David Hendy v The Information Commissioner & Anor

Research on transmission of bovine tuberculosis to cattle is a subject of great interest to many, informing a debate “which engenders very strong feelings on both sides”: R. (NFU) v DEFRA [2020] EWHC 1192 (Admin). That interest includes scrutiny of data that may ultimately affect government policy. On 22 November 2018 Mr Hendy requested information from the Animal & Plant Health Agency, an executive agency of Department for Environment, Food and Rural Affairs, concerning tuberculosis rates in cattle. APHA considered it to be vexatious, so refused to comply with it. On 22 July 2019 the Commissioner issued a decision notice agreeing with APHA (Footnote: ¹), against which Mr Hendy has appealed.

In a decision dated 13 May 2020 (Footnote: ²) the Tribunal allowed Mr Hendy’s appeal in part, but only to the extent of finding that the Commissioner had erred in finding that the request fell to be considered under the Environmental Information Regulations 2004 (“EIR”) rather than the Freedom of Information Act 2000 (“FOIA”). The conclusion that the request was vexatious was upheld. That decision was then set aside by Upper Tribunal Judge Mitchell on 19 November 2021 (Footnote: ³), upon Mr Hendy’s onward appeal. The proceedings were remitted to this Tribunal to be decided afresh.

The following uncontroversial background stands as findings of fact.

Before coming on to the present request for information, it is necessary to set the scene by describing a previous request made by Mr Hendy on 11 September 2017. The evidence before us calls it the “2017 Request”. Upper Tribunal Judge Mitchell called it “Request 1” and described it as follows: (“OTF” means ‘officially TB-free’; “OTFW” means ‘officially TB-free withdrawn’, meaning that the herd’s TB-free status has been withdrawn):

The 2017 Request was rejected by APHA as requiring work above the relevant cost limit of £600 – a sum taken for present purposes to represent 24 hours’ work. It reached that conclusion after commencing a sampling exercise limited to 2016 information only. This was abandoned after the creation of even an incomplete data set for just 2016 took 42 hours of staff time; the request would plainly exceed the cost limit. To comply with its duty under s.16 of FOIA to provide advice and assistance as to how a refined request might come within the cost limit, APHA suggested an alternative request for just:

Dissatisfied, Mr Hendy complained to the Commissioner. In a decision notice of 23 March 2018 the Commissioner agreed that compliance with the 2017 Request would exceed the cost limit, and concluded that APHA had complied with its s.16 duty. Mr Hendy appealed. His arguments included that APHA and the Commissioner had misunderstood the ‘either/or’ conditions within the 2017 Request as expanding his request rather than limiting it, and that compliance would come under the cost limit if APHA were to create or acquire better reporting tools or analysts. In a decision dated 11 October 2018 (Footnote: ⁴) the Tribunal dismissed that appeal, holding that application of Mr Hendy’s limiting conditions would not bring the cost below the limit and that FOIA did not require a public authority to upgrade its staff, equipment and systems in order to reduce the cost of complying with a request for information.

This is the present request for information made on 22 November 2018, with which this appeal is concerned. Again, the evidence before us refers to it as the 2018 Request and the Upper Tribunal refers to it as Request 2. We set it out in full:

APHA’s response to Request 2 was to reject it as vexatious. APHA considered that Request 2 was identical to Request 1, save that it only sought information for 2015 and 2016: ignoring the sampling exercise for 2016 which had taken 42 hours before being abandoned. APHA also considered that its section 16 advice had been ignored and that Request 2 attempted to re-open a matter substantially similar to that which had already been fully addressed, including on appeal. Its email continued:

Following APHA’s decision being maintained on internal review, Mr Hendy complained to the Commissioner on 20 March 2019. In the decision notice of 22 July 2019 the Commissioner first decided that the applicable regime was that contained in EIR, rather than FOIA as applied by APHA. EIR nonetheless contained an equivalent provision to s.14 of FOIA, which applies where a request is deemed to be manifestly unreasonable either by virtue of costs or because the request is deemed to be vexatious. We do not set out the entirety of the Commissioner’s reasoning, but it can be seen that the number of requests made by Mr Hendy, their technicality, and the resulting burden placed on APHA’s staff were all held to amount to vexatiousness. That conclusion was not outweighed by public interest in disclosure, wider statistics already being released online and Mr Hendy’s requested information only likely to be of use to data scientists rather than the general public.

We have summarised this Tribunal’s previous decision on this appeal at paragraph 2. On appeal to the Upper Tribunal, Mr Hendy argued that this Tribunal had failed to properly address the way in which he had narrowed the scope of Request 2 in accordance with the outcome of Request 1. As previously noted, APHA and the Commissioner considered the requests to be identical save for the years requested. After setting out his own comparison of the two requests, UTJ Mitchell’s conclusions were as follows:

We respectfully agree with that analysis, from which neither respondent has asked us to depart. It then informed UTJ Mitchell’s consideration of whether this Tribunal had erred in law, which (so far as material) was as follows:

We have taken careful account of the above context in which the appeal now falls to be re-decided.

