The Metaphysics of Information Separation

For some people, August suggests that there is more to life than blogging about how to approach cases in which requested information is environmental and some is not. At Panopticon we regard such fly-by-night lightweight losers as simply wrong. We know you agree (or at least you will, when you get back from holiday and see this).The case in question is Information Commissioner v Department for Transport & Hastings [2018] UKUT 184 (AAC) and despite it being a judgment of Judge Wikeley there are no entertaining literary quotes anywhere to be found. The original request concerned a meeting between the Prince of Wales and, much to every reader’s surprise, a Government Minister at the Department for Transport. The Department dealt with the request under FOIA. The ICO thought most of the information was environment.

By the time of the appeal, only a small portion of the information remained in issue, which the Department considered to be non-environmental and absolutely exempt under section 37(1)(aa), and the ICO considered to be environmental within the meaning of reg 2(1)(c) EIR. (Judge Wikeley, by the way, believes it is inappropriate to use the shorthand phrase ‘reg 2(1)(c) information’ and prefers ‘category (c) information. If you want to know why, read it at [14]. I don’t see why I should have to write it out.) The FTT agreed with the Department.

How then should one go about a classification exercise where a single document contains information which is environmental and which is not environmental? (‘Who cares?’ is not a permitted answer.) Judge Wikeley had the opportunity to revisit the Court of Appeal’s go at making his answer less clear in DBIS v Information Commissioner & Henney [2017] EWCA Civ 844 (see here), from which he drew four life lessons at [44]-[47]. Of importance in this context is that there a conceptual difference between the concept of “information” under FOIA and “environmental information” as defined in the EIR: the latter is also linked to a relevant measure (although Judge Wikeley accepted he had been sent to the “judicial naughty step” for using the phrase “bigger picture”, whilst gracefully only gently hinting that the Court of Appeal did something basically the same).

One can skip to the endorsed three-stage methodology at [69]-[71], which is worth quoting in full:

First, the starting point for a tribunal’s analysis is that “environmental information” in regulation 2(1) of the EIR must be construed broadly (see e.g. recitals (1), (2), (10) and (16) and Article 1 of Directive 2003/4/EC and the case law discussed above).

Second, the document containing the requested information must be considered as a whole. Tribunals should ask themselves whether the requested information as a whole is information ‘on’ one or more of the matters identified in the regulation 2(1) EIR definition.

Third, where the public authority has disaggregated the information in the document into information which it accepts is environmental information (and so governed by the EIR) and information which it considers is other information (and so subject to FOIA), the tribunal must ask itself whether those component parts are separately information ‘on’ one or more of the matters set out in regulation 2(1) of the EIR.

What you should not do is lose sight of what the information is “on”, considering all parts of the information and the link(s) between them, so as to consider whether it is “on” a relevant measure. You should consider it in terms of a ‘predominant purpose’. You cannot assess status by reference to its impact on already disclosed environmental information (i.e. is withholding it making the disclosed information incoherent). The information itself need not be intrinsically environmental: at [54], [58]-[60].

Remaking the decision of the FTT applying the clarified test, the Upper Tribunal held that the remaining information was environmental. It was not simply administrative, but part of a briefing note on an environmental subject: at [85]. Information on housing policy was self-evidently environmental and it does not matter that the disclosed information is perfectly comprehensible without it. “Greater public access to environmental information is thus an end in itself” and is not dependent on whether it would lead to better decision-making: at [86].

So now you know the answer to that question which has puzzled the sages for ages: how do you solve a problem like potentially mixed environmental and non-environmental information? There must be a musical in that, right?

Christopher Knight