The Administration of Justice in Administrative Tribunals: Zalewski v. Adjudication Office [2021] IESC 24 (Article 37.1)

See my introductory note here and previous post (on Article 34.1) here.

Article 37.1 was designed as a ‘saver’ clause to ensure that administrative decision-makers could exercise some powers and functions of a judicial nature. The historical background is covered in McKechnie J’s dissenting opinion (at paras. 25-30). I think the history is ambiguous (a matter for another day), with elements pointing to both narrow and broad readings of the provision, but there is no doubt that it was intended to put beyond dispute the constitutionality of judicial-type decision-making by some administrative decision-makers.

Remarkably, Article 37.1 had really only been judicially interpreted in one previous instance, Re Solicitors Act 1954, [1960] I.R. 239. The issue here was whether a professional disciplinary body could be created to regulate solicitors and impose disciplinary sanctions for ethical breaches, including the power to strike a solicitor off the register. The Supreme Court held that this was an administration of justice that was not saved by Article 37.1 due to the far-reaching consequences of the available sanctions. As Kingsmill Moore J explained at pp. 263-264:

What is the meaning to be given to the word “limited”?  It is not a question of “limited jurisdiction” whether the limitation be in regard to persons or subject-matter.  Limited jurisdictions are specially dealt with in Article 34, 3, 4°.  It is the “powers and functions” which must “limited,” not the ambit of their exercise.  Nor is the test of limitation to be sought in the number of powers and functions which are exercised.  The Constitution does not say “powers and functions limited in number.”  Again it must be emphasised that it is the powers and functions which are in their own nature to be limited.  A tribunal having but a few powers and functions but those are far-reaching effect and importance could not properly be regarded as exercising “limited” powers and functions.  The judicial power of the State is by Article 34 of the Constitution lodged in the Courts, and the provisions of Article 37 do not admit of that power being trenched upon, or of its being withdrawn piecemeal from the Courts. 

The test as to whether a power is or is not “limited” in the opinion of the Court, lies in the effect of the assigned power when exercised.  If the exercise of the assigned powers and functions is calculated ordinarily to affect in the most profound and far-reaching way the lives, liberties, fortunes or reputations of those against whom they are exercised they cannot properly be described as “limited”. Eventually the question whether any particular tribunal is unconstitutional must depend on whether the congeries that the powers and functions conferred on the tribunal or any particular power or function is such as to involve the pronouncement of decisions, the making of orders, and the doing of acts, which on the true intendment of the Constitution are preserved for judges as being properly regarded as part of the administration of justice, and not of the limited character validated by Article 37.

Re Solicitors has generally been understood as requiring significant judicial involvement in professional discipline matters, and the Oireachtas has indeed legislated on this basis (see further here). In other areas, where courts have reached the issue of Article 37.1, they have readily found that the functions are “limited”: Central Dublin Development Association v. The Attorney General (1975) 109 I.L.T.R. 69 (exempting developments from planning law); Madden v. Ireland (Unreported, High Court, McMahon J., 22 May, 1980) (fixing the price of compulsorily acquired land); The State (Calcul International Ltd. and Solatrex International Ltd.) v. The Appeal Commissioners & The Revenue Commissioners (Unreported, High Court, Barron J. 18 December, 1986) (powers of an administrative tribunal to hear and determine appeals about liability to tax); and An Blascaod Mór Teo. & Ors. v. Commissioners of Public Works & Ors. [1998] IEHC 38 (arbitral power to assess compensation for the acquisition of land).

O’Donnell J considered, however, that the definition given to “limited” functions and thus the scope of the ‘saver’ clause in Article 37.1 was too narrow (and may, indeed, have caused Irish courts to construe Article 34.1 narrowly, so as to avoid the consequences of finding that a given function was an administration of justice). Read in its whole context, and in light of the factual background against which Article 37.1 was drafted — with a clear desire to safeguard the constitutionality of the significant powers exercised at the time of the Constitution’s drafting by the Revenue Commissioners and Land Commissioners — the ‘saver’ provision had to be given a broad reading.

For the majority, O’Donnell J suggested that a variety of matters could be considered in determining whether an administration of justice was “limited” within the meaning of Article 37.1. (Canadian readers will hear echoes here of the majority reasons in the Court of Quebec reference.) The range of subject-matter to be considered by an administrative tribunal, the remedies available to it, the extent of enforceability of the remedies and the availability of a right of appeal are all factors that can be considered in the analysis of Article 37.1:

First, and most obviously, it is limited by subject matter to those areas of employment law specifically identified in the Act.  It has no inherent jurisdiction, and no jurisdiction under, or in relation to, common law.  Furthermore, it does not have jurisdiction to deal with any other type of dispute.  This, in itself, is, in normal language, a significant limitation and, moreover, something that distinguishes such a body from courts established under the Constitution having general jurisdiction.  Second, there is a limitation on awards which can be made by the W.R.C. which, for example, in cases of unfair dismissals, is limited to an award of compensation of 104 weeks’ remuneration.  In some cases, this can, of course, be a substantial sum, but it may equally in some cases fall short of the loss suffered by the applicant.  It is, in any event, a limitation on the powers of the W.R.C.  The Circuit Court has, for example, a limitation on equitable jurisdiction by reference to rateable valuation which captures some very valuable property, but that it is still a court of limited jurisdiction when dealing with such matters is undeniable.  Third, there is the (much reduced) limitation on enforceability coupled with the limited capacity of the District Court to substitute compensation for redress by way of reinstatement or reengagement.  Fourth, the decision of the W.R.C. is subject to appeal.  While the question of appeal or confirmation by the court has tended to be approached under the heading of the enforceability of the order made by the deciding body, it is also, and perhaps more, relevant when considering the question of limitation on the powers and functions of a non-judicial body under Article 37.  A requirement that a decision be confirmed by a court, or which makes it subject to a full de novo appeal in a court, is necessarily a limitation on the powers of the body giving the decision.  Here, the decision by an adjudication officer is subject, firstly, to a full appeal on a matter of fact to the Labour Court.  That body is, in turn, subject to appeal on a point of law to the High Court.  These appeals are available as of right, and do not require permission from either body or the court itself.  Thus, the correctness of the conclusion of the W.R.C. on matters of fact or law may be reviewed and, insomuch as a decision made by the Labour Court is a matter of law (as it can involve the application of law to the facts), it is reviewable, in turn, by the High Court (at para. 116).

The availability of judicial review was also a relevant factor (at para. 117).

Here the dissenters parted company. For MacMenamin J, the majority’s approach was “to disproportionately elevate the administration of justice under Article 37 into a position of near-equivalence to Article 34, which sets out the essence of the principle of where justice is to be administered – that is in courts established by the Constitution” (at para. 132). His view is that there was no need to try to achieve a harmonious interpretation of the two provisions: only courts can administer justice, and that is the end of the matter. Presumably (although MacMenamin J does not spell this out clearly), Article 37.1 allows non-courts to exercise some judicial powers and functions, as long as they are not administering justice. He feared that the majority’s approach would dilute the content of Article 34.1 by permitting all manner of powers and functions previously exercised by courts to be exercised by non-courts, as reliance on a variety of contextual factors would mean there is “no objective boundary” (at para. 135) between those functions the courts must exercise and those administrative tribunals may exercise. McKechnie J was of a similar mind:

Article 37 was grafted as an exception or derogation from a core principle upon which the vast majority of constitutional regimes, throughout the world, function, namely that justice is administered by judges duly appointed as such. Such provision is subservient to Article 34 and must yield to its hierarchical superior of general application. It therefore follows that no matter what interpretive tools are used, such a provision must be narrowly construed and only in that way given effect to: any other application is a hostage to trust and offensive to the equilibrium created by the Constitution: such would severely undermine and damage the judicial institution, even if that is already considered, by some, as the weakest limb of government (at para. 124).

For his part, Charleton J sought an objective boundary in the erection of three permissible categories of “limited” judicial power or function: “technical matters; findings which do not have any result other than the public expression of an opinion with no consequent order; and provisional orders and findings subject to immediate appeal which are so limited until confirmed by a court that are of no lasting effect” (at para. 40). Technical matters would cover issues such as the calculation of taxes (The State (Calcul International Ltd and Solatrex International Ltd) v Appeal Commissioners [1986] 12 JIC 1802) and, in Z’s case, the accurate computing of holiday benefits. As far as ‘opinions’ are concerned, conclusions of tribunals of inquiry, or preliminary decisions by administrative tribunals would fall within this category. Lastly, where matters can be appealed to a court, “that limitation brings the model onto a constitutionally permissible ground” (at para. 48), as long as the appeal is genuinely judicial and does not simply result in the affixing of a judicial stamp to an administrative decision (see e.g. Re Solicitors Act 1954 [1960] IR 239). This case did not fall into any of these categories. And, so Charleton J asked and answered what he considered to be the key question: “does unfair dismissal need an adjudication by a judge, even if only after a preliminary but non-binding unless appealed ruling by an administrator or tribunal? As a matter of justice and of close connection to contract law, the answer is yes” (at para. 54; see also para. 64).

To summarize, the majority read both Article 34.1 and Article 37.1 broadly. The dissenters read Article 34.1 broadly and Article 37.1 narrowly. What is missing from Zalewski is a narrow reading of both Articles 34.1 and Article 37.1. I am not convinced that Article 34.1 must or should be read broadly. As far as Article 37.1 is concerned, I am persuaded by the dissenters’ structural argument (though I do not share their concerns about erosion of the powers of the courts). In my view, there is a strong case for reading both provisions narrowly, grounded in the historical backdrop to the drafting of the Irish Constitution and an appreciation of comparative jurisprudence. I will develop this point at a later date.

This content has been updated on October 18, 2023 at 12:32.