[View without highlighting]
[Printable RTF version]
||Neutral Citation Number:  EWHC 45 (QB)
||Case No: HQ12X01115 and others
IN THE HIGH COURT OF JUSTICE
QUEEN'S BENCH DIVISION
||Royal Courts of Justice
Strand, London, WC2A 2LL
||15 January 2016
B e f o r e :
THE HONOURABLE MR JUSTICE SUPPERSTONE
|| Various Claimants
||- and -
||Sir Robert McAlpine and others
||- and -
||Balfour Beatty Engineering Services Ltd and others
John Hendy QC, Hugh Tomlinson QC, Guy Vassall-Adams
(instructed by Thompsons Solicitors LLP) for the Claimants
David Cavender QC, Edmund Nourse QC and Christopher Stone
(instructed by Macfarlanes LLP) for the Macfarlanes Defendants
Hearing dates: 7-8 December 2015
HTML VERSION OF JUDGMENT
Crown Copyright ©
REASONS FOR THE ORDER OF THE COURT MADE FOLLOWING THE HEARING ON 7-8 DECEMBER 2015
Mr Justice Supperstone (giving the reasons of the court):
- At the conclusion of the adjourned hearing on 8 December 2015 we dismissed the Claimants' application to admit expert evidence on the estimation of their potential career earnings if they had not been blacklisted. By an Application Notice dated 19 June 2015 the Claimants had applied (in the terms of the draft order):
"… to rely at trial on a written report in relation to each Lead Claimant and Reserve Lead Claimant on the estimation of that Claimant's expected career earnings if he had not been blacklisted. Such reports shall be jointly authored by the following experts in the identified fields:
(a) A labour economist, namely Dr Victoria Wass;
(b) An expert in the field of actuarial science and statistics, namely Dr Zoltan Butt;
(c) Experts on working patterns, practices and wages in the construction industry, namely Professor Linda Clarke and Dr Ian Fitzgerald."
Following a hearing on 14 and 15 July 2015 the Claimants had sought to rely only on a single report of Dr Wass. We stated on 8 December 2015 that we would give reasons in writing for our decision, which we now do.
The Legal Framework
- Section 3 of the Civil Evidence Act 1972 ("the 1972 Act") provides, so far as is material, as follows:
"(1) Subject to any rules of court made in pursuance of Part 1 of the Civil Evidence Act 1968 or this Act, where a person is called as a witness in any civil proceedings, his opinion on any relevant matter on which he is qualified to give expert evidence shall be admissible in evidence…
(3) In this section 'relevant matter' includes an issue in the proceedings in question."
- CPR Part 35.1 provides that "expert evidence shall be restricted to that which is reasonably required to resolve the proceedings".
- CPR Part 35.10(3) provides that "the expert's report must state the substance of all material instructions, whether written or oral, on the basis of which the report was written" (see also Practice Direction ("PD") 35, para 3.2(3); and para 55 of "Guidance for the Instruction of Experts" ("the Guidance"), referred to in para 1 of the PD).
- The Guidance (at para 13) states that "Experts should take into account all material facts before them".
- Paragraph 56 of the Guidance states:
"Where tests of a scientific or technical nature have been carried out, experts should state:
(a) the methodology used; and
(b) by whom the tests were undertaken and under whose supervision, summarising their respective qualifications and experience."
- PD 35, para 3.2(6) provides that an expert's report must:
"where there is a range of opinion on the matters dealt with in the report—
(a) summarise the range of opinions; and
(b) give reasons for the expert's own opinion."
(See also the Guidance, para 59).
- In Barings plc v Coopers and Lybrand  EWHC 17 (Ch D) Evans-Lombe J (at para 45) stated the following propositions:
"… expert evidence is admissible under section 3 of the Civil Evidence Act 1972 in any case where the court accepts that there exists a recognised expertise governed by recognised standards and rules of conduct capable of influencing the court's decision on any of the issues which it has to decide and the witness to be called satisfies the court that he has a sufficient familiarity with and knowledge of the expertise in question to render his opinion potentially of value in resolving any of those issues. Evidence meeting this test can still be excluded by the court if the court takes the view that calling it will not be helpful to the court in resolving any issue in the case justly. Such evidence will not be helpful where the issue to be decided is one of law or is otherwise one on which the court is able to come to a fully informed decision without hearing such evidence."
- In JP Morgan Chase Bank v Springwell Navigation Corporation  EWHC 2755 (Comm) at para 19 Aikens J said:
"It is well established that in order to fulfil the requirement of CPR Part 35.1, a court must be satisfied that the expert evidence is properly admissible and will genuinely assist the trial judge in determining the matters which are in issue. The burden of establishing these two requirements rests upon the party that seeks permission to adduce the expert evidence concerned: see Clarke v Marlborough Fine Arts (London) Ltd  EWHC 11 (Ch) at para 5, per Patten J."
- In Singh Mann v Chetty  C.P.Rep 24 Hale LJ (as she then was) observed (at para 17) in relation to CPR 35 that:
"… the court has to make a judgment on at least three matters: (a) how cogent the proposed expert evidence will be; (b) how helpful it will be in resolving any of the issues in the case; and (c) how much it will cost and the relationship of that cost to the sums at stake."
