Editor’s Note: Last month Policy Exchange’s Judicial Power Project published a report by Professors Richard Ekins and Christopher Forsyth on Judging the Public Interest: the Rule of Law vs the Rule of Courts. The Project invited comments on the paper from Professor Adam Tomkins (University of Glasgow), Dr Se-Shauna Wheatle and Professor Roger Masterman (both from the University of Durham) and Dr Jan van Zyl Smith (Bingham Centre for the Rule of Law). Their comments can be found here. In this post, Martin Chamberlain QC offers some observations on Evans v AG.


Professors Ekins and Forsyth in their report make a compelling case that the majority in Evans distorted the clear and obvious meaning of s. 53 and in doing so emasculated a power that Parliament regarded as important, albeit exceptional. Of course, the meaning of “reasonable” differs according to the context. But the words chosen in s. 53 would be a bizarre way to achieve the effect of the constructions favoured by Lord Neuberger or Lord Mance. As Lord Hughes said at para. 155: “If anyone had suggested at the time of the passage of the bill which became the Act that [the majority’s construction] was what was meant, it seems to me that that suggestion would have received a decisive and negative response”. As Lord Wilson notes at para. 170, the inclusion of what is now s. 53 was, as a matter of legislative history, the quid pro quo for the conferral of a power in the tribunal to order, rather than merely recommend disclosure.

Most lay people would, I suspect, be surprised that a court trying to interpret what Parliament has said should adopt an interpretation that Parliament would clearly have rejected if anyone had suggested it. To lawyers, however, constructions that distort – or “read down” – the plain meaning of statutory provisions have become increasingly familiar. Within certain minimal (and difficult to define) limits, they are required where necessary to achieve compatibility with EU law or with the European Convention on Human Rights. In these cases, Parliament itself has licensed the reading down (by s. 2 of the European Communities Act 1972 and s. 3 of the Human Rights Act 1998 respectively). But neither of these applied here (save in respect of environmental information). The distorting construction in this case was a pure artefact of the common law.

Insofar as one can talk meaningfully about trends in the legal grounds for public law challenges, the common law is the latest trend. It started when public lawyers started to anticipate that reform of the Human Rights Act and/or Brexit might deprive them of some currently favoured grounds of challenge. It was given added impetus by Lord Mance’s judgment in Kennedy v Charity Commission [2015] AC 455 and by Lady Hale’s speech to the Constitutional and Administrative Law Bar Association on 12 July 2014. The common law also seems to have commended itself both to Chris Grayling and to Michael Gove, especially in the context of its supposed superiority over international human rights instruments (see, for example, here and here). Nowadays, on-trend public lawyers, conscious of the jurisprudential zeitgeist, tend to put their common law arguments first. This makes it important to articulate in principled terms why and when common law principles can legitimately be used to “read down” the plain words of a statutory provision.

Some “reading down” occurs whenever the courts apply a common law interpretative presumption, such as the presumption against retrospectivity. Most would agree that this is legitimate: as a matter of principle and in general, retrospective provisions are unfair; the Parliamentary draftsman knows this and so knows that, when the intention is to legislate contrary to the principle, this must be spelled out in terms; when it is not spelled out in terms, it can be assumed that Parliament did not intend to legislate contrary to the principle. Here, the presumption is a tool to discover what Parliament actually intended; and the presumption is rebutted if the statutory words, read in the light of the circumstances in which they were enacted, disclose an intention to legislate contrary to the principle. The more unfair the result, the more unlikely Parliament would have intended it and the clearer the statutory words must be to achieve it: see the nuanced judgment of Lord Mustill in L’Office Cherifien des Phosphates v Yamashita-Shinnihon Steamship Co. Ltd [1994] 1 AC 486, at 524-5.

There is a separate line of cases governing the approach to statutory provisions that, on their face, appear to interfere with fundamental rights. Here, the courts apply the principle of legality. The best explanation of the scope of and theoretical basis for the principle was given by Lord Hoffmann in R v Home Secretary ex parte Simms [2000] 2 AC 115, at 131, in a passage cited by Lord Neuberger at para. 56: when legislating contrary to fundamental principles of human rights, “Parliament must squarely confront what it is doing and accept the political cost. Fundamental rights cannot be overridden by general or ambiguous words. This is because there is too great a risk that the full implications of their unqualified meaning may have passed unnoticed in the democratic process.” Again, few would now regard this as controversial. The principle of legality sends a fair message from the judicial to the executive branch about the manner in which it expects Bills to be framed, which promotes rather than undermines Parliament sovereignty: if you are clear and transparent about what you are doing, and accept the political cost, the courts will respect the result as Parliament’s choice; but if you use sneak wide powers through, without coming clean about how they might be used, you can’t expect the courts to regard the result as reflecting a deliberate choice by Parliament at all, and it will be legitimate for the courts to read the powers narrowly.

