E-disclosure: Numbers Too!

July 2, 2009

The Preliminary Report by Lord Justice Jackson on the Review of Civil Litigation Costs[1] addresses, inter alia, the costs of disclosure and, in particular, e-disclosure.  Jackson LJ acknowledges that, while ‘Lord Woolf’s reforms were aimed at limiting the scope and consequently the costs of disclosure…, if anything, this cost centre has spiralled over the last ten years’[2]Many of those involved in commercial litigation believe that ‘E-disclosure is… the single biggest challenge…’[3].

Jackson LJ considers that ‘[t]he obligation [for the disclosing party] to search for and identify disclosable documents remains ostensibly the same as the pre-1999 procedure… [and] the current test does not reduce the extent of the documents which the solicitors on each side are initially required to review’[4].  Since ‘[c]ommunication has become much easier and cheaper…[it] is therefore far more extensive [and] [e]lectronic communications are more readily preserved and retrieved, even after deletion’[5]. Consequently, the number of documents required to be reviewed by the disclosing party’s solicitors has increased significantly. 

Potential Costs of E-disclosure

Jackson LJ was provided by ‘…a City firm of solicitors, who are regularly engaged in heavy commercial litigation…’[6], with estimated typical costs of processing and hosting electronic documents and of junior lawyers reviewing those documents and identifying those that should be disclosed. 

The firm provides an example of a ‘huge’ case involving (after ‘de-duplication’) 75 million documents, which is apparently of an ‘…order of magnitude [that has] been experienced by the firm in question[7].  It estimates, assuming that on average each lawyer reviews 50 documents per hour (or one document every 1.2 minutes) and an average hourly charge out rate of £200, that if all of these documents were manually reviewed (referred to as a ‘linear review’) by a team of 50 lawyers it would cost £300 million and take 14 years[8]!   The estimated costs of e-discovery processing and hosting are, comparatively, very much smaller, being £2.2 million and £1.7 million respectively[9] (although the latter, seemingly inconsistent with the assumed linear review time of 14 years, assumes only eight months of hosting would be required).

Jackson LJ acknowledges that this calculation is ‘…somewhat theoretical…’[10] and ‘[t]his cost would in practice be considerably reduced by the use of electronic review tools… such as concept searches…’[11].  The firm does not give any indication of the typical volume reductions usually experienced as a result of using such tools.  If the reduction were 90%, i.e. 7.5 (10% x 75) million documents, the linear review would still cost £30 (10% x 300) million and take 1.4 (10% x 14) years.  The potential quantum of loss would have to be very large indeed for such costs and time to be considered reasonable[12] and/or proportionate[13].

The firm also does not give an indication of the typical number of documents that would ultimately be disclosed.  If 10% of the documents subject to the linear review, i.e. 750,000 (10% x 7.5 million) documents, are disclosed, further significant costs will be incurred by both the disclosing and non-disclosing parties’ senior lawyers, counsel, witnesses and experts carefully reviewing and considering those documents.  As a result of the higher average charge out rates of senior lawyers and experts and the longer time required for considering each document, these costs are likely to also be in the order of tens of millions of pounds. 

In addition, it is worth noting that this process is frequently much more costly for the non-disclosing party because of the unhelpful manner in which the e-disclosure is often presented.  Jackson LJ considers that e-disclosure ‘…processes do not necessarily lead a party to the relevant or key documents and… it can be easy for an unco-operative litigant to bury relevant material in a mass of material which is too difficult or too expensive to search effectively’[14]Jackson LJ also notes that ‘[t]he costs incurred as a result of the disclosure process are not limited to those incurred in the initial review by the disclosing party and a review by the other side.  The number of documents disclosed has a consequential effect on the rest of the process’[15]

The firm also provides a no less extraordinary example of a ‘medium’ case involving (after de-duplication) 7.5 million documents[16], being 10% of the volume of documents in the huge case example.  The estimated costs are correspondingly 10% of those for the huge case, i.e. £30 million for a linear review (before taking into account the use of electronic review tools).  The estimated time for a linear review of approximately seven years [17] is only half, rather than 10%, of that for the huge case because the firm assumes a mere 10 lawyers, rather than 50, would be required.

