Holding a border poll in Northern Ireland: when does it need to happen and what questions need to be answered?


The prospect of a poll in Northern Ireland about Irish unification, provided for by the Good Friday Agreement and often termed a ‘border poll’, is now widely discussed. But the provisions and wider implications of the law and the Agreement are little explored. The Constitution Unit is considering a project to examine this, and Alan Whysall here gives an overview of the key questions.

Support for a united Ireland appears to be rising. There is little to suggest a majority for unity now, but in the context of Brexit provoking serious strains it might arise. This blog is mainly about process. But the real world risks are high. An early poll, particularly if it takes place in a political atmosphere that is strained following a hard Brexit, could seriously destabilise both parts of Ireland, and put at risk the political gains of recent decades.

Current outlook on border polls

Northern Ireland Unionists have largely ignored or dismissed the prospect of a poll. But the former First Minister Peter Robinson last year urged unionism to prepare.

Nationalists, while looking forward to a poll, have often been vague as to when this might happen. Sinn Féin now appears to favour one immediately after a no deal Brexit. The SDLP propose there should first be a forum to establish the shape of a united Ireland.

The Irish government has been hesitant. The Taoiseach, Leo Varadkar, has suggested that raising the prospect now is disruptive and destructive, and has in the past questioned the wisdom of Irish unity founded on a 50% plus one vote in Northern Ireland.

The UK government has consistently rejected ideas of any early poll. But during recent debate on a no deal Brexit, leaks have emerged of its apparent fears that such an outcome would trigger a poll, dismissed by unionists as ‘Project Fear’.

Recent surveys on Northern Ireland appear to show a marked trend towards a united Ireland. None yet suggests an overall majority, but polling last September suggested 52% of people there would vote in favour in the event of Brexit. However different surveys produce sharply different results and the accuracy of some polling methodologies is questioned. Indeed opinion polling in Northern Ireland has for long thrown up particular problems.

How we got here

In 1973, the UK government called a poll in Northern Ireland on whether to maintain the Union or join a united Ireland. The poll was the subject of a nationalist boycott: 98.9% of those voting, 57.5% of the electorate, voted for the union.

The current provisions for polls are part of an interconnecting, but asymmetric and incomplete, framework established in the Good Friday Agreement and the international treaty accompanying it, and in UK and Irish law enacted in consequence.

Section 1 of the Northern Ireland Act 1998 guarantees that Northern Ireland will not leave the UK without the consent of a majority of the people of Northern Ireland voting in a poll. It gives the Secretary of State for Northern Ireland a general power to call one; with a duty to do so at any time it appears likely that a majority would favour a united Ireland. There is a seven-year minimum interval between polls. If there is a vote for a united Ireland, proposals to give effect to that wish, agreed between the UK and Irish governments, must be put to parliament.

The Good Friday Agreement recognises ‘that it is for the people of the island of Ireland alone, by agreement between the two parts respectively and without external impediment, to exercise their right of self-determination on the basis of consent, freely and concurrently given, North and South, to bring about a united Ireland’. The consent principle is recognised in the new Article 3 of the Irish Constitution.

An incomplete framework

These provisions were little debated at the time of the Agreement: the parties coalesced around the broad approach to consent, without significant probing of the outworkings. So vast cans of worms remained unopened, probably intentionally.

Some things are clear. There must be a poll if the Secretary of State thinks it likely that a majority would vote for a united Ireland. In principle she could be held to this duty as a matter of law. But there is no guidance on forming a judgement

  • A clear majority in a succession of reputable opinion polls may seem the most obvious trigger mechanism. But judging what is a sound poll is not straightforward.
  • A majority of members in the Northern Ireland Assembly, or the general election, from nationalist parties, would be no more than a pointer. A vote by a majority in the Assembly would be more persuasive.

The Secretary of State’s order for the poll would decide the franchise. Some would favour 16 to 18 year olds voting, as in the Scottish referendum on independence – this would likely strengthen the unity vote.

Once it was decided to conduct a poll, the Electoral Commission would assume certain responsibilities – including on the terms of the question.

