A limited case for the closed material procedure: natural justice, open justice and the clear advantage variation

Abstract

The closed material procedure (CMP) – ever since its introduction to English law – has been subjected to a very significant amount of academic criticism. But over time, the CMP has become increasingly settled as a fixture in English law. Whilst the existence of the CMP per se in English law seems settled, the extent of its deployment is not. Given this development, it seems important and constructive to examine whether – and the conditions under which – a CMP can ever be normatively justified, all things considered. Two propositions will be made. First, a common argument for the CMP – the maximising argument – does not demonstrate that the CMP is normatively justified, all things considered, for it does not sufficiently mitigate the two main objections to the CMP, based respectively on the principles of natural justice and open justice. Secondly, where the clear advantage variation is deployed – ie when a CMP allows the excluded party to make use of material that: (a) clearly advantages him; and (b) would otherwise be unavailable for the court's consideration – both objections are sufficiently mitigated. In such a case, the CMP is normatively justified, all things considered. This constitutes a limited normative case for the CMP.

Introduction

This investigation asks: can the closed material procedure (CMP) ever be normatively justified, all things considered? The CMP refers to a set of processes ‘where the court relies on evidence that is not shown to one or more of the parties (closed material)’,¹ and instead ‘disclosure of such closed material is made to special advocates’² – who will represent the interests of the excluded party.³ Three points should be added to this definition. First, although the excluded party may attempt to communicate with the special advocate throughout, the special advocate will only be ‘allowed to communicate with the [excluded party]’ before he has seen the closed material; the excluded party will thereafter be reduced to correspondence without any replies.⁴ Secondly, even where a CMP is deployed, it rarely encompasses the entire case. One will normally expect the court to splinter the proceedings into two, with one part held under the conventional set of adversarial procedures and the part concerning the closed material held under a CMP.⁵ Thirdly, although the special advocate is appointed formally through government processes, he is independent of the UK government⁶ and is expected to act on behalf of the excluded party in a professional manner that best safeguards his interests.⁷

Since its introduction to English law post-Chahal,⁸ the CMP has been subjected to a very significant amount of academic criticism. The academic criticism can be nuanced: critics may accept that in particular circumstances, at least, the CMP may bring with it certain benefits; but the predominant thrust of their criticism is that such benefits are insufficient to normatively justify the CMP, all things considered, given its downsides.⁹ Despite this, the CMP has gradually taken root within English law¹⁰ – through a combination of both legislative¹¹ and common law developments.¹² Especially given its general compatibility with the European Convention on Human Rights,¹³ it seems fair to predict that the CMP is here to stay.¹⁴ The relevant question is no longer whether we should adopt the CMP or not, but the extent to and conditions under which we adopt the CMP¹⁵ in lieu of the conventional set of adversarial procedures.¹⁶ This will be the more realistic question to ask for English law. But if we accept the CMP to any extent as normatively justified, all things considered, this must be argued for: it cannot be right to countenance the CMP simply because it is now a fixture in English law.¹⁷ And whilst the critics have acknowledged some of the CMP's potential benefits,¹⁸ few have gone further to clarify when one may regard the CMP as being normatively justified, all things considered – and if so, why this would be possible in light of the existing arguments against it.¹⁹ This investigation aims to establish a limited case for the CMP: to demonstrate why the CMP can be normatively justified, all things considered, at least in a more limited range of circumstances.

It will be contended that a subset of the CMP can be isolated from the rest. Call this subset the ‘clear advantage variation’. The clear advantage variation allows the court to consider material that: (a) clearly advantages the excluded party; and (b) would otherwise be unavailable for the court to consider. A CMP held when these conditions are satisfied is not only prima facie normatively appealing, but also properly meets the major objections to the CMP. This allows us to see the CMP as normatively justified, all things considered, to this extent at least. This paper is certainly not the first to suggest an argument for the CMP along this line,²⁰ although the pre-existing attempts have, in my view, been relatively undeveloped and leave open several important objections before they can establish why the CMP is normatively justified, all things considered, to this extent. To my knowledge, this investigation will be the first in the academic literature to attempt a detailed defence for the CMP along this line. The argument will proceed in two parts. First, the benchmark for judging when an argument for the CMP is successful will be established by reference to Lord Dyson's objections in Al Rawi.²¹ Using this benchmark, I will critique the maximising argument, an argument commonly made in favour of the CMP.²² Secondly, it will be contended that the clear advantage variation does properly overcome Lord Dyson's objections, and is accordingly normatively justified, all things considered.

Before we proceed further, two caveats ought to be noted. First, the current suggestion is only that a CMP will be normatively justified, all things considered insofar as we are concerned with the clear advantage variation. I do not preclude altogether the possibility that other arguments can justify applying a CMP to a wider range of circumstances. I am neither for nor against this possibility. It is a part of this caveat that I will not address the general merits of Lord Dyson's objections to the CMP in Al Rawi: the point is instead that, assuming their general validity, they can be sufficiently met by the clear advantage variation. Secondly, this paper only addresses the CMP as deployed in: (a) civil proceedings; and (b) English law. Entirely different considerations might apply, once one considers criminal proceedings²³ and using the CMP beyond English law.²⁴ The latter point is noteworthy, since the CMP is regarded as objectionable to the principles of open justice and natural justice, which are both well-established in English law.²⁵ These principles are also underpinned in part by premises closely affiliated with the liberal democratic tradition, which English law embraces.²⁶ But these features may not resonate – at least to the same extent – in other jurisdictions. I am not here committed regarding whether and how the current argument must be modified when one travels beyond English law and beyond civil proceedings.

