The apparently endless APPGER litigation has produced yet another decision of the Upper Tribunal for seasoned FOIA watchers, which amongst some very fact-specific issues, also contains two important clarifications of law: APPGER v ICO & FCO [2015] UKUT 377 (AAC).

As anyone who has ever done any information law ever will know, the APPGER litigation concerns requests under FOIA for information related to alleged British involvement in extraordinary rendition. Some information has been released, some has been released following earlier rounds of litigation, some remains withheld under various exemptions.

Following previous hearings staying various points, the present round of litigation concerned the application of section 23 (the security bodies exemption) and section 27 (international relations). There were two points of wider interest discussed in particular. One is the time at which the public interest is assessed (relevant to section 27), and one is the breadth of the “relates to” limb of section 23.

The time point was one which only really arose because of the Upper Tribunal’s desire to throw a mangy cat amongst the pigeons by suggesting in Defra v ICO & Badger Trust [2014] UKUT 526 (AAC) at [44]-[48] that the correct time to assess the public interest might be the date of Tribunal hearing. As some wise and learned commentators have pointed out, this rather seemed to have been overtaken by the Supreme Court’s – technically obiter – reasoning in R (Evans) v Attorney General [2015] UKSC 21 at [72]-[73] that the time was at the point of the authority’s refusal.

The Upper Tribunal in APPGER (containing at least one member of the panel in Badger Trust) issued a mea culpa and accepted that Evans was right: at [49]-[57]. It did not reach any more specific decision on situations where, for example, the authority has been late in complying. Doubtless the difference in time will often not matter very much. But the principle of the point now seems resolved.

Section 23(1) was not a point answered by Evans, and an argument was run by the requestor that “relates to” should be construed narrowly, as in the DPA. The Upper Tribunal disagreed: at [15]-[19]. The ordinary meaning of the language was broad, it was consistent with the aim of shutting the backdoor to the security bodies, it was consistent with authority, and met the contextual aim of FOIA where the contextual aim of the DPA was very different. The idea of requiring a “focus or main focus” was rejected.

Whilst agreeing that it should not attempt to gloss the statutory language, the Upper Tribunal nonetheless sought to assist future cases by indicating that asking whether the information requested had been supplied to a security body for the purposes of the discharge of its statutory functions (a test attributed to Mitting J) would have considerable utility. It would enable a clear explanation, it would allow differentiation within and without the scope of the exemption, and it was less likely to require a detailed line-by-line approach to redactions: at [33]. The language remains broad, but the practical application of it appears to have been ‘guided’ into a slightly narrower pigeon-hole than might have otherwise been the case.

The judgment as a whole is worth reading on the application of those exemptions to the particular information and the treatment of the evidence by the Upper Tribunal, but those two points of principle are the keys to take away. And about time too.

Timothy Pitt-Payne QC and Joanne Clement appeared for APPGER; Karen Steyn QC appeared for the FCO; Robin Hopkins appeared for the ICO.

Christopher Knight

An important rule of Government is to outsource anything difficult or potentially controversial to an independent body which can then deliver a report to be ignored or implemented as required or the political mood dictate. The recent investigation into new runways at Heathrow was a good example, at least until it came up with an answer the Prime Minister didn’t entirely want to hear, and the Commission on a Bill of Rights was a superlative instance of a very learned study which achieved precisely nothing other than kicking a political football into the long grass.

Now it is the turn of the Freedom of Information Act 2000 to be undergone scrutiny by the Independent Commission on Freedom of Information. Snappy title. It is chaired by Lord Burns (former senior civil servant at HM Treasury) and contains such luminaries as Jack Straw, Lord Michael Howard, Lord Carlisle and Dame Patricia Hodgson (of Ofcom). Just in case anyone was suffering under the delusion that the Commission would be looking into widening the scope and application of FOIA, the terms of reference are set by the Cabinet Office as:

• whether there is an appropriate public interest balance between transparency, accountability and the need for sensitive information to have robust protection
• whether the operation of the Act adequately recognises the need for a ‘safe space’ for policy development and implementation and frank advice
• the balance between the need to maintain public access to information, the burden of the Act on public authorities and whether change is needed to moderate that while maintaining public access to information

One would not, however, wish readers to think that the Government were anything less than fully committed to revealing information. On the contrary, the written statement laid by the Minister, Lord Bridges, opens by saying “We are committed to being the most transparent government in the world.” Well, quite. “We fully support the Freedom of Information Act [could there be a ‘but’ coming?] but [ah yes, there it is] after more than a decade in operation it is time that the process is reviewed, to make sure it’s working effectively.” The new Commission has a webpage here and is to report by November, which gives the grass limited time to lengthen… The Commission won’t, of course, be able to do anything about the EIRs.

