Below is part one of Faha’s much anticipated analysis for the forthcoming Assembly elections – BD

Ballot Box

This is the first of my two part series on the constituencies for the March 2nd Assembly election. In this first series I will cover the 9 constituencies outside of the Belfast area. Here is some of the background I will go over prior to analyzing each constituency.

This chart shows the differential turnout between the Catholic and Protestant community background in the 2016 Assembly election. It is based on the 2016 constituency census data (extrapolated from the 2011 census) and the transfer pattern from nonsectarian parties. As you can see, with a few exceptions, most constituencies had a much lower turnout from the nationalist electorate compared to the unionist electorate. Overall nationalist turnout was 7% less than unionist turnout. If it had been equal to unionist turnout another 45,000 voters would have voted

2016 Assembly Election


This chart shows the difference in turnout for each constituency comparing the May 2016 Assembly election with the June Brexit referendum. In every constituency (except West Belfast) there was an increase in voter turnout for the Brexit referendum.

It is estimated at 110,000 more voters (the number in the chart is 96,209 but since EU nationals could not vote in the Brexit vote the 110,000 is the actual change) voted in the Brexit referendum. From the results it appears that most of those additional voters were nationalist, Alliance, Green and pro EU UUP voters. The pre-election  Lucid Talk poll, which predicted the final result to within 1%, indicated that 85% of nationalist, Alliance and Green voters were against Brexit and 85% of unionist voters were pro Brexit.

These Brexit results indicate that if an Assembly election had been held the same day then the   unionist parties would have received 44% of the vote and nationalist, Alliance and Green parties would have received 56% of the vote.

2016 Brexit Referendum

Faha 2017 Brexit.JPG

This chart shows the Lucid Talk polls from April 2016 and January 2017. I compared the April 2016 poll with the January 2017 poll and calculated the difference in the 3rd column. The last column indicates the difference between the January 2017 poll and the actual Assembly results from 2016. The Lucid Talk poll was very accurate in predicting the overall total vote for nationalist parties, unionist parties and nonsectarian parties. It underestimated the vote for the DUP and overestimated the vote for the UUP, Alliance and smaller unionist parties. Part of the underestimate for the DUP and overestimate for the smaller unionist parties is due to the fact that the smaller unionist parties did not compete in all constituencies. Thus, some of their voters choose the DUP instead by default. Keep in mind that the poll has a +/- accuracy of 3% so it compares very favourably with the actual results.

The January 2017 poll shows little change from the April 2016 poll. The only party change that may be statistically significant is the increase for PBP and possibly the decline for the UUP. All the other changes are too small to determine if they are actually real. However, the overall 3.76% decline for unionist parties is significant. Statistically, the real decline could be anywhere from only 0.76% to as high as 6.76%.

The January 2017 poll did note a few trends. Almost half mentioned the Health Service and RHI as important issues followed by Education, Equality and the Economy. Almost ¼ mentioned Brexit as an important issue followed by Irish Language and Irish Unity. It was also noted that a “large” number of DUP voters in 2016 said they would vote for other parties (primarily UUP) in this election. However, the actual poll results do not seem to indicate this but it could be due to DUP oversampling in this poll. Responses to questions on transfers showed that SDLP and UUP voters are now more likely to give 2nd preferences to each other. The most significant change is that UUP voters are more likely to give their 2nd preference to Alliance and the SDLP rather than the DUP. This is how UUP voters indicated their 2nd preference:

SDLP                      25.3%

Alliance                 23.7%

DUP                          5.8%

TUV                        19.0%

Other Unionist     11.6%

SF                              1.6%

PBP                           2.6%

Green                       3.2%

That trend could have a profound impact on the election prospects for DUP candidates. Unionist voters do not appear to have noticed any increased interest in voting in this election but nationalist voters have noticed increased interest among family, friend and coworkers in voting in 2017. This sounds plausible since 2016 unionist nonvoters have no reason to vote in this election. Brexit was passed in Westminster this week and the RHI scandal would certainly not increase unionist turnout for the DUP. Nationalist and nonsectarian voters appear to be more likely to vote this year with impending Brexit and its impact on Northern Ireland a big issue.

There will be additional Lucid Talk polls on February 13 and February 27th will provide more insight on the state of the electorate.


The following are the constituency profiles.

I have included the demographic data, 2016 Assembly results, the new quota with 5 seats and the Brexit vote.


All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
79,679 59,941 17,644 641 1,453
  75.23% 22.14% 0.80% 1.82%


Total                39,702

Quota                6,617

EU Referendum Turnout   40,987 (+1,285)

Remain        78%

Leave           22%


Both the SDLP and SF should elect 2 on transfers since they are close to 2 quotas (33.3%). In 2016 Dr. McCloskey had 4,227 votes at the last count, almost 1,200 votes behind Eamonn McCann. Gary Middleton of the DUP was elected with 6,641 votes and had a surplus. However with the higher quota in 2017 he would barely be above the quota. He is likely to receive fewer transfers this time but should be close to a quota. There are only 2 scenarios in which Eamonn McCann could be elected. There could be a decline in unionist turnout or an increase in nationalist turnout. If nationalist turnout were to increase by 3,000 votes the quota would be 500 higher and the DUP would be 500 votes short of a quota and Eamonn McCann would be elected. There were 1,285 more voters who did vote in the EU referendum so this scenario is possible.