In Ms Thelen’s skeleton argument provided on 30 May 2022, well in advance of the hearing, APHA additionally seeks to rely on the exemption at section 12 of FOIA: that complying with the request would exceed the relevant cost limit. As confirmed in Birkett v DEFRA [2011] EWCA Civ 1606, subject to the Tribunal’s case management powers a public authority is entitled to invoke additional exemptions at any point during the proceedings. Mr Hendy has raised no objection, and we are satisfied that it is fair to permit APHA to argue this additional exemption.

The appeal was heard on 28 June 2022 by CVP, all participants connecting remotely. No adverse technical issues were encountered, and we are satisfied that the appeal was as fairly and effectively heard as if it had taken place entirely within a physical courtroom. Mr Hendy represented himself, and APHA was represented by Ms Thelen of counsel. The Commissioner did not attend and was not represented. The documents to be considered were agreed as comprising the following:

The Tribunal heard evidence from Mr Hendy and his witness Dr Hazel Wright, and then from APHA’s witnesses Ms Allison White and Mr Ross Akehurst. The parties made their closing submissions and our decision was reserved. The Tribunal apologises for the subsequent delay in promulgating this decision. We are satisfied that no unfairness arises from us now deciding the appeal: a contemporaneous note was taken of all the evidence and submissions, and the panel deliberated upon, reached and noted its principal conclusions shortly after the hearing.

The issues are:

The principles by which the Tribunal decides the above issues were confirmed by the Upper Tribunal in Information Commissioner v Malnick and Anor [2018] UKUT 72 (AAC) at [45] and [90]. Section 58(2) of FOIA provides that the Tribunal may review any finding of fact on which the notice in question was based. This has been authoritatively held to mean that the Tribunal exercises a full merits appellate jurisdiction, making any necessary findings of fact and then deciding for itself whether the provisions of FOIA have been correctly applied. But it does not start with a blank sheet: the starting point is the Commissioner’s decision, to which the Tribunal should give such weight as it thinks fit in the particular circumstances. The issues are to be decided as of the date of the public authority’s response, in this case 22 July 2019: Montague v Information Commissioner and DIT [2022] UKUT 104 (AAC).

Finally, APHA’s written submissions include that it “does not hold the data requested – it would have to produce new data by extracting the information necessary to create the figures requested.” It was clarified at the hearing that “produce” referred here to extraction from APHA’s database, rather than being intended to introduce a discrete issue under section 1(1) of FOIA as to whether APHA holds the requested information at all.

This depends on whether the information sought is “environmental information”.

“Environmental information” is defined by the EIR at regulation 2(1) as including any information in written, visual, aural, electronic or other material form on (so far as relevant in this appeal):

In BEIS v ICO & Henney [2017] EWCA Civ 844 the Court of Appeal upheld a decision by the Upper Tribunal that the communications and data component of the government’s Smart Meter Programme is “environmental information”. The issue between the parties was when and whether information on a measure which does not in itself affect the state of the elements of the environment or the factors referred to in regulation 2(1)(a) and (b) of the EIR, can be information “on” another measure which does.

The Court of Appeal confirmed that the EIR must be interpreted as far as possible in accordance with the purpose of Directive 2003/4/EC, and the Aarhus Convention to which it is intended to give effect, including that the objective that the public may be better informed and better able to contribute to environmental decision-making if given access to the requested information. Those provisions, taken with domestic and European authority, point towards a broad interpretation of the term “environmental information”. In addressing the ‘measures’ to which regulation 2(1)(c) refers, Beatson LJ held that identifying the measure that the disputed information is “on” may require consideration of the wider context, and is not strictly limited to the precise issue with which the information is concerned. In purposively applying regulation 2(1), the information has to be considered in its context. There must nonetheless, he held, be a limit. The provisions are not intended to give a general and unlimited right of access to all information held by public authorities which has a connection, however minimal, with one of the environmental factors mentioned.

Henney was subsequently applied by the Upper Tribunal in DfT, DVSA and Porsche Cars GB Ltd v ICO and Cieslik [2018] UKUT 127 (AAC). That case concerned information regarding a safety test of a car where the car’s throttle was alleged to have been designed to circumvent environmental legislation in respect of noise emissions. This was held not to be environmental information, the connection between testing for safety and the effect upon the environment being too remote. The test was neither intended to assess environmental elements or factors, nor was it conducted by a body with an environmental remit. Even if the test results were to shed light on a vehicle’s performance in environmental tests, that “would be no more than incidental”. The contrary approach, such as that of this Tribunal with which the Upper Tribunal was concerned, would encompass information about any activity which involves running an engine. Virtually anything done by any person or thing has some level of effect on or interaction with the environment. The Upper Tribunal also held that a public authority could not be expected to assess the merits of any “underlying environmental grievance” of the requester in order to determine whether information fell within the regulation 2(1) definition.