Responsibility for FOIA has also been transferred to the Cabinet Office, which at least gives Michael Gove one less constitutional headache to deal with.

Christopher Knight

Mr Brown became a well-known figure in litigation circles when he sought to unseal the Will of Princess Margaret in the belief that it might reveal information showing him to be her illegitimate son. In the course of his unsuccessful litigation, it was revealed that there existed what had been described orally during the court proceedings as a “Practice Direction in respect of the handling of Royal Wills” (although there is dispute over precisely what form this document takes and whether it is really a Practice Direction at all), produced by the-then President of the Family Division following liaison with the Royal Household.

Having failed to unseal the Will, Mr Brown requested a copy of the document from the Attorney General. He was refused, under section 37 FOIA. The First-tier Tribunal upheld that refusal (on which see Robin’s blog here). Mr Brown appealed to the Upper Tribunal on the grounds of inadequacy of the Tribunal’s reasons and a failure to properly apply the public interest test. He was refused permission, but then successfully judicially reviewed the Upper Tribunal for failure to grant him permission (on which, see my blog here).

Much happened subsequently. Having fought hard to prevent disclosure of the ‘Practice Direction’ the AG then released almost all of it to Mr Brown in advance of the substantive appeal hearing before the Upper Tribunal. The unreleased aspect was one paragraph, which was supplied to him in ‘gisted’ form. Nonetheless, Mr Brown sought disclosure of the outstanding paragraph. Perhaps not entirely surprisingly, Charles J in the Upper Tribunal has just refused to give him the final missing piece: Brown v ICO & Attorney General [2015] UKUT 393 (AAC).

The Upper Tribunal decision, in the light of the release by AG, had rather less work to do than it might have done, and the judgment will be of equivalent reduced wider interest. However, Charles J does roundly endorse the proposition that there is a very powerful public interest “against the creation of undisclosed principles and procedures to be applied by the court to an application to seal any will, and this is strengthened when participants in and the decision maker on that application (the court through initially or generally the President of the Family Division) and the normal guardian of the public interest (the Attorney General) have been involved in its creation on a confidential and undisclosed basis, and so in favour of the publication of the principles and procedure to be applied on any such application (particularly if initially or generally the application will be made in private)“. In other words, the AG was right to concede that the material should be disclosed. There was no further interest in the gisted paragraph also being revealed because the essential meaning had been conveyed.

Whether this brings Mr Brown’s campaign to an end is another matter, but whatever one might think of his view as to his parentage, his uncovering of a – to put it neutrally – highly unusual document agreed between the AG, the Royal Household and the President of the Family Division concerning court procedures is a worthy effort.

Robin Hopkins appeared for the ICO; Joanne Clement appeared for the Attorney General and Anya Proops appeared for Mr Brown at some of the earlier stages of proceedings.

Christopher Knight

The principle disadvantage, to the data protection lawyer, of the failure of Esperanto is that every now and then the CJEU hands down a judgment which is only available in French, and even Panopticon cannot blog every entry in Franglais. Such is the problem raised by the Opinion of the Advocate General (Cruz Villalon) in Case C-201/14 Bara v Presedintele Casei Nationala de Asigurari de Sanatate. Readers will have to forgive any failure to capture the nuances.

Bara is a reference from the Romanian courts and contains a number of questions, the majority of which concern the compatibility of national tax authority arrangements with Article 124 TFEU (which prohibits in most cases privileged access for public bodies to financial institutions). Those need not concern us, not least because the AG considered them to be inadmissible.

