While the Secretary of State indulges in some hard dithering on when (or if?) to call a new Assembly election, the parties are already deep into election planning mode. Indeed, it might be said that an unofficial election campaign is already well underway.
As the talks with the SoS continue, each party’s stance may be influenced by how they view their prospects in a coming election.
And when the election is finally announced, a key question for each party will be how to best use its limited resources of money and manpower. Which constituencies should receive extra focus? And which can afford to send some of their activists to canvass elsewhere? Get that right and you may be able to boost your seat tally. Get it wrong, and you could throw one or two seats away.
Unfortunately for party planners, they are to some extent operating in the dark. Apart from any conversations they have recently had with voters on the doorsteps, they have no recent opinion polling to go on. The last Lucid Talk poll was in August and the next is not due for nearly two weeks, by which time the campaign could already be in full swing.
So they will scour the August poll for clues. As well as the headline voting intention shares published in the Belfast Telegraph at the time, they will have the more detailed data published on Lucid Talk’s website about a week later. These will have given them party shares to one decimal place and a table measuring how voters were switching between parties since the May election.
So what conclusions will each party be likely to have drawn from the data?
I have constructed two models of how the party shares could translate into votes in individual constituencies. One is based on vote switching – how each party’s May voters intended to vote in August. The other increases or decreases a party’s share in proportion to the change in its total NI-wide vote share. Between them this produces a high and low projection of party shares in each constituency.
A third set of outcomes is produced by creating a combined average for each constituency.
Sinn Féin won 27 seats last May. LT poll 29.6%. (Up 0.5% on May election.)
26 seats are rock solid safe according to two of the models, with the third seat in FST all but safe – not a major concern for any party in that position but not to be taken for granted.
The model based on Vote Switching is the least kind to SF. It suggests that the 3rd West Tyrone seat could be under significant threat. It gives the party three low chances for gains (EL, Foyle, MU and UB) plus one very low chance (EA).
The Share Change model shows MU and N&A as moderate chances for gains, with 4 others at a very low chance (EL, SD, UB and WT).
The Combined Average model shows a moderate chance for gains in EL and MU; low chances for Foyle and N&A, and very low for SD and UB.
Summary: Worst result – 25 seats. Most likely – 27/28. Best – 29/30
Targets in likely priority order: MU, EL, N&A, Foyle, UB
DUP won 25 seats last May. LT poll 23.6%. (Up 2.2%)
23 seats rock solid safe according to the Share Change and Vote Shift models. Vote Shift has Foyle slightly below safe, but the 2nd UB at medium risk of loss.
Share Change has EL all but safe, with Foyle at medium risk of loss.
The Average Combined shows 24 as safe, but again Foyle at medium risk.
All three models have North Antrim as a possible gain, at a low chance in two and medium in the third.
Two have North Down as a low chance gain from Independent. However, opinion polls cannot give reliable results for isolated independents since the sample cannot be big enough. Given the size of Easton’s vote I see little prospect of him losing his seat.
One model throws up the DUP’s perennial hope of a gain in W Belfast but at a very low level of chance.
Summary: Worst result – 23. Most likely 25. Best 26.
Targets in priority order: NA, ND, WB
Alliance won 17 in May. LT poll 16.4% (Up 2.9%)
14 are safe under the Vote Switch model, which places the 2nd EA as all but safe and UB just below safe. The other two models have the 2nd EA rock solid safe for a total of 15. One makes UB all but safe while the other puts it just below.
All three models have North Antrim as at medium risk of loss.
All three also agree on 4 possible gains – WT (two mediums and a low), Foyle (a medium, a low and a very low), N&A (a medium and two very lows) and FST (three very lows).
EL appears as a potential gain on two models, at high and very high levels. Also on two models are MU (low and very low) and W Belfast (both very low).
Summary: Worst result – 14. Most likely 17. Best 18/19.
Targets in priority order: EL, WT, Foyle, N&A.
UUP won 9 seats in May. LT poll 11.0% (Down 0.2%)
The three models agree that the UUP has 8 rock solid safe seats. One places UB at all but safe, a second at just under safe, while the third has it as a medium risk of loss. It is likely that the exposure given to the party leader during the campaign would reduce the risk.
Two models give a chance of a gain in Foyle – one at medium the other at low level. The third does not show a chance.
N&A appears as a very low chance of gain on one model and EL on another. (This last is based on the poor showing of independents in the LT poll – again a reminder that poll samples are not large enough to measure the performance of individual independents.)
Summary: Worst result – 8. Most likely – 9. Best – 10.
Targets in priority order: Foyle.
SDLP won 8 seats in May. LT poll 6.6% (Down 2.4%)
Two of the models give the party 2 rock solid safe seats – S Belfast and Foyle. Both place S Down as all but safe, while the third has it rock solid safe. The 2nd Foyle seat is deemed all but safe by one – but at high and medium risk by the other two.
All three models put MU and N&A at a medium chance of loss. Two of them also place WT at medium risk, while the third rates it as low.
East Londonderry is rated as a certain loss by all three models.
Summary: Worst result – 2. Most likely 4/5. Best – 7
Targets: None
TUV – won 1 seat in May. LT poll 6.0% (Down 1.6%)
This is rock solid safe with two models. The Vote Shift model reduces it to just below safe.
There are no potential gains.
Summary: Worst result – 1. Most likely – 1. Best -1.
Targets: None.
PBPA – won 1 seat in May. LT poll 1.1% (Down 0.1%)
All three models agree that the seat is as good as safe.
There are no potential gains.
Summary: Worst result – 1. Most likely – 1. Best -1.
Targets: None.
Independent unionists – won 2 seats in May. LT lumps them with PUP and Conservatives at 1.2% (Down 0.9%)
As already stated, a NI wide poll cannot give guidance on the likely performance of individual independents.
For the record in ND Easton ranges from rock solid safe in one model, low risk in second and medium risk in the third. Sugden emerges rock solid safe in one model, and at medium risk in the other two.
In this case previous results are a better guide than the opinion poll. Easton scored 1.4 quotas in May and is most unlikely to have lost half of his vote in the meantime. Sugden won with 0.7 of a quota in both 2016 and 2017. However, her first preferences dropped to just 0.54 of a quota in May which puts her firmly into danger territory.
Summary: Worst result 1 (Easton). Likely 1/2. Best 2.
Greens – won 0 seats in May. LT poll 2.5% (Up 0.6%)
None of the models shows a potential Green gain.
Of course, all of this could look different in the next Lucid Talk poll. And could be different again when, or if, the election actually takes place.
Michael Hehir is a retired sales and marketing manager. He studied in Northern Ireland but now lives between England and Italy.
