Liberal Democrat Target Seats

These are the fifty seats with the lowest percentage majorities over the Liberal Democrat party. This does not necessarily mean they would be the most winnable seats for the Lib Dems in practice, or that they are the seats the Lib Dems party will actually be targetting at the next general election.

1. Cambridge Majority 599 (1.1%)
2. Eastbourne Majority 733 (1.4%)
3. Lewes Majority 1083 (2.1%)
4. Thornbury & Yate Majority 1495 (3.1%)
5. Twickenham Majority 2017 (3.3%)
6. East Dunbartonshire Majority 2167 (4%)
7. Kingston & Surbiton Majority 2834 (4.7%)
8. St Ives Majority 2469 (5.1%)
9. Edinburgh West Majority 3210 (5.9%)
10. Torbay Majority 3286 (6.9%)
11. Sutton & Cheam Majority 3921 (7.8%)
12. Bath Majority 3833 (8.1%)
13. Burnley Majority 3244 (8.1%)
14. Bermondsey & Old Southwark Majority 4489 (8.8%)
15. Yeovil Majority 5313 (9.4%)
16. North East Fife Majority 4344 (9.6%)
17. Caithness, Sutherland & Easter Ross Majority 3844 (11.2%)
18. Colchester Majority 5575 (11.4%)
19. Cheadle Majority 6453 (12.1%)
20. Cheltenham Majority 6516 (12.1%)
21. Berwick-upon-Tweed Majority 4914 (12.2%)
22. Ross, Skye & Lochaber Majority 5124 (12.2%)
23. Portsmouth South Majority 5241 (12.5%)
24. Brecon & Radnorshire Majority 5102 (12.8%)
25. Cardiff Central Majority 4981 (12.9%)
26. North Devon Majority 6936 (13.3%)
27. Wells Majority 7585 (13.3%)
28. North Cornwall Majority 6621 (13.8%)
29. Gordon Majority 8687 (15%)
30. Hazel Grove Majority 6552 (15.2%)
31. Montgomeryshire Majority 5325 (15.7%)
32. Birmingham, Yardley Majority 6595 (16%)
33. St Austell & Newquay Majority 8173 (16.2%)
34. Argyll & Bute Majority 8473 (16.4%)
35. Eastleigh Majority 9147 (16.5%)
36. Oxford West & Abingdon Majority 9582 (16.8%)
37. Bristol West Majority 5673 (8.9%)*
38. Bradford East Majority 7084 (17.1%)
39. Berwickshire, Roxburgh & Selkirk Majority 328 (0.6%)*
40. Chippenham Majority 10076 (18.2%)
41. Inverness, Nairn, Badenoch & Strathspey Majority 10809 (18.8%)
42. Hornsey & Wood Green Majority 11058 (19.1%)
43. Aberdeenshire West & Kincardine Majority 7033 (12.8%)*
44. Bosworth Majority 10988 (20.5%)
45. Maidstone & The Weald Majority 10709 (21.4%)
46. Mid Dorset and North Poole Majority 10530 (22.6%)
47. Newton Abbot Majority 11288 (23.4%)
48. Solihull Majority 12902 (23.5%)
49. Redcar Majority 10388 (25.4%)
50. Watford Majority 9794 (17.5%)*

*Liberal Democrats currently in third place

Comments - 771 Responses on “Lib Dem Targets”
  1. Paddy Power odds on Lib Dems to win individual constituencies

    Westmorland 1/7
    Orkney & Shetland 2/9
    Ceredigion 1/4
    Edinburgh W 1/3
    Leeds NW 4/11
    Twickenham 4/9
    E Dumbartonshire 1/2
    Cambridge 4/7
    Sheffield Hallam 8/13
    Caithness 8/11
    NE Fife 4/6
    Kingston & Surbiton 5/6

