. 3
( 18)


gion? Rich, poor or averagely a¬„uent? Perhaps none of these di¬erences between
people is relevant, for instance, to questions about how a reasonable person would
drive a car, but some or all of them may be thought to be relevant in some contexts.
To complicate matters further, what the law says is that people should behave in the
way that a reasonable person in their position would behave. This formulation allows
the court to invest the reasonable person with characteristics that the defendant
actually possesses. For example, in an action against a surgeon, the standard of rea-
sonableness is that of the reasonable surgeon, not the reasonable GP. However, it is
largely up to the court to decide which of the characteristics of the defendant to
38 Chapter 2

attribute to the reasonable person. In fact, the reasonable person is an abstraction
whose characteristics are invented by the judges. Lord Radcli¬e once said that the
reasonable person was ˜the anthropomorphic conception of justice™;6 which, in the
context of the tort of negligence, means that a person is negligent if they fail to take
that degree of care which justice requires should be taken. However, this tells us
nothing about the characteristics of the reasonable person except that they are a
function of ideas of justice and sound behaviour about which people (and judges)
might justi¬ably disagree.
Would it help to equate the reasonable person with the ˜average person™? A sta-
tistician might reject the idea that there is such a person as the ˜average person™;7 but
in some contexts, at least, it may be possible to ¬nd out how people generally behave
or react to given situations. Suppose, for instance, that an employee is injured by a
machine in a factory, and alleges that it was negligent of the employer to use the
machine without taking certain precautions. Is it any help to ¬nd out whether other
employers do the same? Suppose we ¬nd that the great majority of employers who
use this machine also do not take the allegedly required precautions. Is this not evi-
dence that at least the ˜average employer™ would not regard it as necessary? The
(legal) answer is that it is, and the courts give considerable weight to the practice of
employers in this respect. The courts also give considerable weight to the practice
of professional and business people in arriving at decisions on questions of negli-
gence. To some extent decisions of this kind are probably based on the feeling that
a person should not be blamed for doing what everybody else does, and a court may
be reluctant to consider the issue of whether the common practice was itself negli-
gent. But to some extent also, a decision of this kind may be based on acceptance
by the court of the standards of the community, so that although a judge may per-
sonally feel that the common practice is unsatisfactory, he or she may subordinate
that view to the practice of the community. In this way, the concept of the average
person may be of assistance to judges in helping to set the standard of care in par-
ticular circumstances.
This does not mean that negligence consists simply of the failure to observe
normal or usual precautions in a given situation or, conversely, that observance of
normal or usual precautions cannot amount to negligence. The courts have never
accepted that they are precluded from ¬nding negligence even in the face of wide-
spread and long-standing practice.8 If it were alleged that a driver was negligent in
driving across a road junction without stopping, it would not help much to prove
that 90% of drivers did the same. On the other hand, where professional negligence
is alleged, for example, against a doctor, the fact that the defendant observed the
care or precautions customarily practised by a body of professional colleagues con-
sidered by the court to be reputable, will usually lead a court to hold that there has

6 Davis Contractors v. Fareham UDC [1950] AC 696, 728.
7 R. Powell, ˜The Unreasonableness of the Reasonable Man™ (1957) 10 Current Legal Problems 104.
8 Morris v. West Hartlepool Steam Co. [1956] AC 552; Cavanagh v. Ulster Weaving Co. [1960]
AC 145.
Fault as a basis of liability 39

been no negligence, even if a majority of members of the relevant professional
group would have acted di¬erently.9 The e¬ect of this latter approach is to give pro-
fessionals, especially doctors, considerable protection from negligence liability.
These examples show that at the end of the day the concept of reasonable conduct
depends more on value judgments by courts than on observations about what
people generally do.
Anyway, given the many di¬erences in outlook and behaviour between people
who live in our pluralistic society, the notion of the average person may, in many
contexts, be very di¬cult to invest with any concrete meaning. There may be no
relevant common practice that can be treated as reasonable behaviour, and people
may genuinely disagree about what would constitute reasonable behaviour in
particular circumstances. The standard of behaviour against which the defendant™s
conduct is measured is a standard decided on, and inevitably decided on, by
Why does the law continue to utilize the largely ¬ctional ¬gure of the reasonable
person? The answer appears to be, in order to obscure the role of the judge as
policy-maker. Judges in this country have traditionally eschewed the role of policy-
maker: they continue to proclaim that they are not concerned with policy but only
with law, and it is possible that the public prefers it this way. For many people,
˜impartial justice™ means justice without policy. If a judge were to say to a defen-
dant: ˜You have failed to do what I think you should have done and that amounts
to negligence™, the defendant may come away thinking of the judge, ˜Who are you
to tell me that?™ But if the judge says: ˜You have failed to do what the reasonable
person would have done, and that amounts to negligence™, the defendant may come
away with more respect for the judge and the law.
On the other hand, in recent decades many people have become more aware of
the fact that judges in Britain are overwhelmingly white, wealthy, male, middle-
aged or elderly, from highly or relatively privileged backgrounds, and Judaeo-
Christian in upbringing or outlook even if not actively ˜religious™. If it is correct to
say that it is judges who decide what conduct is reasonable and what is not, then it
may be hard to avoid at least a suspicion that the law embodies standards of rea-
sonableness that may not re¬‚ect the views and expectations of many members of
our society. This inevitably raises the further question of why such people, who are
neither popularly elected nor democratically accountable, should be allowed to
force their standards of justice and reasonableness on the rest of us. This is a
di¬cult question that will not be addressed here. Whatever the answer, the ques-
tion may lead us to ask, further, whether the issue of what sorts of conduct ought
to entail liability to compensate for injury and damage should not be decided by
the legislature rather than the courts.10

9 Sidaway v. Governors of Bethlem Royal Hospital [1985] AC 871; Bolitho v. City and Hackney HA
[1998] AC 232.
10 See generally P. Cane,˜Taking Disagreement Seriously: Courts, Legislatures and the Reform of Tort
Law™ (2005) 25 OJLS 393.
40 Chapter 2

