Oldham by-election

One day I’m going to write a generic post by-election post labelled (insert constituency name here) that I can repost after every by-election. Until that day, here’s my traditional answer to what last night’s by-election tells us about the national political picture: not much.

By-election are extremely strange beasts. They take place in a single constituency that may be completely untypical of the country as a whole, they normally have no impact at all upon who will be running the country the next day, they have far greater campaigning intensity than any other election. After every by-election I post the same conclusion – if they show much the same as the national polls suggest they tell us nothing new, if they show something different it’s probably to do with the unique and different circumstances of by-election. In this case the opposition party has held onto a safe seat. This is exactly what we should expect unless they are tanking in the national polls, and Labour aren’t: despite Corbyn’s poor ratings and the constant news stories of Labour infighting their level of support is still pootling along at around their general election share. There is no reason to expect UKIP surges either – in the last Parliament UKIP had soared from 3% to the mid-teens, so almost every by-election saw them surging, but now we are comparing their support to what they got in the 2015 general election, after their breakthrough. This is a good local result for Labour, but doesn’t tell us much new.

That’s not to say it’s not important. By-elections have a significant effect on the political narrative and in that sense this is a very good result for Labour (or, depending on your point of view, for Jeremy Corbyn). If this by-election had gone differently it would have been part of a different narrative, it would have been all about Labour in crisis, their traditional working class support fracturing to UKIP. It would still have been over interpreting a by-election, but it would almost certainly have happened and it’s been avoided. In that sense, it’s an important victory.

A final note about the polling – there wasn’t any (I don’t know whether to be amused or depressed by the handful of comments I’ve seen about it being a another polling failure. Nothing to do with us mate!). By election polling used to be very rare, then in the last Parliament we were suddenly spoilt, with Survation and Ashcroft polls for most By-elections. This time we are back to having no real evidence to go on, to relying on what commentators have been told by the campaigns, what it “feels like” on the doorstep and in vox pops and all that sort of nonsense. I suspect the collective commentariat have got carried away with what would have made an interesting narrative to report, rather than dull old “safe seat held”. It’s a reminder that without any proper polling By-elections can be pretty hard things to call.

270 Responses to “Oldham by-election”

1 2 3 4 6
  1. I guess it would tell us what percentage of the British public claim to be mind readers, which might be interesting .

  2. There is one thing crystal clear, and that’s UKIP failed entirely to correctly judge the constituency’s demographics.

    This can’t be made clearer, than by the constant mislabelling of Oldham’s British Asian community as “immigrants”. Not only are the vast bulk now born here, the last major British Asian migrations having occurred in the 50s, but it’s kind of dubious to say the bulk of those arrivals weren’t British to start with.

    The majority of British Asian ‘immigrants’ that make up the grandparents and great grandparents of the current British Asian community were British Subjects to start with. And came from all over the British Empire, from Malaysia to Africa to the Caribbean. British Asians had been populating the ‘Colonies’, and when those colonies gained independence, many of those British Asian’s “home” nation was Britain. (Several countries even enacted policies “returning” them to Britain. Some by force.)

    UKIP do appear to have problems with understanding British History as opposed to British Mythology.

  3. MARK W

    Internet voting would be even more open to abuse. If people cannot be bothered to walk to and from the polling station they should not have a vote unless they can prove they are physically incapacitated. My wife and I always vote, GE’s, Local and County elections and referenda. If we can do it at 75 plus other people can too.

  4. What @Thorshammer3 said

    Labour’s expectations management was first-rate, complete with Abby Tomlinson reporting tearfully about the hard time she was being given on the doorstep and the Corbynistas slamming her for saying things like that! I expect the Labour voters in Oldham said “Oh no, it’s going to be tight just like in Heywood, I’d better head out in the rain to vote!”

    We’ve just witnessed a masterclass on the psychological side of getting out your vote, because surely the canvass returns predicted the final result? Very well played on Labour’s side.

