are forecasting far fewer Labour seats then the current ConservativeHome forecast, despite our percentages being similar (though derived in a different way). Our reading of "on the street" opinion through the country tells us that tactical voting will occur and may be significant, even more so than 1997.
Posted by: Peter O | March 19, 2006 at 23:48
I cant see mass tactical voting in the Conservatives favour. It is likely that there will no longer be a significatn level of anti-tory tactical voting, but I see little chance of this developing into large enough levels of anti-Labour tactical voting to chance things all that much.
Posted by: Rob Largan | March 20, 2006 at 00:01
There is a deep disatisfaction among both Labour-switchers (those who voted Labour first time in 97) and old-core Labour. What makes tactical voting a likelihood is that the Cameron led Tory Party is suddenly a viable home for anti-Labour votes in a way that the Hague/IDS/Howard Tory party never were.
Posted by: Peter O | March 20, 2006 at 00:08
I think Cameron's movements will gain in the long-term yes, he has no concrete policy, but the image he presents to people is a great one, which I believe he CAN and WILL capitalize in the future.
I'd say that labour and the tories are neck and neck. I cannot see the Libdems losing any seats in the next general. There will be significant tactical voting against labour. We should expect a massie fall in labour voters in ld-lab marginals.
The labour vote is frankly being leached by the libdems, this might not get the Tories in power, but with an advantage.
There results are very promising.. What it shows is Tory reinvigoration, something that we simply didnt have under Hague/IDS/Howard. The image of the party has changed from the Major era into a new modern "touchy-feely" party.
This is not going to result in overnight massive gains...But I think for the longterm it will benifit the party than most of the rightwingers would care to admit.
Posted by: Jaz | March 20, 2006 at 00:31
"I'd say that labour and the tories are neck and neck"
That is what concerns me. How on earth are Labour holding firm and the Conservatives failing to gain from the current mess the government are in? It is possible that the reason there isnt a significant drop in the Labour support is because of the debacle that the Lib Dem leadership contest turned into. What is worrying is that the Conservatives are failing to benefit from the Labour sleaze. Is it not reasonable to assume that people maybe prepared to express a protest vote against Labour sleaze with the Lib Dems, but because the Conservatives havent offered a clear coherent alternative vision, voters have no real reason to switch.
Posted by: Rob Largan | March 20, 2006 at 01:14
>>>>They're also notoriously difficult to unseat when they do get in, although I would expect their total number of seats to fall back to the 50s at the next election.<<<<
This is because of tactical voting by members of partys where they know that their party can't possibly win - this is the only way that Simon Hughes has been MP for Bermondsey since 1983 and why the Liberal Democrats were able to take Chesterfield in 2001 and hold it in 2005 and why they hold seats such as Yeovil and Westmorland & Lonsdale, and why in 1951 and 1955 with percentage votes approximating to what UKIP got in 2005 they were able to hold 6 seats when the 2 party vote was at it's strongest.
Posted by: Yet Another Anon | March 20, 2006 at 02:18
How on earth are Labour holding firm and the Conservatives failing to gain from the current mess the government are in?
I think it's partly due to fairly low interest rates. If you have a stable income stream, you're mortgaged to the hilt, and you remember what happened in the 1990s, why would you change your vote?
Posted by: Alexander Drake | March 20, 2006 at 05:12
Funny that comment by Alexander Drake - there are 3 times as many savers as borrowers so low-interest rates are causing some people to have an unstable income stream.
At the same time Council Taxes are rising, and annuities from houses like Equitable Life have been cut for existing pensioners.
Presumably in your world only borrowers vote..........?
Posted by: Rick | March 20, 2006 at 07:08
Rick
Alexander has a point - we are on the whole strong in age groups where mortgages are paid off. There is also an overlap between savers & mortgagees. Those with mortgages - especially those who took them out 15+ years ago remember well the pain of high interest rates.
There is also Browns payroll vote & dependency vote.
Between these two groups a vote share of 36% isn't that unexpected.
