How the NI political landscape has changed since 2010

It could be argued that Westminster Elections are now very much of secondary importance and interest in Northern Ireland, compared to Stormont elections. However, this is shaping up to be a very peculiar election for the UK as a whole, and there is a very real possibility that the DUP could play a major power-broking role in the next parliament. Besides, first-past-the-post elections in a multi-party system are gloriously silly, given the prospect that you could “win” some seats with only 15% more of the vote than you need to win back your deposit. This makes them fun to analyze.

The interactive Tableau visualization above shows 2010 Westminster vote shares, performance at the 2011 Stormont election by number of first preference votes, and an estimate of Westminster constituency performance at the 2014 local elections.  Allocating these votes perfectly to Westminster seats is not possible as some council electoral areas straddle constituency borders, so I’ve built a model to allocate electoral areas to Westminster seats based on the number of wards. This is somewhat of an inexact science, so remember; it’s all just a bit of fun.

You can select which Constituency you want to view from the drop down menu under “Constituency Name”, or you can toggle through them.

East Belfast, on the face of it, would appear to be a very difficult ground for the Alliance Party to hold, although it’s certainly possible. Cheerier news for the Alliance Party is to be found in South Belfast, where the model puts them in second, 3% behind the DUP and one percentage point ahead of the SDLP. This is definitely a winnable seat for them.

In North Belfast, the model saw Sinn Féin pip the DUP by 0.1% in 2014. The simplified approach that I used for allocating electoral area results to Westminster seats may have played a part here, but it’s certainly tightening. Nigel Dodds could potentially wake up on 8 May 2015 to find that he has become one of the most powerful men in the UK, given that the DUP support could be vital to forming a new government. However, he has to win first, and this is by no means guaranteed.

I’d previously thought that SF might have been able to sneak through in Upper Bann, due to them getting the most first preference votes in the 2011 Assembly Election by a handful of votes, but I’ve changed my mind. It would require an absolutely perfect storm of an equal DUP and UUP vote, a strong TUV performance and a SDLP collapse. It’s possible, but highly unlikely.

Elsewhere, there has been a rebound in UUP support in East Antrim and East Londonderry, but probably not enough to pose a serious risk to the DUP incumbents. Sinn Féin support has fallen in Fermanagh & South Tyrone, whilst the SDLP’s stock has risen, so a single unionist candidate would have an excellent prospect of picking up the seat. In fact, the UUP have a reasonable shout at picking up the seat even if the DUP do stand, but it would be very hard.

Foyle looks like it could on its way to electing a Sinn Féin MP, whilst the SDLP appear to be holding on in South Down.

If I was forced to make predictions for the election, I would probably plump for the following. Italics denote a gain.

DUP (10): Belfast East, Belfast North, Belfast South, East Antrim, East Londonderry, Lagan Valley, North Antrim, South Antrim, Strangford, Upper Bann

Sinn Féin (6): Belfast West, Fermanagh and South Tyrone, Foyle, Mid Ulster, Newry and Armagh, West Tyrone

SDLP (1): South Down

Independent (1): North Down

,

  • This also presumes people vote same Local to Westminster, which kinda falls down when you look at the election on the same day when votes didn’t cross local to European. So interesting exercise, but whether it means anything depends on what you want to believe.

  • John Gorman

    Agreed. The very first graph shows South Belfast whose voting pattern varies greatly between STV and FPTP elections. To try and predict the 2015 Westminster election we can only really use the 2010 results.

  • John Gorman

    As for the final predictions there maybe a few surprises but Im 99% sure there will be no change other than East Belfast which slightly favours DUP at the moment but Alliance still has a chance.

  • salmonofdata

    The model does show the 2010 incumbents being ahead in 13 of the 18 Westminster seats.

    Disregarding the obviously special case of North Down, and North Belfast where the model is effectively a draw, there are three seats that look like they could flip; Foyle, Belfast East and Belfast South.

    I reckon that a bookie that was offering 100/1 on any one or more of Alasdair McDonnell, Mark Durkan, Michelle Gildernew or Nigel Dodds losing their seat would have a queue down the street.

  • Zeno1

    “I reckon that a bookie that was offering 100/1…………”

    All 4 to win can’t be much bigger that 7/1. The 100/1 will never happen.

