Keane v The Information Commissioner and others: [2016] UKUT 461 (AAC) [2016] UKUT 0461 (AAC)

UPPER TRIBUNAL
ADMINISTRATIVE APPEALS CHAMBER
[2016] UKUT 0461 (AAC)
Between
KeaneAppellantThe Information Commissioner and othersRespondent
Date 17 October 2016Category: Information rightsSubcategory: Information rights freedom of information public interest test, Information rights freedom of information qualified exemptions

The history of Ireland, national security and Government secrecy

[1]There is much still to be written about the history of Ireland since the late nineteenth century. Many events in this troubled era have taken place in the full glare of publicity. Other events have taken place in the shadows. One of the historian’s roles is to shine a light into those dark corners of history.[2]The National Archives (TNA, formerly the Public Record Office) at Kew houses files from various government departments that have been retained as having some Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 2 historical interest (although by no means everything of significance has found its way to Kew: see Ian Cobain, The History Thieves: Secrets, Lies and the Shaping of a Modern Nation (September 2016, Portobello Books)). Some of the Kew files contain information about the use by the police and security services of paid informants in Irish republican organisations at a time when British rule extended to the whole island of Ireland.[3]How far should historians be able to access such files today, more than 100 years after the files were current? Should there effectively be open access? Or are there some details in those files which should still be withheld from researchers, a century on? If so, on what basis should that be? These are questions which have come before the First-tier Tribunal (Information Rights) in the General Regulatory Chamber on more than one occasion.[4]In Metropolitan Police v Information Commissioner [2008] UKIT EA 2008 0078 (“the Butterworth case”) a historian (Mr Butterworth) sought access to Special Branch files from the period from 1888 to 1912 dealing with its investigations into the activities of European anarchists. In a sense, the researcher wanted to inspect the official paperwork underpinning the type of police work depicted in Joseph Conrad’s The Secret Agent. The First-tier Tribunal issued what was, in effect, a consent judgment, ordering the Metropolitan Police to disclose the files in question. However, the Tribunal also directed that the names of any individuals referred to should be redacted before the files were released.[5]In Marriott v Information Commissioner [2011] UKFTT EA 2010 0183 another First-tier Tribunal concluded that Metropolitan Police records from the same era, including details of informants, and relating to the ‘Jack the Ripper’ murders, should not be disclosed. That Tribunal decided unanimously that the records fell within the scope of the qualified exemption in section 30(2) (information held for the purposes of an inquiry) of the Freedom of Information Act 2000 (FOIA). The Tribunal also held (but only by a majority) that the public interest in maintaining the exemption outweighed the public interest in disclosure. That decision was not appealed to the Upper Tribunal.[6]In the present case – Keane v Information Commissioner and Others [2015] UKFTT EA 2015 0013 GRC – a different First-tier Tribunal, again by a majority decision, decided that the details of paid informants referred to in Metropolitan Police records and involved in Irish secret societies in the period from 1890 to 1910 should not be disclosed. The files were accordingly broadly contemporaneous with those in both the Butterworth case and Marriott, albeit the context was different. Disclosure of the names of paid informants was resisted on the grounds of both national security (FOIA, section 24(1)) and health and safety (FOIA, section 38(1)).[7]If nothing else, these three first instance decisions demonstrate how finely balanced such decisions may be and how reasonable people (and especially reasonable judicial office-holders) may reasonably differ. In that context it is important to remember that an appeal to the Upper Tribunal is not a full merits review. The First-tier Tribunal, of course, must conduct a full merits review of the Information Commissioner’s decision notice. The Upper Tribunal’s role is confined to ascertaining whether the First-tier Tribunal’s decision involves a material error of law.

The specific issue arising on this appeal before the First-tier Tribunal

[8]The practical issue raised by this appeal before the First-tier Tribunal was accordingly whether certain information in a National Archives file entitled Activities of named paid informants against Irish Secret Societies (TNA file ref HO 317/38) should Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 3 be released under the Freedom of Information Act 2000. The file covers the period 1890-1910. The requested information in issue comprises the names of such paid informants.

The legal issue raised by this appeal before the Upper Tribunal

[9]The (rather narrow) legal issue arising on this appeal to the Upper Tribunal is whether the majority members of the First-tier Tribunal correctly carried out the public interest balancing test, having already decided that the qualified exemptions under sections 24(1) (national security) and 38(1) (health and safety) of FOIA were engaged.

The background to the complaint to the Information Commissioner

[10]Mr Keane is both an Irish historian and a historian of Ireland. As part of his research he visited Kew to consult the National Archives file referred to in paragraph 8 above. It was evident that a significant number of pages in that file relating to police informants had been removed. He made a request, treated as a request under the Freedom of Information Act 2000 (FOIA), to access the withheld information. He argued (in 2013) that “while there may have been good reason for doing so [i.e. redacting names] when the file was sent to the National Archives it does not appear to have any validity now. As the file ends in 1910 there seems no reason why any papers should be excluded 103 years later.”[11]The Home Office, the public authority which was in effect guardian of the file, declined to release the information requested, having taken the view that the information was exempt from disclosure by virtue of section 24(1) of FOIA, i.e. it was being withheld for the purposes of safeguarding national security. That decision was confirmed on internal review. Mr Keane then lodged a complaint with the Information Commissioner.

