Upper Tribunal Upholds Decision on Refusal of Information Requests

OUTCOME: Dismissed

[NAME_1] v [NAME_2] & The [NAME_4] (Interested Party: Foreign & Commonwealth Office) [2019] UKUT 262 (AAC)

1

IN THE UPPER TRIBUNAL Case No. GIA/973/2018 ADMINISTRATIVE APPEALS CHAMBER

Before: Upper Tribunal Judge Mitchell

Hearing: 1 July 2019, Field House, Bream’s Buildings, London

Decision: 23 August 2019

Attendances:

Appellant ([NAME_5]):

• Mr P [NAME_6] QC, [NAME_7] and Ms J [NAME_9] of counsel.

1st Respondent ([NAME_3]):

• [NAME_10] of counsel (instructed by the Commissioner’s Legal Services Department)

2nd Respondent (Crown Prosecution Services):

• Mr R [NAME_12] Q.C. of counsel (instructed by the Government Legal Department)

Interested Party (Foreign & Commonwealth Office):

• [NAME_13] of counsel (instructed by the Government Legal Department)

Decision: The decision of the First-tier Tribunal, given on 11 December 2017 (tribunal reference: EA/2017/0041), did not involve the making of an error on a point of law. Under section 11 of the Courts and Enforcement Act 2007, the Upper Tribunal DISMISSES this appeal.

GIA/973/2018 2 REASONS FOR DECISION

Introductory matters

1. I record my thanks to counsel for their assistance at the hearing of this appeal. The case was well- argued by counsel who are more than proficient in the law relating to information rights.

2. In these reasons:

- “1961 Convention” means the Vienna Convention on Diplomatic Relations, done at Vienna on 18 April 1961;

- “[NAME_15]” means the decision of a three-judge panel of the Upper Tribunal in All Party Parliamentary Group on [NAME_16] v [NAME_2] and the Foreign & Commonwealth Office [2015] UKUT 377 (AAC);

- “[NAME_17]” means the [NAME_4];

- “Evans” means the decision of the Supreme Court in R (Evans) v AG [2015] UKSC 21;

- “[NAME_18]” means the Foreign & Commonwealth Office;

- “FOIA” means the Freedom of Information Act 2000;

- “[NAME_19]” means the First-tier Tribunal;

- “NCND” refers to those provisions of FOIA that provide for a public authority to respond to a request for information by neither confirming nor denying that the information is held;

- “SPA” means the Swedish Prosecution Authority;

- “US Departments” means the United States Department of State and the United States Department of Justice.

GIA/973/2018 3 Summary

3. Since these reasons are unusually long, I set out here a summary of my conclusions:

(1) for the purposes of the qualified exemption from disclosure provided for by section 30 FOIA, in relation to information relating to investigations and proceedings conducted by public authorities, competing public interests are to be assessed according to circumstances as they stood when a public authority refused a request for information;

(2) [NAME_20] rightly conducted a public interest balancing exercise according to circumstances as they stood in late 2015, when the [NAME_17] refused [NAME_21]’s request for disclosure of the full correspondence between the [NAME_17] and the SPA concerning a criminal investigation into [NAME_34];

(3) a three-judge panel of the Upper Tribunal in [NAME_15] did not misread the Supreme Court’s decision in Evans. I follow [NAME_15], as a three-judge panel decision, in in the absence of a compelling reason not to do so;

(4) [NAME_21]’s appeal against [NAME_20]’s decision is dismissed in so far as [NAME_20]’s decision relates to [NAME_2]’s decision that [NAME_21] was not entitled to disclosure of the full correspondence between the [NAME_17] concerning a criminal investigation into [NAME_25];

(5) [NAME_20] did not err in law in dismissing [NAME_21]’s appeal against [NAME_2] decision notice in so far as it found that the [NAME_17] were entitled to refuse to confirm or deny whether they held information in the form of correspondence with Ecuadorean authorities about the case of [NAME_25]. [NAME_20]’s public interest balancing exercise was not flawed as a result of being carried out by reference to hypothetically held information relating only to the topic of extradition;

(6) [NAME_20] did not err in law in dismissing [NAME_21]’s appeal against [NAME_2] decision notice in so far as it found that the [NAME_17] were entitled to refuse to confirm or deny whether they held information in the form of correspondence with the US Departments about the case of [NAME_25]. [NAME_20]’s public interest balancing exercise was not flawed by treating [NAME_25]’s interest in disclosure as a personal interest.

GIA/973/2018 4 Background

4. These proceedings involve an appeal against a decision of [NAME_20] given on 11 December 2017, following a hearing on 13 and 14 November 2017. [NAME_20] was comprised of a tribunal judge [NAME_67]. and two members of [NAME_20], [NAME_27] and [NAME_29].

The context

5. The background to this case was described as follows in [NAME_20]’s statement of reasons for its decision:

“28. [NAME_31] is a media organisation which publishes and comments upon censored or restricted official materials involving war, surveillance or corruption, which are leaked to it in a variety of different circumstances. Around February to August 2010 it was reported in the media that [NAME_25] and [NAME_31] were the subject of investigation by the US authorities, following publication of confidential US materials.

29. In August 2010 [NAME_25] made a visit to Sweden. From this there arose some allegations against him of sexual offences involving two women. However, he left Sweden in September 2010 with permission from the SPA. Subsequently a European Arrest Warrant was issued for his detention.

30. He was arrested in London by appointment on 7 December 2010. Extradition proceedings ensued. The first direct contact between the [NAME_17] and the SPA for the purpose of progressing the proceedings was on 10 December 2010.

31. When an extradition request is made on behalf of a foreign judicial authority the [NAME_17] acts as the representative of that authority in the extradition proceedings. This function is assigned to it by the Extradition Act 2003, s.190. This allocates to the [NAME_17], headed by the Director of Public Prosecutions, ‘the conduct of any extradition proceedings’...

32. The same section of the Extradition Act assigs to the [NAME_17] also the function of ‘giving advice on any matters relating to extradition proceedings or proposed extradition proceedings’. [NAME_32] [whom the statement of reasons described as ‘a legal manager who manages the [NAME_17]’s team of extradition lawyers] explained in evidence that it was usual for foreign countries to contact the [NAME_17] for advice prior to making an extradition request. ...34. [NAME_25] challenged the extradition proceedings and was granted bail, subject to compliance with certain conditions. The proceedings went to the UK Supreme Court, which

GIA/973/2018 5 by a majority dismissed his appeal and upheld the arrest warrant on 30 May 2012: Assange v The Swedish Prosecution Authority [2012] UKSC 22. 35. [NAME_25] did not surrender as legally required. Instead on 19 June 2012 he sought refuge in the Ecuadorian Embassy in London. On 16 August 2012 Ecuador granted him a form of diplomatic asylum which has a legal status in the law of Ecuador but which is not recognised by the UK Government or by generally agreed international law. [NAME_25] has remained in the Embassy since then. A police presence was maintained outside the Embassy for over three years in case he came out, at a cost in excess of 11 million. In October 2015 the continuous physical presence of police was replaced by less visible measures.

36. He challenged the arrest warrant by legal proceedings in Sweden and also filed a complaint to the UN Working Group on Arbitrary Detention. His proceedings in Sweden were not successful, but some of the allegations against him have expired owing to lapse of time. The UN Working Group decided in December 2015 that in its view (with one dissenter) he was subject to arbitrary detention. [ADDRESS] of Appeal in Sweden, in a judgment issued on 16 September 2016, expressed its reasoned disagreement with the decision of the Working Group.

37. According to [NAME_21]...his concern has been that, if he is arrested and extradited to Sweden, he may be subject to further onward extradition to the United States to face potential charges there, arising in some way from the leak of US documents. The SPA announced on 19 May 2017 that it was revoking the European Arrest Warrant for [NAME_25]. But [NAME_21] stated that he still has a similar concern about direct extradition from the UK to the US.”

The request for information

6. On 8 September 2015, [NAME_21] requested that the [NAME_17] communicate to her the following information: “(1) the FULL correspondence between the [[NAME_17]] and the Swedish Prosecution Authority concerning the criminal investigation against [NAME_34] (2) the FULL correspondence (if any) between the [[NAME_17]] and Ecuador about the case of [NAME_34] (3) the FULL correspondence (if any) between the [[NAME_17]] and the US Department of Justice about the case of [NAME_25] (4) the FULL correspondence (if any) between the [[NAME_17]] and the US State Department about the case of [NAME_25]

GIA/973/2018 6 (5) the exact number of pages of the [NAME_22] file at the [NAME_17].” 7. Since request (5) is not relevant in these proceedings, I need not say anything about it.

8. The [NAME_17] refused [NAME_21]’s requests, initially on 6 October 2015 and then, on 21 December 2015, in an internal review of their initial decision. The [NAME_17]: - in relation to request (1), relied on section 27(1) & (2) FOIA (international relations exempt information) as well as sections 30(1)(c) (criminal proceedings exempt information) and 40(2) (personal data exempt information); - in relation to requests (2) to (4), the [NAME_17] refused to confirm or deny whether the information was held, relying on section 27(4) (confidential information obtained from another State). 9. [NAME_21] applied to [NAME_2] for a decision whether the [NAME_17] had dealt with her requests in accordance with the relevant requirements of FOIA. There was, as this stage, some alteration in the FOIA exemptions relied on. The Commissioner decided that [NAME_21]’s request was to be dealt with in accordance with the following FOIA provisions: - in relation to request (1), a combination of section 21 (information accessible by other means) and 27(1) & (2) applied. The [NAME_17] were not required to communicate the information requested to [NAME_21]. To the extent that section 27 was relied on, the Commissioner decided that the public interest in maintaining that exemption outweighed the public interest in disclosing the information; - in relation to requests (2) to (4), section 27(4) applied and the public interest in maintaining the exclusion of the duty to confirm or deny outweighed the public interest in disclosing whether the [NAME_17] held the information requested. 10. [NAME_21] appealed to [NAME_20] against the Commissioner’s decision notice. The [NAME_17] were a party to [NAME_20] proceedings. The First-tier Tribunal’s decision 11. It is my understanding that, before [NAME_20], the parties agreed that, in relation to requests (1) to (4), section 30 FOIA was to be applied (public authority investigations and proceedings). Part 1 of [NAME_21]’s request: information within correspondence between the [NAME_17] and the SPA 12. At paragraph 47 of [NAME_20]’s statement of reasons, it notes that “the information withheld in reliance on s.30 consists substantially of instructions and advice passing between the SPA and the [NAME_17]”. The Tribunal also found, which is not disputed, that “extradition proceedings are a form of criminal proceedings [for the purposes of section 30 FOIA] which the [NAME_17] has power to conduct” (paragraph 50).

GIA/973/2018 7 13. Regarding the point in time at which the public interest is to be assessed, the First-tier Tribunal’s statement of reasons records: “17. On behalf of the [NAME_17] [NAME_35] emphasizes, without contradiction from the other parties, that the questions for the Tribunal’s decision are concerned with the correctness of the [NAME_17]’ responses, and hence are to be decided (especially with regard to the public interest balance) by reference to the factual situation as it stood when the [NAME_17] dealt with [NAME_68]’s request”. 14. [NAME_20]’s statement of reasons notes that “the most substantial change since that time is that in about mid-2017 the SPA revoked the European Arrest Warrant and decided to make additional disclosures to [NAME_21]” (paragraph 50).

15. The public interest factors in favour of maintaining the section 30 exemption in relation to part 1 of [NAME_21]’s request for informatoin were described in paragraphs 51 and 52 of [NAME_20]’s statement of reasons: “51…the public interest in maintaining the s.30 exemption arises from the nature of the work done by the [NAME_17] extradition unit. It is generally in the public interest that offences be prosecuted and punished. The purpose of the extradition legislation is to serve the interests of justice by making provision for offenders or suspected offenders to be sent to the country which has prosecuted or is prosecuting them. It is also to ensure that the UK does not become a safe haven for criminals. Further, the existence of effective extradition arrangements provides a reciprocal benefit. When the UK wants to extradite offenders or suspected offenders from another country to the UK, this is much more likely to happen where the sending country benefits from effective extradition arrangements with the UK.

52. The question of public interest in maintaining the exemption therefore demands a focus on the practical requirements for the effective conduct of extradition proceedings, in a way which not only serves the particular proceedings but also is in keeping with the wider goals of ensuring that the UK is not a safe haven and of encouraging other countries in their reciprocal arrangements with the UK.” 16. [NAME_20] went on, in paragraphs 53 to 55 of its statement of reasons, to explain why it considered the relationship between the [NAME_17] and a foreign authority such as the SPA to be “akin to the relationship between lawyer and client”. On this analysis, the holder of the confidence is the foreign authority, which is “in effect the client”, and the [NAME_17] remain bound by an obligation of confidence in relation to such of the information sought as had not been disclosed by the SPA itself. The public interest in maintaining such confidence “is strong, as in the analogous case of maintaining legal professional privilege”, and “it is strong both because it is an obligation still owed to the SPA and because of the potential wider impact on extradition proceedings, both outward and inward” (paragraph 55).