Minimising Costs

Other than the development of ever more sophisticated electronic reviewing tools, there are a number of immediate, practical and complimentary approaches that can be taken to minimise the spiralling costs of e-disclosure, including the following:

·               reducing the number of documents ‘captured’ for processing, hosting and initial review;

·               reducing the number of documents disclosed; and

·               ensuring that the documents that are disclosed are disclosed in a helpful manner.

The first two approaches are likely significantly to reduce the costs of the disclosing party and the last approach, while it may marginally increase the disclosing party’s costs, is likely to reduce the costs of the non-disclosing party significantly. 

From the perspective of an accountant, this article identifies problems frequently encountered with reviewing electronic documents and/or e-disclosure, particularly in relation to documents that are likely to be relevant to the assessment of quantum, and sets out practical approaches to minimising costs. 

Numbers as well as Words

Expert accountants are most often engaged in commercial disputes to give an opinion on quantum.  In a typical ‘lost profits’ claim, the claimant seeks compensation for damage to its business, allegedly as a result of the defendant’s actions, and most of the relevant documents for determining quantum will be held by the claimant.  Consequently, access to the claimant’s financial records will be essential to any assessment of loss. 

Unfortunately, accounting experts are frequently either not engaged by the claimant at an early stage or do not have any significant involvement in the initial ‘standard’ disclosure process.  This is, at least in part, because litigation lawyers often focus in the first instance on liability issues; without which, of course, their client would have no claim.  Furthermore, there is no requirement for a claimant to issue an expert’s report on quantum prior to or at the same time as either the Particulars of Claim or initial disclosure. 

Consequently, a claimant’s initial disclosure is most often driven by the alleged liability issues and mainly involves input from the claimant’s internal and external lawyers, IT department and e-disclosure expert.  Most of the documents disclosed are likely to be emails and other ‘word-intensive’ records, such as letters, memoranda and presentations.  In many disputes these documents are likely to be disclosed electronically and can effectively and efficiently be identified and reviewed by both parties using electronic tools such as ‘key word’ or ‘concept’ searches and ‘clustering’.    

Any financial records that are disclosed within the claimant’s initial disclosure are likely to be ‘number-intensive’ documents, such as spreadsheets or databases, which cannot be as effectively or efficiently identified and reviewed using the same electronic tools.  Furthermore, the accounting experts generally require complete sets of financial documents created over the relevant period of the claim; for example, the claimant’s monthly management accounts.  

In the old days of paper disclosure (or ‘discovery’), final versions (drafts being routinely destroyed) of the relevant monthly management accounts typically would have been sent to the defendant collated in ring binders, placed in cardboard ‘bankers’ boxes and clearly labelled ‘Monthly management accounts of [company x] from [date] to [date]’.  There should be no reason why the equivalent collation or ordering of documents, particularly financial records or other routinely produced documents, cannot occur in the new e-disclosure world.  Jackson LJ notes that ‘[i]t has been emphasised by practitioners that it is vital to preserve the context of a document, for example by preserving the folder structure of the data, or the file hierarchy…, rather than disclosing documents separately and not within their original folders.  The paper equivalent would be to disclose files of papers which are labelled by subject or by person (showing their place in the relevant transactions), rather than to disclose an enormous box of papers in unlabelled files with, for example, meeting minutes scattered throughout’[18]. 


Whereas the number of emails and volume of electronic data have increased exponentially over last few decades, the number of monthly management accounts has not; there being still only 12 months in a year.  It does not seem too onerous to require parties to disclose sets of key financial or other documents in separate electronic ‘folders’.  At present, disclosing parties frequently seem to be using e-disclosure as an excuse or cover for failing to do this.  Alternatively, such documents should be disclosed with sufficiently detailed, consistent and accurate descriptions so that they can readily be identified and collated.  In addition, final versions should be readily distinguishable from drafts; the latter usually being relevant in only limited circumstances. 

