The mechanics of a Border Poll

Mark Bassett and Colin Harvey have this week produced an important paper, which works through the mechanics of a border poll. Although as the authors point out, this might not be the best choice of words:

When people use the term “border poll” they are referring to the question of whether Northern Ireland will remain within the United Kingdom or join a united Ireland. Although it is widely used and adopted, we question how helpful it is in capturing the complexity of this constitutional conversation.

They highlight two things straight away – that the simultaneous referendums north and south must have a clear and consistent question, and that they must approve or reject a legally concrete package, rather than an aspirational statement. Bunreacht na hEireann must be amended for any change in the constitutional status of Northern Ireland, and the process of its amendment is strictly defined:

The procedure for holding a referendum in the Republic of Ireland is governed by article 46 of the constitution, and Referendum Acts. In short, the proposal must be approved by both houses of the Oireachtas (or deemed to be passed), submitted to and approved by the electorate and signed into law by the President. Consequently, it can be said, in purely legal terms, that the decision to propose a referendum on unity lies with the Oireachtas, while the approval or rejection of the unity proposal rests with the electorate.

Referendums in Northern Ireland are not so strictly governed, however the simultaneous referendum in the north will naturally be seen as approving or rejecting the same constitutional amendment that the Republic is voting on. It would therefore minimise confusion if this was made explicit:

The critical point, however, is that the [GFA] referendums were essentially the same in substance. The electorate was being asked to endorse or reject a new constitutional arrangement. A rejection in either jurisdiction would have meant an immediate continuation of the status quo. No doubt, this would have prompted renewed efforts to find consensus but this too would, inevitably, have had to find approval in both jurisdictions by way of concurrent referendums.

The traditional vision of a border poll – call one first in order to force unionism to the negotiating table – would simply repeat the mistakes of Brexit. No swing voter will ever vote for a pig in a poke again. Remember that before the GFA referendum, every household got a copy of the agreement through their door. People read it and debated it at the kitchen table before voting on it. If there is a border poll, the electorate will be entitled to nothing less.

The authors drop a few hints about what unity will look like, or more importantly what it will not look like:

A position that demonstrates that unity is consistent with, and furthers the values of the Constitution and GFA, should be adopted. This is, of course, an option and an outcome that the GFA fully anticipates because Irish unity is one logical consequence of basing constitutional status on the principle of consent.


As noted above, in its international legal manifestation, the GFA also offers a normative framework for the unity conversation: there are provisions in the GFA that are clearly intended to continue to have relevance in the event of Irish unity and the principles are of fundamental significance to guiding the approach. This will not be, and it should not be, a “blank page” discussion.


The creation of Bunreacht na hEireann by de Valera’s government is widely, if not universally, seen to have been a good thing for the State. It established institutions, fundamental rights protected by the courts and a method of direct participation, in the form of referendums that have served the State well to date. There is no appetite for wholescale changes to it within the Republic.

This is key. Any change in sovereignty will come with costs, and in an equitable world the more affluent Republic will bear proportionately more of these costs than the north. Combine that with too much constitutional reconstruction in one sitting and Terenure Man may think twice about something that he always believed was a no-brainer.

They go on to warn:

Statements which deride or dismiss the record of Northern Ireland as one of economic decline, political failure and social conflict need to be approached with caution. While these will affirm the views of those who already see the north as a “failed entity”, they will do little to persuade those open to a different constitutional future (and who see the region in more positive terms).

The key to winning a mandate for change is to identify your swing constituency and target your messaging. If your swing constituency is letsgetalongerists and soft unionists, there could be no worse kind of messaging than republicans droning on about the failed sectarian statelet. Change has to be driven from, and won in, the centre ground – and the centre does not respond the same way as the core vote. Often they are diametrically opposed.

