This is the second part of my series on the constituency profiles for the March Assembly election. I will cover the 9 Belfast area constituencies.
In the previous week there has been one major political development. That is the public statement by the UUP leader Michael Nesbitt that he would give his 2nd preference to the 
SDLP after voting for the UUP 1st preference. He also stated that UUP voters should give their lower preferences to whichever candidates would be best for their constituency. His statements attracted much attention, comments and criticism. The criticism even came from within his own party. These are the actual 2nd preferences for voters as indicated in the recent Lucid Talk poll.
Michael Nesbitt is essentially following the preferences of UUP voters. Only 6% will give the DUP a 2nd preference and only 37% to any unionist party. Indeed, half of UUP voters would give the SDLP or Alliance Party their 2nd preferences. The transfer pattern for DUP and TUV voters are similar to recent elections. A significant change from 2011 is that no Alliance voters are willing to give the DUP a 2nd preference. Similar to UUP voters, SDLP voters are now much more willing to give cross community transfers to the UUP. Although 31% of SF voters claim they have no 2nd preference this is unlikely to occur in the actual election since most SF voters do have 2nd preferences. Although much publicity has been given to 2nd preferences it will actually have little influence in most constituencies outside of the Belfast area. In the 9 constituencies that I covered in the past week this is what one would expect. SF 2nd preferences will not matter in 8 constituencies. This is because SF is attempting to win 3 seats in 4, 2 seats in 4 and one seat in North Antrim. The only conceivably constituency where SF transfers could come into play would be East Derry, where the 2nd SF candidate could be eliminated and there would be a SF surplus. Similarly, the SDLP will have a surplus only in Newry and Armagh and the main obstacle for SF in winning 3 seats there is the low nationalist turnout. The UUP transfers will matter only in Foyle and possibly Upper Bann. This is because the UUP will be in the last count trying to elect one in every other constituency. The UUP transfers in Upper Bann (if they occur) could elect the SDLP and defeat SF. In Foyle, the lack of UUP transfers to the DUP could defeat the DUP and elect PBP.
Lucid Talk also did a sub tracker poll this past week where they asked the voters who voted in their January poll if they had changed who they would vote for. Lucid Talk determined that 9% of voters had changed their choice to a different party. The net effect compared to their January poll was approximately
Alliance +1
Green +0.5
PBP +1.5
DUP -2
SF -1.5
It appears that SF is losing votes to PBP and there has been more losses for the DUP compared to the previous drop of 3.3% from the 2016 Assembly election. In practical terms the shift to PBP overestimates their increase because they are only competing in 6 constituencies. Since at most half of their potential electorate is in those 6 then their voters in the other 12 would choose other candidates or not vote.
I will attempt to take into account 2nd preferences in my analysis.
NORTH BELFAST
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 81,477 | 38,342 | 37,864 | 849 | 4,422 |
| 47.06% | 46.47% | 1.04% | 5.43% |
New Quota 6,096
EU Referendum Turnout 39,972 (+3,424)
Remain 50.4%
Leave 49.6%
I believe North Belfast will elect 2 SF, 1 SDLP and 2 DUP. I came to this conclusion by looking at the final count in the 2016 election. In the final count these were the totals for the candidates
DUP
5,909
5,286
4,979
SF
5,225 (5,345)
4,911
SDLP
4,847
Alliance
3,836
The SF vote underestimates what would be their actual final total in 2017 since Gerry Kelly had 120 of his surplus transfer to candidates other than SF. Both the DUP and SF have good balancing. In 2017 the quota is higher at 6,096 so no candidates would be elected until the Alliance candidate is eliminated and their transfers distributed. Since Alliance in 2016 was not eliminated until the last count we do not know where their transfers would go but there are clues from the 2011 election. IN 2011 the 2 SF and 1 SDLP candidates were already elected at the point where the Alliance candidate was eliminated and the Alliance transfers were DUP 20%, UUP 25%, nontransferable 55%. So 55% of Alliance voters were unwilling to give ANY preference to the DUP or UUP. Those 55% would be Alliance voters from a nationalist background who would have given preferences to the SDLP or SF. What we do not know is, of the 45% who did give a preference to the DUP or UUP, how many would have ranked the SDLP above the DUP or UUP. I estimate 10% would have given the SDLP a higher preference than the DUP or UUP. It could be a little higher, it could be a little lower. How was 2016 any different than 2011? Since 2011 there were the flag protests in Belfast and the Alliance Party was the subject of verbal and physical attacks from hardline unionism, including the DUP. We do not know how this exactly affected Alliance transfers to the DUP but we do have a clue from South Belfast. In 2016 at stage 11, the Alliance candidate Paula Bradshaw, had a surplus. 