Hostname: page-component-77c78cf97d-cfh4f Total loading time: 0 Render date: 2026-05-04T18:45:35.544Z Has data issue: false hasContentIssue false
  • English
  • Français

Public interest immunity and judicial deference

Published online by Cambridge University Press:  30 April 2026

Edward Lui Open the ORCID record for Edward Lui [Opens in a new window]
Edward Lui*
Affiliation:
Faculty of Law, The University of Hong Kong , Hong Kong, China
*
E-mail: elylui@hku.hk
Rights & Permissions [Opens in a new window]

Extract

The puzzle underlying the law of public interest immunity (PII) is well known. In ordinary proceedings all relevant evidence should be placed before the court and the parties, including when the government is one of those parties.1 But what if some of the relevant evidence – due to reasons in the public interest – cannot be disclosed to the other parties involved?2 The traditional way in which English law tackles this puzzle is by invoking the law of PII. Ever since the landmark decision in Conway, it has been settled that ‘the courts should balance the public interest in the proper administration of justice against the public interest in withholding any evidence which a Minister considers ought to be withheld’.3 If the public interest favours disclosure of the evidence, the evidence is admitted and made ‘available to both parties and to the court’; but if the contrary is true the evidence will be excluded altogether from the proceedings and made ‘available neither to the other parties nor to the court’.4

Information

Type
Current Developments: Case Comment
Information
Legal Studies , First View , pp. 1 - 5
Creative Commons
Creative Common License - CCCreative Common License - BY
This is an Open Access article, distributed under the terms of the Creative Commons Attribution licence (http://creativecommons.org/licenses/by/4.0), which permits unrestricted re-use, distribution and reproduction, provided the original article is properly cited.
Copyright
© The Author(s), 2026. Published by Cambridge University Press on behalf of The Society of Legal Scholars

Introduction

The puzzle underlying the law of public interest immunity (PII) is well known. In ordinary proceedings all relevant evidence should be placed before the court and the parties, including when the government is one of those parties.Footnote 1 But what if some of the relevant evidence – due to reasons in the public interest – cannot be disclosed to the other parties involved?Footnote 2 The traditional way in which English law tackles this puzzle is by invoking the law of PII. Ever since the landmark decision in Conway, it has been settled that ‘the courts should balance the public interest in the proper administration of justice against the public interest in withholding any evidence which a Minister considers ought to be withheld’.Footnote 3 If the public interest favours disclosure of the evidence, the evidence is admitted and made ‘available to both parties and to the court’; but if the contrary is true the evidence will be excluded altogether from the proceedings and made ‘available neither to the other parties nor to the court’.Footnote 4

PII remains good law today.Footnote 5 But its role has been increasingly overshadowed by another device by which English law tackles the puzzle: it is by invoking the ‘closed material procedure’.Footnote 6 Under this procedure evidence otherwise excluded by PII can still be relied upon by the government in a closed part of the proceedings, from which the other parties are excluded from direct participation (although they may still be represented by a ‘special advocate’ therein).Footnote 7 This procedure has formed the focus of most modern cases reaching the UK Supreme Court vis-à-vis the puzzle.Footnote 8 The same is also true of the corresponding academic commentary.Footnote 9

Against this background, the recent decision by the UK Supreme Court in Re Secretary of State for Northern Ireland Footnote 10 (Re SSNI) has offered a rare opportunity for the principles of PII to be considered in isolation from the closed material procedure. This is because Re SSNI was concerned with evidence to be considered in an inquest, where the closed material procedure is not available.Footnote 11 As a result, the puzzle arising here had to be answered by reference to the law of PII alone. This explains the distinct significance of this case when contrasted with other recent case law.

1. Facts

Re SSNI featured an inquest concerning a murder committed during the Troubles. In April 1994 the deceased (Liam Paul Thompson) was a passenger in a taxi, when the driver was asked to drive near the community peace line in question.Footnote 12 A number of gunmen – who crossed ‘through a hole in the peace line’ that night – then approached the taxi and killed Mr Thompson.Footnote 13 Later a ‘loyalist paramilitary group’ acknowledged that it was responsible for murdering him.Footnote 14 In the corresponding inquest a number of significant questions concerning the death of Mr Thompson were raised.Footnote 15 Included amongst them was whether an informant might have already been aware of the attack in advance or might even have facilitated the attack.Footnote 16