The Factual Background
- Mr Richard Arthur, in his fifth witness statement in support of the Claimants' application, states that following the Court's refusal on 17 December 2014 to allow the Claimants' earlier application for permission "to rely on a generic report which it was hoped could be applied across the whole claimant cohort" (para 8), "the Claimants' legal teams have reflected carefully on the need for expert evidence in general, on the issues which it is necessary for expert evidence to address, and on the particular form which that expert evidence should take" (para 9). Mr Arthur continues (at para 11):
"…The present application, by contrast, is for a series of individual reports which provide valuations tailored to the circumstances of the individual Lead Claimants, whilst retaining as a starting point in each case an analysis that will produce 'benchmark' figures derived from large-scale national data. It is our view that this 'hybrid' approach of combining large-scale data analysis with adjustments that take account of individual circumstances has three key features [see paras 12-14] which make it both appropriate and, indeed, indispensible."
- In a document, in support of the application, entitled "Proposal for Reports estimating the loss of earnings for lead claimants in the CIVIG litigation" dated 18 June 2015 ("the June proposal") Dr Wass explains (at para 1.2):
"The valuation of the potential earnings of each lead claimant had he not been blacklisted is a particularly difficult element of the claim to value. The approach proposed is to develop benchmark wages and employment probabilities in each year from the first entry on the database according to each claimant's characteristics (for example year, age, trade, type of project, willingness to travel) using large scale nationally representative data sources and to adjust these on an individual basis using relevant industry data, for example collective agreements. The former is necessary in order to achieve a nationally representative reference point and the latter to account for deficiencies in the large scale data (for example the industry practice of remunerating large numbers of construction workers outside PAYE and of paying completion bonuses) and to achieve greater precision in individual circumstances than is possible from averages measured across broad sub-groups."
- Dr Wass's proposal involved a two-stage process which she explains in the following terms:
"2.3 In stage 1, benchmark estimates for EE1 [Expected Earnings 1] and WE1 [Work Expectancy 1] will be provided for the lead claimants for each year of their potential loss using best available nationally representative large scale data. The Annual Survey of Hours and Earnings (ASHE) and the Labour Force Survey (LFS) cover the entire working-age population of the UK. Both surveys collect information on wages and the LFS collects information on employment status. Econometric techniques will be used to derive relevant statistics for both EE1 and WE1 for construction workers. These statistics are not published and are not readily available.
2.4 The benchmark estimates will be determined for sub-groups identified by the occupations of the lead claimants. The benchmark estimates for each occupation sub-group will cover workers of the age of the lead claimants in that sub-group, and the periods for which they claim losses. The whereabouts on the distribution of ASHE earnings estimates for sub-groups will be informed by industry information, for example industry agreements, collectively bargained rates of pay, including Supplementary Project Agreements (SPAs) and dispensations, and overtime rates.
2.5 In stage 2 benchmark estimates will be adjusted, where appropriate, on an individual basis for each lead claimant using detailed specialist industry information including on levels of demand, specific skills, employment status and travelling allowances. These factors are likely to be important and they are not recorded in the LFS or ASHE. ASHE only covers PAYE employees though a large proportion (c.40% or more over the past 30 years) of those employed in the industry is classified as self-employed. Qualitative data from industry archives, past surveys, research projects and stakeholder sources will be used to adjust the benchmark estimates for individual factors like employment status (e.g. whether PAYE or self-employed).
2.6 The mechanics of the second stage of individual adjustments to the general model are as follows: The first step will be to draw up a checklist of the key variables and earnings and wage indicators (e.g. employment status, type of project, economic climate, type of collective agreement, project hours, age, mobility, region, experience, qualifications, and promotion possibilities) based on their significance, or lack of significance, for earnings for particular occupations, or for particular lead claimants in different periods. The second step will be to assess whether any adaptation to the benchmark estimate for the relevant occupational sub-group is required in an individual case during any particular period in the light of the particular significance (or otherwise) of one or more variable or factor. For instance, if the individual is mobile in a particular period, they will be classified as a 'traveller' rather than a local. If a claimant's case is that but for the blacklisting he would have been likely to have been employed on a particular project and if wage indicators are available for this or a similar contemporary project, then these will be referred to in the adjustment."
The Parties' Submissions and Discussion
The June Proposal and the hearing on 14-15 July 2015
- Mr John Hendy QC, for the Claimants, described Dr Wass's June proposal as "a refined version of that which was put before Master Leslie in support of the Claimants' earlier application for permission to rely on expert evidence" in December 2014 (Claimants' skeleton argument dated 10 July 2015 ("the Claimants' July skeleton"), para 56). The significance of expert evidence to the Claimants' case is described by Mr Hendy in the following terms:
"The Claimants do not rely on the expert evidence in order to establish causation: they rely on it to show expected earnings and therefore a difference compared with actual earnings; whereas they rely on the appropriate inferences from all the circumstances, together with the legal presumption outlined to support their case on causation. In short, the expert evidence is an essential part of the case which the Claimants are entitled to advance because it is necessary to show loss in the sense of a failure to earn in accordance with expectations, but the evidence and legal arguments as to the conclusion that the court should reach as to the cause of that loss are a separate matter." (Claimants' July skeleton, para 66).
- The June proposal involved Dr Wass and Dr Butt being responsible for the "big data" analysis which would produce the "benchmark"; and Professor Clarke and Dr Fitzgerald would provide some generic industry material to inform the "benchmark", and they would be primarily responsible for expressing a view as to the appropriate individual adjustments for each Lead Claimant (Claimants' July skeleton, para 87).
- Mr Hendy described the exercise that the Claimants' experts proposed as "a disaggregation of ASHE data [emphasis added] into more appropriate sub-categories and informed by relevant evidence as to collective agreements and other data" (Claimants' July skeleton, para 111).