But issues like council tax, low annuities, energy costs, taxes etc are beginning to bite - look at how consumer demand is falling. Add in sleaze and we might see a shift,
Posted by: Ted | March 20, 2006 at 08:18
But Ted,will we benefit from the public concentrating on Labour sleaze?They will also be reminded of our past record as the Grauniad does today and our refusal to name who have lent us money to date contrasts very badly with the open and transparent way the Lib Dems are handling it.
Posted by: malcolm | March 20, 2006 at 09:42
malcolm
Well if we could get Sir Ian Blair to investigate Tony Blair under the 1925 Act which outlawed the Sale of Honours that might change:
While not directly relevant to Acts covering national Government the Courts have construed corruption/bribery in Acts like the the 1889 Public Bodies Corrupt Practices Act and following Acts such as to place the burden of proof on officials or Crown servants to prove a gift and a favour were not related.
"that the burden of proof shifts on to the defendant to show (on the balance of probabilities) that the money, gift, or other consideration was not received corruptly"
and that "Corruptly' in the 1906 Act means not 'dishonestly' but that there is a deliberate offering of money or other favours, with the intention that it should as meaning deliberately offering money or other favours, with the intention that it should operate on the mind of the offerree, so as to encourage him to enter into a corrupt bargain. "
This applies to public bodies (so not the opposition parties or even the Labour Party). It's possible that Blair could argue he was acting in his party role rather than his public one but I think there is a case to answer.
Just the announcement of an investigation would finish Blair.
Posted by: Ted | March 20, 2006 at 10:16
I would agree with you Ted that it looks bad for Blair (Tony) whatever Blair (Sir Ian) chooses to do or not to do.
My contention is that with our record and our behaviour currently, will the votes of those disgusted by the antics of Tony Blair and Labour sleaze in general come to us?
Posted by: malcolm | March 20, 2006 at 10:40
malcolm
I don't think sleaze has helped us - "they're all the same" - but think the ructions in the Labour Party might. I don't hold out much hope of Sir Ian investigating though I think all those (Prescott, Dromey etc) washing their hands might be due to fact that they see there is a possibility of criminal investgation.
It depends on whether after Blair's "Whiter than White" the public accept Cameron as the new, better, detergent. AND it depends on our MPs making sure they keep their business and personal lives clean.
Posted by: Ted | March 20, 2006 at 10:47
Certainly the Lib Dems would hold more than 32 seats on those figures. If, say, they won 50, about two thirds of those would come from us, reducing our total to 223.
As against that, the boundary changes would add about 10 to our total at the expense of Labour.
So the likely outcome would be c.Con 233, Lab 332, LD 50. A decline in anti-Conservative tactical voting could alter this in our favour though.
Posted by: Sean Fear | March 20, 2006 at 10:50
Sean, I wonder whether anti-Tory tactical voting still exists? Does anybody have any empirical evidence that it is still happening?
Posted by: Peter O | March 20, 2006 at 14:25
"Certainly the Lib Dems would hold more than 32 seats on those figures. If, say, they won 50, about two thirds of those would come from us, reducing our total to 223."
Yes I think even if the Conservatives take back some of the marginals they lost to Labour in 1997 they will struggle to win back places like Winchester lost to the Liberals in 1997-due to a combination of disgruntaled Labour voters switching to Lib-Dems, other Labour voters voting Lib-Dem on a tactical basis and the fact that there won't be an 'anti-LibDem' vote to any great extent.
"likely outcome would be c.Con 233, Lab 332, LD 50. "
Tight, but workable.
Posted by: comstock | March 20, 2006 at 14:27
Has anyone found anyone willing to admit they vote Labour recently?
The switchers to Labour claim they regret their switch.
Traditional Labour voters say they don't know who voted for Blair at the last election.
Who are these phantom Labour voters?
Posted by: mac | March 20, 2006 at 14:43
Peter O, it still existed in 2005. Assume that there was a uniform swing across the country from 1992 to 2005. The overall figures in 1992 were Con 41%, Lab 34%, LD 18%. In 2005, they were Con 33%, Lab 36%, LD 23%.
