ISSUE B/04/86
COSTS AND BENEFITS OF
COASTAL VESSEL TRAFFIC.
SERVICES IN EUROPEAN WATERS
R. 0. Goss & J. E. Halliday
Department of Maritime Studies
UWIST
Cardiff
Athrofa G.wyddoniaeth a Thechnoleg
Prifysgol Cymru
The University of Wales
COSTS AND BENEFITS OF
COASTAL VESSEL TRAFFIC SERVICES IN EUROPEAN WATERS
A Report by
R 0 Goss and J E Halliday
of the
Department of.MariUrne Studies
University of Wales Institute of Science and Technology Cardiff
Chapter 1 Introduction
2 Cost-Benefit Analysis: An IntrOduction tohe Technique
3 The identification and Specification of Effect.
Human Accidents
5 Shipping Casua1ties Ships and Cargoes
6 Environmental Effects
7 The Cost of Establishing Shore Support
8 Salvage
Search and Rescue
10 Delays and Acceleration to Ships and Cargoes
13 Summary and a Worked Example
This is the final report of task 2.43 of the COST 301 programme. The task,
the requirements of which are detailed as Appendix 1, is essentially an
exercise in cost-benefit analysis and forms one compOnent in the contribution of Working Group 2 to the COST 301 programme. This remit cornpriséd the basis of a. subcontract between. the City of London Polytechnic (COLP) - who were
main contractors to the EEC and Department of Transport - and Department of Maritime Studies at the University of Wales Institute of Science and Technology (UWIST), Cardiff, and covered the period 1.10.84 to 30.6.86.
There were considerable delays and difficulties in our obtaining the
information we needed from other participants in. COST 301; for this reason
our contract period was extended. The aggregate cost to public funds was £21,800and this was not increased by these delays and difficulties. In
later chapters we describe some of the difficulties we. experinced, In both theory and practice and it will be obvioUs that, with greater resources, more
could be done. We do not believe that more could have been doe within the resources available to us.
The context for this report is provided by the COST 301 programe, the stated objective of which is to "recommend shore-based aids tO navigation which will reduce the potential danger of navigation to human life, the environment, ships and cargoes"1. Within this framework a meeting. of WOrking Group 2 in July 1983 agreed that the U.K. should be responsible for the calculation of
risk' to marine traffic in European waters2. Subsequently the associated
framework for the cost-benefit analysis Is provided by the calculation Of. risk to marine traffic in European waters referred to above and developed under task 2.44 (PolytechnIc of North LondOn (P.N.L.)) and task 2.45 (COLP). Under these terms of reference European waters have been divided into sea areas capable of description by the value of a. problem area Identifier (PAl), which reflects environmental and traffic parameters. Details of ,this
methodology which forrns.the background for the. development of cost-benefit
analysis in this report can be found in the final. re.port of task 2..
The methodology for the cost-benefit analysis had therefore to be appropriate to European waters at the level of these seaareas and tc be capable of using information (sub tasks 2 & 3) from other researchers working within the same
framework. In this context it should be stressed that this work therefore concentrates on shore support relative to collisions and groundings involving merchant vessels of 100 grt and over, it also concentrates on VTS in the open sea, excluding ports .and port approaches from the analysis. The
requirements of the model cost-benefit analysis at the. level of any sea area is then to provide a methodology by which the costs and benefits of various mechanisms for achieving the COST 301 objective can be objectively
evaluated. As will be seen in later chapters, significant difficulties arose, not least in, respect ofthe avaflability and compatibility of
statistics. . . .
COST 301 however does not have a single but rather a tripartite objective, focusing on three distinct topics of concern: safety of navigation, traffic
efficiency and reduction of pollution ri,skL. The. first expresses the change
in the frequency of maritime casualties consequent on an alteration to the navigational environment. The second refers to the changing efficiency with
same changes to the navigational environment. The third relates to the changing level of risk to which the maritime and coastal environment is
exposed. -
-In respect of any one proposal some of the effects experienced in any of these three topics may be negative; some positive. That is, there, may be
associated costs as well as benefits. Different schemes will alter the relative distribution of these costs and benefits within any givensea area; whilst the implementation of similar schemes in different areas will usually be associated with different effects.
We tus need a methodblwhfch
will
adequately reflect both the diversity of sea areas identified by the PAl and different vessel traffic system (VTS) options.
-This study is essentially conceived in a practical rather than any purely theoretical context and thus concentrates upon the costs and benefits of VTS in practical circumstances. As such it needs to ecognise that, over large areas of the world, some rudimentary form of VTS already exists, at least as
a standby, through the widespread use of VHF and other forms of communication for search and rescue (SAR) purposes. Moreover, just as SAR facilitates can
exist in isolation so may a traffic separation scheme (.SS), ie without a
VTS.
Such more or less elaborate forms of VTS as those with which we are concerned (going far beyond SAR facilities on their Own) are only likely to be
instituted where there is evdence of the conflict of sea users, eg in the form of congestion and where,therefore, a TSS has already been established or is likely to be established. This study is not concerned with
the costs
and benefits of TSS's as such and is therefore best considered in terms of the increments of costs and benefits associated with adding a VTS to a traffic separation scheme, whether the TSS previously existed or not.
contribute to the COST 301 programme.
The first stage Is to define
cost-benefit analysis, to provide a background to its use in COST 301 and to examine some of the limitations of the technique. This forms the basis of
the next chapter.
It will then be clear that three distinct processes are
necess'ap4 if
cost-benefit analysis is to be applied to COST 301. First, 'there is the specificatIon, or identification
in qualitative terms, of the effects of a given proposal.
This is based on the three types of effect identified by the COST 301 objective, together with the costs of
establishing, maintaining and operating the necessary shore
support system, and is developed in Chapter 3. Secondly the specified effects need to be quantified in physical
terms'and thirdly these need to be evaluated in order that the total effects
of a
proposal can be expressed and various project options compared.
Chapters 4 to 11 therefore expand upon each of the categories
Identified in Chapter 3, describing the organisatlons contacted, the data or Other response received, the availability of Information inside and
outside COST 301 and the problems encountered in the process of quantification and evaluation.
Chapter 12 then firstly relates the identified effects
to the output of the supporting tasks 2.44 and 2.45, that is to the results
of' the problem area
identifier and risk analysis and. to the results relating
the 'effects of
different levels of V.T.S. to the navigational
environment. It then proceeds to draw 'together all
the factors previously discussed in order to provide a model
cost-benefit analysis which is capable of application to any of the identified sea areas.
