Rod Crosby asks: Will Regional Swings Help the Tories?
April 16th, 2008
Is five seats the most that Cameron can hope for?
There has been much discussion on whether regional swings may provide a hidden bonus for the Tories at the next election. In particular, so the logic goes, the Tories are underperforming in Scotland, so they must be doing above average in England (and/or Wales.)
True enough, but how can we measure the differences, and calibrate any possible benefit? In truth we can’t, for the simple fact that regional poll breakdowns are based on samples of about 100 people – in statistical terms, garbage….
But in the absence of reliable poll data, we can still make reasonable hypotheses, based on quantifiable aspects of the electoral system, and previous experience.
The above graph of Labour’s battleground seats shows, in rank order, the number of marginal Labour seats where the Tories are second-placed to Labour. I have split the seats into those requiring up to a 5% swing, and those requiring up to a 10% swing. A 5% swing would surely see the Tories as the largest party, forming a minority government.
Notice the big difference in the sizes of the orange-coloured columns. In the North-East, no seats at all would fall on a 5% UNS, while in the East Midlands 11 seats would fall. Our own Nick Palmer’s place on the battleground is noted (NP on the graph). The red 10% bands broadly follow this pattern too (apart from the North West, which has a surplus of seats in the 5%-10% band)
What would be the optimum swing pattern for the Tories across the regions?
It should be intuitively obvious that to maximise their gains the Tories should hope for a relatively low swing in the North-East and a relatively high swing in the East Midlands, with swings rising across the graph from left to right. In other words, swing perfectly correlated to the number of marginal seats in each region. For example, in the North-East it makes no difference whether the Tories get a 5% swing, or no swing at all - no seats will fall, whereas in the marginal-rich East Midlands an extra 5% swing would deliver another 7 seats on top of the 11 UNS gains.
How likely is it the regions shown on the graph will magically line-up in ascending order of swing? Those familiar with permutations will realise the answer is 11!:1 due to chance alone.. (or about 1 chance in 40 million…..) In other words, so unlikely it’s laughable…
But, the swing would not have to be perfectly correlated for the Tories to derive some benefit….. So long as it was broadly increasing left to right - even if a few regions bucked the trend - the Tories might make some gains over and above UNS. How many? That is very difficult to answer algebraically - there are too many constraints and permutations. The method I have used is to run thousands of simulations, generating thousands of random regional swings, subject to differing levels of constraints.
What constraints? Well, not all swing patterns are equally likely. Despite regional differences, the nation still broadly moves as one. London is not going to swing 10% to the Tories while the North West swings 10% to Labour. Or even 5% in opposite directions. Such divergences have never occurred in a British General Election, and there is no evidence things are about to change dramatically.
Two common measures of dispersion are Range and Standard Deviation. Range is simply the gap between the highest and lowest regional swing, and standard deviation is an overall measure of how different the swings are from one another. If every region followed the UNS both the range and deviation would be 0, and there would be no regional seat bonuses. Here is a table of these statistics over recent elections.
One thing to notice is that the past two elections have been significantly more uniform than previous elections, with about half the range and standard deviation exhibited during the period 1987-97. Why that is I’m not sure, nor can I predict whether it will continue that way or reverse. I don’t have the stats prior to 1983, but the consensus is elections were relatively uniform until the 1980s.
The model uses the maximum range and standard deviation thus far exhibited in a British general election, with increments up to these levels. Thousands of random swings have been generated within these constraints, and the net seat difference summed across the regions. If ever there was a case of swings and roundabouts, this is it!
In all tests the average net bonus is close to zero, and 95% of results show a difference of less than plus or minus 6 to 11 seats, depending on the range and deviation selected. The higher the deviation and/or range the slightly larger the maximum potential bonus or penalty.
The 95% range of results can be interpreted as a confidence interval. In other words, only about once in every 80 years might we expect a party to obtain a regional swing differential in excess of 11 seats, and the odds indicate it will be considerably fewer. For example, John Curtice calculated that the Tories only obtained a 2 seat bonus in 2005. In 1992 Labour obtained a 5 seat bonus due to regional swings.
A political party clearly does not have the power to force the regions to swing in a way that best suits them. Such outcomes are determined by long-term trends, third-party performance, differential turnouts and a huge amount of randomness. But, as posters have noted, there is a hint the Tories have at least one data-point in the right place. In Scotland the swing will in all probability be below average next time. But look at “next-door” Wales. The long-term trend there is in the Tories’ favour. So if Wales swings above-averagely, that would largely cancel out the benefit obtained in Scotland.
Looking at the top end of the scale, the East Midlands, North West and West Midlands would pay rich dividends if the Tories could boost their performance here. Unfortunately, for the past three elections they have all swung nearly identically to the nation as a whole, and previously when they have diverged it has been in opposite directions! Again, the natural tendency is for variations to cancel out.
So the answer to our question seems to be. Yes, regional swings might help the Tories a bit, but they largely cancel out, and are insignificant in the overall context. Perhaps 5 seats may be attributable to regional swings next time.
MessageSpace Advertising
Thanks Rod - That might need to be read a couple of times but seems to make sense… Presumably when running the models you assumed a Normal Distribution with the Range / Standard Deviations shown?
1, I second that - bit sleepy at the moment. However, looks like a very in-depth analysis. Great work Rod:)
Thanks Rod.
Very indepth analysis.
I suspect we’ll see quite a big swing to the Conservatives in the East Midlands. I don’t think Brown will resonate at all in this region. The West Midlands and North-West England will probably see smaller swings.
1. Correct, using the NORMINV(Rand(),swing,SD) function in Excel, and applying a range constraint (max 8.0, decreasing for each test). Compare the total number of seats that would fall on a national UNS with the total if the regions swung uniformly by the randomly generated regional swings. The difference is the regional bonus or penalty. Run 1000 times for each swing, range, sd. Plot histogram. Calculate averages, sd for results. Estimate confidence intervals.
Rod, many thanks for another very in-depth and comprehensive article.
Excellent anakysis; I hope CCHQ are reading this before deciding where to plough resources in at the next election!