GIA/973/2018 8 17. [NAME_20] identified in paragraph 57 of its statement of reasons considerations “on the other side of the balance”, which it described as “significant public interests in disclosing the withheld information”: (1) disclosure of official information “can promote good government through transparency, accountability, increased public confidence and understanding, the effective exercise of democratic rights, and other related public goods”; (2) “”in support of the more general goals above, there is a public interest in information being made available that can increase public understanding of how extradition proceedings are handled by the [NAME_17]”; (3) in relation to this particular case: “the matter has dragged on unresolved for a long time. The circumstances have also involved a high cost to the public purse. How this came about, and whether the money has been well spent, are matters of legitimate public concern”; (4) “[NAME_25] is the only media publisher and free speech advocate in the Western world who is in a situation that a UN body has characterised as arbitrary detention. It is a matter of public controversy how this situation should be understood. The circumstances of this case arguably raise issues about human rights and Press freedom, which are the subject of legitimate public debate. Such debate may even help to resolve them, which would itself be a public benefit”.

18. However, [NAME_20]’s ultimate conclusion was that the public interest in maintaining the exemption under section 30(1)(c) FOIA, in relation to part 1 of [NAME_21]’s request, outweighed the public interest in communicating the information. In fact, [NAME_20] found that “the balance comes down firmly” on the side of maintaining the exemption (paragraph 67, statement of reasons).

19. On my reading of [NAME_20]’s statement of reasons, it placed particular weight on the following considerations: (1) the position of the [NAME_17] was very similar to that of a lawyer acting for a client and the case law repeatedly emphasises “the importance of maintaining legal professional privilege, and the need for very weighty public interest factors on the other side to tip the balance in a particular case” (paragraph 60); and (2) “the public interest in maintaining the confidence of communications from foreign judicial authorities to the [NAME_17] is important, for the reasons identified above” (paragraph 60).

20. The Tribunal also took into account its findings on certain specific aspects of this case. These are explained in paragraph 64 of the statement of reasons but, for present purposes, I need not set them out. 21. [NAME_21]’s argument that the potentially ‘chilling’ effect of disclosure on extradition proceedings was slight, [NAME_20] found convincing the evidence given by [NAME_32] (the

GIA/973/2018 9 legal manager of the [NAME_17] extradition unit) that “disclosure without the consent of the foreign judicial authority would be likely to damage the functions of the [NAME_17] in extradition proceedings, with the knock-on effects we have mentioned for the relationship with the SPA in particular and with other prosecuting authorities or judicial authorities more generally”. 22. [NAME_20] accepted that the unusual features of this case “make the public interest in disclosure considerably weightier than it otherwise would be” but still not weighty enough to overcome “the strong public interests in maintaining the exemption”. [NAME_20] went on: “the danger from overriding the confidentiality of instructions to and advice from the [NAME_17] in extradition proceedings are real and substantial, and are too great to be outweighed by the general and specific public interest factors on which [NAME_21] relies” (paragraph 67).

23. The Tribunal dismissed [NAME_21]’s appeal against the Commissioner’s decision notice insofar as it related to part 1 of her request for information. Parts 2 to 4 of [NAME_21]’s request: correspondence between the [NAME_17], Ecuadorean authorities and the US Departments 24. In relation to this aspect of the case, as in relation to part 1, the relevant exemption provision was taken to be section 30 FOIA, in particular section 30(3), rather than section 27(4).

25. The Tribunal directed itself that “by definition, the application of s.30(3) does not depend upon what information is actually held. Rather, it depends on consideration of a hypothesis that information of the requisite description may or may not be held” (paragraph 75 of the Tribunal’s statement of reasons). 26. [NAME_20] was informed that “as a matter of long standing policy and practice the UK will neither confirm nor deny that an extradition request has been received until the person concerned is arrested in relation to that request” (paragraph 76).

27. The Tribunal accepted [NAME_32]’s oral evidence that: - the purpose of the policy was to prevent an individual evading justice by learning of an extradition request in advance; - foreign countries usually contact the [NAME_17] before making an extradition request, so the policy applied in those cases as well; - if a foreign authority directed correspondence to the [NAME_17]’ extradition unit it would inevitably concern an actual or contemplated extradition; - the policy needed to be applied consistently “if it was to have the desired effect of not giving any clue to whether an inquiry about or request for extradition had been received from a particular country or not”.

GIA/973/2018 10 28. In relation to parts 3 and 4 of [NAME_21]’s request for information (any correspondence with the US Departments), [NAME_20] found that section 30(3) FOIA was engaged by such correspondence, if it existed (paragraph 80). On the balance of probabilities, correspondence between the [NAME_17] and either Department would be an inquiry about possible extradition or a request for actual extradition (paragraph 80).

29. Regarding part 2 of [NAME_21]’s request (any correspondence with ‘Ecuador’), [NAME_20] found that the application of section 30(3) was less obvious. [NAME_21]’s counsel made a “fair point” that it was unlikely that any such correspondence would be about extradition “given that Ecuador is a country that has proved itself friendly to [NAME_25]” (paragraph 81). 30. [NAME_20] was satisfied on the evidence that the [NAME_17] had no ‘proper role’ in dealing with the Ecuadorian Embassy or other Ecuadorian authorities on behalf of the SPA and that it was unlikely that correspondence within [NAME_21]’s request existed. However, the Tribunal was required to consider the hypothesis that such correspondence might exist and “the unlikelihood of that hypothesis being true is not the point” (paragraph 82).

31. On that hypothesis, the only thing that the correspondence would be about, if it existed, was an inquiry or request concerning [NAME_25]’s extradition to the Republic of Ecuador. Any such information would be held by the [NAME_17] for the purposes of prospective criminal proceedings.

Accordingly, section 30(3) also applied to this part of [NAME_21]’s request (paragraph 82). 32. [NAME_20] accepted the argument of counsel for the [NAME_17] that “making the application of the NCND policy depend upon the likelihood or unlikelihood of a request or inquiry being received from a particular country would tend to undermine [the] usefulness of the policy” (paragraph 82). [NAME_20] then turned to the public interest balancing exercise.

33. The public purpose of the power to bring extradition proceedings would be undermined without a generally consistent NCND policy to prevent express or implied tip-offs. In this case, the purpose of the section 30(3) exemption was to enable such a policy to be followed (paragraph 84). However, the maintenance of a generally consistent policy was not undermined by an occasional exception in appropriate circumstances (paragraph 85). It followed that [NAME_20] needed carefully to consider the public interests favouring making an exception of this case, in order to determine whether “the public interest in maintaining the statutory exemption outweighs them” (paragraph 85); 34. [NAME_20] rejected [NAME_21]’s argument that the unusual circumstances of [NAME_25]’s case justified a departure from the normal policy: “[[NAME_21]] argued this case was a good example of why a blanket refusal was unjustified. [NAME_21] accepted, generally, a strong public interest in avoiding tip-offs. [NAME_25]’s case, however, did not “fit the norm”. He had been in the Ecuadorian Embassy since 2012 and was subject to police surveillance so that, were he to leave the Embassy, his arrest was likely. The public interest in avoiding evasion of extradition proceedings “was fully protected

GIA/973/2018 11 by the ongoing police operation”. Confirming or denying the existence of correspondence with either US Department would not have changed that. The public interest in maintaining the exemption did not begin to outweigh the public interest in favour of the [NAME_17] confirming or denying whether it held the information sought”. 35. [NAME_20] agreed that [NAME_25] had a “strong personal interest” in knowing whether the [NAME_17] had received extradition inquiries or requests from a State other than Sweden. However, [NAME_20] was “unable to see how it would be of more than marginal benefit to the public for that question [concerning extradition enquiries] to be answered”. While the request for information was not expressly linked to extradition, it was necessary to consider the specific question about extradition “because the effect of departing from the NCND policy in this instance would potentially be to answer that question” (paragraph 90).

36. If no correspondence were held, an inference would be drawn that neither extradition enquiries nor requests had been made. If such correspondence were held, the inference would be that extradition had been inquired about or requested. [NAME_20] found that “it is not in the public interest that an individual should be tipped off in such a way” and “the fact that this is a high-profile case does not reduce that public interest”. [NAME_20] considered that it was being asked to decide that the slight public benefits that might accrue from a departure from the NCND policy outweighed the substantial public benefits of maintaining the policy consistently but, in [NAME_20]’s judgment, “the balance indicates that the exemption should be maintained”. There was no special feature concerning [NAME_25]’s position that raised the public benefits of confirming or denying to a level that outweighed the desirability of maintaining the ordinary policy in the public interest (paragraph 91). Legal Framework The general right of access to information held by a public authority 37. Section 1(1) FOIA confers two general entitlements upon “any person making a request for information to a public authority”: (a) the right to be informed in writing whether the public authority holds the information specified in the request. This is referred to by FOIA as “the duty to confirm or deny” (section 1(6)); and (b) if the public authority holds the information, to have the information communicated to the person.

38. The entitlement to information under section 1(1) is to “the information in question held at the time when the request is received”, although account may be taken of amendments or deletions made between the time of the request and the communication of information under section 1(1)(b): section 1(4)).

39. The purpose of much of the rest of FOIA is to place limitations on the general information entitlements under section 1(1).

GIA/973/2018 12 40. Section 1(1) has effect subject to section 2 (amongst other provisions of FOIA). By this means, the general entitlements in section 1(1) are made subject to the system of absolute and qualified exemptions in Part II of FOIA. When the duty to confirm or deny does not apply 41. Under section 2(1), the duty to confirm or deny is potentially inapplicable where “any provision of Part II states that the duty to confirm or deny does not arise in relation to any information”. If the provision confers absolute exemption the duty does not apply. If the provision does not confer absolute exemption, the duty to confirm or deny does not apply if: “in all the circumstances of the case, the public interest in maintaining the exclusion of the duty to confirm or deny outweighs the public interest in disclosing whether the public authority holds the information”. (section 2(1)(b).)

42. In other words, in qualified exemption cases, while the duty to confirm or deny does not ‘arise’, it will nevertheless apply unless the public interest test in maintaining exclusion of the duty to confirm or deny outweighs the public interest in disclosing whether the information is held. To provide, in relation to a single obligation, that it does not arise, but it does apply, does not seem to me to be a very reader-friendly way to craft a qualified entitlement. But this is something that those who rely on, and apply, FOIA need to deal with. When the duty to communicate information does not apply 43. Under section 2(2) FOIA, the section 1(1)(b) duty to communicate information does not apply if the information is exempt information by virtue of any provision of Part 2, and: (a) the information is exempt by virtue of a provision conferring absolute exemption; or (b) if not absolutely exempt, “in all the circumstances of the case, the public interest in maintaining the exemption outweighs the public interest in disclosing the information”.

44. The list of FOIA provisions conferring absolute exemption is set out in section 2(3). None of these is in issue in the present case.

45. It can be seen from the above description that there are two categories of exempt information. One which is absolutely exempt from disclosure under section 1(1)(a) and another which is not.

Refusal of requests 46. Where a public authority relies on section 2(1)(b) FOIA (confirm or deny cases) or section 2(2)(b) (communication of information cases), section 17(3) requires the authority to give the [NAME_36] a notice that states “…the [authority’s] reasons for claiming—

GIA/973/2018 13 (a) that, in all the circumstances of the case, the public interest in maintaining the exclusion of the duty to confirm or deny outweighs the public interest in disclosing whether the authority holds the information, or (b) that, in all the circumstances of the case, the public interest in maintaining the exemption outweighs the public interest in disclosing the information”.

Exemption: public authority investigations and proceedings (section 30)

47. Section 30 FOIA exempt information does not attract absolute exemption.

48. Section 30(1) provides that information held by a public authority is exempt information if “…it has at any time been held by the authority for the purposes of— (a) any investigation which the public authority has a duty to conduct with a view to it being ascertained— (i) whether a person should be charged with an offence, or (ii) whether a person charged with an offence is guilty of it, (b) any investigation which is conducted by the authority and in the circumstances may lead to a decision by the authority to institute criminal proceedings which the authority has power to conduct, or (c) any criminal proceedings which the authority has power to conduct.” 49. A further category of exempt information is provided for by section 30(2) that is information where: (a) it was obtained or recorded by the authority for the purposes of its functions relating to— (i) investigations falling within subsection (1)(a) or (b), (ii) criminal proceedings which the authority has power to conduct, (iii) investigations (other than investigations falling within subsection (1)(a) or (b)) which are conducted by the authority for any of the purposes specified in section 31(2) and either by virtue of Her Majesty's prerogative or by virtue of powers conferred by or under any enactment, or (iv) civil proceedings which are brought by or on behalf of the authority and arise out of such investigations, and

GIA/973/2018 14 (b) it relates to the obtaining of information from confidential sources.” 50. The purposes specified in section 31(2) (referred to in section 30(2)(a)(iii)) include: (a) the purpose of ascertaining whether any person has failed to comply with the law (section 31(2)(a));

(b) The purpose of ascertaining whether any person is responsible for any conduct which is improper (section 31(2)(b). 51. Section 30(3) provides that the duty to confirm or deny does not arise in relation to information which is exempt under section 30(1) or (2) or would be so exempt were it held by the public authority.

Different types of qualified exempt information: consequences of disclosure 52. This case involved the exemption in section 30 FOIA. To fall within this category, information must be “held” by a public authority and, in addition, be information which:

(a) “has at any time been held” for a purpose specified in subsection (1);

(b) “was obtained or recorded” for a purpose specified in subsection (2)(a); or

(c) “relates to the obtaining of information from confidential sources”.

53. Whether information is section 30 exempt information turns on questions of pure or primary fact. Is the information held by the public authority? Has the information at any time been held for a purpose specified in subsection (1)? And so on.