  • John Gorman

    I would be amazed if Durkan lost his seat. For example the recent local elections had numerous independents who wont be standing in May skewing the results likewise Durkan will also get a healthy injection of tactical Unionist votes. McDonnell was only ever in danger if Anna Lo stood as she could pull votes from everywhere including a big pale green vote that Paula Bradshaw will not get as she stood as a Unionist only 5 years ago. Gildernew won last time against a good unity candidate and a good SDLP candidate I would expect her to win comfortably this time. Dodds might be in a spot of trouble but if the UUP pull out it will give him an extra 1500 votes and I think the SDLP have been squeezed to the max here and will retain a core vote so Dodds will be ok just. As I said earlier only East Belfast will be interesting. Upper Bann could throw up a surprise but Im pretty sure a stop SF campaign by the DUP will prevail.

  • Paddy Reilly

    Unionists have lost one seat in each of the last eight Westminster elections, giving them an unbroken record of small but significant defeats extending over three decades.

    That you expect them to win back two and lose none in 2015, just one year short of the grand dépassement of 2016, with the Unionist traditional support base haemorrhaging on every side, flies in the face of common sense.

  • salmonofdata

    That was referring to John Gorman being 99% sure that there would be no change anywhere except East Belfast.

  • mjh

    Salmon Of Data has done us a favour in focusing our election predictions onto solid data. Political parties will do all in their power to win “tactical” votes from other parties supporters. They stress that “only we can beat party X”. This is fair enough if it is true.

    But it can also be a cynical tactic, when the claim is made without a reasonable basis in fact, simply to cause longer term damage to a party competing for similar voters.

    Voters deserve to know what is a fair claim – which they may chose to act on – and what is a trick designed to manipulate them.

    But Dissenter and John Gorman make fair criticism of Salmon’s analysis of the data. Voters consistently vote for different parties in different types of elections – even when they are held on the same day. Of course this does not mean that we have to ignore the elections of 2011 and 2014, we should analyse carefully what they tell us about any change that may have taken place since 2010. But Salmon goes too far in using the last Council elections in isolation to predict that the SDLP will lose South Belfast and Foyle. As John points out, there were a number of nationalist leaning independents in Derry and Strabane who muddy the waters. In fact they took more than 10% of the total vote.

    However, this posting will make me look again at South Belfast where there were no independents in 2014.

  • Isn’t it strange that only one of the partlies in Stormont has been left out of the presentation of the analysis? Has Salmon of data been taking lessons from the BBC about deliberately ignoring the Green Party?

  • Nordie Northsider

    I agree. Another issue is SF’s failure to find a credible candidate in Foyle since Mitchell McLaughlin was transferred to South Antrim. Unless they come with better, that seat is Durkan’s for life.

  • salmonofdata

    I should have saw that one coming. The official excuse is that, because it’s a first past the post election, I only included parties that had higher than a 10% share of the vote in 2014 in their best Westminster constituency, so the TUV qualified (17% in North Antrim) but the Green Party didn’t (8% in North Down).

    The real reason is that because I thought that 3 green lines on a graph made it too difficult to read.

    I’m bullish on the Greens in GB for this election overall, though. I reckon they’ll hold Brighton Pavilion and pick up a couple from Brighton Kemptown, Hove, Bristol West, Lancaster & Fleetwood, Sheffield Central and St. Ives.

  • salmonofdata

    I concede the point on Foyle. Mark Durkan’s personal vote is probably enough for him to see off SF, but an election between “generic” SF and SDLP candidates would be extremely close.

    I still think that South Belfast is a lot more at risk for the SDLP than most people think. Máirtín Ó Muilleoir is a popular candidate, who will probably beat Alex Maskey’s 9% share in 2005 when Sinn Féin last stood, and these votes will mostly come at the expense of Alasdair McDonnell’s vote.

    Throw in the fact that the UUP vote has been migrating to the Alliance Party in the area for years, and you are essentially left with a three-way tie between the SDLP, Alliance and the DUP. It will be close.

  • MainlandUlsterman

    But more specifically, where do you think he has got it wrong?

    East Belfast was always likely to be regained (though who knows) and Belfast South is a knife-edge area … doesn’t seem unrealistic to me that both could go DUP. I wonder if you maybe just assumed unionist representation would continue to go down no matter what? That seems to me to fly more in the face of common sense than Salmon of Data’s analysis.

  • Nicholas Whyte

    Thanks for crunching these numbers, Salmon. I have just finished processing the local election results onto my website, and draw broadly similar conclusions to you, with a few exceptions.