@previous thread/422:
I think that’s implausible too. The ones most likely to care are those least likely to vote, by some margin.
I also reckon their might be a surge in BNP support amongst the white working class who, having voted for another party for Mayor and constituency, will put BNP for their party list to get them over that magical 5%.
For anyone who wants to make a statement on May 1st, a BNP list vote would be by far the most dramatic.
Excellent stuff Rod. Just to confirm terminology, am I right in thinking that a swing of 10% turns over a 20% majority?
Also looking at Scotland (since I don’t think there were boundary changes in Scotland since 2005) I reckon there were only 7 seats where the tories were second to labour in 2005 with the following majorities
Dumfries and Galloway 5.7%
Stirling 10.9%
East Renfrewshire 14.0%
Edinburgh South West 16.5%
Ayr, Carrick and Cumnock 22.2%
Ayrshire Central 24.3%
Ayrshire North and Arran 25.6%
This would suggest that 1 Scottish seat is susceptible to a 5% swing and 3 to a 10% swing. Am I missing something here as the chart shows more. Do you include seats that fall on a Lab-Con swing where Lab-Con are not 1 and 2 ?
Thanks Rod - excellent analysis and just the kind of reasoned analyis that has made this Site essential reading for those interested in political preditions, for betting or any other purpose.
I can’t fault your logic and I certainly can’t challenge your stats!
Thanks again.
Well put together Rod,thank you.
Articles like these will also help the Tories cause.
Palestinian apologist for suicide bombing is among the leaders of a group trying to win Muslim votes for Ken Livingstone in the London mayoral election.
Dr Azzam Tamimi, a supporter of Hamas which is dedicated to the destruction of Israel, is part of a group called Muslims 4 Ken which is aiming to mobilise hundreds of thousands of Muslim voters to help re-elect Mr Livingstone for a third term on May 1.
A website set up by the group vilifies Mr Livingstone’s Conservative opponent Boris Johnson, portraying him as an Islamaphobe.
Very interesting Rob. Thanks.
At the last election, I seem to remember the Conservatives did much better in terms of seats than their national swing would have suggested - mainly due to quite a lot of unexpected wins (well, unexpected to me) in and around the South East. Is it the case then now that any such benefit the Tories have got from that approach has now been squeezed to the maximum? Or that the marginals lined up by region better for them last time? Or was it more due to differential swings within seats rather than within regions?
I’m quite surprised that elections have been getting more uniform in terms of swing - I thought conventional wisdom said the opposite?
4. I concur with this.
Gotta love those Monte Carlos…
O/T Iain Dale running a story where Alistair Darling has criticised GB and the govts way of doing things whilst on a trip to china.
All falling apart at the seams it appears.
http://www.iaindale.blogspot.com/
15. Are we doomed ?
14. Perhaps Darling know the writings on the wall for himself?
Labour do disasterously in the locals, Darlings removed in the resuffle, with Chancellor Balls put into Number 11! Are we just days away from the Brown/Balls dynasty?
16 - Maybe he is practicing for when he gets to knife Brown in the Commons a la Lamont!
What was it? In office but not in power?
O/T Naughty !! Naughty !!
Home Secretary Jacqui Smith was accused of breaching election rules today by making a major anti-terrorism announcement during the run-up to local polls.
Ministers and government departments are supposed to observe a period of silence - or “purdah” - in the weeks before a vote.
The system is designed to prevent the party which is in power from having an unfair advantage during an election campaign.
But Ms Smith today announced 300 new police officers, community support officers and back-room staff will be moved to new duties combating radicalisation.
Conservative shadow communities and local government secretary Eric Pickles reported Ms Smith and other Labour ministers to the Cabinet Secretary for breach of Whitehall election rules
16. It wouldn’t surprise me if Brown axes Darling in a reshuffle, to be honest. In a desperate ‘pin the blame on him not me!’ guesture.
In that scenario, looks like Balls for the Chancellorship (!!!)
I would love Darling to do a Howe, though I don’t think he’s the kind of person who would do it…
19. Conservative stunt!
19. It’s amazing the stupidity that pervades this government.
Excellent article Rod - thanks very much indeed. I agree with post 6 - this should be compulsory reading for whomsoever is planning Tory strategy at CCHQ.
Very good article.
Rod, nice article, mate.
Rigorous and sound. It also marches with the wisdom that the Midlands are often where elections are won and lost.
Cookie at 11: The last elecions “bonus” was mainly due to the way that the seats lined up in order of majority - there were more within a short range of the UNS on the “just over” side that the “just under” side, so when a normal probability distribution randomly around the UNS is taken into account, for 2005 you’ expect a “Tory bonus”(Curtice predicted it in his famous “dead-on”prediction of a 66 majority). I’ the basis for conditional probability seat calculators (Rod and I have separately produced versions) and is unlikely to have as drastic an effect next time.
8 Rod - looking at it further, I wonder whether you have included seats like Edinburgh South, which would the Tories would overcome Labour on a 5% Lab-Con swing but which currently has the Lib Dems in second (and in fact they would win the seat on a 5% Lab-Con swing)
2005 result Edinburgh South
Labour 33.2%
Liberal Democrat 32.3%
Conservative: 24.1%
SNP: 6.2%
Other: 4.2%
Majority: 0.9%
21.No Gabble, its a Labour stunt, apparently the headline announcement becomes very vague and generally falls apart when questions are asked about specific details. Eric Pickles has a valid point.
Is it not also equally important to analyse the distribution of swings within each region and also to look for any possible correlation between size of swing and size percentage Labour majority?
Please read “It’s the basis of …” for “I’ the basis of..” in my last post - am posting on my mobile phone while being lurched around on the Metropolitan Line.
I see Harriet Harman has been reprimanded but not reported to the police over the donations ’scandal’.
http://news.bbc.co.uk/1/hi/uk_politics/7351154.stm
A pretty hilarious stance by the commission, essentially along the lines of “Yes, you broke the rules. Yes, that’s against the law. But we’re not going to refer you to the police because all you deserve is a nice little slap on the wrist.”
Why do we protect our politicians so?