1. Lord Dyson's objections and the maximising argument

(a) The applicable benchmark: Lord Dyson's objections in Al Rawi

It is important to lay down a benchmark for a successful argument that the CMP is normatively justified, all things considered. I begin with the point that it is prima facie normatively justified that all materials relevant to the issues to be decided upon by the court should be available for the court's consideration.²⁷ This point is well-established in English law, well before the CMP phenomenon. It explains the fundamental importance of disclosure in civil litigation,²⁸ for ‘there is the public interest that the administration of justice shall not be frustrated by the withholding of documents which must be produced if justice is to be done’.²⁹ It also explains two entrenched doctrinal features within the law of public interest immunity (PII): PII only serves as an exception to the general obligations of disclosure;³⁰ and the court should consider whether ‘the aspect of the public interest which reflects the requirements of the administration of justice outweighs the aspect of the interest which is against disclosure’ – so when a claim of PII fails, ‘it is the public interest which requires disclosure’.³¹ This point is simply – to use some administrative law parlance – a reflection of the broader point that a decision-maker should prima facie consider all materials relevant to the decision at hand.³² So we have this rather self-evident starting point: if one casts aside the disadvantages of the CMP – discussed below – it is prima facie normatively justified, by ensuring that all the relevant materials can be considered by the court.³³

I suppose few critics can disagree with this normative starting point. I suggest the same will be true if – instead of casting them aside – one can present an argument to demonstrate why such disadvantages will have been entirely defeated or sufficiently mitigated, when a CMP is being deployed. So it is this negative question about the CMP that is controversial: are there arguments that demonstrate why the CMP is not normatively justified, where such arguments are neither entirely defeated nor sufficiently mitigated? As further specified below, the core objections presented in this regard – in both the academic literature and the case law alike – are as follows: the CMP violates ‘the fundamental principles of open justice and of natural justice’ in English law.³⁴ A successful defence of the CMP is correspondingly one that entirely defeats – or at least sufficiently mitigates – these objections. This is then the benchmark for a successful defence of the CMP – to show that the CMP is normatively justified, all things considered.

To apply this benchmark, we should consider these objections more carefully.³⁵ The first objection is that the CMP detracts from the principle of natural justice.³⁶ As Lord Dyson held in Al Rawi:

[T]rials are conducted on the basis of the principle of natural justice … A party has a right to know the case against him and the evidence on which it is based. He is entitled to have the opportunity to respond to any such evidence and to any submissions made by the other side. The other side may not advance contentions or adduce evidence of which he is kept in ignorance.³⁷

The leading case law on the principle of natural justice suggests that it is underpinned by a number of rationales – including to enhance the accuracy of the decision-making process³⁸ and to respect the dignity of the parties involved.³⁹ These aims are not mutually exclusive, and may potentially apply alongside each other.⁴⁰ Through this lens, critics may suggest that the CMP is normatively problematic: it detracts from the principle of natural justice, and this justifiably overcomes the abovementioned normative starting point.⁴¹

We can see the critics’ point more clearly by considering how the CMP can interact with the two rationales underlying the principle of natural justice. Insofar as the principle of natural justice helps ensure accurate decision-making, critics may suggest that a CMP can undermine this by limiting the excluded party's input on the closed material.⁴² Since the excluded party cannot access the closed material, he cannot readily challenge it directly; and although the special advocate is allowed to view and directly challenge the closed material, he cannot take specific instructions from the excluded party.⁴³ For instance, Jackson invites us to consider ‘a source who has a grudge against the [excluded party]. But unless the special advocate can disclose the source's identity to the [excluded party], he may never discover that the grudge exists. How can the special advocate then act in the interests of the [excluded party] if he cannot find out what needs to be represented in his interests?’⁴⁴ Call this the accuracy limb of the natural justice objection.

There is this further point from the critics. By denying the excluded party the opportunity to have his say within the CMP, his views will be to that extent neglected; coupled further with a conception of human dignity that requires the views of the excluded party to be respected,⁴⁵ his human dignity will also be to that extent disrespected.⁴⁶ So if (per Waldron) ‘[a]pplying a norm to a human individual is not like deciding what to do about a rabid animal or a dilapidated house’ and ‘involves paying attention to a point of view and respecting the personality of the entity one is dealing with’ (and the principle of natural justice therefore ‘respect[s] the dignity of those to whom the norms are applied as beings capable of explaining themselves’),⁴⁷ the respect for the excluded party's human dignity will be thereby undermined by the CMP.⁴⁸ Call this the dignity limb of the natural justice objection.

