. 7
( 18)


102 Department for Constitutional A¬airs, ˜Setting the Discount Rate: Lord Chancellor™s Reasons™
(27 July 2001).
103 Warriner v. Warriner [2002] 1 WLR 1703. However, it may be that, assuming investment in ILGS
and allowing for tax, even 2.5% may be too high.
104 R. Lewis et al., ˜Court Awards of Damages for Loss of Future Earnings: An Empirical Study and
an Alternative Method of Calculation™ (2002) J. of Law and Society 406. See also R. Lewis et al.,
˜Loss of Earnings Following Personal Injury: Do the Courts Adequately Compensate Injured
Parties?™ (2003) 113 The Economic Journal F568.
160 Chapter 6

found that US awards are, on the whole, signi¬cantly greater in relative terms than
English awards. Two main explanations are given. One is that English courts
underestimate the extent to which people™s earnings tend to increase over the
course of their career. The other is that English courts underestimate the negative
impact of disability on people™s employment prospects and the amount of time
they are likely to be unemployed in the course of their working life. It is worth
noting that such problems will arise under the periodical payments regime dis-
cussed above as well as under the lump-sum (and associated structured-settle-
ment) system. If, for instance, a court, in making a periodical-payment order in
respect of future loss of earnings, underestimates the amount of time the recipi-
ent is likely to be unemployed, there will be no way of ¬xing this later on. It is true
that periodical-payment orders may be varied in certain circumstances; but the
only ground of variation is a signi¬cant change in the recipient™s physical or
mental condition.
There may also be problems with the adequacy of awards for long-term future
medical expenses and the cost of care. A striking example was given in 1974 by the
Australian Committee of Inquiry into the National Rehabilitation and Compen-
sation Scheme.105 In Thurston v. Todd106 a young woman of 15 su¬ered very severe
injuries in a motor accident, as a result of which she was rendered a quadriplegic.
The accident occurred in 1963. In 1965 she was awarded damages just short of
£120,000, which was an exceptionally high award at that time. Over £50,000 of this
sum was intended to cover the cost of future nursing services and medical expenses,
on the assumption of a weekly cost of some £70. By 1973 the actual cost of these
services had risen so much that the income from the entire award of damages
(which had all been prudently invested) was inadequate to pay for the nursing
expenses alone “ despite the mother™s unpaid services for some 7 hours each day.
Nursing costs alone had nearly doubled in the 9 years since the damages were
awarded. Part of the explanation of such outcomes (in Britain, at least) may be that
since the end of the Second World War the rate of increase of average earnings has
been much greater than the rate of (price) in¬‚ation. In other words, the standard
of living of those in work has, in general, improved. However, as the research dis-
cussed in the previous paragraph suggests, English courts may not make su¬cient
allowance for improvements in the standard of living; or, putting it another way,
for increases in earnings over and above increases in the cost of living. To the extent,
therefore, that the income from a damages award has to be used to buy services
(usually nursing and medical services), it may well prove inadequate if the cost of
these services increases more than the rate of in¬‚ation as re¬‚ected in market inter-
est rates. This problem is not solved by assuming investment in ILGS because they
are linked to increases in the cost of living (as measured by the retail price index),

105 Australian Committee Report, paras. 149“50.
106 (1966“7) 84 WN (NSW) (Pt 1) 231.
Damages for personal injury and death 161

not to increases in wages. Furthermore, it has been expressly held that it is not per-
missible for a court to increase the multiplicand (6.2.2) to make allowance for
expected future increases in the cost of care over and above increases in the retail
price index.107
Another reason why damages awards may turn out to be inadequate is that
although the law assumes that the investment of the lump sum will be risk-free and
index-linked, the recipient (as we have already noted) is not required to invest in this
way but may do what they want with their damages, even if that be investing them
in a highly risky way or even squandering them. In fact, research suggests that most
recipients of large damages awards typically follow their own self-interest and use
the money wisely.108 Even so, there is a signi¬cant public interest that people who
su¬er long-term disabilities as a result of torts be adequately provided for. In this
light, it would surely not be an undue interference with their freedom for the law to
lay down guidelines about the investment and management of large lump-sum
damages awards designed to provide support and care for the future.
Certainly the fact that lump-sum awards may in some cases turn out to be inad-
equate to produce as much income as has been lost should not lead us to conclude
that it would be right to raise the level of such awards even more. In the ¬rst place,
in cases where the injured person has su¬ered large earning losses, substantial
awards are also made for pain and su¬ering and other intangible losses mentioned
below. It is possible to take the view that these are as irrational or excessive as awards
for lost earnings are inadequate, and that the one therefore helps to balance the
other, although the House of Lords has not shown itself sympathetic to such an
approach. Nor does the argument apply to dependency actions where (apart from
damages for bereavement) no award for pain and su¬ering and such like will be
available to augment the award for lost ¬nancial support; and it is possibly in such
cases that awards appear most inadequate.109 But even here, there are often coun-
tervailing considerations “ for example, the fact that life insurance and other
bene¬ts (such as social security) received by the dependants as a result of the death
are ignored in the assessment of damages.
Secondly, there is a great deal of ˜double compensation™. We shall look into the full
extent of this problem later,110 but here it must be noted that many losses are com-
pensated in full or in part more than once. Fourthly, the question of priorities, which
we have stressed so often, must not be forgotten. If more money is to be pumped into
the tort system, the e¬ect will be to increase the comparatively generous provision
already made for victims of fault-caused injuries who are fortunate enough to collect
tort damages while doing nothing for the great mass of the disabled population.

107 Cooke v. United Bristol Healthcare NHS Trust [2004] 1 WLR 251.
108 See 6.1.4.
109 A.F. Conard and others, Automobile Accident Costs and Payments (Ann Arbor, 1964), 179; A.M.
Linden, Report of the Osgoode Hall Study on Compensation for Victims of Automobile Accidents
(Toronto, 1965) (Osgoode Hall Study), ch. 4, 25“6.
110 Ch. 15.
162 Chapter 6

6.5 Intangible losses
6.5.1 Assessing intangible losses
So far we have been considering damages for ¬nancial loss. In personal injury
actions the law also awards damages for certain intangible ˜losses™. Damages for pain,
su¬ering, discomfort, humiliation, indignity and embarrassment are awarded
under the head of ˜pain and su¬ering™. Damages may also be awarded for loss of
the ability to do things and to enjoy life in a way possible before the accident; these
are usually referred to as damages for ˜loss of amenities™ or ˜loss of faculty™. These two
types of injury may merge as, for instance, where an injured person has su¬ered a
loss of sexual potency, or is so badly injured as to impair the prospect of marriage.111
The two kinds of damages may both be recoverable since loss of faculty may be
accompanied by pain and su¬ering; but it is possible to have loss of faculty without
any pain or mental distress at all, as in the case of someone who is rendered perma-
nently unconscious or is incapable of appreciating their situation. It is also possible
to have pain and su¬ering with no physical or mental disability or loss of faculty.
But in most serious cases the two go together. Loss of limbs, paralysis, blindness or
deafness, and so on, are unlikely to be in¬‚icted without considerable pain and
su¬ering; and signi¬cant pain and su¬ering is likely to be accompanied by some loss
of faculty. Damages for pain, su¬ering and loss of amenities are usually referred to
collectively as damages for ˜non-pecuniary loss™.
In fatal cases, until 1981, a small ¬xed amount could be recovered by the estate
of the deceased person to compensate for their ˜loss of expectation of life™. Now, an
amount of £10,000 is recoverable by the claimants in a claim under the Fatal
Accidents Act 1976 to compensate them collectively for their bereavement. As we
have seen, they can also recover for loss of support in money and money™s worth
formerly provided by the deceased. But nothing can be recovered for the loss of the
deceased™s life as such.
The calculation of damages for non-pecuniary loss has an air of unreality about
it. Something that cannot be measured in money is ˜lost™, and the compensation
principle requires some monetary value to be placed on it. There appears to be no
objective way of working out any relationship between the value of money “ what
it will buy “ and damages awarded for pain, su¬ering and loss of amenities. All such
damages awards could be multiplied or divided by two overnight and they would
be just as defensible or indefensible as they are today.112
It is not only lawyers who are concerned with putting monetary values on intan-
gibles. There is a large economic literature dealing with the valuation of ˜life™.
Economists normally value things by looking for a ˜market price™; but there is, of
course, no market in human lives. So they need to ¬nd alternative indications of

111 Moriarty v. McCarthy [1978] 1 WLR 155.
112 For an analysis of the conceptual basis of such awards see A.I. Ogus,˜Damages for Lost Amenities:
For a Foot, a Feeling or a Function?™ (1972) 35 Modern LR 1.
Damages for personal injury and death 163

how much value people put on their lives as such “ the ˜hedonic™ value of life, we
might say.113 The most common method for doing this to gather evidence about
how much people are willing to pay to reduce the risk of death from various causes.
So, for example, economists look at wage di¬erentials between more and less
risky jobs, and at price di¬erentials between more and less safe products. One
problem with this approach is that the amount a person is willing to pay will
depend to some extent on how much money they have. This is because what econ-
omists call ˜the marginal utility of money™ is greater for a poor person than for a
rich person: other things being equal, a rich person is likely to be willing to pay
more for any particular commodity than a poor person simply because they have
more money to spend. We would not want to conclude from this fact that the life
of a rich person is worth more than that of a poor person. It is sometimes suggested
that this problem can be solved by asking not how much a person would be willing
to pay to avoid a risk of death, but how much they would be willing to accept to
incur a risk of death. However, because of the marginal utility of money, a poor
person is just as likely to be more prepared to accept less than a rich person as to
pay less. However, the impact of ability to pay is not quite as serious as it might at
¬rst appear. This is because willingness-to-pay research typically deals with very
small risks. The question is not how much a person would be willing to pay to avoid
certain (or even a high probability of) death, but how much they would be willing
to pay to avoid a small risk of death. So, for example, suppose that 10,000 people
are asked how much they would be willing to pay to avoid a one in 10,000 risk of
death. If each were willing to pay £300, this would suggest that each valued their life
at £3 million. Many more people could a¬ord £300 than £3,000,000, and so the
signi¬cance of wealth is reduced “ but not, of course, eliminated. Moreover, there
are many other problems associated with research of this type arising, for instance,
from the fact that people vary in their attitude to risk “ some people enjoy risk-
taking whereas others are very cautious.114
This willingness-to-pay approach to valuing life is most commonly used in con-
nection with cost-bene¬t analyses designed to support decisions about how much
to spend, for instance, on road safety measures115 or on reducing pollution hazards
or workplace risks. The other use to which this method might be put is the assess-
ment of compensation. However, we have noted that English law does not provide
compensation for loss of life as such. To the extent that a legal value is put on life,
the approach used is what economists call the ˜human capital™ method. This means