19/21 - I thought purdah for local elections applied only to announcements directly on local government matters or having an impact in particular authorities up for election. Terrorism is a grey area as arguably the principal threat is to London. But this does look rather cooked up by the opposition.
Very meaty Rod. And interesting.
Matt1.
You have your wish [last thread] on Brown’s daily embarrassing moment.
Five Live were laughing AT him and his corny comments today on US TV. Roughly what they read out:
‘I love America. I love your TV. We get some of it you know. Thank you for your TV and its contribution to our culture’…
‘Oh dear’ was the anchor’s comment.
Hope they run it here, complete with dazzling inappropriate smile.
Urrgghhhh. Nasty.
28. In fact I should have referred to the correlarion between size of swing and size of majority whatever pary holds the seat.
21. astroturfer!
21.27.
I thought it was a dodgy move by the Tories [even though justified]. No doubt Labour just thought they could do what they liked as its on ‘terror’ and noone would dare to complain [they were right about the LDs].
Pickles made a very good calm matter of fact defense of his position on the BBC.
Smith’s decision was made a while ago and so could have been announced earlier.
It wasn’t urgent and so could have waited [it has so far].
They could have got on with it anyway and then announced it after the locals.
Some of the announcements related directly to areas covered by the locals.
Labour: all spin, no shame.
That’s an interesting analysis - thanks, Rod! your conclusions sound pretty sensible, too. I’ve always thought that analysing polls by brooding on regional variation is a bit of a mug’s game - the local samples are often unbalanced, and if we’re doing a bit better in area X compared with nationally, we’re doing a bit worse in area Y, and it tends to even out. The exception next time could be Scotland, if the impact of having the SNP in local power has a general effect, whether positive or negative.
32. Oh dear oh dear. Just been surfing and found this transcript from the Red Box, presumably from the same interview:
GMA: “You were a television producer, television reporter,” she said, as if he only quit Scottish TV yesterday rather than 1983. “How do you like the other side? Do you ever wish you were here (on this side of the camera)?”
GB: “I spent a little time as a television reporter, I was actually a university lecturer but I worked with Tony Blair in Parliament for 25 years and we worked very closely together on the environment and the economy and to be honest I think you’ve got a difficult job as a television producer. (laughs)
GMA (bemused): Harder than your job?
GB: “Our job is day to day, 24 hours and everthing else. But yours is a difficult job as well and I appreciate it.”
re32 Sally it’s on the BBC’s 6 o’clock news - it is pretty nauseating too.
McCAIN’S UP; OBAMA’S DOWN
——-
The latest Rasmussen Reports daily tracking survey shows McCain narrowly beating both Obama and Clinton. More important, it finds that 53% voters view McCain favorably, while 45% view him unfavorably. Obama’s ratings are 49% favorable and 49% unfavorable, while Clinton’s are 43% favorable and 54% unfavorable.
…
One in four Clinton voters would go for McCain if Obama wins the nomination, while about one in five Obama voters would back McCain if she wins the nomination, according to one survey.
Even worse for Obama, if his small-town slam and his ties to Rev. Wright hurt him in the upcoming primaries of Indiana, Kentucky and West Virginia, it could mean he’s doomed in the Electoral College.
As John Judis writes in the New Republic: “To win in November, a Democratic presidential candidate has to carry most of the industrial heartland states that stretch from Pennsylvania to Missouri. That becomes even more imperative if a Democrat can’t carry Florida - and because of his relative weakness in South Florida, Obama is unlikely to do so against McCain.”
Judis estimates Obama would have to win a minimum of 45% of white votes in the heartland states, and a majority in some to build an electoral majority.
But in Ohio, a battleground state that Clinton carried by 10 points, an exit poll shows Obama got only a third of white voters.
…
Obama lost nearly all the Ohio districts to Clinton that determine who wins the state in the general election, some by as many as 25 points.
Of course, the past doesn’t guarantee the future. That’s what campaigns are about. But they do show that the soaring Obama candidacy is coming back to Earth and landing in an all-too familiar spot for Democrats.
http://www.nydailynews.com/opinions/2008/04/16/2008-04-16_dems_enter_the_dead_zone.html
————
8. I was using Anthony Wells data, and I thought I went for Lab 1st, Con second, although on the face of it for Scotland I must have included the two near third-place prospects. Any errors and omissions are entirely mine. It doesn’t materially alter the picture, in the event. Regarding most of the non-Scottish seats, they are the subject of notional calculations, but I am certain while there may be small differences at the seat level between the various sets of notionals, the overall shape of the distribution won’t change.
11. John Curtice found that on two Tory gains could be attributable to regional variation. The rest of the Tories’ “unexpected” gains were due to the overall uneveness of the marginal distribution.
See http://tinyurl.com/6zg7uj
The overall standard deviation of swing has remained pretty constant since 1974 at around 3.0 (+/-0.4). This anaylsis is only trying to disentagle regional effects from any other effect, eg tactical voting.
37. What? Does Brown listen to what he’s saying or does he tune out?
27. No ChrisD, its a Conservative stunt:
“…But Chris Huhne, for the Liberal Democrats, said it would be “churlish” to criticise the announcement as a breach of election law, as the money had been announced last year, in the Comprehensive Spending Review. ”
http://news.bbc.co.uk/1/hi/uk_politics/7350300.stm
42. Oh well if the lib dems don’t agree it musn’t be a stunt. ot perhaps they’re just doing whatever labour wants them too, again.
10
Yes,great publicity for Livingstone plastered across all the ES billboards.
Interesting ES article by Paddick ‘Ken Livingstone is a nasty little man’,I wonder what he did to piss off Paddick so much.
28/33. Yes, but that analysis is for another day! This one merely isolates the overall regional effects, per the Curtice and Steed methodology.
If anyone has a copy of “The British General Election of ….” (for years other than 2005 and 1992) by Butler and Kavanagh, could you look in Appendix 2: the Results Analysed ? They usually give in the text, under Regional Variation, their estimate of the regional bonus in that election. I would be interested to know what they found in other years. I only have 2005 and 1992 data.
35.”I thought it was a dodgy move by the Tories [even though justified].” Sally, for too long we have been letting the Labour party away with this kind of stunt. Sky wove this story into the mix as well when reporting it, gave the credit for Labour pulling the link to the blogsphere.
