Network Rail Infrastructure Ltd v Handy

There is a sense of “déjà vu” in these proceedings. A previous pair of cases, Network Rail Infrastructure Ltd v Conarken Group Ltd and Network Rail Infrastructure Ltd v Farrell Transport Ltd (which this Court addressed ([2010] EWHC 1852 (TCC)), the judgment being upheld in the Court of Appeal in three reasoned judgments ([2011] EWCA Civ 6440)) were held out by the parties as test cases for the establishment of the proper approach to consequential damages following damage to rail track or other installations caused by negligent road drivers, where those damages related to the sums payable by Network Rail to Train Operating Companies (“TOCs”) for delays occasioned by the physical interruption to the train services caused by the negligent drivers. These sums were and are payable for the consequential delays pursuant to a formula contained in Schedule 8 to the Track Access Agreements (“TAAs”) between Network Rail and the TOCs. The judgment was to the effect that the sums were recoverable. The Defendant’s insurers and solicitors in the current matter are the same as in the Conarken and Farrell cases. The current cases are said to be also test cases and seek to address points either not argued or conceded in the earlier case, in particular the extent to which the formula and aspects of the delay allocation exercise were or are reasonable or unreasonable. There has been what at times appeared to be a micro-analysis of the Schedule 8 formula, sub-formulae, the results of which were built into the Schedule 8 formula, and of academic papers and research. The Defendants seek to make some further concessions in the current case and one wonders whether their insurers seek to keep their “powder dry” on those concessions for later cases which will nominally be against different insured defendants. I am sceptical as to how much of a test case the current set of cases will prove to be, particularly since aspects of this judgment are based on factual findings, although I obviously hope that it will guide similar parties in future.

This was set out in the 2010 judgment, which, updated, I set out below:

Liability for negligence is admitted in each of the five cases and there is agreement that the remedial work costs are recoverable, quantum being agreed. The five incidents and their consequences are set out in the five tables below, the contents of which are agreed:

Rob Hatfield Ltd (driver Mr Sudds)

It is evident from the above tables and, indeed, it is common ground, that not only do TOCs to some extent share rail track but also a delay on one line may well have a direct and consequential effect on trains on another line which interlinks with the line most directly affected by any given incident. Thus, in the Ingram case, although the most directly affected TOC was First Great Western, five other TOCs were affected by the bridge strike in that case. These incidents caused, save in one case (Ingram), cancellations of trains and in all cases delay and disruption to trains. The above tables identify the most affected Service Group.

These five sets of proceedings were commenced between May and November 2012, three cases in the Newcastle upon Tyne District Registry and two in the Mayor’s and City of London County Court and they were transferred to the TCC in early 2013. Originally, the Claims were put in negligence, trespass and (in two cases) nuisance but the nuisance claims were abandoned in October 2013. Aviva is the insurer of each of the Defendants as they were of the defendants in the Conarken and Farrell cases. For Aviva, the Court was told in the written closing submissions of Counsel for the Defendants, that the only reason for these cases being contested is that Aviva believes that payment by it of the Schedule 8 attributed sums for delays to railway services is an inappropriate use of its funds because it believes that the payments are "unjustified subsidies" being "much larger than the true loss suffered by the rail industry when a road or rail incident occurs" and that it would not be actively objecting to the Schedule 8 amounts if it appeared "that they were at a reasonable and realistic level."

The remaining issues in the cases, following the admissions of liability, are in broad terms that the Schedule 8 losses as calculated under the Schedule 8 formula substantially and unreasonably exaggerate the loss attributable to the types of “road on rail" incidents. In essence, the Defendants argue that there are exaggerations by reason of the use in the Schedule 8 formula of unrealistic lateness multipliers, the Schedule 8 formula being said not to be appropriate for the type of single incidents which occurred in the five cases and because the road on rail incidents are very infrequent in passenger experience. There are also said to be different aspects of unreasonableness and unfairness in the systems of attribution of delay some of which arises, not as such within the Schedule 8 formula but, within the computerised systems used for delay allocation and attribution, of which the most important is the Railtrack Insurance Claims Costing System (“RICCS”).

The legal teams have agreed on what the main issues are:

“1.

The extent to which the Schedule 8 losses suffered by and claimed by the Claimant are recoverable from the Defendants. This includes the following sub-issues:

a.

Whether (as the Claimant contends) the payment adjustments required under Schedule 8 reasonably estimate revenues likely to be lost by the TOCs as a result of an individual incident of disruption to the rail network as occurred in the present cases.

b.

If the payment adjustments under Schedule 8 do reasonably estimate revenues likely to be lost by the TOCs as a result of an individual incident of disruption to the rail network, what (if any) legal consequences follow, and in particular whether as a result they are recoverable from the Defendants.

c.

Whether (as the Defendants contend) the payment adjustments required under Schedule 8 do not accurately, reasonably or genuinely estimate or reflect the actual lost revenue suffered by the TOCs caused by an individual incident of disruption to the rail network as occurred in the present cases.

d.

If the payment adjustments under Schedule 8 do not accurately, reasonably or genuinely estimate/reflect the actual lost revenues suffered by the TOCs caused by such an incident, what (if any) legal consequences follow, and in particular whether as a result they are irrecoverable from the Defendants."

There is, additionally, a further issue which remains on the pleading which revolves around whether or not there is an effective cause of action in trespass in relation (in particular) to the Schedule 8 losses. Because the Defendants concede now that "fouling" of the railway lines is sufficient "damage" for the purposes of the tort of negligence and because they concede or accept that the same measure of damages is recoverable both in negligence and trespass, they accept that, if the Defendants are liable for the Schedule 8 losses in negligence, they are also liable in trespass. Notwithstanding these concessions, the Claimant asks the Court to address these issues in the judgement to provide guidance in future; it is particularly concerned with the undoubted risk that Aviva through other defendants or other insurers will seek to argue in future cases that, where there is no fouling or real physical damage but merely trespass which prevents trains running, there will be no liability for the Schedule 8 losses which then, arguably, are purely irrecoverable economic loss.

It is in the TAAs that one finds Schedule 8. A typical TAA is that between Network Rail and Midland Main Line Ltd (the TOC principally disrupted by the Handy incident). It is clear that these are regularly updated and between 1995 and February 2006 it seems that some 27 amendments were made. The Preamble refers to Network Rail being the owner of the Network and to its agreement to grant to the Train Operator permission to use certain track comprised in the Network. Clause 1 was the Interpretation clause. "Permission to use" was, except where the contrary was indicated, to be construed as meaning permission:

Clause 6 set out primary obligations of the parties:

Clause 7 required each of the parties to "perform its obligation, if any, set out in Schedule 8". I will return later in this judgment to Schedule 8 which is entitled "Performance Regime".

Clause 8 contained provisions concerning indemnities and liability for performance both in relation to the TOC and Network Rail. Clauses 8.3 and 8.5 stated:

“8.3

Liability for Late Trains

The rights and obligations of the parties set out in Schedules 4 and 8 represent the parties’ sole entitlement as between themselves to any compensation in respect of any damage, losses, claims, proceedings, demands, liabilities, costs, damages, orders and out of pocket expenses arising from cancellations, interruptions or delays to trains.

8.5

Exclusion of Claims for Consequential Loss

Neither party to this Agreement may recover from the other party any loss of revenue (including fare revenue, subsidy, access charges, Track Charges and incentive payments) or other consequential loss in connection with the subject matter of this Agreement caused to it by the other party, save to the extent otherwise provided in this Agreement or any other agreement between them."

Clause 15 addressed payments. The accounting "Period" was 28 days. Provision is made for the payment by the TOCs of various “Charges”. I understand that at least a number of TOCs pay to Network Rail hundreds of millions of pounds a year. This income is no doubt needed to run, maintain and improve the rail track systems throughout the country.

Schedule 4 provides for compensation for what are called "Possessions", whereby, when Network Rail has to take certain types of “Excess” Possession of parts of the track, it pays compensation to the TOCs. This is referred to in Clause 2. Various formulae are provided to work out what compensation is payable, in Clause 3. A payment regime is provided for. Schedule 4 provides, sensibly, for Network Rail in effect to do planned maintenance and improvement works, amongst other things.

In turning to Schedule 8, with which the issues in this case are primarily concerned, it addresses the financial consequences of delays attributable to matters which are within Network Rail’s risk and within the risk of the given TOC. An example of the latter would be train failures or a passenger being taken ill whilst Network Rail carries the risk for problems including those to do with the infrastructure (for instance failed points) as well as weather and the seasons (e.g. leaves on the line) and, the subject matter of these proceedings, "road on rail" incidents. Schedule 8 is predicated upon the recording of Lateness, Delay and Cancelled Stops by a Performance System which Network Rail is required to operate, which is dependent upon Monitoring and Recording Points which record when trains arrive and pass various points. This System enables Minutes Late and Minutes Delay to be determined, although (by Paragraph 3(c)) any Minutes Delay “which arise from a single incident or series of related incidents and which are less than three minutes in aggregate shall be deemed to be zero”. In broad terms Paragraph 5 allocates responsibility for Minutes Delay and Cancelled Stops. Paragraph 5.5 states:

“Responsibility for Minutes Delay on any day in respect of a Service Group [a "collection of Services provided by the TOC”]” caused by incidents which are unidentified…shall be allocated as follows:

(a)

if there are any Minutes Delay in respect of the Service Group recorded as being caused by incidents which Network Rail or the Train Operator are allocated responsibility:

(i)

50% of the unidentified Minutes Delay…shall be allocated to Network Rail, the Train Operator and joint responsibility incidents pro rata to the aggregate Minutes Delay for that Service Group respectively recorded as being their responsibility under this paragraph 5 for that day; and

(ii)

the balance of the Minutes Delay…shall be allocated to Network Rail…”

Systems and procedures are provided for to produce this allocation and a Delay Attribution Board provided for.

In broad terms for each Minutes Delay, there is a relevant financial rate (which will often be different for each of the different TOCs) and these rates are determined by the Office of the Rail Regulator ("ORR”), typically every five years. Thus the relevant Schedule 8 rates came into force on 1 April 2004 in relation to Market Harborough and Bursnalls Bridge (affecting the Handy and Ingram incidents), on 1 April 2006 for the Strood and Holywell incidents (Rob Hatfield Ltd and Parry) and in relation to the Manningtree (Godley) a specific determination on 19 September 2005.

Paragraphs 7 and 8 of Schedule 8 in effect provide formulae to determine how "Minutes Late" should be allocated to Network Rail and to the TOCs respectively. Based on that allocation mechanism, Clause 9 addresses what is called the "Network Rail Performance Sum (RPS)" for each 28 day period beginning on 1 April of each year. The RPS is in effect what Network Rail has to pay or allow to the TOCs or vice versa (the Train Operator performance Sum) for delay caused by factors for which it or the TOC is responsible as the case may be. The formula is not an uncomplicated one but it is substantially dependent upon the Minutes Late allocated to Network Rail under the allocation provisions. The formula in Paragraph 9 is

“RPS = (RPP-RWAML) x BF x RPR

Where:

RPP is the Network Rail Performance for that Service Group specified in Column B of Appendix 1 for the year in which that Period falls;

To produce a financial result under Clause 9 of Schedule 8, one has to apply the Network Rail "Payment Rate" in Appendix 1 to Schedule 8 which for the Midland Mainline Service Group H101 for instance totals £8,897.90 for the "MRE" (Marginal Revenue Effect), the MRE being expressly referred to in the definition of RPR (above); the "societal rate" also referred to in that definition since April 2004 is irrelevant as it is always zero. The MRE is the estimated effect on the TOC's revenue of one minute of average lateness. Clause 13 provided for Retail Price Index adjustments year on year to the Payment Rate.

Clause 12.1 states:

The MREs in the different TAAs provide a prime area of dispute between the parties in this case and it is necessary to seek to explain the MRE. This is an assessed amount which is supposed to reflect the estimated effect on the TOC’s revenue of one minute of average lateness on its timetable service. The Passenger Demand Forecasting Handbook ("PDFH”) is, historically, the rail industry’s guide to forecasting demand for passenger services; the first edition was created within British Rail in the 1980s and successive editions have been produced every few years since. It "summarises the available research into how rail demand is affected by various factors (including train punctuality) and recommends methodologies for forecasting future demand" (see Paragraph 20 of Mr Angus’ first statement). It is "used by all parties in the rail industry (including TOCs, government, the ORR and Network Rail) as the default methodology for almost all forecasts of rail demand" (ibid). It "has been used as the basis of the methodology for assessing Schedule 8 payment rates in every regulatory review since privatisation".

Based on PDFH methodology, the different MREs are calculated (and updated) for the performance regimes addressed in Schedule 8. Any individual MRE rate is calculated by reference to the effect on the TOC’s revenue of one minute being added to the timetabled journey time of the relevant train service and multiplying this by an appropriate "delay multiplier". This recognises the importance of complying with timetabled journey times. Thus, passengers may resent a delay to the timetabled journey and may make future travel plans by reference to the experience which they have suffered on an earlier train journey.

As Mr Angus has said (first statement - Paragraph 115):

“The effect of timetabled journey time on demand has been well researched by the industry. The effect is usually expressed in terms of:

“a 1% change in journey time leads to a X percent change in demand"”

He goes on to say, uncontroversially, in Paragraphs 116-119:

“116.

The value “X” is known in economic terms as the "elasticity" of rail demand to journey time. A typical value of X might be -0.9; that is, a 1% increase in timetabled journey time typically leads to a 0.9% decrease in demand (and hence revenue) on the relevant services.

117.

In fact there is a little more to it than this. Rather than using the journey time between two stations, the industry uses a concept called "Generalised Journey Time", or GJT. This is a measure of timetable quality that takes into account not only the journey time between two stations, but also the frequency of services and (if applicable) the need to change trains as well.

118.

Put simply, GJT is calculated by adding, to the average journey time between two stations, a number of minutes reflecting service frequency and the need to change trains. The less frequent the service, the more minutes are added; and the more interchanges are needed, the more minutes are added. Thus GJT can be reduced, and demand by implication increased, either by increasing the journey time (making the journey faster); or by increasing service frequency; or by reducing or eliminating the need to change trains. Elasticities of demand are then measured with respect to GJT, rather than to journey time.

119.

The elasticity of demand to GJT varies between different types of service. The choice of the elasticities, based on available industry research, is one of the key issues in calculating MREs”.

There was much talk in this case about elasticities and GJT.

Probably the primary area of dispute between the parties relates to the reasonableness or unreasonableness of the MREs. By reference to the above explanations, it is necessary to delve further into the meaning and background of MREs, GJT and the delay multipliers. The delay multiplier is a figure which is supposed to represent the impact of delays and the higher the figure for the delay multiplier the higher will be the MRE. The logic behind the delay multiplier involves a determination as to the likely impact on passengers of delay. Historically, this has been based on what is called Stated Preference ("SP") research which, essentially, involves surveys of passengers who are presented with various choices about delays ranging from how amenable (if at all) they would be to a longer timetabled journey than hitherto compared with the same length of timetabled journey with a risk of delays of a certain length. Passengers may not object too much if the timetabled journey increases by 1% (say one minute) but may well object increasingly if told it might increase by more, but if it increases by 30 minutes on a (hitherto) 100 minute journey they may be inclined to vote with their feet by no longer using the rail service in question. More recently, some academics and indeed others have relied on or referred to what is called Revealed Preference ("RP") research which is statistically based on analysis of income and passenger numbers overall. However, the PDFHs, which have established the delay multipliers, have historically been based more on SP research. A major area of issue between the parties is whether there has been an unreasonable reliance on SP research. Be that as it may, the MRE is, broadly, established by comparing the established or at least perceived difference between an extended timetabled journey time and non-extended timetabled journey time subject to delay.

To calculate the MRE, which is expressed in units of £1 per minute of average lateness per day, there is a recognised formula:

The Schedule 8 basis of assessment is deployed by Network Rail because the (usually numerous) minutes of delay said to have been caused by the five negligent driving incidents in the five cases has resulted in Network Rail in effect having to pay the relevant TOCs the sums generated by the application of the formula.

So far as delay attribution is concerned, the process is governed by a cross industry panel made up of representatives from Network Rail and TOC representatives, the Delay Attribution Board, which issues and updates what is called the Delay Attribution Guide ("DAG"), which has not been materially altered since 2004. Every train that runs on the rail network is monitored by what are known as Contractual Monitoring Points ("CMPs") and Delay Recording Points ("DRPs"). Minutes Delay represents time lost at or between two consecutive DRPs whilst Minutes Lateness is the relative lateness (relating to actual against timetabled arrival time) at any appropriate CMP. Through various computer systems operated by Network Rail (TRUST and Genius), Minutes Delay and Cancelled Stops can be investigated and recorded. Another archive database, PALADIN, is updated frequently and regularly from TRUST and information from PALADIN is imported into another system PEARS (Paladin Extraction and Reporting System), as a repository for such data. Network Rail then uses its Railtrack Insurance Claims Costing System (“RICCS”) to identify the impact that Minutes Delay and Cancelled Stops attributed to a particular incident has had on the Schedule 8 performance regime.

Although relatively belated concessions have been made by the Defendants that there was present sufficient “damage” for the purposes of negligence in each of the five cases, that appropriate consequential economic loss is recoverable and that comparable losses are recoverable for trespass, there remains an issue on the pleadings as to whether or not in law economic loss is recoverable in trespass even where there is no physical damage to land. This issue is not conceded but it is argued by Mr Bartlett QC that the Court need not and indeed should not comment on it; Mr Onions QC urges the Court to address the issue to discourage one party or the other in future cases (test cases or otherwise) from arguing it. I consider that it is appropriate to address the issue.

The Defendants’ concession that there was sufficient damage to the purpose of negligence was expressly put on the basis that “fouling” of Network Rail property could amount to physical damage, it being conceded that there was at least fouling or actual physical damage. I have been referred to a number of cases. In Attorney General for Ontario v Fatehi [1984]] 2 SCR, the Supreme Court of Canada was concerned with a negligently caused traffic accident on the highway which resulted in gasoline, glass and other debris being strewn on the surface of the highway and the local fire service was called to remove the debris; the issue was whether or not there was damage for the purpose of the tort. Estey J giving the judgment of the court said at page 541:

“It is said by the respondent, and by the majority below [in the Court of Appeal] that the appellant must fail as its loss was pure economic loss. Whether that is so depends upon questions of both fact and law. On the facts as agreed, the road was blocked by the negligent actions of the respondent. It ceased to be a road in the sense of a traffic-carrying facility. Whether the respondent achieved this result by deliberately tearing up a section of the surface, or by negligently operating his vehicle so as to drop a large load of rocks on the road, or so as to strew broken auto parts, debris and gasoline on the road, as was here the case, makes no difference in fact. The road, by reason of the respondent’s wrongful acts, ceased to be a road.

