A bit like buses we had several months with no polling in Northern Ireland on the border referendum issue, and then several came along all around the same time: from Queen’s University Belfast (polling by Ipsos-MORI), the BBC (poll by LucidTalk), the NI Life and Times survey, and Lord Ashcroft Polling.

All four polls asked the key questions: Should there be a ‘border poll’, and what way would you vote if there was one? Just looking at the one key question – What way would you vote in a border referendum? – the results came out as follows:

At first glance there seems to be large differences in these poll results – but are there really? Yes, both the NI Life and Times Survey and QUB/Ipsos-MORI research showed much lower levels of support for a united Ireland than the BBC/LucidTalk and Lord Ashcroft polls. But they don’t show high support for the unionist side either.

Support for Northern Ireland remaining in the UK was 55% (NI Life and Times Survey), 50% (QUB/Ipsos-MORI), with the Lord Ashcroft poll coming in at 49%, and the BBC poll (carried out by LucidTalk) 45% pro-NI-in-UK. So in terms of the pro-union/UK scores, all polls are in the same ballpark, averaging around the 50% mark.

However, the Lord Ashcroft and BBC polls showed the possible border poll results to be much closer than the QUB and Life and Times Survey results. Why might this be the case?

Different methodologies

The differences in the poll results, particularly the large difference in the pro-united Ireland scores, may be down to the different methodologies used for the individual polls. The QUB and NILT projects used direct face-to-face interviews for their research, whereas the BBC poll (conducted by LucidTalk) and Lord Ashcroft poll primarily used an online research approach.

It’s widely accepted within the market research sector that, as a polling methodology, pure face-to-face polling doesn’t have the in-built anonymity that is applicable to online polls, and even (to a lesser extent) telephone polls. Therefore, face-to-face polls (like the two referenced here) usually get abnormally high levels of Don’t Knows/Wouldn’t Votes, because many respondents are reluctant to express their real views to strangers in a face-to-face environment. This is particularly true when controversial issues are being researched including the constitutional question, same-sex marriage, abortion rights, and so on.

The online methodology used by Lord Ashcroft polling, and ourselves at LucidTalk, tend to draw out more of the ‘shy’ voters about any issue – in this case, pro-united Ireland voters. This is because respondents are communicating with their own laptop, smartphone, etc., and thus are in an environment where they feel more comfortable to express their real views.

Plus, it should be noted that the BBC and Ashcroft Poll results are much more in line with Northern Ireland election results over the past 10-15 years. Indeed, if support for the pro-united Ireland side was as low as indicated in the two face-to-face surveys (QUB & Life and Times Survey) then it would mean over 150,000 regular Sinn Féin/SDLP voters would vote for Northern Ireland to stay in the UK in a real life border poll – which doesn’t pass a common sense test.

All the poll results attracted the expected amount of opinion and criticism. However, the criticisms and views from the unionist side were particularly strong, particularly regarding the results of the poll we at LucidTalk carried out for the BBC.

The Brexit factor

This unionist response showed an unrealistic view of the current makeup of opinion in Northern Ireland. Some unionists seem to think that the pro-NI-in-the-UK vote in such a referendum would be (or should be) around 65% to 70% – at least! We got lots of comments like, “Ah, the pro-Union vote would be much higher,” they say, “because all the current non-voters will come out to vote,” (really, would they?) and because “a large section of the Catholic community are happy for Northern Ireland to stay in the UK” (again, is that actually true after the EU Referendum?), and so on.

It’s undoubtedly the case that Brexit has really affected the possible results from a future border poll. After the EU Referendum, LucidTalk’s polling has detected two crucial changes in voting intentions in a hypothetical border referendum.

Firstly, up until the Brexit vote there was a notable minority of nationalists and republicans who were ‘sort of OK’ with the status quo (however, whether this group would actually have gone out to vote to stay in the UK is another question). This group fully supported the 1998 Good Friday Agreement, supported devolution, and were pro-EU. Yes, this group still had Irish nationalist aspirations, but they were prepared to wait for a united Ireland. However, with Brexit, this group has now shifted and become much more assertively pro-united Ireland.

Secondly (and similarly) the vast majority of the Alliance/Green/others voting block (which is currently around 11% of the Northern Ireland voter base) has been pro-NI-in-the-UK, but since the EU Referendum a section of this group has swung to the pro-united Ireland camp, mainly to ensure that Northern Ireland remains within the EU. Plus, based on Brexit, the number of ‘Don’t Knows/Not Sures’ within this group has also grown significantly, moving from their original pro-NI-in-the-UK position. Remember, the vast majority of the Alliance/Green/others voting block is very pro-EU.

A wake up for unionists

Based on this current landscape, unionists should realise that the 1950s and 1960s are gone. That was when unionists had a comfortable majority at Stormont and a unionist held every Westminster seat, usually with comfortable large majorities. For unionism to obtain a successful result in a border referendum – and realistically that has to be somewhere around the 60% pro-NI-in-the-UK mark, at least – they need to adopt a more inclusive agenda, and build a broader pro-unionist coalition.

Indeed, to get to that 60% score in a referendum, unionism needs to ensure they get the vast majority of the Alliance/Green/others block voting pro-NI-in-the-UK, and also a substantial chunk of the current ‘non-voters’ voting that way as well.

The last border poll in 1973 (45 years ago) was 57% pro-NI-in-UK, so the recent poll results – particularly from the BBC and Lord Ashcroft – fit in with what can be assumed to be sensible: they pass a realistic common sense test as to what is ‘felt to be right’ (which is always important when considering polls). This is even allowing for the huge demographic changes that have taken place in Northern Ireland in the past 45 years.

A border poll may still be a long way off, but it’s on the agenda, and on the horizon. However, if unionism wants to get a credible result in such a referendum, these recent poll results show they need to start thinking about how to build a broader pro-NI-in-the-UK coalition – and they need to start thinking fast.