80% of that surplus went to the SDLP and Greens (almost equally), 15% went to the UUP and 5% to the DUP. In 2017 Alliance voters would be even less inclined to give the DUP any preference due to the pro Brexit position of the DUP and the RHI scandal. If this were 2011 we would expect that 2,500 of the Alliance voters would transfer to the SDLP (most likely) with a few to SF. Another 1,300 would have transferred to the UUP or DUP. The SDLP will easily receive the 1,250 Alliance transfers it needs to reach a quota and will likely have a surplus. Of the additional 1,250 Alliance transfers that are going in the direction of the SDLP or SF how many will give SF a preference? Since the Gerry Kelly vote would really be 5,345, even if only 800 give SF a preference then both SF candidates would be in the range of 5,400 to 5,700. It should be noted that the final Alliance vote of 3,836 includes almost 900 transfers from PBP, Greens, etc. who would likely give SF some preference. I doubt that more than 200 to 300 Alliance transfers will give any preference to the DUP. That is enough to put one candidate at a quota but the other 2 would be at 5,300 and 5,100 with the latter too far behind SF to be elected. My calculation does not take into account that the Lucid Talk poll indicated that some UUP voters this year are less willing to give any preference to the DUP so the actual DUP totals would be less on the final count. In 2016 the DUP received 60% of UUP transfers and it could be much less than 50% in 2017. Of course, I am not assuming any increase in nationalist turnout while the Lucid Talk poll does indicate an increased nationalist, Green and Alliance turnout. The DUP can only elect 2 in North Belfast since they will be starved for transfers from the Alliance Party and to a lesser extent the UUP.
WEST BELFAST
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 73,575 | 59,198 | 12,230 | 447 | 1,700 |
| 80.46% | 16.62% | 0.61% | 2.31% |
New Quota 6,047
EU Referendum Turnout 31,191 (-5,059)
Remain 74%
Leave 26%
There will definitely be 3 SF and 1 PBP elected here. Whether the SDLP or SF win the 5th will be dependent on whether or not Alex Attwood is ahead of the DUP candidate in the next to last count. In 2016 the SDLP (4,430) was 90 votes ahead of the DUP in the final count and won the final seat. However, in 2017 there are only 5 seats so there is no possibility of a DUP win. It would seem that the SDLP would receive more than enough unionist transfers to win the 5th seat. SF has good balancing but the SDLP would only need 800 unionist transfers out of the total unionist vote of 4,340 to win the last seat. However, in 2017 this scenario may not occur. That is because the SDLP final vote includes an estimated 350 transfer votes from SF and the PBP. The SF transfers will not exist and PBP have fewer transfers available due to the higher quota. So the SDLP needs either more 1st preference votes or fewer DUP final votes that add up to 350. The SDLP can only win if they are ahead of the DUP in the next to final count. Now there will be fewer UUP, Alliance and Green transfers to the DUP this year (551 in 2016) so the DUP vote total will be less and some of those will transfer to the SDLP. It is not clear whether that 350 vote gap can be closed. Unionist turnout could be lower since there is no hope of a unionist seat with the higher quota. The 5th seat will come down to SF or the SDLP and will entirely depend on whether the SDLP are ahead of the DUP in the next to last count.
SOUTH BELFAST
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 93,699 | 41,615 | 41,353 | 2,583 | 8,148 |
| 44.41% | 44.13% | 2.76% | 8.70% |
New Quota 6,121
EU Referendum Turnout 44,556 (+7,860)
Remain 69.5%
Leave 30.5%
I believe South Belfast will elect 1 SF, 1 SDLP, 1 Alliance, 1 Green and 1 DUP. I come to this conclusion by looking at the stage 9 count in the 2016 election which was as follows:
SF 5,247 (5,400)
SDLP (Hanna) 4,934
SDLP (McKinney) 3,057
Alliance 5,729
Green 4,524
UUP 3,272
DUP (Little-Pengelly) 5,073
DUP (Stalford) 4,368
The new quota is 6,121 so no one would be elected at that point. I estimate the actual SF vote at 5,400 since the Labour candidate with almost 1,000 votes was eliminated in the previous count but SF did not receive any of the potential transfers since the SF candidate had already reached a quota. Some of the Labour transfers would go to SF in 2017. The SDLP
candidate with 3,057 would then be eliminated and Hanna would be elected with a large surplus. Eventually her 3rd and lower preferences would come into play. However, Hanna only needs 1,200 of McKinney’s vote to reach quota and the other 1,850 votes would go to SF, Alliance and Green. SF would need 700 of those and Alliance only 400 so both would be elected. The Green candidate would then be at approximately 5,200. At this point the UUP would be eliminated and approximately 3,300 votes would be transferred. The Lucid Talk poll indicates that many UUP voters will now give their 2nd preferences to Alliance, Green and the SDLP instead of the DUP. The Green candidate only need 900 of those to reach a quota (27%) which will almost certainly occur. Even if all of the other 2,400 transfer to the DUP the 2nd DUP candidate would be over 400 votes short of a quota. The Green candidate would win even with only 700 transfers (20%). The DUP have the same problem here as in North Belfast where they are starved for transfers from non unionist parties and to some extent the UUP. There is an alternative scenario where the Alliance Party, or less likely the 2nd SDLP, could win the 5th seat. This was the vote totals at stage 8.