2.4.2 The nature of negligence
One way of summarizing the points made in the previous section is to say that the
issue of negligence or reasonable conduct is an issue of ˜policy™ which is little
di¬erent from the sorts of policy decisions which public o¬cials are continuously
having to make. The judge™s decision is often a far less momentous decision because
the majority of litigated cases involve a microscopic analysis of conduct at a par-
ticular moment of time. But this is not always so, and there are occasions on which
a judge may have to decide whether action taken ˜on grounds of public policy™ or
˜in the public interest™ was ˜reasonable™ “ i.e. whether the judge agrees with the
policy-makers. For instance, in the 1960s British Railways and the Ministry of
Transport spent a great deal of time debating the desirability of replacing manually
operated level crossings with automatically operated barriers. They took into
account the ¬nancial savings that would result and the risk of possible accidents,
and they decided on balance that the changeover was desirable. Subsequently an
accident occurred at one of these crossings, which formed the subject of a public
inquiry by a QC.11 Had legal proceedings been brought, the judge would have had
to determine a similar question to that which the railway authorities and the
Ministry had already determined, though the question would have been couched
in the language of ˜negligence™.
In more recent years, major disasters such as the sinking of the Herald of Free
Enterprise in Zeebrugge harbour, the ¬re in the King™s Cross underground station,
and the destruction of the Piper Alpha oil platform, have been followed both by public
inquiries, which have investigated the causes of these tragedies and made recom-
mendations as to how similar accidents might be avoided in the future; and also by
tort claims brought by persons injured and relatives of those killed in these disasters.
Such tort claims inevitably raise important policy issues about the desirable balance
between safety and the commercial interests of entrepreneurs, and such inquiries
often lead to the imposition of stricter safety standards by legislation or other gov-
ernment action. Tort actions in which claimants allege that they have been injured as
a result of a defect in the design of a product may also raise fundamental policy issues
about the proper balance between the safety of consumers, the pro¬tability of man-
ufacturing industry and the desirability of innovation in product design.
However, even though the issue of negligence is sometimes essentially an issue
of ˜desirability in the public interest™, it does not follow that in deciding a case a
judge will take into account all those factors that a public o¬cial or a Minister
might take into account. A Minister may, for example, decide that a certain course
of action is likely to win votes for his or her party, and that this consideration out-
weighs all others; a judge would not be in¬‚uenced by such a consideration. In fact
a judge is likely to take into account (whether explicitly or silently, consciously or
subliminally) four main considerations: ¬rst, the degree of probability, judged as at

11 Report of the Public Inquiry into the Accident at Hixon Level Crossing (Cmnd 3706, 1968).
Fault as a basis of liability 41

a point in time just before the accident occurred, that damage would result from
the conduct which is complained of; secondly, the magnitude of the harm (once
again judged as at a point in time just before the accident occurred) which was likely
to result if the conduct complained of took place; thirdly, the value or utility of the
object to be achieved by the conduct in question; and, fourthly, the burden in terms
of cost, time and trouble, of taking those precautions against the risk of damage
which the claimant alleges ought to have been taken.
A famous US judge, Learned Hand, once declared that negligence was a function
of three variables;12 on this view, negligence is shown where the burden of the pre-
cautions needed to avoid a risk of injury or damage occurring as a result of particu-
lar conduct is less than the product of the likely magnitude of the damage and the
probability that the damage will occur. It has been argued that this is fundamentally
an economic test,13 as opposed to the essentially moral notion of fault embodied in
the concept of reasonableness. If it can be shown that the expenditure of £X on
avoiding or minimizing the risk of an accident will prevent accident costs of £X + Y,
then it is clearly desirable that the £X should be spent. On the other hand, it is said,
there is no point in spending £X to prevent accident costs which are less than £X. In
some situations this is a useful perspective, and sometimes (although by no means
always) it may even be possible to put actual ¬gures on the probability of an acci-
dent occurring,14 the damage likely to be caused if an accident does happen and the
cost of precautions. But any attempt to reduce the whole law of negligence to the
form of an algebraic equation must be dismissed because we will normally not be
dealing with precisely measurable values. More importantly, how can we place a
value on the object to be achieved? Signi¬cantly, perhaps, Learned Hand did not
speci¬cally mention this factor; and plainly it cannot be reduced to monetary terms
in most instances. Suppose an ambulance driver is taking a seriously injured person
to hospital and is driving faster than usual in order to arrive sooner and so give the
injured person a better chance of survival. How do we put a value on the life of that
person, and how do we compare it with the value of any lives that may be lost in an
accident caused by the ambulance driver™s speeding? Such things can only be the
subject of a delicate personal judgment, and people may well di¬er in making such
The point is even stronger where, as often happens, the court is required to
compare dissimilar things. Suppose a court is asked to say whether it is negligent to
play cricket on a ground without a fence, so that balls are occasionally hit into the
street. We might be able to assess with reasonable accuracy the likelihood of a ball
being hit out of the ground and injuring someone, the likely severity of those

12 US v. Carroll Towing Co. (1947) 159 F 2d 169.
13 R. Posner, Economic Analysis of Law (4th edn, Boston, 1992), 163¬.
14 One of the easiest risks to quantify is death from various causes. It is also relatively easy to quan-
tify the risk of being injured in a road accident, for example; but determining the risk of su¬ering
speci¬ed injuries or being involved in a particular type of accident may be very di¬cult. Much
will depend on whether reliable statistics are available.
42 Chapter 2

injuries and the cost of taking precautions against this happening. But how can we
compare the value of cricket as a pastime with the value of the safety of passers-by,
and so decide whether playing cricket without a fence was negligent?
At this stage it might be thought that we reach the realm of purely subjective
judgment; but it is often said that what judges ought to do at this point is to attempt
to discern and give e¬ect to community values. If a judge believes that the com-
munity has a high regard for the game of cricket, it is right, so the argument goes,
to give it a high value when weighing it against the possibility of personal injury.
The obvious weakness of this approach lies in the notion of ˜community values™. In
a pluralistic society, although there may be a core of issues on which many people
hold similar views, there will be many more about which a variety of di¬ering and
more or less inconsistent views can be and are legitimately held. The value of the
game of cricket is, perhaps, a good example. Furthermore, there may be occasions
on which judges feel very strongly that widely held values are wrong or misguided,
and there is nothing to stop a judge from trying to change such values by applying
his or her own. In this way the courts may seek to mould opinion, and change the
community™s sense of values. However, as we have already observed, the legitimacy
of such behaviour is questionable given the make-up of the judiciary and the
unelected and unaccountable nature of the o¬ce of judge.
In making a ¬nding of negligence the courts do not generally rely on factual or
statistical or expert knowledge, at least where the facts do not clearly fall within the
realm of scienti¬c knowledge. Where allegations of professional negligence are
made, for example, against a doctor, the courts do rely on expert witnesses to tell
them what is accepted practice and what is not. But in many areas judges rely almost
entirely on their own experience, hunch or instinct. Thus the probability of an
event is almost invariably15 decided without the assistance of statistical evidence;
and the assessment of the amount of damage likely to be done and the burden of
precautions is rarely reduced to arithmetical calculations. Even in cases in which
hard facts could be adduced and measured, courts seem to discourage the use of
empirical evidence. What is known in the USA as the ˜Brandeis brief ™ is rarely used
in this country. Reluctance to hear empirical evidence is partly the result of a desire
not to prolong and complicate trials, and partly of a realization that the ultimate
issue in a negligence action is not a factual one but requires a value judgment.16