The appellant as the owner of the road has thereby suffered damage to its property. The appellants suffered this direct damage in the same manner as any other property owner, and unless there is some disqualifying law operating in these circumstances (which will be discussed later), the appellant would be entitled to recover these as direct damages so suffered."

He then went on to consider a number of the English and other authorities before concluding at page 545:

“In this proceeding it is unnecessary, in my view, to settle this issue because, for reasons given above and illustrated in my view by the cases aforementioned, this is not a case of economic loss but of direct damage to property of the plaintiff occasioned by the negligence of the respondent."

In The Orjula [1995] CLC 1325, Mance J (as he then was) considered a case where a ship’s deck was contaminated with acid such that it had to be decontaminated, albeit that the judge proceeded on the basis that the deck itself had not been physically harmed by the acid. He concluded, without hesitation, that the vessel had been damaged by reason of the contamination (page 1329). In reaching this view, he had particular regard to a decision of Cantley J in an unreported Court of Appeal decision R v Henderson and Batley (29 November 1984) which was a criminal appeal relating to intentional or reckless damaging of a development where lorry loads of soil and rubble were dumped on the site; he said that the question of "whether damage was done to this land was a question of fact and degree" referring to the dictionary definition of "damage" as "injury impairing value or usefulness”. Cantley J also referred to an earlier case of Fisher 1865 LR 1 CCR 7 which held that "an obstruction temporarily rendering a machine useless for the purpose for which it was intended to be used can be damage".

The Appeal Case report of Hunter v Canary Wharf Ltd [1997] AC 655 incorporates the Court of Appeal decision and the judgment of Pill LJ with which the other two judges agreed and which on the point in question was not the subject of appeal. He said this at Page 676E:

“…the deposit of dust is capable of giving rise to an action in negligence. Whether it does depends on proof of physical damage and that depends on the evidence in the circumstances. Dust is an inevitable incident of urban life and the claim arises on the assumption that the defendants have caused excessive deposits. Reasonable conduct and a reasonable amount of cleaning to limit the ill-effects of dust can be expected of householders. Subject to that, if, for example, in ordinary use the excessive deposit is trodden into the fabric of the carpet by householders in such a way as to lessen the value of the fabric, an action would lie. Similarly, if it follows from the effects of excessive dust on the fabric that professional cleaning of the fabric is reasonably required, the cost is actionable and if the fabric is diminished by the cleaning that too would constitute damage. Excessive dust might also be shown to have damaged electrical apparatus and there could no doubt be many other examples. The damage is in the physical change which renders the article less useful or less valuable. On the assumptions we are invited to make, that rather than any general concept of loss of utility is the appropriate test…”

In my judgment and following from the above, property can be damaged for the purposes of negligence where the breach of duty results in substances or physical things being deposited on the property in question in more than a de minimis manner such that the property cannot be used or enjoyed as it otherwise would or could be if the substances or physical things had not been so deposited. So in the Handy case, by so driving his car which ended up straddling the rail track he damaged the track because it could not be used unless and until the car was removed; in the Handy case, he did cause some purely physical damage to some trackside ducting as well which was unarguably physical damage. In the Godley case where a vehicle was driven through level crossing barriers, the primary consequence was consequential fouling of the lines from the wreckage of the barriers; that fouling was damage as of course was the physical damage to the barriers.

The next issue, also (consequentially) conceded by the Defendants, relates to trespass, with the concession being made that, if the damage in the five cases was sufficient for negligence, it must inevitably be sufficient to justify an award of damages in trespass to the same extent as would relate to the cause of action in negligence. If, however, "fouling" of land or other real property is not damage for the purpose of negligence, the question then arises as to whether or not even "fouling" is required for an award in damages (for more than nominal damages) for trespass. Traditionally, conventionally and indeed often, damages for trespass to land are by way of mesne profits. In Ministry of Defence v Ashman (1993) 66 P&CR 195, an originally lawful occupier of land owned by the plaintiff continued to live in the premises after being given lawful notice to vacate; damages were based on the value of the property on the open market based on the proper letting value. In those sorts of case at least, there is no question or issue about damage to the property. The loss is, almost invariably, "economic loss", although as in the Ministry of Defence case there was an argument as to quantification.

No authority has been cited and the textbooks do not clearly appear to suggest that physical damage or even "fouling" of real property is a necessary pre-condition in the law relating to trespass to land for the recovery of related economic losses. One can take a simple example, put to Counsel in this case, of a defendant who, without permission or excuse, parks his fleet of lorries on the claimant’s land on hard-standing, where there is no physical damage to the hard-standing. It would be difficult to describe this even as "fouling" of the land. The damages would, typically, be on this example equivalent to a reasonable charge for parking the lorries; that reasonable charge might well be referable to the income which the claimant land-owner habitually charged or might reasonably charge for lorries to park there and would therefore be equivalent to a loss of income, particularly if income was actually lost as a result of the trespass. Obviously, if there is physical damage caused by the trespass, the costs of repairing the physical damage and indeed also of the direct consequential losses relating thereto would be recoverable. There would be no obvious reason in principle why damages would have to be limited and assessed by reference to the benefit to the defendant (mesne profits, letting value or wayleave fee) or capital diminution or remedial work cost. For instance, a trespasser who causes by his trespass the loss of a wheat crop could expect to pay damages by reference to the income lost thereby. Overall, loss of revenue can be recovered for trespass to land, even when there is no physical damage or even fouling of the land.

The main area of legal issue revolves around the extent to which it is necessary for Network Rail or the Defendant to establish the reasonableness or unreasonableness of a formula based method of establishing loss of revenue as between the claimant and the third party (with whom it is in contract) which is then claimed as damages against that defendant. Part of this issue involves the degree to which the Court can be asked to be concerned with determining the reasonableness or otherwise of the loss of revenue attributable to the tort complained about.

Unsurprisingly, Counsel have primarily based their arguments on the judgments, principally those in the Court of Appeal in the Conarken and Farrell Transport cases. All three judges in the Court Appeal at the very least were in substantial agreement with each other in dismissing the appeal. However, it is said that, on key sub-issues, there may have been material differences. Better to understand the Court of Appeal judgments, it may be helpful to refer to key findings in the first instance judgment:

“54.

Thus it is that one can safely conclude that there is no reason in principle why loss of use, loss of profit or loss of revenue can not be recovered as damages for physical damage negligence, subject to causation being established and provided that it is demonstrably consequential upon the physical damage in question and closely associated with the physical damage.

55.

It almost goes without saying that all losses recoverable in negligence must be of the kinds or types, which, judged objectively, would be reasonably foreseeable by a person in the position of the tortfeasor as the result of the negligence.

56.

Additionally, concepts of reasonableness, justice and fairness, together with policy can assist in the determination of whether any, and if so what, duty of care is owed…

57.

Drawing all these threads together, the following is established in the context of the current case:

(a)

in negligence cases which involve physical damage to property owned by a claimant, loss of use, loss of profit or loss of revenue can be recovered as damages, subject to causation being established and provided that it is demonstrably consequential upon the physical damage in question and closely associated with the physical damage and the work done to repair or replace the damaged property.

(b)

Causation is established on the facts and the evidence.

(c)

The fact that the loss of use, profit or revenue arises because a contract between the claimant and a third party makes the claimant liable to pay or allow it to the third party does not stop it being recoverable…

58.

There can be no doubt that the Schedule 8 sums paid or allowed to the TOCs arose as the direct result of the railway lines in question being closed as a result of the physical damage caused by the Defendants' respective negligence and by what is accepted by all parties as the necessary or reasonable time taken to repair and replace what was required to be remedied…

59.

One then needs to analyse what the scope of the duty of care was in this case. Given that this is a negligence case involving physical damage to a claimant's property, the scope of the duty is to exercise reasonable care and skill in and about driving the heavy motor vehicles concerned so as not to cause physical damage to the property of others on or by the highway. The scope must cover all those heads of loss and damage which are and have been accepted judicially in the past as recoverable; that includes the costs of and occasioned by reasonable and necessary repairs and the loss of profit, use or revenue demonstrably consequential on the damage and concomitant repair of the damaged property.

60.

The Defendants have through their Counsel argued strenuously, and with no little skill, that, to determine whether the scope of any duty of care covers consequential losses and whether the sums paid or allowed to the TOCs were demonstrably or truly consequential upon the two sets of carelessness and physical damage, one must analyse and determine with precision what those sums represent. If, they argue, those sums represent categories of loss which are not consequential, they can not be recovered and are not within the scope of any duty of care.

61.

I said earlier in this judgement that I would return to reach conclusions on the evidence, particularly that of Mr Angus and the documentary evidence. I can summarise those conclusions as follows:

(a)

The bulk of the formula in Clause 9 of Schedule 8 of the TAA involves the determination of the delay for which Network Rail was responsible to the TOC in question in a given Period. The result is applied to the Payment Rate set out in Column E of Appendix 1.

(b)

There is no evidence as such about how the MRE and Societal rates which make up the Payment Rate were themselves calculated.

(c)

It is probable that the product of research and experience on a best assessment basis was put in to the calculation of the MRE rate so that some elements of the loss of fare revenue on the day or days when the railway lines in question were closed or had a restricted use were allowed for. It is probable that a larger element was for the loss of future revenue from passengers' future unwillingness to travel on particular lines affected by any incident which had disrupted travel times…

(e)

Again, there is no doubt that the MRE and the Societal rates were in place in a broad sense to provide incentives or corresponding disincentives to Network Rail and to the TOCs to avoid or limit disruptions to the rail services. The better that Network Rail performed, the greater was the chance that the TOCs would pay them under the Performance Regime in Schedule 8; the better that the TOCs performed, the greater was the chance that Network Rail would pay them. Of course, as between Network Rail and the TOCs, Network Rail took the risk and responsibility under this Performance Regime for damage to its bridges caused by the negligence of drivers which in turn cause disruption to the rail services.

62.

The issue arises however whether any of this really matters. I have formed a clear view that it does not, for the following reasons:

(a)

One has to analyse what the loss to Network Rail actually is.

(b)

On any proper analysis, the sums payable to the TOCs were payable for the loss of use or availability of the rail tracks in question. This was the price, contractually agreed between Network Rail and the TOCs, which was payable in circumstances amongst others in which the rail tracks could not be used by reason of the negligence of drivers who damaged bridges or electricity cables over those rail tracks. One of the main uses to which Network Rail put the rail tracks was to license the TOCs to operate them.

(c)

It cannot be said, and indeed it has not been argued as such, that the sums payable to the TOCs pursuant to the TAAs represented a penalty or were unreasonable or did not represent anything other than a best assessment basis of the loss to the TOCs of the use of the rail tracks for the period in question. It is a material fact that the Rail Regulator approved and endorsed not only the TAAs but also during the review process the Payment Rates, and their constituent elements, the MRE and Societal rates. Obviously, if the sums payable to the TOCs under Schedule 8 were a penalty, unenforceable in law pursuant to such authorities as Dunlop Pneumatic Tyre Co Ltd v Selfridge & Co Ltd [1915] AC 847, the law would find an appropriate way to ensure that the Defendants were not liable for them; that way could be for example that the loss does not flow from the breach or that the scope of the duty of care does not extend to the avoidable payment of unenforceable penalties. Similar considerations would apply if the sums claimed were otherwise demonstrably unreasonable.

(d)

The loss of use can also legitimately be considered as a loss of revenue because in reality, by having to pay or allow the licensee TOCs the Schedule 8 sums, Network Rail's revenue from the TOCs was being reduced by exactly the same amount.

(e)

A key factor in the calculation under Schedule 8 is the amount of rail track or train delay caused in these cases by the negligently caused damage. That delay is made of the times reasonably taken for initial assessment by the emergency services and Network Rail of the accidents and damage and for the necessary and reasonable associated remedial works…

(f)

The Defendants' negligence (that is, their breaches of duty causing physical damage and the need for the concomitant remedial works) therefore directly caused the delays which directly led to Network Rail not being in a position to make available the rail tracks in question for the periods of the delays, the direct financial consequence of which was that Network Rail had to pay or allow their licensees the Schedule 8 sums.

(g)

The fact that within the calculation of those allowances there may have been or were included sums which the TOCs could not have claimed in negligence as against these Defendants is immaterial because it is necessary to consider in this case the position of Network Rail which clearly did have the right to sue for negligence in respect of the physical damage to its property. The reality is that, under the law as it now stands and has stood for many years, these Defendants would never have been liable in negligence to the TOCs for those losses because no property of the TOCs was physically damaged. The losses would therefore never have been recoverable directly by the TOCs. If the Defendants' arguments are right, the one innocent party, Network Rail, cannot recover its very real losses incurred directly in consequence of the Defendants' carelessness…

(i)

Where one has, as here, a duty of care the scope of which includes losses demonstrably consequential upon the physical damage, it must, objectively speaking, have been reasonably foreseeable by tortfeasors in the position of these defendants that, if they seriously damaged bridges over rail tracks or overhead electric cables providing electricity to the railways, the railway services could or would be disrupted for a period of time whilst remedial work was being done. Thus, a loss of use and of revenue was a reasonably foreseeable consequence of the negligence in this case. It is argued that objectively speaking the tortfeasors in this case could not reasonably have foreseen either that payments would be due to the TOCs for the loss of use or that the payments would include incentives. However, this is not material in circumstances where a loss of use is reasonably foreseeable and the loss of use is readily quantifiable in money terms. It is not necessary for the precise loss or machinery by which the loss is ascertainable to be foreseen or foreseeable. In any event, it can hardly be said that a loss of use or indeed a loss of revenue was not objectively speaking reasonably foreseeable by tortfeasors in the position of these Defendants; as was said in the Metrolink case, there was nothing unusual about the facts that Network Rail would receive payment and revenue for providing the rail track for use by rail companies or that it would lose revenue if it did not provide the track. The fact that Schedule 8 provides a somewhat complex formula for determining the value or cost for the non-provision of the rail track is immaterial.

(j)

There is nothing as a matter of policy which can readily be deployed to gainsay either a duty of care existing or the scope of that duty extending to compensation for the loss of use, particularly in circumstances in which previous cases establish an entitlement to loss demonstrably consequential upon the physical damage…

(k)

It is argued by the Defendants that in effect the parties to the TAAs agreed to forego any entitlement by Clause 8.3 and 8.5 to exclude liability for loss of revenue or for losses and liabilities arising from interruptions or delays to trains and that therefore the type of loss, whatever it was, encompassed by Schedule 8 could not have included any (what would otherwise have been) recoverable heads of loss. That in commercial terms ignores the fact that Clauses 8.3 and 8.5 clearly envisage that all such losses shall or may be encompassed within Schedule 8 (or elsewhere in the contract). The parties have agreed that there shall be no freestanding claims for loss of revenue or train disruptions but have effectively agreed that all such entitlements are to be mutually considered as being covered by the entitlements and financial adjustments set out in Schedule 8 and elsewhere. Schedule 8 is either or akin to a liquidated damages or demurrage clause which relates to certain types of delays on the rail tracks to the train services. The rates payable where there is delay attributable to one party or the other were negotiated to reflect what the parties thought realistically represented the likely losses which each should reasonably bear in the case of delay.

(l)

I can not see any or much significant difference in principle between the current case and the negligent driver who drives into a building which houses a shop which is licensed by the owner to a third person. The owner can sue the negligent driver for damages which would include the loss of the sums due from the licensee or compensation payable for the licensee not being able to trade from the premises whilst the requisite reasonable remedial works are done.

(m)

In my view, the Defendants too often in their arguments have confused the losses actually suffered by the TOCs with the losses actually suffered by Network Rail. The former are substantially immaterial in the context of cases brought by Network Rail against the defendants such as those in this case. It is necessary to analyse the losses actually suffered by Network Rail and to determine whether they are truly consequential upon and closely associated with the physical damage which these Defendants negligently inflicted upon Network Rail.

63.

It follows from the above that, in my judgement, the losses representing the sums paid or allowed to the TOCs in relation to the delays caused by the two accidents in question are recoverable by Network Rail from these Defendants. The type of loss was within the scope of the duty owed by both these Defendants and was the reasonably foreseeable consequence of the negligence in these cases; the loss was certainly not too remote. The losses were actually caused by and otherwise demonstrably consequential upon the physical damage to the electrical cables and the bridge respectively and upon the remedial works carried out to put that damage right.”

The appeal was dismissed unanimously and broadly at least, for the reasons given in the first instance judgment. For instance, Moore-Bick LJ said at Paragraph 119 that the first instance judgment "was correct" and his reasons were "substantially the same as those of the” first instance judge. There were three reasoned judgments and it is now suggested that there were some (possibly) subtle but important differences. It is clear that both at first instance and in the Court of Appeal the Defendants were not challenging "the reasonableness of the Sch 8 calculations as a way of assessing how loss is to be apportioned between the parties to the" TAAs (see Paragraph 11 – Pill LJ) but they were not conceding "that whatever parties to the [TAAs] choose to put in them governs the extent of a tortfeasor’s liability to the respondents" (ibid). Pill LJ summarised the issue on the appeal as being "whether the mere presence, in an agreement between [Network Rail] and the TOCs, is enough or whether the heads of damage in the agreements must, before the tortfeasor is held liable, also be tested according to the tortious principles stated by Moore-Bick LJ” (Paragraph 12). Pill LJ summarised some of the Appellants’ Counsel’s arguments:

“26.

Mr Bartlett submitted that the appellants owed the respondents no duty in respect of their financial position under their contract with TOCs, or in respect of the reputation of the railways for reliability in the minds of the travelling public. These include policy-based penalties for lateness and predictions of losses to be suffered by TOCs over an indefinite period anywhere on the rail network by reason of customer reaction to the incident. The sums do not fairly or reasonably relate to the physical loss of use of their tracks suffered by the respondents.

27.

The appellants should not have to pay for the financial consequences of "knock on" effects of the incidents on a range of companies across the rail network, some of which did not even run trains on the damaged track. Moreover, loss dependent on future customer choices driven not by the incident itself but by fear of future incidents was too remote to be recoverable. To define the kind of loss as the respondents' loss of revenue under the contract is too broad. The court has a responsibility to ensure that the scope of liability is kept within proper bounds. That was particularly important when, the respondent being in a monopoly position, there was no market by which the reasonableness of the amounts claimed can be judged.”

Pill LJ’s judgment contains the following, most directly relevant to what is in issue in the current cases:

“32.

The incidents for which the appellants were responsible undoubtedly triggered the payments the respondents were required to make to TOCs under contracts with them. The payments were the direct consequence of the incidents. The contractual loss by the respondents is a direct consequence of the injury to the tracks, as the judge found at paragraph 58 of his judgment. To hold that in itself as establishing liability on the appellants would, however, be to go back to Re Polemis [1921] 3 KB 560. That approach has been rejected. In Overseas Tankship (U.K.) Ltd v Morts Dock & Engineering Co., Ltd. (The Wagon Mound No.1), Viscount Simonds stated, at page 422:

"For it does not seem consonant with current ideas of justice or morality that for an act of negligence, however slight or venial, which results in some trivial foreseeable damage the actor should be liable for all consequences however unforeseeable and however grave, so long as they can be said to be 'direct'".

