R v Rodica Gheorghe

This appeal concerns a claim for Special Preparation under Schedule 1 to the Criminal Legal Aid (Remuneration) Regulations 2013. The relevant Representation Order was made on 14 June 2019, so the 2013 Regulations apply as in force on that date. The relevant provisions are to be found at paragraph 17 of Schedule 1:

“17.— Fees for special preparation

(1)

This paragraph applies where, in any case on indictment in the Crown Court in respect of which a graduated fee is payable…

(a)

it has been necessary for an advocate to do work by way of preparation substantially in excess of the amount normally done for cases of the same type because the case involves a [very unusual or novel point of law or factual issue; or

(b)

the number of pages of prosecution evidence, as defined and determined in accordance with paragraph 1(2) to (5), exceeds—

(i)

in cases falling within bands 6.1 to 6.5 (dishonesty offences, including proceeds of crime and money laundering), 30,000…

and the appropriate officer considers it reasonable to make a payment in excess of the graduated fee payable under this Schedule.

(2)

Where this paragraph applies, a special preparation fee may be paid, in addition to the graduated fee...

(3)

The amount of the special preparation fee must be calculated…

(a)

where sub-paragraph (1)(a) applies, from the number of hours preparation in excess of the amount the appropriate officer considers reasonable for cases of the same type;

(b)

where sub-paragraph (1)(b) applies, from the number of hours which the appropriate officer considers reasonable to read the excess pages…

(6)

In determining a claim under this paragraph, the appropriate officer must take into account all the relevant circumstances of the case, including, where special preparation work has been undertaken by more than one advocate, the benefit of such work to the trial advocate.”

The definition of “pages of prosecution evidence” (“PPE”) referred to at paragraph 17(1)(b) of Schedule 1 is to be found (as paragraph 17 indicates) at paragraph 1, subparagraphs (2)-(5) of the Schedule:

The Appellant represented Rodica Gheorghe (“the Defendant”) in the Crown Court at Southwark. The Defendant faced charges of modern slavery, controlling prostitution, human trafficking and removing from the jurisdiction the proceeds of that criminal activity. The following account of the prosecution is taken from a very thorough skeleton prepared by the Appellant.

The Defendant was charged in June 2019 following extradition from Romania. Her husband and co-accused, Nicusor Gheorghe (NG) faced similar charges and additional counts of Rape.

The Crown first based its case on the evidence of a large number of complainants who had worked as prostitutes under the Defendants’ control. Over time, the Crown analysed 14 mobile phones that had been seized at a police raid of the brothel at the centre of the Defendants’ activities, as well as a Dell laptop holding significant quantities of material linked to the brothel business.

The material on the phones and the computer linked both defendants to the placing of sexual adverts for women said to be controlled by them, and to trafficking and money laundering arrangements. The Crown sought to demonstrate that each defendant was linked to the mobile phones found at the brothel and said to be used by the women and the defendants as part of the business.

The mobile phones held images of travel tickets which supported the trafficking counts and messages between the defendants and others which were relied upon to evidence control, discussions about running the business and about the sexual activities of the Defendant in connection with an individual said to be a customer of the business.

The Crown presented charts of analysis and presentation documents explaining the significance of material extracted from the laptop and the mobile phones. Agreed facts addressed similar issues.

The Defendant claimed that she was herself a prostitute, working in the brothel. She aligned herself with the complainants and denied trafficking, exercising any control over complaints or being party to a joint venture with her husband. She denied attribution of the handsets and any responsibility for the laptop.

The Defendant had not been at the brothel when the Police raided it, having at that time been in Romania for several months. Consequently, she was not present when the handsets and computer were seized. During the trial, she did not support her husband on significant issues in his defence. She agreed that there had been a division of money but she denied being party to the division or to any trafficking, or profiting from the control exercised by her husband.

A first trial date was identified on 29 July 2019 for the Defendant’s husband. An application to dismiss was listed in June. That was argued and was not successful. The Defendant’s defence was not going to be trial ready for July 2019 and so an application to adjourn was heard.

By 27 June 2019, the Crown had identified a large quantity of telephone intercept obtained by the Romanian authorities. This led to the service of a vast amount (over 1 million pages, according to the Appellant) of telephone intercept material which had to be translated and caused significant delay to the case (over a million files and 27,000 calls). A new trial date was fixed for 2 December 2019, which was intended to allow all parties time to review the extensive served material, then moved to January 2022.