- It is essential, the Claimants contend, for the court to have reliable evidence in order to assess what earnings each Lead Claimant would probably have received but for the blacklisting. The Claimants acknowledge that in some cases (they say "a very few cases") it may be possible to provide precise evidence of such earnings in a specific job for a specific time period. However Mr Hendy submits that for the most part this would not be possible for a number of related reasons: first, the relatively short period of projects; second, the frequency with which construction workers changed employer; third, the decades during which the blacklisting subsisted; and fourth, the paucity of documents as to rates and jobs preserved by the Claimants and the Defendants.
- Mr Hendy submitted that there were seven points of inadequacy in relation to the ASHE data (Transcript, 14 July 2015 at pp.97-100): first, it does not cover self-employment; second, the published figures, at least, do not identify the categories of worker in the construction industry on the big sites, covered by the principal national collective agreements which identify the workers and establish the minimum rates; third, the identifiable feature is the longer hours that are worked in the construction industry; fourth, the ASHE figures do not reflect any variation in (a) skills, (b) qualifications or (c) age; fifth, the ASHE tables do not deal with the second part of the equation WE i.e. work expectation; sixth, there is no possibility of calculating the career trajectory of the young worker; and seventh, it is impossible without further calculation to determine the point on the distribution of rates within ASHE in respect of any particular year, for example it is impossible to determine at what point a 45-year-old highly experienced electrician ought to be.
- The Macfarlanes Defendants (and the other Defendants) opposed the Claimants' application at the hearing in July 2015 on a number of grounds. Never, so far as Macfarlanes Defendants' Counsel have been able to discover, has a court permitted expert evidence to go behind the ASHE or other figures, to underlying data, to refine, or disaggregate them so as to be somehow "more specific" to the Claimant. Mr Edmund Nourse QC, for the Macfarlanes Defendants, submitted that "To go down the Wass route will inevitably lead to the court having to consider the experts' methodology and arguments about statistical reliability" (Macfarlanes Defendants' skeleton argument on expert evidence dated 10 July 2015 ("Macfarlanes' July skeleton"), para 21). The conventional approach to make assessments of lost earnings, both past and future, is "on what is often necessarily a broad-brush basis, with arguments about what centile should be used, what Ogden-type discounts should be used, or other adjustment should be made, depending on the particular case" (Macfarlanes July skeleton, para 6). The real problem, Mr Nourse submitted, with the proposal was that Dr Wass and her team would not be giving evidence as to matters over which there are "recognised standards". There are no recognised standards, for example, as to what one might take into account to vary ASHE tables in particular cases. Not only did the proposal contain no explanation of methodology, but Dr Wass did not assert that the proposed methodology for manipulating statistics was generally recognised and would produce statistically reliable results, or that what is to be produced will be meaningfully "better".
- Further Mr Nourse submitted that a serious omission from the Claimants' proposed exercise was that it did not contain any mention of looking at a claimant's prior earnings as a guide to his future earnings, despite their obvious relevance.
- Mr Nourse submitted that the second stage of the process was even more objectionable than the first. Each of the objections to the first stage applied with even more force. In effect, Dr Clarke and Dr Fitzgerald would be making a series of adjustments for various matters which they may or may not have figures for before then adjusting them for particular characteristics for the individual claimants, whether significant or not. These adjustments were, it appeared, to cover such matters as skills shortages and the weather which Mr Nourse observed one might have thought would be reflected in average wages in ASHE.
- Mr Nourse did not accept that either Professor Clarke or Dr Fitzgerald was sufficiently expert to undertake stage 2 of the exercise, judging by their CVs. Their CV's reveal no experience in making adjustments to ASHE, or "improved ASHE" tables in light of particular factual matters (and no such body of "expertise" in the required sense exists).
- At the hearing in July 2015 Mr Nourse (and the other Defendants' counsel) also advanced a number of practical objections to the proposal including timing and cost. Most importantly Mr Nourse noted that Dr Wass and Dr Butt were currently seeking permission to use the underlying micro-data from the ASHE and LFS surveys, which are not published. He made the point that if the Secure Data Service ("SDS") were not to allow this information to be made available to the Defendants, their experts and lawyers and then the court, then the entire exercise would be a non-starter. The Claimants had been asked to clarify the position by letter of 2 July 2015. In his oral submissions in reply Mr Hendy suggested that there would be no breach of s.39 of the Statistics and Registration Service Act 2007 ("the 2007 Act") because Dr Wass would not be identifying any particular person who has filled in or completed the ASHE survey in April of any particular year (Transcript, 15 July 2015, pp.121-122).
- On 15 July 2015 at the conclusion of the parties' submissions Mr Hendy realistically abandoned Stage 2 of the proposal and informed the court that the Claimants would seek only to proceed in relation to Stage 1. Further, he proposed that there be only one expert report and that Dr Wass would be the sole signatory of that report, and the only expert witness. The application to admit expert evidence was then adjourned on the basis that the Claimants would give further consideration as to the content of the proposed report and the matters in respect of which Dr Wass could give expert evidence. The court directed that any expert report on which the Claimants wished to rely at the adjourned application be filed and served by 4pm on 24 November 2015.
The November Report and the Hearing on 7-8 December 2015
- On 24 November 2015 Dr Wass produced a "Report on the stage one benchmarks for earnings in the construction trades 1970-2015" ("the November Report"). Dr Wass states (at para 1.1) that:
"The contribution of the research undertaken in this project is, assuming liability is established and causation is proved (i.e. there was a loss of earnings caused by the claimant having been blacklisted), to value the potential earnings of each lead claimant had he not been entered onto the construction industry blacklist."