Foreseeability is the starting point but, as later cases show, other concepts are involved…

36.

In Voaden v Champion (The Baltic Surveyor) [2002] 1 Lloyds LR 623, Rix LJ, with whom Schiemann LJ and Hale LJ agreed, cited with approval the statement of May J in CR Taylor (Wholesale) Ltd v Hepworths Ltd [1977] 1 WLR 659, at 667:

"But secondly, the damages to be awarded are to be reasonable, that is as between the plaintiff on the one hand and the defendant on the other."

37.

Lord Justice Rix added, at paragraph 88:

"I do not see why in the realm above all of remedies the common law cannot mould its principles flexibly to the needs of the situation, and as so often the test of reasonableness lies to hand as a useful tool."…

49.

Foreseeability and remoteness are, in my judgment, to be considered together. Foreseeability, as an abstract concept, is not determinative. An imaginative person may foresee extremely broad consequences as potentially resulting from his actions. In considering whether consequences are too remote to create a liability in negligence, reasonableness has a part to play…

68.

The losses claimed satisfy the requirement of being a direct consequence of the tort. The liability of the respondents to pay sums to the TOCs is the direct consequence of the tort which occasioned the damage to the tracks. However, it has also to be considered whether the appellants are bound by the assessment of damages in the contracts between the respondents and the TOCs and, if not, whether the damages claimed are reasonably foreseeable (Wagon Mound).

69.

Attractive as is the simplicity of a solution by which the tortfeasor is liable to pay whatever the respondents and the TOCs agree as between them to be reasonable compensation, such an agreement does not, in my judgment, necessarily bind the appellants, as tortfeasors, to pay the agreed sum to the respondents. It is not open to a party to dictate to the whole world the extent of tortious liability and what is reasonably foreseeable and not too remote in order to achieve what it regards as a satisfactory contract with a third party. It could lead to ever more ingenious attempts to attribute possible losses to a tort and would be inimical to the simple solution desired. There could be sound contractual reasons for providing passengers with compensation, for example, for the consequences of lost business, or medical appointments, or job interviews but such consequences are likely to be too remote to charge to a third party in tort.

70.

Making a contract does not confer a licence on the respondents to charge to a tortfeasor whatever types of financial loss, and whatever quantification of financial loss, is included in the agreement provided it is reasonable between the contracting parties. That is particularly so when the terms of the contract are influenced by "government policy aims", as stated by Mr Angus. A policy to impose wide ranging damages on tortfeasors on the roads could be provided by statute but should not be imposed by a contract between a government agency and a third party.

71.

Courts have been prepared to analyse contractual arrangements between the victim and a third party when deciding the extent of the tortfeasor's liability...With differing results, that was done in Dimond and Sandeman. In Dimond, for example, the claimant chose to contract with a specialist vehicle hire company. The additional cost could not be recovered from the tortfeasor. It makes no difference, in my judgment, that the respondents are monopoly suppliers of the use of railway tracks and there cannot be said to be a market. The courts will intervene to assess the reasonableness, as between the claimant and the tortfeasor, of the losses specified in the contract between the claimant and a third party.

72.

The extent of the appellants' liability should be determined on ordinary tortious principles. The issue is not concluded because the loss to the respondents is a direct consequence of the tort. The requirement of reasonable foreseeability must also be established (Wagon Mound). In deciding whether the damage claimed is reasonably foreseeable in the circumstances of a particular case it is necessary to analyse the scope of the duty incumbent on the defendant and the remoteness of the damage claimed. What is reasonably foreseeable cannot in my judgment be determined without having regard to remoteness. The two must be considered together because remoteness is a factor in deciding what is reasonably foreseeable. Whether a particular head of damage is reasonably foreseeable depends in part on how remote it is.

73.

It is in my judgment too simplistic in circumstances such as the present to say that, if a kind of loss, financial loss, is reasonably foreseeable to one who causes physical damage, all financial loss foreseeable is recoverable. Whether one conducts the analysis by considering what is "reasonably" foreseeable or by other tests specified in the cases cited in this judgment, analysis is necessary, and a judgment made.

74.

Differing tests appear in the cases. The scope or extent of the duty was considered in Caparo Industries and Heyman, a value judgment considered in Kuwait Airways, the demands of justice in Naxos, reasonableness in Peabody and Voaden, policy in Metrolink and Spartan Steel. Analysis is required and a judgment made. The court is the arbiter of what loss should be imposed. As Lord Morris of Borth-y-Gest stated in Dorset Yacht Co Ltd v Home Office [1970] AC 1004, at 1039C:

"If the test as to whether in some particular situation a duty of care arises may in some cases have to be whether it is fair and reasonable that it should so arise the Court must not shrink from being the arbiter."

75.

That applies equally to decisions on reasonable foreseeability and remoteness. In performing that task, an objective test is applied to what is reasonably foreseeable. The answer should not depend on how imaginative a particular tortfeasor is, or on whether the driver was the haulage manager, or, for example, a recent immigrant who may have little idea about denationalisation and its effects.

76.

When making that judgment, Rust and West Leigh are the cases most helpful to the appellants because losses claimed to result from fear of the risk of future events was held not recoverable. However, such a principle cannot inflexibly be applied in wholly different circumstances (Ehmler). The courts have been prepared, in appropriate circumstances, to consider loss of income from a loss of reputation resulting from the tort (Andreae) and loss of profit on intended sales (Muirhead).

77.

Circumstances not dissimilar from the present case were considered in Metrolink and the different conclusions reached, following what was, with respect, careful reasoning on each side, illustrates the difficulty of deciding on which side of the boundary the present case falls.

78.

It is in my judgment too simplistic in circumstances such as the present to say that because a kind of loss, financial loss, is reasonably foreseeable to one who causes physical damage, all financial loss agreed between the victim and a third party is reasonably foreseeable. Had there been no such contract, analysis of the headings under which the alleged loss is claimed, and the manner in which it is calculated, would be necessary. The existence of the contract does not, in my judgment, remove the need for such analysis.

79.

It is accepted that the sums claimed are, as between the respondents and the TOCs, reasonable. It does not follow that the figures must be accepted because the respondents are monopoly suppliers of the use of rail tracks and there is no market against which to judge the reasonableness of the figures. However, schedule 8 has obviously been drafted responsibly and with a view to achieving a fair result, in the public interest, as between the parties to it. There is a public interest in a punctual and reliable train service.

80.

The care taken with the exercise must weigh substantially in the balance when deciding whether the public interest in making just provision for compensation by tortfeasors is also served. On the other hand, there are, in my judgment, limits to the influence that "government policy aims" (Mr Angus) reflected in an agreement made between Network Rail, a public body, and a third party should be imposed upon a tortfeasor not party to the contract. Policy aims, as so expressed, do not necessarily determine the extent of tortious liability.

81.

I have come to the conclusion that the appellants should be liable for each of the heads claimed, that is the societal rate component and the MRE component.

82.

It was reasonably foreseeable that, if the respondents' apparatus was damaged, the services of the TOCs, and their value to the public, would be diminished and that arrangements would have been put in place by the franchising authority to penalise the TOCs for the diminution in their services. Two contracts are involved, the contract between the respondents and the TOC, and the franchising arrangement between the franchising authority and the TOC, but the complexity does not render the resulting loss to the respondents too remote from the physical damage.

83.

The MRE component is, in my judgment, also recoverable in the circumstances. It depends on a medium to long term assessment of passenger choices over the network. Whether the "tipping point" of deterring potential rail passengers is reached depends not only on the disruption caused by the tort itself but, as claimed, on an assessment of potential passengers' fears that there will be further disruptions in the future. Provided a genuine attempt has been made to assess future loss of income from this cause, and on the evidence I am satisfied it has, a bar is not placed on recovery by reliance on passenger psychology with its fears of a repeat of the index event. That is enough to decide the issue in this case.

84.

To put it in that way does, however, demonstrate the door that is opened by claims of this kind. When deciding on the "tipping point" for passengers, should account also be taken of other aspects of passenger psychology, for example, the greatly increased price of petrol which would discourage transfer from rail to road or to overcrowding on trains, which would create for passengers a tipping point quite unconnected with the tort? Moreover, if schedule 8 included, in the context of good customer relations, provision for payment of compensation to TOC passengers for the financial consequence of their missing a business appointment or a job interview,, that would not normally be recoverable from a tortfeasor. Analysis of schedule 8 may be required to determine the extent of tortious liability.

85.

In this case, the appellants have not sought to challenge the detail and further issues do not arise. I mention these to demonstrate the problems that arise from a principle that parties to contracts, such as Track Access Agreements, can themselves determine the extent of the tortious liability to one of them of a third party. I consider the appellants' arguments on remoteness to be far from absurd. They are soundly based and cogent but, accepting the care with which calculations have been made, the MRE component in schedule 8 of the Track Access Agreements falls on the side of recoverability.”

Moore-Bick LJ analysed the first instance judgment and then highlighted the “root” of the Appellants’ Counsel’s arguments as being:

Jackson LJ reviewed the law in some detail, including the Baltic Surveyor and CR Taylor cases referred to in the other judgments. He went on at Paragraphs 145 and 146 to summarise relevant principles:

In applying those principles to the appealed case, he went on:

“149.

It is plainly foreseeable that if railway lines are damaged, Network Rail will suffer a loss of revenue. There is nothing remarkable about the revenue losses which occurred in this case. If the main line running from London Kings Cross to the North of England and Scotland is put out of action for most of the day, it is hardly surprising that Network Rail suffers a loss of revenue of about £1 million. Nor is it surprising that if a smaller railway line is put out of action for five days, Network Rail suffers a loss of revenue of about £127,000.

150.

In my view, this action should be characterised as a simple claim for loss of income consequent upon damage to revenue earning property. This is a well established category of recoverable economic loss. Network Rail's loss is the direct consequence of physical damage and it is a type of loss which is readily foreseeable.

151.

Obviously a negligent driver approaching a railway bridge or level crossing, or the employer of such a driver, will know nothing of the detailed arrangements existing between Network Rail and the TOCs. Nor does he need to possess such knowledge, in order to be fixed with liability for Network Rail's loss of revenue while the rail track is unusable as a result of negligent driving.

152.

The present litigation has involved a minute scrutiny of the formulae and calculations which are used to assess compensation for TOCs in respect of periods when railway lines are unavailable. There is nothing unusual or untoward in any of those formulae or calculations. It is hardly surprising that they are complex.

153.

Absent some exceptional circumstance or obviously unreasonable feature in the claimant's business arrangements, in my view it is not appropriate for the court to explore in detail the build-up of any loss of revenue following damage to revenue generating property. It is sufficient for the claimant to prove that the loss of revenue has occurred.

154.

The law of tort should, so far as possible, be clear and simple. This court should not superimpose a requirement for expensive legal inquiry upon categories of case where there is an established entitlement to recover economic loss.

155.

The exploration of the build-up of Network Rail's revenue loss in the present case has not revealed any exceptional circumstance or unreasonable feature of Network Rail's business arrangements. The resultant figures do not appear surprising or disproportionate. It is now accepted that Network Rail's schedule 8 payments to the TOCs are arrived at on the basis of a reasonable assessment of the TOCs' losses flowing from the temporary closure of the lines. It is true that Network Rail's payments under schedule 8 include compensation for the TOCs' future loss of fares. That circumstance, however, cannot detract from Network Rail's claim against the defendants for loss of revenue.”

Although there was much debate to the contrary, I discern little real (let alone practical) differences between these judgments, expressed though they are on occasion in different verbal terms. It is beyond doubt that they were deciding that in principle (and indeed on the facts of the Conarken and Farrell Transport cases) that:

It is clear that the judgments of Moore-Bick and Jackson LJJ do not proceed upon the basis that there is some "independent or overriding principle that damages must be reasonable as between the claimant and the defendant" (see for instance the latter at Paragraph 118). What they both do however is to say that what is reasonable between the claimant and defendant may be relevant to determine the selection of an appropriate measure of damage or may come into play in considering causation or remoteness of damage (see for instance Paragraph 145 of the latter). This is a distinction which is of some relevance. All that Pill LJ was saying is that merely because there is a contract between the innocent claimant (for instance in negligence cases) and a third party which obliges the claimant to pay a certain level of compensation for delays or damage caused by the defendant, that does not mean, automatically and without some analysis and factual consideration, that the claimant can recover that level of compensation, no matter how high that level may be.

In case after case (in negligence), the courts over the years have analysed the elements or exercises which need to be established or done: the existence of duty, scope of duty, reasonable foreseeability, remoteness and measure of damages, linked necessarily by causation considerations. I agree with Moore-Bick LJ that there is no overarching or separate principle which requires damages to be reasonable as between claimant and defendant. Otherwise, one might descend into arguments that it is disproportionately unreasonable, as between claimant and defendant, that a poor defendant should have to pay to a better off claimant the full amount of loss otherwise established. The negligent defendant whose poor driving paralyses a highly successful claimant whose earnings run into hundreds of thousands or possibly millions of pounds every year can, currently, be expected to pay damages for the actual loss of the particular claimant’s earnings, even if the claimant is, say, a banker and, if it be the case, bankers’ earnings (with bonuses) are considered to be extravagantly high: there is no principle that the paralysed banker should not be compensated for the earnings which he or she can prove would have been earned but for the accident or that the Court should reduce those damages to a level which might be considered to be reasonable as between claimant and defendant.

That said, I put to each Counsel a hypothesis as to how the Court should go about addressing say a Schedule 8 type calculation which produced a loss of £1 million per minute of delay caused by the negligence. It would be common ground that, by some route or another, the Court would feel compelled to find some way of finding that the otherwise recoverable consequential loss could not be based on a rate of £1 million per minute. There was some difference as to how the Court would get to the appropriate answer. As indicated above, there is no separate element or exercise which the Court has to find or go through and it is within the established elements of the scope of duty, reasonable foreseeability and remoteness, coupled perhaps principally with causation that the answer will usually be found. In a given case for instance, the Court may well be able to find that the negligence simply has not caused a loss of £1 million per minute of delay but that a large element of it (at least) is attributable to something other than physical damage or consequential delay caused by the negligence. The existence of what appears to be a ridiculously high level of consequential loss payable by the claimant to a third party will act in effect as a red light to the Court to warn it that there may well be explanations based in fact which point to it not being related to the given negligence and its consequences. The claimant driver of a Mini car negligently written off by a defendant can often recover as an appropriate head of loss the cost of hiring an appropriate car whilst waiting for the replacement car to arrive but the Court will refuse compensation for the cost of hiring a Rolls Royce because the cost, over and above what it would have cost to hire a Mini, would not be attributable, reasonably if at all, to the negligence; the head of loss would be recoverable but damages would be limited to what it would have cost to hire a Mini. The main reason for the decision in the House of Lords in Dimond v Lovell [2002] 1 AC 384 which involved the refusal to allow car hire damages was that the particular agreement for the hiring of the car (which allowed for additional benefits) was unenforceable. There is no reason to think that this is exceptionable but the Court, but for the unenforceability of the hiring agreement, would otherwise have allowed for the cost of an ordinary car hire (see, for instance, Lord Hoffman at Page 402 G-H). Such a reduction in damages could therefore legitimately be attributed to either a straight causation test i.e. the cost of the "additional benefits" was not causatively attributable to the tort or that the claimant had failed to mitigate her loss by acting unreasonably in entering into a hire agreement which allowed for additional benefits not attributable to the tort.

It is at least theoretically possible that in an appropriate case the Court could find that the scope of the tortious duty did not extend to cover a type of loss which is likely either to be unconnected with any breach of the duty or which produces a level of loss which is so far beyond what might be considered to be reasonable that the duty should not extend at all to such loss. At the very least, that must be a rare case, if it ever happens at all, because when considering the scope of duty it will or will not to a large extent encompass the types of loss which are reasonably foreseeable and which are otherwise not too remote. What the Court of Appeal decided was that the loss of revenue as such was a recoverable head of loss and it is difficult to see how it could be a totally irrecoverable head of loss just because the amount was thought by one party to be excessive. This is partly because it is difficult to establish criteria by which a loss of £5X as opposed to £X is "excessive" or "unreasonable". One then gets into a debate as to whether the Court should or can allow an award of damages at a level which is not "excessive" or "unreasonable".

What one can extract from three judgments, generally, and in relation to the position of Network Rail, is, in my judgment, as follows:

I should preface my remarks and analysis of the facts with an assessment of the witnesses. Mr Angus is an extremely experienced and intelligent person whose integrity is not impeachable and indeed was not challenged; his knowledge of the history of and reasons for the development Schedule 8 and the PDFHs is encyclopaedic and informed; I have no difficulty in accepting his evidence. There were some objections to the admissibility of limited parts of his statements, largely on the ground of him supposedly giving expert evidence; I do not consider that the objections were well founded on this ground but, even if they were, he is highly expert and I can have regard to what he says in this regard but weighing it up on the basis that, although he presented his evidence objectively, he is not independent. Mr Kenney, who gave evidence about delay attribution, was straight and honest and he was a reliable witness, whose evidence was both credible and indeed was not challenged in many respects. Mr Smith, of Aviva, provided fairly tangential evidence about two TOCs’ claims for lost revenue loss against Aviva’s insured clients (the Handy incident and an accident at Barrow-on-Soar); although he was honest, he was being asked to give evidence which was largely irrelevant, unhelpful and, though no fault of his, not supportive of inferences which the Court was asked to make.

As for the experts, Mr Segal, called by Network Rail, was, head and shoulders, the better of the two. He has had experience over many years of Schedule 8, its background, rail demand research, modelling and PDFH. His evidence both written and oral was comprehensible and articulate, as well as being well researched. He has himself contributed by way of papers and research in the areas considered in this case.

Sadly, the same cannot be said about Mr Palmer who, though honest, I would go so far as to say was not really an expert in the key area with which this case is concerned, namely the rationale, background and workings of Schedule 8. Indeed, he as good as accepted that, before being involved in the current case, he knew little or nothing about Schedule 8, saying in evidence that he was still “learning” about it; he acknowledged Mr Segal’s expertise and experience in the area as established and way above his. He did not help himself by wrongly describing himself in his first report as having “undertaken several rail and public transport projects”, including:

“One (for SDG/ATOC) [which] investigated the reasons why the demand for rail travel was higher than that forecast by the PDFH models. It identified four key weaknesses in the existing approach”.