The model is then tested against reality
in
Chapter 13, where a' worked example is provided for
transport modes it has not been much used in the maritime field apart from the recent and noteworthy work of the Canadian Coastguard5. We have
therefore been conscious that this is, to a significant extent, a pioneering
study. We. have posed new questions, frequently finding the answers are not readily available and we have sought to develop objective techniques for ourselves or to suggest them for the future. In addition therefore to our function of trying to assist the final decision maker, we se'e a secon6 function, no less useful, of exploring the: difficulties which stand in the way of work of this kind. We describe and draw attêhtion therefore to the difficulties we have experienced and in the final chapter, the conclusion of
the report, we. also make suggestions for further research. Because some of
our work is likely to be controversial it is worth emphasising that our general findings are neither to Oppose VTS nor to support it In all
circumstances. The usefulness of our work is rather to clarify the costs
and benefits of a coastal YTS for commercial tra.ffic and there.fore to assist in.
decidin.g when a YTSIs justified. We have, therefore, given a detailed
account of our theoretical, and practical approach to the problems and data in this field, showing how and why we used certain figures. Should revisions, be thought necessary, more appropriate figures may then be inserted.
We are grateful to numerous colleagues in the numerous bodies co-operating in COST 3O1 in United Kingdom Department of Transport and elsewhere for thei'.r
comeflts on early drafts of. this work. We remain, of course., responsible for
1 COST 301/PMS/034 1982: Preliminary Project Plan,
Issue
B/Q9/82.2 See eg Kemp J F 1984: COST 301: Report on the UK Contribution to the
Working roup 2 Progranmie 13.6.84. (Unpublished paper presented to,a
project meeting at COLP, 16.7.84.)
3 KempJ etal 1986: Assessffient
of
Risk tø Shipping Through Collisions and Strandings in the COST 301 Area, COST 301/UK/COLP/T2.46 0P6.COST 301 1984: Objectives and Interim Plan of Work, September 1984.
I
5 See Canadian Coastguard 1984: National Vessel
Traffic
Services Study
they may accrue are In excess of the estimated costs"3.
The technique has since been used and developed extensively In the transport
field. In Britain the method was first applied, for instance, to roads through the London-Birmingham motorway study, to the underground railway system through the Victoria Line project, to a surface railway, the Cambrian Coast line, and to airport location by the Commission on the Thi'rd London
Airport. Largely as a result of such earlier studies it is accepted today
as a standard evaluation technique., widely recognised throughout the WOrld.
Despite the inclusion, however, of navigational projects within the formative Act referred to above, together with work of increasing quality conducted by
the US ArnW Corp of Engineers, there has been little development of the practical application of cost-benefit analysis in a maritimecontext. The
Canadian Coastguard VTS Report referred to in the introduction and in later chapters is a most useful exceptiOn.
The established methodology thus provides this research with a well developed body of technique, yet there is also a clearly defined need for
methodological development and: application in the maritime field.
2.3 The Main Features of Cost-Benefit Analysis and its Limitations
Drawing firstly on the established methodology certain salient features may be mentioned in the context of this application. Thus cost-benefit analysis
is not a substitute for decision making: rather it is an aid to the
decision-maker. It is capable therefore only of identifying and minimising those areas in which subjective judgement is required. 'This is achieved by
effects, their physical quantification, neither of which is necessarily straightforward, and their evaluation. It is this last which.eflables the
comparison of otherwise physically diverse quantities and is the essence of cost-benefit analysis, enabling disparate costs and benefits to be expressed
in terms of the only available comon unit, money. It is, moreàver, only the
last of these which is the particular province of the economist
Certain important principles and several limitations apply to this process of
evaluation.. The first concerns the allocation of the appropri&t values to the costs and benefits. With reference to benefits the crticial concept is
the change in 'consjers' surplus'. This is 'a measure of the' difference
between the value that consumers place on their total consumption of some comodity and the amount they must pay for it'5, that is it is the area b.etween the demand curve and the market price. The benefits calculated are thus a reflection of the aggregate willingness to pay of the beneficiaries.
In the diagram below, for instance, a fall in price from P1 to P2 wIll result in the expansion of cOnsumerss surplus by the area P1ABP2
Price
'per
Unit
P2 0 D Q tyP2
Price
per
P1
Unit
A related and also important concept is producers' surplus. This is the excess Of producers' receipts over the minimum that would have to be paid in order to persuade them to produce a given quantity. It is thus the area between the supply curve and the market price.
In the diagram below, for instance, a rise in price from P1 to P2 will result in the expansion of producers' surplus by the area P1ABP2.
£
MC.
Q ty
In most if not all transport modes the 'price' is best considered as
'generalised transport cost', le SO as to reflect the value oftirne spent in transit as well as such money costs as those of fares and freight rates.
Changes in Other factors, such as risk of loss or damage may also be
included; obvioUsly in COST 301 they are of the essence. These topics are not simple and are discussed later In this Report..
The cost of the project against which the benefits are tO be compared is similarly not necessarily a simple cost, in terms of money spent; rather its
'opportunity cost' is needed, that is the benefits foregone in order to employ the relevant resources in the project in question. In many Instances the money price will fail to reflect this opportunity cost and it may then necessary therefore to use what is termed a shadow price; a price that more
accurately reflects society's valuation of an input or output. This is
particularly important for instance where current market values diverge markedly and obviously from long term costs; or where they are non-existent. An example of the first may be found -in the present state of the charter market; much depressed by excess tonnage relative to the demand for it. An example of the second may be found in the absence. of any value for human
life. In such circumstances, it is no business of the
analyst to impose his
own values on the study: he needs, somehow, to derive those,values
which
appear to be consistent with society's values.
A second important evaluative aspect of any cost-benefit analysis is the need to reflect society's preferences about the incidence of costs and benefits
Over time. The positive time preferences Of society, combined with the Opportunity cost of capital, means that future benefits and costs need to be
'discounted' to reflect present values.. The rate at which the benefits and
costs are discounted is known as the discount rte. In the absence of an official EEC rate our
study
therefore uses the UKgovernment rate of 5%,
which reflects the best assessment known to us of. the. social
opportunity cost
of capital.. The resultant NPV then reflects all identified
components of the calculation, including their timing and may be compared directly with the NPVs of other proposals or with. the 'do nothing' position.
The UK government currently
uses 7% for cOst-benefit analysis but we see no logic in any such distinction, given
t.hat all investment.projects may be
assumed to produce externalities
and there is no knowing whether a project that is marginal when appraised in csh terms has externalities that are, on balance, positive or negative. n this context, however, we do
not think the
There is a danger, nevertheless, tn trying to push our economic appraisal further than our real Understanding would currently justify; but many of the criticisms which have been made i.n this area can be answered by repeating that cost-benefit. analysis is not a substitute for decision making. It does
not attempt. therefore to subjugate all decisions to the dictates of monetary
valuation: it cannot. Those factors which have not been adequately
quantified or evaluated may be listed, together With necessary qualitative
coiwnents. (Income distribution effects are a cornon examp}e of this, though
fortunately of limited relevance here.)