The Information Commissioner’s decision notice

[12]On 13 November 2014 the Information Commissioner issued a Decision Notice (FS50532586), which rejected the substance of Mr Keane’s complaint. In short, the Commissioner decided that the Home Office had been correct to find that the section 24(1) was engaged and had also correctly concluded that the public interest balancing test favoured maintaining the exemption. Mr Keane then appealed to the First-tier Tribunal. By the time the matter proceeded to hearing, the Home Office also sought to rely on the health and safety exemption in section 38(1) of FOIA. In addition, the Metropolitan Police Service (MPS), with its obvious interest both as the historical source of the file and a contemporary State agency using paid informants, was joined as a further respondent.

The First-tier Tribunal’s hearing

[13]The First-tier Tribunal (“the Tribunal”) held an oral hearing on 17 June 2015. As well as receiving assorted documentary evidence, the Tribunal heard oral evidence from Mr Keane himself, from Ms Janet Millar (described as Senior Information Rights Consultant at the Home Office) and from “Officer A”, an Acting Detective Inspector in the MPS assigned to the Counter Terrorism Command, who has first-hand experience of dealing with informants (or Covert Human Intelligence Sources (CHIS), as they are known in the jargon of the Regulation of Investigatory Powers Act 2000 (RIPA)).[14]Mr Keane’s evidence was summarised by the Tribunal at paragraph [9] of its decision. His case was that neither the section 24(1) nor the section 38(1) exemption was engaged, given the age of the disputed information. He also pithily expressed Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 4 the view that “it’s ridiculous to suggest that there would be retribution in 2015 for events in 1910”.[15]Ms Millar’s evidence was summarised by the Tribunal at paragraph [17]-[18]. Her evidence mainly related to the internal processes for reviewing sensitive information and was heard entirely in open.[16]Officer A’s evidence was heard both in open and in (a short) closed session (paragraphs [10]-[16]). His detailed and lengthy open witness statement argued that the issue was not simply whether a particular name should or should not be released. Rather, disclosure would have “far wider reaching implications for the Public Interest and safety of individuals.” In the conclusion to his open witness statement, Officer A contended that there were four reasons why the information should not be disclosed. First, the MPS was unable to identify with certainty which names in the file were true names and which were pseudonyms, and so the police could not conduct a risk management process for disclosure. Second, there was “an unquantifiable risk to the relatives of these informants”. Third, given the MPS’s inability to identify those concerned, the procedure in R (on the application of WV) v Crown Prosecution Service [2011] EWHC 2480 (Admin) could not be operated (i.e. giving those concerned the opportunity to object to disclosure). Fourth, it was argued that disclosure “will cause irrevocable damage to the MPS and Security Services’ ability to recruit and retain future informants”. As the case has unfolded, it is the second and last of these four reasons that has formed the basis of the Second and Third Respondents’ fundamental objections to disclosure.

The First-tier Tribunal’s decision

[17]By a majority, the Tribunal dismissed the appeal. In a lengthy decision, the majority essentially endorsed the position taken by the Commissioner. So the majority found both that section 24(1) was properly engaged and that the public interest balancing test favoured maintaining the exemption, i.e. withholding the information concerned. The majority decision was particularly influenced by two factors. The first was the State’s ability to recruit and retain informants (paragraphs [30]-[38]). The second was the potential risk of harm to the descendants of informants (paragraphs [39]-[56]). Putting these factors into the melting pot of the public interest balancing test, the majority decided that these factors outweighed the arguments in favour of disclosure (paragraphs [57]-[68]).[18]The contrast in the approach of the dissenting member of the First-tier Tribunal was stark. In his assessment, the Respondents’ arguments about the engagement of the claimed qualified exemptions failed “a very basic common sense test” (paragraph [69]). In summary, his view was that neither the section 24(1) exemption nor the section 38(1) exemption was engaged in the first place (paragraphs [69]-[70]). He did not in so many words address the public interest balancing test itself (and, given his conclusion on engagement, perhaps nor did he need to).[19]Mr Keane then applied for permission to appeal to the Upper Tribunal, in part on the ground that while the First-tier Tribunal had purported to apply a qualified exemption by way of the balancing test, in reality it had applied an absolute test. The First-tier Tribunal Judge gave permission to appeal.

The proceedings before the Upper Tribunal

[20]All parties have made written submissions on the appeal. I also held an oral hearing of the appeal at the Rolls Building in London on 29 September 2016. The Appellant attended and was represented by Mr Brian Leahy and Mr Cathal Malone, both of the Irish Bar. The First Respondent was represented by Mr Rupert Paines of Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 5 Counsel. The Second Respondent was represented by Mr Andrew Sharland of Counsel. The Third Respondent was represented by Mr Christopher Knight of Counsel. Of that stellar cast only Mr Leahy and Mr Knight had appeared before the First-tier Tribunal. I am grateful to all counsel (and those supporting them) for their well-focussed submissions and invaluable assistance, both before and at the hearing.

The relevant provisions of the Freedom of Information Act 2000

[21]Section 23 of FOIA provides for information to be exempt information “if it was directly or indirectly supplied to the public authority by, or relates to, any one of a number of named bodies” (e.g. the Security Service). Section 23, of course, is an absolute exemption (see section 2(3)) and was not relied upon in this case. However, by virtue of section 24(1), “information which does not fall within section 23(1) is exempt information if exemption from section 1(1)(b) is required for the purpose of safeguarding national security” (emphasis added). This is a qualified exemption, which necessarily “implies that there may be instances in which it will be in the public interest to disclose information, notwithstanding that the exemption is required for the purpose of safeguarding national security. Otherwise the exemption will be effectively metamorphosed into an absolute exemption” (see Baker v Information Commissioner [2007] UKIT EA 2006 0045 at paragraph [31]).[22]Section 38(1) of FOIA, also a qualified exemption, provides that information “is exempt information if its disclosure under this Act would, or would be likely to —(a) endanger the physical or mental health of any individual, or(b) endanger the safety of any individual.”