He accepted in evidence that he had simply attended a workshop which was part of the study and that he played no part in the report or a later update. Apart from the fact that the report was not nearly as helpful to his thesis as he suggested, it is clear to me that he was pretending to experience which he did not have, such that, although not deliberately disingenuous, his evidence in this respect was misleading. He spoke in his first report of a 1995 report, “Big Delays”, without having read it, going on to give evidence that there was no evidence in a number of respects on various topics when that report gave some such evidence against his thesis; this points, if not to incompetence, to a lack of basic knowledge about the subject in issue in the case. In too many respects, he had to make concessions that what he had said was wrong. This was partly from one report to another; these reports were heavily corrected (and not just with typing errors) which suggests that they were inadequately or at least carelessly prepared; for instance he withdrew a criticism of performance benchmarks which related to the Schedule 8 loss calculation and later he conceded that he could no longer support his assertion that Schedule 8 could be criticised as being both for compensation as well as a penalty type incentive purpose (see his first report, Paragraph 2.1 Paras 55, 57, 84 and 119 and Day 6 Transcript Page 81); he tried, somewhat unsuccessfully to meet the criticism by saying that his initial views were based on “uneasiness”; the problem for him was there is nothing added into the Schedule 8 formula for incentive purposes and he did not seem to have picked that up until giving evidence. He referred to reliance on anecdotal information from work colleagues (not referred to in his reports), which was unhelpful and represented an amateur approach to a serious subject. There were numerous inadequacies in his evidence listed in the Closing Submissions of Mr Onions QC and Mr Drake, the large bulk of which I agree with.

In my view, little or no weight can be given to the evidence of Mr Palmer save where he agreed with Mr Segal or where he made concessions. Concessions he certainly made, for instance in a number of respects accepting that approaches relied on within the PDFHs and by Mr Segal were reasonable or not unreasonable.

The quantification of loss in each case is agreed, not only as to the requisite remedial work or related loss but also as to what the Schedule 8 calculations, as related to the actual delays attributable to the five incidents, produce. There is, rightly, no issue but that Network Rail has actually suffered as a revenue loss the sums produced by these calculations by way of payments or credits to the relevant TOCs.

There is no doubt that Schedule 8 in its various iterations has been drafted responsibly, with care and with a view to achieving a fair result as between Network Rail and the TOCs. They are and were all approved by the ORR and periodic and interim reviews were carried out (2000, 2003, 2008 and 2013) as well as a review of the performance regimes in 2005. These reviews were carried out in consultation not only with Network Rail and the TOCs but also with other interested parties such as Transport for London and the Department for Transport. These reviews involved re-calculations of payment rates under the performance regime and, sometimes, substantial changes such as the removal in 2003 of the "societal rate" which was in place under the TAAs in the Conarken and the Farrell Transport cases. There had been a genuine attempt to assess future loss of income resulting from delay. There is no doubt and in any event I find that the performance regime rates which are the subject matter of Schedule 8 and Appendix 1 to the TAAs were, as between Network Rail and the TOCs reasonable.

There is no real and certainly no reliable evidence that the TOCs truly considered at any stage that these rates were unfair or unreasonable or that they do not effectively provide for compensation for delays attributable to Network Rail’s risk factors under the TAAs. I would have expected the Defendants to call witnesses from the TOCs if they had wished to pursue this assertion. Instead, they sought to suggest inferentially that, because in relation to two incidents (the Handy incident and another at Barrow-on-Soar) the TOCs in question (Midland Mainline and East Midland Trains) presented claims in effect to the insurers of the negligent drivers for loss on the day of the two incidents in question, they must have thought that there was no loss after that day. The claim calculations in question were at best basic and facile and involved a rough and ready and hardly reasoned comparison of income between the day or week of the accident and the same time the previous year. Both claims were withdrawn and they were not expressed obviously or at all to be on the basis that there was no other loss. At best, those claims suggested that the TOCs thought that there was some loss on the day in question; there may well have been, given for instance the length of the delay following the Handy incident and the probability that some actual or potential passengers would have made other arrangements to travel or not travel at all on that day. One of the most obvious reasons why these TOCs may well not have pursued these claims or claims for future revenue losses is that they could well have appreciated that they would be compensated through Schedule 8 by Network Rail for the consequential revenue losses flowing from the incidents so that they could not recover both from Network Rail and the negligent drivers directly as there would be double recovery. There is no hint or suggestion in the disclosed correspondence and documentation that whoever it was who prepared these claims believed that the Schedule 8 basis of calculation was wrong or wrongly founded or that there was no longer term effect. Given that the TOCs have actively participated over the years in the process by which the money rates for Schedule 8 are established and that nothing of relevance has been disclosed which suggests that the TOCs are in fundamental or significant disagreement this suggests that the TOCs and Network Rail were content not just with the money rates but also with the machinery and formulae within Schedule 8.

Although relied on in several contexts, the Defendants sought to rely on the “pleadings” and evidence submitted in an expert determination process in 2005 between London Eastern Railway (“LER”, which trades under the name "one") and Network Rail which sought to have various rates within Appendix 1 of Schedule 8 of their TAA amended. This was referred to as the "One determination". In it, LER essentially argued that the figures in Appendix 1 referred to in Schedule 8 did "not accurately reflect marginal revenue effect of delay" or "the introduction of a new more robust timetable". The actual determination allowed some of the figures to go up and others down. The background as suggested by LER was that, because Network Rail was performing better than the benchmark levels, it was having to pay more than it should have done and the benchmark levels had not been adjusted to reflect changes to the operating timetable. It relied, in one sense against itself, on an incident which had occurred when overhead wires near Liverpool Street were damaged causing what was obviously extremely severe disruption to both mainline and suburban services; it said (Paragraph 5 of Amended Statement of Case) that the sums which it received for that day of disruption were £0.9 million which was said to have exceeded "the total farebox revenue" on those routes that day, going on to say that this "incident cannot realistically represent the real losses to "one" caused by Network Rail’s performance on the day" and that this occurred because the MRE was “too high”.

This "evidence", such as it is, and the evidence relating to the two other incidents relied upon, amongst other things, is said to suggest or lead to the inference that the TOCs do or did not think that there was or is any long-term effect on rail used by passengers as a result of single road on rail or other comparable incidents. However, this “evidence” does not contain any such expressed thought process and it is impossible to infer (from the fact that they do not actually in the claims or in the One determination process say anything about the long-term impact of incidents which delay or disrupt train operations) that they all or indeed any of them consider that there is no long-term impact. The very fact that they subscribed to the review processes relating to the TAAs and the performance regimes, which do assume that there is likely to be a long-term impact, would wholly undermine any such inference.

The above leads me to the conclusion that, as between Network Rail and the TOCs (in this case and generally), the sums claimable under Schedule 8 are reasonable. There is no real evidence that the TOCs in general or the TOCs affected by the negligence of these five defendants in particular had or have any objection to the applicable MREs or delay multipliers; they actively participated in the processes by which they were negotiated and agreed. It has not been suggested, rightly, that the Schedule 8 regime is, as between Network Rail and the TOCs in the nature of a penalty; that is because this regime is not related to breaches of contract. I consider that Schedule 8 produces a fair result in the public interest, there being a public interest in a punctual and reliable train service. There is no doubt that great care was taken in the exercise of drawing up the relevant versions of Schedule 8.

The financial claims as agreed (in terms of quantification and calculation) are not obviously excessive. For instance, the Strood incident (Rob Hatfield Ltd) generated, in relation to South Eastern services, 16 train cancellations (in whole or in part) and 121 other trains were affected and there was an income loss of some £41,000; apart from the Handy claim, the other claims are for even less than that. Even in respect of the Handy claim, at over £259,000, there clearly was major disruption on a busy main line with 38 trains cancelled in whole or in part and 92 other trains affected over a (near) seven hour period, during which for three hours both lines (up and down) were closed.

To adopt Jackson LJ’s approach (Paragraph 153) in the Conarken and Farrell Transport cases, there are no exceptional circumstances or obviously unreasonable features in the business arrangements as between Network Rail and the TOCs. It follows that it is "not appropriate [for this court] to explore in detail the build-up of a loss of revenue following damage to revenue generating property" and it "is sufficient for [Network Rail] to prove that the loss of revenue has occurred". Network Rail has proved that reasonably foreseeable loss has been incurred as the direct result of the negligence of these defendants; it has established that such loss is not too remote so as to be irrecoverable; the loss of revenue from the TOCs is clearly a properly recoverable measure of loss. The Court should be slow to encourage the sort of “minute scrutiny of the formulae and calculations” which are the subject matter of Schedule 8. I am satisfied that the evidence in this case has not revealed any "exceptional circumstance or unreasonable feature" in what lies behind Schedule 8.

In one sense, that should be the end of this case. This more general or less detailed approach is the right one in a case such as this. If there are what one might call "big picture" points to make about the make up of or elements within the contract between in this case Network Rail and the TOCs (particularly within Schedule 8 which fixes financial responsibility for delay which is the responsibility as between Network Rail and the TOCs), they of course can be made and must be considered by the court. That is what both Pill and Jackson LJJ say in effect. It would be absolutely consistent with what Pill LJ said at Paragraph 80 to analyse whether for instance there was a non-compensatory penal or punishment element in Schedule 8 which punished Network Rail for road on rail incidents; that could be said not to be reasonably foreseeable or not within the scope of the duty owed by negligent drivers such as there are in these cases and it could also be said that at least the loss reflected by such a penal element was not in truth caused by the negligence. There are no such "big picture" points in this case.

Lest it be relevant, I will however review the detailed challenges to the Schedule 8 process. This is in the context that the Schedule 8 process is deployed by Network Rail primarily because, in conjunction with the delay attribution process described earlier in this judgment, it can determine both how much delay has been caused by the single road on rail incident as well as the precise loss of revenue suffered by Network Rail. In addressing the detailed challenges, I will necessarily identify other factors which highlight the reasonableness of the Schedule 8 approach and assessment deployed for the 5 cases here.

Although, as will be apparent from the rest of this judgment, nothing will turn on it, the issue of burden of proof was raised. In the ordinary course of events in litigation or arbitration, the burden of proof rests on the claiming party, here Network Rail, to prove its case. However, the law (and practice) recognises that for certain types of defence the burden of proof rests on the defendant. So, for example, a defendant will need to prove that there has been a failure on the part of a claimant to mitigate its loss or that there has been contributory negligence on the part of the claimant. There is nothing untoward about this.

Where, as here, Network Rail has established a duty of care, sufficient scope of the duty of care to cover revenue loss, reasonable foreseeability that revenue loss will or may be suffered by it in relation to the TOCs, the measure of loss as including such revenue loss and causation between the negligent breach of duty and the revenue loss in question, the burden of proof, in my judgment, for establishing that some one or more features of the otherwise established revenue loss was or were unreasonable or otherwise untoward rests on the defendant which is seeking to establish it. In this case Network Rail has in essence established its case and the burden of proof to undermine it in this way rests on the defendant.

The detailed scrutiny to which Schedule 8 and the related MRE calculations was subjected involved not simply mathematics but also extensive analysis of academic papers, presentations of one sort or another to the ORR and speculation as to whether academics could or should have had regard to earlier studies. There was repeated reference to papers by two well respected academics, Professor Wardman and Dr Batley, with whom Mr Segal has worked and co-operated on occasion. Their papers were analysed in cross-examination and in oral and written submissions in massive detail. Neither of them was called as a witness and the Court is being asked not simply to read and understand the reports (which it has done) but also to draw inferences as to what they might have said if asked about some of the issues specifically arising in the current case. The Court should be slow to draw such inferences.

To clarify their challenge, the Defendants through their Counsel provided on Day 4 of the trial their "Headline Points" upon which they rely to challenge the reasonableness of Schedule 8:

“1.

Exaggeration because of use of unrealistic lateness multipliers.

2.

Exaggeration because [it is] not appropriate to use Schedule 8 for [a] single incident.

3.

Exaggeration because road on rail incidents are very infrequent in passenger experience."

In addition, although this does not go directly to Schedule 8 itself, there are challenges because there is said to be "unfairness to the motorist in the systems of attribution and claim formulation". I will address each in turn, although they partly overlap and the reasoning and findings which I adopt and make below in relation to the Headline Points should be read together.

Applying the judgments in the Court of Appeal, one needs to determine whether there is something exceptional or obviously unreasonable in the use and application of the Schedule 8 formulae.

This point was summarily amplified by Counsel as follows:

“Schedule 8 exaggerates the true loss generated in the long term by a deterioration in the general level of reliability, because the AML elasticities implied by the conventional approach of applying a lateness multiplier tend to be much larger [reference: Wardman & Batley 2014], or very much larger [reference: Mr Palmer, having regard to Batley 2011 with Batley 2008 in the context of the other RP [Revealed Preference] evidence], than directly-estimated AML elasticities. The lateness multipliers are largely derived from SP [Stated Preference] research, which provides information on people’s ‘valuations’ (how they say they regard the matter, not how they actually behaved). Econometric RP research and analysis derives directly-estimated AML elasticities from observations of people’s actual behaviour when the general level of reliability has changed."

If one had not participated in the trial, one would be no wiser as a result of this amplification. AML is an acronym for "Average Minutes Lateness". "Elasticities” are changes in demand, relating to journey time, usually expressed by way of a percentage (as indicated in Paragraphs 115 to 116 of Mr Angus’ first statement to which the Court was referred for definition). The lateness multiplier comes into the MRE calculation (for which see earlier in this judgment – Paragraphs 20-24); the MRE figure produced by the calculation for any given Service of Group is part of the Schedule 8 formula. Mr Angus explained lateness multipliers as in effect being “how much worse – how many times worse – is lateness compared with an equivalent amount of timetabled journey time?” (Paragraph 113 of his first statement).

What this area of the dispute is primarily about is whether weight should have been placed in the fixing of the multipliers used to create the MREs on the basis of Stated Preference (SP) survey work referred to in the various editions of the PDFH to which this case relates. It is common ground that the lateness multipliers were largely (albeit not entirely) derived from SP survey work. This is challenged or criticised on behalf of the Defendants on the basis that passengers’ responses to SP surveys are not reliable because they do not correspond to what passengers actually do. The Defendants argue that the assumption that the findings from SP research correspond to real life and the actual behaviour of passengers derived therefrom is “inherently improbable and…involves a leap of faith which econometric research has now shown to be unjustified" (Paragraph 17).

I bear in mind and adopt Pill LJ’s comment at Paragraph 83 of his judgment in the Court of Appeal which actually addressed a number of the points highlighted in this case and on this particular topic:

“The MRE component is, in my judgment, also recoverable in the circumstances. It depends on a medium to long term assessment of passenger choices over the network. Whether the "tipping point" of deterring potential rail passengers is reached depends not only on the disruption caused by the tort itself but, as claimed, on an assessment of potential passengers' fears that there will be further disruptions in the future. Provided a genuine attempt has been made to assess future loss of income from this cause, and on the evidence I am satisfied it has, a bar is not placed on recovery by reliance on passenger psychology with its fears of a repeat of the index event. That is enough to decide the issue in this case.”

It is of some relevance that, although the recoverability of the MRE and the concept of the tipping point (for which see elsewhere in this judgment) were clearly raised by the Defendants in the earlier case (by the same solicitors and insurers, and on appeal the same leading Counsel, solicitors and insurers) and addressed by the Court, these self same points are raised again in the current proceedings by seeking to delve down deeper than they were willing to go in the last set of proceedings.

There is no doubt that the MRE, the PDFHs (Passenger Demand Forecasting Handbooks) and consequently the TAAs (in particular Schedule 8) are predicated upon the basis that delays or changes to timetable journey times can result in medium to long-term loss of passenger business. The Schedule 8 calculation is based upon the impact after the day upon which the incident happened. Thus, although there is bound to be (in most cases of significant road on rail incidents) a loss on the day, that, as such, does not feature in the Schedule 8 assessment because that assessment looks at the longer term impact. The loss on the day can include for instance total or partial refunds to passengers (who have bought their tickets who either can not use them or suffer serious inconvenience as a result of the delays), additional management or staff time, extra salary costs (for instance over-time to the driver, guard, ticket inspectors or station staff), wasted fuel, the cost of buses (to take passengers by road to other stations where they can get onto other trains) and the loss of custom on the day of passengers who turn up at the station to be told that there is likely to be a substantial delay and who therefore do not buy tickets. Of course, there may also be passengers who are either not fazed by the delay (for instance those who are not in a hurry or who have a good book or down-streamed material to look at on their smart phones) or who have season tickets and have no other practical choices other than to stay on the train which has been delayed. In this latter context, there may be less loss of revenue than might be imagined in relation to the day of the incident in question because many ticketholders will have season tickets or have booked cheap fares in advance.

There is no doubt and there is ample evidence to find (which I do) that passengers regard reliability of timetabled train services as important. Indeed a relatively recent National Rail Passenger Survey found that: "Punctuality remains the biggest single influence on satisfaction of passengers". That is not to say that there are not other factors, such as price, which are also important to passengers. Although it probably matters not in relation to the current debate in this case, rail performance data distinguishes between punctuality which relates to services arriving on time and reliability where services are cancelled (see for instance Web based Transport Appraisal Guidance – Tag Unit A1.3 Paragraph 6.5.7). As in the current cases, significant delays lead not only to trains being delayed at their arrival points but also to cancellations.

When one comes to consider overall demand for passenger rail services, there are a number of factors which influence passengers. For instance, a variety of overall economic factors may influence the amount of rail travel, these factors including employment or unemployment levels, competition for instance from bus services and the price of petrol or diesel. Thus, serious levels of unemployment or high petrol prices may tend to lead to less or more use of the railways (see e.g. PDFH Volume 4.1 Paragraph B1.1 and following). When it comes to considering the impact of price or "fares elasticities”, the PDFH lists a number of factors which might influence travel, for instance the higher the price the more sensitive travellers might be expected to become (ibid - Paragraph B2.3). Chapter B3 of the PDFH addresses Generalised Journey Time (GJT), it being recognised that the "timetable related service quality attributes undoubtedly have an important influence on rail demand, and rank alongside fare and external factors as the prime determinants of the volume of rail travel" (Paragraph B3.2). Different GJT elasticities can be and are applied for different rail services and service groups. Indeed different recommendations are made by the document in relation to season and daily tickets for different areas (London Travelcard Area, London Travelcard Area to and from the South East, Rest of Country to and from London Travelcard Area, Non-London and Airport Access).

Chapter B4 relates to "reliability” and emphasises the "increasing awareness of the importance of reliability to travellers" going on in Paragraph B4.1:

“Network SouthEast undertook a major research project in 1988/9 and this was included in the significant updating of the PDFH in November 1989. It broadly confirmed earlier conclusions but also suggested that commuters are more sensitive to the variability of delays than to the average delay itself. Whilst plausible, some inconsistencies were found in its application and so these results have been removed from the PDFH. Since 1994, the PDFSS has undertaken a number of research projects to develop the area. In particular, in February 2000, the seminal work of Bates et al appeared…”

It is clear that the PDFH, as updated from time to time, has been based on genuine research from reliable sources and academics. Elsewhere in the Handbook there are listed all the research and other papers relied upon and they are explained. For instance so far as reliability is concerned, the evidence relied upon is considered and analysed in Chapter C4 where the Introduction states as follows:

“Evidence from surveys has consistently suggested that passengers value reliability very highly. Commuters in the South East have put it at the top of the list of priorities. Yet it has been found to be one of the more difficult areas in the PDFH to research.