SF 5,247 (5,400)
SDLP (Hanna) 4,738
SDLP (McKinney) 2,964
Alliance (Bradshaw) 3,570
Alliance (Morrow) 2,837
Green 4,289
UUP 3,213
DUP (Little-Pengelly) 5,049
DUP (Stalford) 4,355
Both Morrow and McKinney are slightly behind the UUP. If the Alliance vote was more evenly balanced (and with 500 more 1st preference Alliance votes with a higher turnout) and the Alliance vote was higher than the SDLP then McKinney would be eliminated. After his transfers elected Hanna there would still be a surplus of 1,500 votes that would go to SF, the 2 Alliance and 1 Green candidates. SF would be close to a quota and there would be only 800 votes to transfer to Alliance and Green. When the UUP is eliminated then there would be 3,200 votes transferred. The Lucid Talk poll indicates that over 2,000 would go to the SDLP and Alliance with a few to the Greens. Since that poll indicates a much stronger 2nd preference for Alliance (23.7%) than Green (3.2%) then it is possible that both Alliance candidates could be ahead of the Green Party. The Greens still have the edge but a significant increase in Alliance 1st preferences and the willingness of UUP voters to transfer much more heavily to Alliance than the Greens could give Alliance 2. This is much less likely than the first scenario I presented. A similar scenario could occur with the SDLP. At stage 9 the 2nd SDLP was only 215 votes behind the UUP and if the 2nd SDLP were ahead of the UUP then the UUP candidate would be eliminated and many of the UUP transfers would end up with the SDLP candidates. There were also almost 8,000 additional pro EU voters who voted in the Brexit referendum and if these vote in any significant number the DUP cannot possibly win 2 seats.
EAST BELFAST
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 74,979 | 9,571 | 57,695 | 1,067 | 6,646 |
| 12.76% | 76.95% | 1.42% | 8.86% |
New Quota 6,196
EU Referendum Turnout 42,646 (+5,490)
Remain 49%
Leave 51%
The result in East Belfast should be 2 DUP, 1 UUP and 1 Alliance. Alliance is short of 2 quotas but will receive enough Green and nationalist transfers to elect 2. It appears that with perfect balancing the DUP could edge out the UUP this will not occur. The UUP are at 11.1% but there is almost a 6% surplus of nationalist and Green potential transfers after 2 alliance are elected. There is also 13% for the smaller unionist parties that will transfer and the UUP received the equivalent of 4% in transfers from those parties in 2016.
EAST ANTRIM
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 72,580 | 14,952 | 51,646 | 738 | 5,244 |
| 20.60% | 71.16% | 1.02% | 7.23% | |
|
|
New Quota 5,402
EU Referendum Turnout 41,545 (+9,153)
Remain 45%
Leave 55%
For East Antrim I looked at the stage 10 count to determine who will win the 5 seats. The new quota is 5,402.