Henney and Cieslik are primarily concerned with measures under regulation 2(1)(c), but we see no reason why the contextual approach they recommend should not apply to the other elements listed in regulation 2(1).

The Commissioner’s guidance “What is Environmental Information” states that human health and safety at regulation 2(1)(f):

Mr Hendy has not taken any position on the applicable regime.

In the decision notice the Commissioner noted that the requested information relates to a programme which aims to reduce and ultimately eradicate bovine TB. While APHA had argued that the risk to human health from bovine TB is very low, the information nonetheless related to TB testing under that initiative and therefore related to human health and potential contamination of the food chain. The Commissioner has otherwise put forward no submissions on the issue.

APHA argues that the applicable regime is that contained in FOIA. Ms Thelen put forward five arguments. First, the information was plainly concerned with bovine health rather than human health. The requested information merely comprises statistics on the incidence of bovine TB while saying nothing about effects on human health or contamination of the food chain. Second, the Commissioner’s analysis neglected the requirement at regulation 2(1)(f) that human health be affected by one of the elements of the environment at (a) or (b). Third, the Commissioner’s analysis fails to substantiate the claimed connection between the raw statistical data requested and Defra’s anti-bovine TB programme. Fourth, the information cannot relate to the element of biological diversity which, by definition, is concerned with the balance between species. Fifth, the Commissioner’s analysis would require any information relating to bovine TB testing to be treated as environmental information simply because one of the governments objectives is to mitigate the risk to human health posed by bovine TB. That would fail to reflect the principle that information which has only a minimal connection with the environment is not to be treated as environmental information.

As recognised in Henney, where to draw the line in a particular case is not always easy. We reject Ms Thelen’s first, third, and fifth arguments above as requiring far too narrow a focus on the subject with which the information is specifically concerned. The wider context must be considered, being that a purpose of bovine TB reduction is to protect human health. There is no need for it to be the principal purpose, and the connection between bovine TB reduction and the information requested in this appeal is obvious.

We do accept Ms Thelen’s second and fourth arguments. While the requested information may well be “on” the state of human health and safety, regulation 2(1)(f) also requires that the state of human health and safety be affected (or may be affected) by “the state of the elements of the environment referred to in (a) or, through those elements, by any of the matters referred to in (b) and (c)”. None of the elements referred to in (a) appear to us to be engaged by the spread of disease in livestock. It might be wondered whether the element of ‘biodiversity’ might be engaged if a clear enough connection were established between the use of this data and the bovine TB reduction measure of culling wildlife, but there is no evidence or argument before us on the point.

We therefore conclude that the applicable regime is that contained within FOIA.

FOIA does not oblige a public authority to comply with a request for information if it estimates that the cost of compliance would exceed the appropriate limit. In this appeal, the Freedom of Information and Data Protection (Appropriate Limit and Fees) Regulations 2004 prescribe a limit of £600 and an hourly rate of £25 for the work of the public authority’s personnel. If no costs other than time are claimed, the question is therefore whether complying with the request would take more than 24 hours. The statute’s use of the word ‘would’ likely indicates a higher degree of certainty than if read ‘may’ or ‘might’.

The public authority’s estimate must be reasonable, in the sense of being sensible, realistic and supported by cogent evidence. It may only take account of the following activities:

As held in Kirkham v Information Commissioner [2018] UKUT 126 (AAC), the test is subjective to the public authority but then qualified by an objective element. The Tribunal first makes findings as to the nature of the estimate made by the public authority. These will be related to the way in which the public authority holds the information; FOIA does not impose any particular record-keeping practices. The Tribunal may next remove from the estimate any amount which it considers unreasonable, either on account of the nature of the activity to which it relates or its amount.

Notwithstanding the late reliance on this exemption, APHA had already estimated the time needed for compliance when assessing whether the request was vexatious. The estimate now put forward is 39 hours and 9 minutes, well above the relevant limit. We are grateful to Ms Thelen for so clearly signposting – at paragraphs 45-52 of her skeleton argument and at the hearing – which of the bundle’s many exhibits, appendices and annexes, and which parts of the witness evidence, now comprised APHA’s up-to-date evidential case in support of that estimate.