- Dr Wass continues in the "Summary of opinion" as follows:
"1.2 The extent of the Claimants' loss will be approached in two stages. The purpose of Stage One is to value the potential earnings of each lead claimant had he not been entered onto the construction industry blacklist. This is approached by estimating the earnings of a hypothetical comparator (not blacklisted) who shares the key earnings-related characteristics of each claimant to produce 'benchmarks'.
1.3 Stage Two will allow for adjustments to these benchmarks on an individual basis using individual information and relevant industry data. The benchmark is necessary in order to achieve a nationally representative reference point and the latter to account for deficiencies in the large-scale data (for example, the industry practice of remunerating large numbers of construction workers outside PAYE and of paying completion bonuses). Adjustment will achieve greater precision in individual circumstances than is possible from averages measured across broad sub-groups.
1.4 This report is focussed on Stage One and develops a method for the systematic evidence-based reconstruction of earnings for a hypothetical comparator construction worker who is as closely matched to the claimant as is possible in terms of trade, year, age, hours of work, employer size and coverage by collective agreement."
- Dr Wass states that this method relies upon information collected in the New Earnings Survey ("NES") and ASHE, and published in the form of tables with summary statistics (1972-2014) and available as microdata (individual records) (1990-2013) to construct a series of occupational earnings for the construction sector (para 1.5). NES and ASHE are published by ONS as tables of statistics for a limited number of ONS-chosen sub-groups for the period 1972-2014. The results of the surveys themselves are available as microdata only under strict conditions of access and usage (para 1.7).
- The microdata, in Dr Wass's view, allow for "regression analysis" (emphasis added) as an alternative and superior means of estimating earnings (para 1.8). She states that regression analysis, which is a standard approach in labour economics,
"is both more comprehensive and more efficient as a method of estimating earnings in the context of multiple explanatory variables. Regression analysis is more efficient than the quantile matching approach both in statistical terms and in terms of the time taken to undertake each calculation. The regression equations effectively predict occupational earnings in each year depending on employee and employment characteristics (age, hours of work, employer size and coverage by collective agreement)." (para 1.13).
It is Dr Wass's view that
"the importance of this research [emphasis added] is the regression approach to the estimation of annual earnings and the application of an 'Ogden-style' approach to the estimation of the employment risks. In the present context, regression analysis is capable of efficiently estimating annual earnings. It is flexible to the choice of assumptions about the claimant's employment characteristics and, despite the controls imposed by the secure data laboratory, the approach is relatively transparent. The Ogden-style approach is similarly capable and efficient at estimating employment risks. This latter is peer reviewed and tried and tested in the context of valuing awards for loss of earnings following personal injury." (Para 1.20).
- Dr Wass acknowledges that "regression analysis is only possible where NES/ASHE microdata are available" (para 7.6). Hence it cannot be conducted for the years before 1990.
- Dr Wass records that the data has been analysed by Dr Butt who undertook the analysis for the Reduction Factors in the Ogden tables. The analysis used quarterly transitions in the LFS from 1992-2003 (para 2.14). Dr Butt's contribution to this report is in section 8 entitled "Methodology for estimates of employment risks for the construction sector" (see pp.46-47 of the report).
- Dr Wass states that she has experience and expertise in using the NES and ASHE microdata. In order to obtain access to the ONS data for this project she was required to undertake a revised training course which she did on 9 September 2015 and she became an "Approved Researcher" on 14 September 2015. The project also required approval and she submitted a proposal to the SDS on 28 August 2015, approval being granted on 20 October 2015. In the ONS Project Approval application she described "the title of the research" as "Earnings and earnings determination for construction trades 1985-2014".
- In answer to the question in the research proposal application form "How does the project provide a public benefit?" Dr Wass wrote:
"The NES/ASHE series provide the most accurate and reliable source of earnings for construction workers by occupation and potentially therefore the fairest guide to the value of the earnings losses (when actual earnings are subtracted) each has sustained as a result of entry onto the industry blacklist. If approved here, the approach and the estimates will be scrutinised in the High Court in relation to 30 lead cases in The Construction Industry Vetting Information Group Litigation (HQ12X01115). If successful, a further 600 claims are likely to follow.
The existing scheme [that is 'The Construction Workers Compensation Scheme'] has been described as having 'grave flaws' (see section 3 above), offering inadequate compensation (HCSAC 2015: paras 31 and 32) and this may explain the level of uptake on the part of blacklisted workers.
What is proposed here as an alternative is a systematic and evidence-based set of estimates, derived on a transparent basis, using data from a large sample collected and recorded by an authoritative body and with results (once released) shared with all parties at an early stage."
- Mr Hendy submits that the Claimants will simply be unable to advance the case which they wish, and are entitled, to advance without the benefit of this evidence from Dr Wass. Dr Wass is an expert Labour Economist with considerable experience and expertise in disaggregating and analysing large-scale data in order to estimate benchmark fluctuations in earnings and employment probabilities over time. The fact that adjustments may be required to "benchmark" figures in the case of an individual claimant does not, Mr Hendy submits, render the "benchmark" figure worthless. On the contrary, in the absence of anything approaching comprehensive records of actual earnings or particular projects, a reliable benchmark is an essential starting point for the court to track general variations. It is, Mr Hendy submits, in the nature of a benchmark that it will not be a precise match, but that is no reason not to seek evidence of the best possible data relating to the particular industry and trade. The duty of the Court is to award full compensation and to use the best tools and data available to it in order to do so (see Simon v Helmot  UKPC 5, per Lady Hale at paras 60-62).