2.4 Risk and Uncertainty
Cost-benefit analysis attempts to approach an objective evaluation of the decision-making environment. The methodology has, however, finally to recognise and take into account the elements of risk and uncertainty which may characterise this decision naking environment. It has been suggested that whilst these. two words are frequently used inter-changeably they possess distinct characteristics in an economic context, and thus require a different
treatment6.
Risk occurs in a context where a range of outcomes can be identified and the
probability that the. benefit or cost wll.l take on a particular value can also be identified. Uncertainty occurs when the range of outcomes is known but the probability distributionis not7. Other definitiOns are however more
ambiguous. . Risk for instance has been defined as exposure to mischance or peril, whilst uncertainty It has been suggested is Indeterminacy or subject
to doubts.
The following chapters, which discuss the individual elements of the cost-benefit analysis, underline the particular areas where these two
problems are encountered, It is, however, u.eful at this stage to outline the general scope of the problem, its theoretical dimensions and the proposed
solution.
Risk/uncertainty can be identified as posing a problem in four general areas of our analysis.
1. Risk and uncertainty pertain both to the qua.ntificaon and to the
evaluation of each of the discrete inputs into the co. ;nefit
analysis. consider for i!istance firstly the required land uptake
and range of equipment. necessary for a yTs centre and secondly the value of this land and equipment. The unctainty surrounding these. two elements is multiplicative, not additive. The following
hypothetical example graphically illustrates the repercussions. of this factor for the potential margin of error.
io of units req'd (Q) ° of uncertainty (U) Range of 0 under U1 and U2 Price of units (P) ° of -uncertainty (U) Range of PMargin under U1 and U2 of Range (PxQ) Under U1 and U2 . Total Possible Variation Under U1 andU2. 100 Ui ± 10% Mm . 90 units . £10.00 ± 10% . Mm £9.00 Mm -.90 unIts x £9.O0=810 Max 110 unIts x £11.0O=121.0 U1 40O - -:;:. . Max £11.00 Max. 110 U2 ± 50% Mm 50 units -U 50% Mm £5.00 . -Mm 50 units x £5.00=250 Max 50uflitsx £15.00=2250 U2 £2,000 Max 150 Units . ± 1 Max £15.00
The situation is however further complicated because risk and uncertainty are also features of the event and its outcome. One' must alSo consider therefore:
The risk attached to the occurrence of the event. For example., how
often will a collision/grounding involving a ship/ships of a defined size and class, carrying defined cargoesand resulting in serious/non-serious damage occur in a particular area, under particular
environmental conditions?
There is uncertainty áttachèd to the outcome of the event déscribéd above If it occurred now. Consider fOr instance the uncertainty surroundirg the effect of a given event on the natural environment or the tourist industry, due for example to lack of knowledge of the factors affected, or the variation through time of
these, affected factors9.
4, There is uncertaintY attached to the outcome of the event
described above if it occurred in the' future. Consider for instance the likely changes In ship technology, pollution control or the
values and perceptions of,society during. the 'life of the project.
No one can be sure what changes will occur in such fields: if they
could, then they would make the changes now.
-Treatment of Risk and Uncertainty
Several methodologies exist for the treatment of risk and uncertainty. A
comprehensive review of residual, risk is provided'for instance by O'Donnell and Rhodes who differentiate between probabilistic and non-probabilistic
classification adQpted by Pearce, who differentiates between techniques appropriate to risk and those appropriate to uncertainty, the, terms having already been set out above as those where, respectively, a probability distribution is, and is not, available. In subsequent chapters of this report., it. will be seen that not all the techniques described could be
profitably applied in those areas of risk/uncertainty which pertain to the COST-301 project. Furthermore no one technique is entirely
satisfactory11.
Probab ii i stic Techniques
ai) Expected Values and Dispersion Measures
The simplest way in which risk can be described is by recording the distribution and incidence of the probable values, if known.
The expected value fOr instance, indicates the value to be expected on average.
Ifl this
instance the most relevant average is the mean and the expected value is, importantly therefore, not necessarily the same as the most likely value., which would correspond with an alternative measure of average, the mode.
The. expecte.d value expressed alone may not however reflect the risk associated with a project.. Two distributions,
one encompassing a narrow range of values, the Other a wide range, may have the same expected value, yet the risk associated with the greater variation is larger than that associated with the more narrowly bound distribution.
If society is
risk-averse (or cautious), as opposed to risk-neutral (indifferent), or risk-loving, it would tend therefore to choose the project option with
the narrow
spread of values.. This is illutated in
the diagram below. The pread of values can be indicated by the variance (the average (squared)
difference between each possible outcOme and the expected value) and is commonly
expressed as the standard deviation or the co-efficient of variation. It
should also be borne in mind that a Poisson distribution, rather than the depicted Normal distribution1 may, in fact, be a more accurate representation of the incidence of ship casualties where there are a few large casualties and many smaller casualties.
0
Value
In order however to incorporate society's attitude to risk, as reflected by the range of outcomes, as well as the size of the outcome, the expected value has to be extended. The revised measure is termed:
all) Certainty Equivalents
if society Is risk averse the utility that arises from a benefit/cost that occurs with certainty is greater than the utility that arises from a
benefit/cost whose occurrence is the subject of probability. It is
conceptually possible therefOre to express the certainty-equivalent value of any probabilistic outcome and this reflects the cost of bearing the risk.
In order however to realise the methodology we need to know Society's utility function With respect to the benefits and co$ts. Since we know little of this it has been suggested that this is not a practical measure for
aili) TheArrow-Lind Theorem
The problems inherent in measuring society's Utility function may hot be necessary for certain projects if the conditions of theArrow-Lind theorem
are upheld13. This suggests that the larger the group across which the costs/benefits of a project are spread the less the risk to an individual. As the number of individiials approaches infinity, so the project risk
approaches zero and thus for large projects the public may b.assumed to be
risk neutral. Expected values alone would thus sUffice.
The theorem does not however apply to public goods or bads1 and such public.
externalities as potential pollution of the environment, or risk to life and limbs, are manifest in the field of navigational safety. (Indeed, they
represent the point of studies like COST 301). Any increase in the
eflects of these externalities by any one individual does nothing therefore to reduce the consumption by others, so however large the populatiOn exposed, the risk is not reduced. Moreover the groups suffering these externalities are also far frOm infinite (consider for instance seamen or the residents of particular coastal localities) so that there is a Umit to how far the risk,
can be considered pooled..
Other methods also exist. These include beta factors (based on stock market risk premia), Monte Carlo analysis (a stochastic computer model) and decision trees (where the relative probabilities of specific pay-offs can be assigned to a series of events in a strategy enabling different risk options to be
compared). In fact, however, none of these is practicable in the present context and for a variety of reasons, some of which are described below.