The defence sought the service of the full telephone download material so as to be able to consider the various presentation charts and to agree facts (the telephone intercept material, which was reviewed by the Appellant’s junior, is not the subject of this appeal). The evidence was duly served, and forms the basis of the claim for special preparation. The defence produced an audit of the data served, which was provided in XRY format and partially converted to PDF by the Appellant, only to produce a page count for remuneration purposes.

The trial ran its course over a period of 6 weeks. Both defendants were convicted and

Following the trial, a claim was submitted by the Appellant for 206 hours’ special preparation. In support of the claim, a detailed note was submitted, supported by a data audit and a work log.

The Appellant’s case was that there had been over 160,000 pages of served evidence. 20,000 pages counted towards the calculation of the graduated fee and a special preparation claim could be justified for review of the excess pages over 30,000. At one minute per page, he pointed out, that would come to over 2,000 hours of special preparation.

The Determining officer, citing R v Dunne [2013] 6 Costs L.R. 1031, rejected any suggestion that the special preparation claim could properly be assessed on a “time per page” basis. This must, in my view, be right, although in fairness it seems to me that the “minute per page” figure was offered only as an illustration in support of the proposition that the time claimed by the Appellant could be seen to be eminently reasonable.

In support of that proposition, the Appellant has referred to R v Al-Goni, Dandash, O’Gorman & Ataya (SCCO 252/08; 253/08; 254 & 255/08; 08/09; 09/09; 010/09; 11/09; 12/09 13/09; 14/09; 15/09; 16/09 & 17/09; 18/09 & 19/09, 21 January 1990) but the pragmatic adoption in that case by Master Rogers of a minute-per-page measure as the “least imperfect” available measure was a response to a highly unusual case. It was, expressly not intended to lay down any principle and Costs Judges have since consistently declined to adopt any “time per page” measure for special preparation.

The Determining officer noted that the Appellant’s work log provided did not stipulate what data was reviewed on which days, and how long each section of the data took to review. The work log recorded time under generic entries such as “read papers”, “read data” or “read disc data”, falling short of the Crown Court Fee Guidance‘s requirement that the log give dates, times and the nature of the work and in the case of perusal of prosecution evidence, particulars of the documents (Appendix E to the guidance refers).

From a data audit supplied by the Appellant, there appeared to be 6 data files holding 30 subfolders. As the work log did not state which data files or subfolders were considered, or which details of the particulars of the prosecution evidence perused, the Determining Officer based her conclusions upon a reasonable allowance for special preparation upon examination of the served material.

The Determining Officer found the page count of over 160,000 to be inflated and disproportionate, for these reasons.

In her decision letter, the Determining Officer expressed the view that material served in XRY or spreadsheet format should be converted to PDF. Noting that some of the page counts offered by the Appellant number of “items” in XRY format, she found that conversion to PDF reduced the count significantly. For example, the Appellant’s page count for folder KDS/12 was 458, whereas conversion to PDF produced a page count of 112. Folder JAG1410191A had already been converted to PDF by the Appellant, who had in his data audit given a page count of 12,703 pages, whereas the Determining Officer found the correct count in PDF to be 6,233 pages.

In her written reasons the Determining Officer observed that file ALS/260319/10 within folder KDS/5, when converted to PDF gave a page count of 93,999, but that a review of its contents then showed that after page 10046, all of the pages were blank, holding no data whatsoever. File HD3, to which a page count of 24,824 was attributed (I make it 24,828 pages) duplicated subfolders JAG1410191A, JAG1410192A and JAG/141019/2B from folder HD1.

Given that she could derive no assistance from the work log, the Determining Officer could not deem it reasonable for the Appellant to have spent the same amount of time on both folders.

Deducting just the 83,953 blank pages on file ALS/260319/10 and the 24,824 duplicated pages on file HD3, observed the Determining Officer, brought the total page count down to about 52,000 pages. Adding to that the 499 pages of statements, 2813 pages of exhibits and 49 pages of transcripts on the court’s Digital Case System, the Determining Officer derived a total page count of 55,361 pages of actual evidence in the case, of which 30,000 pages were provided for in the graduated fee, leaving 25,361 pages left to support the claim for special preparation.

Of this 25,361 pages, over 10,000 pages held only general and device information, metadata and cookies which, given that the Appellant had a duty to consider all the served evidence, would still not, in her view, merit more than a couple of hours’ consideration, the material being irrelevant and not relied upon by the Prosecution. It did not offer any context for consideration of other, relevant data, nor had the Determining Officer seen anything to show that it could have provided support to the Defence.