    As for UKIP – they threw the kitchen sink at Kent and couldn’t dislodge the Conservatives. They’ve tried to take on Lab in their heartlands and couldn’t manage there either. I think they know that their problem is Farage, but can they get him to see it?

  5. From previous thread

    “They all agree and polls and anecdote prove that UKIP would run Labour close in Oldham”

    could be very differential turnout
    – some wards very low and Lab vs Ukip close.
    – some wards 90% turnout, 99% Lab

    “Assume 25% Muslims in west of Oldham. 40% turnout – 10% of total vote.”

    That’s the key assumption though.

    If it was
    – 100% turnout of 25% = 25%
    – 20% of 75% = 15%
    that ends up as c. 60% of total vote

    If so then it would be less about postal votes as about postal votes coming from roughly half of the voters in a solid voting bloc.

    This kinda fits with the early reports of low turnout and low postal votes and then suddenly boom turnout goes from 20% to 40% and postal votes very high.

  6. CANDY

    I would not rule out UKIP doing well in the Medway if the economy headed south, although I would now be more confident of Labour capitalising.

  7. So if it could be shown there is very substantial differential turnout – not necessarily fraudulent – then regardless of the whys and wherefores the current system isn’t a level playing field.

  8. There’s a lovely picture of McMahon and his wife in the papers. The optics are very good – sort of a Mancunian version of Justin Trudeau and his wife Sophie. Just saying. :-)

    @Mr Jones – no matter how you look at it, the victory in Oldham was delivered by the WWC. If UKIP want to move forward, they need to work out why, not pfaff around with impossible assumptions of 100% postal voting in minority demographics. Farage is a problem, as is the perception in the north that UKIP are Thatcherite – why are UKIP skating around this when they have people who could do his job so much better?

  9. @Mr Jones

    So if it could be shown there is very substantial differential turnout – not necessarily fraudulent – then regardless of the whys and wherefores the current system isn’t a level playing field.

    Of course the system is a level playing field. All voters have the same chance to vote, they can have a postal vote, or even a proxy.

    If there are two wards, A and B, if A contains people who mostly vote for one party and turnout out at 70%, and B contains a more split voting trends turning out at 30%, surely the lazy voters in ward B to get off their backsides?

    It’s entirely fair and level.

  10. Correction

    @Mr Jones

    So if it could be shown there is very substantial differential turnout – not necessarily fraudulent – then regardless of the whys and wherefores the current system isn’t a level playing field.

    Of course the system is a level playing field. All voters have the same chance to vote, they can have a postal vote, or even a proxy.

    If two halves of a constituency/ward/whatever have a big turnout difference, the lazy folk in the low turnout area who don’t vote can’t complain that the people in the other are somehow cheating or have an advantage.

    It’s entirely fair and level.

  11. Is Farage a problem. Even in the north?

    Unless someone knows off the top of their heads I shall have to look it up but I thought his personal approval was usually pretty good.

  12. Candy

    “no matter how you look at it, the victory in Oldham was delivered by the WWC”

    If you mean by not voting i’d agree


    “not pfaff around with impossible assumptions of 100% postal voting in minority demographics”

    Sure, that was just a mathematical example of the effect of differential voting.

    However the point is if there is a very dramatic differences say 80% vs 20% in turnout that totally screws polling predictions unless it is taken into account.



    “Of course the system is a level playing field.”

    Okay scrub that and just look at the polling implications – if on a wet december bye-election there is an 80% vs 20% difference in turnout between different groups then you’d need to weight the polling by expected turnout.

  13. @Mr Jones

    Differential turnout must be a massive headache for AW et la.

    It appears that trying to build a good, working model to create a representative sample of the UK is become very hard.

    I blame the voters ;-)

  14. @MrJones

    Not voting is a valid choice in our democracy – we don’t have mandatory voting.

    If you want people to vote you need to give them a reason to. UKIP didn’t – and it would be more productive for you to work out why than go down an imaginary cul-de-sac where you comfort yourself that you were robbed by one demographic voting 100% (especially as historic info from that constituency shows that teh turnout levels amongst the british-asian side are similar to the WWC side).

