Part 2 of Fahas analysis in the leadup to this weeks assembly election covering the greater Belfast area. I have some caveats regarding the North Down constituency but I’ll leave it to yourselves to comment. BD
This is my analysis of the 9 Belfast area constituencies. The electoral office released the final electoral register on April 28 and there has been an increase of 14,000 voters since December. Over 10,500 of that increase has occurred since March. It appears that approximately 3,000 of those new voters are foreign nationals and 11,000 are native voters. The average per constituency is 777 and the largest increases have been in South Belfast (1,259), Upper Bann (1,203), West Belfast (1,092) and Newry&Armagh (1,005).
North Down has always been an unpredictable constituency. The 2015 Westminster election provides no guidance to the Assembly results since independent MP Sylvia Hermon received half the vote. In 2011 there were 3 DUP candidates elected but this will change in 2016. The 2014 Council elections provide a more accurate indication for 2016. In that election the DUP received 31% which is 2.4% above 2 quotas. It would require perfect balancing and significant transfers to elect 3 this time. The UUP received 17.2% in 2014. However, the UUP have Alan Chambers as one of their candidates and in 2011 he received over 6% of the 1st preference vote. Because he is one of the UUP candidates the UUP 1st preference vote could be as high as 23%. The UUP will receive transfers from UKIP, TUV, Conservatives and other parties so will likely elect 2. There will be 2 non unionists elected. Brian Wilson is standing as an independent and will take 1st preference votes from both the Alliance and Green parties. Alliance should still elect 1 but it is not clear whether Brian Wilson will be a threat to the Green seat. It may depend on where the SDLP and SF transfers go to. I included the 1998 Assembly election results to illustrate a point. In 1998 the SDLP had over 2,000 1st preference votes. Based on the 2016 demographics one would expect the 1st preference SDLP and SF vote to be almost 3,000. However this constituency has been ignored by both the SDLP and SF and few voters from the Catholic community vote and of those that do the majority vote Alliance or Green. Contrast this with West Belfast, where even though unionists have only a small chance of winning a seat, they make a serious effort and go all out to win a seat for a unionist candidate.
East Belfast is likely to see major changes in 2016 if the same electorate that voted in the 2015 Westminster election vote in the 2016 election. If you look at the results of the 201 Westminster election the total unionist vote was 20,467 and the total non unionist vote was 14,021. In the 2015 Westminster election the total unionist vote increased by only 229 votes to 20,696 but the total non unionist vote increased by almost 5,000 votes to 18,986. Who are those extra 5,000 non unionist voters? We have no way of knowing for certain but it is likely that many are from the Catholic, Other religion and No religion communities since these groups have historically voted at low levels in the past. Some could also be liberal Protestants who voted for the 1st time. The big question for 2016 is will these 5,000 voters vote again this year. If they do then there will be 3 non unionist MLA’s elected. Based on the 2014 Council results there will be 2 DUP and 1 UUP elected. The 2015 Westminster results, with the Alliance vote at 42.8%, predict 3 Alliance seats. However, it is likely that there was a large amount of tactical voting by SDLP, SF, Green and independents for Naomi Long so the 1st preference Alliance vote will certainly by less than 42.8%. The Green party will be in contention for that 3rd non unionist seat. Naomi Long will receive a very high 1st preference vote which may be to the detriment of the 2 other Alliance candidates. If one of those has a low 1st preference vote and is eliminated then the Green candidate could be elected. I include the 1998 Assembly results to illustrate an important point. East Belfast had the highest increase in Catholic population between the 2001 and 2011 census of any constituency. It was 5%. The number of voting age I indicate for 2016 only includes those living in East Belfast in 2011. The 2016 number is probably 11,500 when taking into account the rate at which Catholics have been moving in to East Belfast since 2001. That 11,500 is twice the number that was present in 1998 when the total 1st preference nationalist vote was over 2,000. Theoretically it should be twice that in 2016 at 4,000. However, similar to North Down this constituency has been ignored by SDF and the SDLP. SF only compete in 1 ward, the Short Strand area and the SDLP have ignored the constituency since 1998. After the boundary review East Belfast will be adding wards from South Belfast so it is a poor strategy for both SF and the SDLP to ignore East Belfast since there would be a potential nationalist seat in the new Southeast Belfast.