On that basis the Determining Officer allowed 112 hours (14 full 8-hour days) to consider the relevant material, and another 8 hours to consider material of no obvious evidential value, such as metadata.

The Appellant maintains the claim for 206 hours’ special preparation, on these grounds.

It is accepted, as the Appellant puts it, that PDF was the best mechanism to evidence the pages for the LAA. The data was however provided to the Defence on disc, in XRY format. That is the format in which the material was reviewed by the Appellant on a computer, so the claim is not to be determined by reference to PDF pages but on the time reasonably necessary to review the data in XRY format.

The data was partly converted to PDF in order to inform the calculation of pages for the purposes of the Appellant’s graduated fee and special preparation claims. On the hearing of the appeal, the Appellant told me that the exercise of conversions took so long that it was never completed. An audit of the data was however conducted by the Appellant, supplied to all counsel and verified by the Appellant’s Junior. The number of pages, according to the audit, is 167,558. The audit shows, as noted by the Determining Officer, that folder HD3 does duplicate the files on folder. Deducting that duplicated material still leaves a page count of 142,730.

Duplication, submits the Appellant, is a potential issue in PPE claims, not in special preparation claims. In any event, the duplication would have to be “blindingly obvious” to have any significance at all (R v Everett, Costs Judge Whalan, SC – 2019- CRI- 000038, 224/19, SC – 2019- CRI – 000003, SC – 2019 – 000017 and 157/19, 4 November 2019). Only careful analysis would reveal any such duplication (and it was such an analysis that revealed it in this case after it had been considered). Hindsight has to be guarded against, as many costs decisions make clear. It was only on analysis that the duplication was noted. In any event, it is not suggested that every page of the data had to be considered carefully. Some material would plainly take longer to scrutinise.

The Appellant invited me to consider the data in XRY format and provided screenshots identifying the number of pages in folder KDS/5, which contains a download from one of the mobile phones seized at the brothel when the Defendant was abroad. The phone was attributed to the Defendant. It was assigned to ‘Monica”, a name used by the Defendant, and contained messages related to “Alina”, a name used by the defendant when advertising her own sexual services. Who used it, when and how was, says the Appellant, key to the issues in the case.

Screenshots taken by the Appellant from the XRY format shows that folder KDS/5 contains 93,988 items which can be exported to PDF format. The Appellant argues that in seeking to convert the data to PDF and then subjecting it to a process of manipulation by removing blank pages and/or cells, the Determining Officer has gone about the task in an unfair way.

The Appellant argues that Costs Judges have recognised that this approach is flawed in a number of recent decisions. He refers to R v Campbell (SC-2020-CRI-000254 and SC-2021-CRI-0000012, 22 October 2021) in which, he says, Costs Judge Rowley cautioned the LAA against data manipulation of the type seen in this case. In that case, the argument was made that it was inappropriate for the DO to remove blank cells in data for precisely the reason set out in this case. As in R v Campbell, it is not clear how the Determining Officer has manipulated the data to remove blank pages. To do so in any claim is fraught with difficulty and is not transparent. Other decisions, on similar grounds, have criticised attempts to reduce a page count on DCS by removing what are said to be blank pages or cells.

The Appellant argues that the conversion process is subject to error, so that the fairest approach is to look to the pages as revealed in the XRY screenshots provided by the Appellant. If the page count gets anywhere near 100,000 (as the Determining Officer appears to concede they do before blanks are removed), deducting the first 30,000 leaves 70,000 pages to support the special preparation claim. While many pages would take very little time to scrutinise (and it is not suggested every page had to be read in detail), the suggestion that 206 hours is unreasonable cannot be sensibly sustained.

Even if the Determining Officer were correct to reduce the pagination by removing blanks, then 206 hours, says the Appellant, is still modest.

It was, says the Appellant, wrong for the Determining Officer to apply a test of relevance so as to include, for example, call logs, chats, contacts, GPS, locations, SMS, web history, searches, images, cell tower data and travel data, but to exclude for example device information, data files, timelines, accounts etc. this approach, he says, is in error and not supported by the 2013 Regulations or any costs decision. The Appellant points out that in Special Preparation decisions such as R v Dunne, R v Gabbidon (SCCO 45/19, 4 December 2019) R v Syan SCCO 13/17, 19 July 2017) and R v Gheorghe (SCCO 71/18, 21 January 2019) no test of relevance was applied. Once a PPE count of 30,000 has been reached, says the Appellant, what remains is a special preparation claim. The evidence produced in support of that claim is not subject to a test of relevance.