Nearly all of the evidence is based on stated preference studies because of the difficulties in isolating and observing the effects of actual changes and the service reliability. The first study of InterCity service reliability undertaken in 1984 by the Cranfield Institute of Technology…suggested that passengers value each minute late as equivalent to 2.5 minutes of in-vehicle time. Subsequent studies have broadly confirmed this finding. Attempts have been made to value the variability of delay and big delays. And the one revealed preference evidence has looked at the effects of runs of poor reliability at Euston. From these, questions have been raised about the approach to assessing reliability. In February 2000, the seminal contribution of Bates et al appeared which attempted to answer many of these questions. Subsequently a further study by MVA has been produced…”

The reference to “big delays” relates to a survey done in 1995 by SDG Research. Its aim was to understand and quantify "the impact of big delays on future rail patronage and revenue , and "big" delay being 30 minutes or more on a suburban route and 60 minutes or more on an inter-urban route" (Page 1). The study comprised a Stated Preference and Attitude "survey of rail travellers, an omnibus survey of the general population, a telephone survey of big delays sufferers and a survey of trains affected by engineering works". The "Impact of Suffering a Big Delay" was summarised at Paragraph 9 of the Executive Summary:

“Of our sample of customers who had suffered a big delay, 3-4% subsequently stopped using rail altogether and a further 5% reduced their use of rail following the incident.

Taking a weighted average of responses, this converted to a loss of 5-6% of rail trips which the sample would have made without the incident.

The experience had also had the effect of worsening the customers perceptions of the operator with 39% saying that because of the incident they now think of the operator a little or a lot less.

This reflects a generally poor opinion of the way problems are dealt with. Typically, 40-50% of customers said they thought the problems were dealt with not very or not at all well, though this appears to vary substantially from operator to operator.

As further evidence, in the SP & Attitude Survey, information on disruption and the staffs’ ability to cope with disruption were the worst rated of eleven quality of service factors.

There was also strong evidence that how well the delay is dealt with affects the customers subsequent response as illustrated below. Length of delay, where the delay occurs (on the train or at the station) and the reason for the delay or also have their effects on customers’ subsequent response to the delay.

In terms of the longevity of the impact of the big Delay, there was no real evidence that the effect on customers had lessened even after two years have passed since the incident."

Even judges are not immune from rail travel and, indeed, I made it clear to the parties that I was a regular user of the train services, albeit not on the particular lines on which any of the particular incidents happened. Disregarding any rail experiences which I may have had, it is clear that at least on a broad brush basis the results of the "Big Delays" research, although now nearly 20 years old, have some resonance in the light of the evidence given in this case, particularly by Mr Segal and corroborated by Mr Angus, to the effect that long delays to rail services, particularly such as those suffered in the five cases with which the Court is concerned, can lead to passengers (in not insignificant numbers) abandoning or reducing use of the rail service over a not insignificant period of time.

There has been a major area of issue between the parties as to the extent to which it was reasonable or unreasonable for there to be reliance within the PDFH on SP research as opposed to RP research, although it is clear from the quotation above (C4 of the PDFH) that some regard was paid at least to one RP study. Revealed Preference, broadly, involves an analysis of choices considered to have been made, usually, by individuals which may reveal what they would prefer to do. In the area of rail demand forecasting, this often involves “control flow” or “econometric” analysis. Control flow analysis involves consideration of “flows” which relates to particular train journeys on particular lines and often involves a comparison against a "control" flow, where the change or development which one is wanting to study has not occurred, with the flow where the change or development has occurred. I suspect that this type of analysis is not particularly relevant in this case. However, the "econometric” approach certainly is said to be relevant here. Econometrics is said by Mr Segal (Paragraph 3.4.12 First report) to be "a long established technique using statistical analysis to establish the relationship between a dependent variable (the case in interest, rail demand and revenue) and various independent variables such as economic variables, fares, journey time and reliability of train services"; he goes on to say that the technique uses data of the different flows and different time periods. He accepts that it can be "very powerful, but must be used with care" (Paragraph 3.4.15), explaining a significant problem as "multi-colinearity" which arises where "two independent variables (that is variables that affect demand) have changed in a similar way historically (or across different flows), then the analysis will not be able to distinguish the effects separately.

Mr Segal explains SP research at Paragraph 3.4.19 (and following) in his first report saying that it was developed "as a quantitative market research technique to value attributes which could not be measured using revealed preference analysis" explaining that British Rail was at the forefront of the development of SP techniques in the 1980s, it having been continually used since then. He identifies at Paragraph 3.4.23 the principal strengths as being:

“- Almost any attributes can be included in the survey provided it is something that the respondents will trade…

- The survey can be designed to ensure a large amount of trading, meaning that precise estimates of parameter values can be achieved at modest cost;

- Checks can be included to test for game playing and illogical responses;

- It can capture data on passenger characteristics such as journey purpose, gender, income and hence estimate values for each market segment."

He then identifies potential pitfalls which call for care to be taken to avoid:

“- The survey gives respondents specific information which they may not have in real life; this may distort responses and certainly means that SP cannot be used to estimate lags in response;

-

There is a risk of over-sensitising respondents to elements which they would not normally notice or pay much attention;

-

Even genuine responses may not be accurate, as how people think they would behave may not be how they would behave in practice;

-

It is essential that respondents are given realistic choices; if they are given the choice between option A and option B when they would really decide not to travel in this context, then the not travel option should be given."

At Paragraph 3.4.25, he sets out how a properly designed SP study can address these possible pitfalls, namely by "cognitive testing" (checking the respondents understand the questions in the way that the surveyor intended) and piloting (with an initial sample of respondents perceiving that the questions are sensible). He explains later in his first report (Paragraph 3.5.19) that in the recent research both piloting and cognitive testing have been deployed in SP surveys and he explains how the surveys are framed such that respondents are presented with, say, 10 choices as to what they might do in relation to short or long delays. He said at Paragraph 3.4.29:

“The continued use of SP for transport applications from 1980s up to the current day demonstrates that in the view of the industry, if properly designed it is a powerful tool giving robust results."

I accept that observation as logical and right.

SP surveys are not carried out in the immediate aftermath of a serious delay. One can readily imagine (as was referred to by several witnesses) what rail passengers might say having just been delayed by 3-4 hours or more, whilst still irritated if not angry; they might well say that they will never travel again with that TOC or on that rail line and on later calmer reflection some of them may not mean that.

There was what seemed at times during the trial to be a micro-analysis of what Professor Wardman and Dr (now Professor) Batley meant in various papers and submissions which they had written or what they might think if they were asked questions which they had not apparently answered. Thus, Mr Palmer thought that these well-known academics were wrong to take into account a March 2005 study (by OXERA) and that it should have been treated as an outlier to be disregarded as at best anomalous. As indicated earlier, I attach little or no weight to what Mr Palmer says; I do not think that he has the expertise, let alone experience or reliability as an expert, to challenge what they have done.

What is of interest in this regard is that Professor Wardman and Dr Batley have never apparently gone into print or spoken publicly to the effect that the more recent PDFHs or Schedule 8 are wrong or unreasonable. What they have concentrated on is, perhaps, a promotion of the efficacy of more recent RP research. Neither they, nor indeed Mr Palmer, criticise as unreasonable or unrealistic the earlier versions of the PDFHs or Schedule 8, judged by the evidence and information available at the time that they were produced. It is interesting in this context that Professor Wardman and Dr Batley both materially assisted in and indeed carried out a Reliability Review for the update of PDFH v 5.1 as recently as March 2013, Mr Segal being also one of the reviewers of the detailed Technical Note dated 21 March 2013. This was considering amongst other things the requisite multipliers and they endorsed the PDFH approach in their Executive Summary:

“Although we could have opted for a direct elasticity approach, given the evidence that has emerged, we feel that on balance the current approach using late time multipliers driven off the GJT elasticity should be retained since it provides slightly more differentiation between markets. "

They looked at many of the previous studies (many of which were considered in some detail during this trial) and it is quite clear that they were and are prepared to go along with the PDFH approach and indeed the use of the requisite multipliers and other data which were to go into the latest version of the PDFH. This is apparently in spite of the interpretation put on and the conclusions drawn by the Defendants’ team about an earlier paper in 2011 by Dr Batley (contributed to by Professor Wardman). It is said that a later paper in 2014 by them contains evidence which significantly or seriously undermines the reasonableness or reliability of the PDFH and Schedule 8 approach. It would be quite wrong to speculate what their views are or would be on the applicability of MREs to serious delays caused by road or rail incidents for the reasonableness or otherwise of the Schedule 8 regime.

In the 2011 paper ("The impact of lateness and reliability on passenger rail demand" Batley, Dargay and Wardman), the Abstract stated:

“To date, forecasts of the demand impacts of lateness and reliability have been derived largely from individual-level models taken at a snapshot in time. The contribution of this paper is to develop a dynamic model of rail demand at the market-level, yielding short and long-run elasticities with respect to lateness. Whereas individual-level models have suggested a high valuation of lateness and reliability, our market-level models indicate a relatively [muted] demand response. Reconciling these findings, we reason that, whilst rail travellers show considerable disdain for experiences of lateness, such experiences will not necessarily dissuade them from travelling by train."

Stated as blandly as that, it may well be that many passengers and indeed the majority of passengers, although irritated and annoyed if trains are late for any reason, will continue to travel on the trains and "just put up with it" in some sort of classic bulldog fashion. It is obviously the case that many people do not actually have a choice; for instance, many commuters to London in the South East, well outside the areas served by London Bus and Underground, Overground or Dockland Railway systems, who have to be in London for work, say between 7 and 9.30 a.m. every weekday throughout most of the year, may well in practice have little choice than to go by rail, partly because going by car means hitting rush-hour traffic and involves expensive car parking and congestion charges and partly because commuters who often suffer from train delays can just take an earlier train than they might otherwise take to avoid being late at work or for appointments. However, this Abstract does not suggest that there will not be a not insignificant minority of passengers overall who will not be willing to use the train service in question again as a result of a particularly disruptive and lengthy delay. In Chapter 2, the authors acknowledge that there may well be a "diverse range of behavioural responses that could potentially arise".

Relevant parts of the Summary and conclusions of this relatively short (12 page) paper are as follows:

“The primary objective of this paper was to report estimates of the elasticity of demand for rail with respect to changes in service performance…This discussion confirmed the relevance of the lateness distribution as an appropriate measure of performance, but drew an important distinction between the first and second moments of that distribution. Whereas the first moment, namely average lateness, has attracted the focus of many analyses of rail performance (both in Great Britain and in other countries), this has been to the neglect of the second moment, which arguably represents the most theoretically-valid measure of reliability. In this way, the discussion exposed significant limitations in British methods, both with the recording and collation of performance data and the specification of models. Last but not least, the discussion of concepts and definitions serve to clarify the focus of the present analysis which, constrained by the scope and detail of British industry data, was directed more at lateness and less at reliability.

Our analysis involved the development of a dynamic econometric model relating rail demand (in terms of ticket sales…) to performance (in terms of…ALM and the Average Performance Minutes (APM), at service group level; and the Public Performance Measure (PPM) at TOC level)…

Lateness and reliability, as characterised by the metrics used, were found to have only a marginal effect on rail demand. This was true for both static and dynamic model specifications. The long-run elasticities for ALM and APM obtained from the dynamic models were slightly greater than those obtained from the static models, whilst the opposite was the case for PPM. There was no significant difference between ticket type for the ALM and APM elasticities…

In drawing this paper to a close, we note that policy advice on the value of lateness and reliability, and its implications for rail demand, has hitherto been heavily driven by SP research conducted at the individual-level. The evidence emerging from that body of research has tended to suggest that rail travellers place a relatively high valuation against performance (at least when compared to other drivers of demand such as scheduled journey time and waiting time). Given the manner in which this valuation is implemented within standard forecasting frameworks, particularly in Great Britain, a working proposition has evolved that changes in performance could have a marked effect on rail demand. Our own analysis, which yields a dedicated elasticity of demand with respect to performance, would seem to refute this proposition. We find that, at the market-level, the effect of changes in lateness and reliability on rail demand is trivial, both in the short-run and and the long-run.

Why then do the messages from individual-level and market-level analyses diverge? Whilst acknowledging that, on the face of it, the market arises simply from the aggregation of the individuals, we suggest that the two analyses capture quite different behavioural responses to performance. Individual-level analyses tend to restrict behavioural response to an arbitrary service choice, whereas the short-run market-level model could possibly encompass a range of responses including changes to departure time, ticket type, route, operator and/or mode. In the long-run, the market-level could extend to an even wider range of response, such as residential location, car ownership and employment. Moreover, the degree of correspondence between individual-level and market-level analyses of performance will depend on the extent to which these behavioural responses are available to travellers, and the extent to which the available responses are in practice exercised.

In other words, we suggest that the relatively high evaluation of lateness at the individual-level is not necessarily inconsistent with the trivial demand response at the market-level. Another way of putting this would be to say that rail travellers show considerable disdain for experiences of lateness, but that such experiences would not necessarily dissuade them from travelling by train".

This Summary and conclusion sets out, in rather more words, the substance of the Abstract. It is important to note that the authors are not saying that the SP -based or weighted approach is wrong or necessarily wrong. A problem for the so-called "market-led" approach is that there are numerous factors, apparently not yet fed into the RP approach, which may influence or decide what the rail passenger market does as a whole. As has been pointed out in the evidence, the number of passengers travelling on trains over the last 10 or more years has increased significantly. What cannot easily be determined is both whether it would have increased more if trains ran more than insignificantly late and whether, how, when and in what numbers people do not travel or travel less than they would otherwise have done but for significant delays experienced by them on the different rail service lines.

I turned to the paper published on 12 June 2014 by Professors Wardman and Dr Batley: "Travel time reliability: a review of lead time valuations, elasticities and demanding impacts in the passenger rail market in Great Britain". Again the Abstract provides a summary:

“This paper provides an extensive review and reconciliation of British and European evidence relating to the value of, and demand responses to, rail reliability. In particular, we compare the elasticities implied by stated preference valuations of late time with directly estimated lateness elasticities. We find that the implied lateness elasticities are substantially greater than those directly estimated. A possible explanation for this is that lateness has been over-valued, but more sobering explanations would be to suggest that, whilst rail travellers dislike unreliability, they may be unwilling or unable to reduce their rail travel in response to experiences of poor performance, or else conventional economic approaches to deducing elasticities are not appropriate. The findings have been used to update the recommendations of the UK rail industry’s Passenger Demand Forecasting Handbook."

Again, like the 2011 Abstract in their earlier joint paper, this is not exactly prescriptive and merely suggests possible explanations for the difference between SP valuations of late time (which produce implied lateness elasticities) and RP analysis (which produces directly estimated lateness elasticities. There is no suggestion that the PDFH (to which they have materially contributed) is wrong, unreasonable or unfair.

This is a longer paper, running to some 27 pages of text, and I do not intend to set out large tracts as it is not necessary to do so. It contains much theory, reference to academic studies, formulae and calculations. It does show that "train reliability can be seen to be important for all the key players in the rail market" (Page 5) and expressly refers to the importance of reliability within the Schedule 8 performance regime under the TAAs. It refers to the 2012 Passenger Focus report that the importance of arrivals on time was the most important driver of passenger satisfaction. It analyses some of the background to the PDFH. It refers to other "insights" in other European countries. It sets out at table 5 by reference to 6 RP reports between 2003 and 2011 directly estimated elasticities. In a chapter headed “Synthesis", the authors say:

“Turning to the implied elasticities based on our UK review evidence, these are often larger than those implied by PDFH, and some exceed the directly estimated elasticities (substantially in some instances). And where the UK elasticities are less than those from PDFH, it is actually where the latter corresponds well with the direct evidence. Moreover, the review-based evidence seems to be inferior for the purpose of forecasting reliability effects than using the rather crude wAML of 3 throughout. On this basis, we might conclude that the PDFH’s current recommendations are (relatively speaking) defensible."

They go on to set out in Table 10 a list of implied and directly estimated elasticities, upon which the experts to some extent disagreed as to the realistic viability of some of the figures produced, particularly in relation to the directly estimated elasticities. What is clear from Table 10 is that "the implied elasticities tend to be much larger and cannot be regarded as consistent with directly-estimated elasticities” (Page 26); on a comparison of the two sets of figures, that could be said to be obvious.

There is a not insignificant range within Table 10 of directly estimated AML elasticities and it is at least properly arguable that one needs to have regard to the range rather than a mean of the values in the range. This may well suggest that the PDFH approach, based as it is mainly on SP research but with regard had to available econometric based research, is not or is not necessarily unreasonable. Mr Palmer would seek to assert that Professor Wardman and Dr Batley were arguably wrong to base their analysis on all the studies to which they refer because he would seek to suggest that one of them (Oxera 2005) was inappropriate to take into account; I do not accept that approach, not least because these eminent academics, who are way above Mr Palmer in terms of expertise, did not go that far.

In their Conclusions, they acknowledge that the "late time values emerging from the UK evidence…exceed by some margin current PDFH recommendations"; indeed, that appears to be right. They go on:

“The directly-estimated elasticity evidence is also diverse, and a particular issue is that insignificant coefficient estimates for reliability measures are commonplace. These elasticities are unlikely to be zero in practice and hence this adds a significant element of uncertainty into our review. A further issue is that a number of studies used the PPM measure of reliability instead of the preferable Average Minutes Late. Nonetheless, the evidence suggests that reliability does impact on rail demand and we have recovered limited variation by flow type. Noticeably, the elasticity evidence based on mean lateness is more robust than that based on PPM.

We are in the fortunate and original position of being able to compare significant amounts of evidence relating to directly-estimated and implied late time elasticities. Although there are assumptions and approximations involved in this process, it is perhaps the most important aspect of this paper. We find that the late time multipliers imply elasticities somewhat larger than the directly-estimated elasticities. This is consistent with other aspects of the paper, in particular the proposition that the SP-based late time valuations are too large, but we have also discussed other reasons why the implied elasticities might exceed directly-estimated ones.