    I agree with the critiques of mjh and TheDissenter in general, and would add that if the UUP are to gain anywhere it will be Upper Bann. Your figures have them on 24% to the DUP’s 27%, and SF on 21%, in last May’s local elections; my own (very rough) calculations have the three even closer with the UUP also on 24% but the DUP on 26% and SF on 22%. There is another 8% of spare Unionist votes from the local elections to add to whichever candidate impresses the relevant voters more. Not being on the ground, I don’t know whether that is Simpson or Dobson, but I can’t see anywhere else that is as good a prospect for the UUP – like you, I have them much further behind in South Antrim.

  • Nicholas Whyte

    Paddy,

    Excuse my ignorance, but what is the grand dépassement of 2016?

    And what is your evidence that the “Unionist traditional support base” is “haemorrhaging on every side”, given that the combined Unionist vote share increased and the combined Nationalist vote share decreased in both 2014 elections?

  • mjh

    Salmon of data,

    I believe that there is a problem with the model for estimating the 2014 local vote share on a South Belfast constituency basis which is distorting the SDLP share downwards and inflating others.

    Where a District Electoral Area is split between two or more constituencies the model has to assume that each party’s voters are evenly distributed throughout the DEA. I suspect that in Lisnasharragh and Castlereagh South nationalist voters are actually more heavily concentrated within the South Belfast parts of the DEA’s.

    Whilst the figures for all elections in the last ten years show that both the unionist and nationalist vote shares have been declining in the constituency, the model suggest that the nationalist share has fallen from about 36.5% in 2011 to about 31% in 2014,
    which is unlikely.

    Nevertheless the 2014 estimates do show that there has been an undeniable drop in the SDLP share of the vote. Even on the maximum margin of error the SDLP would still only have taken 21.4% of the vote. Down from 23.9% in both the council and assembly elections of 2011 (Nicholas Whyte’s figures.) Almost certainly the drop is greater than this.

    Despite this the data from the last ten years still points to a McDonnell victory as the most likely outcome. It shows:

    1. The total unionist leaning vote has been in consistent decline and was still
    declining in 2011 when it hit 38.2% in the councils and 38.6% in the Assembly.
    It is unlikely to exceed 40% this year. 38% may be more likely.

    2. There is no evidence that the DUP has squeezed the other unionist parties in the two Westminster elections.

    3. Between 2005 and 2010 the DUP consistenly took between 55% and 58% of the unionist leaning vote at each election. It then rose to just under 63%
    in 2011, but appears to have fallen back again this May, to around 55%.

    4. The total nationalist leaning vote has also been declining consistently, if more
    slowly, from a high of 41.7% in the 2005 councils – down to 36.4% and 36.8% in
    2011. On the evidence of last May’s elections is unlikely to exceed 36% this
    year.

    5. Apart from Westminster the SDLP share of nationalist votes has remained at 65/66% in all elections; with SF not moving from 33/34%. It looks like there was a swing to SF at the recent council polls. The split may have changed as far as 60/40,
    but the margin of error makes it difficult to be precise.

    6. There is evidence for a strong squeeze by the SDLP on SF in the 2005 Westminster election. SF dropped to 22% of the nationalist leaning vote. This compared to 35% for the councils on the same day. SDLP share went up to 74%%.

    7. The centre ground has been growing strongly. It reached nearly 25% in the 2011 elections. It looks as if it grew further this May with the entry of NI21 and a higher
    Green total, taking it to a maximum of 29%, but probably nearer 27%.

    8. The centre was squeezed very badly in the Westminster 2005 election – losing half of its vote share. But there was no squeeze at all in 2010. It actually grew at that election – so it seems reasonable to expect it will at least hold between 25% and 27% this time.

    9. The Alliance share of the centre vote has remained between 79% and 83% apart from 2005 when no Green stood. The entry of NI21 pushed its share down this May to somewhere around 66%. Without NI21 the Alliance share should rise back to a similar level.

    10. Transfer patterns show that McDonnell has built a particularly strong personal vote, attracting an unusual number of voters who transfer their second choice to a
    unionist or centre candidate instead of to an SDLP running mate. Indeed the
    absence of McDonnell from the ballot paper in May may have been an important
    factor in the party’s apparent decline. This could make all the difference if the race is tight.