The second objection is that a CMP detracts from the principle of open justice.⁴⁹ The point was again expressed by Lord Dyson in Al Rawi: the principle of open justice requires that ‘trials should be conducted and judgments given in public’; this is ‘not a mere procedural rule’, but ‘is a fundamental common law principle’.⁵⁰ The critics have accordingly suggested a CMP's conflict with the principle of open justice renders the former normatively problematic – thus overcoming the abovementioned normative starting position.⁵¹ We can see their point more clearly by examining how a CMP can interact with two important underlying rationales for the principle of open justice.⁵²

For the critics, whilst the principle of open justice is already important enough in a typical case,⁵³ this is particularly important in cases involving closed materials. Because such materials are by definition withheld to safeguard the public interest,⁵⁴ one can readily envisage a case where the materials concern important issues engaging the public interest – for which political actors are expected to be held accountable in a liberal democratic political system.⁵⁵ But if such materials are considered within a CMP, an important facet of such accountability will be lost: whilst the courts can adjudicate on the issues involving those materials, the general public and other political actors will not be informed of the same.⁵⁶ And by withholding the closed materials in this way, an impugned political actor can significantly constrict accountability over his actions.⁵⁷ A CMP can, therefore, detract from an important function of the principle of open justice: that ‘[i]n litigation, particularly litigation between the executive … and the citizen, the principle of open justice represents an element of democratic accountability’.⁵⁸ Call this the accountability limb of the open justice objection.

Critics may also find the detraction from the principle of open justice problematic, for undermining public confidence in the courts.⁵⁹ Consider Lord Phillips’ suggestion in AF: ‘if the wider public are to have confidence in the justice system, they need to be able to see that justice is done rather than being asked to take it on trust’.⁶⁰ Generally speaking, when political actors are allowed to act with secrecy, other parties in the political community are reduced to a position of blind trust;⁶¹ and frequently, limited argumentation would be available to demonstrate to the others why such trust should be afforded.⁶² This can also be true of the courts, when they do not act openly.⁶³ And although a connection between transparency and public confidence may perhaps need further empirical support,⁶⁴ the critics find this a particularly serious concern whenever a CMP is applied.⁶⁵ For if (per its critics) applying a CMP involves committing an injustice against the excluded party,⁶⁶ the courts will by endorsing the procedure be prima facie committing an injustice in the public's view – whilst being simultaneously hindered from regaining that confidence from the public, by demonstrating that the courts will continue to act professionally and responsibly from within a CMP.⁶⁷ Call this the confidence limb of the open justice objection.

A number of judges and academics have criticised the CMP using the abovementioned lines of reasoning;⁶⁸ I am simply restating their criticisms to offer a clearer notion of the applicable benchmark. But these points were most famously made in Lord Dyson's judgment in Al Rawi. Although Al Rawi itself featured a number of distinct lines of reasoning – in particular from Lord Mance's and Lord Clarke's judgments⁶⁹ – the later cases have come to see Lord Dyson's reasoning as predominant.⁷⁰ I will accordingly, for the sake of simplicity, refer to Lord Dyson's objections to the CMP as the ones that demonstrate why the CMP is not normatively justified, all things considered, according to its critics. The later academic literature has supplied further analytical details to substantiate Lord Dyson's objections: the objection based on the principle of natural justice has two limbs (ie the accuracy and dignity limbs), and the objection based on the principle of open justice also has two limbs (ie the accountability and confidence limbs). I will regard these limbs as being supplementary to Lord Dyson's objections. Per the earlier discussion, the benchmark for adjudging the success of this investigation lies in whether it can establish – in certain circumstances at least – that Lord Dyson's objections to the CMP can be entirely defeated or sufficiently mitigated. If so, the CMP will to that extent not only be prima facie normatively justified – but normatively justified, all things considered.

(b) The maximising argument

Consider here an argument commonly taken to support the CMP: the CMP is normatively justified because it maximises the extent to which all the relevant evidence is to be presented to the court, which best positions the court to adjudicate justly upon the case.⁷¹ Call this the maximising argument. I suggest that this argument does not meet the benchmark,⁷² and explaining why this is so will be significant for understanding why the clear advantage variation does meet the benchmark, as explained below.

The maximising argument invites us to consider the alternatives to a CMP. Before the CMP is available under English law, materials that cannot be disclosed in the public interest are usually considered through the law of PII.⁷³ What is critically different from a CMP is that materials excluded by PII cannot be used by any party to the proceedings.⁷⁴ Obviously, where the materials excluded are significant for either party's case in the proceedings, that party will be disadvantaged by their exclusion.⁷⁵ Equally, PII excludes relevant materials from the court's consideration, and will to this extent be departing from the abovementioned normative starting point.⁷⁶ In contrast, a CMP allows the court to consider such materials.⁷⁷ In Tariq, Lord Mance adopted the maximising argument: he suggested that the alternatives to adopting a CMP would be to exclude evidence crucial to the case of either party,⁷⁸ and ‘[t]he rule of law must, so far as possible, stand for the objective resolution of civil disputes on their merits by a tribunal or court which has before it material enabling it to do this’.⁷⁹ The bite of the maximising argument thus lies in how it avoids adjudication based only on having a partial picture before the court – so that the issues can be properly resolved ‘on their merits’.⁸⁰