    Bermondsey 11/10
    Lewes 5/4
    Richmond Park 5/4
    Southport 6/4
    Eastbourne 6/4
    Carshalton 13/8
    Bath 13/8
    North Norfolk 13/8
    St Albans 7/4
    Cardiff Central 7/4
    Ross, Skye & Lochaber 7/4
    St Ives 15/8
    Hornsey & Wood Green 9/4
    Birmingham Yardley 9/4
    Cheltenham 5/2
    Burnley 11/4
    Sutton & Cheam 3/1

  2. Reading odds remains a strangely confusing experience for me. Would be easier to just do it in percentages.

    But in any case: gaining four Scottish seats would be damned impressive. In fact, gaining seven, even with four losses, would be a big achievement.

    We shall see.

  3. Just take the fraction & convert it into a decimal figure (divide first figure by second figure) – then add 1. Now divide 100 by this number.

    So 6/4 is 1.5. Add 1 = 2.5. 100 ÷ 2.5 = 40 (6/4 is 40%).

    Or 7/2 is 3.5. Add 1 = 4.5. 100 ÷ 4.5 = 22.22 (7/2 is 22%)

  4. No, I know… It’s just tedious. Maths was never my strong suit.

  5. BT – you’re wrong on the constitutional point.

    I will admit that I was wrong on it as well, as I felt that in England and Wales Labour would be squeezed by the more clear-cut choices of pro-leave Conservative and pro-Remain Lib Dem, as far as attitudes towards how Brexit is delivered is concerned. The Tory support slipping somewhat and not going to UKIP clearly demonstrates that the constitutional issue was overstated on the Leave side. Meanwhile, however poor the LD campaign has been, one would have expected support to fall into the LD’s laps in a very noticeable manner from polling, if the EU were the dominant issue we all expected it to be in the run-up to polling day. Especially given their low starting point.

    But no, humble pie is in order before the polls have even opened in my opinion. 2015 was the constitutional election, and I now doubt there will ever be one quite so constitutional in our lifetimes.

    2017 sees the realignment of UKIP voters now that the constitutional matter they were founded on is in the process of being settled, but otherwise tax and spend and health and social care seem to be at the fore.

    The fact that the Tories, Lib Dems and SNP were equally wrong on the extent to which the constitutional matters would dominate is precisely the reason why the gap has narrowed so drastically between the big two, why the LDs look like doing little better than last time, and why the SNP are going to slip back a bit in Scotland.

    The Tories, with some justification, felt that Brexit would carry them home. Not simply because they felt the electorate were that easy to win around, but because with Corbyn at the helm they felt that no policy they came out with could damage them as much as what would ultimately be in Labour’s manifesto. I’m confident they’ll still get an overall majority, but it’s not going to be the obliteration of the hard left and forcing of Labour to find the next centre or centre right reformer that many would have thought likely a very short time ago.

    It was said many a time during the recent by-elections that Labour squirm when Brexit is at the fore, and are desperate to talk about anything else in the belief that they could hold their own when debating tax and spend. Whether you agree with their policies or not (I retain the view that regardless of how Labour do, they would have done better under any other leader), Labour are the only party mentioned in this post who appear to have identified that the public want to move on from the constitutional matters, having had their say on them, and onto the areas of government policy that are most noticeable in day-to-day life.

    Speaking personally, I feel the wrong decision was made on 23 June, but I accept that the decision has been made. That the fact that I retain the view that we would have been better off remaining does not change the fact that we are going to Leave the EU in whatever manner the Government formed this month sees fit.

  6. I honestly think the “how much do voters care about constitutional questions” (or any questions) is something of a moot point. It’s honestly quite hard to tell what voters do and do not care about, and most of the time it ends up being an incredibly ill-informed decision for many, and a strangely based one for a large group as well. There will be people who don’t vote for Corbyn because he has a beard, or May because she’s a woman, or Farron because he’s religious. Silly little things, really.

    Conversely, some will vote only on Brexit/Scottish independence. Others will vote on any issue BUT those. I think the problem is that it’s mixed, and most parties are finding that they can’t quite hit everything at once, which is making them lose ground.