    So if both the UUP and SF contest the election the most likely outcome is:
    SDLP: Min 22% Max 27% Most likely 25%
    DUP: Min 21% Max 27% Most likely 24%
    Alliance Min 18% Max 24% Most likely 22%

    Very close with much depending on the campaign,

    If there is a United Unionist candidate and SF pull out there may well be a squeeze on the centre. Given McDonnell’s track record of appealing outside of the nationalist
    base I am guessing that he would gain more from that.

    If SF pull out but the UUP do not, McDonnell will walk it.

    If there is a United Unionist but SF still stand, all the data tells you is that SF would likely hold at least 8% of the vote however effectively they were squeezed. We don’t know how the unionist leaning voters would react in this constituency – much would
    depend on the candidate. A UUP candidate would stand the best chance. A DUP
    candidate or a candidate seen as a DUP surrogate could have trouble attracting more than half of the UUP voters.

  • salmonofdata

    I’d agree with all of that. I think the same issue is inflating the Sinn Féin vote in North Belfast for 2014.

    From a unionist perspective, I can’t see that a unity candidate would make a lot of sense in South Belfast. There would be no incentive for the DUP to throw away a good chance of a gain to benefit the UUP, whose support has collapsed. A lot of the DUP vote would probably stay at home, or vote TUV or UKIP if they are running a candidate. On the other hand, the UUP standing aside for the DUP would probably help out the Alliance Party just as much as the DUP.

    If the UUP and the DUP were minded to enter a pact with each other, a straight swap between Belfast North and Fermanagh and South Tyrone would make more sense.

    Whilst I agree that the Alliance Party could get close, as you point out it’s the final couple of percent of the vote that would put them on top that is the problem. If, for example, the combined NI21/Green vote was even as high as three percent, it would probably be enough to deny the Alliance.

    In the interests of transparency, I have saved the mapping that I used to map between Council DEAs and Westminster seats as a Google Spreadsheet at https://docs.google.com/spreadsheets/d/1KTXQBig1IUUZutVahHW_2BsXdtkcLsTklV9YAKHn6So/pubhtml?gid=0&single=true Any fellow geeks are welcome to poke holes in it.

  • salmonofdata

    I think you may be being a bit harsh on the Alliance candidate in South Belfast. If I’m reading the results correctly, Paula Bradshaw was elected in Balmoral in 2014 well over quota with a substantial transfer from the SDLP. She also commanded a respectable 17% vote in 2010 under the UCUNF banner.

    Given a strong Sinn Féin vote attracting votes away from the SDLP, if the Alliance were to maintain Anna Lo’s vote in 2010 (not inconceivable given the strength of the Alliance vote in South Belfast in 2011 and 2014 when Anna Lo wasn’t on the ballot), and attract 40% of the UCUNF vote from 2010, then this would be enough to put Alliance there or thereabouts.

    It’s a very big ask, though, and I agree with mjh that Alliance will come a strong third. There are plenty of scenarios where the Alliance come within a few percent of a win, but for them to actually win would require a virtually perfect storm.

  • John Gorman

    MoM is charasmatic and certainly would be more transfer friendly than most SF candidates. In an STV election he has the potential to improve on SF’s usual 12% in South Belfast but who is going to switch to him in a FPTP election when he is starting from a low point, SDLP supporters? Alliance voters? Answer nope. If he stands he will get that 12% max of the vote but no more. Funnily enough he would probably have been more effective in North Belfast as SDLP supporters and even some Alliance voters would be more likely to lend him a vote as opposed to Gerry Kelly. Also the chances of a Unity candidate here are pretty slim. Mike Nesbitt initially suggested FST and North Belfast. Its a bit easier for him to explain a pact to beat an abstentionist SF candidate its a bit harder for him to explain the logic to oust an SDLP or Alliance MP who actually sits in parliament.

  • John Gorman

    McDonnell seems to get votes from a broad spectrum for some reason, it may be due to the least worst option which we see come into play a lot more in FPTP elections. If its just the big 5 standing then McDonnell will get that 27 max you mention or quite likely even higher it will only drop down to the lower figures you mentioned if the Green Party and socialist or PBP stand.

  • Paddy Reilly

    The Unionist traditional support base is the Protestant population of Northern Ireland.

    The Grand Dépassement of 2016 is the point when the Protestant Population of Northern Ireland ceases to outnumber the Catholic population of Northern Ireland. Projecting forward the growth of the Catholic population and the fall of the Protestant population between 2001 and 2011, this will probably be sometime in 2016, probably November or December.

