An extraordinary amount of unionism’s current malaise stems from the DUP panicking over a LucidTalk poll in February 2021, which showed a shock jump in support for the TUV from 4 to 10 per cent.
The DUP should have held its nerve and asked if this was a false, freak or transient result. Instead, it immediately withdrew its reluctant acceptance of the Brexit protocol and went into a tailspin from which it has never recovered.
Now the party is making the same mistake over immigration.
A LucidTalk poll published in Monday’s Belfast Telegraph found unionists are six times likelier than nationalists and Alliance voters to think immigration in Northern Ireland is too high.
The findings were 82 per cent for all unionists and 88 per cent for DUP voters, compared to 13 and 10 per cent for nationalist and Alliance voters respectively.
The DUP has responded by saying “there are deeply held concerns held by many people” and calling for Northern Ireland not to become “a target” for illegal immigrants.
It has tried to soften this by saying the UK benefits from legal immigration and “everyone will know the importance that overseas workers play in our NHS”. However, it is already on the defensive against a harder line from the TUV, which is linking immigration to the protocol due to different interpretations of human rights law either side of the Irish Sea.
Once again, the DUP should have held its nerve and asked why this poll differs so sharply from the latest Northern Ireland Life and Times (NILT) survey, the only other regular source of polling information in Northern Ireland. It found Protestant and Catholic attitudes to immigration are almost identical and overwhelmingly positive.
NILT, a long-running project by Queen’s University Belfast and Ulster University, consistently finds social attitudes to be broadly similar across both main communities – the first reason anyone should wonder why opinions might drastically diverge.
LucidTalk’s poll was conducted after this month’s race riots, while the NILT fieldwork was performed between last September and January, so part of the difference must be due to events. This could abate as quickly as it arose: nationwide rioting has an aspect of mass hysteria.
Each survey asked different questions, requiring careful comparison. LucidTalk simply asked if immigration is too high, too low or “about right”. NILT asked a complex set of 63 inter-related questions, such as “are refugees and asylum seekers respected in Northern Ireland?” and “are people’s perceptions of migrant workers tainted by the media?”
Some questions appeared to invite positive responses, which is no credit to NILT’s method but does show the potential for political leadership on the issue. The public is apparently amenable to positive suggestion.
NILT surveys the entire adult population, while LucidTalk surveys voters in previous elections. This is why LucidTalk has a good track record predicting elections but it is not a public attitudes survey, nor does it claim to be. If the DUP believes only voters count, it should count how many people from a unionist background have been put off voting altogether.
The main methodological difference between both surveys is that NILT approaches people for a response, while LucidTalk issues an open invitation to join its online polling panel, which has over 13,000 members.
- Some basic communication would help address immigration concerns – Newton EmersonOpens in new window
- Brian Feeney: Race riots are an English – and loyalist – problem and political leadership is requiredOpens in new window
- Working class protestant communities have been let down by their supposed leaders - Chris DonnellyOpens in new window
The company has developed a formula to weigh respondents by age, gender and socio-economic status but the one thing it cannot adjust for is that everyone on its panel is the sort of person who volunteers for an online polling panel. They are sometimes described as the politically engaged or the extremely online. You might think of other terms for them – I am in no position to judge.
Monday’s result looks like an artefact of this feature, to an almost absurd extent.
Unionist panellists may have been unusually swayed by anti-immigration content on the internet, while nationalists may have been reacting to what they see as a unionist prejudice. That could also partly explain the north-south discrepancy in nationalist attitudes. Southerners polled on immigration have no unionist neighbours to react against.
TUV support is another possible artefact of a staunch panel. It is impossible to know if LucidTalk was correct in 2021 because the DUP response created its own reality. Pandering to TUV politics boosted the TUV’s credibility and increased its electoral strength, although still not to predicted levels.
Chasing an anti-immigrant vote will likewise become a self-fulfilling prophecy.
If the DUP wants to lead unionism it should show leadership and stop following one online poll.
