Na h-Eileanan an Iar (Western Isles)

2015 Result:
Conservative: 1215 (7.6%)
Labour: 4560 (28.6%)
Lib Dem: 456 (2.9%)
SNP: 8662 (54.3%)
Christian: 1045 (6.6%)
MAJORITY: 4102 (25.7%)

Category: Safe SNP seat

Geography: Scotland, Highlands and islands. The whole of the Comhairle nan Eilean Siar area.

Main population centres: Stornoway.

Profile: Na h-Eileanan an lar covers the Western Isles (the seat was called Western Isles until 2005). It covers the Outer Hebridies, the further reaches of the archipeligo off the coast of north-western Scotland, including the islands of Lewis and Harris, North Uist, South Uist, Barra, Benbecula, Scalpay, Great Bernera, Grimsay and the uninhabited St Kilda. The only town on the Outer Hebridies is the fishing port of Stornoway on the Isle of Lewis, from where ferries sail to the mainland. Stornoway also has an airport with services to the mainland. The seat is socially as well as geographically somewhat isolated: Gaelic is widely spoken, the Western Isles are the only area in Scotland were over 60% of people speak Gaelic. Sunday Observance is also still widely observed on the Islands, particularly in the Northern islands, with a widespread refusal to trade or travel on the Sabbath, due to the continuing strength of the Free Church and Free Presbyterian Church. In 2006 considerable controversey was caused with the opening of a ferry service to Lewis that operated on Sundays.

Politics: Na h-Eileanan an lar has the smallest electorate of any seat in the country with just over 20,000 voters, only a third of the size of most constituencies. Attempts to link the counstituency with others have always foundered on the geographical size of the area and problems of travel and communications for the MP. Politically the seat has been a marginal between the SNP and Labour since the war. In both 2005 and 2010 it was one of only a few seats where the Conservatives failed to retain their deposit. In the 1975 referendum on continued membership of the European Union, the Western Isles and Shetland Isles were the only areas to vote no.

Current MP
ANGUS MACNEIL (Scottish National Party) Born 1970, Barra. Educated at Nicholson Institute, Stornoway and Strathclyde University. Former teacher and BBC worker. Contested Inverness East, Nairn and Lochaber 2001. First elected as MP for Na h-Eileanan an Iar in 2005. A native speaker of Gaelic. MacNeil brough the original complaint that lead to the police investigation into "loans for peerages".
Past Results
Con: 647 (4%)
Lab: 4838 (33%)
LDem: 1097 (7%)
SNP: 6723 (46%)
Oth: 1412 (10%)
MAJ: 1885 (13%)
Con: 610 (4%)
Lab: 4772 (34%)
LDem: 1096 (8%)
SNP: 6213 (45%)
Oth: 1145 (8%)
MAJ: 1441 (10%)
Con: 1250 (9%)
Lab: 5924 (45%)
LDem: 849 (6%)
SNP: 4850 (37%)
Oth: 286 (2%)
MAJ: 1074 (8%)
Con: 1071 (7%)
Lab: 8955 (56%)
LDem: 495 (3%)
SNP: 5379 (33%)
Oth: 206 (1%)
MAJ: 3576 (22%)

2015 Candidates
MARK BROWN (Conservative) Financial services manager.
ALASDAIR MORRISON (Labour) Born 1968, Stornoway. Educated at Nicolson Institute. Journalist. Contested MSP for Western Isles 1999-2007.
RUARAIDH FERGUSON (Liberal Democrat)
JOHN CORMACK (Christian)
Comments - 623 Responses on “Na h-Eileanan an Iar (Western Isles)”
  1. I wonder if the swing in this seat might be higher than average if the ‘Vicar’s daughter’ is regarded favourably. Also not sure if Christian Party (who held their deposit the last two elections) are standing.

    It also appears to be a relatively high swing seat.

  2. Angus MacNeil voted against a general election yesterday, that’s the first time that I’m aware of that any SNP MP has defied the whip.


    As with 4 years, US Presidential elections – I have selected this constituency it’s the nearest one to the USA that has a thread on UKPR (there are No NI threads).

