Colin Murray: A Referendum on a United Ireland: Perspectives from UK Constitutional Law

Editors’ Note: This post is part of an IACL-AIDC Blog Symposium on unification of the island of Ireland. The Symposium is co-hosted by the Constitution Project @ UCC Blog. Selected posts will also be co-hosted by the UKCLA Blog. The full set of posts can be found here.

The Northern Ireland Act 1998, a statute passed by the Westminster Parliament, implements the Belfast/Good Friday Agreement (GFA) within UK law. It declares that Northern Ireland remains part of the United Kingdom and shall not cease to be so without the consent of a majority of the people of Northern Ireland voting in a referendum (the ‘principle of consent’). However, if a majority in such a vote favours a united Ireland, the UK Government must lay before Parliament any proposals to give effect to that outcome which it agrees with the Irish Government.

The Act also provides loose criteria for calling a so-called ‘border poll’. The Secretary of State has a general power to call a referendum on reunification at his discretion. This power is sometimes forgotten against the fraught backdrop of Brexit, but only a few years ago Arlene Foster, Northern Ireland’s current First Minister and leader of its largest Unionist party, the Democratic Unionist Party, was suggesting using this power to affirm its place in the UK. The new Secretary of State, Brandon Lewis, is however under an obligation to institute a referendum if at any time it appears likely to him that a majority of those voting would support a united Ireland.

These provisions were the subject of heated debate in the House of Commons on the evening of 22 July 1998. It was a night of Unionist unity. The Reverend Ian Paisley, the leading unionist opponent of the Agreement, proposed adding a requirement of a cross-community Assembly vote in support of a reunification referendum before it could be triggered. Even David Trimble, who as leader of the Ulster Unionist Party was one of the Agreement’s most important architects, couldn’t resist tabling an amendment which would have required a 60 per majority referendum outcome before the UK Government would take steps towards unification.

The GFA, however, settled the terms of the UK’s obligations word for word; a majority meant a simple majority of those voting in a referendum. The amendments were defeated by overwhelming majorities. Perhaps as a consequence of this distraction, however, the debate neglected various important aspects of the referendum process. Neither the Agreement nor the Northern Ireland Act details how the Secretary of State should form a judgment that there is a likelihood of a majority in favour of a united Ireland. Therefore, although the Act imposes a duty on the minister to act if that condition is reached, the Act appears to hand unrestricted discretion to the minister in how to gauge the opinion of the people of Northern Ireland.

In the words of Paul Murphy, the junior minister tasked with responding to the debate, the ‘spirit of the agreement is that the Secretary of State should, when she or he thinks fit … hold a border poll’. Ian Paisley fulminated about the prospect of such an unrestricted power, and Murphy was obliged to clarify that Parliament would be involved in approving the referendum process: 

In order for the mechanics of that poll to be agreed by Parliament, there would be an opportunity … for both Houses to review the Secretary of State’s decision if it were judged rash and ill-defined and if it was believed that a border poll was being held without basis in reason or common sense. I have not the slightest doubt that hon. Members in this and the other place would point that out during such a debate.

In other words, the UK Parliament would be able to reject the statutory instruments that would be needed to organise a referendum if it thought that such a vote was a bad idea. But on the opposite side of the coin, there is no meaningful parliamentary process for pushing a recalcitrant Secretary of State towards triggering a referendum if the UK Government is unwilling to act upon evidence of a shift in opinion in Northern Ireland towards favouring (re)unification.

In recent years the UK Government has repeatedly refused to be drawn upon how it conducts its assessment of public opinion in Northern Ireland (the weight, for example, it gives to opinion polls or election results in Northern Ireland). Parliamentary questions have been dismissed on the basis that the Secretary of State does not believe that the conditions triggering the duty to hold a referendum have been met. Given concerns over Brexit’s impact upon Northern Ireland, it was almost inevitable that there would be a turn to law to understand the limits of ministerial discretion.

The case of In re McCord, decided by the High Court of Justice in Northern Ireland in June 2018, challenged the failure of the Secretary of State to make public the grounds for determining public opinion on (re)unification. Sir Paul Girvan swiftly dismissed the challenge:

Evidence of election results and opinion polls may form part of the evidential context in which to exercise the judgment whether it appears to the Secretary of State that there is likely to be a majority for a united Ireland. The overall evidential context on how it should be analysed and viewed is a matter for the Secretary of State.

It was always going to be an uphill struggle to convince the courts to tie the hands of ministers over such a decision, but the brusque tone of this judgment calls into question the willingness of the courts to deal with issues around the interpretation of fundamental aspects of the GFA’s incorporation into domestic law.

In re McCord is currently under appeal. Even if that appeal is unsuccessful, the Northern Ireland Court of Appeal might at least take the opportunity to clarify the boundaries of the executive discretion. Even if a decision over a reunification referendum is based upon the Secretary of State’s assessment of the state of public opinion amongst voters in Northern Ireland, that assessment has to be conducted on a rational basis (requiring what Paul Murphy described in 1998 as ‘a proper assessment of the political circumstances at the time’). On this basis, the Court could take the opportunity to clarify that the rationality of a failure to act upon a body of evidence (either opinion polls or election results) which suggested that there could be majority support in Northern Ireland for unity would be open to legal challenge.  

Even after the trigger point for a referendum has been reached, the open nature of the Northern Ireland Act’s key terms generate the potential for political and legal wrangling. The electorate for a referendum is uncertain; the GFA’s definition of the people of Northern Ireland would necessitate the creation of a special franchise which could take years to put in place. And then there is the guarantee of the principle of consent itself; the legislation only requires the UK Government to ‘lay before Parliament … proposals to give effect to that wish’. The legislation falls short of an obligation on the UK Parliament to give effect to a referendum outcome in favour of a united Ireland (even if the notion of such an ‘obligation’ makes any sense in a constitutional order which prioritises the sovereignty of the sitting Parliament).

There is thus no mechanism under the Agreement, or under the Northern Ireland Act, by which the UK Government can be pressurised into authorising a referendum on the (re)unification of Ireland. Sinn Féin can pledge, in its manifesto for Ireland’s 2020 General Election to ‘[s]ecure a referendum, north and south, on Irish Unity’, but even if it enters government in Ireland having secured the largest first-preference share of the popular vote, it has no means of delivering upon this promise. If the Secretary of State for Northern Ireland does not accept that there could be majority support for a united Ireland, he is under no duty to call a referendum.

These shortcomings mean that nationalist politicians should have had as many questions over the workings of a referendum on unity as their Unionist counterparts. But the one-sided nature of the parliamentary debate demonstrated either a trust in the UK Government’s bona fides or a prioritisation of getting the power-sharing institutions in place. This is not to say that the UK Government will renege on its longstanding commitment to having ‘no selfish strategic or economic interest in Northern Ireland’ or that it would seek to thwart majority support amongst the people of Northern Ireland for unity. But there are few legal avenues for enforcing its pledges.

Colin Murray is a Reader in Public Law at Newcastle University (UK). This post forms part of the ESRC Governance after Brexit Performing Identities project and is drawn from the article, ‘Life after Brexit: Operationalising the Belfast/Good Friday Agreement’s Principle of Consent’ forthcoming in the Dublin University Law Journal (co-authored with A O’Donoghue).

(Suggested citation: C. Murray, ‘A Referendum on a United Ireland: Perspectives from UK Constitutional Law’, U.K. Const. L. Blog (18th Feb. 2020) (available at