Anurag Deb: Privacy International: A Matter of Constitutional Logic and Judicial Trust?


The High Court and Court of Appeal decisions in Privacy International v IPT have ignited the debate on the ‘validity’ or otherwise of ouster clauses in the UK constitution (see analyses in the Admin Law Blog and the UKCLA Blog). At one end is the argument that ouster clauses prevent the correction of legal errors by judicial superintendence, while at the other end is the riposte that Parliament’s sovereignty mandates that even an ouster clause should have its ordinary meaning judicially enforced. The baton has most recently passed to the Supreme Court, the much-awaited judgment of which will no doubt be immediately and thoroughly analysed.

In this post, I argue why constitutional logic mandates that the ouster clause stand in this case.

Some brief facts

The crux of the matter for the Supreme Court rests on the construction attributed to section 67(8) of the Regulation of Investigatory Powers Act 2000 (“RIPA”), which states: “Except to such extent as the Secretary of State may by order otherwise provide, determinations, awards, orders and other decisions of the Tribunal (including decisions as to whether they have jurisdiction) shall not be subject to appeal or be liable to be questioned in any court.

The Tribunal is of course the Investigatory Powers Tribunal (“IPT”), established by section 65 of the RIPA to be the only tribunal in which human rights claims against intelligence services or claims for breaches of specified duties under the RIPA are authorised to be heard.

The underlying factual matrix of Privacy International is no less consequential. The decision of the IPT with which the appellant takes issue is its interpretation of section 5 of the Intelligence Services Act 1994 as authorising the issue of thematic warrants by the Secretary of State (which hinges on a class of persons or property) as opposed to specific warrants (identifying specific persons or property), squarely confronting an ancient common law bar on general warrants (see Entick v Carrington [1765] EWHC KB J98). However, this point does not concern the Supreme Court – yet. It may change depending on the outcome of the appeal.

Parliamentary sovereignty and intention

Adam Tucker, in part 1 of a brilliantly detailed post, sets out the doctrinal reasoning for what he calls a ‘duty of hostile interpretation’ on the part of courts when construing ouster clauses, as such clauses are constitutionally offensive in trying to oust the courts’ ordinary jurisdiction. Indeed, Tucker goes further and suggests that parliamentary intention, when construing ouster clauses, is ‘irrelevant’. With the greatest of respect, I would recast the position somewhat.

Parliamentary sovereignty is the logical corollary to the lack of an anterior body of constitutional law in which constitutionalism in the UK has had to evolve. There is arguably a set of anterior principles which have come to define and shape the UK constitution. Chief among these (at least for present purposes) is the separation of powers. It is uncontroversial that each arm of the State has its own domain, in which it is (to varying degrees) a master of its own process. Parliament enacts, the Executive implements and the Judiciary adjudicates and declares the law. In promulgating enactments, Parliament does not stray into the judicial domain by legislating legal effect – this remains, to borrow Marshall CJ’s words, “emphatically the province and duty of the [courts]” (Marbury v Madison, 5 US 137 (1803)). However, as parliamentary enactments cannot be measured for substantive validity against any anterior law, what Parliament enacts (which consequently reflects parliamentary intention) is the logical starting position for determining how such an enactment is to have effect. The antiquity of this rule (see e.g. The Sussex Peerage Case (1844) 8 ER 1034) should not be indicative of any orthodoxy – it is simply, as before, a matter of constitutional logic.

However, an ouster clause legislates for legal effect – in attempting to enshrine only one possible interpretation (ouster), thereby denying courts a role in construal. Thus, such a transgression into the judicial domain must be clear and certain before being allowed. Therefore the clarion call in cases such as Anisminic v FCC [1969] 1 AC 147 and Jackson v Attorney General [2006] 1 AC 262 for ‘crystal clear’ words in ouster clauses is indicative of a duty of constitutional fidelity, rather than hostility to Parliament.