If in each seat, the Conservative vote had fallen by 8%, the Labour vote risen by 2%, and the LD vote risen by 5%, we would now have a hung Parliament, rather than a Labour majority of 64. The outcome would have been something like Con 260, LD 40, Lab 315. The tendency of voters to switch to whichever candidate was most likely to beat the Conservative (and good LD targetting) has hit the Conservatives hard.
We can see that anti-Conservative tactical voting was less in 2005 than 2001, (eg Enfield Southgate or Shipley) but it was by no means dead. Of course, if anti-Conservative tactical voting were to become anti-Labour tactical voting, the Conservatives might make all sorts of unexpected gains next time.
Posted by: Sean Fear | March 20, 2006 at 15:16
I agree that 2005 still saw some (but much less) anti-Tory voting, but the question is is it alive today? I think the local elections (which we might have a stab at forecasting, but is really quite volatile as there are huge amounts of localised factors to take into account) may start to give us some idea whether the anti-Tory tactical vote is still in existence in any meaningful form.
Posted by: Peter O | March 20, 2006 at 15:29
Does anyone else think that there is something wrong if, as this poll suggests, we would get 0.4% less of the popular vote than Labour, but end up with just over two-thirds of their seat count?
I've never been in favour of electoral reform (i.e. a PR system), but if the boundaries are stacked so far against us as this would indicate, what hope do we EVER have?
Posted by: Jon White | March 20, 2006 at 17:15
Jon White.
FPTP - its a bummer, isn't it?
And it ain't going to get better for you boys any time soon, so i hope you like that opposition leather.
Posted by: Orange Booker | March 20, 2006 at 23:37
Rick, as Ted says, people whose houses are paid off are people who are older, and therefore more likely to vote Conservative.
What reason is there for a young couple, that are first home buyers (and coincidentally also disproportionately represented in marginal seats) to change government?
Forget education, education, education. Forget grammar schools, and hospitals. Forget trains, and crime. I think the single most powerful issue that helps to determine how people vote is an individual's perception of how they are doing financially. And for the people we need to gain, that means how they think the Government is impacting on their ability to pay off their house - for most people, a far, far, far more critical aspect to building themselves up financially than tax reform!
I don't think it was sleaze by itself that killed the last Conservative government. The main factor was the searing pain of high interest rates. "Sleaze" was simply the unfortunate garnish.
Of course it's easier for a financially illiterate journo to bang on about that than the dull-but-serious consequences to someone's finances and their home life for having to lose their house because of interest rates.
Posted by: Alexander Drake | March 21, 2006 at 04:50
LOL Orange Booker,
Opposition leather?
You wish...
Posted by: Alexander Drake | March 21, 2006 at 04:52
"Does anyone else think that there is something wrong if, as this poll suggests, we would get 0.4% less of the popular vote than Labour, but end up with just over two-thirds of their seat count?"
On the flip side, under those numbers Labour would have put *on* support (just) since 2005 but *lost* seats.
"I've never been in favour of electoral reform (i.e. a PR system),but if the boundaries are stacked so far against us as this would indicate, what hope do we EVER have?"
If the boot was on the other foot, do you think the Conservatives would take advantage?? In a heartbeat.
OK so you don't like PR, what about 1st/2nd preference?
Posted by: comstock | March 21, 2006 at 11:55
Trends can change - at some point in the future it could be that the boundaries work against Labour and maybe the 2005 distribution of votes would result then in the Conservatives being the largest party or even a majority government.
If you wanted the boundaries to be more favourable to the Conservative Party the best way would be to change the remit of the Boundary Commission to make them take more account of to some extent balamncing constituency size which would increase numbers of rural constituencies relative to urban constituencies.
I rather favour Alternative Vote system but there is a possibility that it could result in a Liberal Democrat landslide, no one would know until such a system was actually tried out - it favours partys that are the least unpopular.
Posted by: Yet Another Anon | March 21, 2006 at 17:03