    Year*Protestant****Catholic
    2001***895377****737412
    2011***875717****817385

    Ergo (projected figures)
    2012***873751****825382
    2013***871785****833379
    2014***869819****841376
    2015***867853****849373
    2016***865887****857370
    2017***863921****865369

    The Unionist vote in the 2014 elections was up because the Unionist turnout was up. Normally it is quite low for Euros and local elections, because these are not perceived as important. But as I remarked to Comrade Stalin at the time, this did not lead to the loss of a single Nationalist seat, so is hardly relevant (He did think there might have been a local council seat which changed hands, but was unable to name it.)

    So both of us (PR & NW) are envisioning the future based on past performance. NW (and CS) are formulating a rule that any percentage loss of vote in the most recent elections will continue in the next elections.

    I have formulated a rule that Unionists lose a seat at every Westminster Election. For the last 8 Westminster elections, including one which was a group of bye-elections, this has held good.

    It should be noted that since the last General Election the Catholic population of North Belfast has risen above the Protestant population for the first time. In 2011 there were 48,126 Catholics in North Belfast and 46,821 Protestants. In 2015 these respective growths and declines will have gone further. So North Belfast is the constituency most likely to confirm Reilly’s rule, that Unionists lose a constituency at every election.

    It should also be noted that in South Belfast, which Alasdair McDonnell managed to win when Catholics and Nationalists were still in the minority, now (as of 2011) has a Catholic majority. McDonnell’s political obituary has been written so many times already: I don’t see it happening in 2015.

    Equally I would go along with the opinion expressed by the majority of Slugger contributors: that Naomi Long will not be unseated in 2015. Alliance is like Japanese bindweed: once it gets in it is almost impossible to eradicate.

    It all depends on turnout. While I cannot say with total certainty who will win North, South and East Belfast, I am on much firmer ground in asserting that turnout in these constituencies will be massive.

  • Zeno1

    Ah Paddy, if only all Catholics were Nationalists instead of about half of them.

  • Paddy Reilly

    And what are the other half?

  • Zeno1

    They are part of the 47% who say they are not Nationalist or Unionist, They are the sensible people who decided they didn’t want to be in either tribe.

  • Nicholas Whyte

    The Unionist vote in the 2014 elections was up because the Unionist turnout was up.

    The votes were up because more people voted???? Something of a tautology!

    this did not lead to the loss of a single Nationalist seat

    There’s certainly room for discussion here, given that we cannot make exact comparisons, but I think most obviously Moyola fits this description, while arguments can be made also for Glengormley, Dunsilly, Craigavon, Castle (Belfast), Ballymena and Bann.

    NW (and CS) are formulating a rule that any percentage loss of vote in the most recent elections will continue in the next elections.

    I don’t know about CS, but that simply is not my position. What I will say is that past election results are a better predictor of future election behaviour than the census statistics are. Even accepting (which I don’t) that the population change over the next few years will be a straight-line extrapolation of the census findings from 2001 and 2011, I don’t see much evidence that the census tells us any more about electoral behaviour than the actual record of votes cast can tell us – especially when you deliberately omit the growing numbers of people who do not wish to identify with either side. It looks rather like adapting the statistics to fit a conclusion, instead of the other way round.

    I have formulated a rule that Unionists lose a seat at every Westminster Election.

    I refer you to this perceptive XKCD piece on electoral precedents:

    http://xkcd.com/1122/.

  • mjh

    Yes I agree that a united unionist candidacy is problematic in South Belfast. I assume that that if there were one that the TUV would not break ranks. UKIP are a different matter. Their leadership may wish to run in every UK constituency.

    I’m not sure that the DUP would have too much trouble getting their voters out for a united unionist candidate that they supported. The historical evidence shows that DUP voters are highly committed to seeing another unionist elected if their own party is not available.

    In 2014 in NI as a whole, about 85% of DUP voters transferred to another unionist when their own party was either elected or eliminated. Only 5% transferred to a centre or nationalist candidate (if available), 10% did not transfer at all.

    Traditionally the UUP transfer to another unionist has been lower. For example in the 2011 council elections 13% did not transfer at all even when there was another unionist available . Even more worrying from the point of view of a united unionist candidate, if there was a centre party candidate available about 22% transferred to centre or even nationalist.

    The evidence from the 2014 election suggested that the average UUP transfer rate improved significantly (from a unionist viewpoint) but with worryingly high variations. And the only example we have from South Belfast was Lisnasharragh, where about 20% of UUP voters failed to transfer to either of the remaining DUP candidates, even though there were no other candidates still available.