113 Or, in other words, the value of life™s pleasures.
114 For a quite accessible overview of this sort of research see W.K. Viscusi,˜The Value of Life in Legal
Contexts: Survey and Critique™ (2000) 2 American Law and Economics Review 195.
115 See e.g. R. Elvik, ˜Cost-Bene¬t Analysis of Road Safety Measures: Applicability and Controversies™
(2001) 33 Accident Analysis and Prevention 9. In Britain, the willingness-to-pay method is the
starting point of assessing the ˜costs™ of road accidents, but other things are also taken into
account, such as police costs and lost output: Department for Transport, Highways Economics
Note No. 1, 2003: Valuation of the Bene¬ts of Prevention of Road Accidents and Casualties. See also
M.W. Jones-Lee, ˜The Value of Transport Safety™ (1990) 6 Oxford Review of Economic Policy 39.
164 Chapter 6

that in a fatal accident claim, the deceased™s life is valued in terms of the loss
in¬‚icted by the death on the deceased™s dependants. But the law does not, for
instance, take into account loss, resulting from the death, to society generally or to
individuals who do not fall within the class of eligible claimants. Some economists
think that the willingness-to-pay method of valuing life is appropriate for deciding
how much to spend on preventing death, but not for the compensatory purposes
of tort law.116 Other writers, who argue that the function of tort law is not (only) to
compensate for harm caused but also to prevent the occurrence of harm, argue that
the willingness-to-pay approach has a role to play in tort law.117
At all events, this type of research is of relatively little use in assessing damages
for non-pecuniary loss because what it values are whole lives, whereas it is (only)
for non-fatal injuries that the law provides compensation for intangible loss. Just
as there is no market in human lives, so there is no market in pain or lost limbs.118
In principle, there is no reason why the willingness-to-pay approach should not
be applied to non-fatal injuries; but in practice, this has not been done, and
the di¬culties of conducting the necessary research to generate willingness-to-
pay values, for each of the great variety of injuries (and combinations of injuries)
that can be caused by tortious behaviour and attract awards of damages for non-
pecuniary loss, would be very considerable. The Department for Transport publishes
tables, estimating the costs of road accidents, that cover not only fatal accidents
but also accidents that result in ˜serious™ and ˜slight™ non-fatal injuries. But this classi-
¬cation is, of course, far too crude for compensatory purposes (for which it was
not designed). For what it is worth, however, we might note that the 2003
¬gure for the ˜human costs™119 of a serious injury was £119,550 “ only about half the
maximum ¬gure that courts award for non-pecuniary loss in the most serious cases
of personal injury.
So we have not made much progress in ¬nding a method for calculating
damages for non-pecuniary loss. There are, however, two noteworthy and related
facts about awards for non-pecuniary loss that may provide a clue: ¬rst, that despite
the marginal utility of money, the amount awarded takes no account of the wealth
of the claimant; and, secondly, that levels of damages for non-pecuniary loss tend
to be roughly related to social prosperity. So, they tend to increase as society
becomes wealthier; and they tend to be higher in wealthier countries than in poorer
ones. These facts suggest that the process of calculating such awards is not entirely
lacking in external reference. It is fair to assume that widespread agreement could
be achieved on the extreme outer limits of what would be regarded as ˜reasonable™
compensation for the intangible aspects of personal injury. For example, we might

116 E.g. Viscusi, ˜The value of Life in Legal Contexts™.
117 E.A. Posner and C.R. Sunstein, ˜Dollars and Death™ (2005) 72 U. of Chicago LR 537.
118 Professional boxing might be thought to get close to a market in injury.
119 Representing pain, grief and su¬ering to the injured person, relatives and friends: Department
for Transport, Highways Economics Note No. 1, 2003, para. 5. Damages for non-pecuniary loss in
personal injury actions take no account of the e¬ects of the injury on relatives and friends.
Damages for personal injury and death 165

con¬dently speculate that few people would think £100 too much or £1 million too
little for the loss of a hand. Perhaps an upper limit of £3 million on compensation
for non-pecuniary loss in cases involving the most serious injuries (quadriplegia
and the like) would be rejected as too high, partly because such a sum is so far
beyond the sort of capital wealth that most people could ever hope to acquire.120
Equally, £1,000 would be rejected as far too low because to very many people, an
extra £1,000 in wealth would be of little moment. Thus it seems that although the
selection of particular sums within such wide limits will be a matter of judgment,
the limits themselves have some external reference point in that they bear some
relationship to the sort of sums that people in general may expect to enjoy as per-
sonal wealth.
In 2000 the Court of Appeal laid down guidelines for the assessment of damages
for non-pecuniary loss in the process of deciding appeals in eight separate cases.121
They were that:
• damages for non-pecuniary loss in cases involving the most serious injuries should be
increased by one-third (from their level at that time of around £150,000 to around
• there should be no increase in cases where the appropriate award for non-pecuniary
loss would be less than £10,000; and that there should be ˜tapered™ increases in cases
falling between that threshold and the most serious;
• damages for non-pecuniary loss should be increased regularly in line with increases in
the retail price index; and
• new guidelines should be not issued unless there was ˜real reason to think that once more
the level of awards is signi¬cantly out of line with the standards we have identi¬ed™.122
Unfortunately, the only standard identi¬ed by the court for judging the adequacy
of awards for non-pecuniary loss was that they should be ˜fair, reasonable and just™.
In 1996 the Law Commission had recommended that awards of damages for non-
pecuniary loss above £3,000 should be increased by between 50 and 100 per cent,
and awards of between £2,000 and £3,000 by smaller percentages;123 but the Court
of Appeal criticized this recommendation as being too heavily based on faulty
empirical research undertaken for the Commission. In recommending a signi¬cant
increase in the most serious cases, the court was in¬‚uenced by the fact that as a
result of advances in medical technology, the life-expectancy of many seriously
injured people has signi¬cantly increased in recent years: the longer an injured
person lives, the greater the pain, su¬ering and loss of amenities they are likely to
su¬er. On the other hand, it argued that losses that would once have been treated
as non-pecuniary “ such as the ability to go on an ordinary holiday or to live in an

120 It has recently been suggested that at least £3 million is now necessary to sustain the life-style
associated with ˜millionaire™ status.
121 Heil v. Rankin [2001] QB 272.
122 Heil v. Rankin [2001] QB 272 at [99].
123 Law Com. No. 257, Damages for Personal Injury: Non-Pecuniary Loss (1999).
166 Chapter 6

ordinary home “ are now treated as pecuniary losses for which speci¬c amounts
can be awarded representing, for instance, the cost of a specially arranged holiday
or of home modi¬cations; and this counted against increases on the scale recom-
mended by the Law Commission. Relevant though such considerations might be,
they do not explain why an increase of 33% is more ˜fair™ than one of 50% or even
100%. Some people might think that even £300,000 would be inadequate com-
pensation for non-pecuniary losses associated with catastrophic injuries.
However, the Court of Appeal was also in¬‚uenced by the likely impact of sudden
large percentage increases in awards for non-pecuniary loss on the total amount
spent on tort compensation. For example, the court was told that the cost of a 100%
increase to the NHS would be £133 million a year in additional compensation
payments, and that it would generate additional insurance pay-outs of around
£1 billion a year. According to traditional understandings of tort law, such ˜social
facts™ are irrelevant to deciding what is fair and just as between individual ˜doers and
su¬erers of harm™. But the court rightly refused to ignore the fact that in a world of
scarcity, choices must inevitably be made between competing calls on society™s
limited resources. This is one reason why it recommended a threshold of £10,000
before there should be any increase: the large majority of tort claims fall beneath
the threshold, and adopting the Law Commission™s recommendations for increases
in all cases above a £2,000 threshold would have added much more to the total tort
compensation bill.
In this respect, it is important to note a di¬erence between judicial law-making
(such as laying down ˜guidelines™ for the assessment of damages)124 and parliamen-
tary law-making. If the Law Commission™s recommendations had been acted upon,
this would probably have been done by an Act of Parliament and by regulations made
under it. Most likely, such legislation would have been purely prospective in e¬ect “
in other words, it would have applied only to claims made after the date the legisla-
tion became operative. By contrast, judicial rule-making operates retrospectively “
that is, it a¬ects not only the claim in the case before the court, but also, in practice,
all unresolved claims that have already been made. Thus the immediate ¬nancial
impact of (retrospective) judicial increases in levels of damages awards is much
greater than that (prospective) legislative increases. This is another reason why the
Court of Appeal was unwilling to adopt the Law Commission™s recommendations.
Indeed, the defendants in the case argued that it was inappropriate for the court (as
opposed to Parliament) to increase the level of damages awards in the way it did.