    Labour, if they improve at all, will be stuck with Corbyn. So, not ideal for them. Even if they go back just slightly, he’ll probably stay.

  7. LDs released a video highlighting their candidates in St Ives, Oxford West, and East Dunbartonshire… Could mean they think those are top targets or very close if they’re pushing them right now.

  8. BritainElects predicts gains in:
    East Dunbartonshire
    Edinburgh West
    Kingston & Surbiton
    LOSS of Southport and Richmond Park.

    YouGov has gains in:
    Kingston & Surbiton
    Oxford West & Abingdon
    Edinburgh West
    LOSS of Southport, Richmond Park, North Norfolk, Sheffield Hallam, Leeds North West, Orkney & Shetland, and Carshalton & Wallington
    (They still predict 9 seats, though that’s 7 by my count)

    ElectoralCalculus thinks they lose everything but Westmoreland, Ceredigion and Orkney & Shetland.

    ElectionForecast has no gains and losses in all but Orkney & Shetland and Ceredigion (yet, somehow, they predict 6 seats…)

    So, there you have it: the only seat that every predictor agrees on is Ceredigion.

  9. So if the Lib Dems stay on about 8-10 seats, there is a decent chance that will include more gains than holds.

  10. Yes. It seems that the predictions are 100% certain only of Ceredigion.

    Notably, Twickenham is the only seat YouGov thinks is “Safe LD.” They give Cable a vote share in the mid-50s. I think they’re overestimating him by a fair bit, frankly. I think BritainElects and YouGov are both too optimistic for the LDs in Kingston, as well.

    My hunch is that Lewes and Oxford West are very, very tight. I think that’s born out here to some extent, but we’ll see; there’s obviously a chance they’re wrong, too.

    Edinburgh West seems the most likely gain, I suppose.

    I wonder if the party after June 8 could be:

    … or something like that.

    If I were the LDs, I would be pushing Swinson, Cooper, Hobhouse, Riches, Jardine, Blundell, Moran, Maguire, and Jardine incredibly hard, as well, of course, as Olney. Getting female MPs is essential for them.

    If they elect Blundell, Moran, Swinson, or Cooper, I think it’s better than evens their next leader will be a woman. The others are either a bit older or less charismatic-seeming, excepting Maguire, about whom I know very little.

  11. Looking at YouGove predicitions for individual seats, it appears that LibDem prospects are becoming concentrated in Southen meidum sized towns where they already have a base. They appear to be narrowly ahead in places like Bath, Eastbourne and Cheltenham. They are coming noweher in the rural South-West and they are at serious risk in provincila urban eats they currently hold such as Southport, Sheffield Hallam and Leeds North-West.

    By the way, the opinion polls have never managed to get Orkney and Shetland right.

  12. Frederic, those aren’t polls. They aren’t “ahead,” it’s just YouGov’s guess based on demographics/national polling (which now has 50-150 results per seat, not representative samples), and it’s base don what would happen if the election were held today, not a prediction for next week. So, big grain of salt.

  13. I’ve noticed that unlike 3 or 4 weeks ago, there are now noticeably more LD>LAB SWITCHERS THAN LAB > LD

  14. @ Mr Pitt.

    re YouGov’s methodology: the constituency projections are not based on small samples from that constituency alone but instead from wider groups of constituencies with similar demographics.

  15. Yes, I realize that; that doesn’t make them a better sample, however.

  16. I don’t think that’s true.

    The sample size feeding into their projection for each constituency is far greater than the number you quoted. And while I don’t pretend to understand all the fine detail of polling methodology, no pollster publishes data based on samples which are not designed to be representative.