2.4.3 Probability of harm
As we have just seen, four factors are taken into account in deciding whether a person
has taken reasonable care: the probability that the claimant would su¬er harm and

15 Haley v. London Electricity Board [1965] AC 778 is an exception; but even here the evidence used
(about the number of blind people in London) was only indirectly relevant to the issue to be
decided, namely whether the Board ought to have guarded against the possibility of a blind person
falling into an excavation in the pavement.
16 For a discussion of the issues in an Australian context see K. Burns,˜It™s Just Not Cricket: The High
Court, Sport and Legislative Facts™ (2002) 10 Torts LJ 234, esp. 247“54.
Fault as a basis of liability 43

the likely magnitude of that harm, the cost of taking precautions to prevent it and
the value of the activity which caused the harm. We will consider each of these factors
in turn. First, probability. The most important thing to note here is that probability
is relative. Some events are so probable that there would be no point in conducting
one™s life except on the assumption that the event will take place “ for example, that
the earth will continue to spin on its axis in its accustomed orbit round the sun.17
Other events may be such remote possibilities that nobody would adjust their
conduct because of them “ for instance, that a major earthquake will occur in Britain.
In between these extremes there is an in¬nite number of gradations, and these
degrees of probability are re¬‚ected in the language of the courts. Events may be
described as ˜very probable™, ˜highly probable™, ˜quite likely™, ˜not unlikely™; events may
be described, after they have happened, as ˜remarkable™ or ˜extraordinary™; risks may,
before the event, be stigmatized as ˜remote™ or ˜fantastic™ possibilities.
There is no ¬xed point at which the law requires people to take account of a pos-
sibility. The point is a moving one because negligence is a function of several vari-
ables. In other words, it may be negligent to disregard a very remote risk in one
situation, but not negligent to disregard a much greater risk in another situation. We
must consider alongside probability the other factors mentioned above “ the utility
of the conduct in question, the magnitude of the damage that may be done; and the
burden of the precautions required to avoid the damage. In Bolton v. Stone, for
example, the possibility that a cricket ball might be hit out of the ground and injure
someone in the street had to be set against the fact that the chance of this happening
was quite small; that the amount of damage, if any, which the ball was likely to do
was limited; and that cricket could not be played without creating such a risk except
at the cost of building a high and costly fence. Taking all these factors into account,
the balance was held to be in favour of no liability.18 On the other hand, The Wagon
Mound involved the very remote chance that oil accidentally discharged from a vessel
in a harbour might ignite on the surface of the water. Set against this the fact that, if
it did ignite, very considerable damage could be done, the fact that a discharge of oil
serves no useful purpose, and that the only precaution required to avoid it was the
turning o¬ of a tap, and the balance was held to be in favour of liability.19

2.4.4 Likely magnitude of harm
This is the second factor taken into account in deciding a negligence issue, and in
some cases it may tip the scales. For example, in Paris v. Stepney Borough Council20
it was held that an employer who knows that one of its workers has only one sound
eye may be negligent if it fails to supply the worker with goggles for work involving
a risk to the eyes, even though the risk is su¬ciently remote that the employer

17 The distinguished mathematician, G.H. Hardy, once wagered ˜his fortune till death™ to a halfpenny
that the sun would rise on the following day.
18 See Bolton v. Stone [1951] AC 850.
19 The Wagon Mound (No. 2) [1967] 1 AC 617.
20 [1951] AC 367.
44 Chapter 2

would be justi¬ed in disregarding it in the case of a normally sighted worker.
Similarly, it has been held that organisers of sporting events must take greater care
for the safety of disabled than of able-bodied participants.21
There is some empirical evidence to suggest that people do take into account the
magnitude of the damage which their conduct may cause in determining the degree
of care with which they will perform some task. In an experiment conducted by the
Road Research Laboratory in 196122 a group of motorists was asked to drive
through three narrow gateways the pillars of which appeared to be made of plastic,
wood and concrete respectively. It was found that they drove at the lowest speed
and with the greatest care when the pillars appeared to be concrete, at the highest
speed and with least care when the pillars appeared to be plastic. Unfortunately
there was insu¬cient evidence to show whether lower speeds and greater care actu-
ally made any material di¬erence to the accident rate.
In order to determine the magnitude of the harm it may be necessary to make
judgments based on community values. Is the death of one individual a harm of
˜greater magnitude™ than the su¬ering of injury by a dozen others; or than the phys-
ical destruction of thousands of pounds™ worth of property? Can the death of one
individual ever be regarded as harm of greater magnitude than that of another or
are all people equal for this purpose? Such questions are rarely discussed openly by
courts or lawyers; and it may very well be that society™s sense of values on such
matters is not always wholly rational. As a society we tolerate many risks that could
be avoided at manageable cost. Cars, for example, could be made safer if people were
prepared to pay more for them;23 roads could certainly be much safer if more money
was spent on them, or if cars travelled at slower speeds. Many accidents could be
avoided every year, but society is evidently not prepared to pay the cost of doing so.
By contrast, if an individual is in actual and immediate peril, society may be pre-
pared to devote very considerable resources indeed to saving that person™s life. For
example, the amount we are prepared to spend to save the lives of miners trapped
underground is much greater, per life saved, than we are prepared to spend to avoid
accidents in coal mines. Although it is hard to ¬nd any rational basis for this,24 it is
also hard to dissent from the way in which society reacts to these situations.