First is the witness statement of Allison White dated 20 February 2020, made for the purposes of the present appeal. She works for APHA as a Records Manager and also manages its Access to Information Team. Her evidence is more relevant to the overall burden of the request on APHA within the context of vexatiousness but does give useful factual background on the way in which the requested information is held and how it would be extracted. At page 9 of her witness statement is a table entitled “Table 1”, which sets out time estimates for various tasks involved in providing the information. She used those figures to reach an estimate of approximately 50 hours. Ms White attended the hearing before us and confirmed her witness statement was true to the best of her belief at the time it was signed. There were practical obstacles to her giving further oral evidence that we need not describe in this decision. Ms Thelen confirmed that she would not object to Mr Hendy’s closing submissions containing points that had not been put in cross-examination, nor would she argue that this reduced the weight they carried. Nor did Mr Hendy object to the procedural course that was adopted.

Second is the evidence of Ross Akehurst, the Team Manager of Management Information and Data Architecture Services – the “MIDAS” team. He has made two witness statements. The first is dated 25 June 2018 and was made for the purposes of the previous appeal in respect of the 2017 Request, dismissed by this Tribunal on 11 October 2018. By implication, the Tribunal’s reasons for dismissing the appeal included that Mr Akehurst’s evidence was accepted. At paragraph 57 of that statement Mr Akehurst describes having produced a document called “Annex 4A” with some detailed workings on one of the alternative options said to have been put forward by Mr Hendy. Annex 4A had then been used by Ms White to produce Table 1.

Mr Akehurst’s second witness statement is dated 8 April 2022 and was made in recognition of the Upper Tribunal’s finding that the 2017 Request had been for one of the two alternatives specified rather than for both (see paragraph 11 above). It therefore exhibits updated versions of the figures in Ms White’s statement, removing time related to the second alternative. One of these – “Table 1 Updated”, an updated version of Ms White’s Table 1 – is a key document in this appeal. It was confirmed by both Mr Akehurst in evidence and Ms Thelen in submissions that Table 1 Updated represents its case in this appeal on the estimated cost of compliance. Its total of 38 hours and 9 minutes is broken down as follows:

Each of these figures is reached by taking the figures for “Ross Akehurst time calculations for 2017 Request (revised)” and then applying them to the work required by the narrower 2018 Request.

The main focus at the hearing was on “Initial Exploration Time”. In Table 1 Updated, that heading is footnoted as follows:

That footnote is confirmed by the way in which Part (a) is described in relation to the 2018 Request in the relevant column of Table 1 Updated:

This commentary speaks for itself. Part (a), the 19 hours and 37 minutes, was necessary to meet the 2018 Request but had already been done. Part (b), 2 hours and 40 minutes, and Part (c), 5 hours and 36 minutes, had been done previously but would have to be done again; further commentary explains that this is due to the possibility of changes to the underlying dataset, which we accept.

In oral evidence it was first clarified with Mr Akehurst that Part (a) had already been done by the time Mr Hendy made the 2018 Request. He confirmed that it had, but explained that it had been included in the time estimate because it was essential to comply with the 2018 Request. He explained that he saw two possible ways of approaching the estimate. If the 2018 Request were seen as a narrowed 2017 Request then the 19 hours and 37 minutes had clearly gone towards complying with it. If, on the other hand, the estimate should be produced on the hypothesis that the 2017 Request had never been made, then that hypothetical scenario would require 19 hours and 37 minutes’ work to be done. Either way, the time ought to be included.

Mr Akehurst was referred to the footnote and confirmed that 6 hours was a reasonable estimate of the time that would have to be spent by APHA re-familiarising itself with the work done under Part (a). He was referred to Ms White’s original Table 1, which had a separate row in the table after “Initial Exploration Time” called “Further Exploration Time” described as follows:

He confirmed that the separate row in Ms White’s table and the footnote in his table were two ways of representing the same thing. This is consistent with paragraph 14 of his second witness statement.

We pause at this point in Mr Akehurst’s evidence to note that neither of his two possible approaches described at paragraph 43 above represent the correct approach to s.12 of FOIA. In her closing submissions, Ms Thelen argued that Request 2 was properly seen as a narrowing of Request 1 and treating them as part of the same request for costs purposes did not offend against the statutory scheme. We cannot accept this. For the purposes of FOIA, Requests 1 and 2 are separate requests and the cost of the former cannot be added to the cost of the latter. Section 12(4) provides that:

When empowering the Minister for the Cabinet Office to make regulations as to when the cost of two or more requests may be aggregated, Parliament saw fit to limit the circumstances to those at s.12(4)(a) and (b). This is incompatible with the public authority being permitted to do so in much wider circumstances simply due to a thematic link, or due to the second request being subsequent to advice and assistance given under section 16. In agreement with paragraph 65 of the Commissioner’s guidance on the cost of compliance, we hold that s.12 and the Fees Regulations cannot be read as allowing a costs estimate to include costs that have already been incurred in relation to a previous request (save where the conditions specified in the Fees Regulations apply). On that basis, the 19 hours and 37 minutes should be altered to 6 hours. That is the time, as the footnote, documents and witness statements make clear, that would actually be spent on the 2018 Request.

