It is October 2018 and Northern Ireland has been without a government for almost 2 years. In that time, the Northern Ireland judiciary has effectively muzzled the legally-dubious ability of Northern Ireland civil servants to govern by proxy. The response of the UK Government, through Northern Ireland Secretary Karen Bradley MP, has been to draft legislation to replace devolved government with a time-limited technocracy. Far from this being the prelude to science fiction, this post is a critical examination of the very real Northern Ireland (Executive Formation and Exercise of Functions) Bill 2018 (“the Bradley Bill”).
Basic Premises of the Northern Ireland Constitution
At its heart, the Northern Ireland constitution (stemming largely from the Northern Ireland Act 1998 or NIA) rests on two key principles: consociationalism and democratic accountability. References to “cross-community support” and requirements thereof for the Northern Ireland Assembly to pass key legislation and for the Northern Ireland Ministers to take key decisions are peppered throughout the NIA in a way which is alien to the Scottish or Welsh devolution settlements. One of the most crucial examples of this consociational lifeblood is the Executive Office, comprising the First and deputy First Ministers, such that if one should resign, the other must follow (s. 16B(2) of the NIA). When in January 2017 Northern Ireland’s last deputy First Minister, the late Martin McGuinness, resigned, his co-occupant in the Executive Office, former First Minister Arlene Foster, was constitutionally forced to follow.
Even democratic accountability in Northern Ireland is uniquely consociational: one third of elected MLAs can force a cross-community vote in the Northern Ireland Assembly on any issue with the widely controversial petition of concern (s. 42 of the NIA).
The above should come as no surprise given that the Northern Ireland constitutional structure is the parliamentary fulfilment of the promise of devolved government contained in the Belfast (Good Friday) Agreement 1998, Article 1 (v) of which exhorts “full respect […] of parity of esteem”.
The Bradley Bill: Promise and Reality
At the outset, it is important to recall that Northern Ireland’s most recent period of self-governance has been no stranger to breakdown and consequent (though time-limited) periods of direct rule from Westminster. The now repealed Northern Ireland Act 2000 (“NIA 2000”) for example, provided for legislation by Order-in-Council when the Assembly was last suspended (para 1 of the Schedule to the NIA 2000).
By contrast, the Bradley Bill is, in its promise, an invidious halfway house between direct rule and devolved government which is also a reordered constitution in miniature without parallel or analogy in the diverse settlements which have come to define modern constitutionalism in the UK. This becomes clear when its provisions are closely scrutinised.
While the first and second clauses of the Bill deal with extending the period of time for the formation of the Northern Ireland Executive and parameterising the power of the Secretary of State to so extend the time, the real meat of the Bill is to be found in the third clause, which permits senior Departmental civil servants to exercise powers constitutionally reserved for Northern Ireland Ministers, so long as the civil servant in question is satisfied that such exercise is in the public interest. What might be in the public interest is informed by (not determined by) guidance required to be published to this effect by the Secretary of State. Clauses 3(4) and 3(5) seek to retrospectively annul the combined effect of the Buick rulings unless matters which fall within these clauses (decisions taken in the absence of Ministers and cross-cutting, significant or controversial decisions taken without being discussed and agreed by the Northern Ireland Executive Committee) are the subject of proceedings which have already begun. Clause 3(8) meanwhile preserves the voidability of civil servant decisions by courts on grounds other than those in the Bill.
So how does the Bradley Bill measure against established constitutionalism? When examined closely, the Bill displaces at least two key tenets of constitutional law long thought entrenched: democratic accountability and the Carltona principle.
Democratic accountability, through which an elected legislature holds the government to account, has long been a facet, not only of the UK constitution, but of all the devolution settlements: Welsh Ministers are as constitutionally responsible to the Welsh Assembly (see s. 48(5) of the Government of Wales Act 2006) as their Scottish counterparts are to the Scottish Parliament (see s. 47(3)(c) of the Scotland Act 1998). Previous forms of direct rule in Northern Ireland even preserved a modicum of democratic accountability during extremely difficult constitutional circumstances, in that all Orders-in-Council (in place of Acts of the Northern Ireland Assembly) were subject to positive resolution procedures in the UK Parliament (para 2 of the Schedule to the NIA 2000) and all functions exercisable by the Northern Ireland Departments were subject to the “direction and control” of the Secretary of State, who was of course responsible to Parliament (para 4(1)(f) of the Schedule to the NIA 2000). Thus, political accountability merely shifted its seat from Stormont to Westminster.
