PolitcsHome have released the full report for their updated marginals polls – it covered just short of 240 seats, with a sample of around 35000. It is a repeated of their 2008 exercise, using exactly the same seats and the same methodology so we can see how the picture has changed in the marginal seats.

Firstly a word about how they (or to be upfront, how I did it, since this was largely carried out by me) did it. The 238 seats are mostly Conservative target seats, held by either Labour or the Lib Dems, though there are also a small group of Lib/Lab marginals and some Plaid and SNP marginals. Last year the choice of seats could have been better – even with 238 seats in some areas the Conservatives got a swing large enough to take them all and we should in hindsight have been polling more distant prospects. Rather than expand the sample, this year we polled exactly the same seats to make sure it was comparable. With a lower Tory lead, we mostly avoided the earlier problems anyway.

Even with a sample of 33,000k+, the number of people in each individual seat is not high enough to give reliable voting intention figures for individual constituencies, so instead we split the 238 seats into 17 groups of seats that shared similar characteristics – so there is a group of Con/Lab marginals in the East Midlands, a group in the urban West Midlands, a group of Con/Lab seats in the London commuter belt. Most importantly we seperated out the seats contested between the Lib Dems and Conservatives (and indeed, between the Lib Dems south west heartland and their more suburban seats elsewhere). Note that the seat predictions in the poll are all based on uniform swings within each group of similar seats – last time lots of people got the wrong end of the stick and through they were based on the 150 or so people in each individual seat, they aren’t, since the sample would be too low.

Finally before we get to the results, there’s how we asked the questions. My theory is that when people are asked voting intention questions they give the name of the party they actually support as their first choice, when in reality they might vote tactically, or be influenced by the particular candidates in their seat. For the marginal seat poll therefore we asked normal voting intention, then asked if this was a tactical vote (just to prompt them to consider it) then asked them to think about the particular circumstances and candidates in their own seat, and how they would vote there. This makes a significant difference, particularly in Lib Dems seats.

So, with that out the way, what does the poll show?

Predicted Conservative majority down to 70.

Last year’s poll predicted a Tory majority of about 145, at a time when they were around 20 points ahead in the national polls. This year’s poll was taken at a time when the Conservative lead had fallen to the mid teens, so unsurprisingly we found a similar drop in Conservative support in the key marginal seats, with a smaller number of gains.

The pattern of support hasn’t vastly changed. The largest swing to the Conservatives is still in the Midlands and the London commuter belt (they are doing comparatively poorly in seats in London itself). Weaker areas are in the North and, interestingly enough, in seaside towns.

Support for minor parties

In most areas the narrowing of the Conservative lead isn’t due to any vast increase in Labour support, but a shift from the Conservatives to “others”. That is, incidentally probably the reason why the seaside towns group is so bad for them – because others are up by 11 points, almost all that to the Greens. The uniform swing in that group is enough for the Greens to take Brighton Pavilion (and given that the support for the Greens is probably actually concentrated in the Brighton seats rather than all the seaside towns polls, that’s probably a very good sign indeed for them in that seat).

Lib Dems holding their ground

A national uniform swing projection for the Lib Dems at the moment normally shows them losing a large number of seats. For example, taking a YouGov poll conducted at the start of the fieldwork for this poll, the national shares of CON 41%, LAB 27%, LDEM 17% showed the Lib Dems losing 26 seats. The PoliticsHome marginal poll last year already showed them doing better than national polls suggested, largely because of the local prompting in the voting intention question. This makes a huge difference in Con-vs-LD seats: about a third of people in Con-vs-LD marginals who say Labour when asked how they would in an election tomorrow, say Lib Dem when asked how they would vote in their seat. There was also some shift from the Conservatives, suggesting the effect of a strong personal vote for Lib Dem MPs.