- Mr Nourse continues to oppose the Claimants application for expert evidence, which he describes as "essentially a fresh one" (Macfarlanes skeleton argument of 4 December 2015 ("Macfarlanes' December skeleton"), para 4). Dr Wass has adopted a significantly different approach to calculating the allegedly "expected" earnings. Now she adopts a "regression analysis". This is not what was proposed by Dr Wass in July 2015. Then, her proposal was simply to disaggregate down the figures in ASHE, allegedly to match the characteristics of the Claimants, and then use those figures for a centile-type analysis. The work that Dr Wass is now suggesting employs a completely new technique to that which she was suggesting in July 2015.
- Mr Nourse notes that the regression analysis uses six different characteristics for the Claimants which include a number of assumptions (see para 9(2) of Macfarlanes' December skeleton). The assumption that the Claimants would always have worked over 50 hours per week has a particularly large effect on Dr Wass's calculation of "expected" earnings. Further, what Dr Wass's R Squared figures (see for example at pages 41 and 44 of her report) show is that her model, using the factors/assumptions identified by Mr Nourse can only explain between approximately 20% and 50% of the variation in wages by her regression i.e. "line of best fit" model. In other words, observes Mr Nourse, between 80% and 50% (her best case) of the wage variations still remain unexplained by her chosen characteristics. Indeed Dr Wass does not say that her approach is sufficiently more accurate than using traditional centile methods so as to justify the significant costs involved. It continues to fail to use the Claimants' prior earnings (i.e. prior to entry onto the database) as a factor in predicting future earnings. Moreover Dr Wass accepts that there is no microdata for the period before 1990. It follows that for the years before 1990 ASHE has to be used. Therefore the centile method is still used for half the relevant period for a significant number of lead claimants.
- Responding to Mr Hendy's seven points (ASHE deficiencies) (see para 18 above), Mr Nourse (Transcript 8 December 2015 at 79-84 and 92-95) contends: (1) the fact that ASHE only covers those on PAYE. In Dr Wass's July report there was the suggestion that Claimants would have earned more on a self-employed basis than they would have done as employees. That was a point therefore that was going to be considered as part of the Stage 2 adjustment. However what Dr Wass now says in her November report is:
"5.3 … In the absence of comparable evidence on earnings under PAYE, CIS and self-employment, it is not possible to estimate the measurement error involved.
5.4 It is my opinion that the earnings information in the NES and ASHE is the best that is available to achieve the aims and objectives set out in section 3 above. It is for the Court to decide whether they are good enough."
Mr Nourse observes that this is not within her area of expertise, and is in flat contradiction of what was proposed in July and what she was instructed to do (Transcript 8 December 2015 at 95). (2) The second point Mr Hendy makes is that the figures do not identify the categories of workers in the construction industry on the big sites; nor do Dr Wass's figures. As for the words "covered by the principal national collective agreements", Mr Nourse makes the point that rather than assume the workers were not covered by collective agreements it could be known whether they were qualified to be covered by looking at the JIB. (3) The third point made relates to the longer hours that are worked in the construction industry. Dr Wass has assumed the number of hours worked. Mr Nourse comments that if somebody works very long hours they are likely to have a higher centile and so the centile can be appropriately adjusted. (4) The fourth point made is that the ASHE figures do not reflect any variation in skills and qualifications. Mr Nourse contends that that is not correct. They do reflect variations in skills because it can be expected that somebody on the 90th centile would be more skilled. Dr Wass does not purport to assess skills or qualifications. However these Claimants must, Mr Nourse suggests, know what level of JIB electrician they were, for example, and one could look at the JIB to obtain information that would assist the analysis. (5) The fifth point made is that the ASHE table does not deal with the second part of the equation: namely work expectation. Dr Wass proposes to do so but it is not clear how she can assist. (6) The sixth point made is that there is no possibility of calculating the career projection of the young worker. However there is none in Dr Wass's proposal either. There will however be 22 generic witnesses of fact who may give some evidence relevant to this issue. (7) The seventh point is that it is impossible without further calculation to determine the point of distribution of the rates within ASHE of any particular point. However this is what Dr Wass purports to have done in her figures for the period pre-1990.
- Mr Hendy does not accept that what Dr Wass is now proposing is novel. At paragraph 3.1 of her supplementary report she says: "The two principal analytical techniques used in my main Report – regression analysis and transition analysis – are established techniques for the purposes for which I have used them here". Mr Nourse accepts that regression analysis is an accepted methodology "but what nobody says, not even Dr Wass, is that there is any acceptance that her approach in this case is an appropriate one" (Transcript 7 December 2015 at 94). What Mr Nourse says is completely novel is applying regression analysis to ASHE figures to produce predicted earnings in the way that Dr Wass has done. She does not suggest that this is an established methodology that can be appropriately used in the present case. The fact is, submits Mr Nourse, that the evidence is not properly expert evidence at all. Further Dr Wass has not set out opposing facts/opinions within the context of established methodologies, as she was required to do (see para 6 above).
- Further there is no indication as to whether what is now proposed by Dr Wass is more favourable to the Claimants than using ASHE, an obvious alternative. Mr Nourse observes: "Because the mere fact that Dr Wass is higher than the ASHE 90th centile, doesn't answer that question, because nobody is suggesting that the Claimants are confined to saying: ASHE 90th centile. If they want to say ASHE 99th centile, you might get a very different answer" (Transcript 7 December 2015 at 95-96).