    Just as fascinating as in 2016. Can Trump hold onto power? Biden leading in the polls at the moment but two months to go. My feeling is for Trump to again lose the popular vote but he’ll get just over the 270 votes needed.

    FYI: Turnout in recent US elections, as a percent of eligible voters:

    2004 56.7%
    2008 58.2%
    2012 54.9%
    2016 55.7%

    The postal votes (Mail-in) system being arranged for the whole country is bound to increase overall turnout but who will it benefit?

  4. It will be an absolute joke if the Republicans yet again lose the popular vote yet win the electoral college. Once would be ok, but three times in 20 years would be laughable.

    I think Biden will flip PA, WI, MI, FL and AZ to win the Presidency. NC and GA will be close but no cigar for the Dems.

  5. I’m beginning to worry that Biden’s lead is diminishing to such an extent it will be eradicated by polling day. Americans have form here – only rarely does an incumbent President get defeated and whilst it seems implausible to anyone with a brain that this, lets be honest, out and out scumbag will get a second term, most such people – myself included – thought he would never get the nod in 2016 – despite his flawed opponent.

    Biden’s leads in Arizona and North Carolina have already disappeared and recent polling from Wisconsin and Ohio suggests he’s almost neck and neck with Trump – Wisconsin one of a handful of states to back liberal Dukakis in 1988

    Biden seems incapable of energising his base to the extent that Trump can and his best – perhaps only – chance of victory is to get his vote out – the way Obama did in 2008 and even more so in 2012. If he manages that he’ll defeat Trump.

  6. Also demographics are in the Democrats favour with those trending Democrat – Virginia, Texas, Georgia – worth more electoral college votes than those states trending the other way – West Virginia, Iowa

    Whilst I doubt that the likes of Texas will vote Democrat in November, the influx of non-whites means it’s likely to go the way of California – which up until the early 90s was reliably as Republican as Texas is today

  7. Three interesting posts…t.y.

    There are 2 good websites…270towin and

    538 gives an updating average of 2020 presidential general election polls and it’s currently
    JB 50.2%
    DJT 42.9%.

    Two goods out polls today for Biden in Wisconsin and AZ. But a shock one in PA for him, showing Trump level with him.

  8. I don’t agree with Tim regarding Arizona…Biden’s lead there has been pretty consistent, including a couple of 10 point leads in the last couple of days. He was never consistently in the lead in NC anyway so I don’t regard the polls narrowing there as any surprise.

  9. AZ is down for me as a Trump hold.

    Hispanic voters make up 25% of the ELIGIBLE VOTERS IN Arizona. Traditionally, nationally around 25% vote REP.
    I believe this time Trump will do fairly well in the Hispanic/Latinx vote, better than what intuititively people might think. He might do BETTER than last time!

    A poll of Hispanics in Jan put Trumps rolling average figure at around 35 percent, and rising.

    If Trump does 12 percentage points better than his 2016 numbers in the Hispanic vote, it pretty much takes Florida, Arizona off the table for Democrats plus, Georgia and North Carolina. They would need to sweep Michigan, Pennsylvania and Wisconsin to reach 270.

    Also done a bit of research and it seems most agree that mail in voting (postal votes) does not make that much of a difference or give a significant advantage to one side or the other.

  10. Also a factor is the TV debates. Trump is a great TV performer. Electrifying. Biden will struggle. Around 80 million will watch. Potential to turn polls on their heads.

  11. Is Trump a good tv performer – to me he’s an incoherent rambler who lies almost every time he opens his mouth. You would have to have an IQ bordering the mentally retarded threshold to be inspired by him and his rhetoric.

    None of that of course gets over the fact that Biden himself is a poor, uninspiring public speaker.

    The polls seem in flux. Last time I looked at Pennsylvania Biden had almost a double digit lead and as unbelievable as it might seem, if the Hispanic vote is more sympathetic to Trump this time round, it does indeed make the likes of Arizona, and more crucially Florida, much more likely to stick with Trump

  12. Trump’s three biggest drawbacks are his general ignorance, his inability to tell the truth and most crucially the fact that he comes across as a thoroughly unpleasant individual – the more you see him the more you dislike him

    In 2016 he was up against a similarly divisive figure with plenty of question marks of her own with regards to personality, so I guess many voters were willing to give Trump the benefit of the doubt

    When you add all his known misdemeanours into equation – whether that be colluding with Russia in 2016 or sleeping with professional whores – it seems unfathomable that such an immoral man could get elected in a supposedly Christian country like the US, but as I say, Americans have form here so be afraid.

