On Tuesday 22 September 2020, Flaux LJ (Chair of the Disclosure Working Group) published an update on the operation of the Disclosure Pilot Scheme (DPS). It was accompanied by the long-awaited publication of the Third Interim Report on the DPS, dated 25 February 2020 and prepared by Professor Rachael Mulheron (of Queen Mary University of London), who has been monitoring the DPS from its start in January 2019. The update also outlined a number of proposed amendments to the DPS, which are expected to be considered by the Civil Procedure Rule Committee at its October 2020 meeting.
The DPS was introduced to address widespread concerns at the perceived excessive costs, scale and complexity of disclosure. Put simply, its aim was to make the disclosure process simpler and cheaper. But it seems those aims are not being achieved. Professor Mulheron’s report suggests that, in the view of court users, disclosure under the DPS has become even more complex and expensive.
Will the proposed amendments put things back on track? Well, they are certainly helpful in ironing out some areas where the DPS’s rules are currently unclear or impractical. In my view, however, they don’t go far enough in addressing the issues of complexity and cost.
Third Interim Report
Professor Mulheron’s report analyses the 71 responses received to a questionnaire that practitioners were asked to complete back in the autumn of 2019. It does not make comfortable reading for proponents of the DPS, so far as respondents’ views on overall outcomes are concerned:
- 85% say the DPS has not saved costs overall (4% say it has, 10% can’t say (or too early)).
- 42% say it has made disclosure less accurate (16% say more accurate, 42% can’t say (or too early)).
- 71% say it has increased burdens on the courts (2% say decreased burdens, 27% can’t say (or too early)).
- 78% say it has not brought about a culture change (6% say it has, 16% can’t say (or too early)).
Of course, it’s now almost a year since the survey underlying these statistics was carried out and it’s possible that, as the DPS has bedded down, the views of court users have become more positive. But anecdotal evidence suggests that significant concerns remain among practitioners in both branches of the profession.
Professor Mulheron’s note accompanying publication of the report states that there will be a further opportunity to give feedback as the DPS enters its third year, following the recent decision to extend it to the end of 2021. There will also be a survey of the judiciary to ascertain their views and experiences of the DPS. It will be interesting to see whether this further feedback is consistent with the views reflected in the report.
Proposed amendments to the DPS rules
Flaux LJ’s update includes a proposed revised version of the DPS rules at Practice Direction (PD) 51U, together with a revised version of the Disclosure Review Document (DRD). Some of the most important changes are considered below.
Known adverse documents
There has been some confusion, since the start of the DPS, as to when the obligation on the parties to disclose “known adverse documents” takes effect. The obligation could be read as arising when a party gives Initial Disclosure (that is, when serving its statements of case) or even, on the most literal reading, as soon as proceedings are issued. The proposed amendments clarify, helpfully, that the obligation does not kick in until the later stage at which any Extended Disclosure is provided.
The obligation to send document preservation notices to “all relevant employees and former employees” has also caused some concerns, particularly as the obligation is not, on its face, qualified by reference to reasonableness. The proposed amendments clarify, again helpfully, that the obligation applies only where there are reasonable grounds to believe that the relevant person has disclosable documents that are not already in the possession of the party to the litigation.
“Model C” disclosure
Model C, or request-led search-based disclosure, is described at PD 51U as “disclosure of particular documents or narrow classes of documents relating to a particular Issue for Disclosure”, by reference to requests set out in the DRD or otherwise defined by the court. This was initially thought to be similar to the approach often adopted in international arbitration, with (potentially quite broad) requests set out in “Redfern Schedules”. On that basis, however, there was some confusion as to the proper boundary between Model C and Model D or “narrow search-based” disclosure (similar to old-style “standard disclosure”).
In Pipia v BG Group Ltd, the High Court expressed the preliminary view that Model C requests should be defined without reference to any criterion of relevance to the issues in the case, so that a request for documents falling within set parameters and “relating to” a specified issue would not be a competent Model C request at all. The proposed amendments confirm that approach, meaning that Model C will be appropriate only where a narrow class of documents can be defined without reference to relevance (for example, invoices between x and y dates).
The DRD is a new creation under the DPS, replacing the (optional) Electronic Documents Questionnaire under CPR 31. As well as setting out the proposed List of Issues for Disclosure and the disclosure model proposed for each, it contains detailed information about how the parties’ documents are stored and how they might be searched and reviewed. The DRD came in for much criticism from respondents to Professor Mulheron’s questionnaire, and the proposed amendments seek to make the document simpler and more flexible as well as dispensing with the requirement to complete it (including the proposed List of Issues for Disclosure) where the parties are not proposing any search-based disclosure models.
Do the proposals go far enough?
While the proposals are welcome, and should serve to clarify a number of aspects of the DPS that have caused confusion or debate, in my view they do not address the fundamental difficulty with the DPS. That is the huge complexity introduced by the requirement to prepare, and seek to agree, a List of Issues for Disclosure and then to map each issue to the appropriate one of five disclosure models.
This can be enormously time consuming, particularly in complex multi-party litigation, and it does not appear that the significant increase in time spent, and therefore cost, is justified by the benefits of the DPS. That point is underlined by the fact that, in many cases, at the end of this process, the parties end up with Model D disclosure (in effect, standard disclosure) for most issues which need to be determined by reference to documents.
A potential way forward
Against that background, I wonder whether there might be a compromise solution, which would seek to achieve the aims of the DPS, while avoiding much of its complexity by dispensing with the need for complex matrices matching disclosure models to issues. I have in mind a system where the parties would seek to agree:
- A list of the issues by reference to which some or all parties (and if only some, which parties) will give disclosure according to the “standard disclosure” test, that is, those documents which (genuinely) support or adversely affect any party’s case on that issue (equivalent to Model D, without so-called “narrative documents” or documents relevant only to background).
- Any specific additional documents or narrow categories of documents that any party will disclose (equivalent to Model C as recently clarified).
Otherwise, a party would need to disclose only the documents on which they rely and any known adverse documents (essentially equivalent to Model B under the pilot) unless otherwise agreed or directed. For example, in a fraud case, exceptionally, Model E or “train of enquiry” disclosure might be appropriate on some issues.
The upshot is that parties would only have to identify and seek to agree disclosure issues to the extent that substantive, search-based disclosure is proposed. For all other issues, the process would default to documents relied on and known adverse documents, without the need to agree a list of those issues, which would in any event be apparent from the parties’ statements of case (and in the Commercial Court at least, the main list of issues in the case). For the issues on which there is to be search-based disclosure, the emphasis could then be on identifying suitably narrow parameters for the search to be conducted (that is, by reference to custodians, date ranges, and key words), and of course adopting technology-assisted review in appropriate cases.
This is not, of course, a complete solution to the problem of excessive costs associated with disclosure. It would still be incumbent on the court, and the parties, to ensure that the list of disclosure issues is kept under control, that search parameters are tightly defined, and that technology is used appropriately. However, it would strip out some of the unnecessary complexity which plagues the DPS in its current form, while still ensuring that disclosure is focused on the issues where it is really needed. It seems to me that it is worth considering.