    In order to win, a united unionist candidate would need about 80% of the unionist leaning vote. The DUP could deliver enough of its votes to a UUP candidate to make this possible. A DUP candidate would risk not only suffering UUP abstentions – but actually handing substantial former UUP voters to Alliance and even to the highly transfer friendly McDonnell.

    But I agree with you that it probably won’t happen and that, if you listen very carefully, behind all the high-minded talk of the importance of unionist unity you can just catch the clicking of calculators totting up party advantage.

  • Paddy Reilly

    But how do they vote?

  • Paddy Reilly

    especially when you deliberately omit the growing numbers of people who do not wish to identify with either side

    I seem to be getting this fallacy from more than one party. If there is to be a referendum about some simple issue, whether we drive on the left or right for example, then the opinions of those who don’t care or have no intention of voting are not relevant. Voting is by those who do care.

    Equally, in a first past the post election, the voting which matters is that for the two largest parties. If, in North Belfast, the SF vote exceeds the DUP vote, or the DUP the SF, it is that fact and that fact alone which will determine who is elected MP, and a surge or otherwise in the voting for Raymond McCord will not affect this.

    Equally in a provincial vote, up to a third of NI’s population could be Zoroastrian Natural Law Party supporters, but unless these can be persuaded to lend their support predominantly to one or the other of the native born contenders, then it is the relevant numbers of Nationalists and Republicans which will determine the issue.

    I don’t see much evidence that the census tells us any more about electoral behaviour than the actual record of votes cast can tell us

    Well look harder. In nearly half of NI’s constituencies there was once a Unionist MP, but he was eventually unseated by a Nationalist one. The actual record of votes would have told us that the seat was fated to remain Unionist: the census would have told us that the majority of voters were Catholic. It was the census that told us that a Nationalist could win the seat. Equally, the census returns tell us that North Belfast can now be won by a Nationalist, having a Catholic majority or maybe just plurality (Zoroastrians excluded) and the census tells us that South Belfast, also now having a Catholic majority, is not a prime candidate for a Unionist recovery.

  • Nicholas Whyte

    mjh – just on a technicality, those transferring UUP votes in Lisnsharragh were actually PUP votes which had subsequently transferred to the UUP. I know you’re not citing them as terribly strong evidence, but they are even less reliable as an indicator of UUP voter behaviour than you suggest.

  • sean treacy

    John , you say MOM would not attract ormeau votes .Surely the lower ormeau and the markets are SF’s stronghold in south Belfast?

  • Nicholas Whyte

    Paddy,

    All I’m saying is that your analysis is the poorer if you identify all census Catholics as Nationalists and all census Protestants as Unionists, and leave out the rest from your tally entirely. It leads you to a whole series of categorical errors, one of which I unpack below.

    It was the census that told us that a Nationalist could win the seat.

    Well, no; the census is (of course!) supporting evidence, but I see no evidence that it is determinative. In most of your eight cases, there was a majority of votes for Nationalist parties, which is a pretty good indication that Nationalists could win, before they coalesced behind a single candidate.

    Going through your eight, that was the case for Newry and Armagh, South Down, West Tyrone, and Fermanagh and South Tyrone. I will agree that Mid Ulster had a Nationalist MP from its creation, but I would submit that the likely voting patterns were as easy to discern from previous election results as from the census findings. South Belfast, on the other hand, elected a Nationalist MP in 2005, before the 2011 census found that there were more Catholics than Protestants in the seat, which is surely not in line with your argument. And East Belfast, which I guess is the last of your eight, has not actually elected a Nationalist MP as yet!

  • MainlandUlsterman

    the old Catholic = Nationalist confusion again. Sectarian headcount politics, though I’m not saying the community numbers are irrelevant either.

    Wasn’t 1916 supposed to be the year by which a united Ireland would be delivered? Good luck with that … one from the George Osborne school of targets I think.

  • John Gorman

    This is true and in 2010 when McDonnell polled 41% the turnout in these areas was very poor (no SF candidate) so as mentioned earlier people are exaggerating the possible impact of MoM on the result.

  • Paddy Reilly

    your analysis is the poorer if you identify all census Catholics as Nationalists and all census Protestants as Unionists

    Obviously I don’t. It was C. Cruise O’Brien who said that the Catholic Unionist Party and Protestant Nationalist could hold a joint conference in a Telephone Box.

