This post is a jointly written effort by myself and Don Keiller.
Readers may remember the Information Commissioner's ruling last year that UEA had to release the CRUTEM data sent by Phil Jones to Peter Webster at Georgia Tech. This had been requested by Jonathan Jones and Don Keiller.
This ruling was obviously very welcome, but in fact it was not the end of the story. UEA had put forward an argument that CRUTEM data was held under agreements with national meteorological services and could not therefore be disclosed to outsiders. Along with his request for the data, Keiller had therefore also requested the covering email that Phil Jones had sent to Webster, which should presumably contain caveats about reuse and disclosure. However, when the Information Commissioner ordered UEA to release the data, UEA's non-disclosure of the email was upheld, on the grounds that the information was, on the balance of probablilities, 'not held'.
However, by the time of the ICO's ruling, it was clear that CRU had its own backup arrangements - the CRUBACK3 server that was at the centre of the Climategate affair - and it was therefore fairly clear that the email did still exist. UEA were in essence trying to argue that since it was no longer on Phil Jones' hard drive, it was no longer legally 'held' for the purposes of the Environmental Information Regulations (EIR), regardless of its presence on the backup server. Keiller therefore decided to appeal the ICO's decision to the Information Tribunal. Firstly he had to provide new prima-facie evidence to support the appeal to an internal ICO Appeals Panel, who decide whether there are grounds for an appeal.
With the Appeals Panel having decided in Keiller’s favour the appeal was then considered by the “First-Tier Tribunal (Information Rights)”. Normally such appeals are dealt with by an exchange of written correspondence, but UEA decided that they wanted to call a witness, necessitating a full court hearing. This was no small step, since judges and barristers are involved and in the event of an unsuccessful appeal it is possible, in theory at least, to end up with a large bill for costs.
This was not the only obstacle placed in the way of the appeal. Parties are required to agree on an “open bundle” of documents to be provided as evidence for the Tribunal to consider. Here Dr. Keiller asked for sections of the Muir Russell Report and the minutes of the House of Commons Science and Technology Committee, which examined emails “leaked” from CRU. He also asked to include the letter to UEA from the Deputy Commissioner of the ICO which famously stated “The prima facie evidence from the published emails indicate an attempt to defeat disclosure by deleting information.”
All these publicly available documents were objected to by solicitors acting for UEA, on the grounds that they were “prejudicial”. In the end Keiller had to write directly to the Tribunal to get them admitted as evidence. This evidence proved crucial.
As the hearing approached it became increasingly clear that Keiller would be contesting not only UEA and the ICO, but a team of solicitors from a nationwide firm and a top London barrister. Help was needed. He contacted me for help, and I pointed him in the direction of David Holland, well known to readers of this blog, and something of an expert in such appeals. David kindly agreed to be a "MacKenzie friend", which allowed him to assist during the hearing and his help, before and during the hearing was invaluable. The hearing was held just before Christmas, with Keiller representing himself and UEA represented by a barrister.
On arriving in the courtroom, solicitors acting for UEA presented Keiller and Holland with a folder containing 384 pages of documents. Also present were the ICO’s representative, counsel for the ICO, counsel for UEA and senior management figures from UEA.
It is often said that a plan of battle fails on first contact with the enemy and so it proved here. On the basis of documents and arguments exchanged before the hearing, Keiller, with Holland’s help, had prepared a set of arguments and questions. However, the Judge started by saying that this was a complex case and that he wanted to keep focus by asking three specific questions. These questions were sequential, each dependent on success of the previous one; Keiller had to demonstrate that each question could, on the balance of probabilities, be answered in the affirmative. Failure on any one of them would mean failure of the appeal.
No pressure then.
The tribunal looked at three questions:
- Is it more probable than not that the email sent on or about 15 January 2009 by Professor Jones to Georgia Tech attaching datasets was backed up onto and retained on the Climate Research Unit’s (CRU’s) back-up server prior to this server being taken by the Police?
- Is it more probable than not that the e-mail contained ‘any instructions or stipulations accompanying the sending of datasets’?