Non-Probabjil stic Techniques
Those techniques which incorporate uncertainty are based on non-probabilistic
techniques. Avery simple measure Is: bi) The Payback Method
This measure is the time needed for the total benefits to equal the total costs, that is the time for which the capital investment i.s at risk. It can
therefore be used (in. conjunction with the NPV and internal rate or return)
as a somewhat crude indicator of project riskiness, though for general
project appraisal it has many disadvantages, eg in ignoring all events after the period's completion.
bii) Decision Rules
A more rigorous approach is based on decIsion theory rules. The first stage is typically a pay-off matrix which records policy choice against the range of possible environmental conditions and assigns an outcome for the project under each combination. Various rules can then be used. as a guide to project selection, each rule reflecting a different attitude on behalf of the
decision maker15.
The most connofl are.
1. Maximax/maxima
This is essentially an optimist's rule Which selects that policy whose best outcome is the best of all possibilities, but ignores the
Maxirrin
In contrast this is the pessimist's rules which chooses the policy whose worst outcome is the best of all worst outcomes (ie the maximum of the minima)! It again. ignores all others.
Maximum 'avr4ge' benefit
This chooses that policy which results in the best outcome when averaged across all possible environmental conditions.
Other rules rest on a matrix ihich records not the cost of making a choice but the cost of making a wrong choice, derived from the difference between the actual outcome and the best value which could have been obtained in that state of the world. Minimax regret for instance selects the minimum of the maximum regrets.
Typically, however, such calculations require a considerable amount of
detail, g as to inputs and outcomes much of which, as we shall see, is not available here. One of the most comonly used techniques capable of
practical. application and the one we intend to use
in this exercise is
therefore:
bili) Sensitivity Analysis
Essentially sensitivity analysis ifldicates what would happen to the overall outcome of a project if the assumed values are varied by sOme given amount,
such as 10%. Intrinsic risk and uncertainty can therefore be expressed in terms of a range of estimates and the consequences of plausible changes for the decision criterion examined. This unables the more important areas Of uncertainty to be identified so that research can be concentrated upon them.
It is thus particularly appropriate where uncertaintY extends over. many Items in the analysis and when there is little information on dispersion. It may subsequently be possible to reject many Of the. variations, eg when they are found to produce insignificant effects. By repeating this process the key variables may be identified and the decision-makers attention thus
concentrateth In the event, as will be seen later In this Report, we had
little chance of using any other technique
A related concept which could have been incorporated and subjected to sensitivity analysis Is the elusive 'catastrophic' factor.. This is
multiplier of indeterminate order which attempts to reflect the apparently greater impact on the public of a major repoted disaster with,; for instance, considerable loss of life or a substantial environmental impact, compared with the same damage caused by a series of isolated instances. The choice
and use of such a multiplier is discussed further in Chapter 6. We decided however, that this is essentially a matter for political judgemént and not
Notes and References
1 Goss R 0 1982: Costs and Benefits of Navigational Aids In Port
Approaches In Goss R 0 (Ed) Advances In Maritime Economics, UWIST,
pp213-237.
2 Thomsofl J M 1974: Modern Transport Economics, Penguin Education.
3 Cited in 'Pearce D W 1983: Cost Benefit Analysis, The. Macmillan, Press
Ltd, p14 (second edition).
4 See, eg
Coburfl I M, Beesley M E & Reynolds D J 1960: The London
trrnJn9ham Motorway, Road- Research Technical Paper No. 46.
Foster C D & Beesley M E 1963: 'Estimating the Social Benefit of Constructing an Underground Railway in London', Journal of the Royal
Sa.ticaI Society,
Series A, pp46-56.Ministry of Transport 1960: The Cambrian Coast Line. HMS0
Comission on the Third London AIrport, 1971: Report. HMSO.
5 Lipsey R G 1963: Positive Economics.
Weidenfeld & Nicolson, sixth, edition. (1983) p175.
7 For futher probabilistic definitions of risk see for instance:
The Royal Society, 1983: Risk Assessment. A study group report p22. and
Harris N C 1982: Problems. in hazard analysis and risk assessment. In the Institution of Chemical Engineers. Symposium series No 71. The
Assessment of Major Hazards, Pergamon Press. p267-283.
8 O'Donnell A I & Rhodes T E 1983: Risk, tlncertainty and Public Sector
Thvestnient Appraisal, Government Economic Service Working Paper No 63.
H M Treasury London.
9 A good example of an attempt to quantify the risk surrounding an event and its outcome is provided by the Carivey Island Studies. See: The Health and Safety Executive, 1978: Canvey: An investigation of
Potential Hazards from Operations in the Canvey Islands/Thurrock area. HMSO Lnndon ppl-192 and
The Heai-th and Safety Executive, 1981: Canvey: A second report.
Review of Potential Hazards from Operations In the. Canvey.
Island/Thurrock area three years after publication of the Canvey Report. HMSO, London ppl-128.
For further methodological examples see:
Barrel A C and Pape R P 1985: 'Risk Analysis and Assessment Of the
Energency'. Paper Presented to theGreénwich Forum Conference on The.
The Institution of Chemical Engineers 1982: op cit and The Institution of Chemical Engineers in association with
The Royal Institution Of Naval Architects 1983: 4th Inteiinational
Symposium. on Loss Prevention and Safety Promotion in the Process
Industries. Vol
2.
Hazardous chemicals Gases - SafeTrasportby Sea,
12-16
September 1983.10 O'Donnell A 1 and Rhodes
T F 1983:
op cit.11 See for example Graham D A 1981:
'Cost-Benefit Analysis Under Uncertainty'. American Economic Review,
71 (4), pp715-725.
Mendelsohn R and Strang W .1 1984: 'Cost-Benefit Analysi:s Under
Uncertainty: Comment', American Economic Review,
74 (5) pp1096-1099.
Graham 0
A 1984:
'Cost-Benefit Analysis Under Uncertainty:Reply',
American Economic Review,
74.(5) ppllOO-1102.
Klausner R .F 1970: 'The. Evaluation of Risk in Marine Capital Investments', Marine Tectirrnlogy, October
1970, pp449-464.
12 Pearce D W 1983: op cit.
13 Arrow K and Lind R 1970: !Uncertainty and the Evaluation of
Public Investment Decisions, American Economic Review,
60 (3)
pp346-378.
14 See. Fisher A
C 1973:
'Environmental Externality and the Arrow-Lind
Public Investment Theorem', lerican Economic Review,
63 (2),
15 In a maritime context see for example: Norman V 1981: 'Market
Strategies in Bulk Shipping' in Hope E (Ed) Studies in Shipping
Economics, BedriftsOkonomefls FOriag A/S Oslo ppl3-28.
3.1 IntroductIon
The first stage in, the cost-benefit analysis has already been defined as the
identification and qualitative specification of effects. It was obviously important at the outset of our work therefore to have a clear idea of the likely areas where any VIS would have an impact. That is. we neaed to
identify, at an early stage, the component parts of the cost-benefit analysis in order that we could progress to consider their quantifjcation and
evaluation.