Although we have covered a significant amount of material, it points to a need for further work in this area. Firstly, the SP values might be too large, but this needs to be tested by obtaining values from well-defined RP choice contexts offering clear trade-offs between reliability and other variables and essentially with large sample sizes. Secondly, the other reasons why the direct and implied elasticities might differ, generalising to other similar contexts, need to be further examined. Finally, it is clear that much further work is needed on obtaining robust directly-estimated late time elasticities. This should not be treated as a ‘side-issue’-, routinely entering some readily available reliability variable alongside other terms in a rail demand model. This review demonstrates that studies which enter the more appropriate mean lateness rather than PPM are more successful in recovering significant effects, as are studies that pay attention to detail in terms of selecting suitable flows and detailed data that supports reliable estimates. In due course, suitable measures representing the variability of lateness in addition to mean lateness as should be entered into these demand models…"

I conclude from this paper that it is not and was not intended to be any sort of condemnation of the approach adumbrated in the PDFH, of Schedule 8 or of the use of SP research to analyse the impacts of reliability or unreliability on demand. Similarly, it is not a ringing or unqualified endorsement of the use of RP research as the only basis upon which to determine such impacts or effects. If anything, it suggests that much more work needs to be done to produce robust elasticities from RP research. It is not being prescriptive, unduly or indeed otherwise, as between SP and RP, even talking of the PDFH’s approach being "defensible", albeit "relatively speaking". Whilst it is undoubtedly clear that the figures produced from RP research show much lower elasticities than those from SP research, that in itself does not mean that the PDFH and Schedule 8 approaches are necessarily wrong or are in any way unreasonable or even unrealistic. It is worth observing that this 2014 paper, whatever its ramifications, was not available to inform those charged with drawing up the PDFH editions current when the incidents in these five case occurred. It can hardly be said to have been unreasonable for these earlier versions of the PDFH not to have taken (the unknown and unwritten) 2014 research into account.

One of the very real problems with both SP and RP research is that the results extrapolate from answers emanating from those surveyed in the case of SP and from what may be wider ranging statistical data in the case of RP. In an ideal or highly theoretical (and unnecessarily complicated) world, one could, in relation to each incident on the railway which delays trains by more than a few minutes, not only track down each affected passenger and ask her or him whether or not and if so to what extent they will continue to travel by rail on the particular service but also to see what he or she actually does in relation to rail use as a consequence of the incident over several years. That of course is highly impractical and, one hopes, will never happen.

The Court was asked by Counsel for the Defendants (Paragraph 41 - Closing Submissions) to find that the "true" elasticities of demand to AML are no greater than those set out in Table 10 of the Wardman Batley 2014 paper or that they are probably in the region of those determined by the 2011 paper. This is simply unrealistic. The reality is that there are no "true" elasticities in the sense that elasticities can never be absolute or constant. Depending on whether one takes SP or RP or even a combination of both approaches, what the authors of PDFH and Schedule 8 were seeking, in good faith and after extensive and intensive work and consultation, to do was to make an informed judgment on what the likely impact of delays or extended timetables could well be. To suggest that the RP-based or directly-estimated elasticities are the "true" ones is, simply, wrong or, with respect to the defendant's arguments, at least naive.

What is reasonable is to make realistic assumptions based on all the generalised evidence available. I am wholly satisfied that the PDFH process and the related Schedule 8 methods of assessment of loss relating to delays were reasonable and realistic. It was obviously done and put together with immense care and following intensive consultation not only with Network Rail and the TOCs and related transport institutions or departments but also with academics and researchers. It was a transparent exercise that was done on each occasion that the PDFH was revised (although it was by no means an open public consultation). It has produced a detailed approach to delay evaluation which is evidence-based where the evidence is and was reasonably robust and forms a logical basis for the delay related evaluation process and formulae which are deployed in the TAAs. The fact that no one method of analysing passenger demand and reliability concerns is perfect should not deflect from the reasonableness of the process which was gone through and of the result adopted. There is wide acceptance that SP and RP have merits and demerits.

The Defendant’s case on this headline point falls down with Mr Segal’s evidence to the effect that the versions of the PDFH current at the time of the five incidents with which these cases are concerned and of the TAAs (including at the time of the “One” determination) were reasonable and provided "robust unbiased estimates of the effect of reliability on rail demand, based on the best of evidence available at the time" (Paragraph 7.3.7 First report), which I accept. The point falls further down when one bears in mind that Mr Palmer was unable to say that they were anything other than reasonable at the time. Even if one was to feed in the latest research and input from Professor Wardman and Dr Batley, there is nothing that significantly undermines the PDFH and Schedule 8 approach. Essentially, the point is a bad one and is certainly not only not established on a balance of probabilities but also positively disproved on the evidence.

The second "headline" point raised by the Defendants to seek to challenge Schedule 8 was amplified by their Counsel as follows:

“Schedule 8 is not designed to estimate the revenue loss resulting from a single incident. When misused for that purpose, it exaggerates the true loss generated by the single incident. The calculation is based on long-run elasticity and ignores the phenomenon of lag. There is a real difference between a permanent change (sustained for long enough to fully overcome lagging factors) and a temporary change (not sustained for long enough to fully overcome lagging factors). Lagging factors make short-run elasticity lower than long-run elasticity. This phenomenon is known from research evidence and explained in PDFH v5.1 and elsewhere. The analogy between timetable changes and changes in the level of reliability (upon which the Schedule 8 calculations depend) breaks down at the level of individual incidents."

Schedule 8 was, clearly, not designed for the specific purpose of estimating actual revenue loss from a single incident, whether road on rail or other; Mr Segal and Mr Palmer both agree that in their Joint Statement. The difference between the parties is whether or not it is suitable so to be used with Mr Segal saying clearly that it is and Mr Palmer saying that it is not.

There is no doubt that the five incidents with which these five sets of proceedings are concerned were each "single" incidents and that the single incidents caused delay and disruption to the running of the various railways. On the evidence, it is clear (and there is no real dispute) that, as a result of each single incident, the various periods of delay to and cancellations of train services occurred as set out in Paragraph 3 above and that in consequence Network Rail had to and did pay the sums set out in the schedules in that paragraph to the various TOCs in respect of such delays and cancellations. There is therefore no real doubt that in that sense the losses represented by those sums were caused by the negligence of each of the respective defendants.

The reference to "lag" or "lagging factors", although not a term of art, is intended to refer to the effect or impact on revenue (or use of the railway in question) over time, say, of a single incident causing delay or disruption to train services. In PDFH v5.1 (April 2013), said by Mr Segal in cross-examination (Day 4 Page 187 line 13) to reflect the intention in previous versions of the handbook, it was stated (B12.1):

Lags can at least partly be explained by reference to the five cases in issue, all of which occurred on week days and, for at least three, the delays were not during morning or evening rush hours (the Handy and Godley cases being the other two). On these types of train, there may be commuters (often with season tickets), business travellers going in effect from Point A to Point B for business-type purposes and leisure related travellers. There is no evidence in the five cases as to the precise numbers of actual travellers affected or into what categories of passenger they fell. However, one can take different examples:

As referred to in the PDFH v5.1, there was evidence of lag effects for all material types of rail travel. Chapter C12 sets out and reaches conclusions on recommended values for lags (effectively to be used in the calculations relating to Schedule 8). The evidence was contained in various papers and studies undertaken by different academics and other data from the 1970s through 2005 and 2006 to 2012. In C12, there is arguably a greater emphasis on the time taken for rail or railway services improvements to be reflected in increased use of or revenue from the railways than on deteriorations in services but is clear that the evidence and recommendations relate to both improvements and deteriorations. C12.3 contains the Recommendations that are prefaced with the observations that the "evidence is very mixed, so the recommendations are relatively simple" and that there is reliance in part upon the evidence but also on "logic" for instance "that indicates that improvements take longer than deteriorations to work through" because for improvements in use or revenue there will have to be "new users" who will have to be informed through advertising or orally and that would take longer than existing users on whom the impact of deterioration will be more immediately experienced. The recommendations cover deteriorations and improvements both in fares or GJT (Generalised Journey Time) and in reliability or punctuality for non-commuters, as well as for changes in all factors for commuters and for the lags attributable to the introduction of major new services.

There may well be, for any given passenger whose travelling plans or arrangements on a particular railway line have been influenced directly by a single rail incident such as the five incidents in this case, some confusion during the actual or potential "lag" period. For instance, the football fans, say, supporters of Ipswich Town FC ( and affected by the Godley incident), may have decided for a long but indefinite period not to travel by train to away fixtures in London and the South East of the country but the only other route is by car or by bus along the A12; if a decision was later made to upgrade that road to a three lane highway and for three years there was likely to be consequentially extensive delays to road traffic, the football fans might decide to return to the train option. Alternatively, the cost of diesel (for buses) might go up by so much, the rail fare structure for cheap day return tickets might be improved so drastically in favour of rail passengers and the Ipswich to London rail line might be so improved that the travel time is reduced by one third that the football fans decide, notwithstanding their decision never to travel by train again to attend away games as a result of a single incident, to revert to the railway. On that example, the lag or period over which the impact on those football fans of the single incident would be cut short, possibly from 4 years to 2 years or more or less.

There may well also be confusion in the determination, with any precision, of the actual impact of any individual single incident. For instance, a single incident which delays travellers by over two hours might lead to 30 passengers out of the 500 affected deciding not to use that railway line in effect for the next four years and they may not be deflected from their decision; however, the overall passengers using that line may increase for socio-economic reasons and it therefore does not appear that there has been an effect on revenue or passenger numbers as a result of the single incident. The Revealed Preference (RP) econometric evidence for that line will not reveal any impact of the single incident although as a matter of fact 30 less passengers are using them than will have been using it if the single incident had not happened.

There are a number of sub-issues which arise under the point raised by the Defendants in this context. The first is whether or not individual passengers are deterred from travelling for a greater or lesser period than they would otherwise have done as a result of a significant single incident (which I might classify as one involving more than 30 minutes delay to a single journey). I have not the slightest doubt on the evidence available and as a matter of common sense and logic that there are in all probability a number of passengers who are so deterred and who decide not to use the particular TOC on the particular route again for a significant period after the incident in question. To suggest otherwise is untenable. It is supported, for instance, by the Big Delays Survey and by the repeated observations by academics, in passenger surveys and in the PDFH that passengers rate reliability and punctuality highly.

An aspect of this is the "tipping point theory" said to have been advanced by Mr Angus in his first witness statement (Paragraphs 138-140). He suggests passengers have or reach a point at which they decide not to use the particular TOC again consequentially upon the single incident in question. This "tipping point" may be the first time that they have experienced a big delay on the railway line in question or on the railways in general; that might apply to some of the younger football fans to whom I have referred above by way of example. For others, the single incident of delay in question may be the last of two or 20 or more incidents such that the passengers in question decide that "enough is enough" and abandon using that TOC generally or on that line. The Defendants’ Counsel team suggests that these comments from Mr Angus are inadmissible as either opinion or expert evidence. However one classifies the evidence, it seems to me a matter of common sense and inference that human beings do and will often behave in that way; one does not have to be a psychologist to reach that conclusion. In any event, Mr Palmer agreed with Mr Angus on this point (Day 6 Page 165).

A second sub-issue is whether or not the Schedule 8 regime can legitimately or properly be applied to determine or assess the loss consequential upon the five incidents in question, in circumstances in which it is suggested by the Defendants that in effect there is no real proof or logic in the lagging factors taken into account in elements of the Schedule 8 regime. Mr Segal in evidence and through his reports was confident that it is suitable or "fit for purpose" to estimate the actual lost revenue resulting from a single incident of delay. I accept his evidence and opinion on this point. The Schedule 8 regime, allied with the delay attribution process, can and does determine in practice how much delay was caused by factors which are the risk and responsibility of Network Rail as between it and each of the TOCs; from the delay attribution arrangements between Network Rail and the TOCs, one can determine with almost 100% precision how many minutes delay was caused by each of the five single incidents in the cases before the Court. To that one can apply the RPR (Relevant Payment Rate) or MRE (Marginal Revenue Effect) for each of the rail services affected to determine what has actually been paid by Network Rail to each of the affected TOCs as a result of the five single incidents in question. There is nothing unreasonable about that.

In effect what the Defendants’ legal team is arguing in this area of the case at least is that (i) either no-one can say that any or many of the passengers undoubtedly affected by the significant (or worse) timetable delay and disruption in the five individual cases (or generally) decided or decide as a result of the incidents not to travel again on that line with that TOC (ii) or that the Schedule 8 formula or the MRE figures consistently over-estimates the number of passengers and/or the amount of revenue likely to be lost as the result of a single incident. Of course, there is no such direct evidence for particular incidents because neither the TOCs nor Network Rail ever seek that sort of evidence in the aftermath of timetable delay such as occurred in the five cases or in comparable circumstances. That is just as well because it is more likely than not that passengers in the immediate aftermath of incidents such as those which occurred in the five cases are, in their annoyed, frustrated, angry or irritated state, likely to respond to an early survey of their reactions that at least many of them would wish never to travel on that line again; that sort of statistical evidence would probably exaggerate to some extent the likely overall impact on revenue or numbers travelling.

The apparent problem is that one (or Network Rail or the TOCs) cannot practically prove in perfect terms what the loss of TOC revenue is from a single incident, which does not mean to say that there has been no loss or that such loss is less or more than the Schedule 8 exercise provides for. If one could say that there is no realistic chance that the TOC would suffer any significant revenue loss from a single incident delay, one could well then say that the Schedule 8 exercise was not compensatory at all and represented a free gift or subsidy to the TOCs in question. That is clearly not the position here. There is at least the real possibility that the actual revenue loss to a TOC for a given delay causing incident could well be more than the Schedule 8 exercise provides. An example could be the (say 100) Ipswich football supporters coming from Ipswich to London by train to watch over a football season, say, 5 away games in the Greater London area who are caught up in the delay caused by the Godley incident who miss the game and decide for the next 4 years not to go by train but to rent buses: one then has 4 times 5 times 100 times (say) £20 (the cheap day return fare) which produces a loss for the individual TOC (carrying the football fans) of £40,000 as well as losses on the day and relating to any other passengers who vote with their feet. The actual Schedule 8 calculation produces some £27,000 which is all the relevant TOCs receive. Of course the actual loss may be less. The whole point of a formula based approach is that it represents a reasonable (if imperfect) attempt to assess a loss of revenue which would be seriously difficult to relate with precision to a single incident, albeit that there is in all probability a significant chance that there has been more than insignificant revenue loss attributable to the single incident.

It merits mentioning what the Office of the Rail Regulator has said from time to time about the Schedule 8 regime:

“[T]he compensation available through [Schedule 8] of the franchised passenger operators’ track access agreements...provides [a] liquidated damages [regime] which [specifies] payments from one party to the other based on a reasonable pre-estimate of the expected loss to [the TOC] arising from [the incident that causes the relevant service disruption]” (“Provisional Conclusions on the Incentive Framework”, ORR, April 2000, Para. 7.16)

“If the regime is to be effective, it is essential that payments between Network Rail and TOCs reflect, as closely as possible, the impact of changes in the level of performance on TOC revenues.” (ORR’s 2005 Performance Regime Review, Final Conclusions Para 2.5)

“One of the key objectives of the performance regime is to provide appropriate compensation to TOCs for the loss of revenue resulting from lateness and any cancellation of their services. To fulfil this objective effectively, the payment rates from Network Rail to TOCs need to reflect, as accurately as possible, the effect on TOCs’ revenues of changes in Network Rail performance.” (ibid Para 3.5)

“A Network Rail payment rate must reflect the effect of one minute change in performance away from the current level. It should therefore reflect the number of passengers associated with a service group and attach a ‘value’ to each passenger. This value is the marginal revenue effect (MRE). The MRE represents a passenger’s ticket value and the likelihood that they alter their mode of transport for making a journey based on changes in performance.”(“Review of Schedule 8 Payment Rates” by AEA Technology for the ORR, December 2005, p 5)

“[T]he Network Rail payment rate is designed to reflect the impact of performance on a train operator’s long term revenue. It is composed of the estimated average marginal revenue effect (MRE) per passenger journey within a service group multiplied by the number of passenger journeys per day in that service group. The MRE represents the impact of a minute’s lateness on fare revenue over time.” (ORR’s 2013 Periodic Review, Draft Determination Para. 20.60)

“The Network Rail payment rate sets the basis for compensation payments from Network Rail to train operators when Network Rail’s performance is worse than benchmark, and bonus payments to Network Rail from train operators when Network Rail’s performance is better than benchmark. Network Rail payment rates are set at a level to reflect the impact over time of performance on fare revenue.” (“Current Compensation Arrangements” in ORR’s “Final determination of Network Rail’s outputs and funding for 2014-19” October 2013 Para 20.15)”

I have no doubt that the Schedule 8 regime represents a reasonable and realistic attempt to assess compensation for loss of revenue attributable to the impact of delays on the TOCs’ revenue caused by any single incident, such as those represented by the five cases with which this judgment is concerned. Although the Schedule 8 regime does not impose in pure terms exactly a liquidated damages regime because there is not necessarily any breach of contract on the part of Network Rail, it is comparable because it provides for compensation to TOCs in the event of delay caused by events for which Network Rail has accepted responsibility vis-à-vis the TOCs. One can then readily conclude that the Schedule 8 regime was a genuine and reasonable pre-estimate of loss of revenue for such delay. It does not represent some sort of “in terrorem” payment or gratuitous donation to or subsidy of the TOCs.

Essentially, the evidence relied upon by those who drew up the PDFH and devised the Schedule 8 regimes with which this case is concerned and the other evidence in this case leads inexorably to the conclusion that the delay and disruption caused to passengers by single incidents such as occurred in these five cases can cause loss of passengers and therefore revenue over time after the incident. This is of course not limited to the passengers immediately affected on the day but also other potential passengers who find out (say from family members who were so affected or from the press) and decide not to travel on that line for a period.

The precise arguments about other factors interfering with the consequences of a single incident, namely the detailed points about lagging factors now taken by the Defendants’ legal team, were not articulated by Mr Palmer in his first three reports and he accepted that he may have come to these views after his third report. That does not provide for any confidence that the arguments now adumbrated in this context have any real validity at all. It is self–evident that there will be some lag in the impact of a single incident on any individual passenger who resolves not to travel with a TOC as a result of a single incident and this may range from a day to a year or more; that depends on the regularity of that passenger’s rail trips prior to the incident: the away-game footballers may travel fortnightly and the rail visit to a relative may be once a year and so the impact will not be felt in terms of loss of revenue for a fortnight or a year respectively. Whether subsequent events alter the decision of the passengers not to travel with the TOC in question is in a real sense random: first, the subsequent events may not happen or the passenger may be unaware of them, secondly, some passengers may change their decisions and others may not; thirdly, other factors may intervene with an individual passenger (such as a move abroad, death or a change of job location) such that the passenger who might or might not have changed the decision does not need to do so.

It was and is overwhelmingly a realistic and reasonable inference for those who were involved in drawing up the Schedule 8 regime and in this case that a not insignificant number of passengers will resolve to abandon or reduce their use of the railway in question (if not generally) by reason of a single incident such as happened in the current cases. It is at least a realistic and reasonable working premise. That abandonment or reduction may well extend to or beyond the times envisaged within the lag periods selected in the PDFH and deployed to support the Schedule 8 assessment. It may of course be less. It seems from the evidence and argument that the Defendants do not and cannot say that there would be no such loss; the argument is more about what the likely loss would be.