54. Many other types of exempt information, by contrast, are established by reference to the consequences of disclosure of the information “under this Act” (these exemptions are sometimes referred to as harm-based exemptions). For example, section 31(1), provides that information is exempt information “if its disclosure under this Act would, or would be likely to, prejudice” various matters such as the prevention or detection of crime. In these cases, determining whether information is exempt still involves findings of fact, but they are of a different character to the findings called for under section 30. Since the consequences of disclosure is key, which of necessity has not yet occurred, the fact-finding task involves having to make a prediction.

[NAME_2]’s enforcement functions

GIA/973/2018 15 55. Any person (“the complainant”) may apply to the Commissioner for a decision whether, in any specified respect, a request for information has been dealt with in accordance with the requirements of Part I (section 50(1)).

56. Unless section 50(2) permits the Commissioner not to make a decision on the application, the Commissioner must serve notice of her decision (referred to by the Act as a “decision notice”) on the complainant and the public authority.

57. If the Commissioner decides that a public authority has failed to communicate information, or to provide confirmation or denial, a decision notice “must specify the steps which must be taken by the authority for complying with that requirement and the period within which they must be taken” (section 50(4)).

The right of appeal to the First-tier Tribunal 58. Following service of a decision notice, the complainant or public authority may appeal to [NAME_20] “against the notice” (section 57(1)).

59. In two cases, [NAME_20] must either allow an appeal or “substitute such other notice as could have been served by the Commissioner”. Provided for in section 58(1), those cases are: (a) “that the notice against which the appeal is brought is not in accordance with the law”, or (b) “to the extent that the notice involved an exercise of discretion by the Commissioner, that [she] ought to have exercised [her] discretion differently”.

60. If neither case applies, [NAME_20] must dismiss the appeal. Section 58(2) also provides that, on an appeal, [NAME_20] may “review any finding of fact on which the notice in question was based”.

The date at which the public interest is to be assessed: case law 61. Various judicial decisions (and views) have been given about the date at which the public interest is to be assessed for the purposes of FOIA’s qualified exemptions.

62. As originally enacted, FOIA provided, in England and Wales, for a right of appeal on a point of law to the High Court against decisions of an Information Tribunal. By this route, the High Court ([NAME_37], as he then was) in Office of Government Commerce v The Information Commissioner [2008] All ER (D) 169 addressed the public interest timing point although his views were clearly obiter. They were preceded by the statement that the judge had reached no “final conclusion” (paragraph 67).

GIA/973/2018 16 63. In [NAME_40], the issue was whether a change of circumstances, namely the coming into force of new legislation after a public authority refused a request under section 1(1) FOIA, “could be taken into account by the Commissioner”. The Commissioner argued that, under FOIA, “questions of disclosure [are] to be determined on the basis of the facts at the date of the request” (paragraph 97). The “circumstances” referred to in section 2 FOIA are the circumstances at the date of the request, a point made clear by section 50’s requirement for the Commissioner to decide whether a request for information ‘has been’ dealt with in accordance with the requirements of Part 1 of FOIA. The tense used in section 50, argued the Commissioner, clearly referred back to the date of the request. If a relevant change of circumstances occurred, it would be open to the applicant to make a fresh request under FOIA. 64. [NAME_37] was “not sure” that he agreed with the Commissioner. He posited a case in which the information requested was relevant to criminal proceedings that began after the request was made but where subsequent disclosure would prejudice the fairness of a trial. In such a case, the information would not be exempt when requested but would subsequently take on the character of exempt information (under section 31). [NAME_37] commented that “it would be undesirable for the Commissioner to be obliged to require disclosure in such a case” and expressed the following view: “98…It seems to me to be arguable that the Commissioner’s decision whether a public authority complied with Pt I of the Act may have to be based on circumstances at the time of the request for disclosure of information, but that his decision as to the steps required to be taken by the authority may take account of subsequent changes of circumstances”.

65. The Upper Tribunal’s decision in [NAME_41] v [NAME_2] and The Badger Trust ([2014] UKUT 526 (AAC) was given by a panel comprised of Warren J, Judge [NAME_42] (a member of the Upper Tribunal rather than a judge). In other words, the Upper Tribunal was not constituted as a three-judge panel. The Upper Tribunal did not rule on the question “at what date should the public interest balancing exercise be conducted” but it did give the following opinion: “45. It seems to us that there is some lack of clarity about the date at which the First-tier Tribunal assesses whether an exemption or exception applies and, if so, where the balance of the public interest lies. Early cases looked at the date of the information request; many Tribunals now seem to look at the date of the public authority’s final decision on the request. This approach has been doubted in the High Court (see OGC v IC [2008] EWHC 774 (Admin) [2010] QB 98 and seems not entirely consistent with the development of the First-tier Tribunal’s role of receiving new evidence and conducting what is now well-established as a full merits review.”

GIA/973/2018 17 66. R (Evans) v AG [2015] UKSC 21 was a decision of the Supreme Court which, again, gave views but did not rule on the correct date according to which the public interest was to be satisfied. A seven-judge panel of Supreme Court justices decided Evans. [NAME_44]’s opinion, at para. 73, included:

“although the question whether to uphold or overturn…a refusal by a public authority must be determined as at the date of the original refusal, facts and matters and even grounds of exemption may, subject to the control of the Commissioner or the tribunal, be admissible even though they were not in the mind of the individual responsible for the refusal or communicated at the time of the refusal to disclose (i) if they existed at the date of the refusal, or (ii) if they did not exist at that date, but only in so far as they throw light on the grounds now given for refusal – see [NAME_6] on Information Rights 4th ed (2014), paras 28-022 and 28-024, and Department for The Environment, Food and Rural Affairs v Information Comr (Birkett) [2011] EWCA Civ 1606…”

67. Two Justices expressly agreed with [NAME_44]. The remaining Justices did not expressly disagree.

68. Finally, and most importantly, I come to a decision given by a three-judge panel of the Upper Tribunal (Charles J, Mitting J and Upper Tribunal Judge Wikeley) in All Party Parliamentary Group on [NAME_16] ([NAME_15]) v [NAME_2] and the Foreign & Commonwealth Office [2015] UKUT 377 (AAC). In the light of the arguments presented on this appeal, I must look at [NAME_15] in some detail.

69. [NAME_15] made a number of requests for information to various Government departments and, subsequently, brought a number of appeals to [NAME_20] against decision notices given by the Commissioner. The Upper Tribunal’s decision in [NAME_15] arose from appeals against three decision notices. For the most part, [NAME_20] had dismissed [NAME_15]’s appeals. [NAME_15] appealed to the Upper Tribunal against [NAME_20]’s decisions on five grounds. One ground succeeded and resulted in [NAME_20]’s decisions being set aside. That ground was concerned with the construction of section 27(1) FOIA rather than the public interest timing point.

70. All parties in [NAME_15] agreed that the Upper Tribunal should re-make [NAME_20]’s decisions rather than remit the case to that tribunal for re-determination. To this extent the Upper Tribunal in [NAME_15] was not exercising its appellate function under section 12(1) of the Tribunals, Courts and Enforcement Act 2007. The appeal had already been allowed and the Upper Tribunal then decided to exercise its discretion under section 12(2) of the Tribunals, Courts and Enforcement Act 2007 to re- make a set aside decision rather than remit to [NAME_20].

GIA/973/2018 18

71. The “public interest timing point”, as the Upper Tribunal described it, arose in relation to the information within two documents. The reasons for the decision record that, before [NAME_20], all parties agreed that the public interest balance was to be assessed, at the latest, at the date of the Foreign Office’s internal review of its initial decision not to disclose the information. The Upper Tribunal said “this shared understanding was in accord with the prevailing orthodoxy” (paragraph 44).

72. Some doubt was cast over the ‘prevailing orthodoxy’ by the views expressed in [NAME_41]. For this reason, the Upper Tribunal in [NAME_15] permitted the parties to make submissions on the public interest timing point although it appears the issue was not argued orally. All parties made “detailed written submissions” but only counsel for [NAME_15] made oral submissions. The Upper Tribunal did not consider it necessary for the other parties to respond orally to [NAME_15]’s submission that the Upper Tribunal should depart from the “traditional understanding” namely that public interest is to be assessed according to circumstances at the date of a public authority’s refusal decision.

73. The Upper Tribunal rejected [NAME_15]’s argument and analysed the public interest timing point as follows:

(a) Evans provided “powerful support” for the orthodox approach; “the Supreme Court observed (and all the parties agreed) that the timing of the assessment of public interest is the date of the public authority’s refusal” (paragraph 48);

(b) while the Supreme Court’s observations were “technically obiter”, they were “at the “almost virtually binding” end of the spectrum of “highly persuasive dicta” from the highest court in the land, not least because…it is very difficult to make sense of [NAME_44]’s leading judgment…if [NAME_15] is correct on the public interest timing point” (paragraph 49);

(c) had the Supreme Court’s decision in Evans pre-dated [NAME_41], “we are confident…the Upper Tribunal would not have raised the question mark it did on the orthodox understanding” (paragraph 51);

(d) Evans “confirms and powerfully supports the view that taken as a whole, the language of the statutory scheme indicates that the Commissioner (and [NAME_20]) is charged with assessing past compliance with FOIA, not with monitoring ongoing compliance” (paragraph 52);

GIA/973/2018 19 (e) while it was well-established that [NAME_20] may consider evidence post-dating the public authority’s decision, “there is nothing unusual about a decision maker taking account of later evidence to inform a historical position” (paragraph 53);

(f) whichever construction were adopted, unfortunate practical consequences might follow (paragraph 54);

(g) “there are clearly disadvantages in the Commissioner and then [NAME_20]…being faced with a moving target on public interest issues” and “Parliament would not have intended that the public authority would effectively be removed as the decision maker because [of] the passage of time and changes in circumstances…Parliament would have intended that the [NAME_36] should make a further request if he wished to rely on changes over time to the public interest factors” (paragraph 56);

(h) “this view of Parliamentary intention and the conventional understanding do not result in any asymmetric unfairness as to the relevance of post-assessment developments”. The Upper Tribunal went on to note the Commissioner’s residual discretion under section 50(4) FOIA not to order disclosure “in exceptional circumstances to avoid a disclosure that should have been made earlier but now should not be because of changes in circumstances” (Information Commissioner v H.M.R.C. & Gaskell [2011] UKUT 313) (paragraph 57).

1961 Vienna Convention on Diplomatic Relations

74. Both the United Kingdom of Great Britain and Northern and the Republic of Ecuador have ratified the 1961 Convention. Article 24 of the Convention provides that “the archives and documents of the mission shall be inviolable at any time and wherever they may be”.

75. Article 27(1) of the Convention provides that “the receiving State shall permit and protect free communication on the part of the mission for all official purposes…”. Article 27(2) provides that “the official correspondence of the mission shall be inviolable” and “official correspondence means all correspondence relating to the mission and its functions”.

76. Section 2(1) of the Diplomatic Privileges Act 1964 provides that “subject to section 3 of this Act [which provides for Orders in Council to withdraw privileges and immunities], the Articles set out in Schedule 1 to this Act (being Articles of the Vienna Convention on Diplomatic Relations signed in 1961) shall have the force of law in the United Kingdom”.

77. Articles 24 and 27 of the Convention are set out in Schedule 1 to the 1964 Act.

GIA/973/2018 20 78. Article 41(2) of the Convention, which is not set out in Schedule 1, provides that “all official business with the receiving State entrusted to the mission by the sending State shall be conducted with or through the Ministry for Foreign Affairs of the receiving State or such other ministry as may be agreed.”

The grounds of appeal

Ground 1

79. The first ground of appeal is whether [NAME_20] misdirected itself in law in determining that the public interest was to be assessed at the date on which the [NAME_17] finally refused [NAME_21]’s request for disclosure.

Ground 2

80. It is not disputed that [NAME_20] correctly proceeded, for NCND purposes, on the hypothesis that correspondence between the [NAME_17] ‘authorities’ existed. However, the actual hypothesis applied by [NAME_20] is criticised because it “only envisaged the existence of correspondence where the [NAME_17] had a proper role” that is a request by the Republic of Ecuador for [NAME_25]’s extradition. That cannot have been the correct hypothesis, argues [NAME_21]. She draws attention to the inherent improbability of a State seeking the extradition of an individual whom it houses in its Embassy after having granted him a form of asylum. Permission to appeal was granted on the ground that arguably the Tribunal erred in law by failing to conduct its NCND analysis according to the actual terms of part 2 of [NAME_21]’s request for information.

81. In granting permission to appeal, I directed that ground 2 was to include certain matters relating to the 1964 Act, described as follows in the grant of permission to appeal:

“Even if the First-tier Tribunal deployed a flawed hypothesis, that might not be material if the Freedom of Information Act 2000 is to be read as subject to the Diplomatic Privileges Act 1964, to the extent that disclosure under the 2000 Act would be contrary to the Convention. But, even if that is not the case, for example if a diplomatic letter is not part of a diplomatic mission’s “archives or documents”, might it be argued that, in carrying out the public interest analysis required by FOIA…the fact that the information requested is comprised within correspondence sent by a diplomatic mission must be given particular weight?... ground 2 is to be read as including arguments that the First-tier Tribunal arguably erred in law by:

GIA/973/2018 21 (a) failing to consider whether the 1961 Convention was relevant to the issues it had to decide; and

(b) in constructing its hypothesis about correspondence between Ecuadorian authorities, including its Embassy, and the [NAME_17], failing to take account of Article 41 of the Convention with its requirement for a diplomatic mission to conduct its business with Ministries of the receiving state.”