A preliminary list: of categories of costs and benefits was therefore
the.
first real output from our research and was put. forward for informa.l discussion within the cosT 301 organisation and later subjected to Wider audiences at the seminar on Safety In European Waters in London1 and the International Law of the Sea Conference in Cardiff2. In the event however,
whilst some coments were received
on individual components, no suggestions were made as to potential additions and no categories were criticised as
irrelevant. These. preliminary presentations
thus served largely to confirm our approach, as did later discussions with other COST 301 participants in Paris and Brussels during the summer of 1986.
(Some helpful suggestions
were, however, incorporated.)
The preliminary specification, outlined
below., thereforeformed the basis of our research into the costs and benefits of navigational, safety.
Chapters
11 which examine these categories in further detail, clearly illustrate how, as the work progressed, our research efforts were, subject
People.killed
People injured.
constraints, able to concentrate on those categories which represent the most important effects of a change in the navigational environment.
It is important to stress at the outset that all categories refer only to those situations which lie within the scope of navigational improvements. This would include, for instance, changes in the incidence of human accidents consequent on collision or stranding but not on cargo handling. Such
distinctions are clear cut in theory but less easy to apply tfl pPactice Where .statistics kept for Other purposes, órnot kept at all, make accurate data
collectfofl a challenge.
3.2 HUman Accidents
One of the, objectives of the COST 301. programme, as stated above (page 1.) is to "reconmiend shore-based aids to navigation which will reduce the potential
danger of navigation to human life"3. Human accident is therefore one
category where any VTS would be expected to reduce future losses below that expected under the do nothing option.
Human accidents fall into two basic divisions:
Ideally, the second component, injuries, needs however to be broken down further to accommodate the different patterns of cost associated with for
3.3 1.ppi_ng CasualtIes (Ships and Cargoes)
Again the inclusion of this category is mandatory, given that the objective
o.f the COST 301 project is also to "reduce the, potential dange.r of
navigation... to ships and cargoes"t". The. introduction or' enhancement of VTS.
would therefore similarly be expected to reduce losses in this category
also.
FoUr sub-components. can 'be identified, relating both to the severity of the
incident and to the distinction between ship and contents..
Ships lost, Ships damaged, Cargoes lost, Cargoes damaged.
As with the human injury category discussed above, a rigorous anal.ysis iould
require further details, including the size and type of shi,p lost/damaged,
the quantity and type of cargo lost/damaged and an indication of the severity of the damage with reference to both ship .andcargo.
3.4 Environmental Effects
This is thethird category of effect' which is lmediatel:y determined'by the
COST .301 objective'. In this' instance 0shore-based aids to navigation which
will reduce the potential danger of', navigation to.. the environment"5.
Essentially this relates to the pollution of the 'littoral, sub-littoral 'and
primarfly Involve pollution by oil but other cargoes such as flammable or toxic chemicals or radioactive materials will also need to be considered.
Specification is difficult because the actual effect will vary considerably, depending on a number of features, Inc-luding location, season, weather and tidal conditions and type Of treatment available. Many of these may be
unknown or inadequately established. Scientific knowledge Is also imperfect. It Is thus often difficult to predict the ecological conseqUences of an
impact and second and third order effects may prove more significant than the first order impact6. Oil pollution may for instance kill salt marsh,
potentially resulting in increased instability of estuari'né deposits, coast erosion, flooding and changes In many other ecosystem components7. A coastal
holiday area which gains a bad reputation may find this hard to shake off, no
flatter what the actual state of its beaches may be.
Bearing in mind the essential quest for evaluation, the environmental impacts divide basically into two groups:
Those with repercussions for consumption, Those with repercussions for output.
The former includes:
i Damage. to marine/coastal biota (excluding the commercial species
included above),
11 Damage to the ablotic environment (again excluding damage which is
commercially accountable),
iii Damage to areas or species of, scientific interest,.
iv Damage to landscape/amenity,
The latter inclUdes:
I The avoidance/treatment Of coastal pollution,
ii Damage to biotic/abiotic elements of commercial significance (eg fish, shellfish, seaweeds, seabed and shore aggregates (sand and gravel) and water quality),
iii Losses of consumers' and of producers' surpluses1 eg in the fishing industry and the holiday industry,
iv Damage to human health.
Again the inclusion of environmental considerations amongst the COST 301 objectives indicates that this too is an area where losses may be expected to
be reduced when. a VIS is established or imposed. Indeed, there are instances
where. this has been a major factor in the decision to establish a VTS. As
indicated above, however, all three objectives are not directionally linked;
thus a system Which produces safer navigational conditions. for ships their
cargoes and personnel, may not neessarily reduce environmental impacts.
It is widely held for instance8 that traffic separation schemes (TSS) may
even have increased environmental risk, that is imposed a cost on
environmentally sensitive and potentially treacherous coastlines by diverting traffic, including potentially dangerous cargoes, close inshore. Conversely relocation of routes, following particularly from the Amoco Cadiz incident, suggests that reducing the risk Of pollution was, in that instance,
considered more important than reducing the risk to life and ship safety. Revisions to both the Ushant and Casquets schemes aimed therefore to reduce
stranding whilst, it. appears, possibly increasing 'collision risks.
Whilst, in terms of COST 301 objectives, it might be hoped to reduce costs (or achieve benefits) in all three areas, trade-offs between them may,
nevertheless, need to be considered. This is one reason why the pioneering work of COST 301 is not straightforward. De Bievre has, indeed, identified
just this combination of fact9rs as calling for extreme caution on the behalf of decision makers9.
One way, it has been suggested, in which VTS may be expected to be effective in all three areas, is by enhancing the general professionalism of seafarers, by increasing their recognition of risks and their appreciation ö'f systematic
approaches. to them. In this one may compare the allied question of road
safety where increased safety measures (eg through road design) have tended to indUce more safety rather than an Increase In speed at the same level of
safety.
3.5 COst of Shore Support
There Is however a fourth category which Is sine qua JIQrJ if navigational safety is going tO be Improved by means of shore support systems, that is the c0sts associated with establishing the VTS control centre or alternatively
adapting existing control functions.
The range of possible areas where expenditure may be required Is considerable
and comprises:
a Land,
Buildings,
C Equipment eg radar and radio. connflunicatiOflS,
d Staff to operate and service the system, e Maintenance of buildings and equipment,
f Other operating costs such as electricIty and telephones,
g Promulgation requirements eg new, amended or over_printed chartsand notices to mariners.
Changes
ifl
the frequencyof maritime casualties. and human accidents may alsohave an impact on allied areas of marine activity. TwO of these appear
impOrtaflt for our work: salvage acti:vities concerfled with the ships and
their cargoes; and search and rescue, primarily concerned with human life.