The micro-analysis adopted by the Defendants in this case, for instance on this particular issue, even if it had merit just on its own, does not address the whole issue as to whether the deployment of the Schedule 8 basis of assessment was reasonable or unreasonable. A good example is the fact accepted by both parties through Counsel and the experts which is the significant under-estimation of the revenue loss endemic in the MREs adopted because the indexation based on RPI built into the Schedule 8 and MRE-related assessment has not kept pace with actual TOC revenues over the period covered by these 5 cases (and before and since); these revenues have substantially increased over the last 12 or more years, compared with the RPI increases. Another example relates to the fact that the Schedule 8 exercise does not allow for what may be relatively significant losses to the TOC on the day of the single incident, such as refunds, bus-hire, overtime and loss of ticket sales on the day in question (people who turn up at a station to be told that the line is blocked and decide not to travel at all; these may be significant for instance on the Handy case with 38 cancellations and 92 other trains (for Midland Mainline) delayed over a near 7 hour period. So even if there was something of an over-estimate of the lag periods over which revenues are lost following a single incident, there is a counter-balancing exercise which would have to be done to take into account these sorts of factors. That exercise cannot be the type of exercise which in the circumstances of these five cases would have been envisaged by any member of the Court of Appeal in Conarken.

Mr Angus gave evidence (in his second statement Paragraph 7-9) to the effect that some 10,000 passengers were affected by cancellations and many more by the other delays to rail services caused by the Handy incident. It takes little to infer that amongst the many thousands of passengers affected there are likely to have been a not insignificant number who would have become so disenchanted with the experience that for a greater or lesser period thereafter they would have been reluctant to use the rail service in question and they would not have used it in fact.

In relation to the issue under consideration in this part of the judgment, Counsel for the Defendants also deployed what was said to be the "available evidence from TOCs" referring to the Handy and Barrow upon Soar incidents and the "One" arbitration which I have already addressed earlier. They do not represent any real evidence that the TOCs in question believe or believed at any material time that there is or was no future loss of revenue from single incidents such as occurred in the current five cases; it is inherently unlikely that, if they truly believed that and had any reliable evidence to support it, it would not have emerged clearly at some stage in the PDFH reviews over the last 20 years or more. There was an attempt to undermine Mr Segal who appeared as an expert in a Commercial Court case before Gross J (as he then was), Great North Eastern Railway Ltd v Railcare Ltd [2003] EWHC 1608 (Comm). The case involved negligent work by the defendant on a rolling stock wheel which resulted in a derailment at high speed of the relevant passenger train. One of the claims was for GNER’s lost passenger and other revenue attributable to the derailment. Mr Segal was the railway consultant expert for GNER. GNER’s quantum case on the lost revenue was put by reference to lost revenue in the three four weekly periods immediately following the derailment and essentially it was asserted that this lost revenue could be determined by a comparison between what revenue was earned in this 12 week period immediately after the derailment and what would have been earned; the argument was mostly around what one compared with what. The judge preferred a comparison between the actual and the best estimate of revenue for that 12 week period. There is no obvious reference in the judgment to GNER arguing that no loss was suffered in the period thereafter, albeit that it was not claiming for it. This case does not provide evidence that GNER did not suffer revenue loss outside the 12 weeks immediately after the derailment; it only provides evidence that GNER did not claim for it. If anything, the case provides some corroboration for Network’s Rail assertion that, at least in the months following a serious single incident, there is likely to be revenue loss; in that case it seems for several days the relevant train fleet was withdrawn from service, with trains being returned to service as and when inspected initially at reduced speeds and with the normal timetable being resumed in full some 6 days after the derailment. It is also consistent with the points made in the two preceding paragraphs.

I conclude that this second point made by the Defendants is bad and unjustified either by any real or relevant evidence.

Counsel for the Defendants amplified this further point as follows:

“Schedule 8 exaggerates the true loss generated by a road on rail incident, because such incidents are unusual in the experience of passengers, so that most passengers would not regard the experience of one such incident as relevant to future travel decisions. (Infrequent travellers, as a class of potential customers on an affected line, are unlikely to experience the event, though some will. Frequent travellers on the line will be more concerned about the general run of relatively frequent delays.) In this respect road on rail incidents differ from delays from causes to which passengers frequently hear delays attributed, such as points failures, signalling problems or train defects."

The argument relates to 2 factors principally, namely the relatively few number of road on rail incidents every year (between 2005 and 2012 some 4,300 as against 800,000 overall delay causing incidents) and what is said to be a common sense point that passengers are much more likely to be put off travelling, not by things which they do not expect to happen (like road on rail incidents) but by things which they do expect to happen such as signalling or points failures. The argument goes on that passengers will not change their behaviour because of something which is unimportant or relatively rare or which is unlikely to be repeated in the foreseeable future.

Counsel for Network Rail say that there is no evidential basis or logic for the Defendants’ arguments. I accept this.

These arguments are, on analysis, based on assertions as to what passengers feel as a result of delays caused by different factors and it pre-supposes that passengers decide not to travel on the railways in general or a particular railway on the basis of accurate information which they have received as to what the causes of delay are; it pre-supposes that they will decide to continue travelling on the railway after a road on rail incident because it was or might be considered by them to be a one-off or very rare incident. There is no suggestion or evidence that passengers at large know that road on rail incidents are relatively rare. What passengers affected by any significant delay causing factor do actually know is and must be that they have simply been delayed by however long the period is and that they have had their plans for the day materially disrupted. There is no evidence to suggest that passengers are more sympathetic towards Network Rail or the TOCs if the delay causing event is a road on rail incident as opposed to a points failure or a driver not turning up to work. There is no evidence that passengers believe what they are told over the train or station tannoy as to what the causes of delay are or that accurate reports are always given as to the causes of delay or cancellation.

It is much more likely that passengers are unsympathetic to any types of significant delay (other than possibly caused by some sort of humanitarian crisis, such as someone falling ill on the train or a bomb), whatever the cause. There is, I presume, rarely if ever any detailed explanation given to passengers as to the causes of delay, although it may well be that there is some evidence of increasing provision of simple and generalised explanations provided by train drivers or at the stations at which passengers arrive. This may well be unsurprising because by the time that information can be provided to customers (say, on the delayed arrival at the destination station) detailed information as to the causes of delay may be unavailable. The proposition that passengers are readily able to make an informed decision as to future use of the railway based on information which may or may not be relayed to them during delay disruption caused by different types of incident or which may or may not be accurate or complete is fanciful.

An argument was deployed by Mr Palmer to the effect that "repeatability" was a factor which passengers would take into account. By this, he meant that, if passengers were aware that the type of incident which had caused the particular delay suffered by them was unlikely to be repeated in the indefinite future, they would not give up the use of the railway as a result of a delay caused by an incident of that type. There is absolutely no real evidence to support this "repeatability" point and it is not obviously right as a matter of logic or even commonsense (to the extent that logic and commonsense are different). One of the problems with this assertion is that, as Mr Palmer accepted (Day 7 Page 6), every incident which delays one or more individual trains can be considered as an individual or single incident (whether it is a points or signal failure or a road on rail accident); all that the passengers know is that they have been seriously delayed or disrupted and, I presume, it is that latest incident which, for many passengers resolving to give up using the railway in question or generally, is the real and proximate cause of the decision not to use the railway. They will say: "Enough is enough I will not/can not take the risk that I will not make it to my destination next time round", whether that "next time round" is the following day, week, month or even year. Mr Palmer accepted (Day 7 Pages 7-8) that there is a lack of research in this area and that the opposing view namely that of Mr Segal was reasonable that passengers would be influenced by "large delays". He went on to accept that it was a reasonable assumption that all delays will affect passengers’ perceptions about reliability and will reduce their willingness to use the rail service in future (Day 7 Page 13).

It was and is wholly reasonable to deploy Schedule 8 and use it as an appropriate basis to determine the revenue losses actually incurred by Network Rail as the result of the five incidents in the current five cases. Schedule 8 and the related MREs were based upon reasonable assumptions, research, evidence and consultation. There is on analysis nothing which can be said to be unreasonable or exceptional in the Schedule 8 approach or its application. The Schedule 8 exercises and fixing of MREs for the periods to which the five cases relate have been done with immense care to achieve a fair balance and to secure that, amongst other things, fair and reasonable compensation is payable to the TOCs by Network Rail, when cancellations of and delays to train services occur as a result of incidents such as those which occurred in the current five cases. Given the substantial amounts of delay and disruption to the various rail services caused by the negligence of the five drivers in each of the five cases, the amounts established as lost revenue are not unreasonable, obviously or at all.

“Lesser Factors”

Counsel for the Defendants identify what they call "Lesser Factors” which are said to be approximations in the Schedule 8 calculations and of which some are in favour of and some against the reasonableness of the Schedule 8 basis. It is said that the effects of these factors are relatively small and, broadly, balance out. The only item which goes to suggest that the Schedule 8 payments might be too small is the fact that rail revenues have outstripped RPI which is factored in to Schedule 8. The five factors said to go the other way are:

“(a)

Schedule 8 ignores crowding effects. Loss of revenue is mitigated or eliminated where a deterred passenger’s place is quickly taken by another.

(b)

Schedule 8 ignores redistribution effects. A passenger lost to one rail line may not represent a loss to the TOC if they transfer to an alternative route operated by the same TOC.

(c)

Schedule 8 ignores adverse effects of road on rail incidents on road users. A lorry stuck under a bridge is likely to block the road and cause inconvenience to road users, who may therefore be deterred from road travel and encouraged to use rail.

(d)

Schedule 8 ignores lag. If (which the Defendants dispute) there is any significant long tail effect of a single incident, the affected TOC receives accelerated payment from Network Rail.

(e)

Schedule 8 ignores franchise termination. The TOC still receives the full payment, even if the supposed long tail effect would occur after the termination of the franchise.” (Paragraph 111 of Closing Submissions)

These five factors are mostly unjustified and at best very minor, for the following reasons (by reference to the sub-paragraph headings in the preceding paragraph):

What the Defendants ignore in these assertions is the fact that there is no evidence that they are germane to the incidents or TOCs which are the subject matter of the current five cases. For instance, there is no evidence that the trains affected by the five incidents were so crowded (or crowded at all) such that over the next few quarters and up to two years after the incidents any passengers who did decide not to travel again with that TOC on the line in question were replaced by other passengers. Further, there has been no evidence that there was available to any of the passengers within the same TOC affected on the day of the five single incidents in question an alternative route operated by the same TOC. Another example is that there is no evidence that on either of the bridge strike cases (Ingram and Rob Hatfield) traffic was seriously delayed; it is a reasonable inference that there was some delay but there is no evidence that the police and other services did not either put in place temporary diversions or traffic controls or promptly notify radio stations and satellite navigation companies of the problem so that road users could avoid the area and the impact on road users in the area would be much more limited than the times represented by the train delays.

Whilst it is right to say that the factor of the increasing revenues over and above the RPI impact arises mostly on the Handy, Ingram and Godley cases and to a much lesser extent on the other two cases (which occurred relatively shortly after a re-fixing of the Schedule 8 rates) that factor is, in general, a genuinely weighty factor in supporting the Schedule 8 rates as reasonable, unlike those raised the other way.

This is the fourth area in which the Defendants seek to undermine the reasonableness of the Schedule 8 process and the related delay attribution process. Three grounds of challenge are relied upon by the Defendants:

“(a)

The motorist is charged for unidentified minutes not attributable to the incident.

(b)

Where a reactionary delay is caused in part by the road on rail incident and in part by a different cause, the higher delay principle attributes the whole reactionary delay to the larger incident. Since road on rail incidents tend to be larger than the average delaying incident, this tends to prejudice the motorist.

(c)

RICCS makes no allowance for delays which did not occur but which would have occurred if the incident had not happened.” (Counsel’s Closing Submissions – Paragraph 114)

It is fair to say that these points would not so undermine the basis of revenue loss calculation in the five cases that nothing would be recoverable; there would at best be relatively small reductions from the damages otherwise recoverable.

I refer to the general remarks at Paragraph 26 above made about the delay recording and allocation arrangements. It is now necessary to expand on that. What is now and since 2004 has been known as the Network Code is incorporated into the TAAs and thus train performance measurement is incorporated contractually. The Performance Monitoring System so incorporated records the times when trains arrive at, and depart from and pass the specified monitoring points against their schedule times, uses them in determining the differences between the actual and scheduled times and also enables Network Rail to record the cause of any delay. The Performance Data Accuracy Code is incorporated into Schedule 8 by Paragraph 4.1 and this Code provides for standards of accuracy and data completeness in relation to times recorded. Network Rail is charged by Paragraph 4 of Schedule 8 with the recording of information in relation to delays and cancellations. Schedule 8 provides for risk and responsibility for delay, it being common ground that the five incidents with which these five cases are concerned would be the contractual responsibility and risk of Network Rail as between the TOC and it. The Delay Attribution Board ("DAB") governs the delay attribution process and this Board is made up of a cross industry panel of equal numbers of Network Rail and TOC representatives; this Board produces every few years a Delay Attribution Guide (“DAG”) which is issued after industry consultation, although changes must be approved by the ORR. The primary purpose of the DAG is to identify the cause of each delay and cancellation on the relevant rail network; it enables each Delay Minute and Reliability Event (described by Paragraph 1.4.1 as arising when "a train is not able to make all the booked calls shown on the train schedule") to be attributed to the main or "prime" incident.

The data relating to the actual performance of trains compared with the relevant timetables is recorded at the Monitoring Points on the TRUST computer system and archived in the PALADIN system. The DAG provides for numerous different scenarios or causes of delay and cancellation which are the risk or responsibility of Network Rail or the TOC, such as for example Bridge Strike or Waiting Train Crew allocated to Network Rail and the TOC respectively. There is no doubt and in any event I find that Network Rail commits substantial resources and expenditure to manage and run the Delay Attribution process. For instance some 160 staff who are effectively trained deal with train delay attribution together with a further 50 staff to manage data quality as well as the dispute resolution process (incorporated in Schedule 8) which is not only thorough and fair but ultimately can be resolved by an "Access Disputes Adjudication". The ORR’s 2005 and 2013 reviews of the Delay Attribution process determined that it was the best way of capturing the delay information. There is no reason to believe in these cases that the delay attribution process has been done anything other than fairly and reasonably. Even when an error is made, which seems to be rare (minor attribution errors were made in the Godley case), they are readily capable of correction.

It is primarily through the Schedule 8 Delay Attribution process that the Minutes Delay are effectively captured and identified. The Railtrack Insurance Claims Costing System (RICCS) is not a part of the Schedule 8 regime and is not used in delay attribution as between Network Rail and the TOCs. It helps Network Rail to identify the impact of the specific incident and the delays and cancellations attributed to it in relation to such Schedule 8 liability as is incurred by Network Rail and to quantify the loss suffered by Network Rail which would not otherwise have been suffered but for the specific incident. As explained by Mr Kenney in his first witness statement (Section 15), RICCS

“…works by making temporary changes to the performance data:

110.1

It temporarily changes the coding of the specified incident to "Planned", so that it will be excluded from the performance regime.

110.2

It temporarily reduces lateness values pro rata by the level of delay caused by the specified incident (this amount of reduction is the Lateness Reduction Factor).

110.3

It temporarily removes any reliability event caused by the specified incident.

110.4

RICCS then performs the calculations normally performed by PEARS [Paladin Extraction and Reporting System] to show what the Schedule 8 liability would have been had the incident in question not occurred.

110.5

The amount calculated by RICCS is compared against the figures calculated by PEARS (i.e. those including the effects of the specified incident). The comparison demonstrates the financial effect of the specified incident on Network Rail’s contractual liabilities/entitlements under Schedule 8."

He goes on to conclude from his explanations that the RICCS produces results to a high degree of accuracy. I accept his evidence as logical. He does identify several "sources of imprecision", addressed principally at Paragraphs 116 to 122 of his first statement. He says that the first source of imprecision relates to the possibility that, if the trains cancelled because of the incident had actually run, they would not necessarily have run on time; a hypothetical example might be in relation to the Handy incident if down line of the incident there had been simultaneously a points failure such that, if the Handy incident had not happened there may still have been delay caused by the points failure. He says, and I accept, that although the RICCS may overstate the impact of the incident such impact will be relatively small (Paragraph 116.1). The second source of imprecision relates to what has been called the "higher delay principle" where the Delay Attribution Guide dictates that, where there has been consequential or reactionary delay to trains other than those immediately and directly affected by the single incident, such delay is to be attributed to the principle incident which has the largest number of Minutes Delay allocated to it. An example might be a bridge strike at Point 1 along the track which causes 5 minutes delay to Train A and later a signal failure causing 9 minutes delay such that when Train A arrives 14 minutes late at a particular destination it blocks another train, Train B, otherwise running on time so that it is delayed by 4 minutes; this four-minute delay is attributed by the "higher delay principle" to the signal failure; the reverse might apply if Train A was delayed by 11 minutes by the bridge strike. As Mr Kenney said, the higher delay principle applies only to such consequential delays and not in practice at all to the direct and usually major delays caused by both cancelled trains and by the trains directly and immediately blocked by the incident in question.

There is sometimes a small amount of "unidentified" delay the cause of which has not been allocated in the TRUST system. Although in logic the few minutes or seconds delay must have been caused by something, the system and process does not identify the cause. In respect of this unidentified delay, Paragraph 5.5 of Schedule 8 sets out how this is to be dealt with (see Paragraph 15 above). So far as is material, sub-paragraph (a) is applicable here. 50% is allocated to Network Rail (sub-sub-paragraph (a)(ii)); I will call this the “2nd 50%”. As for the other 50% (the “1st 50%”), it is allocated to Network Rail, the TOC and joint responsibility incidents pro rata to the Minutes Delay for the particular Service Group as being their responsibility. That is simply the “deal” or arrangement reached by the TOCs and Network Rail. An example might illustrate this: there are 100 Minutes Delay affecting Service Group A, of which 50 are Network Rail’s and 40 the TOC’s responsibility, there are 10 unidentified Minutes Delay and 18 of Network Rail’s 50 minutes are attributable to a particular Defendant’s negligent road on rail incident. Of the 10 unidentified minutes, Network Rail has responsibility for 50% (5 minutes – the 2nd 50%) in any event and for 50/90ths of the other 5 minutes (2.77 minutes of the 1st 50%) whilst the balance of 40/90ths (2.23 minutes) will be the TOC’s.

I will address the three points made by the Defendants separately.