  13. I wonder if we will see any change in the BBC’s reporting on the US election given that new director-general Tim Davie is reportedly a massive Trump fan

  14. Not a Trump fan, just saying he has the ruthlessness to land some big blows and/or spring one or two suprises on Joe, who is a bit doddery. run by Nate Silver are currently saying Biden is a 71% favourite. I’m surprised by that, especially with the Trafalgar poll out today for Florida giving Trump a 3% lead – the first time he has had a lead in a Florida poll since mid March.

  15. If Florida really is swinging behind Trump then I’d say it’s 50/50 and comes down to the industrial states in the Midwest. Ohio and Iowa look set to stay in Trump’s column, on current polling Michigan and Pennsylvania look likely to switch while Wisconsin and Minnesota look too close to call. You would also expect Trumo to build on the recovery he’s making in Texas, Georgia, North Carolina and Arizona.

    Since Clinton’s landslide in 96 and with the exception of Obama’s debut win in 2008, every recent presidential race has gone down to the wire and 2020 looks set to build on that trend

  16. I do wonder whether the chatterati “Trump’s gonna edge it” consensus is simply on overreaction to them all getting it wrong last time. Over the last few years, media pundits have pretty consistently found themselves fighting the last war.

  17. I had a look at the states yesterday and as it was in 2016 it’s very hard to see how Trump could plot a route to victory

    Assuming there will be a not insignificant swing back to Trump – it’s already started with the race now neck and neck in Florida – The best he can really hope for is 259 in the electoral college max – and that assumes he wins Florida, Ohio, Arizona, Iowa, North Carolina and Maine’s second district – and that would still see him losing – although by the sort or margin that would see him turning to the courts.

  18. Yes but it’s doable for him. In all battleground Stars Trump has closed the gap (compared to Aug 17th polls). Biggest catch up in FL:

    Reduction in Biden lead (Aug 17th / Sept 8th)

    Florida +5.3 +2.8 -2.6
    Colorado+13.4 +11.4 -2.0
    Ohio +0.5 -0.9 -1.4
    Pennsylvn +6.4 +5.0 -1.4
    New Hants +9.3 +8.2 -1.1
    Georgia -0.9 -1.7 -0.8
    Nevada +6.8 +6.4 -0.4
    Michigan +7.6 +7.4 -0.2
    Virginia +10.8 +10.8 -0.1
    Wisconsin +6.8 +7.0 +0.1
    Nth Carolina +1.3 +1.7 +0.4

  19. Yes but it’s doable for him. In all battleground Stars Trump has closed the gap (compared to Aug 17th polls). Biggest catch up in FL:

    Reduction in Biden lead (Aug 17th / Sept 8th)

    Florida +5.3 +2.8 -2.6
    Colorado+13.4 +11.4 -2.0
    Ohio +0.5 -0.9 -1.4
    Pennsylvn +6.4 +5.0 -1.4
    New Hants +9.3 +8.2 -1.1
    Georgia -0.9 -1.7 -0.8
    Nevada +6.8 +6.4 -0.4
    Michigan +7.6 +7.4 -0.2
    Virginia +10.8 +10.8 -0.1
    Wisconsin +6.8 +7.0 +0.1
    Nth Carolina +1.3 +1.7 +0.4

  20. It’s doable but he essentially fluked the rust belt states of Pennsylvania, Michigan and Wisconsin and I just don’t see him winning any of those states – or any other state that might compensate for their loss

    When you consider quite how badly Trump has performed as President – by any measure – it’s hardly a ringing endorsement of Biden and/or the Democrats.

    Of course Biden could record an Obama-style victory in 2008 – but at the moment it looks like it will be another tight race

  21. Well, he’s still getting decent polling on the most important issue – the economy and jobs. It’s better than recent previous 2nd term Presidents going into October.