We do not criticise Mr Akehurst for failing to appreciate the legal position when expressing his view on what a reasonable estimate might include. He is a data scientist rather than a lawyer, and it may well be frustrating to discover that such a significant amount of work falls to be discounted. We did, however, identify a watershed in his evidence. Upon suddenly appreciating from the panel’s questions that inclusion of previously-done work might be impermissible, he changed his position. Rather than 6 hours being how long Part (a) of Initial Exploration Time would actually take, he now described it as the amount of time that could be deducted from the original figure. In answer to further questions from Ms Thelen he went further, stating that the 6 hours should be specifically deducted from one particular task appearing on Appendix 2 Updated: “Two staff members – planning / design – 11 hours”. He explained, in essence, that this task was a series of meetings and communications. Ms Thelen asked him if further time could be deducted from the other rows in Appendix 2 Updated that comprise the Initial Exploration Time Part (a) in Table 1 Updated. He scanned the various rows, thinking them through, and found 2½ hours of further deductions from other rows. The remaining tasks, he said, would all have to be repeated. If this second position is correct, it means that the 19 hours and 37 minutes is reduced to 11 hours and 7 minutes instead of 6 hours.

Having carefully considered Mr Akehurst’s evidence, we are satisfied that the second position he expressed was inconsistent with the first. It is the first position that mostly likely represents APHA’s true estimate. Unlike the second position, it is consistent with the actual wording of the footnote and commentary on Table 1 Updated and with the way in which the 6 hours is accounted for in Ms White’s original Table 1. The narrow scope to the 6 hours’ work in Mr Akehurst’s second position is inconsistent with both the wording of the footnote and the wider scope given to that work by Ms White (see paragraph 44 above). The identification on the fly of further deductions from Part (a) was unconvincing. If further deductions could so easily be found, then why was only the 6 hour figure so clearly identified in the written documents? The answer, we find, is that the 6 hours is not a deduction at all. Reinforcing that analysis is our observation that, towards the end of his evidence, Mr Akehurst became more concerned with advocating on behalf of APHA than simply answering the questions put to him; an example being when he disagreed with some figures put to him by Mr Hendy, only to change his mind when it was pointed out that the figures had simply been replicated from Mr Akehurst’s own.

For the reasons above, we find that APHA’s genuine subjective estimate of Part (a) of the “Initial Exploration Time” for the purposes of s.12 is 6 hours. The remaining figures in Table 1 Updated reflect APHA’s genuine subjective estimate. This gives a total estimate of 24 hours and 32 minutes.

The relevant principles are already set out above. Mr Hendy’s criticisms of various amounts claimed by APHA are helpfully set out in his document entitled “Appellant’s Final Submission”. Most concern the tasks that make up the 19 hours and 37 minutes, and we have already reduced that to 6 hours. Given our conclusions on subsequent issues we are content to take a broad brush approach and assess the 6 hour period as reasonable, provided that it has the wider scope described in Ms White’s original Table 1. We therefore do not need to address Mr Hendy’s specific criticisms, but we would otherwise have had concerns on the extent to which the original figure included time recording (and how long was spent doing it) and internal conversations between APHA personnel and Ms White. The panel, with its specialist experience, consider that the latter activity in particular is not reasonably included in an estimate. It is the time spent in location, retrieval and extraction that matters, rather than time spent telling someone else how to do it or how it has been done.

Under the heading “Finalising (compilation)”, APHA specifies 1 hour. Mr Hendy argues that none of the permitted categories of work include composing the response letter to the requester. The Commissioner’s guidance is silent on this particular issue, but as a matter of plain language we agree that communicating information to the requester does not fall within the permitted categories; they all end at the point at the information is retrieved or extracted.

In some cases, explanation of data might properly fall within retrieval and extraction if it is necessary for the information to be understood at all – for example, “on this spreadsheet, column A represents X”. But in this case the need to do this on top of the various other tasks, which devote significant time to conversion of reports into spreadsheets and quality checking, is not cogently explained. Asked about this cost in oral evidence, Mr Akehurst said that it was first to produce an email to Mr Hendy setting out the context of the information “to confirm to him that he is getting what he thought he was asking for” and second to communicate with Ms White’s team so that she could deal with any follow-up questions from Mr Hendy. We cannot see that either task falls under one of the permitted categories of work, and we exclude this hour from the estimate.

Another challenge is to the testing of data extracted. The approach taken by APHA is to apply a 25% uplift to represent time verifying and testing results. Mr Hendy objects to that uplift applying to those tasks which are already concerned with testing, resulting in an additional 40 minutes, describing it as “testing the testing”. Mr Akehurst’s response was to draw an analogy with installing an IT system. Each of the individual computers must be checked to see if it is working, but so too must the overall system once it has been installed. While we understand that analogy, insufficiently cogent evidence is provided to justify its application to the particular tasks. Asked questions on the specific tasks by Ms Thelen, Mr Akehurst demonstrated a degree of uncertainty as to why they did not fall within the general uplift. Given our assessment of his evidence above, we are unable to accept mere assertion that they would have been properly included. Mr Hendy had flagged this concern well in advance, and we cannot find that APHA’s inclusion of this 40 minutes is supported by cogent evidence. We deduct it from the total.