    One has to bear in mind that there are sects such as Jehovah’s Witnesses and Reformed Presbyterians, who abstain from voting, as do some of the Brethren, and that Catholic = Nationalist only works with those born in Ireland of Irish parentage: those who come from Surrey might be Tories: those from Riga might not have a vote.

    But of course, if there are Catholic Unionists, there are probably Protestant Nationalists, so they would cancel each other out. If there are Catholics who wanted nothing to do with (perceived) sectarian politics, there are Protestants in the same boat: their vote can be discounted. The Natural Law and Rainbow George Party vote are irrelevant.

    West Belfast (75,263 Cath, 15,645 Prot) is not going to return a Unionist, and Strangford (15,447 Cath, 65,353 Prot) will not elect a Sinn Fein one. Of this I am certain.

    Where the figures are closer, there is less certainty, but the default position is still that a Catholic majority will return a Nationalist MP. In North Belfast, where the SF vote is creeping up in much the same fashion as the Catholic percentage, I would feel that Gerry Kelly’s boast that Nigel Dodds is just keeping the seat warm for him might not be such an idle one.

  • Nicholas Whyte

    If there are Catholics who wanted nothing to do with (perceived) sectarian politics, there are Protestants in the same boat: their vote can be discounted.

    OK, that’s a decent summary of your position; thank you. I personally prefer to look at all voters, whether or not they fit into traditional boxes, but how you do it is your affair.

  • Paddy Reilly

    Well, yes: in the immediate question of who will be MP for North Belfast, only those who have a political stance which is strongly for (DUP) or against (SF) partition are relevant, and equally in the not so immediate but nevertheless not far off question of whether of not Ireland should be united, only those who have a YES or NO opinion on the matter count.

    But in the context of a United Ireland, this kind of politics is deprived of its relevance. I am reminded of the poor members of the Anti-Apartheid Movement, who gave their life and soul to the struggle against Apartheid. Then some fool abolished Apartheid, and they had nothing left to live for.

    I, however, am not worried: after a United Ireland I can revert to being a Green, which is what I am in other environments.

  • Colin Lamont

    The more I see the evidence the more vulnerable McDonnell appears to me. think a DUP win is possible here Nicholas? Might have a small wager on that as an upset

  • Colin Lamont

    Dodds will win handily. Unionist turnout was much improved in N Belfast vis a vis nationalist last year. That won’t be reversed within a year.

  • Paddy Reilly

    Unionist turnout was much improved in N Belfast vis a vis nationalist last year.

    Funny I made this point but Nicholas Whyte did not seem to understand it. Increasing the turnout is a handy trick when your support base falls below 50%, but it can only be effective for as long as it takes for the other side to get wind of what you are up to. It certainly doesn’t take more than a year. Hence my prediction that turnout in North, South and East Belfast will be stratospheric this year.

    Nigel Dodds, if he wins at all, will not get 50% of the vote. But the last Westminster was in 2010: an election in 2015 ought to show some change: we do not live in total stasis, and N.Belfast is the seat closest to changing hands.

  • Nicholas Whyte

    Colin,

    If SF choose not to stand, then I suspect that McDonnell is safe. If they do stand, however, then I agree that the DUP are in with a racing chance of unseating him, more so than any other party (my mind is still a bit boggled by Slugger contributors deeming Máirtín Ó Muilleoir as his most likely replacement – third place for SF would be a good result, they were fifth and the DUP first in both 2011 elections!)

  • Nicholas Whyte

    Nicholas Whyte did not seem to understand it

    Well, I don’t think you expressed it very clearly. Perhaps it was similar to your suggestion that increased Unionist votes “did not lead to the loss of a single Nationalist seat”, to which I gave several counter-examples; or your imaginative assertion that Nationalists needed census findings as well as election returns to see that they could win in East Belfast, South Belfast, Newry and Armagh, South Down, West Tyrone, and Fermanagh and South Tyrone, all of which are flatly contradicted by the historical record; or indeed your statement that the votes of those who want nothing to do with sectarian politics can simply be “discounted”. This is muddled thinking; you can surely do better.

    On North Belfast, of course Nigel Dodds has a difficult defence; you and I actually agree on that. The political question is whether he can squeeze and motivate Unionist (and moderate) voters more effectively than SF can squeeze and motivate Nationalist (and moderate) voters. The census returns are of limited assistance in determining the answer!