So far so good. But there is a third line of authority that is more problematic. It arises in the context of ouster clauses. The best known example is Anisminic v Foreign Compensation Commission [1969] 2 AC 147, but – as Lord Neuberger observes – the principle goes back as far as the judgment of Lord Chief Justice Denman in R v Cheltenham Commissioners (1841) 1 QB 467. In Anisminic, a statute provided that determinations of the Commission “shall not be called into question in any court of law”. This, the House of Lords held, did not protect from review a determination that was a nullity, and so, in law, no determination at all. Clearer words would have been needed to achieve that, because of the principle that the ordinary courts have a general jurisdiction over the administration of justice: see esp. Lord Pearce at p. 190. So, a determination that Parliament had said must not be called into question in any court could, after all, be called into question – and quashed – by a court; and not just on the basis that it was a forgery or given in bad faith, but on the basis of any public law error.

The solution in Anisminic is elegant, but the result is difficult to square with the words of the ouster clause, read in context. Suppose that someone had suggested in Parliament that determinations of the Commission should, after all, be capable of being questioned on the basis of a misdirection of law (whether going to jurisdiction or not), a procedural impropriety, a failure to take into account a relevant consideration, the taking into account of an irrelevant consideration or an irrationality. What would Parliament have said? To adopt Lord Hughes’ formulation in Evans, the response would surely have been both decisive and negative. The result in Anisminic suggests that this does not matter, at least where the provision in question offends a principle that the court regards as constitutive of the rule of law. Although lip service is paid to the possibility that Parliament might devise sufficiently clear words to oust judicial review, the Anisminic principle is not a mere interpretative presumption; it is something altogether stronger – a constitutional principle capable of licensing the kind of distortion or “reading down” of primary legislation required by statute when compatibility with EU law or Convention rights is in issue.

Parliament has recognised the existence of “the constitutional principle of the rule of law” in s. 1 of the Constitutional Reform Act 2005, but – as I have argued elsewhere – the content and contours of this principle are contested. Perhaps for this reason, Parliament has not required or empowered courts to read down legislation whenever they consider it necessary to achieve compatibility with their understanding of the rule of law. Lord Neuberger’s judgment in Evans proceeds as if it had. He identifies as a fundamental component of the rule of law the principle, “subject to being overruled by a higher court or (given Parliamentary supremacy) a statute… that a decision of a court is binding between the parties, and cannot be set aside or ignored by anyone, including (indeed, it may fairly be said, least of all) the executive” (para. 52). This is presented as uncontroversial. It is not.

If the tribunal’s powers had been limited to making recommendations for disclosure, as had originally been proposed, such a recommendation would not have been “binding between the parties” unless accepted by a member of the executive. Would this have been regarded as offensive to the rule of law? If so, why? (Courts are often empowered to make recommendations that have no legal effect unless accepted by a member of the executive. Indeed, declarations of incompatibility under s. 4 of the Human Rights Act, though not technically recommendations, have no effect between the parties unless acted upon by a member of the executive and subsequently confirmed by each House of Parliament.) If a power to make recommendations would not have been offensive to the rule of law, why should a power to make binding decisions subject to an exceptional executive override?

What happened in Evans is that the scheme chosen by Parliament came to be seen by the judges as offensive to the rule of law. Because of this, the judges felt entitled to read down the statute. They reconciled what they were doing with Parliamentary sovereignty by conceding that Parliament could, by using clearer words, have achieved the supposedly offensive result. But this is not a case where the “full implications” of what the words used by the Parliamentary draftsmen had “passed unnoticed” in the democratic process. The legislative history suggests the implications were understood, but the scheme was not regarded as remotely offensive to the rule of law. On the contrary, it was regarded as promoting the rule of law (by introducing qualified legal obligations with regard to freedom of information when there had been none before). So there was no call to use the kind of language that Lord Neuberger thought would be required for the avoidance of doubt.

What are the implications? For the Parliamentary draftsman, Evans will necessitate more convoluted and defensive language, designed to head off hitherto untried legal challenges. For the practitioner, the absence of this kind of language will give the green light to read down otherwise clear provisions; and even when the draftsman has sought to anticipate the reading down, the question will be whether the language selected expressly rules out every possible construction, however implausible. All this will be tremendous fun for those involved, but it strikes me as more like a complicated word game than a genuine exercise in interpretation.

Martin Chamberlain QC

Brick Court Chambers