82. Following the grant of permission to appeal, the Secretary of State for Foreign and Commonwealth Affairs applied to be made an interested party in these proceedings. The application was granted. The Secretary of State’s interest is limited to ground 2 in so far as it raises questions about the 1961 Convention and the 1964 Act.

Ground 3

83. Permission to appeal was granted on the ground advanced for [NAME_21] namely that [NAME_20] arguably erred in law, in its NCND public interest analysis relating to parts 3 and 4 of the request for information (correspondence with US Departments), by failing to take into account certain matters. The matters were the Tribunal’s own findings that weighty public interest arguments supported disclosure of extradition-related information about [NAME_25], and the public interest in [NAME_25] being able to see such information.

The arguments

84. Without intended to show any disrespect, I shall not refer to the various First-tier Tribunal decisions cited in the parties’ submissions, especially those of [NAME_46] for [NAME_21]. [NAME_21]’s pre-hearing written submissions were drafted by [NAME_48] of counsel. [NAME_47] presented [NAME_21]’s case at the hearing. If I inadvertently ascribe [NAME_48]’s submissions to [NAME_47], or vice versa, I apologise.

85. Adopting the terminology used by the parties and [NAME_15], I use ‘public interest timing point’ to describe the question of law as to the correct date according to which the balancing of public interests under section 2 FOIA is to be carried out.

Ground 1 – whether it should be determined at all

86. [NAME_21]’s application to the Upper Tribunal stated that the “timing of the public interest was not canvassed at the [[NAME_19]] hearing but that should not prevent the Upper Tribunal from considering

GIA/973/2018 22 this important point on which previous authorities have spoken inconsistently”. This was not an academic point. By the date of the hearing, “the situation had changed considerably”. At some point in 2017 the SPA gave the [NAME_17] its express consent to disclose much of the correspondence previously taken to be confidential. An initial disclosure on 3 August 2017 comprised 336 pages of correspondence, followed by a further disclosure (with redactions) on 17 November 2017. The facts ‘around confidentiality’ had moved on considerably by the date of [NAME_20] hearing. At the hearing before myself, [NAME_46], for [NAME_21], placed as much emphasis on the UN Working Group’s report about [NAME_25] as a significant change in the relevant circumstances. As noted above, that report was referred to in the First-tier Tribunal’s statement of reasons.

87. [NAME_49] for the Commissioner and [NAME_50] for the [NAME_17] argue that the Upper Tribunal should consider exercising its case management powers so as to prevent [NAME_21] from relying on a point that she could have taken before [NAME_20]. I think it is fair to say that [NAME_49] did not press this argument with particular force at the hearing, although [NAME_35] did.

88. [NAME_50] submits that, had the public interest timing point been in issue before [NAME_20], the [NAME_17] might have adduced different evidence. It would be inconsistent with the overriding objective of the Upper Tribunal’s procedural rules and unfair to allow an appeal on the basis of a point not taken in the lower court (Crane T/A Indigital Satellite Services v Sky In-Home Ltd [2008] EWCA Civ 978 at [20]-[23]).

Ground 1 – construction of FOIA

89. For [NAME_21], it is argued that section 58 FOIA’s requirement for [NAME_20] to consider “the matter afresh on the evidence before it” indicates that [NAME_20] was not restricted to considering the public interest at the date of the [NAME_17]’s refusal decision in December 2015. Written submissions argued that any case law to the contrary was erroneous and should not be followed.

90. [NAME_20]’s role as a ‘merit-review tribunal’ adds weight, argues [NAME_51]. for [NAME_21], to her case on the public interest timing point. Furthermore, appeals to [NAME_20] are against the Commissioner’s decision notice, not a public authority’s section 17 FOIA refusal notice. The exercise of power by a decision-maker (i.e. the Commissioner) does not exhaust the Tribunal’s ability to exercise that power: see section 12 of the Interpretation Act 1978:

“Where an Act confers a power or imposes a duty it is implied, unless the contrary intention appears, that the power may be exercised, or the duty is to be performed, from time to time as occasion requires.”

GIA/973/2018 23 91. In [NAME_51]’s submission, section 50(4) FOIA recognises the continuing requirement imposed on a public authority by section 1(1). He argues that the Respondents read section 50(4) as if it referred to a public authority having “failed to communicate information…in a case where it was required to do so by section 1(1)”. The wording actually used – “failed to communicate information…in a case where it is required to do so” – reveals the legislative intention namely that the Commissioner’s decision notice addresses section 1(1) questions as matters stand at the date of the notice, not at the date of the refusal decision. In turn, section 58 provides the same result on appeal to [NAME_20]. [NAME_20] stands in the shoes of the Commissioner and is therefore required to consider section 1(1) questions as matters stand at the hearing.

92. Section 58(2) FOIA confers power on [NAME_20] to review any finding of fact on which a decision notice was based. [NAME_51]. submits that ‘review any finding of fact’ is a term of art with an established legal meaning. It describes a re-hearing leading to a fresh pronouncement of the rights of the parties ([NAME_52] v Secretary of State for the Home Department [2015] EWCA Civ 1175), which is consistent with [NAME_20] determining an appeal according to current circumstances. The overall aim of [NAME_20]’s procedural rules, set out in section 22(4) of the Tribunals, Courts and Enforcement Act 2007, also supports [NAME_21]’s case. Section 22(4) provides that tribunal rules are to be made with a view to securing that justice is done, the tribunal system is accessible and fair and proceedings are handled quickly and fairly.

93. [NAME_51]. submits that his argument on the public interest timing point is also consistent with [NAME_20]’s power to determine a request for information vexatious in the light of post-decision evidence as well as its power to refuse to order disclosure due to post-request exceptional circumstances (Sturmer v IC & Derbyshire CC [2015] UKUT 568 (AAC).

94. [NAME_51]. cites the House of Lords’ decision in [NAME_53] v Secretary of State for the Home Department [2007] UKHL 57 in which their Lordships held at [2] that “common sense indicates that the final decision, whenever it is made, should be based on the most up-to-date evidence that is available”. [NAME_54] describes this as judicial authority in a cognate area and submits that general consistency across the tribunal system is commendable. [NAME_47] also cites two Australian decisions about the grant of pilots’ licences but I see no need to refer to these despite his assertion that they concern the operation of the Administrative Appeals Tribunal of Australia a body which he further submits was a source of considerable inspiration for the Leggatt Report (Tribunals for Users – One System, One Service, August 2001). Many of the Leggatt Report recommendations were given legislative effect in the Tribunals, Courts and Enforcement Act 2007, which created [NAME_20].

95. In essence, the Respondents argue that the Upper Tribunal’s construction of FOIA in [NAME_15] was correct and meets all the points now relied on [NAME_21].

GIA/973/2018 24 Ground 1 – the Upper Tribunal’s decision in [NAME_15] and its reading of Evans

96. [NAME_21] submits that [NAME_15] is usually cited as the leading authority for the proposition that the public interest is to be considered as matters stood at the date of a public authority’s refusal decision. However, [NAME_15] relied on a mistaken reading of the Supreme Court’s decision in Evans which did not specifically endorse an assessment of the public interest at the date of a public authority’s refusal.

97. [NAME_21]’s written submissions argue that, if read correctly, it is clear that Evans struck a ‘middle way’ regarding the public interest timing point. While the Supreme Court noted common ground that the relevant date was the date of the refusal decision, it went on to emphasise that facts and matters and even grounds of exemption may, in principle, subsequently be admissible even if they were not in the mind of the original decision maker. This was subject to the following requirements: (i) the fact, matter or exemption existed at the date of refusal or (ii) if not, only insofar as they throw light on the ground now given for refusal (paragraph 73 of Evans). At the hearing, however, it seemed to me that [NAME_46] placed less emphasis on this point. He submits that, in Evans, [NAME_44] was arguably simply ‘stress-testing’ his views.

98. [NAME_21]’s written submissions argue that the practical result of Evans is clear: “this [newly admissible] evidence may throw so much light on the grounds for refusal that, in effect…the Tribunal is considering the public interest factor at the time of its decision”. To hold that a new ground of exemption may be raised but post-decision supporting facts may not be relied on amounts to a baffling inconsistency of approach. It is also inconsistent with the absence of any [NAME_19] power to remit a request for information to a public authority for reconsideration.

99. According to [NAME_21]’s written submission’s [NAME_15]’s reasoning is ‘indicative’ and acknowledges alternative indications. Further, the reliance placed on Evans by [NAME_15] was misplaced given that Evans arose as a claim for judicial review of a certificate given by an accountable person under section 53 FOIA. Upper Tribunal authority supports [NAME_21]’s case: see Soh v Information Commissioner & Another [2016] UKUT 249 (AAC).

100. If one stops beating around the bush, argues [NAME_49] for the Commissioner, [NAME_21]’s submission is that the Upper Tribunal got it wrong in [NAME_15]. And it did so because it misunderstood Evans. However, Evans, at paragraph 73, clearly finds that “a refusal by a public authority must be determined as at the date of the original refusal”. The Supreme Court’s subsequent discussion of the admissibility of evidence, arguments and grounds was all posited on those matters throwing “light on the ground now given for refusal”. The reference to grounds ‘now given’, if considered in its proper context, was clearly a reference to exemptions raised after the public

GIA/973/2018 25 authority’s internal review of an initial refusal and/or the exercise of the section 53 FOIA ‘ministerial veto’. There might be some degree of textual ambiguity in the wording in paragraph 73 of [NAME_44]’s opinion but there is no substantive ambiguity.

101. [NAME_49] submits that, in [NAME_15], the Upper Tribunal considered precisely the same point as is at the heart of ground 1 in this appeal. Further, the Upper Tribunal in [NAME_15] rejected similar arguments to those now relied on by [NAME_21]. [NAME_49] acknowledges that, in [NAME_15], the Upper Tribunal considered the public interest timing issue as part of the process of re-making a [NAME_19] decision that it had set aside. He argues that the Upper Tribunal’s rulings on the public interest timing point were nevertheless part of the ratio for its decision. That Evans involved a ministerial certificate under section 53 does not detract from the general applicability of [NAME_44]’s views. [NAME_21] does not identify any legislative feature upon which a proper distinction might be drawn regarding the public interest timing point between certificate cases and the more typical FOIA case that comes before [NAME_20] on an appeal against the Commissioner’s decision notice.

Ground 1 – the precedential status of [NAME_15]

102. [NAME_51], for [NAME_21], argues that, for two reasons, I am not bound by [NAME_15] on the public interest timing point. Firstly, what was said was obiter (which, if correct, means it was not even binding on [NAME_20]). Secondly, even if the Upper Tribunal’s views were part of the ratio for its decision, they do not bind me sitting as a sole judge of the Upper Tribunal.

103. [NAME_51]. notes that the Upper Tribunal’s findings on the public interest timing point were given during the process of re-making [NAME_20]’s decision and this should be taken into account.

104. The Respondents and the Interested Party dispute that [NAME_15]’s reasoning on the public interest timing point involved obiter expressions of opinion. The Upper Tribunal’s findings of law were part of the ratio for its decision and therefore bound [NAME_20] and should also be followed by myself, sitting as a sole judge of the Upper Tribunal. At the hearing, [NAME_49] made detailed submissions, drawing attention to each relevant paragraph in [NAME_15], to show, he argued, that the Upper Tribunal’s findings on the public interest timing point were clearly an essential part of the reasoning on which its decision was based. The Upper Tribunal’s final decision disposing of the proceedings in [NAME_15], submits [NAME_49], involved it making findings on the public interest as matter stood at the refusal date in [NAME_15].

105. Both Respondents argue that I should consider myself effectively bound by [NAME_15] on the public interest timing point. That follows from the decision of a three-judge panel of the Administrative Appeals Chamber of the Upper Tribunal in Dorset Healthcare [NAME_1] v [NAME_2] & The [NAME_4] (Interested Party: Foreign & Commonwealth Office) [2019] UKUT 262 (AAC)

GIA/973/2018 26 v MH [2009] UKUT 4, which provides, in paragraph 37, that a sole judge of the Upper Tribunal, Administrative Appeals Chamber, should follow the decision of a three-judge panel of that Chamber unless there were “compelling reasons” not to do so. The panel gave an example of such reasons namely a decision of a superior court affecting the legal principles involved. There are no such compelling reasons in the present case. I am effectively bound to follow [NAME_15].

106. [NAME_51]. disputes that I am effectively bound by [NAME_15]. Dorset Healthcare gave only “tentative” guidance about the precedential status of three-judge panel decisions. [NAME_49] responds that this argument cannot be right if one considers the actual wording used by the Upper Tribunal. The principles articulated were firm and clear. Only in the case of compelling reasons should a sole judge of the Upper Tribunal’s Administrative Appeals Chamber decline to follow a decision of a three-judge panel of that Chamber.

107. [NAME_51]. submits that a three-judge panel of the Upper Tribunal could have been convened to determine this appeal. However, I am satisfied that no party made a request for the appeal to be determined by a three-judge panel. I also observe that, on [NAME_21]’s case, there would be little point in convening a three-judge panel. She argues that the precedential status of three-judge panel decisions, so far as sole judges of the Upper Tribunal’s Administrative Appeals Chamber are concerned, is weak at best.

108. In any event, submits [NAME_49], [NAME_15] was correctly decided whether or not I am bound to follow it. The FOIA statutory scheme requires the Commissioner to decide whether the public authority’s response was correct and, in turn, [NAME_20] steps into the shoes of the Commissioner and performs the same exercise afresh (section 58 FOIA).

