There was much talk in the early days of the Human Rights Act 1998– and indeed more recently too – by commentators such as Francesca Klug, Richard Clayton QC, Tom Hickman and Aileen Kavanagh of whether the structure in section 3 and section 4 would facilitate what has become known as ‘dialogue’. This summer’s issue of Public Law contains an interesting survey by Po Jen Yap of academic thinking on the topic. As many will no doubt know, the dialogic model was first propounded by Peter Hogg and Alison Bushell in the context of the Canadian Charter as a half-way house between legislative and judicial supremacy. Their research showed that strike down decisions produced a legislative sequel in the vast majority of cases. I don’t propose discussing the normative claims of the dialogic model, nor the views of those who dispute its application and relevance. Instead my focus is its possible relevance to the UK, alongside some wider issues relating to the role of Parliament in the process of guaranteeing rights.

The dialogue model, under constitutional bills of rights, provides a plausible framework for constitutional adjudication between on one hand the extremes of legislative dictat (and the flaw of ballot-box majoritarianism) and on the other of conferring on unelected and unaccountable judges the power to countermand decisions bearing the imprimatur of democracy. In very brief terms, it assumes the legitimacy of the courts’ power to invalidate legislation but if a judge does decide a statute is invalid, it asserts that legislatures have a choice of responses: they might insist that its say, its legislation, be upheld; they might moderate the terms of the rights-violating law in question along lines ‘suggested’ by the judgment; or they might accept the courts as having had the final word. In each though, parliamentarians must heed what the courts have said on the question of compatibility. There is thus dialogue, a constitutional conversation, between the two arms – about the content of a right, about its scope or reach, about its applicability to given facts, or about legitimate intrusions upon rights.

Though clearly different – since in the UK, a court cannot strike down – it was thought that a court declaring legislation to be incompatible would ignite parliamentary debate and response. From their study of 65 cases in the period 1983–1994, Hogg and Bushell concluded that about 80 per cent produced such a response within two years. The problem is that the comparison simply does not stack up. Much of the original theorising about dialogue had been prompted by the question – worry, perhaps? – of who should have the last word. Canada was different to the United States in that the final word did not necessarily rest with the Chief Justice and judges in Ottawa. Provincial or federal legislatures have always been free to reject judicially-imposed meanings about scope or limitations by re-enacting the impugned law (with a sunset provision of five years) using the ‘notwithstanding’ provision of s 33. In the USA, such a consequence would be impossible – a far more cumbersome constitutional amendment is the only alternative. In Canada then the presumptive last word rests with the judiciary. Taking a recent example, if politicians want to overturn the decision of the Supreme Court of British Columbia invalidating the criminal prohibition on physician-assisted death, they must now positively act. Under the HRA, the opposite holds true: a declaration of incompatibility has no effect on the validity, operation or effect of legislation whenever passed. Westminster needs do nothing and its will prevails still. This is, as Conor Gearty out it, the beauty of the Human Rights Act: just as the judicial appears about to triumph, the political re-asserts its control.

All that only obtains if a court decides to declare legislation incompatible. If it decides instead to rely on its power under s 3 to ‘read and give effect’ to legislation that is patently incompatible then Westminster does need to act to return to the status quo ante. If it does not – or, as I would contend, does not realise it needs to – then the effective last word is in the hands of judges, and not simply as a means to deal with the instant case and for those litigants before the court, and not simply prospectively either. The courts have ‘discovered’ the true meaning of a section or a word by altering its clear meaning, denying Parliament’s clear instruction in the original Act (though arguably, of course, giving effect to its clear instruction in the Human Rights Act itself). It is a welcome sign though (see the AHRC/Oxford Human Rights Hub research Parliaments and Human Rights: Redressing the Democratic Deficit) that Parliamentary involvement in human rights protection is on the up, at least through the medium of JCHR reports being referred to during debates. There is some concern that this is not really parliamentary but a small cadre of either committee members or interested die-hards, and that the Commons is far less active than the Lords. There is evidence too that Parliament is reacting reasonably well to s 4 declarations: at the AHRC two-day seminar discussing that research, Jeff King discussed his research findings. Out of 19 s 4 declarations to date, 14 have produced some response, a hit-rate slightly under the 80 per cent reported by Hogg and Bushell.