21 Morrell v. Owen, The Times, 12 December 1993.
22 Research on Road Safety (HMSO, 1963), 89“90.
23 People™s attitudes to the risks of di¬erent types of transport may not correlate with the degree of
danger involved. For example, many people have a much greater fear of ¬‚ying than of motoring,
even though ¬‚ying is statistically much safer than motoring. ˜Psychologists have identi¬ed . . . a
tendency for people to judge a particular type of event as more likely to happen if particular
instances of the event are easy to remember or, for whatever reason, “come to mind” more readily.
Thus people tend to overestimate the frequency of widely reported causes of death™: M.W. Jones-
Lee, ˜The Value of Transport Safety™ (1990) 6 Oxford Review of Economic Policy 39, 53 n. 22. Such
attitudes may a¬ect the amount we are prepared to spend (or require to be spent) on reducing the
risks involved in various modes of transport.
24 An attempt is made by C. Fried, ˜The Value of Life™ (1969) 82 Harvard Law Review 1415. One
answer is that the time available in such circumstances to e¬ect a rescue puts a limit on the amount
which can be expended. It is sometimes suggested that people who carelessly get themselves into
situations from which they need to be rescued should pay the cost of the rescue.
Fault as a basis of liability 45

2.4.5 The value of the activity and the cost of the precautions needed to
avoid harm
The negligence formula requires the harm caused by the allegedly tortious conduct
to be weighed against the cost of the precautions which it is argued ought to have
been taken and the social bene¬ts ¬‚owing from the conduct. The more valuable an
activity, the more uncompensated harm we might be prepared to accept as a cost
of the continuance of that activity. So, for instance, in Bolton v. Stone the court (in
e¬ect) had to decide whether the playing of cricket on the ground in question was
worth the risk of injury to passers-by in the street, given the probability and mag-
nitude of the risk. Suppose that it would have cost so much to take precautions
su¬cient to remove the risk of injury to passers-by that people would not have been
prepared to pay the entrance charges or membership fees needed to ¬nance it, and
so the cricket club would have had to close. In that case, the court would have had
to decide how important the cricket club™s activities were relative to the interest of
passers-by in personal safety. In fact, the court did not address this issue directly,
but found in the club™s favour simply on the basis that the risk was very small.
However, implicit in this decision was a value judgment that the playing of cricket
under the conditions in question was worth the small element of risk to passers-by
which it created.
In many cases, the taking of suitable precautions will not threaten the very con-
tinued existence of an activity, but may reduce the value of the activity. The speed
at which motor vehicles are driven provides an example. Driving more slowly may
reduce the risk of accidents and injury, but it may also reduce the value of driving
as an activity. A court, confronted with the need to decide whether driving at a par-
ticular speed in particular circumstances was negligent or not, will need (implicitly
at least) to weigh the advantages of higher speed against the greater safety and lower
risk of driving more slowly. For instance, we might be prepared to allow emergency
vehicles to drive at speeds we would not tolerate in other circumstances even
though such speed increases the risk of accident and injury. On the whole, however,
the risks attaching to the use of motor vehicles are great, and the sorts of precau-
tions necessary to prevent the typical motor vehicle accident “ such as driving a
little more slowly, or giving a signal earlier, or sounding a horn, or waiting for the
next straight stretch of road before overtaking “ reduce the value of the activity only
slightly. This is probably why the issue of reasonable care is rarely contested in rela-
tion to road accidents.
Of course, we cannot measure the value of an activity without ¬rst de¬ning it,
and this may not be a straightforward matter. The case of emergency vehicles illus-
trates the point. Transporting a desperately ill person to hospital, sightseeing by car,
driving to work and long-distance trucking are all instances of the activity of using
a motor vehicle. However, because of its special social value, we may be prepared to
pick out the ¬rst and describe it as a di¬erent activity from the other three for the
purposes of tort law. Conversely, we might treat sightseeing and driving to work as
46 Chapter 2

the same activity even though it could be said that the latter is more socially impor-
tant than the former.
In practice, courts rarely decide the issue of negligence by reference to the nature
and value of the defendant™s activity. They tend to focus on the probability and
magnitude of the risk and the cost of precautions. However, decisions about
whether a defendant ought to have taken particular precautions to avoid particu-
lar risks often imply judgments about the value of the defendant™s activity. For
instance, courts are quite unwilling to hold medical practitioners negligent in
respect of the way they treat their patients; and one reason for this approach is the
high value implicitly placed on treatment of the sick. The courts often express a fear
that too readily holding doctors negligent may cause them to practise ˜defensive
medicine™; that is, that they may be led to carry out or not to carry out particular
procedures simply in order to reduce the risk of incurring legal liability and not in
the interests of their patients. In other words, courts are quite willing to hold that
precautions which claimants allege that doctors ought to have taken to avoid harm
are so costly in terms of their e¬ect on styles of doctoring that they would reduce
the value of medical practice to an unacceptable extent. The social bene¬ts of
medical practice are thought to be worth more uncompensated harm than the
social bene¬ts of many other activities.

2.4.6 The function of the negligence formula
The allegation in a negligence action is basically that the defendant paid insu¬cient
attention to the interests of others in deciding how to behave, and has pursued his
or her own objectives at the risk of injuring other people or damaging their prop-
erty. This is perhaps the foundation for the view that negligence is a moral fault.
Whereas the individual looks at the matter primarily from their own point of view,
the judge looks at the matter from the point of view of the public interest and the
need to balance the interests of di¬erent persons. People are entitled to pursue their
own interests and objectives even if by doing so they may endanger other people or
their property to some degree. But there are limits to the extent to which people
may do this, and the judge™s task is to de¬ne those limits with the aid of the negli-
gence formula. People may drive their cars at a ˜reasonable™ speed because the gain
to them and the public from being allowed so to drive is at least worth the risk of
the harm such driving may cause; but people are not allowed to drive at an ˜exces-
sive™ speed because the additional gain that it brings does not outweigh the addi-
tional risk that it imposes on others.
There is another fundamental di¬erence between the judge™s task and that of
anyone else who is called upon to decide between di¬erent courses of conduct, and
that is the purpose for which the decision is to be made. When a person chooses a
course of action, they do so prospectively, before engaging in one or other course
of conduct. By contrast, a court always decides the negligence issue retrospectively:
in a negligence action the judge has to decide what should have been done in the
past, not what should be done in the future. The purpose of this inquiry is to decide
Fault as a basis of liability 47

whether compensation should be paid to those who have su¬ered injury or damage
as a result of the chosen course of action.
What is the justi¬cation for imposing an obligation to pay compensation on the
basis of a judgment that a person should have behaved di¬erently in the past? There
are two main arguments. The ¬rst is that negligent conduct can be stigmatized as
blameworthy and from this it follows that a negligent defendant ought to compen-
sate an innocent claimant for the latter™s injuries. The second argument is that by
holding past conduct to have been negligent and by requiring the negligent party
to pay for loss caused by it, the law might have some deterrent e¬ect on future
behaviour. If, for example, a court were to hold that a cricket club had been negli-
gent in not building a fence around its ground, this might have some in¬‚uence on
the future behaviour of that and other clubs. The assertion that the law can operate
as a deterrent is, in theory, capable of being empirically tested, and if it proved
unfounded then this second justi¬cation would collapse. Each of these arguments
is discussed extensively later in this book (in ch. 17).