Indeed, overtly ignoring parliamentary intention has an insidious effect on constitutionalism: it creates an entirely arbitrary rule of construction, limited only to ouster clauses, which mandates severance of the link between parliamentary enactments and their legal effect. In other words, no matter the language passed by Parliament, ouster clauses should never achieve legal effect, for fear of provoking constitutional crises (Tucker, part 2) thus rendering an entire class of statutory provision almost gratuitous – a concept unknown to UK constitutionalism. It would also make the ‘crystal clear’ requirement effectively unattainable, making the long-established search for such words by courts also gratuitous.

What is thus clear is that although parliamentary enactments are not always construable in exact ‘ordinary language’ terms, particularly in light of principles such as legality (R v Home Secretary ex p O’Brien [2000] 2 AC 115), the intention of Parliament, as expressed through the content and context of any enactment, is highly relevant, if not of paramount importance (R v Home Secretary ex p Daly [2001] 2 AC 532). To that extent, Sales LJ’s remark in Privacy International that the entire issue turned on a “short point of statutory construction” ([2017] EWCA Civ 1868, [24]) is half correct: the issue does indeed turn on a point of statutory construction, but it is certainly not short.

The magic words

It is unnecessary to exactingly parse section 67(8) in order to understand what it means in the ordinary sense: it firstly shields from review or appeal IPT outcomes (“determinations, awards, orders and other decisions”) and secondly it immunises from appeal or review any decisions by the IPT as to its own jurisdiction. There is provision for authorising the Secretary of State to make orders for appeal and/or review of all or some of the above, but this has yet to be realised.

Section 67(8) can thus be contrasted with that famous ouster in section 4(4) of the Foreign Compensation Act 1950 (in Anisminic), and the considerable extent to which this contrast was relied upon by Sales LJ in the Court of Appeal in Privacy International has been severely criticised by Tucker (part 2). To a point, I agree with Tucker’s criticism. Anisminic is authority for only two points: firstly, that ouster clauses should be strictly construed and secondly, that all errors of law preserve judicial superintendence over administrative bodies.

However, even the two cardinal rules in Anisminic do not aid the appellant in this case for two important reasons. Firstly, there seems to be no crack in the enacted words through which the challenged decision could fall, given the protection of IPT decisions as to jurisdiction. Sales LJ interpreted “decisions as to […] jurisdiction” as capable of encompassing both advertent and inadvertent errors of law ([2017] EWCA Civ 1868, [39], [40]). The logic is difficult to impugn: an error of jurisdiction may occur inadvertently (as was argued in this case) or advertently, and there is no doctrinal reason why Parliament should have immunised one over the other from review. More practically, if Sales LJ is incorrect, then judicial superintendence over inadvertent excesses of jurisdiction is a form of severely restricted judicial review which is difficult to plead and adjudicate. Sales LJ bemoaned the “excessively subtle arguments” which would be inevitable if the reviewing court were faced only with inadvertent errors – no doubt practitioners and judges would be less than thrilled at the prospect of adding a layer of esoteric fudge to an already complex field of litigation.

The crucial question in Anisminic had been to ask whether an ouster clause is reasonably capable of bearing two meanings, and if so, the meaning that preserves the courts’ jurisdiction should be adopted. As set out above, the meaning capable of preserving review in this case indeed defangs such jurisdiction, raising the question of whether such a meaning should be considered reasonable.

The second reason why Anisminic does not aid the appellant is because of the nature of the IPT – a judicial, not administrative body, and moreover an analogue of the High Court (R (A) v Director of Establishments of the Security Service [2017] EWCA Civ 1868). In Re Racal Communications [1981] AC 374, while the House of Lords confirmed retention of review for jurisdictional excess over inferior courts it roundly rejected such superintendence over the High Court. This is also a matter of constitutional logic: judicial, rather than administrative bodies, are presumed to have authority to answer questions of law, and reviewability of such answers depends on the corresponding tribunal’s nature (a point Conor McCormick and I have made elsewhere). So for example in R (Cart) v Upper Tribunal [2012] 1 AC 663 the Upper Tribunal was categorically rejected as an analogue of the High Court, unlike the IPT (both by the same judge – Laws LJ, in Cart and A respectively), thereby confirming the retention of review. The point here is simple: the IPT, if an analogue of the High Court, should enjoy an analogously unreviewable jurisdiction. The alternative interpretation would be logically inapposite.

