A close look at the law and policy of holding a Northern Ireland border poll

How the law carefully shapes what will be an essentially political decision

Yesterday Prospect posted something by me on the issue of a potential border poll in Northern Ireland.

Please click (and read!) here.

This post unpacks that Prospect post - a sort of “behind the scenes” perambulation of how that post came together - and a further discussion of the issues.

*

The starting point is that a potential referendum in Northern Ireland has been in the news.

On one hand:

On the other hand:

The Sinn Fein quote was:

"What I firmly believe is - in this decade - we will have those referendums, and it's my job and the job of people like me who believe in reunification to convince, to win hearts and minds and to convince people of that opportunity - part of which, by the way, will be really consolidating our relationship with Britain as our next door neighbour and good friend."

Asked if she meant before 2030, Ms McDonald said "yes".

The Prime Minister’s quote was:

Rishi Sunak has told Sinn Fein to focus on the “day to day” concerns of people in Northern Ireland rather than a referendum on Irish reunification.

The PM said “constitutional change” should not be a priority for the Irish nationalist party, after newly appointed first minister

*

Now we will look at the relevant legislation - the Northern Ireland Act 1998.

There are two key provisions.

First, there is section 1, which should be read carefully:

(And legislative and literary purists will react pleasingly to that “But” at the start of the second sub-section.)

That section 1 needs to be seen in the context of, well, other sections 1s.

Here is section 1 of the Ireland Act 1949 - and look especially at sub-section1(2):

You will see at sub-section 1(2) what can be called the “consent” principle - though it would be for the then parliament of Northern Ireland to give that consent (which at the time would seem very unlikely).

By 1973 that parliament was suspended, and so the Northern Ireland Act 1973 switched the giving of consent to a majority of the people of Northern Ireland:

And as this Substack has set out at length before there had been such a border poll, just before the 1973 Act was passed.

The 1973 poll was heavily in favour of the union - and the nationalists largely boycotted the vote.

At the time the poll seemed pointless from a nationalist perspective, and it was also seen as a sop from the London government to the unionist majority.

However, sometimes, things change - and demographics change.

Those seeking the unification of the island of Ireland no longer see border polls as a necessarily unionist device.

Indeed, the roles are somewhat reversed: it is now the unionists that are wary of a border poll.

*

Having set out that context, let us go back to the text of section 1 of the 1998 Act:

Here there is a declaration - and you will note similar wording was used in the 1949 and 1973 Acts. The use of a declaration is not new. And it really has to be a declaration (or affirmation) as it describes something as it stands, rather than providing for something new. Section 1 does not make Northern Ireland part of the United Kingdom - that status rests on other legal instruments.

The sub-section also repeats the requirement that consent is required for this declared status to change - and like the 1973 Act it then refers to a schedule to the Act.

But.

There is that second sub-section, beginning with a “But”.

And this is where it becomes interesting and departs substantially from previous legislation:

The word “shall” in that provision is highly significant.

For sub-section 2 creates an obligation. If the majority in such a poll supports unification, then the government of the United Kingdom has to bring forward the legislation that would make this so.

If you read carefully, however, you will note that the obligation only goes to putting legislation before parliament. It does not actually place an obligation on the government of the United Kingdom to ensure such legislation is passed. In this technical way, the supremacy of the Westminster parliament is maintained.

But if there was such a poll majority, it is difficult to see how the Westminster parliament would reject such legislation - though presumably some unionist and conservative MPs would oppose it, regardless of the “will of the people” on this one inconvenient occasion.

*

So much for section 1 of the 1998 Act - for it is in the first schedule that the real excitement begins.

(Oh, for what it is worth, lawyers tend to call them sh-edules, not sk-edules when they are in legal documents, I do not know why.)

Here are the first two paragraphs of the schedule:

The two paragraphs are doing different things.

The first paragraph confers a discretion on the government of the United Kingdom (acting through the Secretary of State for Northern Ireland). The key word is “may”.

The Secretary of State may direct there to be a vote; or, then again, the Secretary of State may not do so. It is a choice.

On the face of it, it is an absolute discretion - the Secretary of State may direct as many such border polls as he or she would like. One poll a day, every day even.

But in invisible ink, so to speak, there are limits to be discretion that are implied by law.

The discretion has to be exercised in the public interest and for the purpose of section 1. And given it is part of a statutory scheme implementing the Good Friday Agreement, it would also be necessarily implicit that the discretion should be not be exercised in the selfish interests of the United Kingdom government in maintaining the union.

But those are very wide parameters, and the courts would not intervene if the paragraph 1 discretion is never exercised in favour of holding a border poll.

(If such a poll is held, however, and a majority is in favour of unification, that would mean section 1(2) applies and legislation would have to be brought forward.)

*

Now let us look at the fascinating and significant second paragraph:

This is distinct to and separate from the discretion conferred in the first paragraph, for this paragraph imposes an obligation.

The government of the United Kingdom has to hold a border poll if the condition in that paragraph is met: the government of the United Kingdom cannot choose not to do so.

The condition is framed in wide terms and contains two elements, which I will call (A) and (B): the Secretary of State shall exercise the power under paragraph 1 if (A) at any time it appears likely to him [or her] that (B) 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 first element (A) is about as wide as you can have as a ministerial discretion under public law (the law governing public bodies). This means a court will not intervene readily to gainsay what the Secretary of State considers to be the position.