The scope of this appeal to the Upper Tribunal

[23]Mr Keane’s application to the First-tier Tribunal for permission to appeal to the Upper Tribunal cited two grounds. The first ground was the contention that the Tribunal had erred in law in its application of the public interest balancing test, and in effect had applied an absolute exemption. The second ground concerned the Tribunal’s approach to the evidence of Officer A. The First-tier Tribunal Judge gave permission on the first ground but not the second, commenting en passant that “although it is a matter for Mr Keane rather oddly he does not adopt any of the arguments set out by the minority in the FTT’s judgement but seeks to present quite different arguments.”[24]In their skeleton arguments for the Upper Tribunal all three Respondents argued that the First-tier Tribunal’s ruling was a limited grant of permission to appeal. Accordingly, they argued, the appeal before the Upper Tribunal was confined to the sole ground on which permission had been given, i.e. that relating to the application of the public interest balancing test.[25]At the outset of the oral hearing, Mr Leahy also accepted that the only ground of appeal related to the public interest balancing test, and so it was not open to him to revisit the Tribunal’s finding that the section 24(1) and 38(1) exemptions were both engaged. He suggested, however, if somewhat gnomically, that consideration of the public interest balancing test might “stretch into” the issue of engagement as it was difficult to disentangle the two issues.[26]Normally I would accept any concession by counsel with alacrity and move swiftly on. However, in the particular circumstances of this case I consider that some further comment is in order. In my view the First-tier Tribunal’s permission ruling is ambiguous and can be read in either of two ways. The first (the reading adopted by all the Respondents, and indeed acceded to by the Appellant) is that the Tribunal Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 6 gave limited permission to appeal only on ground one, having expressly rejected ground two. An alternative reading – and one to which I incline – is that the Tribunal gave general permission to appeal on the basis of ground one and any other material issues raised later in the course of the appeal proceedings. Certainly as a matter of good judicial practice if a tribunal is giving limited permission to appeal one would expect it to say so in as many words for the avoidance of any doubt, e.g. “Limited permission to appeal is granted” rather than (as here) “Permission to appeal is therefore granted”.[27]My preferred reading is supported by the contemporaneous documentation associated with the grant of permission to appeal. In that context I note that rule 43(5) of the Tribunal Procedure (First-tier Tribunal) (General Regulatory Chamber) Rules 2009 (SI 2009/1976) provides that the Tribunal “may give permission to appeal on limited grounds, but must comply with paragraph (4) in relation to any grounds on which it has refused permission”. Rule 43(4) in turn requires the Tribunal to give an applicant reasons for any refusal of permission (rule 43(4)(a)) as well as notification of(a) (A) the right to apply to the Upper Tribunal for permission to appeal,(b) (B) the time limit, and(c) (C) the method to be used for any such application (rule 43(4)(b)).[28]In the present case the Tribunal arguably just about complied with rule 43(4)(a) as regards its rejection of ground two (“I do not think that [the Appellant’s] comments on the Tribunal’s approach to Officer A’s evidence found a valid basis for seeking permission to appeal”).[29]However, looking at the file I can see no evidence that the Tribunal complied with rule 43(4)(b). The Tribunal office sent the Appellant a copy of the permission ruling with a standard form e-mail dated 17 September 2015 announcing that “Permission has been granted” (again, no mention of limited permission having been granted). There was no suggestion whatsoever that the Appellant could apply to the Upper Tribunal for permission to appeal on the refused ground (or indeed on any other ground). Nor, indeed, was there any direct statement as to the time limit for lodging an appeal, although a link was provided to the Upper Tribunal AAC website. The Tribunal’s comprehensive failure to comply with rule 43(4)(b) means that I find it very hard to accept that this was an effective grant of limited permission to appeal.[30]Whilst on the procedural aspects of the grant of permission to appeal, there was another problem which needs to be highlighted. On 4 November 2015 the Treasury Solicitor’s representative (presumably acting for the Home Office) emailed the Tribunal office to say that she understood the Appellant had been given permission to appeal (PTA) but she had not been sent a copy of the grant and so would be grateful to receive a copy. The Tribunal office’s reply was (remarkably) that “it is not the Tribunal’s policy to send PTA rulings to respondents in an appeal. If you wish to see a copy of the ruling you will need to contact the appellant”.[31]In truth, the Tribunal office’s response was perhaps not so much remarkable as extraordinary. Rule 43(3) states that the Tribunal, once it has considered an application for permission to appeal, “must send a record of its decision to the parties as soon as possible” (emphasis added; NB to the parties, not to the applicant). I recognise that there may not always be a complete ‘fit’ between a tribunal’s procedural rules and its internal operating instructions (or clerical ‘jobcards’). It may well be that as part of the General Regulatory Chamber’s policy of continuous improvement this problem has already been identified and dealt with. If not, I am sure that the Chamber President and his team will see that it is resolved forthwith. Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 7[32]I did not hear detailed argument from counsel on the issues identified in paragraphs 26-31 above and so will say no (or at least not much) more. In addition, Mr Leahy’s timely concession deprived Mr Knight of the opportunity to mount a headon challenge to Upper Tribunal Judge Jacobs’s decision in DL-H v Devon Partnership NHS Trust [2010] UKUT 102 (AAC), where it was held (at paragraph 3) that an appeal to the Upper Tribunal is not limited to issues covered by the grant of permission. I am sorry to have to disappoint Mr Knight, who at the hearing was evidently (if very elegantly) spoiling for a jurisprudential fight on the issue. That may be an argument for another day.[33]I simply make this final observation on the concession made by Mr Leahy. I do not believe it has made any difference whatsoever to the outcome of the case. While I have serious reservations as to whether the majority of the Tribunal was entitled to find that the section 38(1) exemption was engaged (for reasons which I explain below, and as I hinted at during the hearing), I am more than satisfied that they were entitled to conclude that the section 24(1) exemption was engaged. Indeed, it has not seriously been suggested otherwise. Once at least one qualified exemption was engaged, it follows that the correct focus of this appeal in any event was on the Tribunal majority’s approach to the public interest balancing test. If Mr Keane is to get home on his appeal, he had to persuade me that the Tribunal got this (legally) wrong.

