Rules of English civil procedure have long required parties in legal disputes to disclose all documents potentially relevant to the case being tried. As the volume of electronic documents and communications produced by parties continues to increase at an alarming rate, lawyers are starting to find their clients’ vast volumes of electronically stored information (ESI), increasingly difficult to manage.
In Gavin Goodale v The Ministry of Justice [2010] EWHC 3834 (QB) Senior Master Whitaker published a draft ESI Disclosure Questionnaire. The questionnaire is a document which has been prepared to guide parties in their approach to disclosure. It is currently being considered by the Civil Procedure Rules committee and, if and when it is introduced, solicitors will be required to fill it in to describe their clients’ data, and what they propose to do with it, when faced with a disclosure requirement.
The questionnaire is expected to be officially introduced into practice by a new Practice Direction on disclosure in October this year. It has not yet been confirmed if, and for what types of cases, the questionnaire will be compulsory. It is clear, however, that the courts will be able to exercise their discretion in managing cases by ordering lawyers to fill in the questionnaire where doing so would help make the case run more efficiently.
The aim of the questionnaire is to get opposing parties to confer on what measures are appropriate, and what is reasonable, to get to the information that really matters. But will lawyers possess the technical knowhow to fill in the questionnaire, or guide their clients to do so? I take a broad look at the questions asked, and suggest the best way to answer them whilst maintaining an advantage in the litigation at hand. This article is based on the draft questionnaire as appended to the Goodale case, so the questions may change by the time questionnaire is officially released.
Extent of Reasonable Search
Questions 1 to 5 of the questionnaire deal with the extent of the proposed search for potentially relevant data. Although the heading is ‘Extent of a Reasonable Search,’ there is no scope in this section for arguing that some search parameters might be unreasonable. The section merely requires the party completing the questionnaire to state the proposed date range, custodians, communication types, document types, and databases, without regard for what might be reasonable.
The first question covers date range, which should be reasonably straightforward to answer with reference to the pleadings and knowledge of how the business in question operates. The second question asks for a list of the custodians a party considers to hold the relevant ESI. This sounds straightforward, but the footnote further instructs the party to ‘include names of all those who may have or have had custody of disclosable documents, including secretaries, personal assistants, former employees and/or former participants.’ Careful consideration, not only on the lawyers’ part but probably also from someone with insight into the commercial workings of the business in question, may need to be given to accurately list everyone.
The next question requires a party to create a tabular list of all forms of communication which may contain something of relevance. It suggests e-mail as the primary form of communication, and asks for any others, including instant messaging, voicemail, voice-over-internet protocol, recorded telephone calls, text messages, audio files, and video files. For each type listed, the writer is requested to say whether it is used within the aforementioned date range, whether such documents will be searched for, what type of software, hardware and media the data are stored on, and what the backup situation is.
The most notable point is that the wording doesn’t seem to ask for only those items which the writer proposes to search, but for absolutely all forms of communication available, and a yes/no answer as to whether each one will be searched.
The fourth question requires a similar tabular listing of the formats of electronic documents other than attachments to emails which were created or stored during the relevant time period, with the same sort of questions about whether they will be searched and whether there are any backups. The fifth similarly looks at databases.
An effective way of approaching these types of questions is in a round-table discussion with lawyers, their client’s in-house counsel, commercial managers and IT representatives, and ideally legal technology consultants. During such a discussion, the lawyers can work through data sources, explaining the legal need to disclose each, for example, what evidence is required to prove the case and what must be produced to comply with disclosure obligations. The commercial managers can point out who has what sort of information, and the IT staff can state where that is likely to be stored and help with mapping the data landscape. Legal technology consultants can provide information about how best to extract and preserve such data, and how to deal with it in an efficient and proportional manner.
Method of Search
Questions 6 to 9 ask about proposed methods to help decide which documents to disclose from the list above. Keyword searches, which have proven a popular way of culling data, are suggested first. The party completing the form is asked to say whether keyword searching will be used, and if so, to specify the keywords.
Keyword searching sometimes has a tendency to produce false hits – just because a document contains the word you are looking for, doesn’t mean it is about the subject matter you are interested in. Therefore it is no surprise that a supplemental review of documents hit by keywords is usually necessary. Interestingly, the questionnaire specifically asks to what extent the keyword searching will be supplemented by a review of documents.
Although it is common practice to review documents before disclosure, typically parties agree keywords and leave the review to be carried out using their own preferred method without conferring. The questionnaire seems to require parties to put their review strategy forward from the outset.
The seventh question asks about ‘other types of automated searches,’ and gives clustering and concept searching as examples. Technologies such as these are very valuable for prioritising and speeding up document reviews after the initial data filtering process, but are not typically used in an automated search to decide which documents should and shouldn’t be looked at. The benefit of keyword searching is that the results are black and white – a document either contains a certain word or it does not. Results are reproducible and, with some good sampling, also predictable. The results of concept searching and clustering depend on what software and algorithms are being used, and the way in which they work is difficult to define in an exact manner. If someone proposes culling documents by use of concept searching, it would be reasonable to ask what classes of documents will be included and excluded on this basis, and what parameters are being used to define the concepts. As there is no standard and platform-independent language to describe concept search settings, it would be difficult to agree to use such technology to cull supposedly irrelevant data.
That being said, there is recognition that keyword searching is a blunt instrument and that new technologies need to be embraced to reduce the burden of reviewing large volumes of irrelevant documents. This is an area where lawyers and technical experts will need to work together to develop effective but defensible approaches and provide technical evidence to justify the approach taken.