42.Is Smith Playing Politics With The Police?
“Adding to my suspicion of a political stunt, I am told that Home Office civil servants were unable to tell political correspondents from regional newspapers where these 300 police officers are to be deployed.”
Thanks for the analysis Rod.
Latest Gallup Presidential and Primary Trackers :
McCain 44% .. Clinton 46%
McCain 43% .. Obama 46%
Clinton 42% .. Obama 50%
http://www.gallup.com/poll/106567/Gallup-Daily-Obama-Maintains-Lead-Heading-Into-Debate.aspx
Incidentally for those who wanted to watch the mayoral debate, I notice that it’s available here
39 Phillip are you having a laugh, most of the polls today indicate Obama ahead of Clinton nationally and he has apparently gained a Super Delegate from Indiana.
At this stage gauging a General Election result is difficult until the Democrats have one candidate, that may not be Obama or Clinton.
New ARG Primary Poll for North Carolina :
Clinton 41% .. Obama 52%
http://americanresearchgroup.com/pres08/ncdem8-703.html
Rod - how do you explain the above average swing to Labour in 1997 in London and the South-East?
Your graph clearly shows this. This also helped them win some extra seats.
Also, do you not agree the whole notion of a two-party “swing” (for that is what “swing” is supposed to be, a swing from ONE party to another ONE) is clearly complicated by the introduction of a third-party, or even fourth-party?
I have seen stats on swing which paint a very misleading picture.
Most of us think “swing” to be a direct swing between Labour and the Conservatives. However, this isn’t what it appears to represent in some cases.
Take St.Albans:
2005 RESULT
Anne Main Conservative 16,953 37.3 +2.1
Kerry Pollard Labour 15,592 34.3 -11.1
Michael Green Liberal Democrat 11,561 25.4 +7.5
Richard Evans UK Independence Party 707 1.6 +0.2
Janet Girsman St Albans Party 430 0.9 +0.9
Mark Reynolds Independent 219 0.5 +0.5
Majority 1,361 3.0
Turnout 45,462 70.0 +3.7
SWING 6.6% from Lab to Con
Sorry, but Labours vote declined by 11.1%. Only 2.1% went to the Conservatives. A whopping 7.5% went to the LibDems.
Or in other words, Labours vote declined by 24%. Of that decline 68% went to the Liberal Democrats and only 19% went to the Conservatives. 13% went to others.
Or, put it another way, only 4.6% of all Labour voters in 2001 defected to the Conservatives in St.Albans.
Either way, it does not add up to the 6.6% LAB to CON swing the BBC report. If the Labour voters had not deserted in such large numbers to the Lib-dems, Labour would have held the seat and the reported “swing” would be drastically less.
My point is that the definition of “swing” can be misleading and more seats could fall to the Conservatives on a haemorrhaging of the Labour vote than you suggest in your article. As Nick says, don’t forget the SNP either.
I don’t say you’re wrong, but I’d be very surprised if the regional swing benefits failed to break into double-figures at the next election.
I think another big part of the reason why could be (and this is my opinion, not scientific) because I believe Gordon Brown is a competely toxic product to swing voters in the marginal constituencies, particularly in the South, and I think he will have the REVERSE effect of Blair in 1997.
If I could add me tuppence worth….
When building predictive models of consumer choice for commercial clients we find ‘regionality’ to be a fairly minor factor. We also find that true regional patterns of behaviour tend not to follow the Government Office Region (GOR) boundaries that you are using in your analyses.
What people call ‘regionality’ is usually driven by an underlying difference in distribution of the true forcing factor. In the jargon the data is confounded. This is not to say that region is unimportant, it is just usually not as important as one may imagine. Some markets it does dominate (ale vs lager) but in others it is almost completely unimportant (Tesco propensity).
Building a predictive model based on true forcing factors behind differential swing would be an interesting exercise and eminently doable. Once build you could incorporate Monte Carlo simulations. However my hunch is that you will arrive at a similar conclusion. ie differential swing will only account for +/- c10 seats above and beyond national picture.
46. BBC News: Serious analysis, by Frank Gardner, of the policy to introduce 300 police officers to combat terrorism.
His report followed by a sentence on the mealy-mouthed objection to it’s announcement - the tories look truly pathetic.
re 52 I would say that in a single seat the idea of a national swing from one party to another is completely meaningless. What Rod is trying to get across that with 650 results if you apply the uniform national swing to all of them then you’d pretty much predict the correct result to a handful of seats. Obviously you’d get some seats wrong - like Bob Marshall Andrews last time - but over the whole country it pretty much evens out. Election after election for the last 60+ years have borne this out.
That’s why my money -as I’m sure Rod’s is - is on a hung parliament next time.
I would buy McCain to win the general election.
I get the feeling he will win Ohio, Florida and probably PA.
That ARG poll on N.C is not bad for Hillary considering the AA population for Democrats in North Carolina is near 40%
Coming to a screen near you ….
Hillary Cannot Be Stopped !! Certificate 18
http://www.huffingtonpost.com/2008/04/16/hillary-clinton-cannot-be_n_96986.html
Talk about undermining your political message:
Labour’s latest broadcast was just on ITV - Gordon Brown talking about how the change has just begun and about letting Labour have the chance to continue their good work….followed by ITV News headline that repossessions could be as bad as the early 90’s and Darling’s attack on the government.
46
The additional 300 police according to Jacqui Smith are to ‘address radicalisation’(whatever that means).
Maybe they could make a start with Livingstone’s election campaign manager.
O/T Sky news reporting that the electoral registration process in London has not gone well.
Less people on initial reports registered to vote than in 200 or 2004.
Sky are reporting it as possible bad news for Ken
See it here http://adamboulton.typepad.com/my_weblog/2008/04/london-a-signif.html
56 Jaz. Its a better poll for Hillary, much in line with the last ARG. However most of the other recent polls indicate a slightly larger Obama win. If the data set provides the info look at the turnout filter for black voters. Broadly it should be in the 150% - 160% range.