2.4.7 Foreseeability
Negligence is a form of fault. To say that someone was at fault in behaving as they
did is to say that they should have behaved di¬erently. To say that a person was neg-
ligent is to say that they should have taken certain precautions (which they did not
take) to prevent harm to another. We have seen that the precautions which the law
requires are those which the reasonable person would take in the light of the prob-
ability and magnitude of the harm in question, the cost of the precautions needed
to avoid it and the value of the harm-causing activity. However, we cannot mean-
ingfully or fairly say that a person should have taken such precautions unless we can
also say that they ought to have known about the risk at the time when it is alleged
that the precautions ought to have been taken. In the terminology of tort law, neg-
ligence is failure to take reasonable precautions against foreseeable risks of harm.
Foreseeable risks are those the reasonable person in the defendant™s position would
have foreseen.
The concept of foreseeable risk is a di¬cult one because foreseeability is rela-
tive in three important ways. First, an event may be more or less foreseeable
according to the detail in which the event is described. The fact that most houses
are insured against ¬re is testimony to the foreseeability of damage to or destruc-
tion of a house by ¬re; but it would be a very di¬erent matter to say, after a ¬re has
occurred, that anyone could or should have foreseen when, where and how it
might break out. In general, the more detailed the description of an event, the less
reasonable it would be to say that it should have been foreseen. For instance, it is
reasonable to expect a person to foresee that if they drive negligently they may
injure another road user. It would be much less reasonable to expect them to
foresee (for instance) the sex or age of that other road user or the precise nature of
the injuries su¬ered or the exact sequence of events that led to the accident and
the injuries. When the law says that a person cannot be liable for negligence unless
48 Chapter 2

harm to the claimant was foreseeable, it does not mean that every detail of what
happened must have been foreseeable. A person can be held liable for negligence
provided they ought to have foreseen a risk of harm su¬ciently great to justify
taking the precautions which the claimant alleges ought to have been taken. The
fact that the harm actually su¬ered by the claimant was greater than was foresee-
able, or that it occurred in an unforeseeable way, will not relieve a person of liabil-
ity provided it can be said that the person ought to have foreseen harm which
would have justi¬ed the taking of the precautions in question.
A second way in which the concept of foreseeability is relative arises from the
fact that what a person can foresee depends on what they know. A person who
knows that a vessel is full of petrol vapour, for instance, is much more likely to
foresee the destruction of the vessel by ¬re than a person ignorant of this fact. The
foreseeability of particular events may also depend on the state of scienti¬c and
technical knowledge. This has proved particularly important in relation to negli-
gence claims arising out of the use of pharmaceutical drugs and in litigation
against cigarette companies. Since foreseeability depends on knowledge, the
obvious question is, whose knowledge? The law™s answer is, ˜the knowledge which
the reasonable person in the defendant™s position would have had™. As we saw
above, the concept of the reasonable person in the defendant™s position rests ulti-
mately on value judgments about the amount of care people ought to take for the
protection of others.
A third reason why foreseeability is relative is that people vary in their attitude
to risk. Some people are ˜risk averse™, others are ˜risk-takers™ and yet others are ˜risk
neutral™. The more risk averse a person is, the more likely they are to foresee remote
risks of harm and to take precautions. By contrast, the more ˜risk-taking™ a person
is, the less likely are they to foresee or guard against risks of harm in their activities.
The courts have never explicitly considered the relationship between attitudes to
risk and the legal concept of foreseeability. What is the ˜reasonable person™s™ attitude
to risk? Perhaps it is neutrality. The point is that one person might foresee a risk
that would not occur to another. The law must, even if only implicitly, adopt some
attitude to risk in applying the concept of foreseeability.

2.4.8 The objective standard of care
The question in a negligence action is not whether the defendant personally could
have foreseen the harm or could have avoided it. The general principle is that the
defendant™s personal capacity to foresee and avoid harm is irrelevant. The judge
must decide what the defendant should have done (which is what the reasonable
person would have done), not what he or she could have done. Various reasons for
this ˜objective™ approach can be identi¬ed. First, it would be di¬cult and time-
consuming to determine the relevant capabilities of every defendant; secondly, it
would be very di¬cult to tailor the notion of ˜reasonable care™ to the personal capa-
bilities of each defendant; thirdly, to the extent that the legal concept of negligence
is rooted in morality, it shares with morality the role of setting standards of conduct
Fault as a basis of liability 49

which people are expected to strive to achieve.25 Fourthly, de¬ning negligence in
terms of what the defendant personally could have done would unduly sacri¬ce the
interest of potential claimants in personal security and freedom from injury and
damage to the freedom of potential injurers to engage in risky activities. The objec-
tive standard of care can be understood as the law™s attempt to strike a fair balance
between the competing interests in freedom of action and personal security that we
all share.
The fact that the legal standard of care is objective should be distinguished from
the issue of how demanding that standard is. For instance, prevalence of liability
insurance has perhaps encouraged courts to impose standards of care which are
beyond the reach of many people, because they know that in the typical case, the
defendant personally will not have to pay any damages awarded. This last reason
provides one of the explanations for the rule that a learner driver26 or the inexperi-
enced doctor must conform to the same standard of care as is required of experi-
enced drivers or doctors, and the fact that physical disabilities are generally ignored
in judging whether a driver was negligent:27 if the defendant is insured against
liability (as car-owners are required to be by law) then the law™s aim of compensat-
ing persons injured on the road can be achieved without imposing intolerable
¬nancial strains on negligent drivers.28
Nevertheless, the law must pay some attention to what could have been done: it
would be Kafka-esque to say that the defendant should have done something that
could not have been done by anybody. Conversely if the defendant is a person
claiming special skill, such as a doctor or other professional, the court will take into
account, in deciding what should have been foreseen or what precautions should
have been taken, the standards of conduct commonly achieved by people possessed
of that skill or by members of that profession. Even so, the question in any particu-
lar case is not what degree of care the defendant was capable of exercising, but
whether the defendant exercised the degree of care the law requires.
The objective nature of the legal de¬nition of negligent conduct tells us some-
thing about the aims of the law of negligence: if the law™s main aim was to reinforce
some notion of personal fault, the law might pay more attention to the abilities of
individual defendants, such as learner drivers. Again, if the main aim was to deter
negligent conduct in the future, the law might take more account of the ability of
individuals to avoid the sort of conduct in question. But the prime concern of the
modern law of negligence as it applies to death and personal injuries, is to provide
compensation for loss and injury su¬ered as a result of negligent conduct. In prac-
tice, such compensation nearly always comes out of an insurance fund. This is not
to say that the legal notion of negligence is totally divorced from moral notions of

25 People who are incapable of exercising the required care should avoid situations in which their
lack of capacity may produce adverse consequences.
26 Nettleship v. Weston [1971] 2 QB 691.
27 Roberts v. Ramsbottom [1980] 1 All ER 7.
28 See further ch. 9.
50 Chapter 2

fault, nor that the law is not concerned to encourage care, but only that the goal of
compensation is uppermost in modern law.

