NI GE19 (7) Predictions

Prediction is the surest way of making an idiot of oneself, but it is fun nevertheless. The predictions will be made impartially but from a Remain viewpoint. Thus, the best-case scenario is for a Remain candidate to win and the worst-case, a Leave candidate. Alliance and SDLP candidates count as a point each, SF 0.5 points, UUP 0.3 points and the DUP 0 (nul points), as outlined in post 6 of the NI GE19 series.

Recent data from LucidTalk shows NI-Wide voting intentions at the end of November and are shown in Fig. 1.

Fig.1 LucidTalk Poll.

It is interesting to compare this to the actual votes cast in the 2017 GE (Fig. 2).

Fig. 2 Votes cast in the 2017 General Election. ref

The DUP and SF are down 8 and 5.4 points down respectively on GE2017 and the SDLP and UUP largely unchanged. The main beneficiaries are Alliance up 7 points. Annoyingly, the 10% of ‘Other Parties or Undecided’ has not been split. Assuming the smaller parties get a similar share of the vote as in 2017 (c. 5%) there remains still about 5% to redistribute to the main parties so the DUP and SF vote share drop may not be quite as great.

More granular detail is needed however to help predict actual constituency results, but the LucidTalk MRP constituency level data has yet to be released.

There is good analysis on Slugger with political insight and some detailed number-crunching by Faha on the BangorDub site.

This post will look at demographics and make predictions, drawing from a number of sources. The eight most likely seats to change hands will be examined. Belfast will be looked at in detail as it is a useful microcosm for NI as a whole. Conclusions will follow.

Demographics and Differential Voting Rates

As documented on this blog site, NI is undergoing profound demographic change. If cultural Catholics are not already in a majority, they will be within the next few years. They tend to vote for either nationalist or “Unaligned” parties. Cultural protestants, particularly older ones, tend to vote Unionist. NI Unionist voters are in a majority in the 55+ age range, whereas younger voters lean far more Nationalist or “Unaligned”, than Unionist.

I have not seen NI specific data, but the Warwick Study shows (Fig. 3) that GBR has the highest mismatch of voter turnout between older (55+) and younger voters (18-35). The older group is 37.5% more likely to vote than the younger group. If NI is more like IRL there is still a 17.5% disparity. Either way, this voting pattern favours Unionists and it may take some years for a Nationalist plurality (in the US sense of the largest group but not the majority) to translate into votes.

Fig. 3 Difference in voting rates between those 55+ years old and those 18-35 years old (percentage points)

Seats that May Change Hands

First past the post (FPTP) elections distort things, and in GE17 eleven Unionist MPs were elected (10 DUP, 1 Ind. U), seven Nationalist MPs (all SF) and no “Unaligned” MPs.

The FPTP system has worked in the Unionists favour, particularly in Belfast where Unions have 75% of the seats, with only 39% of the vote. Belfast is a fascinating case and will be looked at separately.

In reality, there are only eight seats with any real prospect of changing hands. These are ranked in decreasing likelihood of change. These likelihoods are educated guesses not calculated probabilities.

SeatLikelihood of ChangeChange
North Down p(1) CertaintyIndU->DUP or Alliance
South Belfastp(0.99) Virtual CertainlyDUP -> SDLP
Foylep(0.75) likelySF -> SDLP
North Belfastp(0.55) leaningDUP->SF
East Belfastp(0.5) evensDUP->Alliance
South Antrimp(0.5) evensDUP->UUP
South Downp(0.3) unlikelySF->SDLP
Fermanagh S. Tyronep(0.25) v. unlikelySF->UUP

Non-Belfast Seats

North Down

Given that popular independent Unionist MP Lady Sylvia Hermon is retiring, the seat will change hands. The constituency was 52.4% Remain in the referendum. The DUP are in pole position. However, it is easily the most affluent constituency in NI and not classic DUP territory. Both the UUP and Alliance are in the running. The Lucid Talk MRP poll is rumoured to make this the most likely Alliance gain. Faha leans Alliance while Slugger calls it for the DUP.

Pessimistic: DUP. Optimistic: Alliance.

Foyle

The seat was profiled in detail in post two of this NI GE19 series. This is a straight fight between Sinn Féin and the SDLP, currently a SF seat since 2017, but that was a shock victory for many. It is the most pro-EU seat in NI (78.3% remain in the referendum). The Lucid Talk MRP is rumoured to be too close to call. The SDLP candidate is the leader Colm Eastwood and easily the bookies favourite. The LucidTalk MRP is too close to call, but Paddy Power has it at as an SDLP gain (SDLP 2/5, SF 11/8).