Essentially, the maximising argument attempts to take the abovementioned normative starting point as a conclusion in itself: ie to contend that the starting point of itself already sufficiently mitigates Lord Dyson's objections to the CMP, and thus demonstrates that the CMP is normatively justified, all things considered. But this is not convincing. Consider first the natural justice objection. Applying the dignity limb of the objection, a critic can insist that there is more to maximising the extent to which relevant materials are put before the court.⁸¹ For instance, the critic can suggest that the court ought to: (i) adjudicate on the basis of all relevant materials, so as to aim at arriving at an accurate outcome; and (ii) in so doing comply with relevant moral limitations,⁸² one of which will be to respect the parties’ human dignity.⁸³ And even when the court decides correctly by meeting (i), its failure to meet (ii) is nevertheless normatively objectionable:⁸⁴ a critic may thus maintain the natural justice objection, in the face of the maximising argument.⁸⁵ So even taking Lord Mance's point that the alternative to deploying a CMP will be to detract from (i),⁸⁶ this does not answer the objection based on (ii). And short of an argument for why (i) is more important than (ii), it is unclear that the maximising argument can sufficiently mitigate the dignity limb of the natural justice objection.

This maximising argument also remains vulnerable to the accuracy limb of the natural justice objection: maximising the materials considered under a CMP does not entail ‘maximis[ing] [the] chances of arriving at the correct decision’,⁸⁷ as such materials may not be reliable.⁸⁸ One can again recall Jackson's example – where the closed material is based on ‘a source who has a grudge against the [excluded party]’, but this is not known to the special advocate.⁸⁹ The critic may remind us that this fact will also not be before the court. Let us assume that – due to this grudge – the piece of evidence is in fact an unreliable guide to the truth. Quite arguably, a court that takes into account this unreliable guide to the truth is more prone to deciding in a way that is substantively inaccurate than a court that is not aware of this unreliable guide to the truth. The critic can then already offer a counter-example: the deployment of a CMP can – even if one leaves aside (ii) – become objectionable, simply considering (i) alone.⁹⁰

Consider also the open justice objection. Both the accountability and confidence limbs of this objection share a common trait: they travel beyond the immediate case before the court. If the closed materials concern subject-matter X, the deployment of a CMP may sacrifice the general ability of the other actors to hold the relevant political actor accountable in respect of subject-matter X – both in and beyond the immediate case before the court.⁹¹ And if deploying a CMP will undermine public confidence in the courts, it is difficult to imagine that this effect will be ‘quarantined’: ie the general public remains as confident as before in the court for all other cases – and indeed, in respect of the open part of the immediate case to which the CMP applies – such that the confidence is only undermined in a limited part of the immediate case. One would instead expect public confidence in the court system to be thereby generally undermined – whether or not other courts are deploying a CMP – if the objection is itself sound.

There is accordingly a mismatch in the scope of the maximising argument and the open justice objection. Let us assume arguendo that the importance of maximising the consideration of all relevant evidence overrides the importance of achieving open justice, when only the immediate case before the court is being considered; let us further assume arguendo that deploying a CMP facilitates a more accurate outcome, in resolving the immediate case concerned. But even if these assumptions hold true – and the validity of either assumption seems at least open to doubt – it is rather difficult to show that arriving at the accurate outcome for the immediate case overrides the importance of achieving open justice generally, the benefits of which will extend far beyond that immediate case and will indeed resonate across the legal system. It seems that this trade-off – of one for the many – cannot satisfactorily explain why a CMP should be preferred, the open justice objection notwithstanding. Accordingly, the maximising argument does not meet the stated benchmark: it does not sufficiently mitigate Lord Dyson's objections to the CMP.

2. The clear advantage variation

(a) Elaborating on the clear advantage variation

A CMP constitutes the clear advantage variation, if and insofar as it allows the excluded party to make use of material that: (a) clearly advantages him; and (b) would otherwise be unavailable for the court's consideration. This subsection aims to clarify these conditions.⁹² It is more convenient to begin with condition (b). A number of important cases concerning the CMP suggest that even if a CMP ultimately applies, it does not entirely displace the need to apply the law of PII. Instead, the question is whether materials that would otherwise have been excluded under the law of PII may nevertheless be taken into account by the court applying a CMP.⁹³ This is usually how condition (b) is satisfied. A further possibility is the controversial⁹⁴ case of Carnduff,⁹⁵ which permits ‘a stay of proceedings’ if ‘it is impossible for the case to be fairly tried’, where crucial material concerning the case cannot safely be admitted for the court's consideration.⁹⁶ It seems unclear whether this should be seen as truly distinct from the law of PII itself: whilst Lord Mance and Lord Clarke in Al Rawi saw Carnduff as a possibility after PII applied,⁹⁷ some other authorities have suggested that it might apply without PII being at play,⁹⁸ and Heaton has also criticised the majority in Carnduff for omitting to apply the law of PII before staying the case.⁹⁹ Regardless of how we should conceive it, Carnduff remains good law today¹⁰⁰ and thus offers a possibility for satisfying condition (b). Condition (b) is also satisfied where the material simply need not be disclosed, because it lacks relevance to the immediate case.¹⁰¹ But in such a case, condition (a) will by definition not be satisfied alongside condition (b), as such material cannot (by reason of its irrelevance to the immediate case) advantage the excluded party, let alone clearly so. This possibility can thus be disregarded here.