This year the position is even stronger for the Lib Dems. There is still a swing to the Conservatives in their South-West heartlands, and a uniform swing in those seats suggests 8 Conservative gains. However, in Con-vs-LD seats elsewhere, mostly the more suburban seats, though also places like Westmoreland and Lonsdale and Norfork North, there is virtually no swing at all to the Tories. The overall prediction for the Lib Dems is 55 seats, so down by only 8.

The full report can be downloaded here, the data tables should be up here at some point later on today.


62 Responses to “More on the PoliticsHome marginals polls”

1 2
  1. AW – Thanks for the detailed analysis. On the previous thread there seems to be some concern from posters who seem in general to be looking for reasons to criticise this poll, presumably because of the sharp drop in Tory fortunes. I don’t see why, as it clearly shows a good majority, ahead of the latest overall running projection from the polls and thus confirms the idea that the Tories are performing better in marginals. The drop in projected majority since the last marginal poll is absolutely expected – as you point out, you wouldn’t expect to have national poll leads shrinking from 20+ to mid teens or less without it reflecting on the majority.
    All in all, 6 month out still pretty good news for the Tories, but signs that shifts may still occur and they need to avoid complacency.

  2. AW

    Was there a split between the individual seat polling and the general category polling when making the predictions?

    I’m curious about Yarmouth being a Labour hold while other seats in the ‘seaside’ category, Morecambe and Blackpool S, are down as Conservative gains despite the majorities being larger.

  3. Trouble is, if they polled 34,000 people polled over 240 seats, that makes just 141.6 people polled per seat.

    That’s a wholly irrelevant sample size.

    Of how much value can such a small polling sample be?

  4. @David in France – From AW’s third paragraph – “Even with a sample of 33,000k+, the number of people in each individual seat is not high enough to give reliable voting intention figures for individual constituencies, so instead we split the 238 seats into 17 groups of seats that shared similar characteristics”.
    Given this, and the fact that the poll mirrors the earlier version, overall the results would in my mind be pretty accurate, all though I’m sure we could all find individual seat predictions that look a little odd. Set against each other though, such variations would in theory balance out if individual seat errors were random, and again I would argue that a significantly reduced Tory majority compared to the last time this was done is exactly in line with the national polls. (I note the running projection currently shows a 34 seat majority, while the marginal pol doubles this).

  5. Anthony if these results are based on uniform swings within a group of similar seats, how can Gt Yarmouth (Con swing needed 3.7%) be a Labour hold when the seaside town group swing Lab to Con is 6%?

  6. After the 1992 G Election, pollsters discovered the ‘shy tories’ who would not declare their true voting intentions due to the Tory governments unpopularity, This gave Labour a false lead in the opinion polls.

    Does anyone know if the reverse could be happening now?

  7. Richard & Runnymede – it is indeed all done on uniform swings within those groups, data on individual seats weasn’t involved at all. Looking back at it, I think Great Yarmouth is a mistake and should be a predicted Tory gain.

    Al J – probably not because of the adjustments the pollsters made after 1992 to address the problem. In recent years these have been helping Labour rather than the Conservatives, so yes- there probably are “shy socialists”, but the poll are already including them.

  8. Excellent piece of work. Thanks for this Anthony.

    I never expected a huge Tory landslide, and infact have said repeatedly I thought a majority between 20-40 was on the cards. To be seeing a predicted majority of 70 six months before the election is about where I’d expect to be. By next April/May that will shrink to around 30. A smallish, but workable majority.

    2014 will be the Tories chance for a landslide, but only if they deliver in turning the economy around. Equally 2014 could see the return of a Labour government. It really will depend how Cameron governs.

  9. @Anthony,

    The “shy Tories”, “shy Socialists” issue, combined with the extreme volatility of fortunes we see in the recent series of daily polls, makes me wonder if perhaps the methodology needs tweaking again.