This requirement should not only apply to monthly management accounts but also to other routinely produced financial records (for example, budgets/ forecasts/ plans, variance analyses, stock reports, fixed asset registers, cash books, accounts receivable/ payable ledgers and invoices) and other business records (for example, minutes of board or other management meetings, together with any related papers/ presentations, routine operations/ marketing/ sales reports and regular project reports)[19] .

Identifying Sets of Documents

Unfortunately, none of the current disclosure rules, guidelines or protocols appears to specifically require or recommend that the disclosing party either:

·               collate sets of such documents; or

·               include an index or mechanism that enables sets of such documents to be easily identified. 

The Practice Direction to Part 31 of the CPR merely states that: ‘Where there is a large number of documents all falling into a particular category the disclosing party may [emphasis added] list those documents as a category rather than individually e.g. 50 bank statements relating to account number__at__ Bank, [dates]; or, 35 letters passing between__and__ between [dates]’[20].

Part 31 does, however, require that each party produce a ‘list’ (essentially an index) of all documents disclosed, whether in paper or electronic form, and the Practice Direction states that in order to comply ‘…it will normally be necessary to list… the documents in date order, to number them consecutively and to give each a concise description (e.g. letter, claimant to defendant)’[21].  Part 31 also states that the list ‘…must identify the documents in a convenient order and manner and as concisely as possible’[22]In the context of e-disclosure and particularly in relation to quantum, claimants appear frequently to be paying lip service to this requirement.  Jackson LJ considers that ‘[i]f a party compiles its list thoroughly, it will save the other parties time and costs.  Conversely, if a party compiles the list inadequately it is likely to cause the other parties to incur additional costs working out what documents the list actually contains’[23]. 


If the volume of documents is large (i.e. in the 10s or 100s of thousands), the ability to sort the electronic documents by, for example, date will provide little assistance to the expert accountant attempting to identify sets of financial documents (such as the monthly management accounts) because each relevant time period (for example, month) is likely to contain a very large number of other documents.  In theory, the description field could be used to identify sets of such documents but, unfortunately, the descriptions given in a typical e-disclosure list usually provide limited or no assistance because they are:

·               based on the document’s original electronic file name, which is often unhelpful because the creator of the document had no need to make it uniquely identifiable from all of the disclosing party’s other records (for example: [‘March 2005.xls’]);

·               automatically (i.e. electronically) generated and usually either meaningless or non-existent when derived from a spreadsheet/ database file; or

·               manually (by the disclosing party’s lawyers) generated and frequently so generic (such as ‘spreadsheet’, ‘report’ or ‘table’) as to be of little use.

As a result, if any financial documents are disclosed by the claimant, it is likely to be extremely time consuming, if not impossible, for the accounting experts to identify sets of such documents within the claimant’s electronic disclosure.  It is equally challenging to determine if a disclosed financial document is the final, as opposed to a draft version.  As Jackson LJ notes, ‘[p]arties often provide multiple copies of one document.  This is time saving for the providing party, but adds substantial cost for the receiving party as they may need to undertake a thorough review of each document to determine whether (a) it is simply a duplicate document or (b) it is different in a way that is material to the case’[24].

These deficiencies could be avoided if claimants and their lawyers focused more on quantum prior to initial disclosure, engaged an accounting expert at an early stage and discussed the location of relevant financial records with their finance personnel, rather than with just their IT department and e-disclosure expert.  This should ensure that complete sets of relevant final financial records are disclosed and allow the accounting experts to start their work without needing to get into the time consuming, and hence costly, ‘needles and haystacks’[25] exercise to determine whether or not complete sets of relevant financial records have been disclosed.