- Is there a valid argument that a back-up of an e-mail retained after the original had been deleted from the computer on which it was composed is not ‘held’ for the purposes of the EIR?
The hearing started with Jonathan Colam-French (Director of Information System at UEA) providing evidence about the procedures in place to record, manage and dispose of staff emails. Unfortunately for UEA, the systems described by Mr Colam-French were not in place at the time Jones sent his email to Georgia Tech, a fact uncovered by the otherwise ineffective Muir Russell Report.
In their decision the Tribunal made some pointed remarks about UEA’s evidence, stating that:
The Tribunal were rather disconcerted by the evidence adduced by the UEA on [Question 1]. Jonathan Colam-French had almost no knowledge of the CRU’s back-up system and was simply unable to answer several pertinent questions...
and
...we noted the complete lack of evidence about anything resembling a coherent deletion/retention policy for emails.
On this basis, the tribunal reached their first ruling, namely that the email was more than probably still on the backup server.
On the second question - about the contents of the covering email - the tribunal was hampered by the fact that they were not able to address Phil Jones directly. Here, counsel for UEA reported on an alleged conversation between Jones and David Palmer (UEA Information Policy & Compliance Manager), which had in turn been relayed to Colam-French, who had in turn reported it to counsel for UEA. Dr. Keiller pointed out that this was “third-hand hearsay” and questioned why Jones was not present to provide this evidence directly or, indeed, why he had not provided the Tribunal with a signed witness statement, adding that UEA’s case was fatally flawed by the absence of Jones. The Judge agreed, saying that counsel for UEA were making “bold assertions”, but these were “unsupported by evidence”. It appears that UEA were keen that Jones should not appear on the witness stand, where he would be required to give evidence under oath. In fact it is noteworthy that, despite all the official “investigations”, Jones has never been required to answer questions under oath or provide a signed declaration.
In reaching their decision, the Tribunal’s starting point was that a covering e-mail was a rather obvious place to set out any caveats in relation to the attached datasets. The Tribunal also took into account that they heard no evidence as to what the relevant email did contain other than the reported assertion that Jones (who did not himself provide any evidence) had said that it didn’t contain anything of the sort. So although the Tribunal took into account the assertion that Jones had said that the caveats were discussed verbally only, they ultimately decided that it was more probable than not that the covering email did contain ‘instructions or stipulations’ relating to the attached datasets.
Two rulings issued, two wins to the appellant.
Finally then, to question of whether information on backup servers is 'held' for the purposes of EIR. Here UEA argued that, since Jones had deleted the email from his hard drive as part of his “normal email management procedures”, there was no requirement for them to search further. However, the tribunal were again unimpressed, stating in their decision:
“Whilst we can see some logic to this position, we noted that the purpose of back-up is precisely to ensure that a document is not lost; the lack of any coherent policy on retention and deletion of documents, and that had there been timeframes in such a policy, we would have expected these to be reflected in the back-up programs operated on the server. In these circumstances, it seemed more logical to us to take the view that if the email existed, it was still ‘held’ by UEA”.
A clean sweep then.
UEA are now required to approach the Police to see if they will provide a copy of the email in question, and to see if they or someone else will extract it for them. (As another aside, we might note that copies of Jones' emails and indeed those of Keith Briffa and Tim Osborn have already been provided to the Russell inquiry. I wonder what happened to those copies of the emails or if UEA also received a copy?).
So, the key outcome of this case is that a further precedent has been set. That precedent is that UEA’s refusal to release information on the grounds that under EIR Reg. 12(4)(a) – Information not held, is no longer valid. Backup servers have to be searched.
As the previous ICO ruling, which resulted in the release of the CRUTEM data, has already demonstrated that refusal under:
Reg. 12(4)(b) – Request is manifestly unreasonable
Reg. 12(5)(a) – Adverse effect on international relations
Reg. 12(5)(f) – Adverse effect on the person providing information
are not valid, it would appear that UEA have no legal means of refusing properly crafted information requests.
The full decision of the Tribunal is available (here)