LucidTalk conducted a poll of Northern Ireland voters and their Westminster voting preferences from November 27th to November 30th. The poll was commissioned by Electoral Calculus and Remain United. I will discuss some of the findings in this poll.
Compared to the previous poll there are increased percentages of voters who state they are 100% certain to vote. In the previous poll unionist voters were significantly less likely to be 100% certain of voting compared to nationalist voters. However, now the percentages are almost equal between nationalist and unionist voters in the 94% to 95% range. However, Protestant voters are 91% certain to vote and Catholic voters 95% certain to vote.
Westminster voting intentions, excluding nonvoters and undecided (5% of voters), showed the results were:
The Other group includes PBP, Aontu, Independents, Green, UKIP and Conservatives. Over 70% of the vote for Others comes from the Catholic community and only 18% from the Protestant community. This indicates that much of that 5% is PBP, Aontu, Greens and Independents.
If only those 100% certain to vote are included the unionist vote would decrease minimally in the poll. The results are much more influenced by the weighting of the sample. The sample of those 100% certain to vote was weighted by Religion as follows:
The percentage of Catholics and Protestants are therefore equal and this is consistent with an extrapolation of the 2011 Census adding in new voters and removing voters (mainly elderly) who have died.
However, the 22.4% shown as None/Other is much higher than the 11% indicated in the census. The higher sample in the poll could be due to the increasing secular identity in Northern Ireland since 2011. It could also be due to over sampling of None/Other voters. A 22.4% None/Other sample, rather than the 11% indicated in the census, has a significant effect on the poll findings. This is because the Alliance voter preference in the None/Other group is over 40%.
I calculated voter preference based on a None/Other percentage similar to the 11% in the census and Catholic-Protestant percentages in the census. The results are:
The total SF-SDLP vote goes up by 1.1% and the total UUP-DUP vote goes up by 1.8%. The Alliance vote declines by 3%.
Electoral Calculus did do seat predictions but the average number of voters per constituency would be 130 which is too small a sample to give accurate predictions. The 5.2% vote for minor candidates may decrease since there is a tendency for such voters not to waste their vote on election day.
Based on the new poll finding and my adjustment I would conclude that:
The DUP vote will be down significantly from their 36% in 2107.
If this is a uniform decline they will not win in North Down, South Belfast and North Belfast. They could also lose East Belfast and South Antrim and possibly Upper Bann.
The Alliance vote will be up significantly from their 7.9% in 2017. It could be over 50% higher than their 2017 vote. This will probably not be a uniform swing and will most likely be concentrated in North Down, East Belfast, South Belfast and South Antrim. In half the constituencies they will poll 3% to 4%.
The vote for the SDLP, UUP and SF are within the margin of error for their 2017 results. SF may lose 2% of their vote to Aontu but this could only be of significance in Foyle and South Down.