    Just checked the Next President betting and it’s
    54% BIDEN
    45% TRUMP
    on Betfair and
    55% BIDEN
    53% TRUMP
    on William Hill online

    So people risking their money think it’s a lot closer that the election websites or what the polls are implying.

  22. Although to be fair when Biden started getting double-digit leads in June, I still spoke to many Democrats who thought Trump would still win – by hook or by crook.

    The obvious comparison is Michael Dukakis in 1988 who maintained a healthy lead over George H W Bush up until the last two weeks of the campaign.

    Whilst Dukakis can be compared to Biden, Trump is no Bush – who was arguably the last President to rule a relatively united America

  23. BIDEN still leads in national polls by about 7% or 8%. State polls also show a relatively stable race, with most — excluding those in Florida — containing good news for Biden. Arizona is the state where Biden’s odds of winning have improved the most.

    A lot of conversation on the U.S right now though is about serious delays in the anouncement of the results and even a contested final result – where Trump is ahead on the night but Biden overtakes him later when all the postal votes are counted.

  24. Trump bring some of the Arabs and Jews together in a ‘diplomatic triumph’. We’ll surely see a small uptick in his polling?

  25. Polling for both candidates has been very stable now for about a month. Neither of the major parties conventions, the recent police shootings and the subsequent protesting and rioting seem to have made any difference with Biden hanging onto a 7/8% lead

    Surely if he hangs onto that type of lead (I doubt he will) then surely the result will be known not long after the polls close.

  26. “Results will be known not long after the polls close.”

    Well, assuming the actual electoral process runs smoothly and fairly. I know we should be able to rule out voting irregularities in a mature western democracy, and yet…

  27. I think as things currently stand, Trump’s expects to lose – but hopefully for him it will be narrow enough for him to claim he actually won and take it to the courts.

  28. With where the votes are being won and lost, and with it being 50 elections remember not 1, Biden will have to be significantly further ahead in the popular vote than Clinton was last time in order to win the electoral college.

    Or else adapt his strategy better to focus more on Arizona, Georgia, Missouri, Texas and on defending Nevada and Minnesota – over-focusing on Pennsylvania, Florida and Wisconsin at the expense of everywhere else is a dangerous strategy.

    Think Biden might perhaps just hang onto Minnesota, but will lose New Hampshire and probably Nevada.

    Think Trump will lose Arizona and Michigan, but hold N Carolina and probably Pennsylvania. Can’t see him losing Florida, though as usual won’t be a massive margin.

    Trump will hold Iowa and Ohio with comfortable margins, Dems should not waste resources in Ohio.

  29. P.S. Think Wisconsin is a toss-up

  30. I agree that Iowa and Ohio are likely to remain with Trump, but I think the leads Biden has established in New Hampshire, Arizona and Pennsylvania should be enough to see him over the line.

    I don’t see Biden winning Texas, Georgia or Missouri and his lead in right-wing North Carolina is likely to evaporate by Election Day. Minnesota is always seen as a swing state yet it even voted Democrat for each of their landslide defeats in the 1980s. It’s almost unthinkable that it might flip Republican

    Florida is still the crucial one. If Biden losses there it’s unlikely he will won the presidency. For Trump, anything than a victory will be curtains

  31. FLA polls give Biden a slender lead – between 1%-3%. Myself, I’ve long had it down as a GOP hold.

    The turnout is US Pres elections are so poor :
    2004 56.7%
    2008 58.2%
    2012 54.9%
    2016 55.7%
    that I just feel if that goes up to, say close to 65% – and with a surge in postal voting that’s foreseeable – we could see more than a few extraordinary results.

  32. In 1996 and 2000 it was even worst, standing at 49% and 51.2% respectively

    If it did go up to something like 65% it would be unchartered territory- and many pollsters believe that would give the Democrats in states like Texas and North Carolina.

    Not too sure myself about Texas but yes evidence does point to there being a much larger turn out this year – I think Trump is that kind of president that you would make an effort to re-elect or dump.

  33. Trump announces Judge Amy Coney Barrett as his nominee to the SCOTUS seat left by Ruth Badger Ginsburg’s passing. No surprise.

    Next Weeks first TV debate is unfortunately for us at 02:00. (Weds). I’ll be getting up however & starting late in the morning. Can’t wait.