6.5.2 The tariff system
In the case we have been discussing, the Court of Appeal increased the existing levels
of damages for non-pecuniary loss. How were those levels arrived at? Until 1934, in
cases that went to trial, damages for non-pecuniary loss were assessed by a jury. In

124 In strict theory, the guidelines are probably not binding rules of law; but in e¬ect, they are.
Damages for personal injury and death 167

the USA, juries still assess such damages. On this side of the Atlantic, however,
damages for personal injury are almost invariably assessed by judges.125 Given the
lack of objective relationship between pain, su¬ering and loss of amenities and
damages for non-pecuniary loss, assessment by a jury might seem a suitable way of
injecting into the legal process community views about how much compensation
particular injuries should attract. On the other hand, because juries were not told
about awards by other juries in comparable cases, there was no way of ensuring a
desirable degree of consistency in awards. Even if we cannot, in any objective sense,
say what a leg or an arm is worth, it should at least be the case that a leg today is
worth the same (in real terms) as a leg tomorrow; that an arm must be worth more
than a hand; a hand more than a ¬nger; two legs more than one; and so forth. Even
here, of course, there is great di¬culty. Is an arm worth more than a leg? Is it worse
to be totally blind than to lose both legs? Is a hand worth more than a foot? With
what can you compare the inability to bear a child? But still, making every allowance
for the element of arbitrariness in the whole process of compensating for disabil-
ities, it is possible to have some internal consistency in the process, and such con-
sistency would not be easily attained if the decision were left to a jury.
Consistency in awards for non-pecuniary loss is desirable not merely in the inter-
ests of justice, to achieve equal treatment of like cases. It is also important for the
smooth running of the tort system because most tort claims are settled by negotia-
tion out of court. Without some consistency in the level of awards, it would be very
di¬cult to predict the outcome of a case and hence to negotiate a settlement. A very
small decrease in the proportion of cases settled, and a corresponding increase in the
proportion of cases going to trial, could seriously overload the court system.
Judicial assessment greatly facilitated the development of a ˜tari¬™ system of cal-
culating damages for non-pecuniary loss. The group of judges who regularly decide
personal injury cases is quite small. Judges may be able to discuss awards with one
another, and judicial assessments that deviate too far from the current norm can be
corrected by the Court of Appeal. Under the tari¬ system, particular ranges of
awards are established for particular injuries and disabilities. For many years, there
was no formal mechanism for ¬xing these ranges. They simply emerged from
reported decisions in individual cases. Indeed, in 1973, the Law Commission came
to the conclusion that the ¬xing of damages for non-pecuniary loss was so arbitrary
that no principles could be recommended on which the courts should work. The
only question, according to the Law Commission, that needed to be settled was ˜who

125 In 1966 the Court of Appeal decided that jury trial would henceforth be permitted only in very
special cases: Ward v. James [1966] 1 QB 273. An example of such a case might be where the
injuries are of so unusual a kind that judicial experience would be of little use: Hodges v. Harland
& Wol¬ [1965] 1 WLR 523 (this is the last reported case of trial by jury in a personal injury
action); or where exemplary damages are claimed: H v. Ministry of Defence [1991] 2 QB 103. The
Law Commission™s view is that assessment of damages for non-pecuniary loss in personal injury
cases should never be left to a jury: Law Com. No. 257, Damages for Personal Injury: Non-
Pecuniary Loss (1999), paras. 4.1“4.5.
168 Chapter 6

ought to decide™.126 In 1992 the Judicial Studies Board published a quite detailed set
of guidelines, for the assessment of ˜general damages™127 in relation to a long list of
injuries, developed by a working party ˜to present a snap-shot of the general level
of . . . damages [for non-pecuniary loss] re¬‚ected by judicial decisions and settle-
ments in¬‚uenced by them™.128 The Guidelines are ˜not intended to promote any views
about what the level of damages ought to be™.129 On the other hand, the working
party gives more weight to decided cases than to reported settlements.130 The ¬gures
given (in the form of a lower and upper ¬gure for each injury) are merely guides: ˜it
is for the courts to set the level of damages and for this book to re¬‚ect them™.131 The
Guidelines are not su¬ciently detailed to be comprehensive, and they have no
authority as such. On the other hand, in his Foreword to the third edition, Lord
Woolf said that the Guidelines should be used as a starting point ˜not only because
it is convenient to do so™ but also because they are ˜the most reliable tool . . . as to
what is the correct range of damages for common classes of injuries™.132 They present
˜a distillation of the awards of damages that have been and are being awarded by
judges in courts up and down the country™.133
Another factor that a¬ects awards for non-pecuniary loss in some cases settled out
of court is the internationalization of the tort system.134 In some cases, especially
those involving large-scale disasters and large corporate defendants, awards for non-
pecuniary loss tend to be higher in the USA than in the UK. If the defendant in a UK
action is a multinational corporation with US roots or US relatives, the claimants may
seek to bring their action in the USA. Whether or not this will be possible depends
not on rules of tort law but on rules of procedure and of the con¬‚ict of laws.135 If there
seems a good chance that an action in the USA might be allowed, the defendant and

126 Law Com. No. 56, Personal Injury Litigation: Assessment of Damages, para. 20.
127 In cases where damages both for pain and su¬ering and for loss of amenities are awarded, the
two are usually combined into a single lump sum called ˜general damages™ and the individual
components are not calculated separately.
128 Guidelines for the Assessment of Damages in Personal Injury Cases, 4th edn (London, 1998), vii.
129 Ibid.
130 Ibid., 2.
131 Ibid.
132 Third edition (1996), viii.
133 Guidelines for the Assessment of Damages in Personal Injury Cases, 7th edn (Oxford, 2004), vii.
134 The importance of this phenomenon is not limited to awards for non-pecuniary loss. For
example, damages recoverable from airline operators by victims of air crashes may be limited by
the Warsaw Convention. Before the Convention, injured persons were free to sue under diverse
national laws and individual airlines were free to limit their liability by contract with passengers.
The Convention was designed to provide a single compensation regime for air accidents. Airlines
accepted a form of strict liability in return for limitations on the scope of their liability in terms
of compensatable harm (limited to ˜bodily injury™: Morris v. KLM Royal Dutch Airlines [2002] 2
AC 628) and amounts of damages. It may be possible to evade such limitations by suing someone
other than the airline operator “ the aircraft manufacturer, for example.
135 The most famous case in which a US court refused to accept jurisdiction was that which arose
out of the Bhopal gas leak: Re Union Carbide Corporation Gas Plant Disaster at Bhopal in India in
December, 1984 (1987) 809 F 2d 195; J. Cassels, ˜The Uncertain Promise of Law: Lessons from
Bhopal™ (1991) 29 Osgoode Hall LJ 1, 17“20. Each US jurisdiction has its own rules on this issue.
Californian courts have been quite generous to foreign litigants: Holmes v. Syntex Laboratories
Damages for personal injury and death 169

its insurers may be persuaded to settle at what are called ˜mid-Atlantic rates™, and to
pay compensation for non-pecuniary loss at a level somewhere between UK and US
norms. This happened, for example, in the Piper Alpha oil rig case in 1988.136
The question has often been discussed whether there is a case for a legislative tari¬
to replace the judicial one.137 At present, the only head of damages ¬xed by statute is
that for bereavement under the Fatal Accidents Act 1976 “ the sum awarded is cur-
rently £10,000. It is acknowledged that any more extensive statutory tari¬ would have
to be somewhat ¬‚exible because experience, both of the tort system and of the indus-
trial injuries system, has shown that very many injuries cannot be neatly labelled and
identi¬ed in a tari¬ schedule. Both the Law Commission (in 1973) and the Pearson
Commission found little support for the idea of a statutory tari¬, and rejected it.
There was, however, more support for a legislative maximum on awards for non-
pecuniary loss, and the Pearson Commission only rejected this idea by a single vote.
It seems unlikely that such a maximum would make a great deal of di¬erence unless
it were set well below the current ¬gure. The Pearson Commission discussed the pos-
sibility of a maximum of only ¬ve times the average industrial wage, i.e. about
£110,000,138 and that certainly is well below the present maximum of about £220,000.
In its 1995 Consultation Paper the Law Commission adopted an agnostic position in
relation to all forms of legislative tari¬. By contrast statutory limits on this and other
heads of damages have, in recent years, been adopted in all Australian jurisdictions.
The Pearson Commission accepted the general basis of most of the present law,
though they made one important proposal, namely that no damages for non-
pecuniary loss should be recoverable for non-pecuniary loss su¬ered during the ¬rst
three months after the date of injury.139 As about 95% of those injured in accidents
are su¬ciently recovered to return to work within 3 months,140 this would have a

Inc. (1984) 202 Cal Rptr 773; Corrigan v. Bjork Shiley Corp. (1986) 227 Cal Rptr 247. But attempts
to sue engine and aircraft manufacturers in Louisiana following the M1 air crash in 1989 (partly
in order to evade limits on compensation imposed by international Conventions) failed; as did
attempts to bring actions in Indiana on behalf of claimants suing in respect of a drug called
Opren (see C. Hodges, Multi-Party Actions (Oxford, 2001), 329). English courts have the power
to issue an injunction to prevent a person bringing an action in a foreign court, but will do so
only in extreme cases: Dicey and Morris on the Con¬‚ict of Laws, 13th edn (London, 2000), 12-
057“12-069. English courts can also ˜stay™ (i.e. stop) claims that are brought in England when
another ˜forum™ could be more suitable. Various factors are relevant to deciding which is the most
suitable ˜forum™ for a case to be heard in. One is the availability of funding for the claim: Connelly
v. RTZ Corporation Plc [1998] AC 854; Lubbe v. Cape Plc [2000] 1 WLR 1545. See also P.
Muchlinski, ˜Corporations in International Litigation: Problems of Jurisdiction and the United
Kingdom Asbestos Cases™ (2001) 50 IC LQ 1.
136 Occidental Oil, the rig operators, paid out over £100 million in compensation and then
sued contractors working on the rig for an indemnity. The action took more than 4 years
to reach judgment: ˜Judgment in the United Kingdom™s longest civil hearing™ [1997] New
LJ 1302.
137 Law Com. No. 56 (1973), paras. 31“6; Pearson Report, vol. 1, paras. 377“9; Law Com.
Consultation Paper No. 140 (1995), paras. 4.53“67.
138 The ˜average industrial wage™ is no longer used in government statistics. Instead I have used
median annual earnings for full-time employees, which were £22,060 in the 2002“3 tax year.
139 Pearson Report, vol. 1, paras. 382“9.
140 Ibid., table 2.
170 Chapter 6

very dramatic e¬ect on eliminating claims for pain and su¬ering and loss of amenity
in minor cases. Moreover, the total saving would be very substantial. The insurance
survey conducted for the Pearson Commission showed that some two-thirds of
damages paid out by insurers was for non-pecuniary loss, and that the proportion
was highest in small cases.141 In its 1995 Consultation Paper the Law Commission
rejected the Pearson Commission™s proposal on a number of grounds including that
pain is often at its most intense in the early stages after injury.142 Indeed, the Law
Commission is against any form of threshold for recovery for non-pecuniary loss.