The Appellant’s submissions on the Tribunal’s approach to the balancing test

[34]Mr Leahy and Mr Malone, for Mr Keane, recognised that both sections 24(1) and 38(1) of FOIA had the potential to raise very serious and weighty issues. However, they argued, it was important not to conflate the de minimis level of evidence needed to engage the relevant FOIA exemptions in the first place with the preponderance required to tip the balance of weighing the public interests against disclosure on the facts of the particular case. The Tribunal majority, they contended, had given into a ‘kneejerk’ reaction – having found the exemptions were engaged, the relevant considerations had not then been properly weighed in the public interest balancing test. In short, the Tribunal majority had transformed what were plainly qualified exemptions into absolute exemptions.[35]Developing this line of argument, Mr Malone drew attention to the structure of the Tribunal’s decision (see paragraph 17 above). He submitted that the only discussion by the Tribunal of the public interest balancing test was at paragraph [57] onwards. Yet he drew my attention in particular to paragraph [56], at the end of the section of the decision dealing with engagement issues:
“56. This means that there is sufficient room for doubt about the safety of traceable descendants, no matter how small that may be. In view of this the majority sees no justification for imperilling their safety, however remote that possibility. In our view, in the absence of certainty it is far better to err on the side of caution than to give rise to such a risk through disclosure.”
[36]Paragraph [56], Mr Malone sought to persuade me, was absolutely fatal to the majority decision. It was not simply that the Tribunal was asking Mr Keane to prove a negative, namely that there was no risk whatsoever to the safety of informants’ descendants – a guarantee which the Tribunal had recognised was impossible to give (see paragraph [50] of its reasons). Rather, or in addition, the last two sentences of paragraph [56] reeked of pre-judgement. Purportedly in the section of the Tribunal’s decision dealing with engagement, they were not really about engagement at all. Instead, they demonstrated that the majority had already arrived at a conclusion on the public interest balancing test before the scales had begun to be weighed (in paragraph [57] onwards). Thus the qualified exemption had been Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 8 converted into an absolute exemption, as Mr Keane could never produce the required guarantee of safety for informants’ descendants. The Tribunal majority’s decision on the public interest balancing test, Mr Malone argued, had been made by that stage (i.e. by paragraph [56]). In effect, the die was cast when the Tribunal found that the relevant exemptions were both engaged. The heavy weight in favour of maintaining the exemptions had been unthinkingly assumed to be decisive, given the nature of the exemptions pleaded, and the balancing test had not been properly applied.[37]In support of that proposition, Mr Malone also criticised paragraph [67] of the majority’s decision, as failing to explain adequately how the Tribunal had weighed the competing public interest considerations. Moreover, it again showed that a nominally qualified exemption had been treated in terms as absolute in nature, as the Tribunal found that the risk to informants’ descendants “to whatever degree … must prevail” (emphasis added):
“67. The Tribunal majority endorses the Commissioner's findings with regard to the public interest. We too acknowledge the importance of historic research and the Appellant's laudable efforts to cast light onto a once dim area of history. However, it is our view that the ability of the United Kingdom to maintain its national security and the safety of descendents [sic], who may be at risk of exposure to harm, to whatever degree, and the consequences this would have on current and future informants, must prevail.”