The closing paragraphs reiterate caution:

In considering this, it should be recalled that the north would be joining a pre-existing state (within the EU) whose constitution anticipates reunification, and in the context of an international agreement that guarantees continuity of protection. It is possible to exaggerate the logistical and practical challenge of achieving this objective: this will simply not be a “blank page” constitutional conversation. Tensions do exist. There are those who view the reunification conversation as an opening for radical and transformative change. In this view, a vote for unity should be consent to a different and new Ireland; a fresh start for the island.

The other view is more conservative: achieving effective reunification would be enough of a task without also re-opening existing constitutional arrangements in a major way. These tensions will need to be discussed as part of the planning and preparation and this is likely to generate quite a debate. Whatever view is taken of this, there is a strong and persuasive case for ensuring that proposals contain robust guarantees for rights, equality and identity in the context of reunification in the spirit of the Agreement’s vision of these safeguarding measures.

It is our view that the basic normative principles to guide the process are already there, and that there is an unhelpful tendency to suggest that everything is “up in the air” or “up for grabs”: full respect for the GFA indicates that this is not the case.

Indeed. The GFA was itself a conservative document. The general thrust of it had been clear since Sunningdale, and the ambitions of Strand Two in particular had been scaled back, yet the GFA was still an enormous step forwards. This was precisely because the GFA was a conservative document. Anything more radical would not have passed.

Even if a change is accepted in principle, it will often founder on the details. The principle of a united Cyprus has been accepted by both sides for decades, and yet still they cannot agree. The principle of a united Ireland is not yet accepted by both sides, so what chance have we of agreeing a radical change, or any change at all? If there is to be constitutional change, three things need to be done in sequence.

Firstly, the need for change must be sold in principle to as broad a constituency as possible. Secondly, the details of the change must be agreed amongst those who already accept change in principle. And thirdly, these details must be approved by a majority of the electorate.

At every stage you will shed dissenters. Not everyone who agrees in principle will agree to the detail. Not everyone who agrees to the detail will be able to sell it on the doors. At all times the electorate will be watching the dissenters carefully, to see what their reaction will be. If you have two thirds backing at the “in principle” stage of the process, you have a decent chance of 50%+1 come polling day. If you have just 50%+1 backing in principle, you have already lost.

So caution is not just advisable; it is mandatory. Given the caution advised by the report’s authors, it is unfortunate that they choose to take a shot at the current Irish government for its own statements urging caution (statements that have been wilfully and persistently misrepresented by the opposition).

But two bullet points from the middle of the document are worth concentrating on:

– Allowance for the continuation of the northern institutions
– Allowance for the continuation, subject to the constitution, of NI laws

Northern Ireland law would have to continue indefinitely after unity for legal and business certainty. Pre-partition Irish law continued (and still continues in many cases) in both jurisdictions after 1922 for the same reason. A legal cliff edge must be out of the question. And since modern law invariably comes with statutory bodies, regulations and enforcement mechanisms, these must all continue also.

To allow these institutions to continue to function without local oversight would be a mirror image of Direct Rule, and no more democratic or just. So inevitably Stormont would have to continue. But to have northern TDs sitting in Dail Eireann voting on issues devolved to Stormont would be a mirror image of the West Lothian Question, and no more democratic or just.

So unless we want to import the constitutional open sore of UK-style asymmetric devolution into a new Ireland, the only option left open is a federal or confederal union of nominally equal jurisdictions. And that will mean some sort of “Council of Ireland” body that can act as an overarching layer of government for the two states that already exist. This is effectively a Sunningdale-plus model of interjurisdictional cooperation, which could be achieved by beefing up the GFA’s existing Strand Two. The various Cyprus reunification proposals (a “bizonal confederation”) are therefore surely the closest analogue and a useful primer.

The details of whatever constitutional proposal will eventually be put to the people are still unclear. But the red lines are being drawn tighter, and it is becoming harder to keep talking about mere procedure without also discussing substance.

What nationalism sorely needs now is someone to discuss that substance with.