The results of LucidTalk’s final poll of 2021 have been released, and the main headline is the steadying of the DUP’s ship. The party is rebounding after a tumultuous summer that saw it oust two leaders. 18% of voters say they would give the DUP their first preference – down slightly on the 20% level of support reported by last month’s University of Liverpool poll, but much improved compared to where the party found itself in August’s LucidTalk tracker poll (13%).

The changes in support for the SDLP and Sinn Féin are well within the margin for error, so basically negligible. Movements of 1% aren’t newsworthy unless they’re part of a clear pattern. Instead it’s usually better to look at polling averages rather than one individual poll. It’s also better to do that if the polls were conducted at a similar time, but in Northern Ireland we don’t get many polls.

So, I’ve collected all LucidTalk’s polls from this year and last month’s poll from the University of Liverpool (reported on The View). The latter offered an even brighter ray of light for the DUP, appearing to claw back support that had drifted to the TUV. It also made for slightly better reading for the Alliance Party, although both Alliance and the TUV would make significant gains if even their poorest polling performances from this year are replicated in the next Assembly election. Four out of the five most recent polls have placed Alliance in either third or joint second. 

If the DUP rebound any further, they stand a reasonable chance of finishing as the largest party – especially if the TUV’s support fails to materialise outside of North Antrim.

The latest LucidTalk poll found that 30% of unionist voters said they would be willing to change their first preference party in order to prevent Sinn Féin from emerging as the largest party. If many of those voters already support the DUP – the only party that pushes that line – then this won’t have much of an impact. 

The SDLP has managed to stay firmly within the margin of error of their 2017 election result whilst Sinn Féin has ticked down a couple of points. The UUP is up a fraction on its 2017 performance.

How could this look in key constituencies

It’s too early to make accurate seat projections so I won’t do one. We can expect some regional variation in the changes to TUV, DUP and Alliance support which will have a big impact in the election itself, so I’ll just mention where the parties are looking at possible gains and losses. 

There’s one selection fight to keep an eye on as it could get very messy, which is the DUP in Lagan Valley. It’s a DUP two-seater that’s home to Paul Givan, Edwin Poots and Jeffrey Donaldson (the constituency’s MP).

Donaldson has said he wants to be an MLA rather than an MP so it would be logical for him to attempt a swap with Givan or Poots, who may not want to move. A strong Alliance showing could then completely upset the DUP’s applecart by leaving them with just one Assembly seat. However, a more likely candidate to fall to an Alliance second seat is Pat Catney of the SDLP. 

The DUP lost most of their marginals in 2017, so they can afford to drop a couple of percentage points and hold relatively steady. However, these latest polls put them at risk in North Antrim (to the TUV), Newry & Armagh (UUP), and North Belfast (Alliance), amongst others. 

Sinn Féin will do very well to hold their fourth seat in West Belfast, but I’d have said that before the last election. They have quite a few marginal seats to shore up: their second in Foyle, third in Fermanagh/South Tyrone, third in West Tyrone and third in Newry & Armagh. All of these were close last time so any decline in their vote share puts all of them at risk. 

Interestingly, it’s the SDLP who is the most likely beneficiary in those seats despite holding steady, but they risk losing a number of seats like Upper Bann and their second in South Down (both to Alliance). It could ultimately end up with a similar seat number but with a very different team.

The UUP has a chance of gains in Newry & Armagh and Mid Ulster but otherwise its seats are amongst the safest of any party. However, East Antrim is at risk, and it’s also the furthest from easy pick-ups. 

The TUV will have high hopes for a second seat in North Antrim and won’t be too far away in East Antrim, if it can muster a candidate who can step out from Jim Allister’s shadow. 

Alliance will be hoping to net upwards of five new seats and none of its seats are in clear jeopardy if it can maintain its current level of polling. A few of its new seats will come at the expense of the SDLP and Sinn Féin in eastern constituencies like Lagan Valley, although it will be targeting unionist seats too, including East Antrim. It even has an outside chance of taking a second seat in Strangford (from the DUP) and one in West Tyrone (from Sinn Féin).

Sinn Féin looks set to win the most first preferences overall, but it’s very possible that the DUP could catch it in terms of seats due to unionist transfers and simply having more wriggle room in key constituencies. 

Views on the Protocol

As well as showing the current state of play among the parties, the latest LucidTalk poll also gives us a sense of current attitudes to the Protocol. Taking a different approach to other surveys, the poll asked respondents which of three options they preferred. The wording of the question pointed out that these are the only three options that have been on the table since Theresa May’s Backstop plan was voted down (May’s deal would have kept the whole of the UK aligned with the EU). The options were:

  1. Keep the Protocol as it is (supported by 29%);
  2. A ‘hard’ border on the island instead of the ‘Irish Sea border’ (supported by 30%);
  3. A United Ireland with EU membership (supported by 36%).

These results will likely be misinterpreted as a ‘border poll’ question, when it isn’t really as some pro-Protocol supporters would likely support a united Ireland if it came down to a binary vote on staying in the UK or joining a united Ireland. However, it does at least show a stark divide over the Protocol itself and the strength of feeling against it from a minority of voters, with 30% willing to erect a physical barrier on the island – which would likely tear up the existing trade deal with the EU and prevent one being agreed with the United States. 

The findings are an example of the Condorcet Paradox: you can reasonably argue that any option is the most preferred option depending on if and how you count second preferences.

On polling in general

In general, Northern Ireland is very under-polled, so getting two separate polls this month was a comparative goldmine. However, every poll we get is now immediately jumped on and attacked by supporters of whichever party went down. This doesn’t exactly encourage new pollsters to enter the field, especially since criticisms often come back to sampling. They’re measures of one point in time and pollsters make adjustments (called ‘weighting’) to help account for sampling issues and ensure that the sample is as representative as possible of likely voters.

However, you can’t adjust polls based on opinions (like unionist/nationalist/neither) because these opinions can change – regardless of how definitive your personal opinion is. If they did, the polls would be way off on election day and then we’d complain that the polls were wrong.

Until then, let’s hope that we do get quite a few more polls – treating each with appropriate caution, but equally not simply dismissing them outright if we don’t happen to like what they say.

Technical details of the LucidTalk poll:

Polling was carried out online from 5 to 8 November 2021. The survey targeted the estab­lished Northern Ireland Lucid Talk online Opinion Panel. Data were weighted by age, sex, socio-economic group, previous voting patterns, constituency, constitutional position, party support, and religious affiliation, with a weighted sample of 3,298 used in the analysis.

All aggregate results are accurate to a margin of error of +/-2.3%, at 95% confidence.