    Is it possible that “shyness” itself is a very volatile variable, and that people’s tendency to conceal their true views depends a lot on the news reporting agenda of the past few weeks/months? In those circumstances, for example, a pollster that was factoring in a structural “shy Socialist” factor based on public hatred for Brown could give a wholly inappropriate weighting to Labour in a poll conducted after an event that had portrayed Brown more sympathetically (say, Sarah Brown’s speech or the onset of the Credit Crunch). Could the shyness move alongside the voting intentions?

    In general I found the poll very interesting and, as a Tory in the Southwest, much more encouraging than some of the views on the ground are. There are some very suprising results, which I suppose is inevitable for a simple “swing vs majority” calculation rather than an actual seat by seat poll. For example, it is generally held in Cornwall that SE Cornwall is the seat most likely to go blue at the GE and yet it is Newquay that changes in this poll. Similarly in my own city of Plymouth I don’t know anyone that expects either of our two seats to be won by the Tories, but the poll shows both going (which would be an incredible result).

    The fuzzy old “Tory Green Initiative” side of me is warned by the prospect of an actual Green MP in Brighton though.

  10. AW

    Thanks.

    BTW please don’t think we’re trying to rip your poll to bits, it really is a most interesting piece of work.

    It’s just that some of us have bets to place so we need to know exactly what we’re dealing with ;-)

  11. Was it ‘shy tories’? in1992?

    How do you know if someone is shy? how do you weight for that.

    Was it not that the pollsters were picking up a disproportionate number of Labour voters in their samples?

    The proof of the current polling is in the eating. The Euros and Locals showed big swings to tories in line with the polls.

    The 145 majority last time always seemed high. The current poll seems not unreasonable, indeed if this is the case going into an election then – always assuming a good campaign – the tories ought to be able to build on it.

    What has changed for me is that the LDs had a bad conference and their numbers came out higher. I suspect they may do better than I thought in the coming GE whereas previously I thought they could see their seats cut by 50%.

  12. Anthony

    Were the 19 Scottish seats in one set? In other words were the “similar characteristics” restricted to that single factor?

  13. Neil & Trevor

    I don’t really like the description shy Tory, since I wouldn’t necessarily accept the reason being “shyness”, but that’s the phrase that’s stuck.

    Part of the reason in 1992 was samples had too many Labour votes (and one reason that was proposed was that Tories were more reluctant to be interviewed – a shyness factor). I don’t think that was shyness (or at least, it wasn’t the biggest factor), and it certainly hasn’t gone away as party strength has shifted. It’s more likely to be demographic or attitundinal differences.

    The other reason was don’t knows – the people who said don’t were then disproportionately likely to actually end up voting Tory in 1992. That reason may be shyness, but it doesn’t really matter, just that evidence from multiple elections shows that people who say don’t know are more likely to stick with the party they voted for the previous time – a tendency which is factored into ICM and Populus’s topline figures.

  14. OldNat – I’m afraid so. It’s the area where I think that assumption is weakest, and if designing it again from scratch I’d have split them up somehow.

    As it is, that’s what we did in 2008, and we wanted to keep things exactly the same to allow comparisons.

  15. An excellent piece of work, at least as far as I am concerned as an interested observer.

    Quite rightly, we should be concentrating on marginals rather than those seats that will not change allegiance in a month of Sundays. There will always be those seats which buck the trend due to the sitting MP being personally very popular or despised, but these should cancel each other out.

    As I recall, in 1979 the Conservatives regained power with a majority of 42; however they needed a ‘popular war’ (the Falklands) to consolidate their position. Wars aren’t that popular any longer but five years is a long time in politics.

    Should David Cameron campaign on the basis of ‘Brace yourselves, this is going to hurt’? On the one hand, he would be credited with treating the electorate as adults but on the other hand it might not be what everyone would want to hear. Furthermore, four years down the line, when it is hurting like hell, the media is likely to have forgotten his honest approach.

    There’s probably never been a better time not to be an MP.