Besides the stipulation for minimum intervals, there is nothing in the 1998 Act or the Agreement to say at what point the poll itself should be held. Allowing a long period, perhaps several years, in which to frame the questions may be legitimate.

The 1998 Act and the Agreement are cast in terms of a simple majority vote in a Northern Ireland poll, and with no minimum turnout threshold: 50% + 1 suffices.

And the choice is a binary one, between remaining in the UK and a united Ireland. Other approaches previously canvassed, for example joint authority or a phased progression to a united Ireland, are not acknowledged.

There must also be a poll in the south of Ireland, and the Agreement speaks of it being concurrent with that in the North. Specific provision for the terms of such a poll would be for the Irish Parliament and Government.

The form of a united Ireland would then have to be negotiated. Would there are also be post negotiation polls? There is no specific provision. But there would be a need for substantial changes to the Irish Constitution, which require a referendum. No northern poll is provided for at the later stage. Is such asymmetry acceptable?

What would a united Ireland look like?

Perhaps the starkest omissions from the Agreement relate to the shape of the united Ireland and how it is to be determined. A few guarantees and principles set out in the Agreement are expressed to be of enduring validity. The European Council has formally declared that the whole of a united Ireland would be in the EU. Beyond that, little is stated.

The 1998 Act speaks of the UK and Irish governments developing proposals to give effect to vote for a united Ireland – but they alone could do little more than outline plans for further process. Much of this is the stuff of constitutional conventions. But the process is unspecified.

Issues to be addressed

This is not the place to develop an agenda for negotiations, but is important to have some idea of their magnitude. Obvious issues may include:

  • The governance of Northern Ireland within a united Ireland: is it separately administered, or does it become part of a unitary state? If the former, do the principles of the Agreement form the starting point, indeed do the existing institutions remain, interacting with Dublin rather than Westminster? Does the whole corpus of UK law carry over? Does that extend to expensive entitlements not matched in the South, like free NHS treatment?
  • The wider Irish context: does the 1937 Constitution continue, with minimum necessary modifications? Or is there some more comprehensive revision, even a new constitution, with an input from the North?
  • Wider relations: does the UK government have any continuing place in Irish arrangements, safeguarding the rights of those who continue to consider themselves UK? Does London continue to fund Northern Ireland’s substantial public sector deficit for a period? Would Ireland rejoin the Commonwealth? Might there be political and practical value in thickening East-West ties between the UK and Ireland?
  • Transitional arrangements. The breadth and abruptness of the changes to be made provoke thoughts of phasing the progression to Irish unity, to permit the working up of new constitutional arrangements, and the removal of the sharp edges.

Principles and realities

How do these questions play out in the real world? The binary nature of the Agreement provision makes the basic proposition a dramatic one: the movement of almost a million people, essentially traditional Unionists, into a united Ireland against their preference. The opposite outcome is in principle even harder to defend as a matter of principle – keeping Northern Ireland within the UK against the majority’s wish. But abrupt change may entail seismic shock.

Many, including probably the two governments, will be inclined to play things long in the interests of stability. But they may not fully control the process. The overall climate may also be conditioned by factors beyond the immediate debate, including:

  • continuing negotiation over Brexit and its consequences – a hard Brexit, and painful negotiations about future EU relations, may leave a brittle and dangerous political landscape;
  • whether it has been possible to resume devolved government in Northern Ireland – potentially a source of stability, because the political parties would once again have to shoulder responsibility;
  • any moves to Scottish independence – influences may be felt in both directions.

Is an early poll desirable?

It is unlikely the Secretary of State at present would call a poll unless obliged to do so. The UK government has traditionally seen a poll as a distraction from the pressing issues of government and society in Northern Ireland, that might interfere with the smooth operation of the Executive, and that it was in any event a foregone conclusion, because there was clearly no majority for change. These arguments are now evidently less valid.