Pessimistic: Sinn Féin. Optimistic SDLP.

South Antrim

South Antrim is a very Unionist constituency and is a straight DUP/UUP fight, currently held by the DUP. The constituency voted very close to 50:50 in the Brexit ref. (50.6% ‘Leave’). This is probably the best chance for the UUP to win a seat. Given the UUP’s new Remain credentials, Alliance voters may tactically vote UUP, which might push it over the line. Faha leans UUP whereas Slugger leans DUP.

Pessimistic DUP. Optimistic UUP.

South Down

This is a very Nationalist constituency and a straight SF/SDLP fight, currently held by SF. It is less close than Foyle but there is a chance of the SDLP regaining this seat. Faha puts the chances at about 50:50 but most analysists think SF will hold on to the seat.

Pessimistic SF. Optimistic SDLP

Fermanagh and South Tyrone

This constituency was examined in detail in post 5 of the NI GE19 series and is a UUP/SF fight. The constituency voted 58.6% Remain and the UUP candidate, Tom Elliott, is a committed Leaver. This may not go down well as Brexit almost certainly will have a devastating effect on this largely agricultural constituency.

Pessimistic UUP. Optimistic SF.

Belfast

Belfast is interesting as it is like the entirety of NI in microcosm, but with demographics that will resemble the entirety of NI within less than a decade.

There are four seats in Belfast, easily remembered as they correspond to the four cardinal points on the compass North, South, East, and West. They are very different in character. Three of the four seats NB, EB and SB are being hotly contested and have been covered in detail in this NI Election series. The fourth, W. Belfast is the safest seat in NI and one of the safest in the UK. Unless there is a major political earthquake it will be a Sinn Féin hold.

Remainers have to some extent made electoral pacts, detailed throughout the series, and there is a good article in the Independent The Battle of Belfast with some more detail.

Table 1 shows the votes of parties that achieved over 1000 votes Belfast-wide in the 2017 GE.

SFSDLPPBPGreenAllianceUUPDUP
N. Belfast19159205806442475021240
E. Belfast894167056115443140823917
S. Belfast714311303022417946152413299
W. Belfast2710728604132073105455
Belfast54303163884132344626595293263911
Table 2 2017 GE Election Results for main Belfast parties.

Both Sinn Féin (SF) and the SDLP identify as Nationalist. People Before Profit (PBP), the Greens and Alliance identify as Other/Unaligned and the UUP and DUP identify as Unionist. Fig. 4 shows these groupings as a percentage in a pie chart.

Fig. 4 Total GE17 Belfast votes by Constitutional grouping

The community sizes are 41:39:20 with the nationalists being the plurality (just). Proportional voting systems would give different results, but the d’Hondt system, for example, would allocate two nationalist seats, one Unionist and one ‘Others’.

The actual results in GE17 were that the Unionists won three out of four seats (75%) with only 39% of the vote. This was a remarkable performance, which is unlikely to be repeated. This is largely because many of the Nationalist voters are located in one constituency West Belfast, where they have an overwhelming majority. In North and South Belfast the numbers are fairly evenly split. In East Belfast, there are few nationalists but many Alliance voters.

Fig 5 Probabilistic outcomes chart from @326Pols Dark Green SF, Light Green SDLP, Yellow Alliance, Red DUP.

Fig. 5 shows the four Belfast constituencies with probabilities of winning. This ‘326 Politics‘ analysis if repeated would give the DUP no seats in Belfast.

Belfast North

This is a very tight fight between Sinn Féin and the DUP. It is a very close call. In addition to the analysis on this blog (No 1 on the NI GE19 series), Patrick Maguire’s article in the New Statesman is well worth reading. New voters, however, give SF an advantage. Detailed analysis by Faha at ward level:

4,000 new voters have been added since 2017 with 2,800 in November. It appears the increased registration is concentrated in certain wards and it is likely that less than 40% of those voters are unionist voters.

Turnout will be key. It is a particularly vicious campaign and the DUP seem to be reverting to appealing to the Loyalist/UDA vote rather than any cross-community outreach. This may totally alienate Alliance voters.

Likely SF Gain

Pessimistic DUP, Optimistic Sinn Féin

Belfast East

Belfast East is a DUP/ Alliance fight. This is covered in detail in the 4th post in the NI GE 19 blog series. The DUP were well ahead in 2017 and there is a lot of ground to make up. However, the Alliance surge seems to be holding. The Lucid Talk polling makes the seat too close to call as indeed does Faha’s analysis.