6.5.3 Subjective factors
The tari¬ approach is largely a result of the demise of juries in personal injury
actions, which has led judges and appeal courts to stress the value of consistency
in assessing damages for non-pecuniary loss. However, the courts still profess to
compensate the injured person for the e¬ects of the injuries on them as a unique
individual. Clearly, there is potential inconsistency between this personalized
approach143 and a tari¬ system; but as already noted, the tari¬ for particular injuries
consists of a range of possible awards (i.e. an upper and lower ¬gure) rather than a
single sum. This allows the individual circumstances of particular victims to be taken
into account. Thus, a claimant with a hand injury may recover more if he or she was
an amateur pianist who took much pleasure in the hobby; a woman with a leg injury
may recover more if she was formerly keen on dancing and is now unable to dance
at all. Here again, we ¬nd a fundamental inconsistency between the tort system and
the social security system. In the latter, only the industrial injuries scheme recognizes
loss of faculty or disability as a ground for compensation under the scheme, and the
assessment is entirely objective: no personal factors (other than age and sex) are taken
into account. Indeed, a committee reviewing the assessment of disabilities in the
industrial injuries scheme thought that it would be ˜inequitable™ to do so, as the
Pearson Commission apparently did despite its acceptance of the common law
system.144 This is doubtless based on the view that everyone places an equal value on,
for example, their hand (leaving out of account loss of earnings, which are separately
compensated). The case for equality of treatment in this respect seems very strong.
One problem which has caused much trouble is that of assessing the damages
awardable to a victim who has been reduced to a ˜persistent vegetative state™.145
Medical science can now keep people with the most devastating injuries alive

141 Ibid., vol. 2, table 107.
142 Law Com. Consultation Paper No. 140, paras. 423“6.
143 The most extreme version of the personalized approach is that of Diplock LJ (dissenting) in Wise
v. Kaye [1962] 1 QB 638, who would have assessed the e¬ect of injuries on each individual
victim™s happiness.
144 Pearson Report, vol. 1, para. 823; see also paras. 379“81.
145 It is lawful in certain circumstances to withdraw life support from a person in such a state: Airedale
NHS Trust v. Bland [1993] AC 789. This possibility should, in theory, be taken into account in
assessing tort damages both for ¬nancial and for non-pecuniary loss. The cost of maintaining
people in such a state is very high, and has been the cause of much controversy.
Damages for personal injury and death 171

in a state of complete coma for many months, or even years, with no hope of recov-
ery. In a case of this nature it is hard to see what purpose there can be in awarding
lump-sum damages for disabilities or loss of amenities, or even for lost earnings
if there are no dependants. There is no question of providing substitute pleasures
for those forgone, because the injured party is unable to enjoy any pleasures; nor is
there any question of providing a solace for pain, su¬ering or mental distress,
because the victim feels none. Yet, the courts have held that although damages for
pain and su¬ering cannot be awarded, none the less, damages for loss of ˜amenities™
or ˜faculties™ must be awarded; and these damages run into many thousands
of pounds. In West v. Shephard146 a majority of the House of Lords, following the
majority of the Court of Appeal in Wise v. Kaye,147 decided that compensation is
awarded for the objective fact of ˜loss™ in cases of this nature. A person who ˜loses™
a leg gets compensation for the fact of losing the leg, and a person who is deprived
of all the pleasures of life gets compensation for the fact of that deprivation. Lack
of consciousness of the deprivation, said the House of Lords, cannot reduce the
objective fact of the ˜loss™; though consciousness of the deprivation can increase the
damages by reason of the mental distress that this would involve.
The result of this approach is that the law draws a very sharp distinction between
death and permanent unconsciousness. If a person dies as a result of personal
injuries, no damages for loss of amenities will be recoverable in respect of the period
after death, even if the deceased has dependants; but if a person is reduced to a state
of permanent unconsciousness, substantial damages under this head will be
awarded even if that person has no dependants and the damages will eventually
accrue as a windfall to the bene¬ciaries of the estate. The Pearson Commission
recommended that damages for non-pecuniary loss should cease to be recoverable
in cases of permanent unconsciousness.148

6.5.4 Should damages be payable for intangible losses?
As we have seen, damages for non-pecuniary loss may be awarded to victims of per-
sonal injury, but otherwise they will only be awarded for the death of a spouse, civil
partner or unmarried minor child under the Fatal Accidents Act 1976. In other cir-
cumstances, no damages can be awarded for non-pecuniary loss. This rules out, for
instance, any damages for the distress and anguish of parents whose child su¬ers
crippling brain damage and whose life may thereby be shattered. Similarly, nothing
is recoverable for the death of someone other than under the Fatal Accidents Act
1976. So no damages will be awarded for the death of an adult child or of a non-
marital or ˜non-civil™ partner; and a husband or wife cannot recover anything for the
e¬ects on themselves of a serious accident to their spouse. All this is not to suggest

146 [1964] AC 326.
147 [1962] 1 QB 638.
148 Pearson Report, vol. 1, paras. 393“8. The Law Commission reached the opposite conclusion: Law
Com. No. 257, 2.8“2.24.
172 Chapter 6

that there should be payment of damages for non-pecuniary loss in these situations,
but to stress the di¬culty of justifying such damages in the cases where they are
presently awarded. As we have seen previously, the Pearson Commission discovered
that something of the order of two-thirds of all tort payments are attributable to
non-pecuniary loss; and much of this sum is paid in relatively trivial cases where a
complete recovery is made by the victim within a short time. The majority of the
Pearson Commission found it ˜hard to justify payments for minor or transient
non-pecuniary losses™, and they went on to say: ˜The emphasis in compensation for
non-pecuniary loss should in our view be on serious and continuing losses, espe-
cially loss of faculty.™149
It is, once again, necessary to remember the remarkable disparity in treatment
between tort victims who obtain full compensation for their pecuniary losses and
damages for non-pecuniary losses as well, and most other classes of victims of acci-
dents and disease who rarely obtain full compensation even for pecuniary losses,
let alone anything extra for non-pecuniary losses. The truth would appear to be
that there is a penal or punitive element underlying damages for non-pecuniary
loss,150 especially damages for bereavement. This is particularly obvious in cases
against corporate defendants arising out of mass disasters such as railway acci-
dents and ¬res. Indeed, in recent years there has been considerable public pressure,
largely generated by such disasters, for increases in the size of awards for non-
pecuniary loss, especially to relatives151 and to survivors of such incidents who may
have su¬ered little by way of physical injury but who, nevertheless, have endured
much mental su¬ering, including the condition called ˜post-traumatic stress disor-
der™. Calls for increases in awards of damages for non-pecuniary loss in personal
injury cases also arose as a response to a number of widely reported defamation
awards which were far greater than the largest awards for non-pecuniary loss given
to victims of personal injuries.152 Such awards supposedly compensate the claimant
for (non-pecuniary) loss of reputation, and there was a common feeling that if
injury to reputation warrants high damages, mental injury resulting from personal

149 Pearson Report, vol. 1, para. 384.
150 See also A. Unger, ˜Pain and Anger™ [1992] New LJ 394.
151 I.e. damages for bereavement under the Fatal Accidents Act 1976. It is also suggested that the class
of entitled claimants should be extended beyond the present married spouses and civil partners
(it is anomalous that some cohabitees can recover damages for ¬nancial loss under the Act but
not for bereavement) and parents of unmarried minor children. Under the Damages (Scotland)
Act 1976, damages for bereavement are subject to no monetary limit and are available to a wider
class of relatives than in England.
152 The size of such awards starkly illustrates the undesirability of allowing juries to assess damages.
The larger defamation awards were also quite out of proportion to awards in e.g. false impris-
onment cases. For instance, a child who received 129 days of ˜treatment™ under the notorious
˜pin-down regime™ imposed in local authority homes in Sta¬ordshire reportedly received only
£42,000: The Times, 29 May 1991. Considerable controversy was generated in the early 1990s by
the size of awards for wrongful dismissal made to women forced to leave the armed forces when
they became pregnant, as compared with damages for personal injuries recovered by service
personnel and civilians injured in the course of action: see e.g. Independent, 29 March 1994; 6
April 1994.
Damages for personal injury and death 173

injury, whether to oneself or to others, deserves more. Steps have now been taken
to bring defamation awards more into line with damages for non-pecuniary loss in
personal injury cases, and this particular cause for dissatisfaction has thus been
Despite all this, however, punitive damages are wholly inappropriate when
damages are normally paid by insurers and not by tortfeasors. It is perhaps only in
the most serious cases of long-term pain and loss of faculty resulting from major
physical injuries that there is a good case for damages for non-pecuniary loss.153

Overall maxima154
While damages for non-pecuniary loss are subject to an informal upper limit, the
operation of the hundred-per cent principle means that there is no upper limit on
damages for pecuniary loss. The result is that awards in cases of serious and long-
term injuries causing severe disablement are very great indeed. The largest reported
lump sum recovered in a personal injuries action to date is in the region of £9
million.155 Awards of this size are worth very much more than the social security
bene¬ts payable even to the most seriously handicapped people. On the other hand,
there is no reason to think that the element of such awards which represents ¬nancial
losses is excessive given the cost of living.156 It is, nevertheless true that recipients of
tort awards are a very privileged class of the disabled, and their position has, if any-
thing, improved over the last 20 years relative to that of other disabled persons.

6.7 Punitive damages
It was argued earlier that damages for non-pecuniary loss may often have a punit-
ive element even though, in theory, they ˜compensate™ the injured person. We must
also make some mention of what are called ˜punitive™ or ˜exemplary™ damages. Such
damages are expressly designed to express the court™s disapproval of what the
wrongdoer has done and have no compensatory component or function but are

153 Contrast the view of the Law Commission: Law Com. No. 257, 2.25“2.28.
154 Damages awards vary widely from one jurisdiction to another within the EC: D. McIntosh and
M. Holmes, Personal Injury Awards in EU and EFTA Countries, 3rd edn (London, 2003). This
may encourage forum-shopping within the Community. So, too, may di¬erent rules a¬ecting
the recoverability of damages awards: A. Geddes,˜Di¬culties Relating to Directives A¬ecting the
Recoverability of Damages for Personal Injury™ (1992) 17 European LR 408.
155 Biesheuvel v. Birrell [1999] PIQR Q40.
156 Long-term care costs represent a signi¬cant proportion of damages in serious cases and may be
the largest single item. In aggregate, such costs have no doubt accounted for a signi¬cant pro-
portion of increases in the cost of tort compensation over the past 30 years as advances in
medical technology have made it possible to keep alive seriously injured people who would for-
merly have died. This is one reason why road-accident fatalities have fallen dramatically in the
past 40 years. The relatively small number of serious cases account for a very signi¬cant pro-
portion of total tort compensation. As increasing numbers of seriously injured people are kept
alive, the cost of compensation may go on increasing in real terms even if the total number of
claims remains constant or even falls.
174 Chapter 6

additional to compensatory damages. They are available only in a limited range of
circumstances. For our purposes the only situation of importance in which puni-
tive damages might be available is where a tortfeasor has acted with deliberate dis-
regard for the safety or health of the claimant in order, for example, to save money.
Such damages may be available against corporate defendants in mass disaster cases
resulting from disregard of safety laws.
It is commonly believed that an award of punitive damages could not be made
in an ordinary negligence action because negligent conduct is, by de¬nition, not
calculated, and because the purpose of a negligence action is solely to compensate
the injured person.157 The ¬rst of these reasons is groundless because the essence of
negligent conduct is failure to take reasonable precautions against foreseeable risks,
and this may be done in a deliberate and indeed callous way. As for the second argu-
ment, there is no reason why the tort of negligence should not be used to punish as
well as to compensate, if this is thought a good idea. Punitive damages are more
common in personal injury cases in the USA, especially where people are injured
by defective products. The best justi¬cation for such damages is that they may deter
the defendant and others similarly placed from taking deliberate risks with health
and safety in the future by stripping them of any ¬nancial gain that may have been
made and perhaps by imposing a penalty over and above any ¬nancial bene¬t
derived. Punitive damages are, however, objectionable, in personal injury cases at
least, because they over-compensate the injured person and encourage vindictive
gold-digging. It would be better to ¬nd ways of forcing enterprises to invest their
˜ill-gotten gains™ in safety than to divert such resources to tort claimants who have
already been fully compensated.