Ground 1 – absurdities and consequences

109. The Respondents’ case gives rise to absurdities, argues [NAME_51]. for [NAME_21]. If the facts change post-decision in a manner favourable to a [NAME_36], would s/he be entitled to proceed with an appeal even though a fresh request would have a greater likelihood of success than the original? Conversely, if circumstances change adversely, from a [NAME_36]’s point of view, should the public authority not be entitled to rely on the altered factual context in its arguments concerning the balance of public interests? And, if a public authority may rely on such new facts, why not an appellant?

110. The Respondents’ case on the public interest timing point, argues [NAME_51], raises a host of practical difficulties. [NAME_20] be permitted to take into account facts in existence at the date of the decision in question but which could not have been known to the decision maker? Or only

GIA/973/2018 27 those facts that should have been known? How does the Tribunal respond to a post-decision change in the law? What happens if, post-decision, large quantities of information are released into the public domain? Would a public authority be entitled to raise a new section 21 FOIA exemption?

111. [NAME_51]. submits that changing circumstances will affect a whole host of complaints and appeals under FOIA. Take the example of the section 26 exemption (defence). Surely if the public interest backdrop alters during the life of a FOIA challenge, such as global events making for a less peaceful world, it cannot be right for a tribunal to order disclosure of information if that would prejudice the Britain’s defence in such a way as to tip the public interest balance against disclosure. The same point is illustrated, perhaps even more starkly, by the health exemption in section 38 FOIA.

112. [NAME_49], for the Commissioner, accepts that the hypothetical changes of circumstances identified by [NAME_51], such as a changing national security context, might, on the face of it, raise doubts as to whether it is always correct to assess the public interest as at the date of the refusal decision. On a closer analysis of FOIA, however, these concerns are capable of being addressed by the Commissioner’s ‘steps discretion’ under section 50(4) FOIA. This is confirmed by various case law authorities.

113. [NAME_51]. disagrees with [NAME_49] that, in some cases, the practical difficulties associated with the public interest being fixed at the refusal date could be overcome by a new request being made, i.e. in those cases where the public interest balance had altered to favour disclosure. This would put the applicant at the mercy of the repeat application provisions of section 14(2) FOIA.

114. [NAME_49] argues that [NAME_51]’s construction itself involves significant practical objections. If [NAME_20] considers the public interest according to current circumstances, what it is to do where, as in this case, the decision was given some period of time after the appeal was heard? Is it supposed the check whether public interest factual considerations have moved on in the meantime? When would it all end? And, if [NAME_21] succeeds and this matter is remitted to [NAME_20], is it really feasible for it to consider the public interest at some point in 2020 (the date it would be likely to re- hear a remitted appeal)? These points illustrate the ‘moving target’ concerns referred to, and dealt with, by the Upper Tribunal in [NAME_15].

Ground 1 – is it academic?

115. Both Respondents argue that [NAME_20] did in fact address the public interest according to the circumstances at the date on which it heard [NAME_21]’s appeal. In other words, it provided alternative reasoning that renders ground 1 academic. In response, [NAME_51]. for [NAME_21]

GIA/973/2018 28 says this is no more than ‘say-so’. The Upper Tribunal cannot be confident that the same outcome would have been reached had [NAME_20] directed itself that the public interest balance was to be struck according to current circumstances.

116. [NAME_56], for the [NAME_17], argues it is not merely ‘say-so’ to argue that, had [NAME_20] focussed principally on the public interest at the date of the hearing, it would still have arrived at the same conclusion and dismissed [NAME_21]’s appeal. The point was expressly considered in footnote (11) of the Tribunal’s statement of reasons. [NAME_35] also draws attention to other parts of the Tribunal’s statements of reasons, such as paragraphs 55 and 57, that used the present tense, showing, he argues, that it gave more detailed consideration to the public interest balance as matters stood at the date of the hearing than might be suggested by footnote (11) alone.

Ground 1 – whether ‘policy’ dictates determination of the public interest timing point according to current circumstances

117. At the hearing, [NAME_51]. for [NAME_21] developed an ambitious argument concerning the general policy underpinnings of the entire tribunal system under the Tribunal, Courts and Enforcement Act 2007 as well as the relationship between citizen and state. This supports, he submits, the argument that [NAME_20] must consider the public interest at the time it decides an appeal against a decision notice. In summary, the chain of reasoning is: (1) [NAME_20]-is not a judicial body. While it is required to act judicially, it forms part of the executive branch of the state; (2) the executive branch inevitably determines entitlements according to current circumstances; (3) [NAME_20] in the exercise of its FOIA jurisdiction must have been intended by Parliament to balance public interests according to current circumstances.

118. As I understood it, [NAME_51]’s argument relies in part on the fact that [NAME_20] is charged with carrying out a ‘full merits-based review’. [NAME_47] submits that, in the absence of contrary statutory provision, an example being section 16(4) of the Finance Act 2000, Parliament is to be taken to have intended a full merits review once a case comes before a tribunal. Such a review is bound to involve determining a case according to current circumstances. There is no such contrary provision in FOIA. This distinguishes [NAME_20] from bodies that are undoubtedly judicial, such as the High Court in the exercise of its judicial review jurisdiction, and support his ‘executive branch’ argument.

119. At the hearing, I found the argument that the First-tier Tribunal exercises the executive, rather than the judicial, power of the state difficult to follow (I still do, to be honest). In response to my questions, [NAME_51]. submitted that his argument is supported by the range of FOIA Scheduled bodies, all of which form part of the executive branch of the state.

GIA/973/2018 29

120. [NAME_49] for the [NAME_17] submits that [NAME_20] is clearly part of the judicial machinery of the state. It is not part of the executive machinery of the state. He is aware of no authority to the contrary.

121. I also had difficulty understanding why, as a matter of legal principle, an executive body would be required to make decisions according to current circumstances. I accept this is very often sensible but there is not, to my knowledge, any Act of Parliament that mandates it. [NAME_51]. submits it is a function of the modern state, a type of quid pro quo between citizen and state or modern social contract. In return for the modern state delving into so many aspects of our lives, we (citizens) expect the state to determine our entitlements according to current circumstances. No legislative provision nor case law authority was cited in support.

122. In relation to the argument concerning the merits-based function of [NAME_20], [NAME_56]. for the [NAME_17] answered by saying essentially ‘so what?’ There is no blanket rule of law that a merits-based adjudication must consider the merits at the date on which a court or tribunal, or some part of the executive branch, determines a case, rather than some earlier date.

Ground 2 – the First-tier Tribunal’s construction of a hypothesis

123. All parties agree that, in determining the NCND aspects of [NAME_21]’s appeal, [NAME_20]-was correct to construct a hypothesis, that is to assume that certain information was held, and then to apply the NCND public interest balancing exercise to that hypothetical information. In relation to part 2 of [NAME_21]’s request (correspondence with the Republic of Ecuador) the parties disagree on the question whether the hypothesis used by [NAME_20] – the range of hypothetical information identified - discloses an error on a point of law.

124. [NAME_20]’s statement of reasons, argues [NAME_51]. for [NAME_21], shows that, in constructing it hypothesis, the tribunal relied on a finding that the [NAME_17] had no proper role in corresponding with the Ecuadorian Embassy on behalf of the SPA. The Tribunal relied on this as a reason not to address “this hypothetical class of correspondence and therefore not to address one of the Appellant’s key submissions on the applicability of the NCND policy” (as it was put in written submissions). [NAME_20] erred by envisaging the existence of correspondence only where the [NAME_17] had a ‘proper role’. This cannot have been the correct way to identify hypothetical information for NCND purposes in this case since (i) at the date of the [NAME_17]’ refusal decision, [NAME_25] was living in the Ecuadorian Embassy; (ii) in 2013-15, the SPA was considering the potential for interviewing [NAME_25] in the Embassy; and (iii) during that period, the SPA was in correspondence with the

GIA/973/2018 30 [NAME_17], that correspondence including a request for “assistance on what measures exist to arrange for interview and the ‘English law and practice in such matters’”.

125. Given the circumstances just mentioned, [NAME_51]. argues that [NAME_20] should have considered the hypothetical existence of correspondence between the [NAME_17] and the Ecuadorian Embassy or authorities, despite the Republic of Ecuador not having made an extradition request. [NAME_20]’s ‘hypothetical’ was inherently unrealistic. It envisaged a state making an extradition request in respect of an individual whom, at that very time, it had given shelter in its Embassy following the grant of a type of asylum. The finding on which [NAME_20]’s hypothesis was based was unsafe and unsupported by evidence. [NAME_20] should not have avoided the exceptional aspects of this case by constructing an unreal hypothesis or, as [NAME_47] put it at the hearing, engaging in conjecture.

126. [NAME_56]. for the [NAME_17] argues that [NAME_51]’s criticisms of [NAME_20]’s hypothesis are flawed because they do not take account of the real world. If, in the real world, the only correspondence the [NAME_17] might receive from a foreign state would concern extradition, any request for information of the type made by [NAME_21] can only sensibly be read in one way – as a request for information relating to extradition. 127. In [NAME_51]’s submission, the correct approach to FOIA NCND provisions is to apply them solely by reference to the terms of the request for information. [NAME_20] erred by acceding to the [NAME_17]’s submission that it should hypothecate correspondence in which the Republic of Ecuador requested [NAME_25]’s extradition. This was no part of the request for information itself, which was not framed in terms of extradition.

128. The more specific any particular request for information, argues [NAME_51], the greater the likelihood that a public authority, in confirming whether it is held, will in substance reveal the information itself. [NAME_20]’s hypothesis therefore increased the likelihood of section 30(3) FOIA being ‘engaged’. [NAME_20] asked the wrong question which, in turn, yielded the wrong answer. [NAME_20] also overlooked the purpose of NCND provisions, which is to prevent requesters from circumventing exemptions by making requests that are so specific that confirming or denying whether the information is held would be tantamount to disclosure of the information. If [NAME_21] argues that the NCND provisions did not apply to her request in the light of its specificity, the argument is not sound, submits [NAME_49] for the Commissioner. Specificity does not displace the NCND provisions ([NAME_57] v [NAME_3] & Attorney General’s Office & Cabinet Office [2016] UKUT 535 (AAC)).

129. [NAME_51]. submits that [NAME_20] should have assumed that correspondence between the [NAME_17] and the Republic of Ecuador existed and then decided whether, on the evidence, it was more likely than not that the correspondence was about [NAME_25]’s extradition to Ecuador. [NAME_20]

GIA/973/2018 31 done so, it would have appreciated that there was no evidence to support a finding that, if correspondence existed, it was about extradition to the Republic of Ecuador.

130. [NAME_20]’s approach, argues [NAME_51], shows that it considered NCND provisions to be harm-based exemptions. They are not. In the section 30 case, subsections (1) to (3) are all pure class- based exemptions. The [NAME_17]’ arguments, accepted by [NAME_20], treated part 2 of [NAME_21]’s request as a request for correspondence between the [NAME_17] and any country about [NAME_25].

131. [NAME_49] submits that ground 2 is really a challenge to [NAME_20]’s finding of fact. The Upper Tribunal should not be ‘bewitched’ by [NAME_51]. into thinking anything else. [NAME_20] accepted the [NAME_17]’s evidence that the only correspondence it received from foreign states concerned actual or contemplated extradition, and made findings of fact accordingly. If [NAME_20]’s finding of fact stands, its hypothesis cannot be considered unrealistic. On the contrary, it was an entirely sensible hypothesis. [NAME_20]’s ‘technique’ was not flawed, as [NAME_47] argues. His argument, intentionally or otherwise, masks the key point. [NAME_20] found as fact that, if a foreign authority corresponded with the [NAME_17]’ extradition unit, the correspondence would concern extradition either actual or contemplated. [NAME_21] cannot dislodge that finding since it is free of any error on a point of law. It follows that the hypothesis deployed by [NAME_20] was not unrealistic and involved no error on a point of law.

132. [NAME_56]. for the [NAME_17] emphasises that the entire focus of ground 2 is a single sentence in paragraph 82 of the First-tier Tribunal’s statement of reasons:

“If we consider [the hypothesis that correspondence between the [NAME_17] and Ecuador concerning [NAME_25] might exist], then on the balance of probabilities the only thing that correspondence would be about, if it existed, would be an inquiry or request concerning extradition of [NAME_25] to Ecuador, or a follow-up to such a request”.

133. If the findings of fact in paragraph 82 of [NAME_20]’s reasons stand, ground 2 must fail, submits [NAME_50]. Unless the findings are perverse, the Upper Tribunal may not interfere. The Upper Tribunal should note that [NAME_20]’s findings of fact were made after hearing the oral evidence of a senior [NAME_17] lawyer, [NAME_32], given under cross-examination. [NAME_21] cannot establish the Tribunal’s findings of fact were perverse.

134. [NAME_49] also argues that NCND responses, where an information request has an extradition context, need to be applied consistently. If confirm or deny requests were answered in respect of some states, but not others, it could readily be inferred that the state/s in respect of which a NCND response were given had made an extradition request or made enquiries about extradition.

GIA/973/2018 32

Article 2 – the 1961 Convention

135. [NAME_49] for the Commissioner and [NAME_56]. for the [NAME_17] argue that [NAME_20]’s key finding of fact in paragraph 82 of its statement of reasons was consistent with Article 41(2) of the 1961 Convention. Any diplomatic correspondence would not be with the [NAME_17] but a Ministry. [NAME_35] argues that, while Article 41(2) of the Vienna Convention was not in [NAME_20]’s contemplation, it provides further weight to his argument that its findings of fact were not perverse. Had the Ecuadorian Embassy wished to make diplomatic representations to the UK, in contrast to a judicial extradition request, for example to urge the UK to treat [NAME_25] leniently, these would, under Article 41(2), go to the Foreign Office or some other Ministry.