3.6 Salvage Activity
A reduction in salvage activity would represent a saving to society of the
use ofreal,.re.sources. This would appear as a saving of costs to those industries involved in the salvage business or supplying to The problem of
change at the margin applies here in attemptingto specify effects
qualitatively, for many salvors claim10 that the present incentive to offer
their. services is miflimal. Given therefore that the expectation of profits in this highly uncertain activity is moderate and any effects Of a. VTS would appear only at the margin, it seems probable that there are no significant changes jn either consumer' or producers' surpluses.
3.7 ChangeS in Search and Rescue Activity
If the establishment of a SITS leads to significantly fewer casualties then it is likely that there will also be savings of time and effort in the various forms of search and rescue (SAR) activity! Although our study is essentially Europe-wide, our lim.ited,resources mk it necessary for us to concentrate primarily on UK practice and institutions, extrapolating from these to the
situations in other countries. The institutions and divisions of
responsibility vary, but the essential act-ivities remain, nevertheless,
rather similar. For reasons explained later this limitation did not prejudice the final evaluation. . . .
In'Britain the likely effects therefore are:
a Effects on the Coastguard service (as coordinator of SAR activity),
.b Effect on Royal National Lifeboat Institution (RNLI) activity.
(lifeboats),
c Effects on MOD activity (helicopters/other aircraft).
3.8 Delays' (and .acceleratlon)to Ships an4 Cargoes
A further category of effect likely to stem from a traffic separation scheme (TSS) is a change in the efficiency and direction of traffic movements, that. is ships' movements are speeded or slowed., either in the. obvious sense, or through following a different route. ThiS has repercussions for ships'
costs.
The effect on distance travelled and voyage time may be either positive or negative and will be felt in three distinct areas.
a Voyage time for ships,
b Voyage distance for ships,
c ' Voyage times for cargoes (largely working capital).
As noted earlier1 however, this Report concerns the costs and benefits,
of
a.VTS: not those of a TSS as such.
Whilst the effects just described may all be relevant to a TSS it isfar from clear how a VTS, whether superimposed on a TSS or not, could typically act so as to delay ships: though it is fairly clear how it might serve to promote the more efficient movement of traffic.
Effect on Other Sea Users
As with the environmentaleffects discussed above, this last category poses
difffculties even at the specification stage.. It should also be bprne in mind that such impacts will become progressively more acute as the incidence of VTS Increases, because any YTS is to a degree exclusive and this imposes a limitation on the optimum use of other marine resources. Indeed, ft may prevent the use of the given sea area by any other than comercial traffic. This factor is certainly significant in congested waters such as the North Sea, where many activities already compete for maritime space1' dditions to existing traffic separation schemes may similarly have an impact out of proportion to the impact of the scheme considered in isolation eg by redUcing sea-use options within a wide area. Cockcroft, for instance, refers to the increased effects that continuous routing would have on maritime uses such as sailing, fishing, offshore explotation and naval exercises. Such activities would be either prohibited or severely retricted over an extensive area'2. As the sea represents one of the few resources remaining unappropriated (as is most of the useful land surface) this effect must be. expected to
increase.
Uses likely to be effected Include:
Fishing,
b Exploitation of marine sand and gravel,
C Extraction of oil and gas,
d Dredging, eg of approach channels for ships,
e Waste disposal,
f Pipeline/cable laying and maintenance,
g Marine recreation, eg yachting, swimming
The next chapters proceed therefore to examine the quantification and
evaluation of these component parts of the cost-benefit analysis. Reference
is made throughout- to the organisat.ions contacted and the chosen approach is
clearly set in the context of the available information and the methodologies used in alliedfields.
Notes and References
Goss R 0. & Hallf:day J E 1985: 'Cost Benefit Analysis of Shore Support.
for Safety in European Waters' In Safety In European Waters
Proceedings of a seminar Organised by the Royal Institute of Navigation and the Nautical Institute, l2.1284. London: The Nautical
Institute.
Goss R 0 & Hailiday J E 1985: 'The Costs and Benefits of Navigational -Safety',Paper presented to the l9thAñnual Conference of Law of
the Sea Institute in The UN Cnvention on the Law of e Sea Impact
aM Implementation, Cardiff, 24-27 July, 1985.
3 COST 301/PMS/034 1982: op cit.
4 ibid.
5 ibid.
See eg Cooper C 1981: Economic Evivation and the Environment, Hodder
& Stough.ton.
See egRanwell D S 1968: Extent of damage to coastal habitats due
to
the Torrey Canyon incident. In Carthy J 0 & Arthur 0 R (Eds):
The
Biological Effects of OH Pollution on Littoral Comunities, A
symposium held at the Orielton Field Centre 17-19 February, Field Studies Council.
Lecture).
Journ. Hon. Co. M, Mar., XXI (81), pplO3-127 and Coékcroft A
N 1981:
'Routelng in the English Channel', Journal of Navigation, 36
(2), pp392-413.
9
De Bievre A 1983:
'Navigational Safety In European Waters', Journal of
Navigation, 36 (2), pp169-182.
10
See e.g Lacey M 1985:
Implications for the Comercial Salv
- Paperpresented to the Greenwich Forum Conference:
The yfrridEniergency,
RUSI, London.
11
See eg Smith H 0 & Lalwani C S 1984:
The North Sea:
Sea UseManagement and Planning.
Cardiff:
Centre for Marine Law and Policy,
UWIST.
CHAPTER 4 DEATHS AND INJURIES RESULTING FROM SHIPS.' CASUALIT.IES
4.1 The Quantification of Life and Injury
The last chapter suggested that human accidents fall into two basic classes, people killed and people injured. In respect of quantification the first category is relatively straightforward, death being unequivocal and
-generally -officially counted. The first requirement th.eréfot is for
statistics to be obtained fOr the number of deaths In situations which could have been avoided by VTS,that is essentially those caused by collisions and
strandings'.
The collection of data on marine, traffic casualties in the COST 301 area formed the basis of a contract with the Maritime Research Institute of the Netherlands (MARIN). In this context data obtained from. Lloyds Shipping
Information Service was obtained for a five year period, 1978-1982, and the findings sumarised in a final report by MARIN presented in May .19852.
This data collection exercise represented one of the substantial inputs to the cost-benefit analysis, theoretically enabling us to obtain a quantified picture of the present navigational activities in European waters. Any
limitations to this database posed therefore serious problems for all the areas of our analysis which relate to maritime casualties, that is ship and cargo damage, consequent pollution and, in the context of this chapter, loSs of life and injury.
The first two limitations relate to lives lost. Firstly the consequence of casualties with regard to human life are often reported as unknown3, that IS
the data Is not comprehensive in terms of the human accident profile. Secondly if death, as a function of collision and grounding, is to be
quantified, it is necessary to define the extent of the period following the incident in which it is legitimate to define the death as a consequence of
the incident. The MARIN data base iS capable only of identifying those
killed or missing on board ships as a consequence of the casualty suffered by
the ship. All but the relatively immediate occurrence of death after an
accident is therefore removed from the analysis; seamen who are Anjured, taken ashore and who subsequently die in hospital are excluded. I addition,
therefore, to the partial availability of statistics on lives lost tnose incidences where it is reported represent only a truncated profile of the actual lives lost. In two respects therefore this input into the
cost-benefit analysis is an underestimate.
