A few people have asked me if I know where there is a spreadsheet of the general election results available so they can crunch the numbers and explore results themselves. Until now I’ve been using results scraped off the BBC website, but the British Election Study team have now released a data set of the election results for download here.

420 Responses to “Spreadsheet of the General Election Results”

1 2 3 9
  1. Thanks very much, Anthony, as ever.


    ^^^ This is just for anyone who missed my comment on a previous thread. (It’s especially for your benefit, Crossbat111).

    Thanks for the link, Anthony. Interesting stuff as always.

  3. Robin
    as my dear departed mother used to say:
    ‘pride comes before a fall’ and ‘people who blow their own trumpet seldom have any friends’.

  4. Just to say that the site’s RSS feed hasn’t updated for a while – not since before the election, in fact! Initially I thought you were taking a well-earned rest.

  5. @Chris Lane
    Hope you’ve had a good day Chris. I really miss posters like
    Cross bat, Amber and of course the pups. Hopefully they’ll be back soon. The place isn’t the same without them.

  6. Are yes, good point at @Valerie, where are those posters like Amber who predicted the Labour victory and death of the Tories. They seem to have disappeared for a bit…seriously I hope they come back and give us some more insight.

  7. Valerie

    I’m sure that your Mum was right about lots of things – but Louis Armstrong had masses of friends.

  8. Robin Hood

    Seeing as the polls were wrong (and wrong for how long?), what method are you using to to measure the autocorrelation constant and what was the standard error you discovered?

    I think the model is a reasonable one for something so simple, although I couldn’t say if it was definitely better than a random walk model as quite simply noone seems to have established the standard errors of the parameters. I don’t feel the need to shout it from the rooftops that it’s somehow 100% correct. There are plenty of parts that look questionable (LD and UKIP scores in particular)

    I’d be highly suspect of going back and “redoing the polls” with a methodology that gave the right figures to see if their model worked. It might be an interesting exercise to see where the polls and reality diverged but quite simply as a forecasting model it might forecast where VI is going, that isn’t really what the team was hoping to predict!

  9. Oldnat,

    I’d also add from the previous thread that there is a possibility that the Tories might pick up some right-wing SNP voters, especially if the SNP don’t promise a second referendum.

    2011 was very disappointing for the Scottish Tories at Holyrood, and 2015 was a curate’s egg (increased vote, falling vote share) so 2016 is a big test for them.

  10. RICH
    “Are yes, good point at @Valerie, where are those posters like Amber who predicted the Labour victory and death of the Tories. They seem to have disappeared for a bit…seriously I hope they come back and give us some more insight”

    I’m not surprised they have disappeared for a bit for they owe me an absolute fortune in Taxi fares ;-)

    Seriously I do miss Amber and BatFink

  11. RICH
    Yes but its the same old thing. In fairness, the polls were so wrong one could certainly be forgiven for expecting a tie, but all this OTT pro Labour stuff which has been totally blown out of the water, was just silly.

    I just want to read some reality in future on the site.

  12. Just for the record, I’m as disappointed as anyone about the result. I am, after all, a Labour Party activist.

    However it needs to be remembered that some people on the front page gave me a pretty rough ride when I tried to explain ‘swing back’ theory and now they’ve all gone as quiet as mice.

    Sometimes swing back occurs gradually (e.g. 1986-7), other times it occurs at the last minute (as in 1992 and 2015).

    Also please do not be fooled by the propaganda of the poll-bashers: the polls were right all along, it’s just that 3% of people changed their minds at the last minute. That is not a massive swing.

    For those who say the polls were incorrect then please explain how they got last year’s Euro elections right when they used exactly the same methodology?

    The fundamental thing is that people behave differently in the sobriety of a general election polling booth to the way they behave when answering a hypothetical voting intention question or registering a protest in a mid-term by-election (as the ex-MP Mark Reckless has found out to his cost).

  13. Rich, Roland,

    One to amuse you, but can you guess who tweeted this at 1:55 AM – 8 May 2015:

    “Still see no reason, whatsoever, to change my view that David Cameron’s career is over.”