  17. James,
    The point is that it is not possible to get a sample from other seats that is representative of Sheffield Hallam, for example, because the only seat which is remotely comparable is Leeds NW (Lib dem incumbent fighting Labour in a northern University seat with a sizeable Tory vote). The behaviour of Lab/Lib Dem switchers in Hallam will be very different from in a Lab/Con marginal like Pudsey, or even in University seats where Labour are the incumbent.
    Currently different models are predicting different results from each other in Hallam and Leeds NW, and I suspect the youth turnout factor is one unpredictable elephant in the room.
    However, it is just plain logic that if the national Lib Dem % is the same as 2015, and the Labour % is up, Labour have a better chance of winning….

  18. My understanding is that their constituency estimates are based on the demographics of the seat in question, so that for example for Hallam they are looking at a large number of respondents from many constituencies, but selected in such a way as to closely reflect Hallam’s unique demographic profile.

  19. James and Andrew are both right. They try to match respondents’ demographics to constituencies. The problem with that is that demographics aren’t the only influence on voting behavior. Sheffield Hallam is a fairly unique seat at this stage: wealthy suburb/rural area mixed with university seat, Lib Dem vs Labour fight with strong Tory history. Leeds NW is slightly comparable.

    So, a voter who fits a certain demographic profile might be expected to vote for Party A, but that voter demographic exists in many, many seats, and few fit the description of Sheffield Hallam. Essentially, the seat is unique enough in terms of political circumstances that a voter in a certain demographic might be solidly supportive of Party A elsewhere, or of Party C, but in Sheffield Hallam, backs Party B.

    Demographics, in short, are predictive, but only in a limited capacity.

  20. If reflective of national splits in each party’s vote, Hallam would have been (roughly rounded) as follows:

    LD Remain 27%
    Lab Remain: 23%
    Con Remain: 6%
    UKIP Remain: 0%
    Green Remain: 2%

    REMAIN: 58%

    Lab Leave: 13%
    LD Leave: 13%
    Con Leave: 9%
    UKIP Leave: 6%
    Green Leave: 1%

    LEAVE: 42%

    Obviously it was 65-35, which I would put down to the Conservative and Labour voters in those seats being more Remain-trending than nationally – a lot of students and public sector among the Lab voters, and a lot of wealthier Conservatives.

    Conversely, I suspect the Lib Dem vote here (which in large areas is an anti-Labour vote) was more Leave leaning than average. The question now is whether those Liberal Leavers stay with the party of Remain post-referendum, and if the quite Remainy Labour vote will stay with their party.

  21. New YouGov election “predictions” very favorable to LDs, but I can’t tell why — any ideas, fellas? They have 7-21 as the 95% range and 13 as their guess. This includes (sorry, Wellytab) O&S now marked as “safe.”

    According to YouGov:

    Orkney & Shetland

    Kingston & Surbiton

    Westmoreland & Lonsdale
    East Dunbartonshire
    North East Fife

    Richmond Park

    St Ives
    Oxford West & Abingdon

    Carshalton & Wallington
    Sheffield Hallam
    North Norfolk
    North Cornwall

    So they’re actually predicting 13, which is safe + likely + lean + tossup ahead + 4 others… so all the “tossup behind” seats or a mix of those and those leaning away?

  22. A few days ago they had Norfolk North as tossup, St Ives Con Lean, Bath as Con lean, Cheltenham as LibDem Lean, **Westmorland** as Tossup, and Lewes & Eastbourne as LD-tossup or LD lean…

    When they first started the model, Orkney was “likely SNP”, and the only “safe LibDem” in the country was non-LD-held Twickenham… which has since reduced to “likely”.

    It’s only with the latest update East Dunbartonshire & North East Fife have had LibDem chances ahead of SNP ones.

    It just doesn’t seems for LibDem seats their model isn’t sufficiently siloing some demographic data.