The Police Service of Northern Ireland (PSNI) confirmed that it ‘held documents relevant or potentially relevant of factual issues within the scope of the inquest’.Footnote 17 These consisted altogether of seven folders of documents.Footnote 18 But the Chief Constable of the PSNI (via the Secretary of State for Northern Ireland (the Secretary of State)) claimed PII over these documents, on the basis of an assessment that disclosing them ‘would give rise to a real risk of serious harm to the important public interest of national security’.Footnote 19 This claim was upheld by the coroner and it was uncontentious that this was right.Footnote 20 The specific issue giving rise to this litigation was: the coroner decided that documents contained in the seventh folder could be disclosed ‘by way of a gist’ under the law of PII.Footnote 21 This was initially opposed by, but later supported by, the Chief Constable on the basis that this ‘would not cause or risk any damage or harm to the public interest’;Footnote 22 but the Secretary of State has consistently opposed this on the ground that this too ‘would bring about a real risk of serious harm to an important public interest’.Footnote 23 The disagreement was not due simply to a difference in legal opinion: the Chief Constable was specifically concerned ‘to achieve broad community support for policing’ in Northern Ireland, by encouraging a more open scrutiny of incidents during the Troubles.Footnote 24 This does not imply that the Secretary of State found this consideration insignificant; a natural inference is that he considered it significant but ultimately outweighed by the considerations underlying the PII claim.

Under the law of PII the court has to conduct a balancing exercise,Footnote 25 whereby it must determine whether ‘the public interest in preserving the confidentiality of the document outweighs the public interest in securing justice’.Footnote 26 Applying this, the coroner decided that the gist should be disclosed: that (a) ‘the information summarised in the gist was “of central importance to the inquest” … [and] was “highly relevant”’;Footnote 27 (b) this has ‘mitigated any real risk of serious harm to the public interest’, contrary to the Secretary of State’s assessment;Footnote 28 and (c) in any event, ‘the public interest in non-disclosure of the information contained in [the] gist … was outweighed by the public interest in disclosure for the purposes of doing justice in the proceedings’.Footnote 29 The Secretary of State challenged this via judicial review.Footnote 30 In a unanimous judgment written by Lord Sales and Lord Stephens, the UK Supreme Court ruled for the Secretary of State: it held that under the law of PII the gist should not be disclosed and that the coroner was wrong in thinking otherwise.Footnote 31 In so doing the Supreme Court has addressed two important aspects of the law of PII vis-à-vis the question of deference, which I shall now discuss.

2. Analysis

(a) Deference to the Government’s assessment

The first important question addressed by the Supreme Court was the extent to which (if at all) the coroner should have deferred to the Secretary of State’s risk assessment, in conducting her balancing exercise. The coroner applied the then-existing guidance from Goldring LJ, that ‘when carrying out the balancing exercise, the Secretary of State’s view regarding the nature and extent of damage to national security which will flow from disclosure should be accepted unless there are cogent or solid reasons to reject it’.Footnote 32 The UK Supreme Court held that this was an insufficiently deferential statement of principle: the court should instead be limited to the ‘normal public law principles’.Footnote 33 This had the result that the Secretary of State’s assessment ‘should have been accepted by the coroner unless there was no evidence to support that assessment or the assessment was Wednesbury irrational’.Footnote 34 And as the coroner did not go so far as to say the Secretary of State’s risk assessment was irrational,Footnote 35 she should have conducted her balancing exercise on the basis that ‘there was a real risk of serious harm to national security’ per the Secretary of State.Footnote 36 The Supreme Court found this deferential position justified on the basis that ‘[i]n assessing national security, it is the Secretary of State who has the democratic authority and institutional competence to make the relevant assessment’.Footnote 37

This does lead to a highly deferential judicial role vis-à-vis the risk assessment.Footnote 38 But as the Supreme Court has mentioned, recent authorities from that court have already repeatedly stressed the need for the court to be deferential to executive assessments in the context of national security.Footnote 39 This need for deference has also been recognised in the law of PII.Footnote 40 In Conway, for instance, Lord Reid explained that under the balancing exercise the courts must give ‘full weight’ to ‘a Minister’s view’; that ‘if the Minister’s reasons are of a character which judicial experience is not competent to weigh, then the Minister’s view must prevail’.Footnote 41 Lord Morris added that ‘[s]ome aspects of the public interest are chiefly within the knowledge of some Minister and can best be assessed by him’, and that ‘[i]f a responsible Minister stated that production of a document would jeopardise public safety it is inconceivable that any court would make an order for its production’.Footnote 42