Lawyers and e-discovery professionals have devised, and are continuing to devise, ingenious solutions for effectively and efficiently disclosing and reviewing the vast numbers of emails and other word-intensive documents that now need to be disclosed by parties involved in commercial civil litigation.


However, most number-intensive documents (for example, monthly management accounts) and some word-intensive documents (for example, board minutes) are not amenable to those solutions and, in any event, their nature and quantity (i.e. 12 monthly management accounts and maybe half a dozen board meetings per year) has not changed since the days of paper disclosure.  Consequently, there is no reason why the method of disclosing such routine business documents in the new e-disclosure world should not replicate the method used in the old paper disclosure world.  The ease and practicality of disclosing sets of documents in a ‘file’ is little different whether that file is a electronic folder in an e-disclosure database or a ring binder in a cardboard box!

The only real difference in the new e-disclosure world is that multiple drafts of each document will have been retained (or can be ‘recovered’), whereas in the old paper disclosure world most, if not all, drafts would have been destroyed.  If it is appropriate or necessary to disclose drafts of such documents, these should be placed in the same electronic folder as the final versions but should be clearly distinguishable as drafts.

Finally, it is worth noting that considerable efficiencies can be obtained by potential litigants if they adopt good processes for managing their electronic information irrespective of whether or not a dispute is contemplated.  Jackson LJ states that ‘[t]he first step in the EDRM [Electronic Disclosure Reference Model] is information management.  This is the organisation by potential litigants of their electronic information, so that the relevant categories can easily be found.  The equivalent traditional approach is a good filing system, as opposed to a room full of unlabelled boxes of papers.  In relation to electronic disclosure, the management can for example take the form of ‘filing’ or labelling documents, so that they are easy to search for and organise.  The obvious result of good information management is that the savings can be huge for a potential litigant, as the pool of potential information is already smaller and better organised’[26]. 

We very much agree.

Simon Anthony is a chartered accountant and a Managing Director in Navigant Consulting’s Disputes & Investigations Group in London. Phil Beckett is a chartered certified accountant and the Director of Navigant Consulting’s Computer Forensics and E-Disclosure Group in London: santhony@navigantconsulting.com and pbeckett@navigantconsulting.com

[1] http://www.judiciary.gov.uk/about_judiciary/cost-review/preliminary-report.htm

[2] Paragraph 4.1, Chapter 41, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[3] Herbert Smith’s Litigation E-Bulletin dated 27 April 2009

[4] Paragraph 4.4, Chapter 41, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[5] Paragraph 1.1, Chapter 40, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[6] Paragraph 6.5, Chapter 40, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[7] Paragraph 6.9, Chapter 40, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[8] Appendix 19, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs (note that Jackson LJ at footnote 6 on page 383 incorrectly indicates that the calculation represents ‘…one lawyer spend[ing] 24 years… reading the documents’)

[9] Appendix 19, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[10] Footnote 6, page 383, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[11] Appendix 19, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[12] Rule 31.7 of the CPR

[13] Rule 31.3 of the CPR

[14] Paragraph 5.1, Chapter 40, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[15] Paragraph 4.3, Chapter 41, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[16] Appendix 19, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[17] Appendix 19, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[18] Paragraph 5.11, Chapter 40, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[19] See article by Messrs Charlton and Lavy, ‘E-disclosure, Needles and Haystacks’, in SCL Magazine April/May 2007

[20] Paragraph 3.2, Practice Direction to Part 31 of the CPR

[21] Paragraph 3.2, Practice Direction to Part 31 of the CPR

[22] Rule 31.10(3) of the CPR

[23] Paragraph 2.8, Chapter 41, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[24] Paragraph 4.5, Chapter 41, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs

[25] See article by Messrs Charlton and Lavy, ‘E-disclosure, Needles and Haystacks’, in SCL Magazine April/May 2007

[26] Paragraphs 3.11 and 3.12, Chapter 40, Part 8, Jackson LJ’s Preliminary Report on the Review of Civil Litigation Costs