Let us then turn to condition (a): what is meant here by an advantage to the excluded party, let alone a clear one?¹⁰² It is important to note that where rule 31 of the Civil Procedure Rules 1998 applies,¹⁰³ a CMP can only apply to material that ‘is disclosable’ under ‘the ordinary principles that apply to disclosure’; the issue of the CMP is irrelevant when no disclosure is required of the material.¹⁰⁴ The English law of disclosure is complex in modern times,¹⁰⁵ but a good starting point will be rule 31.6 of the Civil Procedure Rules 1998.¹⁰⁶ It provides:

‘Standard disclosure requires a party to disclose only—

1. (a) the documents on which he relies; and

2. (b) the documents which—

   1. (i) adversely affect his own case;

   2. (ii) adversely affect another party's case; or

   3. (iii) support another party's case; and

3. (c) the documents which he is required to disclose by a relevant practice direction’.¹⁰⁷

Rule 31.6 should be read alongside a long-standing principle in English law, expressed by the House of Lords in Air Canada:¹⁰⁸ that ‘[i]n an adversarial system such as exists in the United Kingdom, a party is free to withhold information that would help his case if he wishes … He cannot be compelled to disclose it against his will’.¹⁰⁹ This means ‘that a party who seeks to compel his opponent, or an independent person, to disclose information must show that the information is likely to help his own case’¹¹⁰ – in that ‘the material sought be such as is likely to advance the seeker's case, either affirmatively or indirectly by weakening the case of his opponent’.¹¹¹ It is, perforce, insufficient for a material to be relevant in an unspecified and muddled sense, before disclosure of it may be compelled.¹¹² It follows that when a material must indeed be disclosed, there will be a specified basis as to why this is so.¹¹³ And as explained above, this must be answered as a logically prior step in every case involving a CMP where CPR rule 31 applies.¹¹⁴

Consider then a case where rule 31.6 is specifically applicable, before a CMP arises.¹¹⁵ The foregoing analysis means we can always interrogate whether a material (for which a CMP is contemplated) has to be disclosed on the basis that it ‘support[s] another party's case’¹¹⁶ and, more specifically here, the excluded party's case. We can similarly question whether the said material has to be disclosed, on the ground that it ‘adversely affect[s] [the UK government's] own case’.¹¹⁷ I suggest that this offers to us a familiar, and indeed analytically prior, test of what constitutes an advantage to the excluded party, when r 31.6 already applies to the immediate case – before the question of a CMP arises. Consider further cases where although rule 31 does apply, rule 31.6 does not specifically apply.¹¹⁸ Even in such cases, it should be stressed that rule 31.6 contains a statutory test that has been designed to apply to a broad range of civil litigation in lieu of Peruvian Guano;¹¹⁹ its versatility and familiarity to judges and practitioners alike can still offer us a helpful definition of ‘advantage’ under condition (a), whether rule 31.6 has been applied to the immediate case or not. Nothing precludes us from meaningfully asking the same statutory questions.¹²⁰ I will therefore define ‘advantage’ under condition (a) as when the material itself has satisfied r 31.6(b)(i) or (iii), or at least will have done so if r 31.6 does not apply to the immediate case.

We must then turn to examine when there is a clear advantage, under condition (a). A reader may initially find this baffling: for why must the material in question not only offer an advantage to the excluded party, but also that the advantage be clear? This qualification is inserted to cater for cases where a material is disclosable on multiple grounds. For instance, a material relied upon by the UK government may simultaneously assist the excluded party's case, and so satisfy both r 31.6(a) and (b). In such cases, the material will offer an advantage to the excluded party in the sense defined above. But the advantage to him is nevertheless not clear, for the material may also be highly prejudicial to him. The argument for the clear advantage variation below relies on how the excluded party is – in common parlance – simply better off than if the material is excluded from the court's consideration. One cannot say that the same is true when the material works to the advantage of both litigants, and this should be reflected in how condition (a) is itself defined. To express this point with greater precision, we can draw from the administrative law lexicon: the advantage occasioned by the material to the applicant must be ‘sufficiently clear and unambiguous’,¹²¹ and this will not be the case where the litigants may plausibly offer ‘competing’ reliance on the same material to positively support their respective cases.¹²²

One must also elaborate on the institutional question: how and who should decide whether the two conditions for the clear advantage variation are satisfied?¹²³ One issue may be that the excluded party will by definition lack knowledge of the contents of the impugned material, and so cannot meaningfully contest whether these conditions are satisfied;¹²⁴ it also may be questioned whether it is for the courts to decide this point.¹²⁵ Here the law of PII offers us a helpful reference point for addressing both issues. In the law of PII, an important point concerns whether or not the courts should inspect the impugned material before ruling on the PII claim.¹²⁶ In Air Canada, Lord Fraser held:

[I]n order to persuade the court even to inspect documents for which public interest immunity is claimed, the party seeking disclosure ought at least to satisfy the court that the documents are very likely to contain material which would give substantial support to his contention on an issue in the case, and that without them he might be ‘deprived of the means of … proper presentation’ of his case.¹²⁷ (Emphasis added)

