The Brexit referendum is over and Northern Ireland voted 440707Remain (55.75%) and 349,422 Leave (44.2%). The Total vote was 790,149 which was far above the Assembly vote in May.
Who are these additional voters and how many were there? This requires quite a bit of detective work.
To begin with we know that 694,314 voted for candidates in the May Assembly election. The breakdown by party, candidate, and sectarian vote was:
Unionist Parties and Candidates 343,732 49.5%
Nationalist Parties and Candidates 274,328 39.5%
Nonsectarian Parties and Candidates 76,254 11.0%
I included People Before Profit in the nationalist vote since their entire vote appears to be from the nationalist community.
How many of these people voted in the EU Referendum? There is one group who voted in the Assembly election but who were ineligible to vote in the EU Referendum and these are EU nationals who are not UK citizens. There are over 80,000 EU nationals old enough to vote but only 30,000 were on the electoral register as non UK citizens. Assuming the turnout was similar to other voters then 14,000 to 15,000 voted. Removing these voters would change the party vote. I assume that few voted for unionist parties since most of these EU nationals are of Catholic background AND it is unlikely they would vote for unionist parties that want to leave the EU and deport them. So excluding these voters the vote for the Assembly was approximately:
Unionist Parties and Candidates 343,000 50.5%
Nationalist Parties and Candidates 262,000 38.5%
Nonsectarian Parties and Candidates 75,000 11.0%
The total Brexit vote was 790,000 so an additional 110,000 voted. I am assuming that all or almost all the 680,000 that voted in the Assembly election also voted in the Brexit referendum. I believe this is accurate for unionist voters since unionist voters are very reliable voters and consistently vote in all elections. However, as I will show this is true for the net unionist vote but some did stay at home in some constituencies but were counter balanced by others who voted in other constituencies. This is also true for the nationalist vote and nonsectarian vote.
What would have the Brexit result if only the 680,000 Assembly voters voted? The clues are found in the Lucid Talk tracking poll for Brexit. The Lucid Talk tracking poll was very accurate for the Assembly election. There was some pre-election skepticism regarding their poll since it was predicting a nationalist vote of only 38% but given the final result it was very accurate. In the month prior to the election Lucid Talk found the following results;
Since the final result was 55.75% Remain to 44.25% Leave it appear the undecided broke more towards leave. Lucid found this was true in their final tracking poll one week prior to the election when there was a strong move of undecided unionist voters towards Leave. These were their findings:
Remain Leave Undecided
Unionist 13.6% 83.4% 3.0%
Nationalist 86.4% 12.2% 1.2%
Nonsectarian 83.2% 13.0% 3.8%
Clearly the unionist voter preference was the mirror image of the nationalist-nonsectarian voter preference. After dividing up the Undecided the final preference would be approximately:
Unionist 15.0% 85.0%
Nationalist 87.0% 13.0%
Nonsectarian 85.0% 15.0%
For the 680,000 Assembly voters the Brexit vote would have been
Unionist 51,500 291,500
Nationalist 228,000 34,000
Nonsectarian 64,000 11,000
Total 343,500 336,500
Obviously the vote would have been very close with only a slight edge for Remain. Since the actual vote was 440,707 Remain and 349,422 Leave the additional 110,000 voters who voted in the Brexit referendum voted
Remain 97,207 88.25%
Leave 12,942 11.75%
It is clear that those additional 110,000 voters must include few if any voters who would vote for unionist parties since the vote distribution is similar to the preference of Nationalist and nonsectarian voters. Indeed they were slightly more likely to favour Remain than nationalist and nonsectarian voters in the Lucid Talk Poll. However, I believe that a significant number did originate from the Protestant community though they appear to be liberal Protestants who would vote Alliance, Green or even SDLP. This will become more evident when I analyze the constituency vote.
These are the turnout figures by constituency for the May Assembly election
The nationalist turnout was much less than the unionist turnout in all constituencies except Foyle and Belfast West, South and North. Overall, nationalist turnout was 7% less than unionist turnout for all of Northern Ireland.
The following table shows the turnout by constituency for the Assembly and Brexit Referendum and the increase that occurred in each constituency for the Brexit Referendum. The vote for the Brexit Referendum is also shown by constituency.
I estimate that of the additional 110,000 voters that voted in the Brexit Referendum that approximately 75,000 originated from the Catholic community, 15,000 from the None/Other community and 20,000 from the Protestant community. I will now look at the constituency vote for Brexit and refer to the turnout figures for the 2 elections as needed.
I will start with the 6 nationalist constituencies of Foyle, West Tyrone, Mid Ulster, Fermanagh South Tyrone, Newry & Armagh and West Belfast. 5 of these constituencies had SF MP’s at the beginning of 2015.There was a large drop in the Brexit vote in West Belfast of over 5,000 and this was the only constituency that had fewer voters than the Assembly election. These would have been mainly SF voters who stayed home. The net increase for these 6 was only 3,100. The vote breakdown in Foyle, West Tyrone, Mid Ulster and Newry & Armagh indicates that the net additional voters were nationalist voters. The native nationalist turnout in the latter 3 would still have been slightly lower than unionist turnout. Fermanagh South Tyrone is different. The Brexit vote can only be explained by a decline of 2,000 unionist voters and an increase of 3,000 native nationalist voters. The unionist turnout was 71% in May, the highest of any constituency and with the decline it would still be higher than the native nationalist turnout. There would have been approximately 5,000 EU voters who did not vote in Brexit so I am estimating that there was a net increase of 10,000 native nationalist voters, 1,000 None/other and a decline of 3,000 unionist voters for all 6.