2.4.9 Negligence in design and negligence in operation
In practice, although not in legal theory, there is an important di¬erence between
negligence in the operation of an object or an activity and negligence in the design of
an object or activity. The distinction is not always easy to draw, but in general terms
it is much easier to establish negligence in operation than negligence in design.29
This is especially noticeable in the case of road accidents. There is no reason in
theory why an injured person should not sue a motor manufacturer for the negli-
gent design of a vehicle, or a highway authority for negligent design of a road junc-
tion or a roundabout, but in practice such actions would be unlikely to succeed. Bad
vehicle design is undoubtedly a factor in the causation of many injuries, and in the
USA it has been the source of much negligence litigation. So far in Britain no judg-
ment in favour of a claimant in a motor accident case has been based on bad vehicle
design.30 One hardy litigant sued the manufacturers of a bus alleging that the failure
to provide a central pillar on the platform was negligence, but the action failed in
the House of Lords.31 English courts are unwilling to decide cases on design issues
of this sort and tend to base their judgments on other grounds.32 This unwillingness
appears also to be a feature of cases involving injuries caused by defective products
in which negligent design is alleged: courts tend to decide such cases on the issue of
failure to control or warn against the hazard rather than on that of negligent cre-
ation of the hazard.33 In general there is a strong tendency to attribute injuries to
some act or omission occurring close in time to the event causing injuries (such as
speeding or failure to protect a worker from some health hazard) rather than to
some design feature of the environment in which the act or omission occurs (such
as the state of the road or the design of equipment).34

29 There are some Australian examples of successful design negligence claims: e.g. O™Dwyer v. Leo
Buring Wines [1966] WAR 67 (design of a wine bottle stopper); Suosaari v. Steinhardt [1989] 2
QdR 477 (design of a trailer); Flynn v. Commonwealth of Australia (1988) 6 MVR 186 (design of
a median strip). English examples are hard to ¬nd. One is Winward v. TVR Engineering Ltd [1986]
Business and Trading Law Cases 366. The case of the Abbeystead pumping station in Lancashire
(in which there was a methane gas explosion in 1984) is another: the Court of Appeal held that
the station was negligently designed (Guardian, 19 February 1988) and the House of Lords refused
the designers leave to appeal the decision (Independent, 10 June 1988). Another possible example
is Wood v. Bentall Simplex Ltd, The Times, 3 March 1992 (design of a slurry tank). See also
J. Stapleton, Product Liability (London, 1994), 251“2.
30 In 1992 an action was launched against Ford in respect of the design of centre rear seat-belts: The
Times, 8 December 1992.
31 Scottish Omnibuses v. Wyngrove, The Times, 24 June 1966.
32 Perhaps the major exception to this generalization concerns obligations of employers to provide
a ˜safe system of work™.
33 J. Stapleton, ˜Compensating Victims of Disease™ (1985) 5 Oxford J. of Legal Studies 248, 253.
34 J. Reason, Human Error (Cambridge: Cambridge University Press, 1990), ch. 7. It is probably
the case that both human behaviour (especially deliberately risky behaviour): D. Parker and
S. Stradling, ˜In¬‚uencing Driver Attitudes and Behaviour (No. 17)™ (Driver Behaviour
Research Group, University of Manchester, undated) and system design contribute to many
Fault as a basis of liability 51

A number of factors may account for the lack of litigation on design issues.
First, the fact, for example, that most cars are designed with the same basic defects
might help manufacturers in that it would enable them to argue that they built all
the customary safety features into their cars and that there is no reason why they
should be required to do more. As we have seen, however, the courts have never
accepted customary practice as completely precluding a ¬nding of negligence, and
there is nothing in law which would prevent a holding that customary design was
Another factor is that a decision by a court that a vehicle (or other product) was
badly designed, though technically a decision about a ˜question of fact™, would
e¬ectively be a legislative act. If there has been negligence in the design of an article,
there must have been negligence in the manufacture of all other articles made to
the same design. When dealing with motor vehicles a court would doubtless be reluc-
tant to make a decision of this kind, because in doing so it would be competing with
the statutory powers of the appropriate Minister to make regulations prescribing
requirements for the construction of vehicles. There is certainly no legal reason why
a court should not declare a design to be negligent, even though a safer design has
not been prescribed by regulation. But there are grounds for regarding the legislative
powers of the Minister to be a more appropriate way of dealing with this sort of
problem, partly because the Minister can take into account wider issues of public
interest that would be ignored in the courts, such as, for instance, the e¬ect of par-
ticular design requirements on the export trade, and the need to give the makers time
to change their designs.35 Similarly, if the courts were to hold that it is negligent of
the Home O¬ce to maintain ˜open™ prisons from which the inmates can easily escape
and do damage, they would be pronouncing on complex and politically sensitive

accidents “ although this distinction itself may be misleading because humans design faulty
systems. It may also be the case that the relative contribution of these factors varies from area to
area and activity to activity. But hard evidence may be di¬cult to ¬nd. For instance, one survey
concluded that ˜despite a wealth of literature . . . there is no reliable evidence either for or against
a relationship between car crash injury risk and any . . . measure of fatigue or sleepiness [other
than sleep apnoea] from current research™: J. Connor, ˜The Role of Driver Sleepiness in Car
Crashes: A Systematic Review of Epidemiological Studies™ (2001) 33 Accident Analysis and
Prevention 31. Another study found that road improvements in the USA between 1984 and 1997
did not reduce, and may even have increased, the total number of injuries and fatalities. Reduction
in total fatalities was attributed to factors such as increased usage of seat-belts and reduced alcohol
consumption: R.B. Noland, ˜Tra¬c Fatalities and Injuries: The E¬ects of Changes in
Infrastructure and Other Trends™ (2003) 35 Accident Analysis and Prevention 599. Yet another
recent study concluded that there are so many methodological defects in studies of the causal role
of ˜psychological™ factors in road accidents that we know have only ˜a very vague knowledge of
what psychological variables can actually predict accidents™: A.E. af Wahlberg, ˜Some
Methodological De¬ciencies in Studies on Tra¬c Accident Predictors™ (2003) Accident Analysis
and Prevention 473. For a thorough review of the research see Health and Safety Executive,
Di¬erences in Accident Liability, Contract Research Report 175/1998.
35 For the same reason there may be objections to allowing actions for damages in respect of fault-
ily designed houses, although it appears that negligent design accounts for a signi¬cant number
of accidents in the home (8% according to one inquiry: see Personal Factors in Domestic Accidents:
Prevention through Product and Environmental Design (Consumer Safety Unit, DTI, 1983), 17
(reporting research by the Building Research Station in 1964).
52 Chapter 2