This point was affirmed by Lord Mance in Al Rawi, where he cited with approval Air Canada and suggested: ‘[w]here a prima case of public interest immunity is made out, a party who wishes to invite the court to inspect material before determining whether it should after all be deployed must show that it is likely to give substantial support to his or her case’.¹²⁸ It will then be the court that determines whether the Air Canada test is satisfied.¹²⁹ Notice in particular two features of the Air Canada test:

1. (a) The court will essentially ask whether the material advantages a particular litigant, which naturally can attract disagreement between the litigants on this precondition.¹³⁰

2. (b) The Air Canada test has been criticised by Zuckerman as too stringent for the litigant seeking disclosure, particularly since ‘the party seeking disclosure is ignorant of the contents of the documents’.¹³¹ But Zuckerman's point underscores precisely why the Air Canada test would offer an apt analogy here, as the excluded party will also lack knowledge of the materials upon any contest concerning the two conditions.¹³²

I suggest that both these features of Air Canada have clear parallels with the current institutional question: how and who is to determine whether the conditions for the clear advantage variation are satisfied? For this question shares with the Air Canada test a highly similar core: how and who should determine whether a said material advantages one litigant, when that very litigant will – by reason of his lack of knowledge – have very limited possibilities to demonstrate this? It is true that the case law generally resists drawing a parallel between the law of PII and the CMP.¹³³ But this disanalogy is predicated on the assumption that a CMP – when held – will violate the principles of natural justice and open justice, whereas applying the law of PII will not;¹³⁴ and as explained below, the clear advantage variation is not objectionable by reference to these very principles. Once this point has been dismissed, little remains to explain why – given the clear parallels that exist – the institutional question should be answered differently. That is, it seems difficult not to apply an analogous approach here, short of overruling Air Canada itself. I will therefore suggest that, by analogy to Air Canada, it will be for the court to determine whether the two conditions of the clear advantage variation are satisfied; and this is despite the fact that the excluded party will be in a limited position to contest this.¹³⁵

A final point should be added. The argument below will only hold insofar as the two conditions for the clear advantage variation are satisfied; so if a CMP is held in reliance upon this current argument, both conditions must be continuously satisfied by the said CMP.¹³⁶ This will thus require the court to continuously assess if this is the case throughout the CMP. This point will no doubt remind many readers of the Justice and Security Act 2013, section 7(2), which posits that ‘[t]he court must keep the declaration [for a CMP under the JSA 2013] under review, and may at any time revoke it if it considers that’ a statutory precondition for authorising the CMP no longer applies.¹³⁷ A similar requirement will apply here: if a CMP is held solely in reliance upon the current argument, the courts must continuously review this matter throughout the litigation – and cease applying the CMP whenever either of the two conditions is no longer applicable. This will likely offer significant safeguards for the excluded party, since a faithful application of this requirement will preclude the possibility of material prejudicial to the excluded party being used within a CMP, solely relying upon this argument.

The strict insistence on these two conditions being met may lead to questions about whether the clear advantage variation may even be useful. Why would the UK government even wish to apply for a CMP? But this objection cannot be sound. When the two conditions are met, the material is by definition of significance to the case at hand. The consequence of ignoring such material will be to detract from the judicial aim of attaining ‘the objective resolution of civil disputes on their merits’,¹³⁸ the general importance of which is underscored both by the normative starting point and the maximising argument,¹³⁹ the latter's flaws notwithstanding. Assuming that the UK government acts in good faith when conducting its litigation (and very strong evidence must be available to demonstrate the contrary),¹⁴⁰ this too will be aligned with the UK government's overarching incentives in litigating, which must extend beyond simply ‘winning every case it fights’.¹⁴¹ So this strict approach towards the clear advantage variation does not deprive it of all meaningful utility.

(b) Meeting the benchmark

Now that the clear advantage variation has been elaborated upon, we may consider whether it is normatively justified, all things considered. Its intuitive appeal is that it is directly concerned with advantaging the excluded party.¹⁴² This distinguishes it from when a CMP is deployed under the maximising argument. In the latter case, there is an important sense in which the excluded party is being sacrificed for other purposes: what matters is that all the relevant materials are placed before the court,¹⁴³ and in a way that accommodates the governmental need for secrecy;¹⁴⁴ although the excluded party may be thereby disadvantaged, this all is a worthwhile trade-off.¹⁴⁵ So in Tariq, Lord Hope spoke of a ‘balance to be struck’: although the ‘[excluded party] will be at a disadvantaged if the closed procedure is adopted’, ‘the disadvantage to the Home Office is greater’ – such that ‘the balance is in favour of the Home Office’.¹⁴⁶

This balance evidently does not necessarily favour the excluded party: the need for government secrecy can become much weightier.¹⁴⁷ And a critic may contend that tension exists between this approach and ‘the … Kantian principle that individuals are ends and not merely means; they may not be sacrificed or used for the achieving of other ends without their consent’.¹⁴⁸ But within the clear advantage variation, the excluded party is not neglected or diminished just because some other matters outweigh his importance. Instead, the clear advantage variation is dedicated to clearly advantaging his position: it provides him with weaponry to fight his case, when the weapons would not otherwise have been available.¹⁴⁹ The Kantian principle does not appear to have been breached: the excluded party has not been sacrificed for others’ ends; rather, he has been facilitated in ‘being able to set and pursue [his] own ends’,¹⁵⁰ at least when compared with the position if the clearly advantageous material is not made available at all.