    Although they can’t get the sample sizes for proper constituency polling, you think it’d be worth bundling up say:
    Bath/Richmond/Twickenham/Kingston for a decent enough sample that they didn’t need to rely on national/wider-regional trends of ‘what remainers do’ ‘what ABC1s do’ ‘what 2015LibDems do’ etc to supplement the otherwise high MOE samples.
    Similarly Eastbourne/Colchester/Wells/StIves, for southern rural-esque seats with 2nd place “popular” incumbents running again.
    And the 3 Scottish targets: EdinWest/NEFife/EastDunbar

    I’d be really interested to see some “seat group” polling generally. In fact I think it could be a sounder (but still commercially viable basis) for seat-predicting models going forward – as you can basically cut bothering to survey anyone in the 400+ seats that wouldn’t change hands regardless.

  23. Are you mainly saying there’s a lot of variability? Because generally that seems to be true of their LD predictions, yes.

  24. I think its fair to say that the LD’s will still be in DEEP TOILET, either way on Friday morning.

  25. Commenting on the two seats on that list I know best, the Lib Dems are not going to win either Eastbourne or Lewes. I suspect they’ll disappoint in SW London though Sir Vince might scrape in. It doesn’t fill me with confidence that the view on the others is very accurate. St Ives in particular is surely a long shot at best.

  26. I stand by thinking that Lewes will be close, probably a better shot than Eastbourne.

  27. They might win Lewes, the incumbent is a weak candidate, but Eastbourne, where that’s not the case, seems increasingly unlikely

    And besides as much as i agree with him as much as anyone on here about most things, when it comes to predicting elections, H Hemelig has form – and nearly all of it bad!

  28. Sadly for them, it will be adieu to the L-i-e Dems (as well as UKIP) on 8/6/17.

  29. I think the Lib Dems will be lucky to get much more than 10 seats. I can see them losing Carshalton & Wellington, Southport and their by-election gain in Richmond Park. They’ll probably gain Twickenham and maybe one more seat from the Conservatives (Bath or Kingston & Surbiton would be the most likely IMO). They also look set to get 2 or 3 seats off the SNP (most likely Edinburgh West, East Dunbartonshire and NE Fife). Even if they do manage to gain all those constiruencies, I really can’t see them doing any better than 12 seats.

  30. LDs were out in force today – in 50 mph winds – in Southport.

    But I’m told from insiders in both LDs & Tories that the Tories should win it. It may even become a 3-way marginal in 2022 or a Tory v Lab fight.

  31. Of course, Southport isn’t as much a one-man seat as some of the LDs’ old strongholds. Pugh and Fearn together built it into a strong local organization. They’ll lose it, but it’s not so much because Pugh is leaving (though obviously he’d be favored to hold, I’d think) as because of the national tide.

  32. It would be curious if the Lib Dems lost Richmond Park.

  33. Mr Pitt,
    You can tell the YouGov model is rubbish by the fact they have Labour miles out in front in Leeds NW…

    I suspect they have failed to notice that most of the students will not be there… If Labour win this seat it will be via a recount.. Mulholland is very popular..

  34. Why would it be curios if the LDs lost Richmond

  35. Tim Farron and the Lib Dem battle bus were in Sheffield today. As the party doesn’t have a single target seat in Yorkshire (indeed the nearest geographically is probably Birmingham Yardley) it suggests they might be worried about Leeds North West or Sheffield Hallam and want to get on Look North.

  36. Tim Farron and the Lib Dem battle bus were in Sheffield today. As the party doesn’t have a single target seat in Yorkshire (indeed the nearest geographically is probably Burnley) it suggests they might be worried about Leeds North West or Sheffield Hallam and want to get on Look North.

  37. when you produce a multilevel regressive and post-stratificative model of 50,000 people you can go about rubbishing other people’s models

  38. Walt White

    “They might win Lewes, the incumbent is a weak candidate, but Eastbourne, where that’s not the case, seems increasingly unlikely

    And besides as much as i agree with him as much as anyone on here about most things, when it comes to predicting elections, H Hemelig has form – and nearly all of it bad!”

    What about a small friendly bet then? £25 that Lewes is a Tory hold.

  39. The LibDems seem to be in real trouble; particularly if the Conservatives are as good at campaigning in LIbDem seats/targets as they were in 2015.