    This result was similar to the one Meacher got in May – and Meacher was a life-long Corbynite don’t forget, so Farage’s focus on Corbyn wouldn’t have cut much ice there, and he didn’t offer anything other than a focus on Corbyn did he? The loss is because he misread the culture of that constituency – that’s a serious political failing.

    You might also need to face the fact that we’re going back to two party politics – and that’s the reason UKIP couldn’t dent the Cons in Kent either, and both the LibDems and Greens are on life-support.

  15. Candy

    Maybe so

    or maybe there’s some wards with 80% turnout and some with 20%.

    In which case polling done in the first should weight the expected vote times 4.

  16. Matt Singh has an interesting piece in the New Statesman, suggesting that the win was due to Labour getting their vote out well.

    I understand that in America, the current thinking is that getting your own vote out, rather than triangulating policies is the way to win elections. You have to enthuse people, which is where Labour fell down in May 2015.

  17. In last month’s Ipsos Mori Poll (where UKIP got 7% VI). Farage had a personal approval of minus 12 (33 positive, 45 Negative), which was better than Cameron or Farron, but nowhere near as good as Corbyn (-3). So I am not sure that he is a drag on UKIP. Although I would expect him to be a lot less popular in Lancashire than Kent.

    Although, without Farage, would Carswell be leader? I doubt he would be any more popular in the NW

  18. @ Hawthorn

    You have to enthuse people, which is where Labour fell down in May 2015.

    I think that’s exactly right. As my son put it: When you try to appeal to everyone, you end up being trusted by no-one.

  19. An improved showing for Labour here despite no improvement in national polling, and in the context of a new, polarising leader, may suggest that Labour are piling up votes in their heartlands and that that’s masking a decline elsewhere – although that would be reading far too much in an individual poll, of course.

  20. Hawthorn,

    American politics is a lot more polarised with few swing voters and a lot of locked-down interest groups for each party. Whereas here there are people who vote between Tory and Labour much more freely than people switch between Democrat and Republican.

  21. “Discussed before, adopted after.”


    S’pose that makes a change from the more usual state of affairs, whereby they rush to do stuff then discuss the folly of it later in hindsight…

  22. Though if they discuss it first and still can’t see the problem prolly doesn’t help any…

  23. @Welshguy

    I agree.

    It was a genuinely good result for Labour. Any other interpretation would plain daft from where I sit.

    Of course, the by-election had a number of things that were good for Labour:

    1 – They had a genuine, good and local candidate, with a good established profile.

    2 – The seat was held well by Michael Meacher, so the electorate knew a what a Labour left-winger looks and sounds like, and therefore Corbyn as Leader would cause little fear.

    3 – The Labour vote is strong in communities where Corbyn’s and the majority of the PLP’s view on not wanting to bomb Syria is an asset.

    The real test lay if we have a by-election in a typical swing seat, the sort Labour struggled to win in 2015.

  24. catmanjeff

    “It was a genuinely good result for Labour. Any other interpretation would plain daft from where I sit.”

    Depends if it was 40% turnout across the whole constituency or not – if it was then you’re right.

    If it was say 80% turnout in some wards vs 20% in others – as maybe indicated by the twitter confusion last night – then it’s a bit different – still good in a lot of seats but maybe not so good in others.


    seems to me it ties in a bit with the analysis of the general election polling narrowing in on low turnout among a lot of Lab identifiers.

    from the sound of the posts on here it seems there is no clear standard number to down weight working class votes by – if there is very substantial differential turnout between diff chunks of Lab identifiers then knowing the ward turnouts from Oldham might be the extra piece of info needed to make the general election polling numbers come right.

    Lab identifier type A – weight by 20%
    Lab identifier type B – weight by 80%

  25. @MrJones

    This can be checked, as Candidates can buy marked registers post election.

    It won’t confirm who people voted for, but if certain wards do have a high turnout, it can be calculated.