The next 4 constituencies I looked at are East Derry, North Antrim, Upper Bann and South Down. The net increase in the Brexit vote was 27,700 and in East Derry, Upper Bann and South Down it appears to have been almost entirely from the Catholic community. In East Derry the Brexit turnout was 51%. In May the unionist turnout was 51% and the nationalist turnout only 34%. If the nationalist turnout increased to 51% it would add 6,000 additional voters which is what occurred. It is also the only explanation for a Remain vote of 48%. The native Catholic electorate is 43% and the other 5% came from None/Other and Protestant Alliance-Green voters. A similar analysis applies to Upper Bann and South Down where the low nationalist turnout increased to the level of the unionist turnout and accounts for the increased vote in Brexit. North Antrim is a bit different. Overall turnout there was 58% and if the nationalist turnout increased from that of 36% in the Assembly election to 61% in Brexit that would add over 6,000 votes. Since the increased vote was 8,742 the additional 2,742 votes would have come from the None/Other and Protestant communities, probably equally from each. Overall for these 4 constituencies there would have been an additional net 22,000 Catholic voters but probably 27,000 native voters after allowing for EU nationals who voted in the Assembly election. The turnout from the None/Other voters probably increased by at least 4,000.
The next 4 constituencies I will look at are South Antrim, North, East, and South Belfast. In these constituencies the increased vote appears to have been mainly from the Catholic community but also significantly from the None/Other community. In East Belfast the increased vote appears to be entirely due to an increased turnout of Catholic and None/Other voters. Indeed, the Remain and Leave vote is almost identical to the Unionist and non-unionist percentage vote in the 2015 Westminster election. Some of those voters did not vote in the Assembly election but then voted again in Brexit. In South Belfast the increased vote of 7,860 appears to be from the Catholic community with perhaps 2,500 from the Other/None community. Only in South Antrim does there appear to be some increase from the Protestant community. 5,000 appear to be from the Catholic community, 1,500 from the None/Other community and 2,000 from the Protestant community. The additional Protestant voters may be those who voted in the 2015 Westminster election for Danny Kinahan but did not vote in May 2016. I suspect many of the latter voted Remain. The small increase in the North Belfast vote appears to be mainly Catholics and None/Other.
The final 4 constituencies are North Down, Strangford, East Antrim and Lagan Valley. North Down had an increase of almost 12,000 voters compared to the May Assembly election. Indeed, it had the highest percentage turnout of any constituency. This is quite amazing since North Down usually has the lowest turnout of any constituency. It appears that 3,500 came from the Catholic community, 2,500 from the None/Other community and 6,000 from the Protestant community. The majority of those additional voters from the Protestant community appear to have voted Remain. Lagan Valley had an increased vote of 9,572 and I estimate that 5,000 were from the Catholic community, 1,500 from the None/other community and 3,000 from the Protestant community. For East Antrim I estimate 3,500 from the Catholic community, 1,500 from the None/Other community and over 4,000 from the Protestant community. For Strangford I estimate 3,500 from the Catholic community, 1,500 from the None/Other community and 4,500 from the Protestant community.
The increased turnout for the Brexit referendum appears to be entirely due to Catholic, None/Other and liberal Protestants. The increased Catholic vote of 75,000 may seem high but the May Assembly turnout was only 45% and increasing that turnout to 57% would result in that high an increase. This is the same turnout from the Catholic community that occurred in the 2005 Westminster and 2003 Assembly elections. The nationalist voter apathy that existed from elections beginning in 2011 disappeared in this election. The increased turnout from the Catholic community appears to be mainly middle class Catholics who would vote SDLP, though a significant minority would vote SF, Alliance or Green. This is because of which constituencies the increased turnout occurred in. The None/Other group is 70,000 potential voters. Opinion polls have shown that this group would vote 50% Alliance/Green, 25% nationalist and 25 % unionist. However, they also vote at very low rates since they do not identify with the usual sectarian elections. At most 20,000 voted in May but this probably increased by 15,000 for this election. It could have been more but this is a young demographic and young voters have a lower turnout than older voters. The additional 20,000 from the Protestant community that voted appear to be liberal Protestants who voted Remain. They were exclusively located in the suburban Belfast constituencies.
If the Assembly election had been held the same day as the Brexit referendum the results would have been very different than in May. Alliance would have picked up an additional seat in North Down and possibly East Antrim and the Greens would add one in East Belfast. There would have been an additional SF seat in East Derry and additional SDLP seats in Upper Bann, Lagan Valley, South Antrim, North Antrim, Strangford and possibly South Down though one would have been lost to the DUP in West Belfast.
Although there was a large increase in the vote from the Catholic and None/Other communities I do not expect that this will recur in future Assembly and Westminster elections. Most of the Westminster elections outside of Fermanagh South Tyrone, North Belfast and East Belfast are not competitive and most of the increased vote was in the noncompetitive seats. These voters are also unlikely to vote in an Assembly election as it makes no difference in their lives whether there are 56 unionist, 42 nationalist and 10 Other MLA’s or 48 unionist, 48 nationalist and 12 Other MLA’s. The nature of the Executive is such that the outcome is already pre-determined prior to the election.
However, whether the UK is in or out of the EU potentially has a major impact on their lives which is why they voted in the Brexit Referendum.