The Appellant argues that the Determining Officer did not apply this approach consistently or correctly in any event, as can be seen by reference to the contents of file KDS/5. According to the XRY screenshot supplied by the Appellant, of the 93,988 “items” in in KDS/5, 46,307 are images, each of which, in XRY format, count as a page. The Determining Officer has not included those images within the PPE count.

On the hearing of this appeal, the Appellant stressed the importance of the laptop and telephone download data in relation to the case against the Defendant. The Prosecution used extracts from that data, such as identities used by the Defendant and advertising material apparently created and distributed by the Defendant, to persuade the jury that the Defendant was controlling and not just engaging in prostitution.

In preparing a data audit for the Determining Officer, the Appellant did exclude irrelevant material, as for example in folder DWG/1, from which the Appellant counted only 1,922 pages of image and video data from a much larger body of irrelevant data. This can be contrasted with, for example, folder HD/1, a download from the Defendant’s telephone, which according to the Appellant’s data audit comprises some 37,940 pages.

I entirely accept that much of the telephone download data reviewed by the Appellant was of real importance in identifying the role of the Defendant in controlling prostitution. The difficulty lies in identifying an appropriate page count, and I am unable to agree with the Appellant’s methodology in that respect.

I start with the proposition that the Determining Officer was wrong to apply a test of relevance to the data relied upon in support of the Special Preparation claim. I am unable to accept that. This is a claim under paragraph 17(1)(b) of Schedule 1, in which the Special Preparation fee is based upon the number of hours considered reasonable for reading PPE in excess of 30,000 pages. The starting point, accordingly, must be to establish the total PPE count, so as to identify the figure in excess of 30,000 pages. One can then allow a claim for time reasonably spent in reviewing those excess pages.

Electronic evidence served on disc, in accordance with paragraph 1, subparagraph (5) of Schedule 1 to the 2013 Regulations, will not be included within the PPE count unless the Determining Officer decides that it would be appropriate to include it, taking into account the nature of the document and any other relevant circumstances. A key criterion, in coming in that decision, is whether the evidence in question is of central importance to the case (Lord Chancellor v SVS Solicitors [2017] EWHC 1045 (QB)).

In other words, “relevance” is very much to the point. Duplication may well also be an issue, depending upon the nature and extent of the duplication.

Those conclusions, which seem to me to be unavoidable, are in no way contrary to the logic of the decisions referred to by the Appellant. R v Gheorghe was concerned with a claim under paragraph 17(1)(a), and so was not based upon the PPE count. In R v Dunne the PPE count was agreed, and in R v Gabbidon and R v Syan it does not seem to have been in issue.

For those reasons, it seems to me that the Determining Officer cannot be said to have been wrong to attach less importance to material such as device data and timelines, which are commonly excluded from the PPE count by both Determining Officers and Costs Judges on grounds of irrelevance and duplication. Strictly speaking she could have justified excluding them from consideration altogether, rather than attributing to them a lesser amount of time for review.

With regard to duplication, the phrase “blindingly obvious” is a red herring. It was used by the Respondent’s advocate in R v Everett in support of the proposition that duplicated material could be identified so readily that it should not be included within the PPE count. The point of Judge Whalan’s decision was that in that case, it was not possible to identify the duplication without first thoroughly analysing the underlying material. He did not offer “blindingly obvious” as any kind of test.

In this case, the Appellant’s data audit identified what seems to me to be rather obvious duplication between HD1 and HD3, given that the relevant file names and page counts are identical. I am not persuaded that it would have taken any degree of analysis to spot the duplication.

As for conversion into PDF, identifying an appropriate page count for data served only in electronic form can be problematic. That is because the concept of “pages of prosecution evidence”, as the term indicates, dates from a time when all prosecution evidence was served on paper. It can be difficult to apply it to evidence served only in electronic form.

In many decided cases, the same data has been served in spreadsheet (usually Excel) and PDF format, and the consensus among Costs Judges in such cases has been that although legal representatives are likely to work from the spreadsheet format, PDF, being designed to mimic presentation on paper, is nonetheless the best format from which to derive a reliable PPE count.

This, as I have indicated, seems in principle to be conceded by the Appellant. He argues however that the Determining Officer’s conversion of the data from XRY to PDF has produced an outcome which understates the volume of the material actually considered, and in consequence the time needed to consider it.