The Respondents’ submissions on the Tribunal and the balancing test

[38]The three Respondents sung in harmony from the same hymn-sheet, albeit with slightly different shades of emphasis. I hope I can fairly summarise the thrust of their respective submissions as follows.[39]Mr Paines, for the Information Commissioner, submitted that once the Tribunal had found that sections 24(1) and 38(1) were engaged, it was entitled to assume that those factors would carry considerable weight in the public interest balancing test. It did not necessarily follow that a qualified exemption had been turned into an absolute one. He argued further that the high point of the Appellant’s case was Mr Malone’s attack on paragraph [56] of the Tribunal’s decision. This challenge, Mr Paines contended, was unpersuasive for four reasons. First, it simply made the point that section 38(1) carried significant weight. Second, even if the passage had been phrased in rather loose terms, it disclosed no error of law. Third, the section 38(1) exemption was only a supplementary line of analysis; the bulk of the Tribunal’s decision was devoted to the national security exemption. Fourth, and crucially, paragraph [56] had to be read in the context of the Tribunal’s decision as a whole.[40]Mr Sharland, for the Home Office, agreed with Mr Paines. In particular, he laid considerable emphasis on Mr Paines’s third point about paragraph [56] of the decision. Thus the primary focus of the Tribunal’s decision was on section 24(1) and the risks as regards the recruitment and retention of informants, now known as CHIS. Moreover, the Tribunal, which was the principal fact-finder, had accepted Officer A’s evidence as compelling (see paragraph [37]), and the Upper Tribunal could not go behind that factual finding now. He argued that the Tribunal’s decision showed that the public interest balancing test had been properly carried out and the exemptions had not been treated as absolute in nature.[41]Mr Knight, for the MPS, predictably agreed with both Mr Paines and Mr Sharland. He submitted there was nothing surprising or erroneous in the Tribunal’s reasoning that the risk of harm to national security was a very significant matter Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 9 which would require compelling countervailing evidence for the public interest balancing test to tip in favour of disclosure. This was, said Mr Knight, citing that wellknown jurist Basil Fawlty, no more than a statement of “the bleeding obvious”. The Tribunal, Mr Knight argued, had not fallen into the trap of saying “this is a national security case so the Government wins”. The Tribunal’s decision showed it had properly weighed the various factors in the public interest balancing test; the qualified exemptions had not transmogrified into absolute exemptions. As regards both sections 24(1) and 38(1), the majority had been entitled to conclude that a small risk of a big harm necessarily carried significant weight. For example, Mr Knight argued, if a tribunal got the balancing test wrong in another case on section 35 (formulation of government policy), the worst that could happen is that in future a civil servant might draft advice to ministers more cautiously and less frankly. If a tribunal got it wrong on section 24 – Mr Knight paused for dramatic effect – the risk was that people might die. The Upper Tribunal’s analysis Introduction

The Upper Tribunal’s analysis

[42]In general terms I prefer the submissions of the three Respondents to those of the Appellant. My main reasons for reaching this conclusion are set out below at paragraphs 51-60. At the outset, however, it is only right to mention some misgivings I have about some of the findings and reasoning of the majority of the Tribunal.