    Followed by this a mere 4 hours later

    “Voters have spoken, Cameron wins. Voters always correct imho. Was wrong to place faith in polls. Sorry I got it wrong folks. Off to work now”

  14. @Rich

    I don’t know if anyone predicted the death of the Tories. If they didn’t die after 1997, it wasn’t going to happen in or after 2015.

    It is also (for much the same reasons) premature to predict the death of the Lab party now. For a start, Lab only lost 26 seats, and had a net gain against the Tories, whereas in 1997 the Tories lost 178 seats – most directly to Lab who won 145 seats.

  15. JM has resigned …..developing

  16. Thanks very much for the link.

    I was planning on crunching numbers today but the PSU on my PC died. So I went to see Mad Max instead.
    Hopefully once I get my machine back I can get some numbers to post here.

    I’m mostly interested to see how well the Tories did in the South, especially the South West.
    If there’s no coming back for the LibDems there then it’ll be almost impossible for Labour to push the Tories out of power.

  17. ROBIN HOOD :


    Survation-yes it has !
    Anthony Wells-No it hasn’t !

    Fascinating-and totally inconclusive !”


  18. VALERIE and all; Good Evening to you, and I also miss Nick Poole, of London, who predicted a landslide.

    The only thing on polls and swing back is the latest polls seemed to be showing a pro Labour swing.

    Maybe the Stone and Ed’s fall metaphorically and physically on the leader show played a role in swing back.

    Apparently at the hustings today Dan Jarvis was impressive.

  19. @Colin


    The only thing we know for sure if that the Tories have a majority. At this stage do we need to know anything else?

  20. Er sorry 3 hours late with that one-teletext is a bit slow these days.

    Will he return ,prob not ,when next tory leader does the deal with nicola it will be 2018.So 2020 plus before normal bread and butter politics resumes.


    Unlike you I am not a Labour sympathyser but I certainly agree with your post above, especially the last paragraph, a point I made time and again over the last year.


    “Also please do not be fooled by the propaganda of the poll-bashers: the polls were right all along, it’s just that 3% of people changed their minds at the last minute. That is not a massive swing”

    Totally agree with you and I also said the same.

    However the polls were almost spot on in terms of Scotland because as one of the panel members correctly said on election night…”polls are better at picking up avalanches”

  23. RAF

    I don’t -but I suppose anyone who intends to visit this regularly in the future will wish to be assured that it is providing polling results which are actually measuring public opinion.

    One imagines too that the Press & others who commission & pay for these polls will be asking questions about value for their money.

    Finally-as Lord Foulkes indicates on that podcast-politicians may take an interest in the accuracy of the 2015 GE polls if they fell that voting behaviour was unreasonably affected by inaccurate polls.

  24. RAF

    Totally agree, I think Labour are far from finished.

    I think the next five years are going to be difficult for all the current top three parties.

    The Tories managing a small majority with such major problems to face up to.

    Labour in finding a way forward that will have mass appeal.

    and the Scots Nats who are probably going to be on the way down from their high point after 2016.


    I don’t think the 2015 figures for the Tories in Scotland are very indicative of what may happen in the future. In Ediinburgh many long standing Torries voted for other Unionist parties to oppose the SNP e.g.a lot of Tories in Edinburgh West told me that would as a one-off tactically vote for the Lib Dems & less logically (in my view) I believe many Tory voters assisted Labour’s only successful candidate in Scotland.

  26. @ Alan

    I think they have been wrong ever since they thought they established the pattern of the LibDem deserters, and probably used this for sampling and cleaning the sample. But the latest is probably summer 2014.

    I’m still only at Dewsbury (going alphabetically … ) – 153 constituencies, looking at patterns of LibDem breaks. It’s statistically not valid, of course, as I’m making judgements on the basis of past election patterns (e.g. systematic ABT or ABL votes, holding up voting shares in previous elections, etc.).