  34. Biden doing ok but looking and sounding frail. However, I dont think this is going to change much….so in that sense a win for Biden as he’s likely to hold onto his lead in the polls.

  35. Betting markets are shortening odds on Biden. A small contraction but perhaps confirmation that Biden did enough to assuage fears that he was gonna be a disaster on the night.

  36. So, Yougov have done an MRP. Biden leads by 53% to 45% in the popular vote, and 350 to 188 in the electoral college.

    I’m not sure how useful this is, it would have been better for Yougov to announce their result in terms of probabilities of each candidate winning the presidency, as 538 has done. We all know Joe Biden is in the lead, but so was Hilary Clinton at the same point four years ago, for all the good it did her.

  37. Biden would have to win back all the Midwest industrial states including Ohio, and the likes of Florida, Arizona and North Carolina to get to 350.

    I personally thought Trump came over horrendously last night – yet again – whereas Biden didn’t look like someone who would be incapable of being President. If Trump’s the measure, he’s surely in good health and of sound mind.

    I hope it’s enough to at least stall Trump’s modest recovery in the polls

  38. Have had a good rethink and check and on my current thinking and using one of these interactive maps I come up with a TIE! 269-269 each.

    Trump wins all in South apart from NM. Wins rust belts Michigan, Ohio, Wisconsin and Indiana.

    Biden takes Minnesota and Illinois.

    This is despite I suspect Trump will win…possibly by taking Minnesota and/or New Hampshire.

  39. All the South. I do not see Trump winning Virginia – which is well on its way to becoming a safe democrat state – and whilst I suspect he recover enough to win North Carolina, Georgia and Florida – polls there have been remarkably consistent

    And I don’t see Trump winning either New Hampshire or Minnesota although I think he’ll probably retake Ohio and Arizona – despite current polls predicting he won’t

  40. The biggest problem in calling the US election is the limited polling in swing states. I keep an eye on real clear politics and their updates swing quite heavily based on a single poll where there may not have been many in that state to start with.

    It does seem though that whenever I look Biden seems to be ball park to where Clinton was on eve of poll. So I guess that the main question is whether the polls are more accurate this time round- if they are then Biden looks good for it.

    Also hard to imagine that Trump won’t have put some people off and that soft Dems are more likely to get out and vote (not because of Biden but because of Trump) but that could be countered by voter suppression.

  41. I think Trump’s dislikability no longer masks the fact that on his own merits, Biden is a fairly average Presidential candidate – more John Kerry/Mitt Romney – just older – than Obama/Bill Clinton – so historical precedent doesn’t necessarily hold well for Democrats.

    Their hope has to be than Trump’s innate nastiness and general incompetence is enough to get people to the polls to vote. At the moment it looks like it is, but I still think things could change.

    Having said that Biden’s lead has been remarkably consistent over the past 2/3 months or so and he’s still recording the sort of narrow leads in Republican leaning state’s like North Carolina, Arizona and Georgia that he was back then

  42. “It does seem though that whenever I look Biden seems to be ball park to where Clinton was on eve of poll.”

    That’s not correct. Biden is polling markedly better than Clinton and, as Tim says, the polls have been showing a pretty consistent picture for many months now.

    Right the way back to the Spring, Biden’s polling average has never once dipped below 50%. Clinton’s polling rarely topped 50%. By eve of poll in 2016 the polling gap was in the region of Clinton 48ish Trump 45-46ish which is pretty close to the actual result. Currently Biden is 51-52ish with Trump still at around 46 as in 2016.

    Yes there’s still time for things to change, yes the COVID diagnosis might still upend the election one way or another, but if current polls are reflected in the result then a 51-46 victory to Biden in the popular vote makes it almost impossible for Trump to win the electoral college.

    Trump’s big problem is that everybody has an opinion about him, good or bad, so pretty few undecided voters this time.

  43. @ HH

    I was talking about swing states not national picture:

    If you scroll down you will see the current swing states so for example…

    PA- Biden 0.3% smaller lead than Clinton eve of poll
    FL- Biden 1.3% smaller lead than Clinton eve of poll

    It’s 50:50 on the 12 shown whether Biden is higher than Clinton but some of the large states he has lower leads.