  16. A useful exercise with much useful data .
    I agree that the caveats for interpreting the results from the seat groups to individual constituencues .
    I understand the reasons for polling the same constituencies as the poll last year but it is a shame that this perpetuates some inconsostencies .
    For example the choice of Lab/LibDem marginals is poor , Leicester South is not even a realistic target seat . I also do not see the logic of including the Hants LibDem seats in the South West grouping , they have far more in common with say Lewes and Oxford West .

  17. Anthony

    Thanks for the response.

    I still have problems with the idea that pollsters try to incorporate Scottish data, which derive from a very different political structure into a GB pattern.

    Why does this continue? Is it tradition (that’s the way we’ve always done it)? I”Political” (minimise differences between Scotland and other parts of the GB)? Some other reason?

    Whatever the reason the methodology needs to be re-examined.

  18. @Gin – I think you are probably right with your analysis re big Tory majorities, There is always an uncertainty factor when a party returns to power from a long period of opposition, even if the incumbent is viewed very poorly. It is interesting that there are increasing concerns around the margins of Tory policy now appearing in parts of the media as the GE date looms. Some issues, like Hunt’s attack on the BBC as a threat to democracy and the apparent desire to hand over to Murdoch just what he wants in terms of media regulation, are now causing real waves in the liberal/left sections, and I would expect more similar threads to open up. It gets harder to maintain the shiny unblemished image as the scrutiny levels get tougher. There are uncomfortable issues for Labour and LDs to attack Cameron on if they can get organised, and Labour’s conference suggests they might to a degree. If polls narrow a little more we could be in for an interesting GE, rather than a coronation, and that could get interesting.

  19. Thanks Anthony for the really interesting report & for the reply re ‘shy Tories’

    Is there a way to gage how solid the Tory lead is?

  20. I’ve been saying for about two and a half years that it wouldn’t surprise me to see the LDs take 50 seats which a lot of people have ridiculed, although it wasn’t based on any real evidence. This report seems to confirm that that is still a sensible prediction to make. The main reasons I’ve been saying that was because: (a) I think most LD MPs have a strong personal vote, and (b) I think the LDs will probably poll around 20% which is only down 2.6%, not a huge decline.

    Although the Lab majority is 2% lower in Poplar&Limehouse than Hammersmith, I think Hammersmith is more likely to be a Tory gain, and I’m sure quite a lot of other people would think the same; but I’m aware that the uniform swing has to be applied in an exercise of this kind (within each group of seats).

  21. Thanks to Anthony and YouGov for this excellent report.

    One of the most striking things in the report is the drop in the Labour vote in some areas such as:

    London commuter belt: -16%
    Labour’s southern bastions: -14%
    Urban West Midlands: -19%
    West Midlands hinterland: -15%
    East Midlands: -16%

  22. Alec – the poll shows the Tories doing worse in the marginals not better, with a predicted majority of 70 against a swingometer prediction of 92 based on the 41/27/19 national figures at the time. Looks like the LibDem vote holds up much better in seats where they have a chance.

  23. Anthony
    Many thanks for an interesting report which will keep me busy for a day or two.
    There seems to be discrepancies between the maps and the charts. For example Nottingham South appears as Labour on the map and Conservative in the charts; there are other similar examples. I live in Brigg and Goole constituency and feel that Goole itself does not fit into the East Midlands area very comfortably; Goole is of course former West Riding and now in East Yorkshire. I appreciate that the Brigg area is more at home in the East Midlands. The former Humberside constituencies tend these days to be included with Yorkshire and sometimes even in the North East !