    What I can see is that polls probably couldn’t cope with the huge variety of LibDem voters. While it certainly happened that Labour lost votes to UKIP and gained LiBDem, but in most cases the scale of it suggests straight LibDem UKIP switch (especially in the North), and I haven’t seen any discussion about that.

    Hawthorn’s remark (and an article I read) suggests that at the price newspapers likely paid, they got their “value” in terms of sampling or research design – both online and phone.

  27. Regional results that I’ve calculated so far:

    Greater London region:

    Lab: 1,545,080 (43.69%)
    Con: 1,233,386 (34.88%)
    UKIP: 286,946 (8.11%)
    LD: 272,544 (7.71%)
    Greens: 171,652 (4.85%)
    Others: 26,622 (0.75%)
    TOTAL: 3,536,230

    Lab: +7.07%
    Con: +0.35%
    UKIP: +6.37%
    LD: -14.39%
    Greens: +3.26%
    Others: -2.65%

    Swing, Con to Lab: 3.36%

    North East region:

    Lab: 557,100 (46.89%)
    Con: 300,943 (25.33%)
    UKIP: 198,823 (16.73%)
    LD: 77,095 (6.49%)
    Greens: 43,051 (3.62%)
    Others: 11,201 (0.94%)
    TOTAL: 1,188,213

    Lab: +3.33%
    Con: +1.60%
    UKIP: +14.03%
    LD: -17.08%
    Greens: +3.30%
    Others: -5.18%

    Swing, Con to Lab: 0.87%

    East Midlands region:

    Con: 969,379 (43.46%)
    Lab: 705,787 (31.64%)
    UKIP: 351,777 (15.77%)
    LD: 124,039 (5.56%)
    Greens: 66,239 (2.97%)
    Others: 13,201 (0.59%)
    TOTAL: 2,230,422

    Con: +2.28%
    Lab: +1.89%
    UKIP: +12.50%
    LD: -15.26%
    Greens: +2.45%
    Others: -3.86%

    Swing, Lab to Con: 0.20%

    West Midlands region:

    Con: 1,098,110 (41.77%)
    Lab: 865,075 (32.91%)
    UKIP: 412,770 (15.70%)
    LD: 145,009 (5.52%)
    Greens: 85,653 (3.26%)
    Others: 22,322 (0.85%)
    TOTAL: 2,628,939

    Con: +2.23%
    Lab: +2.30%
    UKIP: +11.70%
    LD: -14.94%
    Greens: +2.69%
    Others: -3.98%

    Swing, Con to Lab: 0.04%

  28. Stella Creasy standing for Deputy Leader.

  29. @AndyJS

    You’re slow! I had all these calculated by Sunday last week….

  30. The electorate change figures make interesting reading from those figures.

    For example if you look solely at the most marginal Lab / Con tossups in the Midlands / East-Midlands / East of England / SE / SW / West midlands and Yorks and Humber then the electorate dropped by about 46,000 voters in the top 60 seats, despite record levels of immigration over the past 5 years.

    The picture is very patchy – so 2000 less electorate in Morley and Outwood (2015 maj 422), 9000! less in Lincoln (2015 maj 1443) and 7000 (2015 maj 3082) less in Halesowen and Rowley Regis.

    Did the new individual voter registration system help the Tories in some of these seats and disproportionately harm Labour?

    Overall the electorate is down 100,000 on 2010 but that hides the fact that the electorate in Scotland has increased almost universally, making the reduction in England and Wales much greater – arpund 520k.

    Certainly the electorate figures for Scotland are much higher almost across the board.

  31. Bill Patrick

    I was quite surprised that so few right wing Unionist (but SNP voting) folk seemed to have switched in the existing SNP seats Con change since 2010

    Dundee East -0.2%
    Angus -1.9%
    Banff and Buchan -2.0%
    Perth and North Perthshire +2.2%
    Moray +5.1%

    I had expected more.

  32. @ Robin Hood & others.

    It’s an equally valid argument that 3% of people did not tell the truth when asked their VI for what ever reason, although I still support the ‘easy life’ theory.