    These leads to swing around a lot though as one more recent poll can make a big difference.

    All healthy enough leads for Biden if you can believe the polls this time but the polls on the swing states were seriously wrong last time.

    Personally I think Biden should be good for it for all the reasons I mentioned and the fact than many states Trump only just squeezed in last time. I’ve still no certainty which way this will go though and with the postal votes issue there could be significant voter suppression.

  44. .. DO swing around

  45. Yes, Biden’s poll figures are good. Anyone expecting a Trump win, as I do, is swimming against most pollsters and pundits.

    But remember Brexit, remember Clinton 2016, Corbyn 2017…recently been a lot of polling screw ups. And I am suspicious that increased turnout and mass voting by post will unnecessarily benefit the Democrats. I think it could be the reverse! I think there’s something in “the silent majority” argument.

  46. I think in all of those cases, they weren’t really polling screw-ups as much as the chattering classes screwing up their interpretation of polling.

    Look at the polls in the couple of weeks before Brexit. On any objective reading, the vote was 50-50. A lot of “metropolitan elite” pundits (apologies for the cliché) read those same polls and confidently declared that remain would edge it.

    Biden is like 10 points ahead. Absent either a big swing in the next month (which would have to be even bigger as people are already voting), or the polls being out by multiple standard deviations, he will be the next president. 538 has him at 81% likelihood of winning, and I think even that’s too cautious.

  47. It’s probably true that foreigners are more confident about a Biden victory than most Americans are. My American wife pretty much agrees with Mark Felt / Deepthroat’s view. But maybe that’s just because as a committed Democrat she is expecting the worst so the bad news won’t be so devastating this time.

    One issue that the foreign media has got totally wrong is this lazy blanket narrative that “America has opened up too soon”. Yes that’s true in some states but the opposite is the case in others. My relatives in Illinois remain quite heavily locked down – my nieces still haven’t been able to return to school having last been there in March. It’s the same across much of the midwest. I was talking to a business contact in Michigan the other day where it is a similar story. In these locked down midwestern states, Trump blaming “Democratic governers” keeping the states locked down for political reasons is a very fruitful narrative. Or at least it was, whether Trump’s own diagnosis changes things is another uncertainty.

    So I think Trump will probably overperform again in the midwest and there is probably something in what Shevii is saying on polls in those states. However what Shevii didn’t mention is Trump’s awful polling in other key states where it wasn’t even close last time – Arizona for example. Because of this, even if the midwest holds up fairly well for him, it’s hard to see a path to a Trump victory in my view.

  48. It was hard enough to map out a Trump route to victory in 2016 – but it’s even harder this time round because even in the unlikely of his vote holding up in the Midwest – as Hemmelig says there’s plenty of Republican-leaning states that have been showing leads for Biden now for a good few months

    And it looks as if Trump catching corona does t seem to have generated any political sympathy beyond his base

    Whilst I can understand why somebody like Ronald Reagan would inspire mass respect amongst the Right – it beggars belief that somebody as incompetent and dislikeable as Trump has any base at all

  49. Hemmy: “It’s probably true that foreign observers are more confident in a Biden victory than most Americans are.”

    True, but I would dispute the unstated assumption that the latter are more reliable. Information “on the ground” counts for little in a nation as large and geographically polarised as the US. Meanwhile, American liberals obviously have more skin in the game than their European counterparts, and this can lead to an emotional, rather than objective, interpretation of the data.

    I mean, it would be weird if those people weren’t fearful/desperate/ whatever, but whatever the merits of emotion in politics (and there are plenty), a dispassionate reading of the electoral runes isn’t one of them.

  50. I was really annoyed actually when I read an article Online about 2/3 weeks ago which had been published in the The Guardian.

    It was saying how Trump had no chance of winning, all the polls were swinging against him (when if anything the reverse was true) and that he just didn’t have a path to victory in the electoral college. It was quite clear the writer – obviously a liberal – was projecting his view of Trump onto the American people

    This is a paper with form of trying to influence American elections and failing – remember 2004.

    Europeans underestimate Trump at their peril- but it’s so easy to do because to nearly everyone outside the US he seems utterly horrendous.

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)