However, the fourth question referred was in the following terms: “May personal data be processed by authorities for which such data were not intended where such an operation gives rise, retroactively, to financial loss?” The context appears to be that people deriving their income from independent activities were called to pay their contributions to the National Fund for health insurance, following a tax notice issued by the Romanian health insurance fund. However, that tax notice was calculated on the basis of data on income provided National Tax Administration Agency under an internal administrative protocol. The complaint was that the transfer by the Tax Agency to the Health Insurance Fund of personal data, particularly related to income, was in breach of Directive 95/46/EC because no consent had been provided to the transfer, the data subjects had not been informed of the transfer and the transfer was not for the same purpose as the data was originally supplied.

The Advocate General answered the fourth question by saying that the Directive precludes national legislation which allows a public institution of a Member State to process personal data that has been supplied by another public institution, including the data relating to the income of the persons concerned, without the latter having been previously informed of this transmission or treatment. This was despite the fact that the AG recognised that the Romanian bodies had a legitimate interest in being able to properly tax self-employed persons; the informal protocol did not constitute a legislative measure setting out a relevant national exemption under Article 13. The AG stressed that the requirement of notification in Article 11 had not been complied with, and that the data subjects accordingly had been unable to object to the transfer. The data subjects had not given their unambiguous consent. Although Article 7(e) (necessary for the performance of a task) could apply to a transfer of income data, it had to be shown that it was strictly necessary for the realisation of the functions of the Health Insurance Fund. (This appears to be a higher test being imposed than the usual interpretation of necessary as ‘reasonably necessary’, as per the Supreme Court in South Lanarkshire). The AG did not consider that test met.

It remains, of course, to be seen whether the CJEU will take the same approach; but it seems fairly likely that Bara will produce a judgment which confirms the illegality of inter-institutional transfer of personal data without express consent or a carefully defined need which is prescribed by law. There is nothing ground-breaking in that conclusion, but it is an important reiteration of the need for data controllers anywhere in the EU to think carefully about the authorisation they have to hand over personal data to other bodies; informal agreements or policy documents are not sufficient without a legal underpinning (through the DPA) or consent of the data subject.

The forthcoming judgment in Case C-582/14, Breyer will also raise issues over consent in the context of IP information retained by websites, along with the vexed question of whether an IP address can constitute personal data when combined with other information available to a third party (issues similar to those raised in Vidal-Hall v Google, on which see here). When the final judgments in Bara and Breyer appear, so will the analysis of some intrepid blogger of this parish.

Christopher Knight

Following a period of considered reflection, or laziness depending on one’s view, it is worth noting the decision of the Supreme Court in In the matter of an application by JR38 for Judicial Review [2015] UKSC 42. The case is all about Article 8 ECHR, and is of particular interest because of the dispute about the breadth of the correct test for the engagement of Article 8. The context is also one which will be familiar to English data protection and privacy lawyers: the publication by the police of photographs seeking to identify a suspect. If anyone remembers that famous picture of a youth in a hoodie pointing his fingers like a gun behind an awkward looking David Cameron, JR38 is basically that, but with Molotov cocktails and a sprinkling of sectarian hatred.

In JR38, the suspect in question was a 14 year child whose photograph was published by the PSNI as someone involved in rioting in an area of Derry in 2010 which had particular sectarian tensions. The judgment of Lord Kerr makes clear that JR38 has by no means been a well-behaved young man before or since the riots of 2010. Moreover, and amusingly, it is apparent that he and his father failed to correctly identify his own appearance in pictures published, and originally sued on the basis of images which did not show JR38 at all. However, a correct image was eventually alighted upon.

The judgments contain a lengthy and detailed discussion of the domestic and Strasbourg case law on the engagement of Article 8, but there was a 3-2 split in the Court between the correct approach. Lords Toulson and Clarke (with both of whom Lord Hodge agreed) considered that the overwhelming approach of the existing domestic law was to apply the touchstone of the reasonable/legitimate expectation of privacy test: see Lord Toulson at [87]-[88]. The test (originating, of course, in Campbell) focuses on “the sensibilities of a reasonable person in the position of the person who is the subject of the conduct complained about…If there could be no reasonable expectation of privacy, or legitimate expectation of protection, it is hard to see how there could nevertheless be a lack of respect for their article 8 rights”. The warning in Campbell not to bleed justification matters into the engagement analysis was stressed.