136. [NAME_56]. argues that [NAME_20] correctly declined, as part of its NCND analysis, to consider the likelihood of a particular state having made an extradition request. If NCND responses were only given where it was likely a request had been made, a fugitive could infer that his extradition was likely to be sought, which is nearly as bad as a direct top-off. [NAME_35] also questions the point of ground 2. Even on [NAME_21]’s case, the Republic of Ecuador would not have made an extradition request or enquiry in respect of [NAME_25]. It follows that [NAME_21] must have some other correspondence in mind but, if such correspondence exists, it would not have been with the [NAME_17].

137. [NAME_58] for the [NAME_18] submits that the correspondence sought by [NAME_21] could not have been of a type that attracts inviolability under the 1961 Convention. Correspondence of a mission, held by a receiving state, is not inviolable as a general rule. The exceptions to this general rule are limited and none could have applied to the (hypothetical) correspondence under analysis in this case (R (Bancoult No 3) v Secretary of State for Foreign and Commonwealth Affairs [2018] UKSC 3). Articles 24 and 27 of the 1961 Convention could not have been relevant to the issues that [NAME_20] had to decide. [NAME_20] could not have made a material error on a point of law by failing to consider either Article.

138. Alternatively, argues [NAME_58], had the (hypothetical) correspondence been of a type attracting inviolability under the 1961 Convention, the absolute prohibition in section 2(1) of the 1964 Act would have come into play. In such circumstances, arguably section 44 FOIA (prohibition on disclosure provided for by some other enactment) would have rendered the information absolutely exempt from disclosure. If [NAME_20] erred, it was not an error in [NAME_21]’s favour.

139. [NAME_58] points out that Article 41 of the 1961 Convention has not been incorporated into the law of England and Wales ([NAME_59] v Department of Trade and Industry [1990] 2 AC 418). In

GIA/973/2018 33 R v Lyons [2003] 1 AC 976, at 27, Lord Hoffman held that the courts of England and Wales “have no jurisdiction to interpret or apply” unincorporated international treaties. It would not therefore have been open to [NAME_20], submits [NAME_58], to seek to give effect to Article 41(2) in the construction of its NCND hypothesis. [NAME_20] would, however, have been entitled to take Article 41 into account as a matter of evidence ([NAME_59] at 501A). Had it done so, it could only have provided further support for [NAME_20]’s findings of fact, which were consistent with the practice anticipated by Article 41.

Ground 3

140. [NAME_51]. for [NAME_21] argues that, in considering the NCND aspect of her appeal in relation to parts 3 and 4 of her request for information (correspondence with US Departments), [NAME_20] failed to take into account its earlier recognition that various factors supported the public interest in disclosure. These considerations were not only relevant in relation to part 1 of [NAME_21]’s request for information (see paragraph 17 above for the factors identified by [NAME_20] as supporting the public interest in disclosing the information within part 1 of [NAME_21]’s request). The UN body’s determination that [NAME_25] was subject to arbitrary detention was based on its acceptance that he feared persecution connected to the possibility of his extradition to the USA. [NAME_20] should have taken this into account. The substantial general public interest in relation to press freedom should also have been factored into the analysis, taking into account the US Government’s well- known antipathy towards [NAME_25].

141. [NAME_20], submits [NAME_51], relied on a finding that it would not be of more than marginal public benefit for the [NAME_17] to confirm or deny whether the information sought was held. Other than that, the only consideration taken into account was the public interest in not seeing a personal interest (i.e. [NAME_25]’s) defeated. [NAME_20] left out of account factors in favour of disclosure which it had itself identified in relation to part 1 of [NAME_21]’s request for information. [NAME_47] does not dispute the Tribunal’s weighing of competing considerations. Rather, his argument is that a host of relevant factors were simply not put in the balance.

142. [NAME_51]. submits that [NAME_20] drew a false ‘dichotomy’ between [NAME_25]’s personal interest and the benefit to the public of the [NAME_17] confirming or denying whether the information was held. What was left out of the balance was the public interest in [NAME_25] seeing information relating to himself. The public interest in public understanding of [NAME_25]’s case was rightly taken into account but the overlooked ‘facet’ of the public interest was in an individual not being deprived of information relating to himself. Having been asked to elaborate on this point by myself at the hearing, [NAME_47] submits that, while [NAME_20] acknowledged [NAME_25]’s strong personal interest, it left out of account his frustration at information not being disclosed. I suppose that (the

GIA/973/2018 34 frustration) may be taken as read but, on my reading of the First-tier Tribunal’s bundle, it was not supported in evidence.

143. The public interest in any matter is single and indivisible, argues [NAME_51]. That information relates to a single individual does not diminish the public interest in its disclosure. An acute public interest exists in not depriving an individual of access to information relating to himself. That public interest found expression in the very earliest legislation about access to information such as the Data Protection Act 1984 and the Access to Medical Records Act 1990. The public interest in an individual seeing particular information about himself is not graduated according to his popularity or the level of curiosity in his predicament.

144. Ground 3 is without merit, argues [NAME_49] for the Commissioner. Paragraphs 90 to 91 of [NAME_20]’s statement of reasons show that it did recognise a public interest in either confirming or denying that the [NAME_17] held the information albeit only “marginal” weight was given to those interests. Only a misreading of [NAME_20]’s decision could lead to the conclusion that [NAME_20] treated the public interest as no more than that of the public in not seeing a personal interest defeated. On a fair reading of [NAME_20]’s statement of reasons, it acknowledged [NAME_25]’s personal interest but found that confirming whether the information was held would be of only marginal benefit to the public. The statement of reasons does not suggest that the latter was a function of the former. [NAME_56]. for the [NAME_17] agrees. [NAME_20] accepted a public (and private) interest in revealing whether the US Departments had corresponded with the [NAME_17] about [NAME_25] but gave this consideration little weight because it would provide only a “modest” increase in public understanding of [NAME_25]’s case. It is for [NAME_20] to determine the weight given to any particular consideration. Absent perversity, the Upper Tribunal may not interfere. The First-tier Tribunal’s reasoning was not perverse.

145. In substance, ground 3 is a perversity challenge argue both [NAME_49] and [NAME_56]. Such challenges face very high hurdles (DWP v Information Commissioner and Zola [2016] EWCA Civ 758). The Upper Tribunal should generally adopt a light touch approach in scrutinising [NAME_20]’s weighing of competing public interest considerations. [NAME_51]. disputes that ground 3 amounts to a perversity challenge. He repeats that the essence of his argument is that [NAME_20] left out of account relevant matters, including matters it had itself identified in an earlier part of its reasoning.

146. Regarding the argument that the matters identified by [NAME_20] as supporting disclosure of the information within part 1 of [NAME_21]’s request for information were not taken forward and factored into [NAME_20]’s NCND public interest balancing exercise, [NAME_56]. warns against a counsel of perfection. The reference in paragraph 91of [NAME_20]’s statement of reasons to the case

GIA/973/2018 35 being ‘unusual’ indicates, on a sensible reading, that [NAME_20] had these matters in mind in determining the NCND aspect of [NAME_21]’s appeal.

147. If ground 3 argues that the Tribunal gave inadequate reasons for its decision, [NAME_49] submits it cannot succeed. [NAME_20] “has done enough to show that it has applied the correct legal test and in broad terms explained its decision” (UCAS v Information & Lord Lucas [2014] UKUT 557 (AAC). [NAME_49] reminded me of the often-given warning against taking too ‘nit-picking’ an approach to [NAME_19] reasons. On a fair reading, there was no error of law in the form of inadequate reasons.

148. [NAME_49] argues that, in substance, [NAME_51]’s submissions come dangerously close to an argument that the public was entitled to be ‘tipped off’ about a proposed extradition. There is no public interest in anyone being tipped-off about a proposed extradition. [NAME_56]. goes further and submits that, in relation to parts 3 and 4 of [NAME_21]’s request, [NAME_20] reached the only conclusion reasonably open to it. It is in the public interest that [NAME_25] should not be tipped off about enquiries the USA might make about his extradition. In substance, that is the supposed public interest factor that, on [NAME_21]’s case, [NAME_20] failed to take into account. If [NAME_20] treated this as a personal interest, not carrying any weight as a public interest factor, it was fully entitled to do so.

149. [NAME_49] argues that [NAME_51]’s submissions lost sight of the fact that [NAME_21] made the request for information. The request was not made by, or on behalf of, [NAME_25]. Separate legal mechanisms outside FOIA govern individuals’ right of access to information about themselves.

Conclusions

Ground 1 – should this ground be determined?

150. There is some force in the Respondents’ argument that [NAME_21] should not be permitted to rely on a ground that turns on a point of law that could have been, but was not, taken before [NAME_20].

151. [NAME_21]’s notice of appeal to [NAME_20] against the Commissioner’s decision notice made no submissions on the public interest timing point, nor did her written reply to the Respondents’ responses to her appeal. The notice of appeal and reply were drafted by counsel. [NAME_20] skeleton argument drafted by counsel for the [NAME_17] clearly stated, at p. 16, that the live issues on the appeal involved considering the balance of public interests in late 2015, that is at the time when the [NAME_17] refused to comply with [NAME_21]’s request for information. It seems that [NAME_19] case management

GIA/973/2018 36 directions did not provide for simultaneous exchange of skeleton arguments because [NAME_21]’s skeleton argument made reference to the [NAME_17] skeleton argument.

152. Having now considered [NAME_20] papers in depth, in my view the public interest timing point was undisputed before [NAME_20]. There was no dispute that the public interest was to be assessed according to circumstances at the date of the [NAME_17]’s refusal decision. [NAME_21] knew the [NAME_17]’s case was that the public interest balancing exercise should be conducted according to circumstances at the date of the [NAME_17]’s refusal decision. Despite that, neither [NAME_21]’s skeleton argument nor her oral submissions to [NAME_20] disputed the [NAME_17] contention that the public interest was to be assessed at the date of their refusal decision.

153. It was not readily apparent from [NAME_21]’s application to the Upper Tribunal for permission to appeal against [NAME_20]’s decision that, before [NAME_20], she did not dispute the [NAME_17]’ case on the public interest timing point. The application states that “the issue of the timing of the public interest was not canvassed at the [[NAME_19]] hearing”.

154. Perhaps I am at fault for not reading the statement ‘not canvassed at the [First-tier Tribunal] hearing’ to mean ‘not argued because the public interest timing point was undisputed’. Be that as it may, the application should have made it clear that permission to appeal was being sought on a point of law that was not disputed before [NAME_20]. Had it done so. I might not have granted permission to appeal on ground (1).

155. Despite having some reservations, as just explained, I decline to re-visit my grant of permission to appeal. I now have the benefit of detailed arguments from counsel all of whom have significant experience of information rights cases. There is also in my view some merit in the Upper Tribunal re-examining the public interest timing point under FOIA if only to put the matter finally to rest.

Ground 1 – construing FOIA free of authority

156. If the public interest timing point were free of authority, I suspect that I would construe FOIA in the manner contended for by the Respondents, by applying conventional principles of statutory interpretation.

GIA/973/2018 37 157. A [NAME_36] of information is entitled, under section 1(1)(b) FOIA, to have the requested information communicated to her. The general rule in section 1(4) is that the information to be communicated is that which is held at the date of the request although “account may be taken of any amendment or deletion made” after the time of the request.

158. Section 1(1)(b) does not apply to any exempt information which is also absolutely exempt. If an absolute exemption is not applicable, section 1(1)(b) does not apply where “the public interest in maintaining the exemption outweighs the public interest in disclosing the information” (section 2(2)). Since the process has only just started, at this stage the public authority is bound to address the public interest according to current circumstances.

159. Section 14 FOIA is concerned with vexatious or repeated requests. These are addressed not by enacting a category of exempt information but, in the case of vexatious requests, simply disapplying section 1(1) FOIA. In the case of repeated requests, as defined, section 14 provides that a public authority is not obliged to comply with the instant request. In both cases, there is no associated public interest balancing exercise.

160. Section 30 FOIA provides for a category of exempt information relating to investigations and proceedings conducted by public authorities. To fall within this category, information must be “held” by a public authority and, in addition, be information which:

161. The question whether information is section 30 exempt information turns on questions of primary fact. Is the information held by the public authority? Has the information at any time been held for a purpose specified in subsection (1)? And so on. To use [NAME_51]’s terminology, section 30 does not enact a harm-based category of exempt information.

162. I now come to [NAME_2]’s enforcement functions. Section 50(1) FOIA permits a complainant to apply to the Commissioner for a “decision whether, in any specified respect, a request for information made by the complainant to a public authority has been dealt with in accordance with the requirements of Part I” (my emphasis).

GIA/973/2018 38 163. In section 50(1) FOIA, Parliament uses the present perfect simple tense to set out the nature of the Commissioner’s task (see the emphasised words above). The present perfect simple tense denotes an activity that has been completed. When used in conjunction with an actor – in this case, a public authority - it clearly informs the reader by whom the activity was completed. Ordinary linguistic canons of construction therefore indicate that Parliament’s intention was for the Commissioner to inquire into the way in which a public authority completed the activity of responding to a request for information made under FOIA. And the public authority, when it completed that activity in a qualified exemption case, will have addressed the public interest according to current circumstances.