This point about the CMP's intuitive appeal – that its use may advantage the excluded party – is not entirely novel; it has been expressed on some earlier occasions, when a CMP has been sanctioned by the court.¹⁵¹ But mere intuition clearly cannot suffice to meet the abovementioned benchmark; this investigation seeks to establish this with detailed theoretical care. Let us, therefore, begin by considering the natural justice objection. The critic may object that even within the clear advantage variation, the principle of natural justice will also be violated: it remains that part of the contentions and evidence considered by the court will not be known by the excluded party.¹⁵² This is true. Nevertheless, a critical difference exists here. Within the clear advantage variation, the closed material will not otherwise have been available to the excluded party. In such a case – given that the material is of clear advantage to the excluded party – a very significant contrast emerges:

1. (a) Scenario 1 (if a CMP is unavailable): The excluded party will not be heard on his arguments based on this clearly advantageous material, since it will be disregarded altogether by the court.¹⁵³

2. (b) Scenario 2 (if a CMP is available): The excluded party's interests will be represented by the special advocate concerning this clearly advantageous material,¹⁵⁴ and the special advocate will be able to make submissions for the excluded party. Such material – alongside these submissions – will be heard by the court.

When these two scenarios are compared, scenario 2 seems positively less objectionable than scenario 1 – even from the perspective of the principle of natural justice.¹⁵⁵ For, in fact, the excluded party will have more of a say – albeit indirectly – before the court on the materials that clearly advantage him.¹⁵⁶ The dignity limb of the natural justice objection posits that complying with the principle of natural justice is normatively important, as it (as explained) accords respect to the human dignity of the excluded party: and yet the respect accorded to the excluded party in scenario 2 does not seem any less than that accorded to him in scenario 1. The accuracy limb of the natural justice objection posits that the principle of natural justice is important for ensuring the accuracy of decision-making, as explained above: but this too seems likely to be better achieved in scenario 2 than in scenario 1.

Crucially, in scenario 1, materials clearly advantageous to the excluded party's case – and thus necessarily relevant to the accurate determination of the case – will be neglected by the court.¹⁵⁷ But in scenario 2, such materials will be considered by the court; and in addition, they will be subjected to adversarial argumentation based thereon.¹⁵⁸ The special advocate will not receive instructions from the excluded party on such materials,¹⁵⁹ but such materials are – by definition – clearly advantageous to the excluded party. So the worst that can happen is that such materials will be entirely disavowed, since the special advocate cannot take instructions from the excluded party. But even if so, the excluded party will be in no worse a position than in scenario 1. And where such closed materials are accorded any force by the court – which is far from unrealistic¹⁶⁰ – the excluded party will be in a positively better position than in scenario 1. Consider these points too from the perspective of the court. In scenario 2, the court will also have received adversarial argumentation, on materials that are by definition relevant to the accurate adjudication of the case – whereas this would not have been possible under scenario 1.

These points provide a potent answer to both limbs of the natural justice objection, as it can no longer explain why scenario 2 is preferable to scenario 1.¹⁶¹ Although the clear advantage variation breaches the principle of natural justice, the latter's underlying rationales are in fact better served than otherwise.¹⁶² I therefore consider Lord Dyson's natural justice objection sufficiently mitigated. Two further points buttress this conclusion. First, this argument cannot be used by a proponent of the maximising argument, for in such a case, the closed materials may not favour the excluded party's case, and so the excluded party may not be better served by invoking a CMP than if the material is simply excluded. The foregoing logic therefore no longer applies. This argument therefore does not prove too much: it does not contradict the earlier critique of the maximising argument.

Secondly, can the critic suggest that scenario 3 exists: that the excluded party can argue explicitly based on the material in question, in the open? The critic may suggest that compared with scenario 3, scenario 2 still clearly places the excluded party at a disadvantage; the clear advantage variation is accordingly problematic.¹⁶³ But this is not what the entire discussion about the CMP is or has even been about; the discussion has always assumed that the alternative to adopting a CMP is that the material will be excluded from the court's consideration, on either PII or Carnduff grounds.¹⁶⁴ Whilst Carnduff has been regarded with more academic scepticism,¹⁶⁵ the same is not true of PII itself: few have questioned that the courts may thereby exclude materials if a CMP is unavailable.¹⁶⁶ Given this, scenario 3 simply will not arise. The clear advantage variation is therefore ‘mitigating’ in nature – by offering a better solution to the alternatives that will otherwise apply for the excluded party;¹⁶⁷ in this regard, ‘[i]t can only enhance the rights of [the excluded party]’.¹⁶⁸ This is quite different from a situation where without any CMP, ‘the material would have [otherwise] been disclosed’ (eg where PII is inapplicable);¹⁶⁹ the use of a CMP then will not satisfy condition (b).¹⁷⁰