Reservations

[43]As intimated above (at paragraph 33), I have some reservations in particular about the Tribunal’s approach to section 38(1) of FOIA in the context of this appeal. Section 38(1), it will be recalled, provides by way of a qualified exemption that information “is exempt information if its disclosure under this Act would, or would be likely to —(a) endanger the physical or mental health of any individual, or(b) endanger the safety of any individual.” My reservations are two-fold.[44]First, in my view the Tribunal majority failed actually to make an explicit finding that section 38(1) was engaged in this case. At the outset, the Tribunal certainly identified this as a relevant issue raised by the appeal (see paragraph [8] of its decision). At a later stage the Tribunal again acknowledged that this was an issue on which the Home Office and Mr Keane had joined arms:
“26. The Home Office also sought to rely on the ‘late claimed’ exemption under s.38(1) of FOIA in the event that the Tribunal did not find that the exemption in s.24(1) was engaged or if the Tribunal found that the PIBT in relation to s.24(1) favoured disclosure. The Home Office contended that the persons likely to be ‘endangered’ (as referred to in s.38(1)) by disclosure of the information would be the descendants of informants. The Home Office accepted that the risk of such endangerment was small but submitted that the nature of the harm that might flow from disclosure was potentially very serious. In relation to the late claimed exemption under s.38(1) FOIA Mr Keane submitted that again this was not engaged as the risk of harm to descendants was speculative and unsupported by any evidence.”
[45]Having established that both sections 24(1) and 38(1) were potentially in play, the majority then stated (at paragraph [29]) that “the appeal should be dismissed for the following reasons”. Those majority reasons were organised under two heads, namely(1) the ability to recruit and retain informants (paragraphs [30]-[38]) and(2) the risk of harm to descendants of informers (paragraphs [39]-[56]). In the former discussion there is an express finding by the majority that section 24(1) was engaged (see paragraph [37]). There is, however, no such explicit conclusion in the latter Keane v (1) IC (2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 10 section of the decision dealing with the risk to descendants. Indeed, the waters are muddied to some extent as at several junctures in their discussion of the latter point the majority made it clear that they regarded concerns for the safety of informants’ descendants as a supplementary reason for finding that section 24(1) was engaged, and not simply a freestanding argument in support of a conclusion that section 38(1) was engaged in its own right (see e.g. paragraphs [39], [42] and [46], and reiterated at paragraph [62] in the context of the public interest balancing test). All that said, the tenor of the discussion at paragraphs [39]-[56] undoubtedly suggests that the majority’s (implied) conclusion was that section 38(1) was engaged. Indeed, the very fact that the minority member dissented by arguing that neither section 24(1) nor section 38(1) was engaged indicates the contrary conclusion on the part of the majority. My concerns in this regard are therefore essentially ones of form, not substance. All parties have understandably proceeded on the basis that the majority found section 38(1) engaged. I do so likewise.[46]Secondly, however, I also have real reservations as to the basis for the majority’s conclusion that section 38(1) was engaged on the facts of this case. The Tribunal summarised the Appellant’s reasons for disputing that section 38(1) was engaged. These included some of the arguments advanced in the Marriott case, i.e. that there was no evidence of either(a) any case in which an informer’s descendants had been targeted long after the informant’s death; or(b) any problems resulting for descendants as a result of the disclosure of informants’ names from parallel historical records held in Ireland itself. Mr Keane also told the Tribunal that as part of another phase of his research he had interviewed the grandchildren and great-grandchildren of Irish informants active in the early 1920s, none of whom had expressed any concern for their welfare today.[47]The Tribunal majority also referred to the Respondents’ arguments, encapsulated in this account of the Home Office’s submission on section 38(1):
“47. The Home Office contends there is a real risk that upon disclosure of the identities of the informants, their descendants will be able to be traced. They go on to say that groups or communities within which those informants operated are likely to seek retribution against descendants. They claim that this is particularly the case in Ireland and Northern Ireland where there remain threats to individuals from paramilitary dissident groups. Whilst the risk is acknowledged to be a small one, the nature of the harm that could result (serious injury or death) is so serious that identification should be avoided.”
[48]Having reviewed the evidence, the majority members reached a finding of fact that at least some descendants could be traceable, not least given the opportunities provided by the internet for genealogical searches. That finding in itself seems to me unassailable, not least as Mr Keane himself had advised the Tribunal that he had successfully traced some descendants of informants named in other public records, even on occasion using a tried and trusted gumshoe method in smaller rural communities, namely “to go to the local public house and ask around” (paragraph [49]).[49]The majority of the Tribunal then concluded their discussion of the issue of risk to such descendants as follows (omitting paragraph [56], which is referred to at paragraph 35 above): “53. Having accepted this, we must now consider the likelihood of harm to traceable descendants. We note the Appellant's firm belief that no group with an interest in events in Ireland which pre-date 1910 poses a risk to descendants of Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 11 informants. We acknowledge his expertise in matters of Irish history in relation to events of the period in question; we similarly acknowledge the quality of his research and the depth of knowledge he has acquired through interviews with descendants of informants.[54]However, we are not satisfied this guarantees that no harm could possibly come to such descendants following release of the disputed information. Despite his extensive research, the Appellant cannot possibly have interviewed every descendant of every informant from the period in question, particularly because there may be names within the disputed information of which he is unaware.[55]Neither can the Appellant be sure of the good nature of every disaffected group in response to the revelation of informant identities. Indeed, much has been said about the possible actions of such groups but little has been said about the possible reaction of local communities. It is by no means fanciful to suggest that on revelation that a person's ancestor was an informer, elements of the local community might choose to shun him or her, causing them distress. Whilst obviously not to the same degree as physical harm or even death, as has been suggested, mental distress is just as undesirable an outcome and an impact on safety.”[50]My concern in this regard is the evidential basis for the majority’s conclusion that section 38(1) was engaged. Obviously I did not hear the witnesses the Tribunal heard. However, I have read their open and closed evidence. I have also read (several times) the Tribunal’s account of their evidence. I am struggling to see any basis on which the Tribunal could properly reach a finding that disclosure would endanger the physical or mental health of any individual or endanger the safety of any individual. I am also in a similar difficulty in identifying any basis on which the Tribunal could be satisfied that disclosure would be likely to result in either type of harm. At best the majority’s findings would appear to justify a conclusion that disclosure might just conceivably lead to such a harm. Given that “likely to” in this context means a real and significant risk, albeit a risk that may well fall short of being more probable than not (see e.g. R (on the application of Lord) v Secretary of State for the Home Office [2003] EWHC 2073 (Admin)), the majority’s conclusion seems at best problematic. However, I recognise both that the evaluation of evidence is for the First-tier Tribunal and that I did not hear detailed argument on this point.