We have reduced the estimate by a further 1 hour and 40 minutes, which takes it to 22 hours and 52 minutes. This is below the cost limit and the exemption at s.12 does not apply.

For the principles that apply to deciding whether a request is vexatious, we first turn to Information Commissioner v Dransfield [2012] UKUT 440 (AAC). The Upper Tribunal emphasised that the request must be vexatious, not the requester. A request may be inconvenient, irritating, or burdensome to the public authority without necessarily being vexatious; holding public authorities to account by giving access to information is one of the purposes of the legislation. That creates a balancing exercise, where the distress, disruption, irritation or burden caused by a request that must be weighed against the justification for making it. It is important to adopt a holistic and broad approach that considers all the relevant factual circumstances. They are likely to fall under four headings: (a) the burden on the public authority and its staff; (b) the motive of the requester; (c) the value or serious purpose of the request; and (d) any harassment or distress. The Upper Tribunal gave more detailed guidance on each of those topics, to which we shall refer within our own analysis.

An appeal against the Upper Tribunal’s decision in Dransfield was dismissed. In Dransfield v The Information Commissioner [2015] EWCA Civ 454, Arden LJ (as she then was) approved the Upper Tribunal’s analysis and guidance subject to some clarification of her own. While the aim of s.14(1) might be to protect the authority’s resources from being squandered on disproportionate use of FOIA, as held by the Upper Tribunal, that aim would only be realised if the high standard set by vexatiousness was also satisfied. Parliament had chosen a strong word in ‘vexatious’, meaning that the hurdle of satisfying it is a high one: consistent with the constitutional nature of the right. It is an objective standard, primarily involving making a request where there is no reasonable foundation for thinking that the information sought would be of value to the requester, or to the public or any section of the public. This, and a relevant motive that could be identified with a sufficient degree of assurance, such as vengeance for the public authority’s actions, might both be evidence from which vexatiousness could be inferred. Arden LJ nonetheless added that this “could not be said, however vengeful the requester, if the request was aimed at the disclosure of important information which ought to be made publicly available.” At [85], Arden LJ also agreed that a request might be vexatious in part because of, or solely because of, the costs of complying with it. The preservation of the Upper Tribunal’s guidance in Dransfield has subsequently been confirmed in authorities such as Cabinet Office v ICO and Ashton [2018] UKUT 208.

We divide our consideration into the list of topics suggested by the Upper Tribunal in Dransfield, before setting out our overall conclusion.

We have regard to all the circumstances, and particularly under this heading to the the parties’ previous course of dealings and the number, breadth, pattern and duration of requests.

APHA points to the time that this request would take to fulfil, together with that spent already. The evidence of Ms White and Mr Akehurst gives more specific detail on the burden that this is apt to cause. First, the request must be dealt with by data scientists in the MIDAS team as the data must be extracted from a complex database, requiring bespoke SQL queries to be written. It is a small team, having only 4 personnel at the time of the 2018 Request. While personnel are engaged in that activity for Mr Hendy, they cannot do it for APHA. Utilisation of APHA’s systems to produce the reports can also prevent or destabilise their use for its own work. We accept this, and find that the burden caused by Mr Hendy’s requests is greater than if they could be fulfilled by employing administrative staff without specialist qualifications or skills. For the purposes of s.14, it is also legitimate to consider the amount of time that was spent on the 2017 Request. They can, as Mr Akehurst attempted to do for the purposes of s.12, be looked at together.

Mr Hendy made either 31 or 32 requests (it does not matter which) between 2012 and 2019, and none since. It is accepted that around half of these relate to bovine TB. Nonetheless, it is not suggested that any of those were as complex or burdensome as the 2017 and 2018 Requests. Indeed, there is scant evidence from APHA on whether they considered the requests prior to 2016 to be especially burdensome. In the Decision Notice and the Commissioner’s Response it is recorded that APHA had claimed that a total of 468 working days had been spent on Mr Hendy’s requests.

In his rule 24 Reply to the Commissioner’s Response, Mr Hendy disputed the figure of 468 days. This would amount, he observes, to an average of 111 hours on each of his requests. At page 236 of the supplementary bundle he presents a bar chart showing the date of each of his requests, how long it took APHA to issue a response, and the outcome. It can be seen that the requests are evenly spread out over the seven year period and that APHA was able to comply with most of them within 20 days or so. A minority were rejected, mostly under s.12. Mr Hendy therefore argues that the 468 day figure cannot be correct – not only would APHA have been entitled to refuse far more requests under s.12, but it must also be questioned whether there is even room for so many days within their response times.