The ftrther constraint indicated by MARDI, that such casUalties may not necessarily be a consequence of the initial incident alone but may be the
result of the subsequent chain of events" does not result in any compensating overestirnate of attributable death, for the causal link, relating the death tothe initial collision or grounding,remainS strong.
We made attempts to remedy this deficiency. The problem in a British context at least is however well illustrated by the casualty and accident statistics produced by the UK Department of Trade5 and relating to vessels registered in
the United Kingdom.
Apart from the two obvious difficulties, thatthe statistics do not relate solely to European waters and that they are restricted to the registry of One country, the reports highlight another problem. "Complete statistics are not available as deaths of seamen occurring ashore, after discharge in the United Kingdom - even though due to accidents sustained at sea - are not reported to
the Registrar General of Shipping and Seamen"6. It appears that such
statistics have, in fact, never been compiled, though the point has repeatedly been noted.
Nevertheless, the reports produced by the Department of Trade do clarify one
impOrtant and relevant point. This is the fact that deaths caused by
collisions and groundings are very infrequent. In the five years, 1978-82, with which this project is cOncerned, for instance, 119 out qf 674 deaths
(17.7%) amongst crew were due to casualties to vessels. A detailed breakdown of these statistics shows, however, that none of these deaths we,. due to
stranding and only 9 due to collisions.. Hence only 9 out of 674 deaths
(1.3%) were due to the cause with which we are concerned, collisions and
strandings7. Furthermore of the 9 collisions 5 occurred in river or harbour
(the remaining 4, more pertinently to this project, occurring at sea)8.
The majority of deaths due to casualties to vessels in this period were in fact due to missing vessels. Fbunderings, a category, which as noted above, may be influenced by VTS, accounted for 29 deaths, a more substantial
proportion than strandings and collisions, and explosions and fires were also
important.
The majority of deaths (82.3%) relate instead either to disease or to
accidents aboard shp, in the engine room for example, or on deck unconnected with collisions or groundings.
This has two imediate implications. On the one hand the numbers involved are small; and the numbers dying at a later date and omitted from the statistics are likely therefore also to be small. On the other hand with such small figures, the omission of even a few deaths in this manner would give a wide margin of error.
301 area, in order that extrapolation could be made from the largest number
of records. These are sunnnarised in table 4.1 below which indicates that, based on the historical recOrd, 0.44 fatalities occur/collision and
0.07/stranding.
Table 4.1 Numbers Killed in Collisions and Strandings: European Waters (1978-1982)
wo anomalies became evident once this data was analysed and both are pertinent not only to this chapter but to the remainder of this report,.
The first is thatcasualty data supplied by MARIN includes entries relating to vessels falling in COST 301 categories 14-18, that is fishing vessels, naval vessels, tug supply vessels, 'unknown' vessels and
hydrofoils/hovercraft, that is vessels outside the original 'commercial'
definition. In total 9.2% of the casualty data on collisions refers to non-commercial vessels in the strict sense, similarly 8.6% of the data on
strandings. Type of casualty Number of casualties Number 'killed' Number 'missing' Total killed Number killed per casualty Collision 446 68 128 196 0.44 Stranding 649 28 16 44 0.07
:-In contrast, however, traffic flow data supplied to our work, another important input Into the fitial cOst-benefit analysis, includes only hydrofoils and hovercraft from these additional categories. In effect
therefore casualties. ocurringon the other vessel types are being
distributed over the remaining COST 301 categories as represéntöd in the traffic flow.
This is obviously unfortunate. BecaUse, however, of the detail of breakdown supplied, it is nOt poSs1ble to remove the influence of these vessels from all casualty data. Moreover, the same casualty data has been used as an
input into the PAl methodology and the influence of this 'other' traffic is
therefore already incorporated into aflOther area Of our .work.
A second anomaly, apparent only when this report was in the final stages of preparation is that the PAl rnethodO1oy is based on incidents occurring in the open sea and restricted.waters. In contrast the casualties data (and presumably traffic flow data) supplied to our work relates to open sea areas
only. Given that the expected collision and stranding rates are specific to
our 3 study areas it is not clear whether this alo has introduced a
potential source of errOr Into our work or, if so, how great an effect this
has.
Such anomalies seem to be derived largely from the original nature of the whole COST 301 study, in which independently-produced data have had to be
used in a variety of flew contexts. Given double the resources and
twice the
time tO complete this study we could, of course, have gone back and sought their resolution: as it was we could only accept the resulting (small) margin of error.
It Is generally acknowledged that it is yet more difficult to collec.t reliable statistics on the incidence of injuries, for published reporting
tec'hniques in this area are generally less comprehensive9. The problem of adequate data Is further compounded because of the range In severity of injuries which can be expected following such accidents, a range which again is not found on record.
The problem emanating from the casualty base as prepared by MARIN Is however not one of degree but rather absolute: there is no mention in their final report of any data on Injuries, and it has subsequently been confirmed that no such data is available from this source for input into our work10.
One study, acknowledging the lack of knowledge as to the distribution of the severity of injuries, resorted instead to the use of an average,
extrapolating the severity distribution of the accident and hence the average cost, from a study of road accidents in Finland11.
The assumption that the severity distribution of injuries In maritime
casualties Is the same as in road accidents, or that at least their means are the same, seems however questionable. Speed, frequently a contributory
factor on roads, is very different for the two modes, as is unit size.
Impact with a hard object inside a car is a frequent cause of Injury in road accidents but drowning is very rarely a component, neither is hypothermia.
In marine accidents the cause of injury is therefore likely to be qUite different and therefore so is the consequence. Whilst an average may be the best we can attemptwhen evaluating the cOst of InjurIes it should at least therefore be based on an appropriate range of injuries.
Other studies are similarly not particularly useful In indicating how this shortfall can be overcome. The Canadian Coast Guard Study for instance uses the Coast Guard's own casualty investigation database which is specific to Canadian waters12. It does however tend to confirm the view that collisions,
groundings and strafldings rarely cause serious injuries and deaths. The 534
casualties recorded by the Coast Guard as occurring between 1975 and 1981 thus resulted in only one fatality. The cutoff point in time Is, however, not made explicit that is, it is not clear whether this includes only those
deaths occurring immediately on board ship, or later., consequential deaths eg in hospital; and once indirect effects such asfire and explosions consequent in collision are included the potential for a fatality or injury increases,
Giziakis made use of the same essential database as the. MARIN study., that is
Lloyds Casualty Returns, this time for the four year period 1975497813. Again, therefore, these components are likely to be undercounted and, as the published data relates to 31 zones worldwide and to total loss, situations only, his suggestion that 0,42 people are killed or missing for every 1000 grt lost as a result of collision and 0.05 are killed or missing for every 1000 grt lost as a result of grounding is not useful'. Again accidents
resulting in injuries only were omitted..