157 See Kralj v. McGrath [1986] 1 All ER 54. The terms used in this case was ˜aggravated damages™, but
these are essentially the same as punitive damages. The law is di¬erent in Australia: Lamb v.
Cotogno (1987) 164 CLR 1. There is Privy Council authority for awarding exemplary damages for
˜gross™ or ˜outrageous™ or ˜¬‚agrant™, although non-deliberate, departure from the standard of care
(Bottrill v. A [2003] 1 AC 449), but this decision is hard to reconcile with the House of Lords case
of Rookes v. Barnard [1964] AC 1129. Decisions of the House of Lords are binding on English
courts, but decisions of the Privy Council are not. However, the decision in Bottrill might indi-
cate that the House of Lords would modify the restrictive approach in Rookes if it got the chance.

An appraisal of the fault principle

The fault principle has traditionally been understood as a principle of morality,
which can justify not only the imposition of liability for death and personal injury
but also the assessment of compensation according to the full compensation and
hundred-per cent principles. Grosser fault may even be seen as justifying the award
of exemplary or punitive damages. But in moral terms, the fault principle might be
thought to su¬er from serious defects. It can also be attacked on social and practi-
cal grounds. In this chapter we consider various arguments that might be made
against the fault principle as a basis for the payment of compensation to victims of
personal injuries by those who in¬‚ict them.

7.1 The compensation payable bears no relation to the degree of
Under the fault principle, being required to pay compensation is a sort of penalty
for bad conduct. In the criminal law, it is seen as a basic requirement of justice that
˜the punishment ¬t the crime™ in terms of the seriousness of both the o¬ender™s
conduct and the consequences of that conduct. In tort law, on the other hand, there
is no such idea that the compensation payable should be proportional to the tort-
feasor™s fault. Fault is like a magic talisman; once it is established, all shall be given
to the injured party. It is generally immaterial whether the fault was gross or trivial1
or whether the consequences of the fault were catastrophic or minor. A degree of
fault on the part of someone justi¬es compensating the injured person for all the
losses su¬ered, provided the claimant was in no way personally at fault. Yet the seri-
ousness of the consequences of a negligent action often bear no relation to the
degree of fault which gave rise to it. A piece of momentary thoughtlessness on the
road may cost someone their life and cause great loss to their family; but similar
acts of thoughtlessness may be committed by scores of others every day with only

1 There are some exceptions. For example, people are expected to take more care for the safety of
others than for their own safety, and adults are expected to take more care than children. In some
cases, too, a defendant will be held liable for negligence only if their conduct was so unreasonable
that no reasonable person in their position would have engaged in it: see P. Cane, The Anatomy of
Tort Law (Oxford, 1997), 41“2.

176 Chapter 7

minor or even no adverse consequences. It has been estimated that for every acci-
dent on the roads there are 122 near misses,2 and a US study found in a test under
normal driving conditions in Washington, DC that even ˜good™ drivers committed
an average of nine driving errors of four di¬erent types in every ¬ve minutes.3 Yet
in this country, in 2004 only about two car drivers in every 1,000 were injured in
road accidents,4 and most road accidents cause only minor property damage. Thus,
it seems that whether an act of negligence ends up in the accident statistics or as a
near miss, and whether it causes much, little or no harm, are largely matters of
chance, outside the control of the person at fault. They would certainly appear to
have little correlation with the defendant™s culpability.
On the other hand, in applying the fault principle courts do sometimes explicitly
recognize a distinction between negligence, on the one hand, and error or mistake,
on the other.5 Not every mistake constitutes fault, because even reasonable people
can make mistakes. For instance, when dealing with allegations of medical negli-
gence, and in the context of contributory negligence, courts are apt to insist that not
every mistake should be treated as grounds for imposing liability or reducing the
claimant™s damages, as the case may be. But in other cases courts often appear to
assume that the reasonable person never makes a mistake. On the road, for instance,
almost any driving error is apt to be treated as negligence without argument, despite
the evidence that the typical driver commits driving errors every few minutes. And
in other situations it often happens that acts of casual or momentary carelessness
can be treated as negligence, even though most of us regularly commit such acts
without thinking ourselves to be guilty of fault or blameworthy conduct.
It is true that there is some evidence for asserting that in road accident cases there
is some correlation between accident involvement and driving ability. Transport and
Road Research Laboratory (TRRL) Reports show, for instance, that there were
signi¬cant di¬erences between the driving ability of two groups of ¬fty drivers, the
members of one group all having been convicted of careless or dangerous driving.6
Consequently, it is going too far to say that accident involvement is entirely a ques-
tion of bad luck; careless drivers are more likely to have accidents than careful drivers.
Moreover, it may well be that many road accidents, and perhaps the more serious
accidents, are the result of acts of carelessness which are seriously and undeniably

2 M. Austin, Accident Black Spot (Harmondsworth, 1966), 33.
3 Driver Behaviour and Accident Involvement: Implications for Tort Liability (Automobile Insurance
and Compensation Study: US government Printer, Washington, 1970), 176“80. In a British survey
of 300 drivers in 1996, the respondents admitted to making an average of 50 serious errors a week
and to being careless at least once on 98% of their journeys. More than half the motorists said they
had had an accident; only 4% said their crashes were genuine accidents with no human error
involved: The Times, 2 December 1996.
4 Department for Transport, Road Casualties in Great Britain 2004, table 30.
5 A. Tunc ˜Fault: A Common Name for Di¬erent Misdeeds™ (1975) 49 Tulane LR 279.
6 Laboratory Report (LR) 70 (1967) and LR 146 (1968); see also LR 395 (1971) and LR 449 (1972).
More recent research stresses the relationship between individual accident risk and psychological
and social factors (such as social deviance): Research Report 306 (1991); Contractor Report 309
An appraisal of the fault principle 177

culpable, such as driving while intoxicated or speeding. There is also evidence to
suggest that drivers who commit many ˜violations™ (i.e. deliberately unsafe driving
behaviour) “ as opposed to ˜errors™ (such as misjudgments and failures of observa-
tion) “ are more likely to be involved in accidents.7 Nevertheless, it is a matter of
everyday observation and experience that extreme carelessness, and even deliberate
violations, are very frequently committed with no, or only minor, ill consequences.
Tort law™s lack of concern with the relationship between culpability and liability
for consequences is a re¬‚ection of the fact that tort law is much more concerned with
victims than is criminal law. Although the idea of personal responsibility might
seem to require that attention be paid to the relative culpability of the tortfeasor™s
conduct, tort law has an equally strong concern with compensating for harm
su¬ered. The basic idea is that as between a tortfeasor and a totally innocent victim,
it is only fair that the harm su¬ered be borne by the former rather than being shared
between them. However, even as between the victim and the tortfeasor, doubts are
often felt about the justice of imposing liability on the latter for the most catas-
trophic consequences of a negligent act. We have already seen how various attempts
are made to limit liability in extreme cases by invoking causal or risk principles, or
by denying liability for unforeseeable consequences. But even if justice between
victim and tortfeasor demands that liability be imposed on the latter however
extreme the consequences, and however tri¬‚ing the negligence, it may nevertheless
be felt unjust that tortfeasors should be left to bear this bill as between themselves
and society. Since the tortfeasor may be no more culpable than many others, and
may only have done what others are constantly doing, it may seem inequitable that
the few whose negligence results in injury or loss to others should be required to
bear this burden while the majority of negligent people go free. From this perspec-
tive, liability insurance, which spreads the burden of compensation amongst a pool
of potential tortfeasors, may be seen not only as a means of ensuring that victims
are compensated, but also as a way of reducing the injustice of the law™s lack of atten-
tion to degrees of fault.