Ms White’s response to Mr Hendy’s analysis is found at paragraphs 40-41 of her witness statement. She says that Mr Hendy misinterpreted the 468 day figure as how long had been spent working on his requests whereas it was always intended to be the total time taken to respond, measured from the date of request to the date of response. The letter to the Commissioner dated 12 July 2019, setting out the 468 day figure, is provided. We accept Ms White’s evidence that she had not intended to mean that 468 days’ working time had actually been spent working on fulfilment of Mr Hendy’s requests. We do consider, however, that the Commissioner misunderstood her letter in the same way as Mr Hendy. As it is, the figure does not materially contribute to the issue of burden. We prefer the analysis conducted by Mr Hendy, as described in the above paragraph.

Our conclusion on burden is that we do not consider the overall number of requests to be manifestly excessive, given the period of time involved and that APHA were able to either promptly comply with the requests or notify Mr Hendy that to do so would exceed the cost limit. As suggested by Dransfield, we do take into account whether the volume and duration of requests is unreasonable in light of the past and anticipated future burden upon APHA, but there is no basis upon which to consider that Mr Hendy’s future requests will routinely replicate the complexity of the 2017 Request. Furthermore, Mr Hendy’s evidence is that he deliberately spaces out requests and considers their complexity before making them. We accept his evidence on this point alongside that concerning his motive in making the requests, as set out below.

While FOIA is both “motive blind” and “applicant blind”, a proper application of s.14 includes the underlying rationale or justification for the request and its connection with the body of requests as a whole. As discussed in Dransfield, it may be entirely proper for a journalist to find that a response opens up a new line of inquiry that then justifies a new request, and so on. In other cases, a request might be properly made about an issue of concern but then expand to cover a whole series of unrelated matters and grievances that are unconnected to the original aim – the so-called “drift”. Mr Hendy’s requests bear far more resemblance to the first example. We see no hint of any purpose to the 2017 and 2018 Requests beyond scrutiny of government policy concerning bovine TB. APHA has not suggested that Mr Hendy is deliberately setting out to inconvenience or disrupt its operations, or acting out of malice, nor could the evidence justify such a conclusion. We note in this respect Dr Wright’s evidence, discussed below, as well as a letter of support from Mr Roger Blowey, who is the ‘Specialist in Cattle Health and Production (dairy)’ at the Royal College of Veterinary Surgeons. It is very unlikely that either would support Mr Hendy if it was thought that he was acting out of any other motive than increasing understanding of bovine TB transmission.

As recognised in the judgment of the Upper Tribunal, and as now acknowledged by APHA, when making the 2018 Request Mr Hendy made a genuine attempt to narrow the 2017 Request. APHA continues to criticise Mr Hendy for not revising his request in the precise way that it had suggested, but puts forward no basis other than its now-rejected case on s.12.

As already noted under the heading of burden, several of Mr Hendy’s previous requests had been rejected as exceeding cost limits. He has not brought routine and meritless challenges against those decisions, and has limited his appeals to where he genuinely disagrees with APHA’s conclusions.

There is nothing concerning Mr Hendy’s motive in making the request, we find, that tends towards it being vexatious.

APHA accepts that bovine TB is an issue of public concern, but argues that the information requested makes no contribution to public understanding of the issue. It points to its routine online (Footnote: ⁵) publication of data that enables public scrutiny, including in conjunction with other organisations (Footnote: ⁶), as well as active collaboration with academic researchers in which more data is released. It enters into arrangements with universities that enable it to support research without (unlike FOIA compliance) affecting its operations. Mr Hendy does not dispute this, and we accept in principle that the wide range of information made available by APHA may reduce the likelihood of the burden of compliance with Mr Hendy’s request being proportionate.

Ms White and Mr Akehurst give evidence that APHA does not consider the requested information to be useful. Mr Akehurst’s statement contains the following:

Inevitably, the weight that can be attributed to this evidence is seriously reduced by the way in which it is given. If APHA wishes to rely on Dr Parry’s evidence then a witness statement ought to have been provided, or at least correspondence in her own words rather than as paraphrased by Mr Akehurst.

As well as the usefulness of the data itself, APHA has concerns over it being provided to Mr Hendy in particular. He is not a published academic, and APHA expresses concern that he may misinterpret the data he receives to draw false conclusions. We accept, hypothetically, that a request from a published academic is less likely be vexatious, but that principle should not be over-extended. The entitlement at s.1(1) of FOIA would be undermined if a public authority reserves academic-grade data to those it deems able to properly understand it. If release of particular information would prejudice the conduct of public affairs, then the corresponding exemption in FOIA can be argued. In this appeal, we place little weight on APHA’s concerns over Mr Hendy’s academic qualifications and ability.