But a poll would risk draining the remaining energy and will from efforts to resume the Executive. Any UK government, like its Irish counterpart, would fear the consequences for stability of a narrow vote for unity. A narrow negative might mean that the focus of politics would change from making things work in Northern Ireland towards seven-yearly replays.

Some nationalists will continue to see an early poll as the longed for opportunity. But many within nationalism, including most probable future Irish governments, will continue to doubt the merit of a poll in current circumstances, and the potential disarray that might ensue from a vote in favour.

It is also possible that a southern poll would reject unification, even when the northern one favoured it, the voters there believing the admission of so many recalcitrant Unionists would bring more disruption than they wish to contemplate. Such a result might seriously shake both parts of Ireland.

What might the governments do?

The Secretary of State might in various ways seek to engage others in informing her judgement on a poll. There might usefully be a process of consultation either to establish the state of opinion, or to propose procedures for doing so. She would clearly need to engage with the Irish government, which would have its own polls to organise.

And she might legitimately say that she would be assisted by the considered views of elected representatives expressed through the Assembly. More enlightenment is likely to emerge when the parties are talking to each other within the institutions, rather than through megaphones. And such an approach would deter the parties from letting the prospect of a border poll inhibit a return to devolved government.

If the duty to call a poll is engaged

In various contexts, though, perhaps particularly that of a variant of Brexit that provoked acute discord, calling a poll may become a legal duty. Foot dragging would appear to be a repudiation of one of the key pillars of the Agreement. At that point, though, it may be a widely held view that a period of several years was needed before the poll, in order to seek to resolve some of the questions above before it was conducted.

A benign dynamic might set in. If a poll is imminent, parties on both sides of the community ought to feel themselves under an obligation to appeal not only to their own side, but also to the other. In the course of this, Unionist and nationalist “packages” might emerge. And it is possible a wide measure of agreement would develop about staging of any eventual move to Irish unity, with the aim of reducing immediate shocks and their potential for division.

But in the increasingly polarised atmosphere of Northern Ireland politics such constructive outcomes are not to be counted on – especially with the UK and Irish governments, whose partnership has been the motor and the guardian of the process in the past, less able to work effectively together.

The prospect arises of the whole process set out above being triggered at the time of great tension with key questions unanswered. Potentially a second overlay of chaos is added to one stemming from the consequences of a difficult Brexit. It is hard to be confident given the history that all in the community will adjust to this with perfect serenity.


Until a few years ago, these issues appeared to be resolving themselves –the border no longer mattered for most purposes, and everyday politics, though profoundly imperfect, was beginning to get a foothold. The border poll increasingly seemed a form of, as it might be termed, backstop that either would never happen, or if it did, might not cause great upheaval.

Recent political developments mean that it may rapidly come to life. It is, perhaps, the least indefensible of all the alternatives, but it is a brutal and indiscriminate mechanism, and, triggered without preparation, may have many casualties.

There is a need for reflection and preparation; to establish what the process might be, and then to find ways in which, on the way to the vote and beyond, the aspirations of as large a part of the community as possible are accommodated and the worst ill effects are avoided.

The Independent Commission on Referendums, sponsored by the Constitution Unit, laid heavy emphasis on the need for such preparation. But it requires serious consideration, among the parties, and indeed society, in Northern Ireland; in the south; and certainly in London.

The Constitution Unit is considering a project to pursue further some of the questions about the process raised in this post, building on the work of the Independent Commission on Referendums. Alan Whysall is preparing a background paper, which the Unit will publish later this month. All thoughts will be particularly welcome.

About the author

Alan Whysall is a former senior civil servant who has worked on Northern Ireland for most of the last 35 years. In his retirement he has become an Honorary Senior Research Associate at The Constitution Unit.

2 thoughts on “Holding a border poll in Northern Ireland: when does it need to happen and what questions need to be answered?

  1. Pingback: Post-election negotiations in Northern Ireland must set the Belfast Agreement on a firmer footing and re-establish constructive politics | The Constitution Unit Blog

  2. Pingback: Renewing and reviving the Belfast/Good Friday Agreement | The Constitution Unit Blog

Comments are closed.