IMO Obama will win in the 20-25% range.
http://www.realclearpolitics.com/epolls/2008/president/nc/north_carolina_democratic_primary-275.html#polls
Like others, I will have to look at the detail of Rod’s analysis. However, I am convinced that the real issue will not be whether there are regional swings but the extent of any improvement in the Conservative share being concentrated in seats with a particular psephological profile.
It will be the performance in England that will determine whether the Conservatives form a minority or majority government, if indeed they do so. In short gains in Wales and Scotland will be any icing on the cake.
The Conservatives have 98 English seats where their share of the vote in 20% or less - Labour have 91. Even if the share of these were to increase uniformly by 10%, then there will be very, very few gains if any from these seats. Such an increase would imply the share at least increasing by 50% to more than doubling. Over 80% of these seats are in the North West, North East, Yorkshire & Humberside and London. Most are either inner city or are / were heavily industialised. Some of these seats in reality could see very little improvement or even a reduction in share for the Conservatives for a variety of reason, particularly if there is a strong anyone but labour vote.
If the improvement in any Conservative national share is concentrated in those seats where their share in 2005 was 25% to 45%, then detailed work suggests this could result in substantial gains in those seats over and above those expected from any uniform national swing.
54.Gabble, the introduction of 300 police officers to combat terrorism is to be welcomed. But you are also trying to defend blatant politicking at election time, and lets not forget the looming 42 day detention battle.
The faux outrage when anyone dares to criticise this government’s motives and timing of these announcements just doesn’t work anymore. Its as cynical as Brown’s trip to Iraq during the Conservative conference to announce troop withdrawals, another headline which did not stand up to scrutiny.
Noticeable move back to Boris and away from Ken since the position after the debate last night.
Latest - Boris 1.61, Ken 2.7.
JackW, indeed, the black population will make sure Obama wins big in NC. He could not compaign there at all, call North Carolinans dumb, and still win.
64-After this voter registration story there should be a big move back to Boris.
If Ken has not persuaded the BME vote to go and register then the advantage is surely in BoJos court.
60. Yes - I’ve posted on here before that despite a substantial increase in population in England since 2000, the total electorate has barely increased at all.
There is a rapidly increasing number of people not registered to vote and not surprisingly these are disproportionately located in areas of:
1) high mobility
2) low income
3) poor education
4) people not speaking English
ie London will have a lot.
52. I think what you have to remember is swing is a useful measure for this reason.
Whether or not there is a direct “swing” between two parties is irrelevant. Why? Because under FPTP it’s the plurality (the margin of victory) that counts, not any particular vote share. That is all that has to be overturned for a seat to change hands. Within a wide range of outcomes it matters not whether Labour fall 4% and the Tories rise 4%, or Labour fall 8% and the Tories stand still, that is a 4% swing in both cases… All 8% Labour majorities will (in theory) be overturned.
Consider the 1983 general election. The Tory vote fell by 1.5% points…. but the Labour vote fell by 9.5%. Result: a 4% “swing” to the Tories, producing a landslide which whould have been near-identically replicated if the Tories had risen by 4% and Labour only fallen by 4%…
Although I agree with you that there are variations around the mean. These tend to be normally distributed and thus largely cancel out.
62 Richard - welcome to the forum, and an excellent point.
15% swing from a low base (say 25%) in the North West could yield much larger numbers of seats than we might otherwise suspect. My hunch is that if the Tories are to win more than a slender majority (and I doubt they can win a majority at all), it will be from seats which they lost something like 46%-25%(Con)-14% in 2005.
52 I need to spend some more time re-reading your post. It makes a crucial point about how we define ’swing’, but I need to mull over whether I agree with you or not. Great post.
53 Completing my ‘complements quota’ for the day, the commercial angle on regionality is fascinating, and something that fascinates me, so thanks very much for making that point. I assume you work for a consultancy of some description - what sort of area? Marketing/Sales or Supply Chain?
65 Jaz. The importance of the size of the win in NC will be measured in the delegate count. In Penn it would appear that a Hillary win in the 55/45 range will net her around +10 delegates. It’s highly likely that Obama will more than erase that figure with a big win in NC. In addition there is much speculation of a bulk NC super delegate announcement after the result ; likely then not a great day for Hillary !!
“But look at “next-door” Wales. The long-term trend there is in the Tories’ favour. So if Wales swings above-averagely, that would largely cancel out the benefit obtained in Scotland.” - Ok. So say Wales swings 5% to the Conservatives and Scotland 3% only. How exactly would that stop the Scottish eats that might go on a 3% swing from going just because the Tories might have a 5% swing in Wales
Blimey. My head’s still hurting after reading that. I’m surprised that the swings have apparently been pretty uniform at the last 3 elections, as one of the points always being made is that Tories have only made progress since ‘97 in the South East.
Also seems to show that the Tories did relatively well in Scotland and Wales in ‘97, but very badly in London and the South East.
Let me be a simpleton here:
Where are most of the swing voter types?
In the known swing seats, not the usual heartlands.
If the Tories were were getting swings in the known battleground swing seats of say 10%-12% how many many seats would they win with say:
A 10% swing in them all?
A 12% swing in them all?
63. ChrisD. This announcement is not directly related to local government and Smith was careful not to identify individual areas to avoid influencing the local elections.
It’s a useful reminder of just how creepy the tories actually are. Once again, their desperate opportunism has been exposed.
JackW, I get the feeling that if she can do better than +10, and then win Indiana, it would ultimately balance out the gains in North Carolina.
Then there is Kentuky and West Virginia, both should be leaning Clinton. Puerto Rico is Clinton also. Obama after NC is Oregon, and he will have to make some kind of victory in one of the dakotas on montanas.
I personally have no wish to see Obama in the white house, I have deep sceptism of people who believe they are change makers because ultimately they fail, including Cameron…eventhough I’m conservative.
