General Election: Big changes in Upper Bann and South Belfast
So we've now got the final results from our April Opinion Panel Poll, and the last before the general election on 7th May.
If you've been following our monthly polls since January you will know that our Northern Ireland (NI) Opinion Panel has 440 participants and is carefully constructed to provide an accurate representation of Northern Ireland opinion - via gender, area of residence, age-group, community background, socio-economic group, and employment group. The opinion panel has a pool of approximately 1,200 members who regularly take part in LucidTalk poll projects, and for this project a representative sample of 440 opinions has been collated.
Our Opinion Panel members were asked: (a) Likelihood of voting in May 2015, and (b) What party they currently plan to vote for, and (c) Their current opinion regarding who would win in each of the 'non-100%' certainty seats (see results table).
As you may have seen from our previous monthly Opinion Panel reports, the data from each monthly panel poll is entered into our predictor models, which then forecast the results (as a % probability) for each of Northern Ireland's 18 Westminster constituencies, and track any trends up and down. For a full description of our Opinion Panel Methodology see: http://www.lucidtalk.co.uk/april-opinion-panel-results/
This is similar to the prediction models we ran for last May's European election when three weeks before election day we (in the Belfast Telegraph) predicted Sinn Fein for a 26.2% vote share with the actual result coming in at 25.5%; and the DUP for a 20.8% share with the actual result turning out to be 20.9%. So (in our humble opinion!) our Euro election prediction models weren't that bad! Results are presented as a % likelihood of the named political parties winning in each of the 18 Westminster Parliamentary seats. Our forecasts do not predict vote share, size of vote, and/or size of majority etc., as to do this would require a much higher sample and more comprehensive poll. In this context, it should be noted that if the % scores go up or down for the main prediction, then it's the alternative party or parties (right-hand column in the table) that is retreating or advancing by that same score. The full results are detailed in the attached table along with any changes from our March Opinion Panel poll, which enables us to see the trends, and any changes in opinion.
As an approximate guide the % predictions can be interpreted as follows: +55% = too close to call, +60% borderline, +65% = slight favourite, +70%-75% = moderate favourite/favourite, +80%-85% = favourite/strong favourite, +90%-95% = strong favourite/near certain, 100% = full certain. As can be seen the biggest change since our March poll has been in Upper Bann which has dropped from an 80% score for the DUP to 65%. This seems to be mostly due to the strong challenge that the UUP's Joanne Dobson has been mounting against the sitting MP the DUP's David Simpson. However, what should worry the unionists is that our models now show there is not just one alternative to the DUP, but two i.e. Sinn Fein and the UUP. Some comments from our Opinion Panel members in Upper Bann say that Catherine
Seeley the Sinn Fein candidate could 'come through the middle' if the DUP and UUP evenly split the Unionist vote.
Sinn Fein would need a very particular set of circumstances to occur, for them to win this seat, and the UUP say that the DUP's line that they're the only party that can stop Sinn Fein is scaremongering. But do Sinn Fein have a chance in this seat? - let's look at the figures: At the last Westminster (2010) and Assembly (2011) elections the total Unionist vote came in at around 55-60%, with the nationalist/republican vote on approx. 40%, and Alliance/others on 5%. The DUP scored 34% in 2010, with the UUP (called UNCNF then) on 26%. So if say the DUP drop only 4 % points, and the UUP gain 4 % points, and Sinn Fein get about 3/4 of the possible nationalist/republican vote (which is 30% i.e. 3/4 of 40%) then Sinn Fein could narrowly snatch this seat from the Unionists in a very tight finish between the three parties. Yes, all these factors would have to come into play for Sinn Fein to win - that's unlikely, but it could happen! Remember this is a First-Past-the-Post election and one more vote than your opponents is enough.
South Belfast is the other major change in our April poll with the SDLP jumping from a 65% score to 75% (i.e. a good favourite, but still not certain). Our Opinion Panel think that the Unionist vote is so split that as a panel member said 'the SDLP vote could drop to 9,000 and McDonnell could still win' - NB that would be 5,000 votes less than McDonnell scored in the 2010 general election. Interestingly Ukip seem to be scoring well in S. Belfast particularly in middle class areas and with the over 65's in our panel. The Sinn Fein candidate former Belfast Mayor Mairtin O'Muilleor is mounting a strong challenge for the nationalist vote but South Belfast is one of the last seats where the SDLP are still ahead of Sinn Fein and this should be in McDonnell's favour. However in our models you need to get to 80%+ to get comfortable, i.e. that you're going to win, 55% - 75% is borderline going to 'moderate favourite' - so at 75%, yes the SDLP are favourites, but they're by no means a certainty for this seat.
Other smaller changes are South Antrim which has been drifting downwards for the DUP for the past couple of months. They're still at the favourite level at 75% to hold this seat, but the trend seems to be against them, but maybe that trend is too late to stop William McCrea holding this seat for the DUP. Fermanagh and South Tyrone is heading back in Sinn Fein's direction again with a 5% gain from our March panel poll to an 80% score. Our panel think the Sinn Fein machine will be too much for the UUP to overcome on election day.
East Belfast has stayed at a 75% score for the DUP, the same score as in our March poll. However our panel noted that Naomi Long is putting up 'a heck of a fight' and remember this DUP score of 75% is with a Unionist pact - in this context, it's surprising that the DUP candidate Gavin Robinson isn't hitting 90% with our models. However, our Opinion Panel think that with the Unionist pact, beating Robinson is too big a mountain for Alliance to overcome - we'll see!
You may think it strange that South Belfast is the same score for the SDLP's Alasdair McDonnell at 75%, as it is with East Belfast and the DUP's Gavin Robinson. This is because of the incumbency weighting built-in to our models, which is probably keeping Gavin Robinson from hitting 80-90% because he isn't the sitting MP, and boosting McDonnell in South Belfast because he is the sitting MP. Incumbency is always an advantage at First-Past-The-Post Westminster elections.
As our Forecast table shows, out of the eighteen Westminster seats there are only eight that are 'in play' i.e. have winning probabilities for a particular party at less than 100%. Ten seats are at 100% certainties e.g. West Belfast, North Antrim, LaganValley etc. Apart from the seats we've covered the other three 'non-100%' seats are Foyle and South Down, which have shown little change from around the 90-95% probability scores for the SDLP to hold both of these seats, and North Belfast where the DUP's Nigel Dodds is on 95% to hold this seat for the DUP.
So that's it for our Opinion Panel and the 2015 General election. Not long to go now to see if the thoughts and views of our Opinion Panel over the last four months reflect what the voters actually say on election day!
Bill White is Managing Director of Belfast polling and market research company LucidTalk, polling partners to the Belfast Telegraph. You can follow LucidTalk on Twitter: @LucidTalk.