One would expect declarations under s 4 to be followed up in Parliament. They are a clear judicial sign that something is rotten in the state of human rights. Ministers will have been joined to proceedings by virtue of s 5. It’s also likelier that declarations of incompatibility will make the news whereas transformative readings under s 3 tend not to – or not in the same contextualised way and to the same extent. A non-scientific, non-exhaustive study of contemporaneous UK newspaper reporting of key human rights cases, using Nexis, does seem to bear this out.  There were only three that covered the House of Lords judgment in Mendoza (one in The Times and two in The Daily Telegraph respectively) and all three were little more than Law Reports or summaries. None make any explicit reference to the wide use made of s 3. Similarly, the rape shield decision in R v A in May 2001 made the news only in The Guardian, The Independent and The Telegraph (though there were also a couple of newspaper pieces referring to it later in the year as the HRA approached the  first anniversary of coming into force). By contrast, the House of Lords decision to hold provisions of the Anti-Terrorism, Crime and Security Act 2001 incompatible with the ECHR, in R v A and Others, produced over 120 newspaper hits for the single week from 17-24 December 2004. Clearly, not all of this coverage was concerned with the intricacies of the scheme in ss 3–4, given the importance of terrorism as a narrative, but it is also clear simply from a cursory look that there was certainly much more media discussion surrounding the wider constitutional framework than was ever the case in the reporting of any of those major s 3 cases.

It is instructive too that all the focus is on s 4. There is not so far as I can tell, any work comparable to Jeff King’s being done for the interpretation cases under s 3 for the simple reason that they are difficult to track down. There is nothing similar to Ministry of Justice’s annual report on human rights judgments, including s 4 declarations (the most recent of which was published this week), or the follow-up work of the JCHR. Indeed if one were to read the former, there’d be no way of knowing how widely used – or not – the new interpretative power is. There is simply no mention in the JCHR 2009/10 guidance that ministers inform them, let alone keep track themselves of how s 3 cases have fared. It is hard to see why Parliament, acting either by MPs individually or through the collective strength of the JCHR, should be any less inclined or any less entitled to know, as Lord Millett in Mendoza put it (at [70]), that cat is being read as including cat or dog than they are to know that in Morris discriminatory aspects of the Housing Act 1996 relating to those who became unintentionally homeless were held to have breached the ECHR. The only collation of s 3 readings that I know of in the public domain is that contained in the appendix to Lord Steyn’s speech in Mendoza, way back in 2004. The problem for practitioners, academics and politicians is simply keeping abreast. Given that s 3 can arise and be used in any legal context, it would mean reading the judgments of any court or tribunal (not simply, for s 4, the High Court and above) hoping to spot one. Of course I do not do this – I doubt anyone does – but it does mean I spend a fair bit of time each week skim reading summaries from Lawtel, or relying on colleagues to let me know if one crops up in their own areas of expertise. I keep a running folder in my email inbox – O v Crown Court at Harrow, anyone, reading down provisions relating to bail?  The same is true of the use of s 19. This was used most recently when the House of Lords Reform Bill was introduced in early July as a result of the non-extension of the franchise to prisoners – in clear breach of Hirst and Scoppola. It took a discussion on the UK Human Rights blog to establish that this was not the second time the power had been used but the third though, as Elin Weston pointed out, it arose for the Local Government Bill 2000 (because of the ban on promoting homosexuality in schools) only when the Bill reached the Commons. The only time it had previously been used on presentation to Parliament was the Communications Bill of 2003, given its absolute ban on political advertising. Maintaining and publishing a database of this sort of information is exactly the role of government, surely?