69 Re-paste, “What odds the LDs failing to take Cardiff Council (resurgent Tories and a useless leader)? - Hmm say 15% chance. No more. Because while I think you’re right the LDs maybe stalling I think Labour are collapsing so fast they will benefit by default. The Tories hold all the seats in Cardiff North anyway and will annhiliate their opponents there. After that it’s tricky to see even with resurgence where they go. It’s possible their vote may edge up in central but to delivering seats or taking them from Lib Dems doubtful. South Cardiff err no still stony ground for them. That leaves really just the Bay area and Cardiff West. In the latter I think could be where if anywhere the Tories surprise, but as I say I think Labour’s problems nationally are of such magnitude that Cardiff will drop most likely into Nick Clegg’s lap anyway. Thoughts?
by Punter April 12th, 2008 at 5:03 pm
Excellent article Rod.
This is scientific journal material.
“A boy was cautioned by police after pretending to sneeze and then wiping his hand on David Cameron’s jacket as the Conservative leader visited Sussex. ”
http://news.bbc.co.uk/1/hi/england/sussex/7351320.stm
The data used to draw the squiggly graphs can be found here..
http://www.titanictown.plus.com/regions.xls
78. Well he’s had his 5 minutes now then.
The boy, that is.
62. Excellent post. Agree entirely.
68. Thanks Rod, but I don’t see how it all cancels out?? Do you disagree that Labour did disproportionately better in 1997 in London and the South-East? Wouldn’t Portillo have held his seat otherwise?
There is a significantly larger Conservative activist base in both London and the South-East.
Surely you don’t think this campaigning base makes no difference to seat retention and ability to fight and win marginals?
Maybe regionalism is the wrong phrase to use. Maybe the correct analogy is the strength of the Conservative activist base and residual Conservative vote.
I don’t know.
72. “Also seems to show that the Tories did relatively well in Scotland and Wales in ‘97, but very badly in London and the South East.”
You see, I find this point peculiar, seeing as the Tories lost ALL their seats in these areas in 1997.
I’m no statistican; although I am educated to university level in Maths, very rusty; but I would have thought it clear that marginals swing much more than safe seats and a high activist base, more so.
re 66 I seem to remember that it used to be a criminal offence not to register and not just a matter of persuasion. Is that no longer still the case?
76 Strangely Punter , I have never rated the chance as very high of the LibDems taking overall control of Cardiff council . If you look at the results closely from 2004 , there is in fact only 1 ward Butetown which Labour won where the LibDems were close and here there is a strong Ind presence . I have no knowledge of how the campaign is going or which wards are being targetted by which party .
75 Jaz. With recent conflicting polls Indiana is too close to call and will likely not net either candidate any significant delegate traction and essentially the other races will cancel each other out or give Hillary a small boost in delegates.
Effectively in terms of pledged delegates the game is up for Hillary. She must rely on the remaining super delegates to provide her with such a one-sided lift that all present evidence shows is not likely to happen, indeed quite to the contrary as events since Super Tuesday have shown.
Hillary is the ‘Micawber Candidate’ - Something will turn up. Except it hasn’t and in all likelyhood isn’t going too. To all intense and purpose, despite all the froth she is playing Huckabee to Obama’s McCain.
re 78 well Gabble they’re your laws. He’s probably banged up under some anti-terrorism law.
46 The BBC you say. Really! Not being nice to the Tories, you say.
Sky seem to have her number - this time, and last time and the time before that.
But if she gets away with it its fine. Great. New Labour. Our rules [they brought these in] are always for others, not us.
I understand Gabble.
The Times has reported that the 42 day detention is causing concern amongst Labour MPs because it was only done to make the Tories look weak on terror and its backfiring.
Oh but lets give them the benefit of the doubt for this time and the next and the time after….
68. Sorry Rod, forgot to add, I understand your justification for the term “swing” in your post, but I’m not sure I agree with how it’s defined.
For example, I don’t think there *WAS* a swing to the Tories in 1983/1987, it’s just the opposition divided itself so as to give the appearance of one.
I think it’s highly misleading to say there has been a Lab to Con swing when the overall vote for the governing party declined, on a similar support.
It hardly suggests an increase in its popular support, as “swing” does, does it?
Swing made much more sense in the 1950s and 1960s, when it was straight up Conservative vs. Labour - no funny business.
65. Which suggests that blacks are voting in the Dem race overhwlemingly on race.
If whites did that in this case they’d be accused of racism.
Fantastically impressive……
@74:
Astroturfing at this hour, sir?
Look, Labour have been caught out indulging in their usual cynical electioneering stunts, like Brown’s trip to Iraq during the Tory conference where he lied about troop withdrawals.
One of the things really killing Labour now is your inability to let it go, even when you’ve been busted. And busted you are.
You were caught out. Admit your cynical error and move on.
86 should say 54
83 Yup. But all you seats are so safe now save Cardiff North that your Councillors can play charades while your activists focus solely on attacking Labour
83. I’live in the city and can’t say I’ve heard much. The council don’t get a huge amount of criticism in the local media, which is probably a good sign.
81. Looking at the figures again it may be that the ’supposedly’ strong performance in Scotland and Wales in ‘97 was just a consequence of the ground tey had to make up after ‘87, when they did very badly in both regions. After ‘87 the Tory performance got a lot worse nationally, but only a small amount worse in Scotland and Wales, so in relaitive terms they did well in those regions in ‘92 and ‘97.
42: ‘…But Chris Huhne, for the Liberal Democrats, said it would be “churlish” to criticise the announcement as a breach of election law, as the money had been announced last year, in the Comprehensive Spending Review.’
Well Mr Huhne should tell that to his leader cos Mr Clegg has just been lambasting the announcement on C4 News.
74
Get over it,you don’t do the indignation bit well,sounds very false.
92. Quite.
Once again “swing” is misleading. As is relying entirely on a pure statistical analysis to understand what is happening.
And, the most important point, the Tories lost ALL their Scottish and Welsh seats in 1997.
93. oh dear oh dear oh dear. Huhne says no problem, clegg says there is, whoops!
http://www.usatoday.com/news/politics/election2008/2008-04-15-obamainside_N.htm
Obama’s claim of independence questioned.
96. Yes, Clegg described it as a re-announcement. If he believes that, how does that breach the rules??
Relentlessly bad coverage for the government in ITV news this evening.