The results in South Belfast will be determined by which electorate shows up on May 5th– the 2014 Council electorate or the 2015 Westminster electorate. If it is the 2014 Council electorate then the SDLP risk losing a seat to either Alliance or the DUP. In the 2015 Westminster election the number of voters increased by 6,000 and almost ¾ were SDLP or SF voters. 2015 also had the highest turnout of any election in South Belfast since the 1998 Assembly election. SF should win their seat here as they need very few transfers. If the SDLP balance their candidates evenly they should elect 2 also since they would receive significant transfers from the Green Party and Alliance surplus. Alliance and Green transfers will go at least 2 to 1 towards a nationalist candidate rather than a unionist. The change in the electorate since 2011 also favour nationalist candidates since the Catholic community electorate has increased by 1,900 while the Protestant community electorate has declined by 900. It will be difficult for unionist candidates to maintain their vote with a declining electorate. On the unionist side the DUP could win 2. The 2015 DUP Westminster vote at 22.1% was over twice the UUP vote of 9.1%. The UUP should pick up enough Conservative, Alliance and Green transfers to be competitive with the DUP. The DUP could still win 2 if the 2 candidates are perfectly balanced and they receive twice as many transfers from the PUP, TUV, UKIP and Ruth Patterson as the UUP do.
West Belfast will have some major changes compared to 2011. Similar to South Belfast it will be partly determined by which electorate shows up on May 5th, the 2014 Council electorate or the 2015 Westminster electorate. One certainty is that Gerry Carroll of People Before Profit Alliance will be elected as his 19.2% vote in 2015 is far above a quota with a 5% surplus. The SF vote has been in the 54% to 55% range in 2014 and 2015 which is slightly less than 4 quotas. However, they should receive enough transfers from PBP to reach 4 quotas. Nevertheless, they will lose 1 MLA compared to 2011. Theoretically, they could keep 5 with a poor showing by the SDLP and with all 5 of their candidates perfectly balanced at 11%. However, this is unlikely since the SDLP received almost 10% in 2015 and would receive some PBP transfers as well as transfers from Alliance, Green and the UUP. It is also possible that the DUP could win a seat. If the turnout is similar to 2014, when the total unionist vote was 16.0% then the DUP would certainly win. However, in the 2015 election the total vote increased by 5,000 and these were all nationalist voters. In 2015 the total unionist vote was 13.2% which is only 1.1% less than a quota. The last nationalist candidate (probably SDLP) would also be less than a quota. It would have been very close. The one disadvantage the DUP have is Frank McCoubrey may not receive enough UUP transfers due to his previous association with the UDP. In 2011 over 100 UUP transfers went to nationalist candidates and 200 did not transfer. The number that do not transfer could be higher this year.
It was thought before the final candidate list was known that the SDLP seat could be under threat from Alliance or a 3rd SF candidate. If you look at the 2014 Council results the SDLP vote was only 1% more than the Alliance vote and SF was less than 2 quotas. However, 2015 was a different result with the total nationalist vote equaling 45.7% which is 2.8% over 3 quotas. Since there are only 2 SF and 1 SDLP candidates by default all 3 will be elected. There is still a scenario where the Alliance candidate could win. The Alliance vote was 7.25% in 2015. If the total nationalist vote were to increase to 47% or more then there would be at least a 4% nationalist surplus. This is possible since there are 1,500 EU nationals on the electoral register who were not eligible to vote for Westminster and if even half those vote for pro EU parties the nationalist vote could reach 47%. Furthermore, since 2011 the Catholic community electorate has increased by 2,400 while the Protestant community electorate has declined by 200. If Alliance receives 3% of the 4% surplus in transfers the Alliance vote would be 10% plus. If the Green and Labour candidates bring out additional voters then those transfers could benefit Alliance also. The 2014 Council vote shows a DUP vote of 28.4% which is 2 quotas so only 2 DUP will be elected. The last seat could come down to Alliance, UUP or PUP. The PUP vote was 7.4% but with TUV and UKIP transfers should be 10%. If the UUP is at or slightly less than 10% then the UUP candidate would be eliminated. If 3% of the UUP transfers go to Alliance, 3% to the PUP and the rest do not transfer then it could be very close between the PUP and Alliance on the last count.
In 2011, the SDLP lost here by only 458 votes. The main obstacle for the SDLP in winning a seat is nationalist voter apathy. Despite a potential electorate over 13,000 the total SDLP and SF vote barely reaches 3,200 with perhaps another 1,000 from the Catholic community voting Alliance. There are 3,500 voters from the Catholic community in Newtownards and Comber DEA’s and the Moneyreagh ward but turnout there is less than 20% and the majority of that 20% vote Alliance. Based on the 2015 vote there should be 3 DUP, 1 UUP and 1 Alliance elected easily with the final seat between the SDLP and either the TUV or UKIP. This constituency is similar to East Derry where basically stay at home nationalist voters are gifting a seat to the TUV or UKIP. There are 11 unionist candidate on the ballot so unionist turnout will probably be as high as it was in 2015. On the other hand UKIP and the TUV have few sources for transfers other than one of these parties to the other so one could be elected with as little as 10% or 11% of the vote on the final count. The Conservatives received over 2,000 votes in 2015 and in the 2014 Council election Conservative transfers were just as common to Alliance as they were to the TUV and UKIP.