The difference between the majority and minority of Lord Kerr (with whom Lord Wilson agreed) was explained by Lord Clarke at [105]. Does the reasonable expectation of privacy test provide the only touchstone? The majority thought that it did, it being the only test set out clearly in the cases, and it being a broad objective concept to applied in all the circumstances of the case and having regard to the underlying values involved, unconcerned with the subjective expectation of the individual, be they child or adult (see at [98] per Lord Toulson and [109] per Lord Clarke).

In essence, the majority did not consider this context to be one which Article 8 was designed to protect. The identification of a suspect was not within the scope of personal autonomy, although publication of the same picture for a different purpose, other than identification, might be: at [98] (and at [112] where Lord Clarke did not consider the mere fact of criminal activity took the matter outside Article 8). Historic or re-published photos may alter the situation: at [101].

By contrast, Lord Kerr took a broader view, holding that the reasonable expectation of privacy test might be the ‘rule of thumb’, but could not be an inflexible, wholly determinative test. The scope of Article 8 was much broader and was contextual, requiring consideration of factors such as: age, consent, stigmatisation, the context of the photographed activity and the use of the image. Reasonable expectation of privacy failed, in his view, to allow for these factors to be considered: at [56]. Rather than shoehorning such factors into the test, they should bear on the issue in a free-standing footing: at [61]. The focus must be on the publication – i.e. the infringement – rather than the activity the photo displays. For Lord Kerr, the fact that JR38 was a child, taken with the potential effect publication might have on the life of the child, was more than sufficient to engage Article 8 (in the way that it might not for an adult): at [65]-[66].

The debate is an interesting one, but there is a very strong chance that the flexibility of the majority orthodox approach is likely to mean very little difference in substance between the two. It will, however, be worth emphasising the importance of context, particularly in child cases under Article 8.

The Court was, however, unanimous in agreeing that publication was justified in any event; rioters had to be identified (and other methods had been tried internally first), with the peril in which inter-community harmony was placed being particularly important in the fair balance.

Where, readers of this blog might ask, was the DPA in all this namby-pamby human rights discussion? Why is there no mention of schedules and data protection principles and all the other black letter statutory stuff that so gets the blood pumping? Well, it was mentioned, at [70], by Lord Kerr who considered that compliance with the DPA would mean that the limb of proportionality which requires the act to be in accordance with the law would be met. In very brief reasoning, Lord Kerr concluded that this type of case was within section 29 because publication was processing for the purposes of prevention and detection of crime, and that the relevant condition met in both Schedule 2 and 3 (because he agreed it was clearly sensitive personal data) was that of the processing being necessary for the administration of justice. Unfortunately, there was no analysis of the way in it was necessary for the administration of justice, or the extent to which this is the same as the prevention and detection of crime. Nor is it quite the same reasoning as adopted by Lord Woolf CJ in the well-known ‘naming and shaming’ case of R (Ellis) v Chief Constable of Essex Police [2003] EWHC 1321 (Admin), which, at [29], appeared to apply the conditions in Schedules 2 and 3 whereby processing was necessary for the performance of functions by or under any enactment (without further specification). Where the Supreme Court speaks, we follow, but it might have been helpful to detail this aspect a little more, although it is another example of a case in which Article 8 is presumed to do all of the work and the DPA be raced through in a paragraph to avoid having to think about it too much. That Article 8 and the DPA are ensured to be pulling in the same direction is, however, a relief to us all.

 Christopher Knight

You know how it took you years to get your head around what the Data Protection Act 1998 meant? Well, the new general Data Protection Regulation took one step closer towards ripping up Directive 95/46/EC (and therefore the DPA) today. The Commission’s proposals for the new Regulation have now been signed off by the Justice Ministers (see press release here). The Commission informs us that the first meeting in the trilogue between the Commission, the European Parliament and the Council of the EU will take place on 24 June, and the aim is to have a text agreed before 2015 is out. The trilogue is where all the nitty gritty stuff has still to be hammered out, and plenty of changes are likely through that process, but at least we have a little more of an idea now of the timescale.

Christopher Knight