There is as noted above, now a requirement for non-fatal injuries to merchant
seamen on UK-registered ships to be reported to the. Department of Transport:
but only since Qctober 1982, that is after the endof our study period. Statistics relating to fishing vessels, admittedly very different craft, do however refer to non-fatal injuries over a longer time span. These suggest that the conclusions reached by the Canadian study are, in general, valid and that non-fatal injuries are again likely.to be only a very minor component in the cost-benefit analysis. In 1977 and 1978 fOr instance there were 1182 non-fatal injuries on UK registered trawlers) but none of thee arose from
casualties to vessels. The vast majority were accidents on deck and a large percentage of the remainder occurred in engine rooms and stokeholds.
Various atttempts were made to collect suppleipentary information in this
area. Since 1982 for instance the General Council for British Shipping
(GCBS) have kept occupational safety statistics which are now reported by rule to the Department of Transport and. submitted to the National Maritime
Board15. These cover all kinds of accidents to people.. Whilst tiowever,
injuries sustained as a result of casualties to vessels should be included they are not attributed to such causes as collisions and groundings and are thus not capable of separate Identification.
Approach was also made concerning a representative cost for such injuries. Again, however, it was found that figures received from member companies in
th costs Of occupational accidents were in the form of annual totals.. As we were told: "These are incomplete and used simply to produce an aggregate figure to. illustrate the considerable cost to the industry of such
accidents".
Indeed they do, but they are not useful In this context. Moreover, while these figures would include such elements as medical costs, costs of
repatriatiOn1 time lost etc, it is extremely unlikely that they would include
costs related to collisions, grounding etc.. In any event. it would be
impossible to separate any Such data from the broad information provided, if
included16. . .
Accident statistics, as noted above, have also been recorded, for recent years, by the Occupational Health and Safety Division of the Department of
Transport. These.refer to all Injuries, which result in more than three
Documentation. provided shoWed that in 1983 and 1984 respectively 1,200 and 1,100 such injuries occurred. These are broken down by type, 9ne category amongst many being Injuries resulting from collisions, foundering and
strafldiflgs, that is the area of injury With which this project is concerned.
In 1983, however, only ? of the 1,224 injuries fell into this category, in 1983 4 of the 1,090, that is 0.2% and 0.4% respectively. The vast majority of accidents are related to slips and falls the main injury being bad
bruising and cuts and lacerations, followed by fractures. Again therefore
evidence suggests this category is, of minor importance.
-Enquirieswere also made of the National Union of Marine, Aviation and Shipping Transport Officers (NUMAST), the National Union of Seamen and the British Petroleum Company (BP), concerning both the numerical incidence of injuries and deaths due to collisions and groundings and any data relating to
their cost.
BP searched their records for. th period 1978-82 with. regard to injuries and
fatalities caused by collisions and groundings of ships in their.fiéet. Although some of their vessels were involved in this type of incident there were, however, 'no injuries or deaths attributed to them'18. This therefore reinforces the idea of a cost cateory of relatively minor impact.
This was further reinforced by the response received from the National Union
of Seamen who pointed out that 'shipping casualties are so infrequent that we
do not keep separate records about them'., Furthermore they are dependent on
members reporting accidents and 'this does not necessarily happen'19.
Whilst ideally, therefore we would have incorporated a factor to reflect the numbers of injuries in our model cost benefit analysis., this proved to be impossible; although the evidence strongly suggests that this class of injury would, In any case, have been virtually negligible. A brief cOnsfderation of the possible evaluation of injuries is, nevertheless., given in the following
sections.
Meanwhile, and despite the curious contrast. between the dearth of statistical information described above and the existence of an international convention on the Safety of Life at Sea (SOLAS), we are satisfied that our treatment of this subject leads to no significant margin of error in our overall results.
4.2 The Evaluation of Life and Injury
E;isting methodologies are far more useful when ft comes to considering the evaluation of these two components. Three distinct elements can be
identiffed, the direct costs or the incident, including for instance medical treatment and administration, the loss of output and the value attached to
the life or limb. Different conceptual problems are associated nevertheless with each.
42!1Pirect Costs
Most of the evaluation of life, and limb for cost-benefit analysis has tended to follow Dawson's work on road accidents21 and certain elements in that work, the direct costs, remain undisputed.
These include:
Administrative costs,
..b. Medical treatment,
c Police and ambulance costs.
Just however as Kostilainen's work was criticised above for. extrapolating from land transport to the maritime mode, so these elements too need
transformation into maritime equivalents. Police costs for instance are not
apprOpriate. Mibulances, if used at all, will represent only the final leg of a longer journey whjch may have Involved the RNLI, helicopters, other ships, own ship diversion to port or treatment on board using either existing staff or supplemented by help from land, again transported to the. ship by
some method.
Ideally a scenario needs to be developed which reflects the causal chain of events consequent on human accidents and takes into account such
variables as severity of injury, size/type of ship (and hence expected on board facilities), and distance from shore. The cost of medical treatment will similarly vary depending on whether the casualty is treated in a
hospital or on board ship.
The only sensible course therefore would be to follow Dawson's approach and cost a sample of cases In detail. This has not been possible in this study.
It remains, however, a legitimate requirement if cOst-benefit analysis is to be applied more widely in a maritime context, whether in respect of VTS or
4.2.2 Loss of Earnings
With injuries there Is a cost equivalent to the loss of earnings whilst
recovering. It is recognised that this factor should include the social cost of employment, that Is elements such as the. employer's expenditure on
national Insurance and pension funds, as well as the gross pay, so as t
represent, the, marginal product of labour.
With death this samefactOr has been used as a proxy for the v3lue of life per se in the 'gross output' approaches. It will therefore be cieIt with more extensively In the next section which considers the values that may be attached to life and limb
4.2.3 The Value of Life an4 Limb
Two poitits need making at the outset, before the available methodologies are
considred. Firstly, controversy remains as to whether, and how, the loss of
life and limb can be valued. Some would question whether we can place any
values on such matters, even though such a view is equivalent to Implying that the proper value is infinity. We may, however, Observe that neither public authorities nor people in their individual capacities act in this way. Moreover, "however such decisions are taken, some form of implicit valuation
is literally ünavóldable"22. An example of this appeared in an earlier study which derived the value of seamen's lives implicit in decisions made by what
is nOw the Marine Directorate of the Department of Transport. This wás found to be £34,000 in 1972 and may be compared with the value of £25,400
calculated in that paper, using the first of the two techniques discussed
below23. Methods available and used by various researchers
include the net
and gross output approaches, the life insurance approach, valuatio?S implicit in public sector decisiOnS by regulatoryauthOrities and in safety' at work,., court decisions and the 'willingness, to pay' approach.