  • John Graham

    While Alliance have taken much of the McGimpsey-UUP vote in South Belfast the last few elections have seen a substantial growth in the Roman Catholic support base for them and Anna Lo. Paula might get the Alliance vote in Balmoral – which is more Unionist than Botanic/Laganbank, but not so sure how she will do in Botanic for example given that it is more Roman Catholic. She won’t be pulling out support from Unionist areas in their droves from the Village, Donegall Pass, Sandy Row or Belvoir either. It’s a nice Churchy Prod vote with a nice Castle-Catholic Terrance O’Neill inspired type vote (Radios, houses and a job!) she has! She could do the Unionist candidate/s some damage while a lot of Roman Catholic Alliance voters opt for Alasdair because this is sectarian politics.

  • Paddy Reilly

    A seat in my parlance is at Westminster or Stormont. A defective turnout of Fenians on Slieve Gallion Mountain (by how much? 200 votes?) is a single swallow which definitely does not betoken a summer.

    Local elections are imperfectly sectarianised: there are loads of Independents, which is as it should be: they are not a reliable indicator of provincial politics.

    I said that census returns could be an indicator of who would win a seat, often better than the past political history: I said nothing about politicians anxiously consulting the census before standing. Such information may have influenced them, when it was available, but politicians have a strong vein of foolish optimism, which carries them forward regardless, which is why we have so many elections: in bygone days many MPs were returned unopposed.

    In a vote which concentrates on the issue A or not A, the votes of those who refuse to be drawn by the issue, the don’t knows or don’t cares, can be discounted, in matters which pertain to this issue. However, when some other issue is at stake, they may be crucial. ‘A Plague on both your houses’ may be an honourable stand to take, but in Northern Ireland it tends to disqualify you from participation in what is perceived as the most important issue.

    Nigel Dodds already has nearly the whole of the Unionist vote (less than 3,000 UUP voters in North Belfast.) The question is whether, by sending his Orange goons to harass N. Belfast’s Catholics, he has upset enough SDLP voters to hold their nose and vote for Gerry Kelly in the hope of being rid of them. But the census returns suggest that Kelly may still get in if he hasn’t.

  • Nicholas Whyte

    imperfectly sectarianised

    Says it all, really.

  • mjh

    Yes it is possible that the high non-transfer to DUP is disproportionately affected by the earlier PUP transfers to the UUP. This does not effect the strength of the risk that the Lisnasharragh result indicates for a united unionist candidate associated with the DUP, but it introduces another possible explanation for it.

    Strictly speaking the figure of about 20% which declined to transfer to the DUP is the percentage of the 1362 first preference votes for the UUP plus the transfers in from other unionist leaning candidates (300 from PUP and 2 from DUP) who did not transfer to the DUP.

    I should have considered whether this section of the PUP vote was especially unwilling to transfer to DUP. If it is PUP voters not UUP it would mean that up to around 40% of the original PUP voters had refused to transfer to DUP.
    Unfortunately there is little evidence to go on.

    If it were so the question would arise whether the phenomenon extends to TUV voters.

    The implications would be greater for East Belfast than South.

    But that’s another hare, which I have not time yet to persue.

  • Nicholas Whyte

    Strictly speaking the figure of about 20% which declined to transfer to the DUP is the percentage of the 1362 first preference votes for the UUP plus the transfers in from other unionist leaning candidates (300 from PUP and 2 from DUP) who did not transfer to the DUP.

    With respect, mjh, that’s not the case. The bundle of votes transferred was not the 1362 first prefs for the UUP, but only the 300 received from the PUP on Stage 7. The 1362 first prefs and all other subsequent transfers bar the last remained in McGimspey’s pile. Those 300 PUP->UUP votes went 124 to Sandford and 81 to Hussey, the remaining 95 having no preference for either DUP candidate. But they were the only ones in play at that stage of the count.

    it would mean that up to around 40% of the original PUP voters had refused to transfer to DUP

    I make it more like 30% – of 684.72 PUP votes on elimination, 270.72 went directly to the DUP and 205 indirectly via the UUP, which leaves 209 – 30.5% of their final tally, 33.2% of their first prefs. But either way, your point is sound: there is a substantial chunk of PUP vote that simply won’t transfer to the DUP.

    the question would arise whether the phenomenon extends to TUV voters

    Good question!

  • mjh

    Nicholas
    Thanks. Penny has dropped.