In the 2014 and 2015 elections Lagan Valley had the 2nd lowest turnout from the Catholic community (only North Down is lower). It was approximately 25% whereas turnout from the Protestant electorate is near 50%. This is at a level that is seen only with an organized election boycott but no one has been able to discover what group is behind the boycott. Lagan Valley also had the 2nd largest increase in the Catholic population between 2001 and 2011 which was 4%. Since many of these people are moving in from West Belfast where nationalist voter turnout is high it is puzzling why the turnout is so low. In view of the in migration I estimate the Catholic electorate is 17,000 and with potential votes from the Other and None communities the potential electorate is 18,000. If even 40% voted there would be a nationalist quota even allowing for more than 1,500 voting Alliance. Based on the 2014 and 2015 elections there will be 3 DUP, 1 UUP and 1 Alliance. The last seat will be down to the last count between the SDLP and either the TUV or UKIP. There are also 1,500 EU nationals on the electoral register and this may be a constituency where EU voters vote at a higher rate than native nationalist voters. The SDLP do seem to be making a serious effort in this election so it could be close for the final seat.
South Antrim is difficult to predict because the Assembly and Council election results are so different than the Westminster results. This is due to tactical voting by Alliance and SDLP voters for the UUP in Westminster elections which was most noticeable in 2015 for Danny Kinahan. This is not the entire explanation since Westminster elections seem to bring out additional unionist voters who do not vote in Assembly and Council elections. In 2011 the SDLP lost on the final count by 1,250 votes. The demographics since 2011 show a net of 1,750 new voters in the Catholic community and 500 in the Protestant community. There was a 4% increase in the Catholic population between 2001 and 2011 indicating significant in migration so the true increase between 2011 and 2016 may be at least 2,500. The DUP and UUP are both standing 3 candidates but the Council results indicate that there are only 2 for the DUP and at most 2 for the UUP. The Alliance vote is always the highest for Assembly elections. The final seat will be between the SDLP and UUP and will depend on whether nationalist turnout increases from its usual low level. There are also 1,500 EU nationals on the electoral register and they will influence the final result.
East Antrim is one constituency where SF could lose their seat. It was only by luck that SF won the seat in 2011. In that election at stage 8 Gerardine Mulvenna of Alliance was only 68 votes behind Rodney McCune of the UUP when she was eliminated and 318 of her transfers went to SF and SF won on the last count by 500 votes. However, if the UUP had been eliminated instead the final SF vote would have been only 3,090 and Gerardine Mulvenna would have needed only 700 of the 2,400 UUP transfers available to win. Now East Antrim should elect a nationalist MLA but nationalist turnout is very low here- only 32%. The main reason it is so low is that SF and the SDLP ignore most of the constituency. The 15,000 Catholic electorate is equally divide in 3 areas- 5,000 in the Glens, 5,000 in Larne Town and 5,000 in Carrickfergus and Jordanstown area. The turnout in the Larne Town area extending to Jordanstown is only 20% to 25%. Based on the 2014 Council results there will be 2 DUP, 1 UUP and 1 Alliance elected. One of either UKIP or the TUV will also be elected. The last seat will be between the UUP and SF and will depend on turnout. There is a chance the SDLP could win the final seat. The SDLP were almost 700 votes behind SF in the 2015 election. However, they do not need 700 1st preference votes to overtake SF since if one of the Alliance candidates is eliminated then they would receive approximately 400 Alliance transfers (based on analysis of the 2011 vote). So the SDLP only need to increase their 1st preference vote by at most 400 to win. This is no small task considering the level of nationalist voter apathy.
The 3 most important factors that will determine the final results are
If the turnout is similar to the 2014 and 2015 elections then the number of nationalist MLA’s could decrease from 43 to 40 with a SF loss to a unionist or Alliance in East Antrim and a SDLP loss to the DUP in West Belfast and a SDLP loss to Alliance, Green or a unionist in South Belfast. If turnout from the Catholic community equaled turnout from the Protestant community in all constituencies then there would be 50 to 51 nationalist MLA’s elected. I hesitate to predict any increase in turnout for this election. Unionist voters are reliable when it comes to voting and in looking at the votes for unionist candidates in Westminster and Assembly elections going back as far as 2003 I expect the vote for unionist candidates to be between 340,000 and 370,000. The 319,000 in 2011 was an anomaly due to scandals within the DUP. The vote for nationalist candidates has been in the narrow range of 285,000 to 287,000 since 2010. Since EU nationals can vote in this election and there are 30,000 on the electoral register the nationalist vote should increase to 300,000. It is difficult to identify any issues that would increase turnout further. There has been much publicity in the press about abortion and gay marriage. The latter affects only a very small minority of the electorate and would have minimal influence on turnout. Abortion is an emotional issue and could increase turnout among voters who are either pro life or pro choice. The publicity surrounding the anniversary of the Easter Rising may increase overall interest in politics among nationalist voters but that is probably limited to people who are already voting anyway. It is also possible that Brexit could have a slight influence among nationalist voters who are concerned about the border or economic implications of Brexit.