  24. ‘2014 will be the Tories chance for a landslide, but only if they deliver in turning the economy around. Equally 2014 could see the return of a Labour government. It really will depend how Cameron governs.’ GIn

    Does one ever get a landslide when one is in Government? It would have to be extremely unlikely. Landslides occur when everything is going right for an opposition and wrong for a govt. Once in power you start to alienate people as not all dreams can come true, people start to get bored with certain faces and outside influences impact on the country. As such, a govt may do well-and perhaps as well as before but I can not see it as at all possible that they go from say a comfortable majority to a landslide the following election.election…

  25. Jack, Margaret Thatcher, 1983. Unlikely, but entirely possible.

  26. @Jack – 1983 stands out in my mind. A not very popular and somewhat scorned Mrs T defeated the much more popular Jim Callaghan by a reasonable but small majority in 1979, and then swamped the opposition in 1983. Labour in 2001 also achieved a second massive majority, although this was not an increase on the previous win.

  27. The scottish figures are misleading and seem to have been taken in such a way as to make the figures look less conservative than they may otherwise be- They have skipped the margional scottish seat currently held by the current tory mp, and ye have included other far less margional seats (moray for example) this must totally skew results in favour of snp votes and against tory votes. As such it makes the scottish results totally worthless, whith political manipulating having been the basis of the poll rather that statistical evidence.

  28. Superb piece of work.
    Whilst no doubt there will be some inaccuracies the overall majority projected will be closer than national polls and UNS.
    It does mean we are 2-4% swing either way from noc or a land-slide and we have an interesting 7 months in store.
    Any chance of a repeat early March?
    Re 1983, was the con vote share not down despite the Falklands? It was the SDP breakaway and high alliance score that produced a land-slide.

  29. I am sure that the Tories will be either the largest party in a hung parliament (about 25% likely IMO), have a working majority of up to 70 (50% likely IMO), or have a large majority of 70+ (25% likely IMO).
    Mrs T’s 1983 landslide was not only based on the Falkland’s War, but an opposition that was in complete disarray, like The Tories in 2001.
    The broad thrust of this poll points to an outcome that feels right to me at the moment. A second Cameron Government 2014 might not get a landslide if The Labour Party learns from its own painful past. There are enough old heads in the party who will should not allow the post 1979 bloodshed and civil war in The Labour Party to occur. However, I do feel that the changes in style and policy required by The Labour Party to become electable again may take sometime.
    Great work Anthony

  30. Jack,
    Surely landslides were obtaine by incumbents in 1959 – 1966 – 1983 – 1987 and 2001.

  31. Thank you Anthony for a very worthwhile analysis of an important polling exercise. Reading both together is obviously essential to appreciate the implications.

    In terms of longer term forecasting I’m not at all sure it is sensible to think about 2014. If Cameron wins, even with a small majority, he will undoubtedly change the boundaries and reduce the number of seats. The consequences of that cannot, in my view, fail to be very significant indeed.

    In polling terms I suggest there will need to be a serious review of forecasting and in some cases methodology whilst comparisons with previous elections and polling may have a very limited value indeed.

  32. Graham is right. Landslides by incumbents is not unusual at all.

    I’m looking forward to the time when each marginal might have about 1,000 YouGov volunteers willing to answer polling questions, so we can get accurate predictions for every seat. As Anthony says, in this poll the average was about 150 per seat. It’s just a question of gradually increasing that number.

  33. 1,000 Yougov volunteers in each marginal seat would almost certainly give inaccurate forecasts as they would not be a statistically accurate cross section of the electorate as a whole .

  34. Notwithstanding the Scottish caveat, this is the most reliable projection for Scotland that I have seen and an excellent piece of work.

    I said to a PPC the other day that if there were fewer than ten changes overall, that I would be the least surprised person in the country.

    Nobody doubts that the SNP will take Ochil, and the Conservatives, Dumfries but Alex Salmond’s pre-Glasgow East claim of a total 20 or Conservative claims of up to 12 have never seemed credible to anyone who looks at each constituency to consider whether a change is likely.

    There will be fewer than four SNP gains in addition to those predicted which will surprise only those who live outside the constituency. Similarly, the Conservatives have three other winnable seats and they might reasonably hope to gain one of them. That would show that they were on the way back, but the should not be disappointed if they gain only one.

