Every so often a politician will say that the time is right for a Border poll, triggering a flurry of debate, opposition and support.
The pattern was repeated recently by British prime minister Andy Burnham, who rejected it saying: “That is one issue that will be off the table,” and then US president Donald Trump saying it would be a “great thing”.
Yet British, Irish and American politicians of a different era anticipated this behaviour.
Those who masterminded the 1998 Belfast Agreement signed on Good Friday that year stipulated that “the Secretary of State shall exercise the power ... if at any time it appears likely to him that a majority of those voting would express a wish that Northern Ireland should cease to be part of the United Kingdom and form part of a united Ireland.”
The decision is ultimately down to the “wish” of the people of Northern Ireland.
But how exactly is that measured?
Demographic proxies
Northern Ireland is changing and demographic data reflects that.
In politics, the gap between unionists and nationalists has narrowed to one percentage point, according to an analysis of Assembly election party voting patterns over the past 50 years.
In religion, similar to the republican vote, the percentage of Catholics in Northern Ireland has increased incrementally over time, becoming the dominant religion in the 2011 and 2021 censuses, while those who state “Protestant” as their current religion has declined dramatically.
However, most of this decline among the Protestant unionist loyalist (PUL) community appears to have shifted into the “neithers” categories – the don’t knows and unsures – a religiously agnostic and politically complex group that reflect the atypical views of an emerging generation of “ceasefire babies”.
This makes it harder to use identity markers as a proxy for a Border poll vote.
The truest measure of how the Northern Irish public would feel about the question of a Border poll is simple: just ask them.
Polling the public
Multiple pollsters have posed the question to the public over the last 50 years.
An amalgamation of polls over the past five decades shows the gap between unionists and nationalists has narrowed significantly.
Most of these polls, going back to the 1980s, were collated by the Working Group on Unification Referendums on the Island of Ireland, a cross-community group of academics who examined the question of how best to execute a referendum. Data on polls since 2021 has been updated by The Irish Times.
Including the latest polls, 48 per cent of the public favoured remaining in the UK when polls were averaged over the previous three years.
Polls show 35 per cent of the public in Northern Ireland were in favour of a united Ireland. A further 17 per cent said they didn’t know or were not sure.
As it stands, just over a third of the Northern Irish public are in favour of joining a united Ireland, suggesting that it does not appear likely that the threshold for a majority has been reached.
Still, the evidence shows that since the UK’s Brexit vote in 2016, a substantial shift has been recorded.
Unionist support is no longer in the majority, falling below 50 per cent in recent years. There is also an increasing “don’t know” group.
Bill White, director of pollster LucidTalk, says all three results are important and have to be considered equally.
“A key score is the pro-United Kingdom result. If that drifts down to 45 per cent that could indicate 55 per cent are pro a united Ireland or could be if the ‘don’t knows’ sway,” he says.
“These are all subjective judgments for the Secretary of State.”
Paula Devine, director of the annual Northern Ireland Life and Times Survey, says taking into account the “don’t knows” is important but harder to decipher.
“The don’t knows are very important in surveys but there’s no ‘don’t know’ option in a referendum,” she says.
“That’s what makes it difficult to predict, to translate to a vote.”
[ Andy Burnham raises prospect of UK rejoining European Union after next electionOpens in new window ]
However, Katy Hayward, a professor of political sociology at Queen’s University Belfast and a member of the working group that collated the polling, says some polling data points towards a unionist sway among “don’t knows”.
“A lot of the ‘don’t knows’, when pushed, shift toward what they already know: remaining in the United Kingdom,” she said.
Post-Brexit, people are especially “more nervous of change”, she adds.
Projecting forward
In the last 10 years, outright support for a united Ireland rose 11.4 percentage points, according to a rolling three-year average since 2016.
A simple projection forward would suggest that if that pace of change were to continue, rising 1.14 percentage points per year, outright pro-unification support would reach over 50 per cent by the year 2039. Presuming a year-long campaign in the run-up to a ballot, a Border poll could then take place in the year 2040.