- At the start of the hearing on 7 December 2015 Dr Wass responded in writing ("the 7 December Response") to paragraph 20 of Macfarlanes Defendant's skeleton raising questions as to Dr Wass's involvement (and expertise) in the work done by Dr Butt:
"Zoltan Butt was employed here to undertake the data manipulations on the Labour Force Survey and to produce a set of employment probabilities for the construction sector using similar methods to those developed during the Ogden project. He has the technical skills and experience to do this manipulation and he worked under my direction. I told him what was needed, we discussed how this could best be achieved and I decided on the course of action. I am familiar with the data and with how it needs to be manipulated in order to achieve the required outputs. It would have taken me longer to undertake the manipulation and I would have used different software. Decisions regarding matters such as variables, time periods and required outputs were taken by me after discussion with Dr Butt…
All the data manipulation for this case was conducted in the period from the middle of October 2015. The raw LFS quarterly panels 1992-2003 originally downloaded for the Ogden project were re-used in this project but a new model was developed to compare the employment probabilities within and outside the construction sector for this same period 1992-2003. Dr Butt and I worked closely together to adapt the analysis to the construction sector and then to include employment outside the sector as an alternative employment status. This is a different exercise to that conducted for Ogden although many of the underlying principles are the same… data used in this project were manipulated separately from the Ogden project…"
Mr Nourse asks, where is there a description of 'the data manipulation'?; where are any of the matters referred to in these passages in the 7 December Response set out?; Where is the "new model"?
- Dr Butt's part of the report, in particular, does not comply with the requirements of CPR 35, PD 35 and para 56 of the Guidance (see paras 5-7 above). His methodologies, manipulations and adjustments apparently have been discussed with Dr Wass but Mr Nourse observes that none of that is in the report. Material facts and his instructions are not set out in any detail. The exercise conducted by Dr Butt involves the use of a number of equations, which are set out at pages 46-47 of the Report, with very little explanation. What is not explained at all is the "extensive and complex exercise in data manipulation" (para 6.2), or "the dynamic multi-state modelling approach" (para 8.1). Still less are other opinions considered.
- A further objection to the November report is that it assumes that there is a Stage 2 to come and that there may need to be subsequent "adjustments" at some point if her assumption should be shown to be wrong. Paragraph 7.38 of the report states:
"It is important to recognise that all the assumptions (employer-size, coverage by agreement and a 50 hour plus working week) are flexible and can be changed in either the calculation of the benchmarks (because they are incorrect on average) and/or in the individual adjustments to be made at Stage Two (because they are incorrect for the individual claimant)."
Mr Nourse comments that this suggests that the Stage 1 analysis could be recalculated "if they are incorrect on average". That, Mr Nourse says, could only be by way of further expert evidence from Dr Wass, presumably as part of some exercise to be conducted after trial. Other than in this respect, Mr Nourse observes that it is nowhere explained when this Stage 2 is to be conducted, or by whom, or how, or using what information. Mr Nourse describes Stage Two as now being "almost completely opaque" (Macfarlanes skeleton argument, para 5(7)). Not only will this affect the timetable for trial, but the Stage 2 adjustments will, Mr Nourse submits, negate the value of the supposed "increased accuracy". Mr Nourse suggests that what it appears Dr Wass is proposing is that after the court makes findings of fact she will conduct a further regression analysis for each of these claimants prior to Stage 2 taking place (Transcript, 8 December at 35). Mr Hendy has not said this is not so.
- Dr Llewellyn in a Preliminary Report (which we admitted without objection from the Claimants) commenting on the method proposed by Dr Wass in her November report observes at paragraph 12:
"Highly complicated The overall method proposed by Dr Wass is highly complicated, requiring both a demanding programme of econometric analysis and a considerable degree of data-cleaning and manipulation.
12.1 If this method were to be adopted it would be necessary—as Dr Wass tacitly acknowledges—to develop the calculations further, and quite possibly significantly further.
12.2 This would almost certainly be inevitable were it to be accepted that the results ultimately have to be produced on an individual by individual basis.
12.3 Even if further work were to be undertaken, there can in my opinion be no guarantee that this would produce results that would be compelling and of sufficient quality to be able to warrant their being used operationally. Such is the nature of research."
- On 8 December 2015 (the second day of the hearing) Mr Hendy applied to rely on a supplementary report prepared by Dr Wass in response to an e-mail sent by Mr Arthur to her at 22:46 on 7 December. Mr Arthur wrote:
"I am writing on behalf of the CSSG in order to provide instructions for a witness statement that we request you to prepare in order to address a number of points made by the Defendants and their expert to the court during the course of today's hearing (7 December 2015)."
- Mr Arthur asked questions under nine headings: "(1) Instructions; (2) Use of benchmark figures versus published ASHE percentiles; (3) Basis and standards for the techniques used; (4) Regression analysis; (5) Transition analysis; (6) Full explanation of calculations; (7) Adjustments to benchmark data at 'Stage 2'; (8) Number of experts in the relevant fields; and (9) Intended use of this work".
- Dr Wass replied to Mr Arthur's e-mail in an eleven page document entitled "Supplementary report to the Court" ("the Supplementary Report").
- Mr Nourse objected to the introduction of this "Supplementary Report" on the basis that it was in breach of directions given by the court on 15 July 2015 as to the filing and serving of any expert evidence on which the Claimants wished to rely by 24 November 2015 (see para 24 above). Having heard submissions from Mr Hendy and Mr Nourse we granted relief from sanctions under Rule 3.9, notwithstanding there was no evidence as to why the default occurred, on the ground that the overriding objective required us to admit the report for the purposes of the substantive application before the court.