But.

Element (B) limits that subjective discretion.

Consider the following: that there is, over a period of time, an accumulation of evidence that the nationalist parties are securing majorities both of the popular vote and of the seats on representative bodies, in successive elections, and especially for seats in the Assembly and the Westminster parliament. And that such support is not a blip but a sustained trend.

There will come a point - a tipping point - where it will be come impossible for a Secretary of State to plausibly maintain that it is unlikely that a majority would express a wish that Northern Ireland should cease to be part of the United Kingdom and form part of a united Ireland.

It would become literally incredible.

At that point, the so-called Wednesbury doctrine (named after this case which this Substack has examined before) will apply.

The Wednesbury rule is that, as a matter of law, it is not open for a public law decision-maker to make a decision so unreasonable that no reasonable public law decision-maker would make it.

A refusal by a Secretary of State to direct that a border poll take place in certain circumstances would be Wednesbury unreasonable.

And that would then make it potentially a matter for the courts.

*

The courts will not want to deal with it.

The courts will view this as a political question.

And the two judgments in the McCord litigation (here and here), where an application was made to make the government of the United Kingdom publish (and thereby abide by) a policy on when it would call a border poll, show that judges really REALLY do not want to get involved.

But.

The judges may not get to duck out of it, as much as they would like to do so.

For that second paragraph was placed on a statutory basis deliberately, so that it could have effect in a certain situation.

Here we need to look at the Good Friday Agreement.

The parties to that agreement, including the governments of the United Kingdom and Ireland, agreed the following:

Section 1 and Schedule 1 of the 1998 Act are both express parts of the Good Friday Agreement.

And the United Kingdom undertook to place the provisions on the statute book as part of their obligations under the agreement.

The sentiments behind the provisions could have been put in a mere political declaration, or a recital, or something else without any (real) legal effect - but no. The other parties to the Good Friday Agreement required the United Kingdom to place these provisions into law, and the United Kingdom freely accepted that requirement.

The other parties were wise to do so.

For by placing the obligation into legislation, a legal backstop was created where, if - if - the point was eventually reached where there was simply an abundance of evidence that a majority supported the unification of the island of Ireland, the United Kingdom could not maintain an unreasonable refusal to hold a border poll.

*

Of course, in that extreme scenario, the judges may still wish not to get involved - and it is unlikely that the courts would grant a so-called mandatory order requiring the Secretary of State to direct a border poll.

But there would be no reason why the court could not grant a declaration saying instead that a refusal would be Wednesbury unreasonable.

*

Another ground on which the courts may intervene is if the second paragraph was used by the government of the United Kingdom to cynically hold a snap border poll hoping that they would win, so as to gain the protection of the stipulation that another such poll could not be held within seven years.

Such a ploy would be in bad faith, and for a collateral purpose, and this would mean that a court could quash such an order.

Wednesbury and bad faith are among the very widest parameters in the public law of the United Kingdom - and they only apply in exceptional cases. But they are there, and this means that paragraph 2 of the schedule is not a law-free zone - as much as politicians (and judges) may want it to be.

There will be circumstances where nationalists can and will apply to the courts to enforce a provision freely agreed to and legislated by the British state. And so the government of the United Kingdom cannot just refuse a border poll forever, regardless of what happens in Northern Ireland.

*

Of course, the condition in paragraph 2 may never be met. It may well be that the evidence never becomes that overwhelming and stark, and that support for unification (like support for independence in Scotland) never solidifies into an ongoing, sustained majority.

But that a condition is not (easily) met does not make it any less of a condition at law. The government of the United Kingdom knows it is there, and the nationalist community knows it is there.

It is no longer an absolutely free choice by the government of the United Kingdom whether there is a border poll or not.

*

Holding a border poll for Northern Ireland is essentially a political matter.

And in the real world, it is a matter that will undoubtedly be decided by politics, and not by courts.

Yet it is not a situation where the law is entirely absent. The 1998 Act sets a longstop where, if a certain condition is ever met, a border poll has to be held - even if the United Kingdom does not want to do so - and that a majority in that poll for unification has to be respected.

And this is a key and express part of the Good Friday Agreement, that carefully crafted (and extraordinary) document which has had so many long-term effects on our polity.

*

Going back to the quotes at the top of this post:

The suggestion is that Sinn Fein believes they will be able to show a sustained majority for unification within ten years; while the British government wants the political majorities in Northern Ireland to signify other day-to-day things, unconnected with the unification question.

You can see why, for both, these are the lines-to-take.

An objective of the nationalists is to create a “majority” situation where it would be unreasonable for a border poll not to be held; while an objective of the British government is to have a situation where a Secretary of State in good faith can reasonably believe that no such majority (yet) exists.

But if and when a political decision is made for the poll to take place, there will be some regard to the ultimate legal position under the first schedule to the 1998 Act.

As always: law shapes policy and politics; and politics and policy shape law.

And the policy and politics that shaped the extraordinary and consequential Good Friday Agreement (and the 1998 Act) in turn continue to shape the policy and politics of the United Kingdom.

***