However, to project based on the shifts of the post-Brexit decade is to presume public opinion follows a linear path. It can accelerate, reverse, fluctuate or even flatline.
If we were to look back to the 2010s, similar shifts in the opposite direction could be observed in the years after the recession as Ireland’s economy suffered significantly.
It also presumes that outright support for a united Ireland is the primary metric; a decline in explicit support for the union with Britain and an increasing number of “don’t knows” could also be a factor.
[ Trump weighs diesel export ban as US fuel crisis deepensOpens in new window ]
However, if support for the union with Britain were to continue to decline at the same rate as support for a united Ireland is rising, pro-unification support would surpass unionist support in the early 2030s, presuming “don’t knows” stay roughly the same. This could serve as an alternative trigger for a Border poll
Even if the Secretary of State does require a majority of the public to state explicitly that they want to join a united Ireland in polls, the Belfast Agreement also stipulates it must “appear likely”.
So how much of a majority is “likely”?
“It could be argued that opinion polls will have to show significantly more than 50 per cent support for a united Ireland,” says White of pollster LucidTalk.
A simple majority may need to be sustained over a number of years to assure the Secretary of State to plan a Border poll referendum.
They could even wait for a super majority of 60 per cent (projected on the current path for the late 2040s) or 70 per cent (in the mid-2050s) to ensure a referendum is not won or lost on a knife’s edge.
Yet to delay a vote until a wider majority polls in favour could anger nationalists who feel the promises of the Belfast Agreement are not being honoured.
“It would be unfair to require one group to be in a supermajority to have their constitutional preference filled while another can be in a minority and maintain the status quo,” says Hayward.
[ Drumcree row is a useful smokescreen for cynical Sinn Féin and DUPOpens in new window ]
It may also be in the nationalist interest to be cautious – too narrow an edge in polls that don’t materialise at the ballot box could scupper their hopes for almost a decade – as the Belfast Agreement stipulates once a referendum happens it cannot be repeated for seven years.
“The pro-united Ireland camp need to get a result that is at least 40 per cent to 50 per cent in an initial referendum, then they could try again after seven years,” says White.
“But if they lost by a lot – say a result less than 40 per cent – it would be disastrous for them.
“There wouldn’t be a lot of appetite for a repeat vote within seven years as it would be used to argue it was ‘not likely’ to change that dramatically.”
A larger majority in the polls could also account for the wide margin of error in polling on hypothetical questions compared to making a decision at the ballot box.
“A NI Border referendum will be very dependent on the campaign. What is a united Ireland? Nobody knows yet,” says White.
“A lot of this depends on how the British and Irish establishment work within their campaign.
“The campaign in Scotland changed everything – started at 25 per cent pro-independence then ended up 45 per cent. You’ll have to argue about the economics, about the health service and education. People have to experience that campaign. ”
“Polling on this topic at is like the long-range weather forecast”
— Bill White
None of this makes projecting the public’s wish easy.
“Polling on this topic at the moment is like the long-range weather forecast,” White says.
Yet the stakes are high.
The Secretary of State also needs to worry about preventing a potential boycott of a Border poll if unionist voters feel alienated.
Such an outcome may sound outlandish but it played out in 1973 when the British government held a Border poll to try to quell the rising violence of the Troubles.
This resulted in nationalists boycotting the referendum they knew they had no chance of winning at the time.
Ballots were burnt on bonfires, polling stations were raided and one British soldier was killed.
The final vote was 591,820 in favour of remaining in the UK and 6,463 opposed, in an electorate of more than one million people.
All of these data questions will need to be considered by the Secretary of State, who at the same time must contend with politicians, including prime ministers and presidents offering public views on the topic.
Yet, the roadmap ahead is laid out in the Belfast Agreement and in the hands of one individual.
The agreement also stipulates that consent for a united Ireland must be “freely and concurrently given, North and South”, which is generally accepted to mean that referendums must be held on both sides of the Border at roughly the same time.
So the role of the Northern Ireland Secretary of State remains pivotal.
“There’s a whole range of data,” White says. “But the Secretary of State over a long period of time has to assess the realpolitik.”