- The November report contained only an extremely brief summary of Dr Wass's instructions. Attached to the Supplementary Report was the letter of instruction to Dr Wass dated 12 August 2015. The first issue Dr Wass was asked to address was:
"1. Methodology for producing benchmark earnings based on national data sets
Please describe and explain a methodology which you would recommend for analysing, adjusting and applying large national data sets of earnings and employment probabilities, in particular data from the Annual Survey of Hours and Earnings ('ASHE') and Labour Force Survey ('LFS'), in order to produce the best available match for what a non-blacklisted construction worker sharing the characteristics of each Lead and Reserve Claimant would have expected to earn over the period of his claim, and for the three years preceding the start of his claim.
Please include an explanation of your opinion as to:
(d) Any alternative methodologies which, in your opinion, might usefully be considered and their strengths and weaknesses in comparison with the methodology that you recommend."
- Further, Dr Wass was asked to ensure that her final report
"(3) contains a statement setting out the substance of all facts and instructions which are material to the opinions expressed in the report or upon which those opinions are based (you may wish to append these instructions and/or any further instructions, information or evidence either supplied by us or obtained by you in the course of preparing your report);
(6) where there is a range of opinion on the matters dealt with in the report—
(a) summarises the range of options; and
(b) gives reasons for your opinion."
- Mr Nourse makes three points on these parts of her instructions: (1) Dr Wass has not looked at the actual characteristics of the Claimants; (2) she has not looked at the three years prior to the start of the claim (prior earnings being the most obvious indicator of what a person would have earned); and (3) she does not give her opinion as to any alternative methodologies which might be considered.
- Taking one lead Claimant, Mr Ruffel, as an example why would his prior earnings be ignored? He is an electrician, qualified, for the various different stages under the JIB. He may or may not have been approved under the JIB at different stages but that is a fact that he must know or he could ask his union about. (See Mr Ruffel's pleaded case in CB2 at 630, supplemented by an RFI and his work employment history in HB7).
- Further in her Supplementary Report Dr Wass states that in relation to adjustments to benchmark data at "Stage 2" "they are straightforward for the lawyers to make" (para 6.1). Mr Hendy said this can be done by the lawyers by way of a provision of schedules of special loss (Transcript, 7 December 2015, page 12). However in relation to Stage 2 there will be 22 trade witnesses of fact who may, for example, cover the issue of bonuses. How, Mr Nourse asks, will the lawyers take account of bonuses in circumstances where Dr Wass has assumed that the Claimants will be working more than 50 hours. On page 3 (at the bottom) her instruction states:
"It is necessary for the evidence about expected earnings to address the circumstances of all the Lead and Reserve Claims".
The instructions continue:
"It is our understanding that, for the purposes of producing relevant series and selections of benchmark data based on the available national data sets, the individual characteristics that will be relevant are only the occupation, age and period of claim for each Lead and Reserve Claimant".
Mr Nourse observes, there is no reference in her instructions to the assumptions she has made that (1) the Claimants would always have worked for companies with over 50 employees; (2) they would always have worked over 50 hours per week; (3) coverage by collective agreement (See paras 1.15 and 7.38 of her report). The only assumptions that she is asked to make are set out at page 4 of her instructions.
- Importantly, Mr Nourse submits there are very serious practical problems caused by the ONS conditions for access to the microdata used by Dr Wass. In her November report Dr Wass notes:
"1.9 The conditions for access to the microdata, and the advantages they convey, are particularly strict and impose significant constraints on the way this research has been undertaken and the way in which it can be scrutinised. Access to the data and to any analysis which is not included in the output release is restricted to ONS Approved Researchers who are registered on this ONS Approved Research Project and whose work is restricted to the secure data laboratory. I have made my analysis as transparent as it is possible to do under these conditions. …
1.10 Outside scrutiny of the data and analysis would be possible by another 'Approved Researcher' who would have access to the secure data laboratory. I would be happy to add an Approved Researcher to this project. This would provide access to all the data and the analysis used to generate the results reported here."
- The practical concerns that the Defendants have about access to the data used by Dr Wass, and whether any expert they may instruct would be able to obtain any alternative data from the ASHE microdata server is, Mr Hendy submits, entirely answered by the report of Dr Wass (see para 52 above). He acknowledges there is a difficulty with access, but does not accept that it is an insuperable problem. The commands she has used to produce the data reports that she has relied on are in Appendix 2 to her report and therefore available for any expert instructed by the Defendants to scrutinise. Moreover if any expert instructed by the Defendants wishes to have alternative reports run using different combinations of commands, Dr Wass can run such reports and provide them, subject only to the requirement that they satisfy the non-disclosure criteria for release.
- The suggestion that the Defendants are only able to have access to the NES/ASHE microdata via Dr Wass is, Mr Nourse submits, untenable. Mr Nourse says that his clients are not going to pretend to anybody that they are going to be conducting some wider piece of research. The suggestion that the Macfarlanes Defendants would get access via Dr Wass is "fanciful". Further, any work product of the Defendants' expert is privileged. Dr Wass cannot possibly have access to such privileged material. Moreover there will be considerable problems with access if approval is required every time the Defendants wish to conduct a new regression analysis for each different type of claimant.
- In reply Mr Hendy submitted that one solution to the access problem would be for the court to make an order under s.39(4)(e) of the 2007 Act. However, as Mr Hendy acknowledged, any such order would have to allow the Defendants' researcher to access any data that he required in the microdata for the ASHE material and to run a model and do other investigations. Moreover the Board would have to be made a respondent to any such application, which would take time.