issues about the design of the penal system which are more appropriately decided by
the executive and the legislature.36 There would be no such di¬culty in a ¬nding that
prison o¬cers had performed their custodial task negligently.
A third factor militating against judicial resolution of design issues arises from
doubts as to whether courts are able to weigh the social costs and bene¬ts of
di¬erent designs and whether they ought to do so. For example, if it would cost £X
to install a new safety device in all cars, are judges the right people to decide that
everybody should pay £X more for their cars? Again, is a court the right body to
decide how much a drug company should spend on testing a new drug to ensure
its safety? The problem of balancing risks against gains in such cases is very di¬cult
indeed, and probably beyond the resources of the courts. It would involve an assess-
ment of the risk of accident “ which may be di¬cult enough “ but it would also
involve consideration of how much the public gains through being able to buy
cheaper cars or from the availability of a particular drug sooner rather than later,
and less rather than more expensively.
A fourth relevant factor arises out of the fact that litigation which considers
design issues is likely to be much more complex, lengthy and costly than litigation
which concentrates on speci¬c acts or omissions.
The question of whether the courts ought to be making decisions in ˜design™
cases that certain precautions to avoid loss or damage ought to have been taken is
of particular importance in relation to actions against public bodies. Suppose, for
example, that a local authority is sued in respect of a road accident on the ground
that it should have installed tra¬c lights at a dangerous intersection; or in respect
of someone™s death by drowning at a dangerous beach because it neglected to
provide warning ¬‚ags or a lifeguard. Installing lights or providing lifeguards costs
money. Furthermore, a decision that lights ought to have been installed at one
intersection or a lifeguard provided on one beach might lead this and other local
authorities to feel that in order to avoid liability in negligence it would be necessary
to install lights at many dangerous intersections or to provide lifeguards on many
dangerous beaches. To meet the cost of such precautions other public projects, such
as the provision of new hospitals or extensions to schools, might have to be starved
of funds if extra revenue cannot be raised. As a matter of constitutional theory, it
is widely accepted that such policy choices between, for example, safer roads, better
schools and more hospitals, ought to be made by elected representatives of the
people and not by judges.
Another important example of this problem relates to the prevention of crime.
Criminal attacks on individuals may sometimes be facilitated by ine¬cient police
patrols or investigations; or by refusal of the police to protect someone who has
been threatened by thugs or a vital witness in a case against a well-organized gang of
criminals.37 In England a court would be very unlikely to ¬nd the police negligent

36 Home O¬ce v. Dorset Yacht Co. [1970] AC 1004.
37 See Schuster v. New York (1958) 154 NE 2d 534.
Fault as a basis of liability 53

for failing to take adequate steps to prevent this or that crime, because questions
about expenditure on the prevention of crime and about the level of policing are
left to the police or their political masters.38 The existence of a scheme, separate
from the tort system, for the compensation of the victims of criminal injuries
(the Criminal Injuries Compensation Scheme) provides the courts with another
ground for refusing to impose negligence liability on the police, at least in respect
of personal injury and death. The argument that the law of negligence might be
used in this context to encourage the police to take greater care in detecting and
preventing crime has been turned on its head: to hold the police liable would be,
it is said, to risk ˜overkill™; that is, it would tend to make the police unduly cautious
in doing their job for fear of being sued rather than for any good operational
reason. In other words, to impose liability too readily on the police or other public
authorities would be to risk over-deterrence. The overkill argument is commonly
used not only in actions against public authorities but also in actions against
professionals. Its main weaknesses are that there is very little empirical evidence
to support the idea that tort liability has the sort of e¬ects the argument assumes;
and that no court has ever de¬ned in any meaningful way how much deterrence
is too much.39

2.5 Conduct of the claimant
So far in this chapter we have been considering the notion of fault (in the sense
of negligence) in relation to the conduct of the defendant. There are circumstances
in which a claimant may be deprived of part or all of a damages award because
of his or her own conduct. Circumstances in which we would say that a person™s
injuries were wholly or partly their ˜own fault™ are mostly40 dealt with in the law
by the ˜defence™ of contributory negligence. The defence of volenti non ¬t injuria
(or ˜assumption of risk™), by contrast, exonerates the defendant not because the
claimant was at fault but because the claimant accepted the risk of injury; although
in many cases, such conduct is little di¬erent from contributory negligence. Finally,
the defence of illegality deprives the claimant of damages not because the injuries

38 Hill v. Chief Constable of West Yorkshire [1989] AC 53 (police owed no duty of care to a woman
who was the penultimate victim of a serial rapist and murderer).
39 See further R. Dingwall, P. Fenn and L. Quam, Medical Negligence: A Review and Bibliography
(Oxford, 1991), 44“51; M.A. Jones and A.E. Morris, ˜Defensive Medicine: Myths and Facts™ (1989)
5 J. of Medical Defence Union 40; D. Tribe and G. Korgaonkar, ˜The Impact of Litigation on Patient
Care: an Enquiry into Defensive Medical Practices™ [1991] Professional Negligence 2; P. Cane,
˜Consequences in Judicial Reasoning™ in J. Horder ed., Oxford Essays in Jurisprudence, Fourth Series
(Oxford, 2000); and
40 The rule that a claimant must take reasonable steps to ˜mitigate™ (i.e. reduce to a minimum) their
loss is also underpinned by some notion of fault. Contributory negligence is pre-accident fault,
while failure to mitigate loss is post-accident fault. In relation to economic loss, the courts have
adopted a general principle that people should take reasonable steps to protect themselves against
such loss. This principle is used as the basis for denying tort liability for negligently caused eco-
nomic loss.
54 Chapter 2

were his or her own fault but because of an objection to compensating a person for
loss or damage arising out of criminal behaviour.