    I can see the LibDems as being confined to a small enclave of seats in South-West London, with literally one or two maverick results elsewhere.

  40. The Tories had a much better campaign in 2015 this one is one to forget me thinks

  41. I agree CON will hold Lewes. I don’t believe Lib Dems will GAIN any seats south of the border.

  42. I think the Lib Dems might lose five MPs. Three to the Tories and two to Labour.

  43. Nameless: nearest geographical target seats would be Hazel Grove and Cheadle, not Yardley.

    I agree with Matt Wilson as well: May is, when it comes to campaigning, no Cameron. The 2015 Tory campaign was frankly a masterpiece, on the level of Obama’s 2008 campaign in the U.S.

    It’s also worth noting the Tories are (somewhat by necessity) focusing on Labour seats this time. I think their message is more tailored to northern, working class voters — where they stand to make gains — rather than the progressive, small-L liberal remainers in some of their southern seats. In other words, losing a Twickenham, a Bath, or a Cheltenham wouldn’t be so dire for them if, as is likely, they’re making two or three gains up north for every seat they lose like that.

  44. Or maybe even six MPs. But their Welsh and Scottish seats should be safe.

  45. I think they’re probably slight favorites in Edinburgh West and about evens on Fife NE and E Dunbartonshire, and almost certain in O&S (though who knows, with that seat, I suppose). That’s at least two there. Then a hold in Ceredigion, and at least two SW London seats (maybe three, possibly four).

    So, five seats. I think Hallam and Leeds NW are probably at least tossups, not leaning Labour (I could be wrong; but that’s my current hunch. Most Labour predictions in Hallam and Leeds NW are based on UNS or YouGov-style “similar demographics” polling, and I don’t know how well those will hold up there; yes, their students might be Corbynite, but will the wealthier voters in the leafy suburbs be at all receptive to him?) So let’s say they hold one of those at minimum.

    Then, North Norfolk and Westmoreland are both obviously a bit challenging, but I’ll say one there, two. So that’s seven. I think the LD minimum is seven seats, practically speaking, unless it’s an absolutely unmitigated disaster for them (and so far, polling suggests a mitigated disaster, I’d say — I also think that, with fewer places to concentrate, their targeting is a bit better this time).

    I also think they have chances in Lewes, Bath, Cheltenham, Oxford West, St Ives, and very, very outside hopes still in Cambridge and Eastbourne.

    So I’d say I think their absolute maximum is 17. Absolute minimum probably 5-7 (seven is my low-end guess, but obviously we’ve been surprised before). I’d wager on 9-12.

  46. PT – they came within 1500 of winning Southport in 2015 (the closest since 1992).

    But I take your point that their share fell as did the LDs’ and it was UKIP and Labour who surged. From memory the Tory PPC was only selected 6 months before the 2015 GE so it didn’t surprise me in the least.

  47. Lancs is right on Southport.

    Tories will gain it this time; LDs should be hoping to contain Labour so it doesn’t become a three-way.

  48. Mr. Pitt is bing optimistic from a LD point of view.

    In particular, is there any reason to think that Scotland will buck the trend with a considerable or even a modest swing to the LDs?

  49. Oh, I think Scotland will be a net swing away. Just that unionist votes are consolidating. The Holyrood election proved that. That’ll make Scottish seats more winnable, but also mean dozens of lost deposits in Scotland.

    Not at all a fringe view.

  50. Totally agree. It’s possible the LDs might even fall below 5% in Scotland (as some polls have shown) but make 2 or 3 net gains because of tactcial voting and targeting the right seats.

Leave a Reply

NB: Before commenting please make sure you are familiar with the Comments Policy. UKPollingReport is a site for non-partisan discussion of polls.

You are not currently logged into UKPollingReport. Registration is not compulsory, but is strongly encouraged. Either login here, or register here (commenters who have previously registered on the Constituency Guide section of the site *should* be able to use their existing login)