Incredibly dumb politics by the Tories. The more they go on about the announcement by Labour about more police against terrorism, the more it gets impressed on the public that Labour are doing something about terrorism. The public aren’t concerned with the niceties of parliamentary convention.
An excellent analysis below for Penn indicating a net gain of 12 delegates and 120K in the PV for Hillary :
http://www.fivethirtyeight.com/2008/04/pennsyvania-prediction-clinton-to-net.html
12 Blimey stjohn, I’m impressed - you’re knowledge of Statistics rivals mine of Italian!
99. Typically dishonest coverage of the ‘credit crunch’ by ITV news.
They featured a tragic story of a mother who had to give up her job to look after her daughter who has been diagnosed with leukemia. Now the family have no income and cannot pay the mortgage. They are living under the threat of repossession.
This story is worthy of coverage and raises issues as to what level of help and support is/should be available to a family in these circumstances.
However, it has absolutely nothing to do with the ‘credit crunch’ and it is dishonest to portray it as such.
Punter - 15% chance of the Lib Dems failing to win Cardiff Council seems about right. I think it would be exactly what we expect if they secure it (and I agree Labour’s collapse in the city makes it easy), but would be a major failure for Nick Clegg if they fail to - it would be a significantly missed target, at the same time that Clegg has nailed his fortunes to the ridiculous target of doubling parliamentary seats.
Ignored if he does win Cardiff, damned if he doesn’t methinks.
Gabble struggles to get out of the pit he and the LDs have dug for him.
Is Hulne mounting a leadership bid.
What were the odds on Cameron beong the only one in place at the next GE
61/56 - Its ARG… a good rule of thumb with any poll they produce is to toss it, utter joke of a pollsters frankly.
Interesting trend in PA (where we continue to get a deluge of polls) between some showing a rebound for Clinton and some continuing to show that much more narrower margin we’ve seen emerge recently, of course Obama is outspending her massively in the state but you’d expect the ‘bitter’ comments to bite him (although it’ll hurt him most in Appalachia - where he already was very weak, hence why he’s likely to lose the WV and KY primaries by wide margins).
NC (bar something unexpected) favours Obama heavily, though it remains to be seen if he manages to wrack up the kind of margin he did in neighbouring VA and SC… meanwhile IN, no one seems to have a clue about, with polling infrequent and somewhat contradictory.
As it is I think gaffes or no gaffes there is a good chance that Clinton wins by an insufficient margin in PA (although it’ll be worth seeing what kind of legs the ‘bitter’ remarks have, although so far they don’t seem to have had a big impact, despite the Clinton campaigns best efforts). Another interesting bit of research that came out recent (don’t have a link) suggested that even with a fairly emphatic win Clinton wont net more than 5-10 delegates out of PA. Following PA, Clinton will probably struggle on to NC and IN, but a big Obama win in NC seems likely making IN key, Clinton will need a convincing win there and I’m not sure she’ll get it…
71. I could have expressed that a bit differently. What I mean is:
imagine a line superimposed on the first graph. It starts high on the left and falls to low on the right. That would be the Tories getting a high swing in the North East and a low swing in the East Midlands. Result, the Tories would not get as many seats as UNS suggest. They would be penalised because their high swings were in places that had few marginal seats, and their low swings would be in places that had lots of marginal seats. Not the pattern the Tories want. They want the exactly opposite pattern, rising from left to right (in step with the columns.) They’d get a lot more “bang for their buck.”
If the swing was a straight line across the regions, they’d all swing the same as the average so there would be no bonus or penalty above or below what UNS predicts. If it was an essentially random line that showed a flat trend, then also there would be no significant difference from UNS.
Now, if you’ve followed the above, you should see that the regions in the left half of the graph should have low swings, and the regions in the right half should have high swings (from the Tories point of view.) But Wales and Scotland both being in the left half, if one goes low and the other high, that breaks the pattern, and thus tends to cancel out the Tory-favourable “trend.”
Since it’s zero sum, all votes have to add-up to the same however they’re distributed. Doing (relatively) well in Wales has to be at the expense of doing badly someplace else, and that place could deliver more seats than Wales could. So by the Tories doing (relatively) badly in Scotland they have a chance to do relatively well someplace else. But Wales doing the opposite would tend to “cancel out” the Scottish benefit.
Hope that helps. Maybe someone else can put it better?
re 64.. quote
Noticeable move back to Boris and away from Ken since the position after the debate last night.
Latest - Boris 1.61, Ken 2.7. end quote
As a matter of interest,historically speaking are Bookies odds on political betting a more reliable guide to the outcome than say odds on horses etc???
101 - Just saw, Jack W has provided the PA delegate calculation of which i spoke…
Rod and Andy Cooke - thanks for your answers to my questions - very interesting.
107 - Makes perfect sense to me, but then I understood what you meant initially…
Rod Crosby:
In other words, only about once in every 80 years might we expect a party to obtain a regional swing differential in excess of 11 seats, and the odds indicate it will be considerably fewer. For example, John Curtice calculated that the Tories only obtained a 2 seat bonus in 2005. In 1992 Labour obtained a 5 seat bonus due to regional swings.
What happened in 1997? As Casino points out there was a significantly above UNS swing in London (5%) and there was a significant change in seats :- 1992 Con (48) Lab (35) to 1997 Con (11) Lab (57).
How likely is it the regions shown on the graph will magically line-up in ascending order of swing? Those familiar with permutations will realise the answer is 11!:1 due to chance alone.. (or about 1 chance in 40 million…..) In other words, so unlikely it’s laughable…
Hypothetically, that may be true but the regional polling information that is available (you can pooh pooh it all you like) does suggest that the Conservatives are making proportionately greater improvement against Labour in the Midlands, North and London than they are in the South. Given as you point out the North East is a Labour Stronghold it would suggest that any gains will appear either in Yorks/ Humber or the North West.
So here’s a scenario for you. The South, labour is down to core vote, swing voters either vote Libdem or Conservative. Polls suggest they are swing ing to the Conservatives
Rod:
Posted prematurely I’ll continue shortly….