This section goes on further to clarify whether the searching methodology adopted will capture attachments, compressed files, imaged text and other potentially difficult to search forms of data, and what kind of software is going to be used for processing and de-duplication. A competent e-disclosure software vendor should be able to easily answer these questions in relation to their processing platform.
Potential Search and Accessibility Problems
Up until this point, it seems that the questions have to be answered for all data, without any room to argue that certain methods or sources would not be proportionate. Only at question 10 does the questionnaire ask whether ‘any of the sources and/or documents identified in this Questionnaire raise questions as to the reasonableness of the search which ought to be taken into account?’ The questionnaire also asks about hard-to-access data such as encrypted files.
It is easy to imagine the disputes that might arise as a result of framing the questions this way. For example, a party might identify a vast array of sources of data, but then limit the search to a much smaller subset on the basis of proportionality. The party who receives the questionnaire then argues that a larger proportion of the data universe is potentially disclosable, with it being reasonable and proportionate to do so.
This is arguably a good thing – as it will make parties come to some agreement (or settle it in court!) about which approach is appropriate, without the risk of having already done it and being ordered to repeat the process properly, as happened in Digicel (St Lucia) Ltd & Others v Cable & Wireless Plc & Others [2008] EWHC 2522 (Ch). It also ensures transparency in the disclosure process and flushes out contentious issues early on in the proceedings. On the other hand, it is possible that arguing about such matters early in the litigation – before disclosure has begun – will increase the costs significantly.
How best to argue, then, what is disproportionate and what is just too hard to deal with? In terms of arguments about processing tricky document types (such as password protected, encrypted or deleted documents) or restoring backups, the best way to approach it may be to be readily armed with a quote from an e-disclosure provider. This can set out how much work will be involved and how much it will cost to allow parties to assess the reasonableness of the proposed solution.
In relation to arguments about what types of keyword searches should be applied, the most justifiable way to argue that a search parameter is disproportionate, or just ineffective, is to load a good sample set of data into a review platform or early case assessment tool, run the proposed keywords or other filtering parameters and provide a list of the results. Often it is immediately clear that some keywords return so many results that they are clearly too wide, and that some are completely superfluous because other terms bring up the same data. This sort of analysis helps you adopt a more scientific approach to searching and allows you to have sensible discussions with other parties about the keywords that should be used to select disclosure documents.
Preservation of ESI
Legal hold is a concept that is becoming better known in the English jurisdiction, and the questionnaire seems to promote acting quickly to preserve potentially relevant data. Question 14 asks ‘have you given an instruction to preserve ESI, and if so, when?’ The answer may well be no, but at least it gives the other side a chance to demand preservation. In England and Wales, at minimum, the duty to preserve arises when litigation has commenced (see Rockwell Machine v Barrus [1968] 1 WLR 693). It was, however, noted in Earles v Barclays Bank [2009] EWHC 2500 (Mercantile) that ‘there might be cases where it was appropriate to draw adverse inferences from a party’s conduct before the commencement of proceedings.’ In practice, an organisation would therefore do well to treat the obligation as arising as soon as it is reasonably believed that a dispute (and therefore litigation) might arise.
The other question asks whether there is a document retention policy. It does not seem to ask for any detail, but hopefully lawyers will be sure to advise clients that, where there is a document retention policy, it should be reviewed to ensure that any information subject to the legal hold is not routinely destroyed.
Inspection
There is nothing new about the idea of agreeing with the other party as to what the disclosure format will be. Although it may be helpful to decide on the disclosure format early, agreeing it before seeing all the documents could be difficult, when one cannot fully anticipate the format and content of the documents. For example, it may be agreed to exchange TIFF images of all documents together with an index of metadata fields and OCR text. But complex spreadsheets – or other file types which are hard to read in printed format – may be better presented in native formats. But what if these require redactions? There needs to be room to move from the agreed disclosure format depending on the documents found. Lawyers should talk to their technology consultants to familiarise themselves with the strengths and weaknesses of various disclosure formats and associated settings – before committing to a disclosure format.
Disclosure and Other Parties
This section of the questionnaire allows for the expression of opinions on what the other party (or parties) should be doing in relation to their ESI, if any such opinion has been formed at this stage. This may be a difficult set of questions to answer without giving away one’s case strategy and without having seen the other party’s answers to the previous questions. Nevertheless, it provides an opportunity to say what you want or expect from the other side, at an early stage.
There is no doubt that further argument is likely upon receiving the other side’s answers to the questionnaire.
Statement of Truth
The completer of the questionnaire is required to declare that he or she believes that the facts stated in it are true. A statement of truth is of course not a new concept, but it is interesting on a form such as this one, where it is unlikely that any one person will know the answers to all of these highly technical questions. On this basis, parties filling out the questionnaire will need to be careful about what they claim.
Learning the Ropes
The completion of the ESI questionnaire needs to be a collaborative process, involving lawyer, client and technology consultant, whether internal or contracted or both. A prior understanding of the technology available to support disclosure and the best way to use it will give lawyers the basis to efficiently carry out exercises such as answering this questionnaire, and to manage their client’s case generally. As Lord Justice Jackson pointed out in his report, e-disclosure should form a substantial part of legal education for solicitors, barristers and members of the bench. The better prepared lawyers are to answer questions such as these, the less confusion will occur – and the smoother the disclosure process will be. Whatever its current status the questionnaire acts as a useful checklist of the issues parties should consider when approaching eDisclosure. As a matter of best practice, parties that consider the issues raised in it at an early stage are better placed to reduce the costs and risk associated with e-disclosure.
Daniel Kavan is a legal technology consultant at Kroll Ontrack in London: www.krollontrack.co.uk