164. Section 50(4) FOIA sets out when a decision notice is required to specify steps to be taken by a public authority to comply with a FOIA requirement. Steps must be specified if a public authority “has failed to communicate information…in a case where it is required to do so by section 1(1)”. Section 50(4) is silent concerning the type of steps to be specified. In my view, this provision, consistently with section 50(1), focusses on what the public authority actually did (‘has failed’). Section 52 also uses the term ‘has failed’ in conferring power on the Commissioner to serve an enforcement notice on a public authority.

165. Once a decision notice is served, the complainant or public authority may appeal to [NAME_20] against “the notice” (section 57(1)). Section 58 FOIA, which describes when [NAME_20] must allow an appeal, also refers to “the notice” (i.e. “the notice against which the appeal is brought is not in accordance with the law”). Section 58 further provides for [NAME_20] to “substitute such other notice as could have been served by the Commissioner”. If the Commissioner addresses the public interest at the date of the public authority’s refusal decision, then the range of notices that the Commissioner could have given is limited by those properly open to the Commissioner in the light of the circumstances at the date of the public authority’s refusal decision. On this analysis, the public interest timing question, in relation to [NAME_20], is answered by identifying the relevant date for the purposes of the Commissioner’s functions.

166. I suppose it might be argued that, in relation to the predictive (or harm-based) set of exempt information categories (see paragraph 57 above), the requirement to consider the consequences of “disclosure under this Act” provides some support for the ‘moving target’ approach to the public interest since disclosure would only occur as and when a decision maker so decides (Parliament could not possibly have intended for the public interest and questions of fact relevant to whether information is exempt to to be determined according to different temporal reference points). However, section 30 exempt information is identified by answering questions of primary fact. Section 30 does not require the consequences of disclosure “under this Act” to be considered.

GIA/973/2018 39 167. I should point out that the comments in the previous paragraph should not raise anyone’s hopes for a re-visitation of the public interest timing point in the Upper Tribunal. The Upper Tribunal in [NAME_15], which was concerned with a predictive (or harm-based) category of exempt information under section 27(1) FOIA, clearly did not consider the requirement to address ‘disclosure under this Act’ as establishing a public interest ‘moving target’.

168. To conclude this section, if the matter were free of authority I would be likely to hold, on ordinary principles of statutory construction / accepted linguistic canons of construction, that, in a section 30 case, FOIA’s intention was for the public interest to be assessed throughout the challenge and appeal process according to the circumstances at the date of the public authority’s refusal decision.

Ground 1 – am I expected or required to follow [NAME_15]

169. A central issue on this appeal is whether I should follow the decision on of the three-judge panel of the Upper Tribunal in [NAME_15] in relation to the public interest timing point.

170 The Dorset Healthcare ‘guidelines. concerning the precedential status of Administrative Appeals Chamber three-judge panel decisions cannot properly be described as ‘tentative guidance’. The Upper Tribunal said: “37…those guidelines are as follows: (iii) In so far as the AAC [Administrative Appeals Chamber] is concerned, on questions of legal principle, a single judge shall follow a decision of a Three-Judge Panel of the AAC or Tribunal of Commissioners unless there are compelling reasons why he should not, as, for instance, a decision of a superior court affecting the legal principles involved”.

171. Dorset Healthcare does not deal with the topic of three-judge panel decisions and their precedential status tentatively. The Upper Tribunal was not carefully probing the boundaries of some new principle. It was in fact carrying on, with appropriate adaptations, a practice followed by many years by the Social Security Commissioners (who became judges of the Upper Tribunal, assigned to the AAC, as part of the Tribunals, Courts and Enforcement Act 2007 reforms).

172. The Upper Tribunal may set aside a decision of [NAME_20]-tier if it “finds that the making of the decision concerned involved the making of an error on a point of law” (section 12(1), (2)(a), Tribunals, Courts and Enforcement Act 2007). If the Upper Tribunal sets aside [NAME_20]’s decision, section 12(2)(b) requires the Upper Tribunal either to remit the case to [NAME_20] or re-make [NAME_20]’s decision. In [NAME_15], [NAME_20]’s decision was not set aside because it involved an error on a point of law relating to the public interest timing point. The public interest timing point was considered only

GIA/973/2018 40 after [NAME_20]’s decision had been set aside, as part of the Upper Tribunal re-making [NAME_20]’s decision.

173. The submissions on this appeal did not include developed arguments about the implications, if any, of the fact that, in [NAME_15], the Upper Tribunal did not consider the public interest timing point in exercising its appellate function under section 12(1) of the 2007 Act. I was, however, asked to take this feature of [NAME_15] into account. Perhaps no party argued that the [NAME_15] findings on the public interest timing point might fall outside guideline (iii) in paragraph 37 of Dorset Healthcare because such an argument was considered to have little merit. If so, I would agree.

174. In [COMPANY_61] v [NAME_63] and others [NAME_64] and [COMPANY_65] v [ADDRESS] [2007] All ER (D) 408 (May), the Court of Appeal rejected the argument that a county court judge sitting at first instance was not bound by a decision of a High Court judge given in the exercise of a first instance jurisdiction. The ratio of such a High Court decision had the same precedential status as the ratio of a decision given by the Court in the exercise of some appellate jurisdiction. The hierarchical relationship between the Upper Tribunal and [NAME_20] is akin to that between the High Court and the county court. [NAME_20] must therefore be bound by any decision of the Upper Tribunal whether given under section 12(1) TCEA 2007 in determining an appeal or, subsequently under section 12(2)(b), in the course of re-making a decision of [NAME_20]. The Upper Tribunal in Dorset Healthcare NHS Trust did not put different types of Upper Tribunal finding into separate categories. It must have intended to refer to all findings that bind [NAME_20] on questions of legal principle. It follows that guideline (iii) in paragraph 37 of Dorset Healthcare NHS Trust applies to questions of legal principle determined by a three-judge panel whether the principle was determined in setting aside a decision of [NAME_20] or subsequently in re-making a decision of [NAME_20].

175. The ratio of [NAME_15] concerning the public interest timing point falls within guideline (iii) in Dorset Healthcare NHS Trust. This is a black and white matter. Guideline (iii) cannot accommodate any ranking of three-judge panel decisions. There is nothing relevant to ‘take into account’ in determining whether the ratio of [NAME_15] falls within guideline (iii).

176. The remaining questions are (a) whether [NAME_15]’s findings on the public interest timing point were part of the ratio for the decision; and (b) if they were, whether there is nevertheless some compelling reason not to following them.

177. I agree with [NAME_49] that the Upper Tribunal’s findings on the public interest timing point in [NAME_15] were part of the ratio of its decision and, if it makes any difference, were also the Upper Tribunal’s answers to questions of legal principle. The Upper Tribunal in [NAME_15] set out its

GIA/973/2018 41 findings of law on the public interest timing point under the heading ‘Issue 3’. The subsequent ‘Issue (4)’ was “the proper application of the public interest balancing test”. Within the issue (4), there is the sub-heading ‘The public interest balancing exercise in respect of Document 59’, immediately followed by paragraph 98’s statement that “This is directed to the position in June 2009”. June 2009 was the public authority refusal date in that case. I am satisfied that the Upper Tribunal’s findings in [NAME_15] on the public interest timing point were part of the ratio for its decision and, if it makes any difference, were also the Tribunal’s answers to questions of legal principle. They were not obiter comments.

178. Is there a compelling reason not to follow [NAME_15]? [NAME_21] submits that [NAME_15], misreads Evans and that supplies the necessary compelling reason not to follow [NAME_15]. The high- water mark of this submission relates to the following passage from [NAME_44]’s opinion in Evans:

“73…although the question whether to uphold or overturn…a refusal by a public authority must be determined as at the date of the original refusal, facts and matters and even grounds of exemption may, subject to the control of the Commissioner or the tribunal, be admissible even though they were not in the mind of the individual responsible for the refusal or communicated at the time of the refusal to disclose (i) if they existed at the date of the refusal, or (ii) if they did not exist at that date, but only in so far as they throw light on the grounds now given for refusal.”

179. In [NAME_21]’s application to the Upper Tribunal for permission to appeal against [NAME_20]’s decision, she argues that [NAME_15] failed to appreciate that, in paragraph 73, [NAME_44] envisaged newly applicable grounds of exemption being relied on before [NAME_20]. If a newly applicable ground of exemption may be relied on, it follows that the public interest balancing exercise in relation to that ground is undertaken according to circumstances post-dating the refusal decision. This opens the door to the ‘moving target’ approach to assessing the public interest.

180. If the relevant passage from [NAME_44]’s opinion is cut down, it is true that one ends up with: ““although the question whether to uphold or overturn…a refusal by a public authority must be determined as at the date of the original request…even grounds of exemption may…be admissible even though they were not in the mind of the individual responsible for the refusal or communicated at the time of the refusal to disclose…(ii) if they did not exist at that date…”.

GIA/973/2018 42 181. [NAME_49] for the Commissioner concedes a degree of textual ambiguity within paragraph 73. I would, with the greatest of respect, agree. However, only an ultra- literal and selective reading of [NAME_44]’s words would lead to the conclusion that his Lordship envisaged newly applicable grounds of exemption being relied on before the Commissioner or [NAME_19]. That cannot be a correct reading. If one returns to [NAME_44]’s entire wording, it is clear that:

(a) [NAME_44] is discussing three features of FOIA cases: ‘facts’, ‘matters’ and ‘grounds of exemption’;

(b) clause (i), towards the end of paragraph 73, is intended to operate on all three matters. In other words, new facts, matters and grounds of exemption may in principle be relied on if they existed at the date of the refusal decision; and

(c) clause (ii) is not intended to operate on grounds of exemption, only on ‘facts’ and ‘matters’. While, on the face of it, clause (ii) begins with a reference back to all three things (see the word ‘they’), it concludes with “but only in so far as they throw light on the grounds now given for refusal”. Only ‘facts’ and ‘matters’ are capable of throwing light on a ground for exemption. A ground of exemption does not throw light on itself.

182. [NAME_66]’s argument is also inconsistent with other parts of [NAME_44]’s opinion. I do not need to explain why here because the issue is fully dealt with by sensibly construing paragraph 73 itself. I am satisfied that any reasonable reader of [NAME_44]’s full opinion will appreciate that he did not envisage a system in which a ground of exemption could be relied on before [NAME_20], or Commissioner, even though the ground was simply inapplicable at the date of the original refusal decision. The Upper Tribunal in [NAME_15] did not misread Evans.

183. I also reject the argument that, in [NAME_15], the Upper Tribunal should not have relied on Evans, or should have accorded it less weight, because it involved a challenge to a certificate given by an “accountable person” under section 53(1) FOIA. A section 53(1) certificate relates to a decision notice. It must do since the operative provisions in section 53(2) provides for the decision notice to which the certificate relates to cease to have effect. The certificate in Evans was challenged by way of a claim for judicial review in the High Court. I do not see that the relatively rare legal route by which a FOIA issue came to be adjudicated upon made any difference to [NAME_44]’s construction of Part I of FOIA and the exemption provisions of Part II. As I have said, in a certificate case, a FOIA request must at least have got to the decision notice stage. On my understanding of the submissions on this appeal, the decision notice stage is the key stage so far as the public interest timing point is concerned. Whatever the correct reference date for the Commissioner may be, it determines the correct date for [NAME_20]. If the Commissioner looks to circumstances at the date of a

GIA/973/2018 43 refusal decision, so must [NAME_20]. That is why [NAME_44]’s views are as relevant to proceedings before [NAME_20] as to certificate cases even though Evans itself was a challenge to a section 53(1) certificate.

184. To conclude this section, I decide that the Upper Tribunal’s findings on the public interest timing point in [NAME_15] were part of the ratio of the three-judge panel’s decision and, for the purposes of guideline (iii) in paragraph 37 of Dorset Healthcare NHS Trust, constituted answers to questions of legal principle. I should therefore follow [NAME_15] unless there is a compelling reason not to do so. There is no such compelling reason. I follow [NAME_15] and so ground 1 must fail.

185. Even if I were not expected to follow [NAME_15], and Evans had never been given, I would still, in all likelihood, have rejected [NAME_21]’s case on ground 1 (see the above ‘free of authority’ analysis) . I doubt the arguments for [NAME_21] would have persuaded me to depart from the construction of FOIA contended for by the Respondents:

[redacted] taken to be a legal term of art. Even if it is, in the absence of a statutory definition the term must still construed in its own legislative context. I find it difficult to see how the asserted generally accepted legal meaning of “review any finding of fact” could overcome other statutory indications in FOIA as to the correct date by reference to which the public interest is to be assessed;

(b) the enabling power for the First-tier Tribunal’s procedural rules is of no assistance in construing those provisions of FOIA that bear on the public interest timing point;

(c) section 14(1) of FOIA (vexatious requests) is not an exemption from the duty to apply with section 1(1) FOIA. It disapplies the duty completely and there is no associated public interest balancing test. Section 14(1) does not assist;

(d) the correct adjudicative reference point in asylum or other immigration-related cases cannot be read across to FOIA. These are very different systems of law;

(e) the approach taken by the Administrative Appeals Tribunal of Australia cannot influence the construction of FOIA even if it were a source of inspiration for the Leggatt report;

(f) section 12 of the Interpretation Act 1978 is neither here nor there. The power reposed in the Commissioner is not the same power as is reposed in the First-tier Tribunal. And the fact that, under section 12, any particular power may be exercised from time to time as occasion requires says nothing about the nature of a power in other respects.