But Re SSNI featured here an interesting scenario where there were disagreeing executive assessments about the applicable risks: whilst the Secretary of State found the gist to be sufficiently risky, the Chief Constable did not.Footnote 43 If the court was to be deferential to executive assessments about the risks to national security, it seemed that ex hypothesi it should be deferential to both the disagreeing executive authorities here. One answer to this was that the Secretary of State (by enjoying both ‘democratic authority and institutional competence’)Footnote 44 merited more deference thereto than the Chief Constable, who of course was not elected. The Supreme Court seemed to be proceeding upon this line when it suggested that ‘the Secretary of State is in the best position to act as the clearing house to draw upon, consider and weigh up the views of relevant public authorities (which may be in conflict) so as to arrive at a final overall view of the national security aspect of the public interest’.Footnote 45 Indeed the Supreme Court further suggested that the governmental authorities should ideally have settled the matter between themselves,Footnote 46 but short of a consensus the Secretary of State’s assessment would be adopted as the overall executive position.Footnote 47 But this view seems flawed. Even if the Secretary of State is in a better position than the Chief Constable to assess the applicable risks (on the basis that the former has not only ‘institutional competence’ but also ‘democratic authority’)Footnote 48 the Chief Constable remains a better decision-maker than the courts on this matter. The Supreme Court’s reasoning seems to ignore the fact that there are two better decision-makers involved that each merits deference from the courts.

Imagine we are sitting in a university committee deciding on a matter relating to the welfare of female students, and before us are considered opinions offered by an institutionalised female student group and those offered by some individual female students beyond the group. The Supreme Court’s initial suggestion will be to say that we should be deferential to the assessment of the institutionalised group, which is in a better position than we are in terms of both direct knowledge about female student welfare and of being directly responsible to the female students themselves.Footnote 49 But the Supreme Court’s further suggestion will be to say that if the institutionalised group and the individual students disagree, we should be deferential only to the institutionalised group and not at all to the individual students. This is the wrong way to go about it. As a university committee, we should accept that the individual students know more than we do about their own welfare, even if the institutionalised group is in the best position in this regard. We may give more weight to the views of the institutionalised group, but it does not mean that the views of individual students do not merit any deference at all. The same seems true in the immediate context, although the precise doctrinal device by which this should happen (on this alternative legal picture) is a matter to be further explored.

(b) Deference to the initial judge’s assessment

The second important question addressed by the Supreme Court concerned the extent to which an appellate judge should defer to the PII assessment of an initial judge. This question arose from the fact that ‘at first instance … it is the first instance court which makes the decision’ about PII – in this case, the coroner.Footnote 50 But the initial judge’s assessment can (depending on the precise status of the judge) be challenged via judicial review or via an appeal to an appellate court.Footnote 51 Here the coroner’s PII assessment was challenged via judicial review.Footnote 52 The question is this: should the appellate court render a PII assessment de novo, or should it to any extent defer to the assessment of the initial judge?

The Supreme Court held that the former was correct.Footnote 53 According to it the initial judge was ‘applying a substantive part of the law of evidence’ and ‘not exercising a discretion’; and correspondingly, ‘[t]he appellate or reviewing court … has to determine whether the first instance court has identified the relevant rule of substantive law and applied it correctly’.Footnote 54 In particular, the Supreme Court was of the view that ‘an appellate or reviewing court is as well placed as the first instance court’ to undertake the balancing exercise.Footnote 55 It suggested, significantly, that there was a uniquely correct answer to a PII balancing exercise: ‘[t]here is not a range of views about the public interest which the court has to consider so as to ensure it takes a view falling within that range. If the [initial judge] … misidentifies the public interest, it has gone wrong in law and an appellate or reviewing court is required to correct its error’.Footnote 56

Crucially, it seems that the Supreme Court has conceptualised the PII balancing exercise as a question of law: this explains its statement that if the initial judge disagreed with the appellate court ‘it has gone wrong in law and an appellate or reviewing court is required to correct its error’.Footnote 57 This reading is further confirmed by the fact that the Supreme Court did not distinguish between the roles of an appellate court on a judicial review challenge and on a statutory appeal challenge respectively: for whilst the two roles are not identical,Footnote 58 they share the common premise that the appellate court can overrule the initial judge without deference if it discovers an error of law.Footnote 59 That no such distinction was drawn by the Supreme Court indicated that, in its view, the PII balancing exercise raises a question of law and going wrong on this leads to an error of law.

But this does not seem compatible with the court’s earlier reasoning on judicial deference to an executive assessment under the law of PII. This is for the simple reason that if the PII balancing exercise is conceptually a question of law for which a uniquely correctly answer exists, the same is true vis-à-vis the executive authorities. That is, ex hypothesi the executive authorities are in applying the law of PII engaged with a question of law. But on ordinary principles of administrative law, the courts will decide de novo a question of law without a need to defer to the administrative decision-makers.Footnote 60 This shows that the Supreme Court’s discussion on deference has not been internally coherent. The relationship between the two issues examined by the Supreme Court seems therefore to merit further consideration in the case law, before we can obtain a coherent account of deference in the law of PII.