Reasons

[51]Notwithstanding these lingering doubts about the engagement of section 38(1), I conclude that the appeal must be dismissed. My three main reasons are as follows, and do not turn on any narrow jurisdictional issue as to the proper scope of the appeal to the Upper Tribunal.[52]First of all, it is trite law that the Tribunal’s decision must be read as a whole. Appellate courts and tribunals must be careful not to seize on a particular phrase or even a specific passage and take it out of context. As the Court of Appeal held in Re F (Children) [2016] EWCA Civ 546 (at paragraph [23]):
“It is not the function of an appellate court to strive by tortuous mental gymnastics to find error in the decision under review when in truth there has been none. The concern of the court ought to be substance not semantics. To adopt Lord Hoffmann’s phrase, the court must be wary of becoming embroiled in ‘narrow textual analysis’.”
Keane v (1) IC (2) Home Office (3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 12[53]There is plenty of other authority to similar effect, but it would be otiose to cite it here. Added to this line of authority is the observation that one can normally assume that a specialist tribunal knows what it is doing. As Lloyd Jones LJ put it in Department for Work and Pensions v Information Commissioner and Zola [2016] EWCA Civ 758 at paragraph [34] (dissenting on the outcome but not on this point):
“Given such expertise in a Tribunal, it is entirely understandable that a reviewing court or Tribunal will be slow to interfere with its findings and evaluation of facts in areas where that expertise has a bearing. This may be regarded not so much as requiring that a different, enhanced standard must be met as an acknowledgement of the reality that an expert Tribunal can normally be expected to apply its expertise in the course of its analysis of facts.”
[54]The fundamental problem with Mr Malone’s submissions is that they invite me to do just that, namely to take one passage (and in particular paragraph [56]) out of its context. His arguments presuppose that the Tribunal’s decision is to be read as a rigidly linear analysis, with each step in the argument in a hermetically sealed compartment, isolated from the next stage in the reasoning. But this is a decision by a specialist First-tier Tribunal, not a High Court Judge sitting in the Chancery Division. The drafting may be a little rough and ready around the edges – as also e.g. in the failure, as already noted, actually to spell out in as many words that section 38(1) was indeed found by the majority to be engaged – but that does not necessarily mean that the underlying process of reasoning was deficient. I cannot accept that paragraph [56] demonstrates that the majority approached the public interest balancing test with a closed mind or as otherwise having prejudged the issue. At worst, the majority members were simply foreshadowing their later assessment of that test. As Mr Paines rightly observed, this is no more than some rather loose drafting.[55]Mr Malone’s attack on paragraph [67] of the Tribunal’s reasons (see paragraph 37 above) faces the same difficulty. This is no more than the majority’s final substantive paragraph in the decision, confirming and following the reasoning and approach of the Information Commissioner. It would be wrong to subject the drafting of this short passage to a forensic process of dissection as a means of seeking to undermine the Tribunal’s reasoning on the public interest balancing test which leads up to that point. The use of the expression “must prevail” does not lead me to think the majority treated the exemption(s) as absolute in nature; rather, it was a recognition of the significant weight accorded to e.g. the national security qualified exemption (see paragraph 41 above).[56]Secondly, and again reading the Tribunal’s decision as a whole, I am satisfied that the Tribunal applied the public interest balancing test properly and reached a sustainable decision (whether or not it is one that I would have reached on the same evidence is neither here nor there, for the reason indicated at paragraph 7 above). The section in the Tribunal’s decision from paragraph [57] onwards rehearses the competing public interest arguments for maintaining the exemptions and for ordering disclosure. I have to say I have read better analyses of the application of the balancing test in other First-tier Tribunal cases. The discussion here certainly might have been better organised. But again that is not to the point, and goes to form and not to substance.[57]Neither Mr Leahy nor Mr Malone was able to point me to any particular public interest factor which had been omitted by the Tribunal when it should have been included in its analysis, or vice versa. Rather, their argument was essentially that the Tribunal should have attached a different weight to particular considerations. Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 13 However, that is plainly an invitation to me to trespass on the fact-finding function of the first instance Tribunal – and an invitation, tempting though it may be, I must decline. In any event, as Mr Knight argued, the public interest balancing test is ultimately an impressionistic rather than an arithmetical assessment. The Tribunal’s approach here was consistent with the guidance of Upper Tribunal Judge Turnbull in Cabinet Office v Information Commissioner [2014] UKUT 0461 (AAC) at paragraph 67 (and indeed the general practice of the First-tier Tribunal), namely (with emphasis as in the original):
“(i) to consider to what extent the public interest factors potentially underlying the relevant exemption are in play in the particular case and then (ii) to consider what weight attaches to those factors, on the particular facts.”
[58]Nor am I persuaded by the Appellant’s arguments that the Tribunal treated either exemption as absolute in nature. The framework of analysis as set out at the start of the Tribunal’s reasons make it plain that they were well aware they were dealing with qualified exemptions, as did the organisation of their reasoning, notwithstanding some rough edges. Whilst it may well be wise to avoid characterising particular exemptions as carrying “inherent weight” (see Upper Tribunal Judge Turnbull’s decision in the Cabinet Office case at paragraph 66), the reality is that the public interest in maintaining the qualified national security exemption in section 24(1) is likely to be substantial and to require a compelling competing public interest to equal or outweigh it (as recognised in the First-tier Tribunal decision in Kalman v Information Commissioner [2010] UKFTT EA 2009 0111 (GRC), [2011] 1 Info LR 664 at paragraph [47]).[59]Thirdly, the Appellant’s arguments as to the public interest balancing test were in effect exclusively devoted to attacking the Tribunal majority’s findings and reasoning on the issue of the potential risk to informants’ descendants. However, the issue of the risk to informants’ descendants was always a supplementary line of argument in the context of section 24(1). Neither Mr Leahy nor Mr Malone took me to any passage in the Tribunal’s assessment of the State’s interest in the ability to recruit and retain informants (paragraphs [30]-[38]) with a view to arguing that it showed some error of law. Indeed, the Appellant accepted both that the national security qualified exemption was broad enough to encompass both direct and indirect threats and also that informants perform a vital role in protecting the UK from such threats. It was accordingly accepted on all sides that any adverse impact on the ability to recruit and retain informants posed a threat to national security. In that context the Appellant’s decision not to challenge the finding that section 24(1) was engaged was entirely sensible.[60]Moreover, I remind myself that the invocation of section 38(1) in these proceedings was something of an afterthought on the part of the Home Office. Thus the Information Commissioner’s original Decision Notice did not consider the possible applicability of section 38(1) in its own right; the complaint was treated as exclusively a section 24(1) case. The fact remains that the Tribunal’s decision could be shorn of all discussion of the risk to informants’ descendants and it would still sustain a conclusion that the public interest is maintaining the section 24(1) exemption outweighed the public interest in disclosure. I accordingly find that any deficiencies as regards the Tribunal’s approach to the issue of the risk to informants’ descendants, and in particular in the context of weighing the various factors in the public interest balancing test, were not material to the outcome of the appeal. Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 14[61]I therefore dismiss this appeal.

How long is long enough?