Alliance 5,234
UUP (Beggs) 4,708 (4,800)
UUP (Stewart) 2,622
DUP (Hilditch) 4,631 (5,400)
DUP (Lyons) 4,631 (5,000)
DUP (Ross) 3,108
UKIP (Jordan) 2,986
SF 3,470
The numbers in parenthesis are the actual total for a 2017 election due to a higher quota and in 2016 these candidates had their surplus transferred. Stewart of the UUP would be eliminated and enough of his votes would transfer to Alliance (168 needed) to reach a quota. Most of the remainder would go to Beggs of the UUP and he would be elected. There would still be another 1,700 available for eventual transfer to the DUP or UKIP. In 2016 they did transfer 33% to DUP and 18% to UKIP with almost 50% not transferring. UKIP would be at 3,300 but there would be only 600 votes to transfer to the remaining 2 DUP candidates. Lyons could be at or just below a quota but Ross would also be at approximately 3,300. Either the DUP or UKIP would win the 5th seat on the others transfers most likely reaching a quota. SF would be 2,000 votes short of a quota. This is where UKIP has their best chance for a seat and poor balancing by DUP or fewer voters willing to transfer to the DUP could result in a UKIP win. Nationalist turnout is very low here but there were over 9,000 more voters who voted in Brexit and an estimated 4,000 were nationalist voters. If 2,500 of those vote in 2017 there would be a nationalist seat here. The difficulty for the SDLP winning here is that they need an additional 1,000 1st preference votes in order to be ahead of the 2nd Alliance candidate, whose transfers would put them ahead of SF.
SOUTH ANTRIM
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 77,818 | 24,699 | 47,700 | 704 | 4,715 |
| 31.74% | 61.30% | 0.90% | 6.06% |
New Quota 5,856
EU Referendum Turnout 43,553 (+8,430)
Remain 49.5%
Leave 50.5%
The DUP should elect 2 as their total vote is over 2 quotas. Even with a decline to less than 33.33% they would receive enough transfers from the smaller unionist parties to elect 2. The UUP are above 1 quota and will elect one. There will be 2 non unionist seats. Or will there be? At stage 5 in 2016 the vote was
SF 4,739
SDLP 3,547
Alliance 3,764
UUP 4,452
UUP 4,031
DUP 5,067
DUP 5,036
DUP 4,212
At the next stage in 2016 the SDLP was eliminated because they were 217 votes behind Alliance. If the SDLP end up with more votes than Alliance then Alliance would be eliminated and the SDLP would win a seat on transfers. However, there would not be enough Alliance transfers to elect SF also. A 2nd UUP would be elected on Alliance transfers and the DUP surplus. Nationalist turnout is very low here, 37% versus unionist 52%. An increase in nationalist turnout to 45% would add another 2,000 nationalist votes which would be more than enough to elect both the SDLP and SF. An additional 8,430 voters did vote in the Brexit referendum and it appears from that vote that at least 5,000 were nationalist voters.
LAGAN VALLEY
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 80,861 | 15,331 | 59,343 | 818 | 5,369 |
| 18.96% | 73.39% | 1.01% | 6.64% |
New Quota 6,476
EU Referendum Turnout 48,414 (+9,572)
Remain 47%
Leave 53%
This is a difficult constituency to predict. From the 2016 vote it appears there would be 3 DUP elected since the DUP at 47.2% are close enough to 3 quotas to elect all 3 on transfers from the smaller unionist parties. The UUP would elect one and Alliance one. This was the stage 6 count in 2016
SDLP 4,012
Alliance 4,994
UUP 5,004
UUP 4,737
DUP 5,635
DUP 4,986
DUP 4,518
DUP 4,309
The SDLP were eliminated and in 2017 they have enough transfers to elect Alliance and one UUP. In 2016, 60% of SDLP transfers were eventually nontransferable to Alliance or the UUP. Lucid Talk indicates a greater willingness for SDLP voters to give the UUP a preference in 2017. However, there may not enough to elect 2 UUP. The lowest DUP candidate would then be eliminated and the total DUP vote is just shy of 3 quotas but certainly ahead of the 2nd UUP. The major difficulty in predicting 2017 is that there were 9,572 additional voters who voted in the Brexit referendum and from the results it appears they were all nationalist, Alliance, Green and a few pro EU UUP voters. Half were nationalist voters and nationalist turnout for Brexit was almost twice what it was for the Assembly election. If only 3,600 of those vote in 2017 then everyone’s quota increases by 600 (new quota 7,100). If Alliance are then at 6,500 and the SDLP at 6,000 then the 2nd UUP is eliminated and Alliance and possibly the SDLP are elected on transfers. The SDLP 1st preference vote may be helped by two factors. The former Alliance MLA Seamus Close endorsed and signed Pat Catney’s nomination papers. There are also 2 Alliance councilors on the Lisburn Castlereagh council who recently resigned from the Alliance Party and endorsed Pat Catney. Michael Nesbitt the leader of the UUP has also publicly stated that he will give his own 2nd preference to the SDLP.