Mr Hendy has adduced evidence in response to APHA’s case on the information’s value. We heard evidence from Dr Hazel Wright, Senior Policy Officer of the Farmer’s Union of Wales. In a letter, she states as follows:

Dr Wright’s evidence was subject to competent and comprehensive cross-examination by Ms Thelen. Dr Wright explained her academic qualifications and experience as a statistician. We accept that she is well placed to understand the potential use of data such as that requested by Mr Hendy. She gave a thorough description of the public data sources made available by APHA and DEFRA, and their benefits and shortcomings. The raw data publicly released by APHA is particularly useful to her farming policy work but omits information on herd numbers and movements. This is a point that is also made in Mr Blowey’s letter of support. The information would be useful to understand whether a decline in infection rates is a result of measures taken against bovine TB by DEFRA, or is actually attributable to a the decline in the number of herds. Given the significant consequences to the farming industry of anti-TB measures, there is a strong public interest in there being as full an evidential picture as possible. Contrary to Dr Parry’s concern that limiting data to Devon reduced its usefulness, Dr Wright considered regional data to be potentially useful because bovine TB spreads in different ways in different parts of the country. She accepted that the requested information would only provide a small part of the picture and could not justify drawing any conclusions on the effectiveness of government policy by itself, but considered that it would have the potential to indicate where future research might be warranted. Asked about misinterpretation of data, Dr Wright responded that this was a risk in any data analysis, but the risk of error was not increased simply because the underlying topic is controversial. Dr Wright expressed gratitude for the amount of work that APHA does with external researchers and for the information that it does make publicly available.

We considered Dr Wright to be an impressive witness, clearly qualified to give the evidence she did. She candidly accepted the limitations of the data, while giving a cogent explanation of how it nonetheless had the potential to contribute to further research on a topic of considerable importance to a section of the community. We also take into account Mr Blowey’s letter of support, which also suggests that there may be value in the requested data.

Considering all the evidence, we find that the requested information would be of value to Mr Hendy as a person interested in the research underpinning government policy concerning bovine TB, and to the public, given the effect of that policy on the farming industry. There is a strong public interest in disclosure.

It has not been suggested that Mr Hendy has engaged in any harassing or abusive behaviour. He has certainly been persistent, and on occasion may have taken a somewhat exasperated tone in his correspondence (for example, describing APHA’s position as “rubbish”), but he has remained civil. APHA say that his requests cause “distress” due to the disruption they cause.

Save for the lack of clarity in the 2017 Request observed by UTJ Mitchell, no real criticism can be made of Mr Hendy as to the way in which he has made his requests or their underlying motive. We have also found that there is a strong public interest in favour of disclosure. In this case, vexatiousness therefore really depends on whether the burden imposed by his requests is proportionate to the value of the information sought.

In Cabinet Office v ICO and Ashton [2018] UKUT 208, the Upper Tribunal considered the authorities and concluded as follows:

We carefully take a broad and holistic view in accordance with that guidance, taking into account all the circumstances we have set out above. The burden upon APHA’s resources imposed by Mr Hendy’s requests is certainly significant. Tempering that observation is the somewhat unusual nature of the 2017 and 2018 Requests. If they are taken together then, as argued by APHA, compliance is well above the s.12 cost limit. The 2018 Request only falls below the cost limit because of the abortive work that had already been carried out on the 2017 Request. There is no suggestion that such circumstances have arisen before, nor that they are likely to arise again. An entirely novel request in the future that requires a similar scale of work will quickly be identified as engaging s.12. Nor is Mr Hendy free to make unduly frequent or burdensome requests for similar information to that presently sought, for example for other regions or years, simply because APHA can now fulfil them with less than 24 hours’ work – the overall burden of such requests could well justify a finding of vexatiousness in the way described in Ashton.

In this appeal however, we find that the high standard of vexatiousness has not been reached. The public interest in disclosure of the requested information outweighs the resource implications of compliance.

As a final observation, we note Mr Hendy’s suggestion that the misunderstanding of the 2017 Request may be to blame for any distress felt by APHA’s personnel. We see no reason to expand on what has already been said by UTJ Mitchell concerning that misunderstanding, but Mr Hendy may have a point. It is important that his future requests are entirely unambiguous as to the data he seeks, proportionate in their scope and frequency, and capable of the same weighty justification as was provided in this appeal by Dr Wright. It also important that APHA does not let the burden they impose cloud its judgement on how best to respond to them. The way in which its case on s.12 was presented in this appeal may represent such a lapse of judgement. As stated in the Commissioner’s guidance:

A more conciliatory and constructive approach by both parties might well have avoided considerable expense of resources by APHA, and this is a feature to which the Commissioner and the Tribunal will no doubt be alive in any future dispute.