7.2 The compensation bears no relation to the means of the
In tort law, the tortfeasor™s wealth or ¬nancial means are usually irrelevant to lia-
bility.8 The fact that a tortfeasor is rich is no ground for imposing liability, and the
fact that they are poor is no ground for not imposing liability. Most people would
probably accept as morally right this principle of equality before the law regard-
less of wealth, which is implicit in the fault principle. But when we take into
account the fact that once liability is imposed, the compensation payable will bear

7 Driver Behaviour Research Group, University of Manchester, In¬‚uencing Driver Behaviour and
Attitudes (No. 17) (undated).
8 But, as we have seen (, in some circumstances the wealth of the defendant can a¬ect the
standard of care required.
178 Chapter 7

no relationship to the means of the tortfeasor, we may begin to doubt whether it
really is fair to ignore their ¬nancial position. So, for example, a parent might feel
morally obliged to pay a neighbour a few pounds for a window broken by their
child; but it is doubtful whether parents would feel morally obliged to sell up house
and home and impoverish themselves and their family if the child were to blind a
neighbour™s child with an airgun and were held liable for damages of tens or hun-
dreds of thousands of pounds. A person who loses a book borrowed from a friend
would not hesitate to pay for the book even if its loss was not the borrower™s fault;
but a person who borrows a friend™s car may be very reluctant to pay out the whole
value of the car if it was completely wrecked in an accident while they were driving
it, and it turned out to be uninsured.
No criminal court would think of imposing a ¬ne for culpable conduct of the
amounts that civil courts award as damages every day, without serious inquiry into
the ability of the defendant to pay.9 The fact that tort law ignores the wrongdoer™s
means is justi¬ed by saying that the ˜purpose™ of the civil law is to compensate and
not to punish. But the ˜purpose™ of the law is irrelevant to the tortfeasor who is
made to pay the damages “ what matters to them is the e¬ect of the law, not its
˜purpose™. So far as the wrongdoer is concerned, deprivation of money by the court
is precisely as painful whether the ˜purpose™ is to punish the wrongdoer or to com-
pensate the victim.
There are strong social grounds for not placing crushing legal liabilities on people
of modest means. Most people would experience the utmost di¬culty in paying a
damages award of any appreciable size. The only asset of any real value that very
many people own is the house in which they live “ or, more accurately in most cases,
the value of the house over and above the value of any mortgage secured by it. To
impose a liability on a person which would require that person to dispose of their
house (or to borrow large amounts of money using it as security) would plainly
cause them and their family a great deal of dislocation and misery. Of course, the
victim must not be forgotten, and as between a needy victim and a tortfeasor of
limited means, justice may seem to favour the former. But it hardly seems fair or
socially desirable to strip a person of everything because of what may have been a
venial act of negligence. It is exactly for this reason that the law allows liability insur-
ance, and without it the tort system could not operate e¬ectively as a compensation
mechanism. However, liability insurance con¬‚icts with the rationale of the fault
principle in that it relieves the faulty person of the burden of paying damages.
Moreover, like tort damages themselves, liability insurance premiums are unrelated
to the means of the insured.
Suppose the tortfeasor is not insured against the liability; if there is no way of
compensating the victim except at the expense of someone who has caused the

9 In criminal law, the idea that monetary punishments should re¬‚ect the means of the o¬ender is
called ˜the principle of equal impact™. It was embodied in the Criminal Justice Act 1991, ss. 18“21,
but these provisions have since been repealed. See A. Ashworth, Sentencing and Criminal Justice,
3rd edn (London, 2000), 210“11.
An appraisal of the fault principle 179

injuries, we may well feel that justice is on the side of the victim even if the tortfea-
sor has to sell up house and home to pay for the damages. Even in this situation,
however, a case could be made for dropping damages for non-pecuniary loss. If the
victim™s economic losses are made good it would arguably be more harmful to
society to require the tortfeasor personally to pay substantial damages for intangi-
ble losses than it would be for the victim to forgo them.

7.3 A harm-doer may be held legally liable without being
morally culpable and vice versa
As we will see later, the fault principle cannot be justi¬ed on practical grounds, such
as convenience, e¬ciency, speed or cheapness of operation. The traditional justi-
¬cation is that the legal concept of liability for fault embodies a moral principle to
the e¬ect that if a person, by blameworthy conduct, causes damage or loss to an
innocent person, the former should compensate the latter for that damage or loss.
But there are at least two grounds on which people have questioned whether tort
law actually does embody such a moral principle. In the ¬rst place, it is said, if tort
law was based on fault would it not prohibit liability insurance, vicarious liability
and other loss distribution devices by which the burden of paying compensation
can be shifted from a party at fault to another party not at fault?

7.3.1 Collective liability
Consider collective liability in this regard. As a moral concept, the idea of fault
applies most straightforwardly to conduct of individuals. When it is argued that a
company, or a local authority or some other organization or group was at fault, the
moral content of the allegation may seem rather more attenuated. Suppose a claim
is brought against a corporation for negligent failure to appreciate and guard against
a danger in the workplace. Often, the real complaint is not that some particular indi-
vidual was at fault, but that as a result of some failure of organization in the
company, no individual had responsibility for anticipating and preventing the acci-
dent that occurred. We might be perfectly happy to hold the company responsible
even though we cannot point to any individual who was personally to blame. In
Carmarthenshire County Council v. Lewis10 a young child wandered out of a nursery
school maintained by the council, down a lane, through a gate and on to a busy road,
where a lorry driver, trying to avoid the child, crashed into a tree and was killed. The
Court of Appeal held that the child™s teacher was negligent in failing to keep a
su¬cient eye on him, but the House of Lords exonerated the teacher from the charge
of negligence while still holding the defendants liable. ˜They™ (that is, the County
Council) were negligent; ˜they™ should not have allowed an unlocked gate at the end
of a lane near a nursery school bordering a busy road. But who actually was ˜at fault™?
Was it every councillor who ought to have proposed a resolution at a meeting of the

10 [1955] AC 549.
180 Chapter 7

Council for the appointment of someone whose duty it was to prevent such acci-
dents? Or the town clerk? Or the head-teacher of the school? And why hold the
Council liable rather than the person(s) who ought to have taken precautions to
prevent the accident?
If our purpose in asking such questions were to prevent similar occurrences in
the future, we would have a strong incentive to try to ¬nd a responsible individual
so that we could repair the defect in the organization™s risk-management system.
Similarly, if our purpose were to discipline or punish someone for what had hap-
pened, we would want to be able to identify the individual who should have taken
steps to prevent the accident. But the main aim of a negligence action is to com-
pensate injured persons; and so long as we are satis¬ed that appropriate precautions
ought to have been taken within the organization to prevent the harm, there is no
good reason not to impose liability to pay compensation on the organization, even
though there is an obvious sense in which the (moral) fault must have lain with indi-
viduals within the organization rather than the organization as such.11 Traditionally,
the law has been much less willing to impose criminal liability on organizations than
to impose civil liability, partly at least because organizations were thought incapable
of the sort of (morally) culpable conduct that justi¬es criminal punishment. But
when it comes to civil liability to compensate for harm, the law has, for centuries,
had no qualms about organizational liability. In this respect, it re¬‚ects attitudes and
values held widely in the community at large. For this reason, to say that tort law
sometimes imposes liability in the absence of fault is not a criticism but only an

7.3.2 The objective de¬nition of fault
A second ground on which tort law™s adherence to a moral principle of responsi-
bility for fault has been questioned is this: if the law really re¬‚ected morality, it
would not adopt an objective de¬nition of fault which, on the whole, ignores the
personal qualities of the persons involved and which does not require that
the harm-doer should have had any consciousness of moral wrongdoing, or even
of the risk they were creating or of the dangerousness of their conduct.12
Negligence, as we have seen, is de¬ned as the failure to take reasonable care; that
is, the care which the reasonable person would have taken to avoid risks which the
reasonable person would have guarded against. It does not matter that the injurer is
not a ˜reasonable person™ but is clumsy or stupid or forgetful or has bad judgment.
It does not matter that the injurer is inexperienced or young or old or (probably)

11 See also Cassidy v. Ministry of Health [1951] 2 KB 343.
12 As explained in the sixth edn, my views on this topic have developed and changed. However,
because I still agree with the basic thrust of what follows, and because to take matters further
would launch the discussion into deep philosophical waters, I have decided again to leave this
section largely untouched. My current views can be found in ˜Retribution, Proportionality and
Moral Luck in Tort Law™ in P. Cane and J. Stapleton eds., The Law of Obligations: Essays in
Celebration of John Fleming (Oxford, 1998); The Anatomy of Tort Law (Oxford, 1997), ch. 7; and
Responsibility in Law and Morality (Oxford, 2002), ch. 3.
An appraisal of the fault principle 181

even that they are handicapped or disabled. It does not generally matter that the
injurer could not personally have foreseen the risk or avoided the accident. Even
those who have wholeheartedly supported the principle of ˜no liability without fault™
have also subscribed wholeheartedly to the objective de¬nition of fault. A reason
often given for this approach is that the injury in¬‚icted is the same whether the
injurer could or could not personally have avoided the accident. So, for instance,
Mr Justice Holmes declared, in a celebrated passage:13

If for instance, a man is born hasty and awkward, is always having accidents and
hurting himself or his neighbours, no doubt his congenital defects will be allowed for
in the courts of Heaven, but his slips are no less troublesome to his neighbours than if
they sprang from guilty neglect. His neighbours accordingly require him, at his proper
peril, to come up to their standard, and the courts which they establish decline to take
his personal equation into account.

This is a weak argument because it does not go far enough: the damage or injury is
the same whether or not there has been fault at all, even as objectively de¬ned by
the law. If the reason for adopting an objective standard of fault is that when
damage is done the victim has been hurt and deserves to be compensated whether
or not there has been subjective fault, it is hard to see why it does not also follow
that an injured person should be compensated whether or not there is fault at all,
whether objectively or subjectively judged.
A di¬erent approach would be to argue that the law™s objective de¬nition of fault
is not actually out of line with morality at all. Morality does not always acquit a
person of blame for acts traceable to defects of personality or capacity. If an adult14
behaves badly as a result of stupidity or forgetfulness or bad judgment, we would not
necessarily hold that person morally blameless. We might say that they should try
harder next time or that they should not put themselves in situations in which their
personality faults are likely to produce adverse outcomes. If a person su¬ers from
some physical disability about which nothing can be done, we would not normally15
blame them for accidents resulting from that disability, but we might well blame
them for putting themselves in situations where their disability might cause acci-
dents.16 A blind person is not to be blamed for being blind, but could be blamed for
attempting to drive a car on a public road and for any accident resulting from the
attempt. On the other hand, if a person is suddenly and unexpectedly overtaken by
a physical disability (e.g. if a person has a heart attack while driving) and as a result

13 The Common Law, Holmes™s most famous extra-judicial writing, was originally published in
1881. This passage can be found at pp. 86“7 of the edition edited by M. DeW. Howe (Boston,
14 Children must be given time to develop physical skills and a sense of moral responsibility before
being subjected to the full rigours of our moral code. The law, too, treats young children more
leniently than adolescents and adults: McHale v. Watson (1964) 111 CLR 384.
15 But it would be di¬erent if the disability was self-in¬‚icted as a result e.g. of consumption of
alcohol or drugs.
16 See Jones v. Dennison [1971] RTR 174.
182 Chapter 7

causes an accident, we would not call that person morally blameworthy; nor would
he or she be legally liable for negligence.17
It would appear, therefore, that a dichotomy between objective legal fault and
subjective moral fault is too simple.18 The gap between law and morality is not as
great as might at ¬rst appear. However, there are some areas in which the gulf seems
quite wide. The most obvious is the way the law deals with lack of skill resulting
from inexperience. In general, the inexperienced driver, for instance, is held to the
same standard of care as the experienced, even though it is clear that inexperienced
drivers as a group cause more accidents than experienced drivers.19 But drivers must
learn to drive somewhere, and they must acquire experience of driving on the roads
amidst tra¬c; to expect inexperienced drivers as a group to display the same skill
and judgment as experienced drivers does seem morally unfair. The courts have, in
several cases, also taken a harsh approach to accidents caused by mental impair-
ment that the injurer could do nothing about.20
But even if the law is out of step with morality, it does not follow that this is a
bad thing. If we think of the law as designed to regulate the conduct of people to
whom it is addressed, it does indeed seem unreasonable to treat as negligence some-
thing which a person could not avoid. On the other hand, if we think that the main
purpose of the law is to compensate injured persons, there is no reason why moral
fault should be the criterion of liability to pay compensation. Indeed, if this is our
aim, the criterion of whether a person is entitled to compensation ought to be
whether they have been injured, regardless of how they were injured. From this
point of view, the chief shortcoming of the tort system is not that it sometimes
compensates people whose injuries were not the result of moral fault, but that it
fails to compensate very many other people who have su¬ered injuries in circum-
stances that do not fall within the tort system at all.