Support for this point can be found in wider Commonwealth jurisprudence. In Attorney General of Quebec v Farrah [1978] 2 SCR 638, a Québécois statute which excluded “questions of fact or law” from review was interpreted to necessarily exclude from review any excess of jurisdiction (although the provision itself was ruled ultra vires section 96 of the Constitution Act 1867 for lacking in legislative competence).

Finally, the Supreme Court in A [2010] 2 AC 1 has already recognised section 67(8) as an “unambiguous ouster” (although as obiter dicta). If I am incorrect and the rule of law requires clearer language to allow a non-reviewable judicial body to commit mistakes of law, then Parliament could enact words making such authority explicit. However, I do not think such formalism is required given the question of reasonable meanings posed in Anisminic. It cannot be reasonable to demand Parliament underline in red the writing already on the wall.

Before finishing on the language, it is useful to consider whether, should the Supreme Court uphold the ouster in this case, it would preclude future challenges of varying types against the same ouster. The appellant relies on authorities such as Rex v Cheltenham Commissioners (1841) 1 QB 465, in which the ouster clause was considerably more detailed and explicit. Lord Denman CJ upheld the retention of certiorari due to the actual bias in the inferior court. It is incontrovertible that certiorari against bias and mala fides is a control mechanism par excellence, and more importantly, both may lie outside “decisions as to jurisdiction” as interpreted by Sales LJ. This is because mala fides (and in certain circumstances bias) requires ex post facto legal inquiry to be determined, and cannot be assumed to have infected any decision ab initio (Smith v East Elloe Rural DC [1956] AC 376), thereby lying arguably unprotected by the ouster. However, the existence of a control mechanism does not necessitate its availability in every case, including the present one – a point made far more cogently by Joanna Bell.

A matter of judicial trust?

A major plank of the appellant’s case in the Supreme Court is the need for the common law to be interpreted and applied uniformly by the High Court as an independent arbiter and the successor to this historic role of the Court of King’s (or Queen’s) Bench (see Day 1, Morning session). The underlying case being a human rights claim, the appellant submitted that it was of paramount importance that the application of the Human Rights Act 1998 be uniform, taking into account the way in which the ordinary courts have interpreted human rights law. However, this is arguably what Burton J did in his judgment – firstly, by taking into account the case law of Strasbourg, and secondly, by accepting the applicability of the Ullah principle (Burton J at [50]), despite the idea, advanced by the appellant, that section 67(8) effectively places the IPT in an autocratic ‘legal island’.

I would see this as the rule of law properly in play – despite it being apparently acceptable, in principle, for the IPT determining for itself in a complete or even partial vacuum the parameters and application of human rights law in claims brought before it, in practice, it proceeded along anterior and precedential legal lines, both at Strasbourg and domestically, thus recognising the overriding importance of cogency and uniformity in the determination and application of the law.

A question of judicial politics thus forms a common thread in the fabric of this case: how much trust is the Judiciary willing to repose on the judges of an apparently unreviewable tribunal?


Privacy International has ignited a worthwhile debate, as ouster clauses deserve the most anxious scrutiny from the courts. However queasy the notion of an unreviewable body may be, when looking at the nuances inherent in this case, it is difficult to ignore the ouster. More hopefully, even in the absence of judicial superintendence, it is the rule of law, sharpened over the centuries, that should assuage any lingering fears about the emergence of an IPT Star Chamber.

Anurag Deb, paralegal at KRW LAW LLP and BPTC student (University of Law)

(Suggested citation: A. Deb, ‘Privacy International: A Matter of Constitutional Logic and Judicial Trust?’, U.K. Const. L. Blog (8th Jan. 2019) (available at