Rod - The various English regions are delineated artificially for convenience, but am I right in thinking that where there has been a discernable difference in the extent of a political shift between the North and South of England, it has been the case that the Midlands area, taken as a whole, generally falls in between the two? Simple logic suggests this would be the case, unless there are special factors at play, for example regional Government grants or, on the negative side, major regionally based job losses.
109 Ben. There’s also this one :
http://www.electoral-vote.com/evp2008/Comparison/Maps/Apr15.html
103. Thats what I thought. But anyway, it was probably the most ghastly ten minutes coverage for this government that they have had.
So here’s a scenario for you. The South, labour is down to core vote, swing voters either vote Libdem or Conservative. Polls suggest they are swing ing to the Conservatives
Midlands. Libdems have never done so well in the Midlands. Current polls suggest swing voters moving from Lab to Conservative.
North again swing to the Conservatives from Labour. Harder to identify which region because polls do not differentiate that well but intuitively it won’t be the North East.
Scotland little change in Con vote
Wales little change in Con vote.
I know it is all intuitive but I do not think we can totally diregard the polling information. yes it is little more than anecdotal but it is there.
To follow on from 117.
So that suggests
Midlands above average swing
London above average swing
North above average swing (but smaller)
South below average swing
Scotland no swing
Wales no swing
Good lead Rod- pbCOM’s highest IQ
101 Jack W - The net gain of 12 delegates is actually slightly less than Rod Croby’s projection of a 16 delegate gain, based on his 55%-45% assessment of her victory.
Morale boosting yes, but it really doesn’t change the bigger picture at all.
Marc Ambinder reporting that super delegates Reps Mel Watt and David Price of NC will endorse Obama later today :
http://marcambinder.theatlantic.com/archives/2008/04/reps_watt_and_price_to_endorse.php
118 - There will not be no movement in Scotland or Wales. In Scotland at least the situation will be complicated by the fact that the votes will be going every which way and the Conservatives could end up with seats simply because votes are switching all around them.
72/81. Yes, the Tories did relatively well in Scotland (1992 and 1997) and Wales (1997 and 2001.) But you are forgettng the magnitude of the swing in 1997. Despite the relatively good Tory performance, the absolute performance was still dire, and resulted in the loss of all their seats. The Scottish performance in 1992 is an example of why the Tories don’t want to do well in Scotland (at the expense of doing well someplace else.) A massive rebound from the “Thatcher effect” 1979-87 produced only 1 gain (Aberdeen South.) If those votes had been in the East Midlands or the NorthWest instead, Major’s majority would have been higher than 21, and would have consequently saved a lot of grief.
Sorted courtesy of The London Communications Agency article in the Evening Standard !
The London Communications Agency report suggested the result would depend on second preference votes.
Crucially, it could turn on whether Lib-Dem Brian Paddick endorses either of his main rivals, which he has so far refused to do.
It comes after a YouGov poll for the Evening Standard showed Mr Johnson’s lead over his Labour rival had halved in a week. “Both candidates can win,” the study said.
“Ken needs to explain how his record to date has benefited London and adopt a more consensual tone of voice. He needs to remind people of the cheeky chappy Londoners liked in the first place. Boris needs to convince voters that he is up to the job as well as just being likeable.” It concluded that only the Labour or Tory candidates stood a realistic chance of winning.
London local government expert Tony Travers said: “There is some evidence there is a small lead for one candidate but this is so small you have to call it a ‘to the wire’ contest.”
The LCA said there would be 10 new faces on the London Assembly although its overall make-up would remain almost the same - nine Tories, seven Labour, five Lib-Dems, with the Green party up one to three seats and Ukip gaining one seat.
I refer any budding statisticians to my question at 73.
119. Not really. Statistics are fairly straightforward if you do them regularly and have a rudimentary understanding of Excel.
I don’t think a man who denies the holocaust and calls Winston Churchill a “bastard of the first order” can be described as having pb.coms highest IQ either.
120 PfP. Indeed. Tinkering at the edges but effectively the games up !
72/81,
It’s very possible for the Conservatives to do “relatively well” in swing terms but still lose all seats. If 8% swing will lose every seat and the UNS is 12%, beating the swing by 3% in that region may provide some moral comfort - but you’d still be eradicated (despite doing “relatively well” in terms of regional variation.
New Turning Point Presidential Poll for Colorado :
McCain 52% .. Clinton 40%
McCain 51% .. Obama 39%
http://schaffervudall.blogspot.com/2008/04/schaffer-udall-tied-in-latest-poll-at.html
39 Monsieur
Hope you read Jack W at 48. Must be so dispiriting for Obama haters everywhere.
Malcolm
115 - Cheers Jake that’s actually very interesting, the relative margins in the Philly burbs and places like Scranton will be interesting.
131 - Sorry, dont know here ‘jake’ came from…
RodCrosby - Is there not some danger in mixing up ‘relative’ with ‘absolute’.
I mean surely in absolute terms, the Tories DO want to do well in Scotland. Clearly if the latest polls suggest the biggest shift to them has beeen in Scotland Wales and the North East, it’s not that helpful.
After ‘Huffington’ broke the ‘bittergate’ story Hillary had better hope this ’screw em’ story doesn’t get the same traction :
http://www.huffingtonpost.com/2008/04/16/hillary-clinton-on-workin_n_97017.html
……………….
132 Ben. Your ‘ding a ling’ ??
126 - Can we please not go there?
97 - See this article from Daily Kos
http://www.dailykos.com/storyonly/2008/4/5/10050/11057
I don’t disagree that the Obamamania needs to be countered with the downside, but I hope you don’t mind if I open-mindedy critique the attacks on him, which I think (up until now) have been pretty unfounded.
74.”63. ChrisD. This announcement is not directly related to local government and Smith was careful not to identify individual areas to avoid influencing the local elections.
It’s a useful reminder of just how creepy the tories actually are. Once again, their desperate opportunism has been exposed.”
Did you watch C4 news tonight Gabble? A Labour stunt.
The Pope comes to town …. woof woof.
http://www.outsidethebeltway.com/archives/2008/04/pope_benedict_our_german_shepherd/
124 I think Paddick’s attack on Ken “Nasty Little Man” Livingstone rules out him suggesting his second preferences go that direction
http://tinyurl.com/6xgkd9