We can see not only this lack of transparency but also a wider problem associated with the use of s 3 in the House of Lords decision in Hammond in 2005 where the procedure for review of mandatory life sentences post-Anderson (in the Criminal Justice Act 2003) were effectively re-written. I am guessing that those with an interest in criminal justice/penal policy might have heard of it – few others will have done. It certainly doesn’t appear in the indices of any of the major texts on the HRA. Despite the legislation being clear – these were to be ‘determined by a single judge of the High Court without an oral hearing’ – the House decided it should be read subject to the following implied condition: a High Court judge has discretion to order an oral hearing, where such hearing is required to comply with a prisoner’s fair trial rights under Article 6. What is intriguing about Hammond is that it did not make it into the list of section 3 readings in the Lord Chancellor’s Review of the Implementation of the Human Rights Act in July 2006 (p 17). This might be because, in a very technical sense it was not a s 3 case at all – the adopted reading had been agreed on by all parties. Correspondence with the DCA indicated the list was drawn up on counsel’s advice bearing that view out… but that supports the case being made here about transparency. Neither an interested lay reader nor an MP worried about Parliament’s carefully crafted words being flouted (reversed in fact) would ever have known of it. That must be a concern. The greater concern though stems from that agreement between counsel. As Lord Hoffmann noted ([29]), the House was not being asked to decide whether such a bold exercise in ‘interpretation’ was permissible. Only two years before, in 2003, Parliament had made its views very clear – clear one must assume in knowledge that the ECHR protects fair trial rights (though I have not gone back to Hansard in writing this piece). There is a real problem with our constitutional fundamentals if, with the consent of the other side in lis pendens with an obvious vested interest in a favourable outcome, the executive is able to have, using Lord Millet’s words again, black read as including not black. That risks subverting the internal boundaries between the two political arms. Does it not offer an incentive to ministers to agree to term X on the floor of the House or in Committee, knowing when it is argued in the courts – at the instigation no doubt of someone who stands to lose by term X being applied – they will simply succumb and agree to term Y instead? There are grave difficulties with the constitutional suitability and propriety of such a course – and who knows in how many other cases it has occurred?

This piece has highlighted some problems with the way sections 3 and 4 work and interrelate. There is a real need for some institutional and mechanical tinkering with the HRA’s scheme of protection, if it is to remain an aspirational template for the parliamentary model. Most obvious is to see that the threat from section 3 is as great as from section 4, albeit of a different flavour. Courts should not be able to allow the scheme to be subverted in the manner of Hammond. Irrespective of possible agreement between counsel, s 5 might be amended so that it covers not simply the Crown for s 4 but (as occurred in Wilson (No 2) in relation to the use of Hansard) entitles the Speaker too to intervene whenever a court is considering using s 3. Secondly, sufficient resources should be available to create a live database of s 3 readings, populated with data from courts themselves. Thirdly, it should be a requirement under the HRA that a responsible Minister report formally to Parliament what course of action is planned in response to a declaration of incompatibility. That would make more concrete the recommendation of the JCHR in 2009/10 in its Guidance for Departments that ministers report to it within 14 days. It would thus also complement the requirement in section 19 during the legislative process, though as an aside an amendment there would ensure on-going statements during passage and amendment not simply on one day, at second reading and not simply for government-sponsored bills either. Last, either through s 19 or otherwise, there needs to be an ongoing commitment by the executive to consider the human rights implications of legislation it proposes and to make parliament aware of its concerns and its position. This means much more than that the minister simply make a bland statement one way or the other. Change does seem afoot here; we might contrast the relatively lengthy discussion of relevant ECHR principles in the Explanatory Notes to the Defamation Bill with the paucity of analysis in the those accompanying SOCPA 2005 assessing the Strasbourg position on demonstrations around Westminster. This tallies, as Elin Weston pointed out in the blog discussion above, with Ken Clarke’s indication in evidence to the JCHR that he would expect to see free-standing Human Rights Memorandums with perhaps greater regularity (q 32). More food for thought for the Bill of Rights Commission perhaps?

David Mead is a Senior Lecturer in Law at the UEA Law School.

Suggested citation: M. Mead, ‘Talking about dialogue’ UK Const. L. Blog (15 September 2012)(available at http://ukconstitutionallaw.org).