STRANGFORD
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None |
| 72,001 | 12,496 | 53,673 | 684 | 5,148 |
| 17.36% | 74.54% | 0.95% | 7.15% |
New Quota 5,440
EU Referendum Turnout 42,110 (+9,484)
Remain 44%
Leave 56%
In 2016 the final count was:
SDLP 3,338
Alliance 4,662 (4,624)
UUP (Nesbitt) 4,673
UUP (Smith) 3,958 (3,248)
DUP (Bell) 4,663 (5,164)
DUP (Hamilton) 4,663 (4,983)
DUP (McIlveen) 4,663
The numbers in parenthesis represent what the actual totals would be with the higher quota. 2 DUP candidates had a significant surplus that transferred to the UUP. Smith of the UUP would be slightly behind the SDLP and his transfers (777) would elect Nesbitt. There would still be 2,471 UUP votes available to transfer and it is likely that Alliance would receive the 800 needed for a quota. Based on previous elections the SDLP would receive enough to reach 3,600 but would be far short of a quota. So there would be 3 DUP, 1 UUP and 1 Alliance elected. This is another constituency where the Brexit vote was much higher than the Assembly vote. An additional 9,500 voters voted in the Brexit referendum and it appears that they were mainly nationalist, Alliance, Green and pro EU UUP voters. At least 4,000 appear to be nationalist voters. Nationalist turnout was only 35% for the Assembly election. If 3,000 of those Brexit voters vote in 2017 the quota would be close to 6,000. The 1st scenario above where the 2nd UUP candidate is eliminated would then come into play but Alliance could already be at a quota with the additional voters. The SDLP could be at 4,500 to 5,000 prior to the UUP transfers and would likely be ahead of the 3rd DUP after UUP transfers. The role of Jonathan Bell of the DUP in revealing the extent of the RHI scandal could affect the DUP vote. Since he is a candidate he will be a constant reminder to voters of the RHI scandal. It is unlikely that many of his votes will transfer to the DUP. It is also possible that his allegations may result in fewer UUP transfers to the DUP. Michael Nesbitt has also publicly stated that he will give his next preference to the SDLP after the 2 UUP candidates. This is a shrewd move since if Joe Boyle is eliminated the bulk of the excess SDLP transfers could end up with the UUP after the Alliance candidate is elected and a 2nd UUP could be elected.
NORTH DOWN
| All usual residents | Catholic | Protestant and Other Christian (including Christian related) | Other religions | None | ||
| 72,633 | 9,420 | 55,164 | 870 | 7,179 | ||
| 12.97% | 75.95% | 1.20% | 9.88% | |||
New Quota 5,372
EU Referendum Turnout 44,177 (+11,982)
Remain 52%
Leave 48%
North Down will elect 2 DUP, 1 UUP, 1 Alliance and 1 Green. The DUP vote is well above 2 quotas and even with a small decline in their vote would be above 2 quotas. The UUP are only 1% shy of a quota which they will receive on transfers from the smaller unionist parties. Alliance is at a quota and the Greens should reach a quota on SF, SDLP and Other transfers.
The final results will be strongly influenced by any change in turnout. If the exact same electorate votes that voted in the 2016 Assembly election then the worst case scenario for nationalist parties (SF, SDLP, PBP) is 34 seats with 47 unionist seats and 9 Alliance-Green. The electorate will be different in 2017. Lucid Talk has picked up increased interest among nationalist voters for this election and a slight decrease in interest among unionist voters. I estimate the unionist vote could be down by 5,000 due to disillusioned DUP voters who will stay home. That would bring the total vote down to 690,000. It is difficult to quantify how many more nationalist voters will vote this year. It is probably enough to increase the total vote to 700,000. I doubt the total vote will be anywhere near 800,000 which is what one would expect with all the Brexit voters voting this year. 750,000 is probably the maximum. However, even a decline of 5,000 unionist voters and an increase of 10,000 nationalist votes would have a profound effect on the outcome. There would then be 3 more additional nationalist seats, 1 more Green and 4 fewer unionist with a final result of 43 unionist, 37 nationalist and 10 Alliance-Green. If this were to occur then a government could be formed at Stormont if a few changes were made to the GFA. These would include:
Abolition of the Petition of Concern.
End community designation and cross community requirements for any votes.
Voluntary coalition with one First Minister elected.
No weighted majorities.
If these changes were made the Alliance and Green parties would hold the balance of power. Nationalist parties would have no objection to Stephen Agnew or Naomi Long as a First Minister. The DUP would object to both but would be unable to prevent their election since there would no longer be a unionist majority to do so.
The alternative is no government at Stormont and a return to direct rule. Since the Conservative Party will be in power at Westminster for a long time this would essentially be a return to unionist rule with no nationalist representation.