    If there was a ‘swing back’ then what event or series of them caused it, and why did it only persuade such a small amount of floating voters?

  33. @ Gary O

    In the first 153 constituencies (in alphabetical order) the result was heavily influenced by absolute turnout, although it affected the outcome only in three cases – 2 for Lab one for Con.

  34. 8 of the 153, where is the edit button for tablets?

  35. Re the previous point and the size of the electorate – net migration stats suggest that the electorate should have increased by over 1 million people since 2010.

    In actual fact it decreased by 100,000 (500,000 in England and Wales).

  36. Evidence perhaps that at least 1.5 million people were disenfranchised by individual voter registration?

  37. @ COLIN

    Well, I’ve often suspected that dear Anthony was a swing back denier.

    If there was really no swing back then could someone please explain this:

    Consistent double-digit Labour poll leads throughout much of 2012 and early 2013…

    … Followed by single- digit Labour poll leads throughout 2014…

    … Followed by a neck-and-neck situation for the first few months of 2015…

    … Followed by a 6% Tory lead on general election day May 2015…?

    – If that little lot is not “swing back” then I don’t know what is.

    And do you know what is really maddening? It happens in nearly every parliament with almost the same regularity that people deny it’s going to happen.

    Sadly, when the Conservatives fall behind in the polls again next year (and doubtless start losing by-elections) we will have the same bunch of people on here predicting the inevitability of Tory disaster at the next general election.

    Still, it’s good fun to watch.

  38. Correction.

    Migration stats for Scotland suggest they had net migration of around 100k in the last 5 years so actually if we deduct that from the 1.5 million its only 1.4 million that have been disenfranchised.

  39. Robin – what happened to swing back to the Lib Dems?

  40. @ Jim Jam

    “Robin – what happened to swing back to the Lib Dems?”

    It knocked the LibDems out and swung all the way to the Cons and UKIP.

    At Constituency 209 (in alphabetical order): there was a complete misunderstanding of the nature of the LibDem vote of 2010 … Primarily, the majority of them were not liberal, and/or social democrats.

  41. @Jim Jam

    Or swingback to LiS?

  42. Sorry, correction they were not liberal, anti-con liberal, pro-labour social democrat.

  43. Individual voter registration would not have effected Scotland as Scotland used a hybrid system which meant that anyone registered to vote in the referendum and who qualified would get a vote in the GE.

    The polls were very accurate in Scotland.

    In 1992 it was eventually decided that the hangover from non-registration because of poll tax caused Labour to lose could individual voter registration have had the same effect in 2015?

    Maybe polling companies should cross check with the electoral roll.

  44. According to Oddschecker, Betfair are offering odds of 5/6 (yes five to six on) for Oona King as next leader of the Labour Party. Now, there’s a curved ball! (don’t know if Labour’s constitution allows a Leader in the House of Lords)

  45. Burnham calls for Euro referendum in 2016

  46. Let’s not forget that Labour’s popular support actually rose since 2010, just not quite as much as the Tories. Still, it’s hardly evidence of a party in terminal decline.

    Tory Labour
    2010 10,703,754 8,609,527
    2015 11,334,920 9,347,326

    @Gary O
    “Evidence perhaps that at least 1.5 million people were disenfranchised by individual voter registration?”

    Or 1.5 million fraudlent registrations no longer got through? Realistically, the truth might be somewhere in between, though to say people were disenfranchised because they couldn’t be bothered to register seems to be stretching things a bit.

  47. @Daibach
    Either you or Oddschecker have made a mistake. I’ve just checked Betfair and she’s not even mentioned.

  48. Thanks Pete B. I didn’t believe it either!

  49. Correction to my earlier post, Labour’s vote actually rose slightly MORE than the Tory vote since 2010.

  50. Pete B

    To say people couldn’t be bothered to register is a bit of a stretch too. Many people may not have realised they weren’t registered.

    Bizarre anecdotal fact that my own parents never received polling cards and didn’t think anything of it until it was too late. Which constituency? Morley and Outwood.

1 2 3 9