Readers who have kept their eyes on the Northern Ireland statute book since the collapse of its devolved institutions almost 2 years ago will be aware that Northern Ireland Departments have continued to make secondary legislation in vast quantities in the absence of Assembly oversight since then. This has been possible with respect to all legislative proposals that can be progressed by way of the negative resolution procedure. That procedure is still available because the Assembly has not been formally prorogued. This means that although no MLAs have taken their seats since January 2017, draft statutory rules can still be laid before the empty chamber and take effect when (inevitably) they are not resolved against. The constitutionality of this troubling practice is legitimised to a very limited extent by the fact that the ‘statutory period’ during which the Assembly can annul secondary legislation made in this way is defined as either ten sitting days or thirty calendar days, whichever is the longest (see s. 41(2) of the Interpretation Act (NI) 1954). Thus, there remains a theoretical possibility that secondary legislation made in this way could be annulled by a future Assembly during the first ten sitting days of its return. The reality, of course, is that no such Assembly could feasibly apply proper scrutiny within this time period to the vast array of statutory rules that have been made to date. The extent to which these practices have normalised low expectations of democratic accountability over law-making for Northern Ireland is a serious cause for constitutional concern.
The Bradley Bill pours fresh fuel on such concerns. Northern Ireland Departments are not placed under the control of the Secretary of State, as the Bill does not introduce direct rule by suspending devolved government in the manner of the NIA 2000 outlined above. The senior civil servants for Northern Ireland Departments are instead enjoined to “have regard” to guidance published by the Secretary of State in the exercise of otherwise Ministerial functions. What this might mean in practice is that the Secretary of State could not be held to account for the actions of the civil servants because they would not be constitutionally responsible to her, and thus nor could she, in turn, be responsible to Parliament on their behalf. In this regard, it may be useful to recall the distinction which was made in Miller v Secretary of State for Exiting the European Union  UKSC 5 between freely exercisable powers for which Ministers are answerable to Parliament, and those which are not at all exercisable without prior parliamentary conferment on Ministers (Miller, ). While the Bradley Bill confers certain appointment powers on the Secretary of State (which are consequently amenable to scrutiny in Parliament), it places the bulk of the quotidian governance of Northern Ireland outside any political accountability – whether at Stormont or at Westminster.
Even more insidiously, the Bill creates a test of “public interest” according to the satisfaction of the civil servant who must exercise Ministerial functions in the public interest, after only having regard to guidance from the Secretary of State. This raises two crucial questions of law: first, whether the civil servant’s satisfaction on public interest ought to be construed subjectively or objectively (a question which had opened a public law Pandora’s box in Liversidge v Anderson  AC 206) and second, what weight ought to be accorded to the Secretary of State’s guidance. It would be astonishing if the duty to have regard to such guidance was subject to the almost impassable heights of Wednesbury irrationality given the statutory requirement for the guidance and the fact that it is handed down to an administrator who is simultaneously inexperienced in exercising actual Ministerial power and greatly immunised from the constitutional consequences of such exercise. It may be useful here to draw parallels with Lord Bingham of Cornhill in R (Munjaz) v Mersey Care NHS Trust  2 AC 148 and suggest that the guidance be normally followed, unless cogent reasons for departure from the guidance were shown.
The Carltona principle, a rule of constitutionalism handed down in Carltona v Commissioner of Works  2 All ER 560, elides the separate identities of civil servant and Minister, such that civil servants work with the authority of the relevant Minister. Crucially however, this elision is necessarily contingent on the rule of democratic accountability, as Lord Greene MR himself observed (Carltona, at 563B).
As previously discussed, the Bradley Bill greatly immunises civil servants from political accountability for exercising Ministerial powers, raising the question of whether, as a matter of constitutional law, civil servants would be exercising per se Ministerial powers, rather than powers conferred under the Bill, which cross-refer to the actual Ministerial powers in the NIA as a convenient statutory shorthand. The distinction is of large consequence given that the Bill does not immunise civil servants’ decisions generally. The reason is as doctrinally important as it is practically consequential: when reviewing public authority decisions, courts are loath to interfere with Ministerial powers to the same extent as non-Ministerial powers as Ministers are not merely public servants – they are constitutional actors. In Re Johnstone’s application  NIQB 33, Deeny J (as he then was) characterised this judicial approach as one of “mutual respect” between “the different arms of the State”, implicitly invoking the doctrine of separation of powers which is a foundational aspect of UK constitutionalism.
While the Bill is silent on the above distinction, it must be necessarily implied, as the alternative would be to sanction the creation of a new class of quasi-constitutional actor with near zero constitutional accountability. It is also telling that a significant number of the suggested amendments to the Bill involve conferring explicit Ministerial powers and duties on civil servants, rather than the laissez-faire approach currently in the Bill.
Northern Ireland has found itself at a governance crossroads again, but unlike previous episodes of this ilk, the core tenets of UK constitutionalism itself are in jeopardy. Whether direct rule is preferable to the Bradley Bill is a matter of political as well as constitutional importance but is outside the scope of this post. What is clear however is that we find ourselves in uncharted and seemingly bottomless waters.
Anurag Deb is a paralegal at KRW LAW LLP (Belfast) and BPTC Student at the University of Law; and Conor McCormick is a Lecturer in Law at Queen’s University Belfast.