It seems to me that the difficulty of extracting a PPE count from XRY format lies at the heart of this appeal. Before returning to the facts of this particular case, I should refer to some of the case law relied upon by the Appellant.

Judge Rowley in R v Campbell was not addressing the consequences of conversion from another format to PDF. He was considering evidence served in spreadsheet format, and the determining officer’s attempts to identify an appropriate PPE count in that format. Notably Costs Judge Rowley did not criticise in principle the determining Officer’s attempt, in R v Campbell, to “manipulate” the served data (by which I understand him to have meant that the determining officer rearranged the data within the spreadsheet) so as to eliminate blank cells. He saw that as a matter of degree.

Judge Rowley found that the determining officer had taken a rather overzealous approach to eliminating from the PPE count blank spreadsheet cells. He accepted that that approach overlooked the fact that, even on paper or PDF, not all pages are completely full of data.

He was also concerned with inconsistencies in the page count produced by this over-strict approach on determination and redetermination. He fell back on an approach that had been applied to data from a different device and which had not been challenged as unfair.

Judge Rowley did observe that conversion to PDF did not seem to be a solution to identifying an appropriate count, because it often “simply encapsulates many blank pages as part of the PDF” (paragraph 26 of his judgment refers), but that was merely an observation, upon which his decision did not turn, and he was referring to conversion from Excel spreadsheets, not XRY.

There does not seem to me to be any question of data “manipulation” in this case. Some of the data provided to the Determining Officer had already been converted to PDF and in respect of that data, the Determining Officer merely noted that the page count produced in that way amounted to a fraction of what had been claimed. In respect of the rest, the Determining Officer appears to have converted to PDF and then made a note of the resulting page count, excluding completely blank pages.

That takes me to file ALS/260319/10. On the hearing of this appeal the Appellant pointed out that conversion of that file to PDF produced a total page count of 93,999, close to the “item” count of 93,988 of the same data in XRY format. He suggested that this supports the conclusion that in effect, the PPE page count and the XRY “item” count are the same. This overlooks however the fact that of the 93,999 pages produced on conversion, 83,953 (just over half of the total page count claimed) were completely empty. It also overlooks the Determining Officer’s findings, in respect of other files, to the effect that the number of pages produced by the Appellant’s own conversion to PDF did not come close to the Appellant’s page count (which was presumably based upon the XRY “item” count).

Whatever the shortcomings of the conversion process, one could hardly expect the Determining Officer to ignore such clear indications that the body of data under consideration was significantly less than a page count of 167,558 would suggest.

This leaves the question of how to identify an appropriate page count. Unlike Judge Rowley in R v Campbell, I do not have an uncontroversial counting method to fall back on. One cannot simply adopt the XRY “item” count. I have no basis for equating an “item” with a “page”, and to do so plainly gives a figure far in excess of the PDF page count, which is accepted in principle as offering the least unreliable page count for PPE purposes.

Nor, incidentally, is it obvious to me that the Determining Officer has forgotten to include 46,307 images from file ALS/260319/10. Depending upon size, all of the images might well fit within the 10,046 non-blank pages she has included within the PPE count.

This leaves me with little or nothing to go on by way of reviewing the Determining Officer’s decision. As the Determining Officer pointed out, a more detailed work log might have helped in this respect. The Appellant accepts that the work log might have identified the phone number or handset was being looked at, but describes that as a ‘best practice’ approach which ignores the reality of how counsel works.

I cannot agree. It is for the Appellant to produce the information needed to support his claim, which might have been stronger had his work log for example identified which particular file was being examined at a given time, along with a brief indication of its contents.

I do have some reservations about the outcome of this appeal. It seems evident to me that the Appellant did have to review a very substantial body of data (too large, I understand, to be provided to the LAA via its file sharing platform). I do not believe that the Appellant would have made a claim for 206 hours of special preparation if he did not believe it to be justified, and I have no analysis upon which I might be able to judge any weaknesses or distortions in a page count produced on conversion from XRY to PDF.

I bear in mind however that it may be necessary in some cases for counsel to review a large body of data that does not necessarily support a special preparation claim. The obstacles to a successful appeal in this particular case have been first that the Appellant’s approach to formulating the claim seems to me, for the reasons I have given, to be wrong in principle and second that there is nothing in the information I have been given that seems to offer any sound and reliable basis for increasing the allowance made by the Determining Officer.

For those reasons, this appeal does not succeed.