    It boils down to GOTV. Clearly UKIP did not do this well enough. BTW, I live in an area where it happens to be the Tories that are king of the postal vote. In one Council election where I stood for the Greens, I was neck and neck with my Conservative rival on the papers out of the boxes I was observing.

    When the postal votes came in, this doubled the Conservative vote, leaving me third by quite a margin.

  26. Catmanjeff

    “This can be checked”

    That will be interesting.

  27. I believe the cost is about £25-40 ish.

    If UKIP are serious about fighting this seat in the future, it’s a no-brainer to buy one.

    Of course, there is one hazard with checking the facts in this case. If it ends up with a broadly similar turnout in all wards, then Mr Farage will look very silly, having made serious claims.

  28. indeed – good opportunity for other parties to make him look silly then

    if it’s wrong

  29. messing about with the numbers

    say you had 70,000 electorate split between 25% with a fixed 80% turnout for one party and 75% with *either* a 20% turnout in wet bye elections and 40% at general elections and split between various parties.

    then in both the bye election and general group A would produce 14,000 votes

    the rest would produce a total of either 10,500 or 21,000 votes

    (head sums so could be wrong)

    so no way of winning at a bye election but what would that say about the general?

    if the 14,000 number was correct the vote among the 75% was roughly

    6 1/2 Ukip
    3 Lab
    2 1/2 Con
    1 Lib

    so at 40% turnout instead of 20% that would be

    13 Ukip
    6 Lab (+ the 14) = 20
    5 Con
    2 Lib

    compared to the
    23 Lab
    9 Ukip
    8 Con
    1 1/2 Lib
    at the general

    which would be close if a there were a lot of Con->Ukip switchers

  30. “which would be close”

    i mean close to the GE figures not close result.

  31. Farage’s main problem to me seems to be that the longer he’s around and the more exposure he gets, the more he comes across to voters as the one thing he says he isn’t. Part of the political establishment.

    In addition, whenever his party does badly, he comes across as what we in the north would call a ‘mardy bum’. No one up north likes a mardy bum!

  32. Suzanne Evans would project a softer image and comes across less macho, less harsh and delivers her points in a more subtle fashion.

    She’s clearly a smart lady.


    My otherwise placid and rational wife is reduced to screaming obscenities whenever Suzanne Evans appears on anything. It’s her ‘slappable face’ that’s the subject of much of the anger. I’m not so sure…

  34. Forth Road Bridge closure until New Year could have political ramifications in Scotland (… or not, but I just wanted to shoehorn in some fun stuff!).

    Here’s a hilarious contribution from UKPR regular, Statgeek.


  35. @Sven Hassel Schmuck

    You can’t please all of the people….

  36. Amber

    Your stoicism at being cut off from civilization does you great credit. :-)

  37. Just come in from the cold to ask how many votes in Oldham were postal? How many were handed in and are that sort counted as postal or not?

    Most grateful in advance.for any reply.

  38. @ Old Nat

    I know! I have friends/ family in Perth so I’ll just have to let the train take the strain (as the advert used to say).

    I do feel sympathy for people who are affected by it. I don’t know whether there’s the possibility of mitigating action being taken to increase trains.

  39. Amber, I am into train stuff and if there were any spare, they would have been used to strengthen the already overwhelmed ‘Borders’ service.

  40. Amber

    It’s going to cause misery to lots of folk, and even the 25% increase in train capacity between Fife & Edinburgh can’t mitigate all of the disruption.

    Ferry provision (and linked buses either side) is being looked at.

    Just as well that the Queensferry Crossing will be completed by the end of next year. Can you imagine the long term chaos if the likes of George FFoulkes had been listened to!

    Good Evening to you.
    I fully agree with you. I think evidence points to Labour doing well in core areas. The undoubtedly good result in Oldham is on top of a large swing in May 2015.

    The same thing happened in 1950 and 1951, with Labour piling up votes in relatively ‘deprived’ areas, but not doing so well in prosperous areas, where the marginal seats were, and are still, located.