Pessimistic DUP. Optimistic Alliance.

Belfast South

This seat has been discussed in detail in post 3 of the NI GE19 series and will be a safe SDLP win. There is almost no likelihood of the DUP candidate winning and indeed may be in 3rd place behind Alliance.

Belfast Map Post Election?

If it is a bad night for Leave and the DUP the map could be very different. Belfast could have four Remain seats which would be a stunning result.

Fig. 6 Best case scenario in Belfast. Left 2017 Right 2019. Red DUP, yellow Alliance, Light green SF, dark green SDLP. (Graph by the author).

Conclusion

Many of the seats are too close to call, but in the most optimistic scenario, the overall result will be DUP 6. SF 6. SDLP 3. Alliance 2. UUP 1. Result: A Remain score of 8.3.

How: The DUP loses all three Belfast seats and S. Antrim. SF gains North Belfast but loses Foyle and S. Down. The SDLP gains South Belfast, Foyle and S. Down. Alliance gains East Belfast and N. Down. The UUP gains S. Antrim.

In a more pessimistic scenario, there are only two seat changes. The SDLP gain South Belfast and the DUP gain N. Down. This would leave the parties at:

DUP 10. SF 7. SDLP 1. Result: A Remain Score of 4.5.

The likelihood will be somewhere between the two extreme scenarios, but my hope is that it is closer to the former than the latter.

Irrespective of Brexit, I asked today if the DUP had any redeeming features, but nobody could think of one apart from the fact they had made a United Ireland far more likely. Rather than standing on their record they have as in Fig. 7 resort to dog-whistle politics. Then again, DUP like politics seems to have come to England.

Fig. 7 DUP Belfast election material

One other bit of news today: The latest (2019) edition of the UN Human Development Index has been released. This is one the best indicators of how well a country is doing on a broad range of indicators, though our preferred measure on this blog is the IHDI, which adjusts for inequality. Not much change in the IE and UK positions, but IE has gone up one point to 3rd (behind Norway and Switzerland) and the UK has dropped one point to 15th. Maybe Brexiters will take consolation from the fact that the UK’s position is now tied with the US? Perhaps NI Unionists should be less worried by an economic United Ireland?

Comments

  1. davy green -

    Sean-It is widely reported that the majority of new voters in North Belfast are from Unionist areas-several seasoned sources reporting this.There have been many vote registering meetings in North Belfast.This would sway me towards a DUP hold.
    On a different note Houghton and Sunderland South is usually the first seat to be declared-Usually a rock solid Labour seat-Be interesting to see if the blonde maestro has much of an effect here.If the Tories win here Johnson will get his majority.We should know round 11pm if not before

    1. Sean Danaher -

      Davy thanks. I have heard this also and it is possible Faha is wrong. He is normally highly accurate though. Both the DUP and SF have formidable and equally matched electoral machines and I suspect the Nationalists have also had a big registration push.

      I note Paddy Power had SF ahead by a whisker yesterday but now have the DUP marginally ahead.

      Agree Sunderland is a place to watch. I used to always get a lift when Chris Mullin was first elected. There is a bit of a NE derby with Newcastle Central hoping to pip them to the post. Probably less interesting as Newcastle was a ‘Remain Island’ un a sea of Leave areas.

    2. Samuel Johnson -

      So, a tipping point of sorts, at last. Unionist seats now outnumbered by the combined total for nationalists and others. This trend will continue incrementally with each election.

      The DUP are, no doubt, regretting that they didn’t back Theresa May’s deal while they could.

      A bad night for the union. A watershed perhaps; potentially, the start of a new direction for both Scotland and NI.

      1. Sean Danaher -

        Well, all the seats with a probability above 0.5 were called correctly. Happy with that. main disappointment E. Belfast, the 15.5% swing to Alliance was too little.

  2. Richard Bond -

    For someone who previously knew little about NI politics, this has been a brilliantly informative series. Here’s hoping for a best case scenario. Thanks Sean.

    Given that NI voted Remain and the DUP are Leave, but have been shafted by Johnson, I have hope.

    1. Sean Danaher -

      Thanks, Richard. It is extraordinary how loyal the Unionist voter base is to the DUP. Given the age profile, with Unionist voters being older and sometimes much older I hope a December Election may work against them.

Write a reply or comment Comments Policy

Your email address will not be published. Required fields are marked *

Name *