                                              EAST  DERRY

All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
79,650 33,062 43,080 566 2,942
  41.51% 54.09% 0.71% 3.69%


New Quota          5,733

EU Referendum Turnout   40,563 (+6,164)

Remain        52%

Leave           48%

The DUP should elect 2 here even with a decline in their 1st preference. There should also be 1 SDLP and 1 SF elected. If you look at the stage 9 count last year this is where the SDLP and SF vote was.

SDLP                             4,172

SF (Archibald)             4,139

SF (O’Hoisin)               3,532

Since the SDLP was ahead of both SF candidates the SDLP candidate will be elected. Since the total nationalist vote equaled 11,843 and 2 quotas is 11,466 both SF and the SDLP should elect one. The only complication factor this time is that there is essentially a 2nd SDLP candidate as the current SDLP MLA, Gerard Mullan, has decided to stand as an independent. John Dallat’s 1st preference vote was 2,000 votes higher in 2011 compared to Gerard Mullan in 2016 so the latter will likely be eliminated. The 5th seat will come down to the UUP and Claire Sugden. Claire Sugden seemed unwilling to stand up to the DUP during the recent RHI debate at Stormont so this could hurt her chances for receiving transfers from Alliance or nationalist voters. If those additional 6,164 Brexit voters turn out there could be significant surplus of nationalist voters who will determine whether the UUP or Claire Sugden wins the 5th seat.

The following 3 constituencies I group together because they all have one demographic in common. The Protestant voting age population is in the 31% or lower range which is less than 2 quotas (33.33%).



                                           WEST TYRONE

All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
71,747 48,604 21,926 286 931
  67.74% 30.56% 0.40% 1.30%


New Quota        6,468

EU Referendum Turnout   40,039 (+1,252)

Remain        67%                                              

Leave           33%

I included a former SF member candidate in the SF total and 2 former SDLP councilor candidates in the SDLP total since these 3 candidates left their parties just prior to the election. The SDLP should elect one here since even in the 2015 Westminster election their vote was 16.67%. SF would need to be very close to 50% to elect 3. They are short by almost 6% but there are some possible transfers from the 5.76% nonsectarian voters (Alliance, Green, others). The main problem for SF here is the low nationalist turnout which was 52% in 2016 versus 65% for the unionist electorate. Since this is a border constituency Brexit will be a big issue here. The prospect of a hard border with Donegal could increase nationalist turnout which could potentially elect a 3rd SF MLA. If the nationalist turnout reaches 60% there will be 3 SF elected otherwise there will be 1 UUP and 1 DUP elected with only 2 SF.

                                                    MID ULSTER

All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
76,858 51,112 24,033 358 1,355
  66.50% 31.27% 0.47% 1.76%



New Quota        6,791

EU Referendum Turnout   42,411 (+1,683)

Remain        60%                                              

Leave           40%

The SDLP should elect one here as they are just short of a quota but should receive enough Alliance and Green transfer to reach a quota. SF would be 2.5% short of the 50% they need to elect 3. Low nationalist turnout (50% versus 60% unionist) is all that is preventing SF from electing 3. In the 2011 Assembly election SF had 49.2% and the SDLP 14.7% and there was 3.3% for independent nationalist and 1% for Alliance. With the 2011 results SF would have elected 3 and the SDLP one with the DUP winning the 5th seat. Another 2,000 nationalist voters in 2017 is all that is needed for 3 SF. The presence of Michelle O’Neill as a candidate and new SF leader could be enough to bring out those additional 2,000 nationalist voters.

                                        NEWRY AND ARMAGH

All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
88,081 58,598 26,966 522 1,995
  66.53% 30.62% 0.59% 2.26%





New Quota        7,953

EU Referendum Turnout   50,622 (+2,929)

Remain        63%                                              

Leave           37%

The SDLP should elect one here since they are above a quota. SF are further away from electing 3 here as they would need to increase their vote from the current 41% to over 45% to be close enough to elect 3 on transfers. Again, the problem for SF here is low nationalist turnout which is 11% less than unionist turnout. If nationalist turnout approaches 60% then SF will elect 3 here. If that occurs the DUP and UUP will be in a close battle for the final seat. Brexit should be a big issue here since obviously a hard border will affect this constituency in a significant way.