7.3.3 Moral culpability without legal liability
So far we have been discussing cases of legal liability that may not involve moral
culpability. The converse “ where a person may be morally blameworthy without
having committed a legal wrong “ is less likely; but there is one sort of case which
does raise the possibility, namely where a person has been guilty of an omission. We
have already discussed liability for omissions (3.2.2), and it remains only to notice

17 Waugh v. James Allan Ltd [1964] 2 Lloyd™s Reports 1.
18 The main point can be put starkly by saying that ability to comply with morality, as much ability
to comply with the law, may be a matter of luck.
19 One survey showed that 6.8% of drivers with over 5 years™ experience were involved in single-
vehicle accidents (i.e. those least likely to be the result of the fault of anybody else), while 19.4%
of drivers with less than 6 months™ experience were involved in such accidents: Ministry of
Transport, How Fast? (HMSO, 1968), para. 77. See also TRRL Laboratory Report 567 (1973). It
has been estimated that motor cyclists have twice as many accidents in their ¬rst 6 months as
drivers as they cause in their second 6 months: Austin, Accident-Black Spot, 60.
20 E.g. Adamson v. Motor Vehicle Insurance Trust (1957) 58 WALR 56; Roberts v. Ramsbottom [1980]
1 WLR 823.
An appraisal of the fault principle 183

here that this is one further di¬culty in the way of equating legal liability with
moral fault.

7.3.4 The fault principle and popular morality
Criticisms of tort law based on its divergence from moral principles assume that
there is a clear distinction between law and morality, and that morality provides the
proper standard for judging the law. However, there is a certain amount of empir-
ical evidence that casts doubt on these assumptions. Research21 has shown that per-
sonal injury victims do not always think that being responsible for an accident
entails moral culpability, or that either of these entails an obligation to pay com-
pensation; nor, conversely, that absence of moral responsibility entails absence of
an obligation to pay damages. It seems that whether an injured person thinks that
someone else should pay depends much more on what he or she knows (or is told)
of what the law says about when compensation is payable than on independent
ideas of morality and fault. So, for example, in the industrial context where employ-
ers have for a long time been subject to liability in certain circumstances regardless
of fault, injured workers are quite likely to think that their employer ought to pay,
without basing that judgment on an attribution of fault.
In relation to accidents in the home, it seems that injured persons are very
unlikely to attribute fault and even less likely to think that they ought to be com-
pensated. It is arguable that this has little to do with morality and more to do with
a desire not to disrupt harmonious domestic relations by the aggressive act of liti-
gating; and perhaps with a realization that since the party responsible will rarely be
insured, litigation would be pointless. In brief, it may well be that for many victims
of personal injury, thinking that someone is (morally) responsible for injuries is
neither a su¬cient nor a necessary condition for thinking that the person ought to
pay compensation for the injuries.
It does not follow from this that the fault principle is not defensible as a princi-
ple, but only that it may not be defensible on the ground that it is based on popular
conceptions of who ought to pay.

7.4 The fault principle pays little attention to the conduct or
needs of the victim
If no-one can be identi¬ed as being to some extent at fault for the victim™s injury
or loss, no compensation will be available under the fault principle. In this event, it
is immaterial whether the claimant was injured while in a drunken stupor; was
driving a car with a slight degree of negligence; was indulging in some perfectly

21 S. Lloyd-Bostock, ˜Fault and Liability for Accidents: the Accident Victim™s Perspective™ in Harris
1984 Survey. See also by the same author, ˜Propensity to Sue in England and the United States of
America: The Role of Attribution Processes™ (1991) 18 J. of Law and Society 428 in answer to H.M.
Kritzer, ˜Propensity to Sue in England and the United States of America: Blaming and Claiming
in Tort Cases™, ibid., 400.
184 Chapter 7

ordinary activity with all due care; or was engaged in an heroic attempt at rescuing
someone in great peril, at risk to their own life. On the other hand, where, for
instance, a person has been killed in an heroic attempt to rescue another, judges will
make every e¬ort to ¬nd someone at fault if they possibly can. Conversely, a person
injured as a result of a piece of utter folly on their own part would ¬nd a judge
somewhat unreceptive to the argument that another person was partly responsible
for their injuries.22 But when courts take account of the victim™s conduct in this way,
they often do so in spite of the fault principle and not because of it.
It is also true, of course, that where a negligent party can be found, the law
does pay some attention to fault on the part of the injured person. We have seen
before how the treatment of contributory negligence as a relative doctrine requi-
res comparison of the injured person™s degree of fault with the injurer™s; and how
this leads to results which, however justi¬able as between injured and injurer,
appear indefensible in a wider context. Though it may be justi¬able in terms of the
fault principle to refuse compensation to a wholly innocent victim because they
were injured without fault on the part of anyone else, while giving part compen-
sation to another victim, who may have been 80% to blame for their own injuries,
because someone else was partly to blame, it seems quite inequitable in light of the
fact that in most cases of fault it is not the injurer who will pay but an insurer and,
ultimately, the public at large. Why should fault on the part of the injurer be a pre-
condition of an award of compensation when it is not the injurer who will pay the
Popular sentiment is much more sensitive than the law to the perceived merits
of victims of injury or disability. We are generous to rescuers regardless of whether
the rescue was precipitated by someone™s fault. We are much more prepared to help
people who su¬er injury through no fault of their own than people who bring mis-
fortune upon themselves, regardless of whether anyone else was to blame. It
would, in theory, be possible to construct a compensation system based on the
fault of the injured person alone, if that were felt to be desirable. A claim against a
compensation fund could be admitted wherever a person was injured without
fault on their part, even if no other person at fault could be found; and reduced
compensation could be given to an injured person who was partly to blame, if that
was felt to be just. This would also be a ˜fault™ system, and perhaps a more fair
system (though it would be open to similar objections as the present doctrine of
contributory negligence).23 But it is not the fault system we have now.
In addition to paying little attention to the injured person™s conduct, the fault
principle largely ignores the injured person™s needs “ just as it ignores the injurer™s
capacity to pay. Here again, this may be just as between injured and injurer, although
even this is open to doubt in some cases. For example, it is by no means obvious that
justice requires a working person to provide an annuity for a young, healthy and

22 E.g. ICI v. Shatwell [1965] AC 656.
23 2.5.1.
An appraisal of the fault principle 185

childless widow whose husband they have killed (say) by negligent driving.24 But
even if this is thought to be just, our opinion might change if it was not the injurer
personally who was to pay the damages. As between the widow and the public, it
seems di¬cult to say that justice requires compensation for the death of the husband
without considering the widow™s needs.

7.5 Justice may require payment of compensation without fault
Neither in law nor in morality is fault the only ground on which a person may be
required to compensate another. For example, where a person has been overpaid
by mistake, both morality and the law (of restitution, not tort) say that the over-
payment should be returned. No question of fault arises in this sort of case, but we
require repayment because otherwise the recipient would be ˜unjustly enriched™ at
the expense of the payer. This sort of argument for liability without fault may at
¬rst sight appear to have little place in the ¬eld of accidental damage to person or
property, because in such circumstances there is rarely any ˜gain™ in an obvious or
tangible sense to the harm-doer. But this depends on what we mean by ˜gain™. Let
us look for a moment at a leading US case in the law of torts, Vincent v. Lake Erie
Transportation Co.,25 which illustrates the struggle between the principle of no-
liability-without-fault and other grounds for requiring compensation to be paid.
In this case the claimants were the owners of a dock in which the defendant
shipowner™s vessel was anchored. A storm was threatening and both parties were
anxious for the safety of their property. The defendant was requested to remove the
ship, but the shipowner declined to do so for fear that damage would be done to it
in the storm. The result was that the vessel remained at anchor, and in the ensuing
storm it was repeatedly hurled against the dock, and the dock was damaged. The
question was whether the claimants were entitled to compensation for the damage.
The court conceded that perhaps the shipowner would not be considered morally
culpable or blameworthy. A person whose property is in jeopardy may not be
thought of as acting wrongly by trying save it, even at the expense of creating a risk
of loss to someone else. However, the court said in Vincent, even though we might
not morally blame a person who chooses to save their own property at the expense
of risk to someone else™s property, we might nevertheless think the former ought to
pay for the privilege thereby enjoyed. Even if life and not just property had been at
stake, would fairness not require compensation to be paid?
The idea that compensation ought to be paid even though the person paying it
is not morally culpable is well established in public law contexts. For example, if a
government authority compulsorily acquires private land in order to build a road
or a school, few would regard this as in any way morally reprehensible (assuming,
for the sake of the argument, that the power to acquire has been exercised wisely and

24 6.1.2.
25 (1910) 124 NW 221.
186 Chapter 7

reasonably), but most people would think it only fair that the deprived landowner
should be paid compensation for loss of the land. The payment of such compensa-
tion is usually provided for by statute. The idea that it can be fair to require com-
pensation to be paid even in the absence of wrongful conduct plays very little part
in private law. Even the principle of Vincent would probably not be applied in
England.26 But why should it not be applied?
Admittedly, in cases of negligence we cannot usually say that the injurer has been
unjustly enriched at the expense of the injured.27 But we may be able to say, as in
Vincent, that the injurer has ˜gained™, or furthered their own ends, by taking a risk
at the expense of the injured person. For example, in Bolton v. Stone28 the House of
Lords held that the cricket club had not acted unreasonably in not building a higher
fence around their ground, because the risk of a ball escaping and injuring someone
was very remote. Thus the club bene¬ted by not having to spend money on a higher
fence, but at the cost of a risk of injury to the claimant and others. In such circum-
stances one might think that it would be fair for the club to pay compensation to
the person injured by the escaping ball, even though their actions were not negli-
gent. Or take the case of the installation by British Railways of automatic half-
barrier level crossings, which was the subject of investigation at the Public Inquiry
into the Hixon crossing rail crash.29 The inquiry found that these half barriers were
not as safe as the supervised gates they replaced.30 The reason for installing the new
barriers, despite this ¬nding, was that they saved time and money. British Railways
estimated that they would save £2 million a year by installing automatic barriers at
all level crossings; and such barriers were much quicker in their operation than the
old ones, thus reducing delays to people using the crossings. So, the installation of
the barriers bene¬ted certain people but created risks of injury for others. If the
bene¬ts are thought to outweigh the disadvantages, it seems only fair that those
who bene¬t should compensate those who are injured, regardless of whether instal-
lation of the barriers was in some sense negligent or blameworthy.
The point becomes even clearer when the impact of insurance is taken into
account. In a situation such as that in Bolton v. Stone, for example, the law would not
require the club to build a fence even if it was held liable for escaping balls. It would
be open to the club simply to insure against the risk of liability and, except in cases
where the risk of balls escaping was quite high,31 this course would usually be
cheaper than building a new fence. Once the issue is reduced to terms of who should
insure against a particular risk, it seems clear that it might be fair for a person to do
so even if that person was not at fault in creating the risk.