  35. The variation in the change in the Conservative vote in different categories of seat is interesting. For example in the East Midlands the increase is 11% but in the category entitled “Conservatives v LDs elsewhere” the Tory vote is said not to changed at all since the 2005 election.

  36. Makes sense… as Labour weaken; so do the Tories, and people feel safer voting for the LDP, and Others.

    I wish the Greens every success in taking Brighton, and perhaps a couple of others… more diverse voices in parliament can only be a good thing… though I think the GP could do with some better quality candidates.

  37. If you look at Neil’s list of SNP marginals on the Ochil site, you can see that they are in third place and in many cases up against the LibDems. Labour, not the LibDems, will be the big losers in this election, and the vote for rural incumbent LibDems is more secure than any other.

    Yet under FPTP, the SNP with a share of the vote catching up on Labour, because it has its support more widely distributed in four party Scotland, gains far fewer seats.

    Next time will be different, there will be many more marginals, a Conservative government and Labour in disarray even more so than at present and not looking like an alternative government after a parliament which will probably last less than the full term.

    The possibility then is that the SNP may have FPTP working in their favour and that they could win a majority of the Scottish seats. That was their pre-devolution dream: a majority of Scottish MP’s demanding the recall of the Scottish Parliament.

    The person who can do most bring that about is not in Scotland, but the next Prime Minister. He needs Ms Goldie at his elbow every waking hour if he is to have any chance at all of avoiding that outcome.

  38. John B Dick

    Isn’t Scottish politics delightfully complex! Not only does it have a fault-line (Independence) which isn’t congruent with party, but an experience of tactical voting which would be difficult to analyse – even if we had polling methodology not based on the Tweedledum/Tweedledee dichotomy!

  39. Aye, the threat of a Tory victory should do wonders for the SNP – what good are Labour to the average socialist Scot if they are demolished in Westminster south of the border?!

  40. It looks to me like Anthony answered Oldnat’s question about why it continues to be done the same way in Scotland–because they want to keep the same for comparison purposes even though it’s “mince” as far as accuracy for that nation.

    I wonder if that’s really a good idea. From the point of view of a foreigner, it’s an interesting poll and I’m not criticizing it in general. But I’d love to see results in Scotland that actually was more meaningful. That’s where I lays my money down if nothing else. Maybe I should just flip a coin. ;-)

  41. OldNat

    You are right. After three SP elections the electorate are splitting their votes every possible way. They fancy a pragmatic and competent SNP government for the SP but don’t want independence and don’t see the point of sending an SNP MP to the Westminster playground.

    I first noticed so-called tactical voting in the North East fifty years ago and it has spread steadily since. I often wonder if any of the parties has more than 10% voting for it rather than against the others. Here in Argyll, the SNP claim not to have found ANY positive LibDem voters, yet the LibDem MP has a majority that would bring comfort to many facing re-election.

    The electorate is quite content with minority government in the SP, but the party leaderships of the three Westminster parties, both in London and Edinburgh havn’t got the hang of it with the exception of Annabel Goldie and even she disappointed over Megrahi.

    It’s a pity “Alex Salmond for First Minister” so damaged the Greens and that both sorts of socialists are spending more time with their lawyers, but they will be back. I want my grandchildren to live in a country with a parliament where there is a place for the likes of Denis Canavan and Margo Macdonald.

    Independence would not be my first choice, but it is now clear that I will not live to see the second part of Donald Dewar’s vision as he explained it to me when we were at school. The Scottish Parliament is not just for the better governance of Scotland. Donald also intended that it should be the model for the reform of Westminster.

    If you are ever in Rothesay I can tell you what he DIDN’T say to the Nationalist who put the “slippery slope” argument to him. Later he said that “Scotland will be independent when people vote for it.” Can you see Blair/Howard/Brown etc (take your pick) saying that?