References

1 See eg Al Rawi v Security Service [2011] UKSC 34, [2012] 1 AC 531.

2 See eg R (Roberts) v Parole Board [2005] UKHL 45, [2005] 2 AC 738.

3 Conway v Rimmer [1968] AC 910 at 951.

4 Al Rawi, above n 1, at [41]; see also R (Haralambous) v Crown Court at St Albans [2018] UKSC 1, [2018] AC 236 at [22].

5 See eg R (Jordan) v Chief Constable of Merseyside Police [2020] EWHC 2274 (Admin).

6 See eg Justice and Security Act 2013, s 6; Haralambous, above n 4, at [59].

7 Justice and Security Act 2013, ss 6(4)(b)(i) and 9. See eg Jordan, above n 5.

8 See eg Al Rawi, above n 1; Haralambous, above n 4.

9 See eg M Chamberlain ‘Special advocates and amici curiae in national security proceedings in the United Kingdom’ (2018) 68 University of Toronto Law Journal 496; E Lui ‘A limited case for the closed material procedure: natural justice, open justice and the clear advantage variation’ (2024) 44 Legal Studies 668.

10 Re SSNI [2025] UKSC 47, [2026] 2 WLR 109.

11 Ibid, at [20].

12 Ibid, at [13]–[14].

13 Ibid.

14 Ibid, at [15].

15 Ibid, at [24].

16 Ibid, at [25]–[27].

17 Ibid, at [43].

18 Ibid, at [43], [45].

19 Ibid, at [52].

20 Ibid, at [56]–[57].

21 Ibid, at [57], [72].

22 Ibid, at [58], [70].

23 Ibid, at [72].

24 A Deb et al ‘Legacy issues: In re Secretary of State for Northern Ireland [2025] UKSC 47’ UKCLA Blog (18 December 2025), available at https://ukconstitutionallaw.org/2025/12/18/anurag-deb-colin-murray-and-gabriel-tan-legacy-issues-in-re-secretary-of-state-for-northern-ireland-2025-uksc-47/.

25 Re SSNI, above n 10, at [129].

26 R v Chief Constable of West Midlands Police, ex p Wiley [1995] 1 AC 274 at 280; see also Re SSNI, above n 10, at [125].

27 Re SSNI, above n 10, at [90].

28 Ibid, at [92], [94].

29 Ibid, at [94].

30 Ibid, at [73].

31 Ibid, at [165].

32 Secretary of State for Foreign and Commonwealth Affairs v Assistant Deputy Coroner for Inner North London [2013] EWHC 3724 (Admin) at [57]. See Re SSNI, above n 10, at [90]–[92].

33 Re SSNI, above n 10, at [130].

34 Ibid, at [138].

35 Associated Provincial Picture Houses Ltd v Wednesbury Corpn [1948] 1 KB 223.

36 Re SSNI, above n 10, at [138].

37 Ibid, at [38].

38 Deb et al, above n 24.

39 See eg R (Begum) v Special Immigration Appeals Commission [2021] UKSC 7, [2021] AC 765 and U3 v Secretary of State for the Home Department [2025] UKSC 19, [2025] AC 1510, cited in Re SSNI, above n 10, at [38]. See also Deb et al, above n 24.

40 See eg Conway, above n 3; Balfour v Foreign and Commonwealth Office [1994] 1 WLR 681.

41 Conway, above n 3, at 952; see also Balfour, above n 40.

42 Conway, above n 3, at 955–956; see also Balfour, above n 40.

43 Re SSNI, above n 10, at [70], [72].

44 Ibid, at [38].

45 Ibid, at [37] (emphasis added).

46 Ibid, at [60].

47 Ibid, at [61].

48 Ibid, at [38].

49 See eg A Perry ‘Wednesbury unreasonableness’ (2023) 82(3) Cambridge Law Journal 483. See also M Teo ‘Administrative expertise and Wednesbury unreasonableness’ [2025] Public Law 359.

50 Re SSNI, above n 10, at [124].

51 The Father v Worcestershire County Council [2025] UKSC 1, [2025] 2 WLR 155, at [59]–[60], [62].

52 Re SSNI, above n 10, at [73].

53 Ibid, at [124]–[126], [158]–[160].

54 Ibid, at [124].

55 Ibid, at [125].

56 Ibid, at [126] (emphasis added).

57 Ibid (emphasis added).

58 See eg General Medical Council v Michalak [2017] UKSC 71, [2017] 1 WLR 4193.

59 See eg T Endicott ‘Questions of law’ (1998) 114 Law Quarterly Review 292; Edwards (Inspector of Taxes) v Bairstow [1956] AC 14; R (Finch) v Surrey County Council [2024] UKSC 20, [2024] PTSR 988.

60 See eg R v Lord President of the Privy Council, ex p Page [1993] AC 682.