GIA/973/2018 44

Ground 1 – remaining matters

186. It is not strictly necessary to deal with the remaining ground 1 arguments so I shall keep my comments brief.

187. [NAME_51]. for [NAME_21] advances ambitious arguments about the nature of [NAME_20] and how this connects to the citizen-state relationship. In my opinion, [NAME_20] cannot properly be characterised as part of the executive branch of the state. The fact that, in its information rights jurisdiction, [NAME_20] will always adjudicate upon a decision taken by a branch of the executive does not make it part of that branch. All generally accepted characteristics of a judicial body are seen in the First-tier Tribunal. I need not enumerate them. Moreover, the Tribunals, Courts and Enforcement Act 2007 amended the Constitutional Reform Act 2005 so that judges and members of the First-tier Tribunal are part of the “judiciary” for the purposes of section 4 of that Act (as was recognised in paragraph 85 of [NAME_44]’s opinion in Evans).

188. The first plank of this argument not being sound, the second fails too. But, even if I had agreed that [NAME_20] is part of the executive branch, it would not necessarily follow that it would be bound to adjudicate according to current circumstances. At the hearing, [NAME_49] described this part of [NAME_51]’s argument as his ‘Reith Lecture’. A cutting observation, perhaps, but an apposite one drawing attention, as it does, to the absence of discernible legal moorings. The arguments were about what the law should be rather than what the law is.

189. The practical difficulties arising from whichever date is used as the public interest assessment reference point were dealt with in some detail in [NAME_15]. There is no need for me to add anything. Nor is there any need for me to say anything about the submissions that ground 1 is academic.

Ground 2

190. In my judgment, [NAME_20] was entitled to accept the senior [NAME_17] lawyer’s ([NAME_32]) oral evidence and make findings of fact accordingly. In particular, [NAME_20] was entitled to find that, if any correspondence with Ecuadorian authorities were held by the [NAME_17], it would only be about extradition. At the hearing, this finding was not of itself seriously challenged on [NAME_21]’s behalf.

191. The remaining issue is whether [NAME_20]’s use of its finding of fact concerning the extradition subject matter of any correspondence that might be held led it to stray from the terms of section 30(3) FOIA. In particular, did [NAME_20]’s use of that finding in constructing the hypothetical

GIA/973/2018 45 information, by reference to which the public interest balancing exercise under section 2(1) would be undertaken, lead it to determine this aspect of [NAME_21]’s appeal in a manner contrary to the terms of section 30(3) FOIA?

192. I think it assists to begin with the FOIA entitlement at stake. Under section 1(1)(a) [NAME_21] had a right, in principle, to be informed by the by the [NAME_17] whether it held all the information described in her request – the whole lot of it.

193. [NAME_21]’s right under section 1(1)(a) was jeopardised by two provisions, section 30(3) and section 2(1).

194. Section 2(1) provides:

“(1) Where any provision of Part II states that the duty to confirm or deny does not arise in relation to any information, the effect of the provision is that where either— (a) the provision confers absolute exemption, or (b) in all the circumstances of the case, the public interest in maintaining the exclusion of the duty to confirm or deny outweighs the public interest in disclosing whether the public authority holds the information, section 1(1)(a) does not apply.” 195. Section 30(3) FOIA provides, in a not particularly reader-friendly manner, that the duty to confirm or deny does not arise in relation to information which is exempt under section 30(1) and (2), or in relation to information which, if it were held, would be so exempt. I say it is not particularly reader-friendly because the overall position under FOIA is that, even if the duty to confirm or deny does not ‘arise’, it will nevertheless apply if the public interest balancing exercise under section 2(1) so dictates. 196. Section 2(1) FOIA enacts that it gives effect to section 30(3). But this does not mean section 30(3), standing alone, is an ineffective provision. Section 30(3) is the gateway to the public interest balancing exercise under section 2(1). To the extent that the information requested does not pass through the gateway, the applicant’s right to be told whether information is held is preserved. In other words, the right to be told is preserved in relation to so much of the requested information as does not match the descriptions in section 30(1) and (2). It must follow that the section 2(1) public interest balancing exercise can only ever involve so much of the information requested as amounts to (or would if held amount to) exempt information within section 30(1) or (2). To this extent, at least,

GIA/973/2018 46 that FOIA does not require a NCND analysis to be conducted by reference to the entirety of the requested information.

197. [NAME_20]’s finding of fact that, if there were correspondence, it would concern extradition, I cannot but avoid the conclusion that [NAME_20] was also satisfied that, to the extent that [NAME_21]’s request went beyond extradition, the information was not held by the [NAME_17]. It should not come as a surprise that the [NAME_17] does not hold information that is irrelevant to functions. Although expressed differently, this analysis really makes the same point as did [NAME_50] in arguing that a NCND hypothesis needs to be linked to the real world.

198. [NAME_20] did not err in law by applying the public interest balancing exercise to a reduced range of hypothetical information, as compared with [NAME_21]’s request for information, comprised of extradition-related information. If information is confirmed not to be held by a public authority, it cannot pass through the section 30(3) gateway and has no role to play in the section 2(1) public interest balancing exercise. That would clearly be pointless because the relevant entitlement under section 1(1)(a) has already been satisfied.

199. For the above reasons, ground 2 fails. I do not accept [NAME_51]’s argument that NCND provisions must in all cases operate by reference to the terms of the request for information. [NAME_20]’s approach discloses no legal mistake (or mistake of ‘technique’).

200. [NAME_20] rightly held that section 30(3) FOIA does not inevitably operate by reference to the entirety of the requested information. In limiting the scope of the hypothetically held information as it did, [NAME_20] was simply following through with the natural consequence of its finding that, if any information were held, it would relate to extradition. I find it very difficult to see how [NAME_20] could have taken a different approach in the light of that finding. If, despite having found that the only Ecuadorian information that the [NAME_17] might hold would concern extradition, [NAME_20] then subjected the entirety of [NAME_21]’s request to the public interest balancing exercise, it would have only been storing up difficulties. How could it have determined whether the public interest in maintaining the exclusion of the duty to confirm or deny outweighed the public interest in disclosing whether the information was held when, on its findings, much of the information requested would not have been held? But [NAME_20] did not put itself in that position, no doubt because it appreciated that it would have set up an impossible task for itself.

Ground 3

201. [NAME_21] argues that an element of the public interest relating to [NAME_25], going beyond a personal interest, was left out of account. At the hearing, [NAME_51]. identified this as the public

GIA/973/2018 47 interest in an individual not being deprived of information relating to himself. At times, the arguments seemed to treat [NAME_25] and [NAME_21] as one and the same. But the fact is the request for information was not made by [NAME_25], it was made by [NAME_21] on her own account. [NAME_47] further argues that this facet of the public interest finds expression in various pieces of legislation conferring rights on individuals to obtain information about themselves and a public interest analysis under FOIA needs to acknowledge this. In my judgment, these arguments do not work.

202. I am sure that there is, in a general sense, a public interest in not depriving an individual of access to information relating to himself or, at least, promoting such access. But that does not require the interest is to be recognised in a FOIA NCND public interest balancing exercise. Separate legislation confers rights on individuals to access information about themselves. Broadly speaking, FOIA cannot be used by an individual as a means of obtaining personal information about himself; the individual generally needs to rely on his rights under the Data Protection Act 2018 (see section 40(1), FOIA).

203. In my judgment, the public interest in an individual seeing information about himself is served by the existence of separate legislation that confers specific rights for that purpose. It is not a factor requiring recognition in a NCND public interest balancing exercise which will connect to a request made by someone other than the subject of the information. I fail to see how a [NAME_36]’s use of FOIA to obtain personal information for and about another person could rank as a relevant public interest factor favouring disclosure or, even if it did, how it could attract more than a minimal weighting. Refusing under FOIA to communicate information to an individual so that s/he can then communicate the information to the subject of the information is not, to my mind, a matter of public interest or concern. This is for the simple reason that what the public interest or concern is really directed towards is the information subject’s rights of access and these are provided for in separate legislation.

204. [NAME_51]’s argument must be based on an assumption that, if information were communicated to [NAME_21], she would in turn communicate it to [NAME_25] or at least disseminate it in such a way that it came to his attention. In the absence of such an arrangement, presumably there would be consideration of this type to be placed in the [NAME_36]’s side of the public interest scales. That does not seem to me to be a principled basis on which to differentiate between FOIA requests.

205. Furthermore, this aspect of the case could not have resulted in [NAME_25] seeing any information held by the [NAME_17] about himself. At best, it would have resulted in him knowing whether certain information was held, not its content or, at least, nothing specific about its content beyond

GIA/973/2018 48 that which might be inferred. Even if there was a public interest consideration referable to [NAME_25] it was not the consideration identified by [NAME_51].

206. For the above reasons, if [NAME_20] failed to recognise a public interest in [NAME_25] seeing information about himself, it did not err on a point of law.

207. The other strand of [NAME_21]’s case on ground 3 argues that factors in favour of disclosure identified by [NAME_20] itself, in determining her appeal in relation to part 1 of her request, were overlooked when [NAME_20] came to the NCND part of her appeal.

208. It is worth remembering that, had this aspect of [NAME_21]’s appeal succeeded, she would not have obtained any information. All she would have obtained was confirmation that certain information was or was not held by the [NAME_17].

209. In [NAME_20]’s determination, a ‘slight public benefit’ might have accrued from requiring the [NAME_17] to depart from its usual NCND policy. This would have taken the form of a “modest increase in public understanding of an unusual case”. The considerations favouring disclosure of the information sought by part 1 of [NAME_21]’s request, on the other hand, were set out in paragraph 57 of [NAME_20]’s statement of reasons:

(1) there were onsiderations which, in every FOIA case, favour disclosure, described by [NAME_20] as: “disclosure of official information can promote good government through transparency, accountability, increased public confidence and public understanding, the effective exercise of democratic rights, and other related public goods, including fostering constructive public debate”;

(2) there was also the public interest in information being made available that might increase public understanding of extradition;

(3) the fact that “the matter has dragged on unresolved for a long time”, involving a cost to the public purse, also supported the public interest in disclosure. How this happened and whether public funds were well spent was a matter of legitimate public concern;

(4) [NAME_25] is the only free speech advocate in the Western world whose situation has been described by a UN body as arbitrary detention and “it is a matter of public controversy how this situation should be understood”;

GIA/973/2018 49 (5) the case raised issues about human rights and Press freedoms, which are the subject of legitimate public debate and thus supported the public interest in disclosure. Further, such debate might help to resolve the issues.

210. In my judgment, these considerations could not have operated with the same force within the NCND public interest balancing exercise. [NAME_21] succeeded, public knowledge about [NAME_25]’s case would only have increased through making it known that the US Departments either had or had not corresponded with the [NAME_17] about [NAME_25]. I warn myself that my function is not to re-determine the merits but I have doubts whether [NAME_20] could properly have found that the increased public knowledge flowing from disclosing whether the information were held could have said anything relevant to public understanding of extradition or whether public funds had been well- spent.

211. The arguments advanced on [NAME_21]’s behalf do not meaningfully address the different natures of the disclosure sought by the NCND aspect of the appeal as compared with the disclosure sought in the failure to communicate aspect of her appeal. It seems to be argued that [NAME_20]’s paragraph 57 considerations were of at least equal weight but that cannot be right. Each consideration was a function of the anticipated public response to the release of information (actual correspondence) about the [NAME_17]’ and the SPA’s dealings in relation to [NAME_25]. The anticipated response to that disclosure, had it occurred, cannot reasonably be equated with the anticipated response to disclosure that the [NAME_17] either did or did not hold information in the form of correspondence with either of the US Departments. I am not going to make my own findings about the likely public response to such a disclosure but there was a clear issue here that I would have expected [NAME_21]’s submissions to address.

212. [NAME_51] for [NAME_21] argues that [NAME_20] left its paragraph 57 considerations out of account at the NCND stage. They were simply ignored. [NAME_50] for the [NAME_17] counters with the argument that this expects too much. I prefer the somewhat more sophisticated analysis of [NAME_49] for the Commissioner. In referring to a modest increase in public understanding of [NAME_25]’s case, [NAME_20] must have been creating a link with its earlier (paragraph 57) findings all of which were the likely result of the proportionately greater increase in public understanding flowing from disclosure of correspondence between the [NAME_17] and the SPA. The ‘modest increase in public understanding’ referred to by [NAME_20] was intended as a contrast with that likely had [NAME_21]’s part 1 appeal succeeded.

213. I agree with [NAME_49] that a fair reading of [NAME_20]’s statement of reasons shows that it did not forget about its paragraph 57 findings. [NAME_20] considered that confirming or denying whether the information was held would not generate the significant public goods that were referred to in

GIA/973/2018 50 paragraph 57 of [NAME_20]’s statement of reasons. Given the different natures of the disclosures that would ensue, [NAME_20] was entitled to make this finding.

214. Ground 3 fails. I am not persuaded that [NAME_20] left out of account any relevant considerations in carrying out the NCND public interest balancing exercising in relation to [NAME_21]’s appeal insofar as it concerned parts 3 and 4 of her request for information.

Conclusion

215. None of the grounds of appeal are made out. This appeal is dismissed.

(Signed on the Original)

E Mitchell

Judge of the Upper Tribunal 23 August 2019

Clerical mistakes and other accidental slips corrected under rule 42 of the Tribunal Procedure (Upper Tribunal) Rules 2008.

(Signed on the Original)

E Mitchell

Judge of the Upper Tribunal 3 September 2019