    I used to try to pick a different argument with him every week to support my case that no person of intelligence and integrity could keep his place and his sanity as a Westminster MP. In over a hundred conversations I deplored every one of the abuses that are spoken of today and others like the Trade Union block vote that are gone.

    He always had an solution and that solution without exception is now to be found in the Founding Principles of the Scottish Parliament, its procedures and processes and the values on the mace.

    Alex Salmond’s only contribution to the expenses debate was to propose copying the practice of the Scottish Parliament. Well might he do so. That’s what the SP is for.

  42. The Tory share fell in 1983 yes,

    1979 13,697,923 44.9%
    1983 13,012,610 43.5%

    but Labour suffered a big loss of votes to the Alliance, who increased their vote but very evenly and in a very poorly targeted way, thereby creating a Tory landslide.

    (Labour still had 11,532,218 in 1979 – similar to the two elections in 1974)

  43. The total Labour vote in 1979 was actually up very slightly, by 75,000 votes.

  44. Yes it was.
    I think in the churn of things, they did get some more core voters to the polls.
    The Tory increase seems to have come from turnout, from the Liberals, and from C2/C1 voters mainly.
    I know two former Labour councillors who remember canvassing in 1979 and they were however convinced they lost votes to Thatcher directly, but clearly, Labour did manage to find some votes from somewhere else.

  45. Andrew

    Your desire for comfort by finding a favourable spin from your point of view cannot be sustained in view of Anthony’s explanation above.

    The tide ois flowing from Labour to the SNP at the moment and the Conservatives are in a backwater and hardly affected.

    For the same reason the Conservatives sole seat is not particularly at risk and I don’t have it very high up in my list of possible changes.

    Next time, with a Conservative government, it could be different couldn’t it? If a Conservative government is elected with double the number of MP’s from Scottish constituencies, do you think it will pay attention to what they have to say?

  46. “Even with a sample of 33,000k+, the number of people in each individual seat is not high enough to give reliable voting intention figures for individual constituencies, so instead we split the 238 seats into 17 groups of seats that shared similar characteristics”

    That’s quite a heavy generalisation, so I’d take these results with a pinch of salt to be honest.

  47. Very useful work.

    Is there anywhere where I can see a list, rather than the interactive map, of the seats that will be won by Labour?

  48. “last time lots of people got the wrong end of the stick and through they were based on the 150 or so people in each individual seat, they aren’t, since the sample would be too low.”

    I’m slightly confused about this – why did PolHome offer up individual seat polling data for sale and use the “example” of Brighton Pavilion in which the sample was nearly 300 and solid conclusions were drawn from the individual seat data?

    http://www.politicshome.com/Landing.aspx?Blog=2961&perma=link#

    I realise there is slightly oblique disclaimer at the bottom, but still…

    Is the whole data set, broken down by individual constituency, going to be available this time, even if (given the sample size) it’s “just a bit fun”? Forgive me if it’s already up – I clicked the link above and coudn’t see it, but I only had a quick look around so might have missed it…

  49. Anthony,

    Thanks for an excellent job.

    This is far more useful than extrapolating opinion polls (evwn WMA) onto existing seats using UNS.

    As others have commented, Scotland is harder to deal with. The problem there is that it really is a separate country (no that does not mean I am in favour of cottish independance) with its own regional variations and patchwork quilt of 2, 3 and 4 party contests. Frankly I don’t see how it can dealt with to the level of accuracy some people expect without polling at least 500 electors in each and every seat.
    That would give us about 30k polled just for Scotland.

    The position in Wales on the other hand seems far closer to that in England than in Scotland. Maybe that explains why Plaid are not leading a minority administration in Cardiff.

  50. Question – last time round the report was based on the groups of marginal seats same as this time.

    Back then PoliticsHome also marketed the splits for individual constituencies to candidates, PPCs and so on, on an “as seen” basis. This time round I think they’re only selling it in bulk, but it’s really a question for Freddie.

1 2