This part of the judgment and Paragraph 133 above have been partly rewritten following the submission to Counsel of the judgment in draft. As a result of further submissions, I have ascertained that I had made some errors in the first draft. I was not particularly assisted either by the pleadings or by the opening submissions of Counsel on the issue of "unidentified minutes". The issue was addressed to a limited extent by Mr Palmer in one or more of his reports and by Mr Kenney principally in his 7th statement. It emerged more clearly when the Defendants’ Counsel articulated this issue in their written Headline Points produced on Day 4 of the trial; the closings did not delineate the issue very clearly. It has only become clearer since the most recent submissions. Essentially, the Defendants argue that as a matter of fact and causation the unidentified minutes or at least Network Rail’s half share of them cannot be shown to have been caused by the negligence of the five individual Defendants and that it was unreasonable for them to be charged as damages for these minutes. Mr Kenney accepted that, if the unidentified minutes could not be proven to be "the responsibility of the incident", it would not be reasonable as between Network Rail and a particular Defendant.

Network Rail argues that, although the cause of unidentified minutes has eluded reasonable attempts at discovery and it is in practice impossible to say whether this allocation is correct in any particular case, it remains liable to the TOC in question for the Schedule 8 financial consequences of its share of the minutes. It points to the fact (which I find) that the unidentified minutes are almost invariably very small. Mr Kenney said in unchallenged evidence that, for instance, of the overall revenue loss on the Handy incident of £259,733 only £188.07 is sought to be charged as damages for the proportion of unidentified minutes attributed to Mr Handy (0.0724%). There can be no objection as such to the fact that the TOCs and Network Rail share responsibility for the unidentified minutes on the basis set out in Paragraph 5.5 of the TAA; I accept that and indeed it is not challenged as such. It argues that by this splitting of responsibility for these minutes a third party tortfeasor such as one of the Defendants in this case is only asked to compensate Network Rail for an actual loss representing part of what might defensibly be added to the delay minutes positively allocated to the incident in question. It goes on to argue that the allocation of the Network Rail share of the unidentified minutes for any given day in proportion to the delay obviously attributable to a given defendant is not unfair because it undoubtedly suffers the related loss under the terms of Schedule 8, this is a contractual consequence of the incident, the contractual rules under which this arises are not in themselves unreasonable and it is likely that the sums involved are financially insignificant

On analysis, the issue raised by the Defendants does not properly raise an issue of reasonableness; if it did, the reasonableness is mostly on Network Rail’s side. The issue is really one of causation. The loss of revenue quantum is based on the amount of delay in minutes suffered by trains (and passengers) as a result of the five incidents in question. Schedule 8 and the related Delay Attribution process secures in a robust and careful way an attribution of this delay to amongst other factors incidents such as occurred in these five cases, which are the risk and responsibility of Network Rail. Because sometimes it is just not possible to attribute some few minutes to any particular factor, responsibility for these unidentified minutes is, pragmatically, split between the TOCs and the Network Rail. Up to this point, this is the function and result of the Schedule 8 Delay Attribution process.

It is Schedule 8 which mandates that the 2nd 50% of the unidentified minutes is to be the responsibility of Network Rail. Of the balance, the 1st 50%, this is to be split pro rata as between the TOC and Network Rail; if that day, all the other Delay Minutes are the responsibility of, say, Network Rail, it will be responsible for the whole of the 1st 50%, and vice versa if the responsibility is all the TOC’s. On the example given in Paragraph 133 above, of its share of 2.77 minutes, 0.9972 minutes (18/50) would actually relate to the fact that 18 of the 50 minutes for which Network Rail is responsible to the TOC is attributable to the relevant negligent defendant’s road on rail incident; I will call this the “Pro Rata share of the 1st 50%”. What then happens is, as I have understood the evidence, is that Network Rail, with its share of the unidentified minutes or indeed seconds and possibly without more, allocates or claims for the Pro Rata share of all such (and possibly other) unidentified minutes to the negligent (actual or potential) defendant who caused the incident on the day in question. What I have found difficult to ascertain is to what parts of the unidentified minutes Network Rail seeks to apply the Pro Rata share; I will therefore deal with the issue in principle and leave the parties to ascertain the impact.

As for the 2nd 50%, no part of this is recoverable by Network Rail, even though it is Network Rail’s responsibility under the TAA, because that share of the minutes has not been proved on a balance of probabilities to have been caused by that negligence, at least in the five cases here. The fact that the Delay Attribution process, robust, well-staffed and carefully done as it was, has been unable to identify that these 2nd 50% unidentified minutes delay were caused by the incident giving rise to the established negligence points to it being difficult to conclude with any confidence that they were probably caused by the negligence. Therefore, and unless there is some other evidence which is able to point to the probability that this share of the unidentified minutes was caused by that defendant’s negligence, it is not proved to the requisite standard. It is not proved here that any part of the 2nd 50% was caused by any defendant’s negligence.

As for the 1st 50%, there is a sufficient causative link in relation to the Pro Rata share of the 1st 50% because Network Rail has to pay and has here paid out such share of the unidentified minutes as the direct result of the negligence of the relevant defendant. It is justified as reasonable because (a) but for that negligence that Pro Rata share would not have had to be paid out by it and (b) given the robust and fair nature of the Delay Attribution process, the unidentified minutes will invariably be few on any given day and Network Rail takes responsibility for most of the unidentified minutes. It follows from the above that Network Rail is only entitled to recover revenue losses in relation to unidentified minutes for the Pro Rata share of the 1st 50%. For the rest of the unidentified minutes, causation has not been proved in these five cases. Whilst Network Rail’s approach is pragmatic and reasonable, it does not prove that the causative link between negligence and the loss represented by these other unidentified minutes is established. For those minutes, the probability of the cause of those minutes delay being the negligence of the given defendant could in logic be anything between zero and 100%; that the cause could in theory well be the negligence is, without more, immaterial.

It follows that Network Rail has not proved those parts of the revenue loss attributed to unidentified minutes, save in relation to the Pro Rata share of the 1st 50% of the unidentified minutes. I have become wholly unsure as to what this share is in relation to any of the cases, and I will leave the parties to agree what, if anything, relates to the this, which is to be deducted from the otherwise established loss.

The “Higher Delay” principle, as it has been called, is applied because Schedule 8 and the incorporated Delay Attribution Guide calls for it; it is part of the base calculation used to calculate the Schedule 8 liability (in this context being a different source and mechanism to the unidentified minutes process). It addresses “Reactionary Delay”, which is knock-on delay caused to a train by delay to another train or trains, and how it is to be attributed. Paragraph 4.15 of the Delay Attribution Guide calls for such delay to be attributed to:

“…the principal incident (i.e. the one that has the largest number of Minutes Delay allocated to it that contribute to the lateness at that point). Where two or more incidents have had the same effect then the Reactionary Delay must be split equally between them.”

This is a relatively minor point. For instance, Mr Kenney said in his 7^th and 8^th statements (as corrected in the later statement) that for the service groups which represent the largest single contributions for each of the five incidents only at most 18.2 minutes delay out of 3,120.5 minutes delay (0.58%) arises out of the application of the Higher Delay principle.

The Defendants argue that, although this arises out of the application of the Schedule 8 calculation, it is unfair and unreasonable because it may result in the Defendants having to pay for minutes delay which on analysis may be attributable to the lesser cause of delay, when the cause of the greater or “Higher” delay is one of the five incidents in the current cases. Their Counsel argue that it would be more fair to adopt a ‘proportionate attribution principle’, such that the “Reactionary Delay” is addressed with each incident which has affected the train in question giving rise to an apportionment such that a pro rata proportion is applied. In the example given at Paragraph 132 above, the 4 minute delay to Train B would be attributed as to 9/14 to the signal failure (2.57 minutes) and as to 5/14 to the bridge strike (1.43 minutes).

Another example was given by Mr Kenney in evidence, relating to the Godley incident which happened on 7 April 2011. The same rail service groups were affected that day by three major incidents including the Godley incident (referenced 843094), the two others being referenced 842664 (coincidentally a bridge strike) and 843238). 136, 324 and 123 minutes were allocated by the PEARS system respectively to these three incidents. There was a substantial overlap between the three incidents because some of the affected trains are the same for each incident. Delays attributed to 842664 included ones of 53, 30, 24, 20, 18, 15, 13 and 10 minutes compared with those of 29, 22, 16, 15, 12 and 10 minutes attributed to 843094. The application of the Higher Delay approach is not (invariably or possibly ever) often going to result in an exaggeration of the delay caused by road on rail incidents such as occurred in the five cases. This suggests that there is an element of “swings and roundabouts” in the application of the Higher Delay approach and in any event, almost invariably this approach is only going to relate to Reactionary Delay, which is not the direct delay where the bulk of the loss tends to arise.

I do not consider that there is anything in this point. There are two or possibly more ways in which the attribution might be made, by the “Higher Delay” or by the “proportionate attribution” approaches. I see nothing unreasonable about the use of the Higher Delay approach. If the causes of delay are equally causative of the time delay, the time is in any event split equally between them. There is a good chance that in many cases the Higher or longer delay will be the prime cause of the Reactionary Delay. There is no reliable evidence in the five cases that any injustice or unreasonableness has been or would be inflicted on any of the current Defendants by the application of the Higher Delay approach or that the selection of one approach rather than another makes any significant difference; it does not necessarily or at all produce some obvious bias in favour of Network Rail and against the five Defendants. Indeed, defendants such as these Defendants may do rather better out of the application of the Higher Delay approach. I agree with Counsel for Network Rail that the Higher Delay approach is a reasonable one, albeit to some extent a “rule of thumb”; proportionate attribution is no more fair or reasonable. The application of the Higher Delay approach does not lead to anything other than a reasonable assessment of the loss of revenue suffered by Network Rail as a result of the five incidents in question or so far as I can ascertain of comparable incidents.

This point made by the Defendants emerges in two ways:

These arguments were not positively supported by Mr Palmer because, although, in his first report, he said that if the given incident had not occurred other delaying events would probably occur, he qualified that in his second report by saying that he saw the force of the observation that this was unlikely to introduce any significant overestimate of the impact of an individual incident, going on in evidence to say that this was all outside his expertise. He did however under cross-examination (Day 6 Page 95) accept that this would have only "a marginal…or small impact". In their Closing Submissions, the Defendant’s Counsel eschew this. They base their arguments, in part upon what Mr Kenney (principally) said and in part upon what they consider to be logic. The fact that the Defendants were unable or unwilling to deploy any expert evidence in support of their arguments undermines their case on these arguments.

The arguments stem from a consideration of the way in which the RICCS software is used by Network Rail in calculation of the delays and losses attributable to the single road on rail incidents. What the software does is to assume that none of the delays attributed to those incidents and all of the delays attributable to other factors on the days in question would still occur but also assume that no other delays would have occurred on those days other than those which did actually occur.

In summary, Network Rail and its Counsel argue that these points are illogical and wrong and in any event speculative and unrelated to the facts of the current five cases.

Addressing the first argument at Paragraph 145 (a), there is no reliable (nor, so far as I can ascertain any real) evidence that, for the five cases with which the Court is here concerned, it arises or is supportable as a matter of fact. Even if it did arise, the immediate and proximate cause of the train cancellations and delays (all as set out in the Tables in Paragraph 3 above) in the five cases was the respective negligence of the five Defendants. This is even clearer in the case of cancellations which have only been caused in all probability by each of the relevant incidents addressed in this trial. In truth and in fact the negligence was the cause of the cancellations and delays.

The second argument is a little more complicated by the factual scenario upon which it is based. What the Defendants sought to do in argument and through cross-examination of Mr Kenney was to seek to compare the day on which each of the five incidents occurred with a weekday average. To illustrate the point in cross-examination, Counsel produced a graph for one service group relating to each of the Handy, RHL and Parry incidents which showed over the four week period (during which each such incident occurred) the delays in minutes (termed "performance” or “lateness” minutes by Counsel), including the delay on the day in question attributed to the incident in question. It shows a weekday average over the four week period excluding the day of the incident. The four week period is the accounting period for which there are doubtless 13 in any given year so that the day of the incident in question may be in the first, second, third or fourth week. The point in effect being made was a statistical one to the effect that allowance should be made or credit given against revenue loss damages otherwise being claimed for what is said to be the statistical probability that, even if the incident in question had not happened, other delays could well or even would have happened. The argument does not work particularly well on the graph produced for the Handy incident because the average delays of two minutes occurred in any event on that day. It works better on the Parry graph because almost 50% of the performance or lateness minutes attributed to the incident in question are purportedly shown by the graph as likely to have been incurred in any event.

On any logical analysis, the second argument is far-fetched and unjustifiable. Taking the accounting four week period as the appropriate period is random in that the four week period is not necessarily representative of what normally happens. For instance, if one is trying to look at an incident on a weekday as opposed to a weekend, because there might be different problems over a weekend, it might be arguable that one should bring in only the weekdays from the averaging exercise; one might need to exclude from the averaging bank holidays; if the incident in question occurred at the end of the four week period (such as the RHL incident), it may be more sensible to take the four weeks around the incident if one is going to do the exercise at all. Secondly, the weekday average is inflated by other serious incidents which are unrelated to the incident in question. Thus, for example on the Parry graph, the Parry incident occurring on 26 September 2005 and recording 6.757 performance minutes, some serious delay occurred on 2 October when 15.936 performance minutes are recorded. There are several other days in that period such as 6 and 9 October (and arguably more) for which the graph indicates other serious delays. It is at least not uncommon when trying to work out a sensible statistical average to remove what is sometimes called "outliers" from the calculation. The average for the four week period, when one takes out those three outliers as well as the performance minutes attributable to the Parry incident, is close to the other delays which occurred on the day of the Parry incident. The same applies to the other two graphs relied upon by the Defendants. I do not accept that this exercise is or has begun to be demonstrated as realistic because, as ever, it depends what one feeds into the averaging. Thirdly, the reality is that the Schedule 8 and the consequential RICCS exercises do effectively account for all the causes of delay on the day in question (except for the "unidentified" minutes, for which see above); therefore it is a reasonable assumption and inference that but for the incident in question there would only have been the other recorded delays and no minutes worth of loss is claimed for those delays. In relation to this, one can have regard to the Godley incident (in respect of which no graph was produced) on the day of which there were three major and a number of other causes of delay, only one of which was Mr Godley’s negligence where the record documentation referred to by Mr Kenney identifies delays in particular sections of locations between 3 minutes and 324 minutes delay for a particular Service Group operated by Anglia Inter City. Every Minute Delay is identified and there is no reason to assume or speculate that anything different would have happened in relation to the other causes and duration of delays if the Godley incident had not happened.

It has not effectively been demonstrated that this weekday averaging approach is sensible or appropriate. There is no reason to believe that it is any more appropriate or reasonable than the approach adopted by Network Rail in the five cases. It would be wrong to seek to guess that this averaging approach is realistic when inadequate explanation and no reliable expert evidence has been provided by the Defendants to support it. Indeed, one can say with confidence that Network Rail’s approach is realistic and reasonable and effectively establishes in general and specifically in these five cases that the delays upon which their revenue loss quantum is based did in all probability occur and were caused by the negligence of the five Defendants.

In my judgment, Network Rail has established in the five cases that, as a result of the admitted negligence and of trespass, it has suffered the losses set out in the Tables at Paragraph 3 above, subject to a reduction in relation to part of the "unidentified minutes". Apart from the losses which relate to the inspection and repair costs which have been admitted and paid, I am satisfied that it incurred the Revenue Losses, as adjusted and as referred to in those Tables, and that such Losses were caused by the negligence and trespass of the five Defendants. Those Losses were sufficiently consequential upon the negligence and trespass, they were reasonably foreseeable and they represent a proper measure of loss in negligence and in trespass. Although I do not consider it is necessary for Network Rail in this case to establish that these Losses are objectively reasonable by way of some type of separate legal test, I am satisfied that the Losses as calculated and assessed through the Schedule 8 and RICCS processes are in this case objectively reasonable and that those processes were reasonable ones. I have indicated above that, if there is some sort of objectively reasonable test to establish in cases like these, the burden of proof is on the Defendants; not only have they not discharged this burden but also it has been positively established that the Losses are reasonable, reasonably calculated and based on reasonable assumptions. Contrary to what Aviva believe, they were at a “reasonable and realistic level." The whole basis of establishing such losses by reference to Schedule 8 is reasonable, not simply because Schedule 8 is logical, well researched, carefully constructed and well supported but also because, when I have come to consider the detailed challenges, they are not made out and, indeed, are positively wrong.

I have indicated in this judgment that the proper approach of the Court in cases like this, for instance where there are claimed revenue losses consequential upon negligence on the part of a defendant which arise by way of the application of a formula or basis agreed upon by the claimant with a third party, is and should be a general or "big picture" approach. Thus, the Court should be directed to clear, obvious and readily provable or exceptional matters which demonstrate that the agreed formula or basis is obviously unreasonable or unsustainable. A micro-analysis (as I have called it) or a "minute scrutiny" (as Jackson LJ referred to it) will rarely, if ever, be an appropriate exercise for the Court. I have considered the case on both the general and more detailed basis and have reached the same conclusion by way of each.

In relation to the "unidentified" minutes applied by Network Rail where they have sought to allocate a proportion of these minutes delay to one or more of the five Defendants, this small element of the claims does not fail as being unreasonable as such. Save in relation to to the Pro Rata share of the 1st 50% of the unidentified minutes, it does however fail because, as a matter of simple causation, the link between the negligence or breaches of duty on the part of the Defendant and the unidentified minutes of delay has simply not been proved on a balance of probabilities. Network Rail has established entitlement to the loss of revenue attributable to to the Pro Rata share of the 1st 50% of the unidentified minutes. I will leave it to the parties to agree the amounts to be deducted, if any, from the five claims and, in what I hope will be the unlikely event that agreement can not be reached, the Court will have to decide upon such deductions.

In answer to the issues formulated (see Paragraph above 7), the sums claimed (as agreed) are recoverable against the Defendants save in relation to the unidentified minutes and:

Subject to the minor exception relating to part of the unidentified minutes, there will be judgment for Network Rail in each of the five cases.

I am conscious that in one sense this is a disproportionately long judgment in that the amounts in issue are not substantial being in the region of some £400,000 in total. There was however a mass of evidence, not limited to the detailed records of the five incidents and the related delays and cancellations of trains. The expert evidence, particularly from Mr Segal and Mr Palmer, but supplemented by extensive background evidence from Mr Angus, ran to many thousands of pages. I am conscious that the trial in general and this judgment in particular has descended (or possibly ascended) into the type of detailed, extensive and expensive investigation, the likes of which were expressly discouraged by Lord Justice Jackson in the Conarken and Farrell Transport judgments; it has been necessary here to engage in an even more “minute scrutiny of the formulae and calculations” (Paragraph 152 – Jackson LJ) than in those cases. The length of the judgment is, I hope, justified by reason of the need in justice to address the many detailed arguments and nuances raised by the Defendants, many going to those formulae and calculations, and by the possibility (at least) that it will be treated not only by the Defendants and their insurer but by other similar insurers as a test case so that the type of detailed investigation which has proved to be necessary in this case will not have to be repeated.