Let us then turn to the open justice objection: does the clear advantage variation at least sufficiently mitigate it? As explained above, the problem that the maximising argument faces is one of scope: it cannot satisfactorily explain why the accurate determination of the immediate case warrants undermining accountability and public confidence in the courts generally. The critic may similarly criticise the clear advantage variation. Once more, the clear advantage variation does violate the principle of open justice.¹⁷¹ But there is also a significant difference here. Contrast once more scenario 1 and scenario 2: either way, the impugned material will not be revealed to the general public and other political actors.¹⁷² And given the irrelevance of scenario 3, it has never been a realistic option that the material will be dealt with in open court.¹⁷³ So if secrecy in dealing with the materials in question will – per both limbs of the open justice objection – entail a reduction in accountability and public confidence in the courts, it will occur either way. The clear advantage variation cannot worsen the situation. From this we may already conclude that neither limb of the open justice objection presents a real normative obstacle to the clear advantage variation: for whilst the principle of open justice is indeed violated by the clear advantage variation, its underlying rationales are not undermined by its use.

Taken at face value, the foregoing argument may apply to justify both the clear advantage variation and the maximising argument alike. But some further, important distinctions exist between the two – vis-à-vis the open justice objection. A CMP that relies on the maximising argument remains vulnerable to the criticism that it is inimical to ensuring accountability of the political actors and the public confidence of the courts – ie the abovementioned underlying rationales for the principle of open justice. For the maximising argument does not preclude materials adverse to the excluded party from being used, without disclosure to that party,¹⁷⁴ such that:

1. (a) The CMP positively undermines the excluded party's ability to succeed against the UK government¹⁷⁵ and so to at least achieve accountability in the immediate case;¹⁷⁶ it also seems to convey the judicial approval of an approach whereby entitlements designed to ensure respect for a litigant's human dignity¹⁷⁷ can be sacrificed as a means to further ends – a tension mentioned above.

2. (b) There may therefore be questions about whether ‘justice is done’, when a CMP is so deployed to the evident detriment of the excluded party.¹⁷⁸ And as explained, the courts are also hindered from dispelling this public perception of prima facie injustice.¹⁷⁹

3. (c) Accordingly, the maximising argument remains vulnerable to the open justice objection – particularly by reference to the confidence limb:¹⁸⁰ whilst it does not necessarily withhold more materials from disclosure than on the alternatives,¹⁸¹ deploying a CMP in reliance upon the maxmising argument per se generates new and distinct problems in terms of the open justice objection. The abovementioned problem of scope therefore continues to apply.

But the open justice objection will be significantly weakened when the clear advantage variation is deployed. Although the clear advantage variation will still not permit the impugned materials to be openly dealt with – and to that extent it detracts from upholding the accountability of the political actors¹⁸² – it does not share the previous concern that this will undermine the excluded party's ability to hold the UK government accountable in the immediate case; rather, it will bolster it – and is indeed precisely aimed at bolstering it.¹⁸³ This provides a significant counterweight to the accountability limb of the open justice objection: for accountability of the UK government in the immediate case is thereby facilitated, when much less will be offered in scenario 1.¹⁸⁴

Further, as the clear advantage variation operates to the clear advantage of the excluded party, it can help dispel the perception that the excluded party has been sacrificed as a means to the ends of others.¹⁸⁵ It is unclear if the public – once it properly understands the strictures appertaining to the clear advantage variation – will see a prima facie injustice being committed, when the court endorses it; or at least, it is questionable whether the public will see the prima facie injustice as unanswered: this time, the court will have a ready answer at hand to regain the confidence from the public, even though the impugned proceedings remain secret. It seems fair to expect that the detrimental effect on public confidence in the courts will likely be thereby significantly mitigated. It is also possible that public confidence in the courts will indeed be bolstered, based on its independence from government,¹⁸⁶ if the courts are perceived to be facilitating the excluded party in testing the government's case, especially when the excluded party within a CMP is frequently in a de facto weaker position than the UK government as fellow litigants.¹⁸⁷

I therefore consider clear advantage variation to have sufficiently mitigated the open justice objection: when compared to its alternatives, the clear advantage variation does not undermine the general objectives of accountability and public confidence – and they may indeed in some respects be promoted. Accordingly, the problem of scope does not arise. But the same cannot be said of the maximising argument, since its use per se is prone to generate new problems vis-à-vis the open justice objection. This conclusion – combined with the above conclusion concerning the natural justice objection – means that both of Lord Dyson's objections to the CMP are sufficiently mitigated by the clear advantage variation. This meets the stated benchmark. One may therefore conclude that the clear advantage variation is not only prima facie normatively justified, but also normatively justified, all things considered.

Conclusion

This paper seeks to investigate whether a CMP can ever be normatively justified in English law, all things considered. We have set out to establish a benchmark: a CMP is shown to be normatively justified, all things considered, if it entirely defeats or sufficiently mitigates Lord Dyson's two objections to the CMP in Al Rawi – ie the natural justice and open justice objections.¹⁸⁸ If this benchmark is met, the normative starting point – which favours a CMP – will not be displaced; the CMP will accordingly be normatively justified, all things considered. It has been contended that: (i) a commonplace argument for the CMP – the maximising argument – is unsuccessful in meeting this benchmark; and (ii) the clear advantage variation does meet this benchmark. Accordingly, the clear advantage variation is normatively justified, all things considered. This achieves the objective of this investigation.