26 See further P. Cane, Tort Law and Economic Interests, 2nd edn (Oxford, 1996), 224“7.
27 But see ibid., 324.
28 [1951] AC 850.
29 Cmnd 3706 (1968). See also 2.4.2.
30 Later research found otherwise: Level Crossing Protection (HMSO, 1978), para. 10.10. See also
Railway Safety 1983 (HMSO, 1984), para. 50; Railway Safety 1991/2 (HMSO), table 6.
31 As, perhaps, in Miller v. Jackson [1977] QB 966.
An appraisal of the fault principle 187

7.6 It is often dif¬cult to adjudicate allegations of fault
So far we have considered suggested ethical objections to the fault principle. Its
value can also be doubted on the basis of the practical di¬culties to which adjudi-
cation on fault gives rise. There are three distinguishable problems. The ¬rst arises
out of the nature of the legal test of fault; the second out of problems of proof; and
the third from concentrating too much on one speci¬c cause to the exclusion of sta-
tistical and other evidence about accidents of the kind in question.
Looking at the ¬rst problem, the essence of the legal concept of fault is unrea-
sonable failure to take precautions. This concept is both abstract and fact-
dependent. As a result, it may be di¬cult for a person to determine what they must
do in order to meet the standard of reasonableness. This has serious implications
for the utility of the concept of fault as a guide to conduct.32 The nature of the neg-
ligence test may also a¬ect the settlement of tort claims. There is, for instance, evi-
dence in relation to medical injuries that claims may succeed in the absence of
negligence and fail despite its presence.33
Turning to the second problem, in the case of many accidents, the events that
cause the injury occur in a very brief period of time, often in a fraction of a second.
Adjudication on the fault issue requires witnesses to be able accurately to recall what
occurred in that fraction of a second if we are to have any con¬dence that the
¬ndings of fact made by a court correspond with what actually happened. Similarly,
if a case is settled by negotiation, the parties™ advisers need to be able to assess with
reasonable con¬dence the likelihood that a court will ¬nd fault on the basis of the
evidence of the witnesses. The unreliability of observations of eyewitnesses (even
highly trained and experienced ones) has often been demonstrated by experiment.
To the inaccuracies of observation must be added the di¬culties of recall produced
by the considerable period that often elapses between the time of the accident and
the time when witnesses are asked to give an account of what happened; and also
the fact that people do not always tell the truth. If, as a result of such defects, the
version of some witnesses con¬‚icts with that of others, what chance does the court
have of reaching a correct conclusion? In addition to all this, we must not forget that
in a not-insubstantial number of cases, suitable evidence is simply unavailable at all,
and ˜real™ evidence (i.e. objects) may disappear.34 Di¬erent, but equally di¬cult,
problems may arise in proving fault in cases of illness or disease as opposed to trau-
matic accident.35
Other serious problems arise out of the need to prove that the injurer™s fault
caused the victim™s loss. This may be di¬cult in the case of accidents that happen

32 See further
33 M.M. Mello and T.A. Brennan, ˜Deterrence of Medical Errors: Theory and Evidence for
Malpractice Reform™ (2002) 80 Texas LR 1595, 1618“20.
34 The courts can make orders to facilitate the preservation of evidence (see Civil Procedure Act
1997, s. 7), but the time taken to obtain such an order may deprive the power of much practical
use in many cases.
35 J. Stapleton, Disease and the Compensation Debate (Oxford, 1986), ch. 4.
188 Chapter 7

in a moment, for all the reasons we have just mentioned. Problems of proving cau-
sation also arise acutely in cases of medical negligence, for instance, and in cases
where it is alleged that a person has become ill or contracted some disease as a result
of exposure to some chemical or the taking of some drug.36 Our knowledge of the
way many diseases and illnesses come about is inadequate, and this fact may present
an impenetrable barrier to much tort litigation. Of course, this problem is not
unique to the fault-based tort system, but is shared by any compensation system in
which entitlement depends on proving the cause of loss. But all of these problems
of proof do suggest that fault, at least, is not a satisfactory criterion of entitlement
to compensation. And if this is true of cases tried by a court, how much more is it
likely to be true of cases settled by negotiation. Insurance companies frequently
have only the witnesses™ statements to go by, and one experienced senior barrister
once said that ˜more often than not™ the evidence a witness gives in court di¬ers sub-
stantially from that in pre-trial statements.37 Moreover, such di¬culties of proof
mean that the process of deciding what caused an accident and who was at fault is
extremely expensive and time-consuming in many cases. Several surveys have
found that di¬culties of proof are one of the major reasons why people either
abandon tort claims or do not make them in the ¬rst place.38
Concerning the third problem, it may be true, despite the above di¬culties, that
if we concentrate exclusively on the behaviour of the principal parties involved in
an accident, we can, in a reasonable proportion of cases, arrive at a workable con-
clusion on fault and causation. However, this exercise may often be misleading
because it omits to take account of factors that would not always, or indeed often,
be thought to be responsible for accidents. Statistical investigation of the causes of
accidents generally, as opposed to legal investigation of the causes of individual
accidents, often throws an entirely di¬erent light on matters. The point is well put
in the following extract from a volume on road safety published in 1963:39
The statistician does not think so much of the individual accident and its causes, but of
the probability of accidents and whatever may a¬ect this probability. Now such things
as the width of a street, its curvature or gradient, the quality of its surface, the ¬‚ow of
tra¬c and its speed, all in¬‚uence the probability of an accident in a street. Such things,
since they in¬‚uence the probability of accidents and therefore the number of accidents,
should appear in the statistical picture of factors important in accident causation . . .

36 Ibid., ch. 3.
37 C.P. Harvey, The Advocate™s Devil (London, 1958), 67.
38 E.g. Harris 1984 Survey, tables 2.12, 3.12; S.B. Burman, H.G. Genn and J. Lyons, ˜The Use of Legal
Services by Victims of Accidents in the Home: A Pilot Study™ (1977) 40 Modern LR 47, 57.
39 Research on Road Safety (HMSO, 1963), 3“4; see also J.J. Leeming, Road Accidents: Prevent or
Punish? (London, 1969); J. Reason, Human Error (Cambridge, 1990), ch. 7; J. Mosedale, A. Purdy
and E. Clarkson, Contributory Factors to Road Accidents (Department of Transport, 2004). An
interesting example of this phenomenon is the road accident in which Princess Diana died. At
¬rst, this was attributed to the conduct of the driver of the car and of certain paparazzi. However,
it was later pointed out that a major factor contributing to the seriousness of the accident was the
fact that the central columns in the tunnel were unprotected by any sort of guard rail.
An appraisal of the fault principle 189

When individual accidents are studied and ˜causes™ sought it is not, in general, these
factors that will be cited. Then only the unusual or abnormal are usually noticed: not
the width of the road but only whether it narrows suddenly, not the visibility allowed
by the size and shape of the car™s windows but only the obstruction caused by pennants
or a dangling doll. Ignoring the normal gives rise to a tendency to ascribe most accidents
to human factors such as error or carelessness, since it is usually possible to believe that
there would have been no such accident if someone had acted di¬erently.

This is not to say that driver ˜error™ is never the cause of road accidents. Indeed, the
Pearson Commission noted research which suggested that as many as 65% of road
accidents were the result of human error alone.40 Nor do statistics by themselves
prove what caused any particular accident. On the other hand, statistical informa-
tion about the causes of particular types of accidents may alert us to factors, other
than the conduct of those involved in the accident, which might have caused or
contributed to it. The more we appreciate the signi¬cance of factors other than
driver behaviour in the cause of road accidents, the less does the almost exclusive
concentration of the tort system on driver conduct make sense.
For example, as a result of research it is now known that skidding accidents can
be greatly reduced by altering the surface of the roads. One survey compared ¬fty-
¬ve skidding accident sites with an average length of a quarter of a mile before and
after the sites were treated with a non-skid surface.41 The treatment produced a dra-
matic reduction in the number of accidents at those sites. Who, then, was primar-
ily to blame for the accidents that occurred before the sites were treated?42 And
suppose that lack of funds had held up treatment of other sites and skidding acci-
dents had continued to occur there, who would be to blame for them? Suppose that
a local authority prefers to spend its income on building a new school rather than
treating skid-prone sites, who would be to blame for the accidents that would
inevitably occur? Or indeed, suppose that it is simply not appreciated that the road
surface is contributing to accidents so substantially. Consider also the case of the
motorist who tries to reduce speed on approaching a roundabout, skids and crashes
into a bollard in the centre of the road. If this case ever came into court the motorist
would almost certainly be found entirely responsible for causing the accident by
negligence. Yet this sort of accident is so common at some roundabouts that it has
been found cheaper to treat the road with a non-skid surface than to replace the
bollards every time they are damaged. If this is not done, who is more at fault, the
motorist or the highway authority?

40 Pearson Report, vol. 2, table 42. Similarly, it has been said that driver behaviour (as opposed to


. 7
( 18)