  42. @ Bristolian Howard

    Postal votes are counted as postal votes, even if they’re handed in on the day because they still go through the verification process.

    There were 7k postal votes, 25% of the total votes cast.

  43. @ Old Nat

    I’m not getting into the politics of the new bridge vs reducing traffic on the old one (to extend its life) by not making it free & by encouraging much more public transport use/ park & ride/ rail & sea freight etc.

    It’s a past argument & revisiting it might undo our current détente. ;-)

  44. First we have expectation management, now comes the outcome management. I suppose it was ever thus, but it doesn’t stop some of the commentary on last night’s by-election result straying into the realms of hilarity. UKIP cry foul and claim fraud and the Tories think Cameron will be happy because it shows UKIP losing momentum ahead of next year’s referendum. All we’re lacking is the Lib Dems and Greens saying that they were unfortunate victims of a classic two-party by-election squeeze!

    It seems to me, however happy or sad it may make some of us, that last night was a pretty good night for Labour and its leader. It may well be that a popular local candidate increased the Labour vote, and clever and concerted local campaigning assisted too, but I sensed this was a disaster in the making here and the fact that it was averted, and turned into somewhat of a triumph, has some genuine political importance and significance.

    A perfect storm was brewing for Labour in Oldham West and Royton. Dire national opinion polls for both the party and leader and a disastrous few weeks that were bookmarked by McDonnell’s inept response to Osborne’s spending review at one end and then the Syrian vote shambles at the other. The seat was also heavily targeted by UKIP who threw the kitchen sink at it, convinced that Corbyn’s leadership would drive white, working class Labour voters into their hands. They thought it was Heywood and Middleton all over again but they got their comeuppance. A poor showing from the Tories too in a seat where they used to routinely get 30-40% of the vote. They actually won it in a by-election in 1968, albeit without the Royton bit.

    As for UKIP, I never quite bought this threat to Labour in the North theory and with the exception of Heywood, they were comfortably seen off in all the other Labour held seat by-elections in the last Parliament, even when they were riding high in national opinion polls. There was no great evidence either that they took chunks out of the Labour vote in the North in May 2015, or in London. They tended to make some hay in suburban seats in the Midlands but, there again, I’m not sure they tipped many seats into Tory hands.

    No harbinger for elections to come, but a very decent result for Labour nonetheless. It would be silly to deny that.


    You can’t please all of the people

    Butnin politics, to achieve real success you must please a great many.

  46. I heard on the news, that they would be laying on extra carriages on the trains across the Firth of Forth, if that’s any help. Sounds appalling to be facing a 15ml diversion to the next bridge and presumably 15ml back on the other side :(

  47. Amber

    “It’s a past argument”. Indeed, though why some would suggest that “Forth Road Bridge closure until New Year could have political ramifications in Scotland” is less clear.

    Perhaps you were referring to the Green’s opposition to the Queensferry Crossing?


    @David Colby
    “…but why can’t we be told how the postal voters voted?”
    The (Acting) Returning Officer is prohibited from counting the votes given on the postal ballot papers separately from the votes given on the ballot papers from the polling stations. They have to be mixed together first.
    Rule 45(1A) of the Parliamentary Elections Rules (contained in Schedule 1 of the RPA 1983) states that:
    The returning officer shall not count the votes given on any ballot papers until—
    (a) in the case of postal ballot papers, they have been mixed with the ballot papers from at least one ballot box, and
    (b) in the case of ballot papers from a ballot box, they have been mixed with the ballot papers from at least one other

    THANK YOU for explaining that. It does seem that it would be helpful if these rules were re-examined. Why shouldn’t all postal ballot papers be put together and counted as a single block?

  49. A YG London Mayor poll is being mentioned in the Graun.

    An Evening Standard/YouGov poll showed Sadiq Khan, the Labour candidate, well ahead of his Tory rival Zac Goldsmith for the London mayoralty elections in May. Given a choice between Khan and Goldsmith, the former held a 53% to 47% lead among likely voters.

    Does anybody have more info on this poll?

1 2 3 4 6