- In response to this suggestion Mr Nourse observed there is no public benefit in his clients having access to the data that he can see. There is plainly a privacy issue to be considered. Accordingly it is by no means certain that the Defendants would obtain access to the material that Mr Nourse submits they might need.
- At the conclusion of his submissions in reply Mr Hendy raised a suggestion made by Mr Tomlinson QC, for the Claimants, that one possibility might be that permission is granted conditional on access being obtained to the microdata by a date to be agreed; the Claimants will then make an application to obtain the microdata, for the specified purpose, from the Board and if that fails by the fixed date then the permission would lapse.
- On the issue of costs the Claimants estimate that the costs of, and associated with, the admission and use of Dr Wass's report, together with the report of any expert the Defendants may instruct in response to be approximately £500,000. Within the context of the overall value of this litigation, as well as the additional value which Dr Wass's report potentially adds to the Claimants' previous estimates, such a sum, Mr Hendy submits, is plainly proportionate.
- Mr Nourse criticises the Claimants' costs figures as being plainly too low. The Macfarlanes Defendants' estimate of costs (including experts and lawyers) is in the region of £700,000.
- We have given careful consideration to the November report of Dr Wass (together with the 7 December Response and the Supplementary Report), and the submissions of counsel.
- The whole approach in the November report is very different from the approach in the June proposal. In her Supplementary Report (at para 4.5) Dr Wass said:
"In advance of access to the NES/ASHE microdata I did not know whether regression analysis would be possible."
Mr Hendy says Dr Wass is not to be criticised in achieving the task she set out to achieve in June by a technique which she did not realise was possible until the Autumn of this year. We observe that since before December 2014 Dr Wass has in fact been considering how best to estimate the Claimants' potential earnings if they had not been blacklisted. However the real problem with what is now being proposed is that it is, as she describes it, "research" (see para 28 above). We are not satisfied that "there exists a recognised expertise governed by recognised standards and rules of conduct" (see para 8 above) in respect of the proposed methodology.
- Mr Hendy made the point that this is an application for permission to adduce expert evidence; it is not a trial of the merits of Dr Wass's report. That is correct. However we have felt it necessary to give the report the degree of examination we have, first, in the light of the detailed competing submissions from counsel; second, in order to test its cogency (see dictum of Hale J in Singh Mann at para 10 above); and third, in order properly to assess whether this expert evidence is "reasonably required to resolve the proceedings" (see CPR 35.1 at para 3 above).
- Further we are not satisfied that what is proposed will assist the court in determining the matters which are in issue. We are not persuaded that Dr Wass's new approach is sufficiently more accurate than using traditional centile methods in circumstances where (1) she has ignored the Claimants' prior earnings, (2) there is no microdata for the period before 1990 and therefore the centile method is used for half the relevant period for a significant number of lead claimants; and (3) there is still the need to conduct Stage 2 which we are not confident will be as easy to do as Dr Wass suggests (see para 66 below). We do not consider that the time and expense that will be involved in conducting the analysis proposed by Dr Wass is proportionate.
- In our view Dr Butt's part of the November report does not comply with the requirements of either the PD or the Guidance. It does not set out properly the facts, the modelling or the methodology used, or the discussions between himself and Dr Wass. All that Dr Butt sets out is the equation that he has used; and none of his workings are exhibited. Further the Defendants may well, as Mr Nourse thinks, have considerable difficulty in identifying an academic with specialist expertise similar to that of Dr Butt who, given the limited explanation of Dr Butt's contribution to the report, can assist them in adequately responding to that part of the report in time for the start of the trial.
- The different specialisms of Dr Wass and Dr Butt are likely to require the Defendants to instruct two persons with similar expertise, both of whom will require access to the ONS data. Mr Nourse raised in clear terms the Defendants' concern about access to the data at the July hearing (see para 23 above). Access to the data is even more critical now that Dr Wass has adopted a new approach in the November report. In our view the Defendants have legitimate concerns about access to the data. We do not think that any of the suggestions made by Mr Hendy or Mr Tomlinson meet Mr Nourse's point that the Defendants are entitled to have access to any data they may wish to use which may involve multiple applications between now and the conclusion of the trial. We consider that there are likely to be real practical problems in obtaining access to the microdata.
- In the June proposal Dr Wass set out what was to be done at stage 2, then, as she proposed, by Dr Clarke and Dr Fitzgerald (see para 13 above). We are not satisfied that that work can as easily be done by lawyers as Dr Wass suggests. Further, for the reasons put forward by Mr Nourse (see para 41 above) we do not consider that the work that will be done on stage 1 will assist the Court materially without the work on stage 2. The June proposal involved a two-stage process to "provide valuations tailored to the circumstances of the individual Lead Claimants" (see paras 11-13 above).
- Finally there is no consideration in the November report of any alternative methodologies. Obvious alternatives are doing what Dr Wass was suggesting in the June proposal, or using ASHE 90 with some adjustments. In June she was proposing to disaggregate ASHE figures to be adjusted with the particulars of the particular claimant. There is no explanation in the November report as to why this does not remain a possible alternative approach. We agree with Mr Nourse that a serious omission from Dr Wass's proposed exercise is its failure to use the Claimants' prior earnings (say for three years prior to the start of the claim) as a factor in predicting future earnings.
- For the reasons we have given this application is dismissed.
Copyright Policy |
Donate to BAILII