                                   FERMANAGH SOUTH TYRONE


All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
81,074 46,746 31,962 532 1,834
  57.66% 39.42% 0.66% 2.26%




New Quota        7,863

EU Referendum Turnout   48,158 (+1,010)

Remain        59%                                              

Leave           41%

This constituency is one of the most difficult to predict for obvious reasons. Based on the 2016 vote there would be 2 DUP and 1 UUP elected with 2 SF. The 2016 vote showed that after transfers the nationalist electorate was 51% and the unionist electorate 49%. SF could not elect 3 since they would need all of the SDLP vote as well as those Alliance, Green and Other transfers that go to the SDLP to transfer through to SF. Starting with a base vote of 39.95% they are too far away from 50% to elect 3. The DUP were just short of 2 new quotas and the UUP would elect one on transfers. The nationalist turnout was a full 20% less than the unionist turnout. However, it is now 2017 and the RHI scandal is now major news with Arlene Foster of the DUP the main candidate here. How this will affect the unionist vote and overall vote is unknown. Brexit and the prospect of a hard border will also be an important issue here. The UUP MP here, Tom Elliot, voted in favour of Brexit in Westminster this past week.  One thing is certain over the years is that unionist voters are very reliable voters and will vote in every election so there will be 22,000 to 23,000 unionist votes. The question is how those votes will be distributed among the DUP and UUP. The Lucid Talk poll is indicating a shift from the DUP to the UUP. In the 2011 Assembly election the UUP received more than 3,000 1st preference votes compared to 2016. If 2,000 of those return to the UUP then the UUP candidate will be elected with Alliance and Green transfers. If it is 2,500 then even the transfers will not be needed. Whether the 2nd DUP will be elected will depend on balancing between Foster and Morrow and nationalist turnout. It will be very difficult for the DUP to balance their candidates here since they do not know how the RHI scandal will affect Arlene Foster’s 1st preference vote. Obviously they need to push the 1st preference vote in Arlene Fosters’ direction since her defeat would be a major defeat for the DUP. For SF there is a sweet spot for increasing nationalist turnout. If it increases to 52% to 54% then they will elect 3 on transfers. However, if it increases to 55% or more then this scenario favours the SDLP. That is because the 3rd unionist candidate would then be polling 10% or less and would be eliminated. Enough of their transfers would go to the SDLP to elect the SDLP candidate. If nationalist turnout is in the 52% to 53% range then the SDLP candidate would probably poll less than the 3rd unionist candidate and would be eliminated instead.

                                              UPPER BANN


All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
92,513 40,880 46,748 809 4,076
  44.19% 50.53% 0.87% 4.41%



New Quota        7,614

EU Referendum Turnout   51,812 (+6,159)

Remain        47%                                              

Leave           53%

The combined SF-SDLP vote in the final count was 17,236 which is 2,000 votes over 2 quotas so there will be 2 nationalist MLA’s elected. O’Dowd of SF finished 167 votes ahead of Kelly of the SDLP but there were 1,260 Alliance and UUP voters who did not transfer to anyone. There will be 3 unionists elected but in 2016 the DUP were 2% short of 2 new quotas on their 1st preference vote. This has traditionally been a strong constituency for the UUP and any significant decline in the DUP vote could result in 2 UUP MLA’s elected. However, the Lucid Talk poll is not indicating such a dramatic shift. There were over 6,000 additional voters who voted in the Brexit referendum and it appears that nationalist turnout may have been close to unionist turnout. If those nationalist voters turn out for the Assembly election then 2 UUP could be elected with nationalist transfers.

                                                 SOUTH DOWN

All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
84,986 58,906 23,152 477 2,451
  69.31% 27.24% 0.56% 2.88%


New Quota        6,846

EU Referendum Turnout   47,701 (+6,649)

Remain        67%                                              

Leave           33%

At the Stage 4 count in 2016 these were the party vote totals:

SF                   13,321

SDLP               14,957

DUP                  6,470

UUP                  5,567

The SDLP are 1,267 votes above 2 quotas and SF are 371 votes short of 2 quotas so 2 SF should be elected with SDLP transfers. Subsequent stage counts in 2016 only involved DUP and SF transfers which will not occur in 2017 due to the higher quota. The DUP is at risk of losing out to the UUP for the 1 unionist seat if their vote declines in this election. There were also an additional 6,649 voters who voted in the Brexit referendum and if significant number of these vote then the UUP should win on nationalist transfers.


                                            NORTH ANTRIM

All usual residents Catholic Protestant and Other Christian (including Christian related) Other religions None
86,546 24,618 57,871 644 3,413
  28.44% 66.87% 0.74% 3.94%


New Quota        6,833

EU Referendum Turnout   49,720 (+8,742)

Remain        38%                                              

Leave           62%

This constituency is straightforward. The TUV will elect one since they are above a quota. SF will elect one on SDLP transfers. The UUP 1st preference vote is almost 6% less than a quota but with unionist transfers (including TUV) as well as Alliance, Green and SDLP transfers the UUP candidate should be elected. Only 2 DUP will be elected. However, if the Brexit electorate was the electorate for this election then there would be 2 nationalist candidates elected, both SF and SDLP. In 2016 the nationalist turnout was only 36% and the unionist turnout 52%. In the Brexit referendum they appear to have been equal.

Obviously many of the 5th seats will be determined by which electorate votes in the election: the 2016 Assembly electorate or the 2016 Brexit referendum electorate.