[62]There was some discussion at the hearing as to how the MPS’s policy of perpetual anonymity and confidentiality for informants could ever be reconciled with the qualified nature of the FOIA exemptions relied upon. Mr Leahy and Mr Malone cited in support of their submissions the observations of the First-tier Tribunal in Marriott (at paragraph [42]) about the possible disclosure of the identity of informants in the English Civil War:
“42. The difference arises from the significance to be given to the age of the information. All agree that there must come a time when the disclosure of the identity of an informant who operated in the distant past would not have an effect on the confidence of a current day informant. Or at least one whose inherent paranoia was not so great as to make him or her totally unsuitable to perform the role in any event. To take an extreme example, if a potential informant were to be discouraged from co-operating by the fear that his or her activities would be disclosed after, say, three hundred and fifty years (the equivalent of the disclosure today of those who may have acted as spies during the English Civil War), then one might conclude that his or her paranoia was so intense and irrational that it would not be safe for the police to pursue the recruitment process. Conversely, as the MI5 policy referred to above suggests (supported by the conclusions of the Investigatory Powers Tribunal in the Frank-Steiner case) it would certainly be premature to disclose today information about those acting as informants or agents during the Second World War. But, as one extends further back in time than that, those seeking to retain confidentiality must shoulder a greater burden of demonstrating that the risk of real danger, or of a rational perception of danger, has not diluted to such an extent that the public interest in maintaining secrecy loses much of its weight. In that context it is not just the seniority and experience of those giving evidence that must be considered. The Tribunal must assess the reasoning of an expert witness, no matter how eminent, experienced and knowledgeable he or she may be.”
[63]The argument for the Appellant, in short, was that the requested information in the present case fell into the same category as information about the identity of State informants in the English Civil War. In my view there are two responses to this.[64]First, this is a classic issue of judgment on the facts for the First-tier Tribunal. The majority of the Tribunal in the present case plainly took that into account and explained their assessment:
“38. We say further, in relation to the Appellant's point about the age of the disputed information that there is a defining difference between a revelation of this nature (i.e. about informants in Irish history from the late 19th and early 20th centuries) and similar historical matters from say the 17th century (i.e. the time of the English Civil War). It is that totally different systems were in place. Informants in the Irish conflict were operated by agencies still in existence today and still in full operation. Furthermore, there are lingering embers from this conflict, however tangential they may be. Revelations about informants in 17th century affairs cannot be held the responsibility of anyone or any organisation still existing, whereas a revelation about Ireland could easily be linked to the MPS or MI5. As long as the MPS is operating in this way, anything they did in the past retains strong protection because it would be exactly they and no one else that would be responsible for redeeming the promise of perpetual secrecy.”
Keane v (1) IC (2) Home Office (3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 15[65]Second, the observations of the majority of the First-tier Tribunal in Marriott (at paragraph [42], as set out above) need to be read in their full context. That Tribunal went on to record as follows: “43. The majority were satisfied, on the basis of what those experts said, that the importance of the informant programme to modern policing work is so great that a very cautious approach should be taken before doing anything that those most closely involved with it consider might discourage informants or potential informants. This is not reduced by the fact that some disclosure has taken place via the Clutterbuck and Lowdes publications, or that the identity of an informant is occasionally leaked inadvertently. The deliberate disclosure of a batch of names by the MPS itself, albeit under direction from a tribunal, would have a greater impact than the occasional loss of control over a single name and would be seen as an important precedent. The majority view is that the risk of descendants being traced and targeted should not be ignored. It may be quite small, but the nature of the harm that could result (serious injury or death) is so serious that even a small percentage chance of identification should be avoided. This is so because of both the danger to those descendants and the fact that current day informers would be justified in fearing that at some time in the future their own descendants may be harmed, and their reputation within their community tarnished. The majority say that the potential value of even a single informant in preventing a terrorist outrage is so great that no step should be taken that might conceivably deter him or her from co-operating with the police.[44]On balance the majority view is that the small public interest in disclosure is not outweighed by the also fairly small, but very important, public interest in maintaining the exemption.”[66]Mr Knight acknowledged that the outcome might well be very different if the present appeal had concerned historical records of informants from the English Civil War. Notwithstanding the official policy of perpetual confidentiality, he could envisage that in a case concerning records of such informants from the mid-seventeenth century either the relevant exemption would not be engaged at all or, if it was, the public interest balancing test would favour disclosure. Thus Mr Knight accepted that even if the State’s policy was one of permanent blanket anonymity for informants, a judicial determination by a First-tier Tribunal would not necessarily come to the same answer, irrespective of the period concerned. There would be a tipping point at some stage, at which e.g. the countervailing public interest factors in favour of disclosure would outweigh those supporting the maintenance of the relevant qualified exemption. Where that tipping point occurs is ultimately a question of fact, not law.[67]I agree with that analysis. The present case necessarily turned on its particular facts. This was, on any assessment, a borderline case. I can see, for example, that a respectable case might be made out for saying that the Easter Rising of 1916 marked a step-change in modern Irish history, and that records of State informants preceding that date would have no contemporary repercussions. However, that would be to seek to re-argue the application of the public interest balancing test on the facts, which is not permissible on an appeal confined to errors of law. Whilst some of the Tribunal’s reasoning in this case might have been better expressed, I repeat that I am not persuaded that it erred in law in any material respect.

Conclusion

[68]I therefore conclude that this appeal must be dismissed. The decision of the First-tier Tribunal stands. Keane v(1) IC(2) Home Office(3) MPS [2016] UKUT 0461 (AAC) GIA/3119/2015 16 Signed on the original Nicholas Wikeley on 17 October 2016 Judge of the Upper Tribunal

Cited in 7 later judgments