2.5.1 Contributory negligence
Contributory negligence is failure to take reasonable care for one™s own safety as
opposed to failure to take reasonable care for the safety of others; or, put another
way, failure to take reasonable precautions against risks of injury to oneself, of
which one was aware or ought to have been aware. Until 1945, a ¬nding that
the claimant™s injuries were wholly, or even partly, the result of his or her own
(contributory) negligence was a complete defence in the sense that it resulted in
the claimant receiving no damages at all. This defence was originally based on the
same general idea of fault that justi¬ed liability for negligence, although
the legal justi¬cation for the defence was expressed in terms of ˜cause™.41 It may
also have been partly based on some idea of deterrence: people should be encour-
aged to take care for their own safety even when imperilled by the negligence of
others. It has been argued that neither of these rationales is very satisfying.42 As
for the ¬rst, the kind of ˜fault™ which justi¬es liability is not the same as the kind
of fault embodied in the notion of contributory negligence. The ˜fault™ of a defen-
dant can, in a broad sort of way, be treated (in many cases) as involving self-inter-
ested or unsocial risk-taking at the expense of others. A claimant™s ˜fault™ is not of
this kind: such fault is not so much sel¬sh as just foolish, and it is not clear that
one can equate, or even compare, foolishness with sel¬shness. The second ratio-
nale is looked at in some detail later,43 and it is enough to say here that the instinct
for self-preservation is likely (in most circumstances) to be quite su¬cient to
deter most people from taking risks with their own safety. It is, therefore, not at
all evident that the deterrent function of the doctrine of contributory negligence
is of any real value.
In the course of time, a rule that denied a negligent claimant any compensation
however slight the claimant™s fault and however serious the defendant™s, appeared
unjust and led the courts to invent devices to mitigate the e¬ects of the doctrine.
The courts did not feel able to take the sensible course of reducing the claimant™s
damages to re¬‚ect the fact that he or she was to some extent to blame. What was
done, in suitable cases, was to deny that the claimant™s contributory negligence had
been a real cause of the loss, and to insist that the defendant™s negligence was the
˜sole™ cause of the damage. The problem with this approach was that the claimant™s
negligence was always in some sense a cause of the damage, and so to say that the
defendant™s negligence was the sole cause only meant that the court was prepared
to ignore the claimant™s negligence and treat the defendant™s negligence as solely
responsible. The device for evading the contributory negligence doctrine was the

41 See e.g. Bowen LJ in Thomas v. Quartermaine (1887) 18 QBD 685, 694; and Caswell v. Powell
Du¬ryn Collieries [1940] AC 152.
42 G. Schwartz, ˜Contributory and Comparative Negligence: A Reappraisal™ (1978) 87 Yale LJ 697.
43 See
Fault as a basis of liability 55

so-called ˜last opportunity rule™, i.e. the rule that the person who had the last oppor-
tunity to avoid the accident should be treated as its sole cause.44 The courts were
never very happy with this rule, partly because it ran counter to prevalent ideas
about causation, and partly because it seemed such a crude method of mitigating
the harshness of the contributory negligence rule. Eventually the law became intol-
erably subtle and complex so that it was well nigh impossible to direct a jury in intel-
ligible terms “ and cases were still being tried by juries when these di¬culties were
at their peak.
The Law Reform (Contributory Negligence) Act, enacted in 1945, empowered
the courts to apportion the responsibility for an accident, and to reduce the
damages awarded to a claimant who had been guilty of contributory negligence,
while not denying a remedy altogether. At ¬rst there were doubts about the extent
to which the Act had done away with the complexities of the old law,45 but in prac-
tice the Act of 1945 has removed most of the di¬culties from this part of the
law. Now the only questions that arise in the typical case are whether the claimant
acted negligently, whether the damage or loss was wholly or partly the result of that
negligence, and by how much (if at all) the claimant™s damages should be reduced.
One of the few points of general application to arise since the passage of the 1945
Act has involved the application of contributory negligence rules to passengers in
cars. Failure by a passenger in a car to wear an available seat-belt is contributory
negligence,46 and if it can be shown that the passenger™s injuries would not have
occurred or would have been less serious if a seat-belt had been worn, the passen-
ger must bear some share of the responsibility. In Froom v. Butcher 47 the Court of
Appeal laid down as a general guide that the claimant™s damages should be reduced
by 25% in cases where the injuries would have been prevented altogether, and by
15% where they would have been less severe if an available seat-belt had been
worn. A passenger who consents to be driven by a driver clearly the worse for
alcohol may also have their damages reduced for contributory negligence.48
The Act of 1945 has greatly simpli¬ed the law of contributory negligence and
made it much fairer. Judging from the reported cases, it appears to work smoothly in
practice and few di¬culties have been encountered in its application.49 Nevertheless,
there are important questions about the whole doctrine of contributory negligence,
and about the relationship between the doctrine and liability insurance, to which
attention should be drawn.

44 See R.F.V. Heuston and R.A. Buckley eds., Salmond and Heuston on The Law of Torts, 21st edn
(London, 1996), 486“8.
45 See e.g. Davies v. Swan Motor Co. [1949] 2 KB 291, 310; Stapley v. Gypsum Mines [1953] AC 663,
46 As e.g. is failure by a (motor) cyclist to fasten the chin strap of a crash helmet: Capps v. Miller, The
Times, 12 December 1988.
47 [1976] QB 268.
48 Owens v. Brimmell [1977] QB 859; see N. Gravells, ˜Three Heads of Contributory Negligence™
(1977) 93 LQR 581.
49 For some evidence about the role of contributory negligence in the settlement of cases see Harris
1984 Survey, 91“2.
56 Chapter 2 The di¬erence between negligence and contributory negligence
In the ¬rst place, it is important to understand the relationship between negligence
and contributory negligence in practice. At ¬rst sight contributory negligence
appears to be a sort of mirror image of negligence itself. There is an apparently sat-
isfying balance in the idea of the negligence of the injurer being counterpoised by
the negligence of the injured. But the practical e¬ect of a ¬nding of contributory
negligence is very di¬erent from the e¬ect of a ¬nding of negligence. To ¬nd a defen-
dant guilty of negligence shifts a loss away from the claimant and typically spreads
it by means of insurance or other processes. A ¬nding of contributory negligence
usually has precisely the opposite e¬ect, which is to leave part or all of the loss on
the claimant, who will typically be without relevant insurance. Thus, reduction of
damages for contributory negligence typically falls much more heavily on the
claimant than liability for negligence bears on the defendant. In practice, negligent
people do not pay for the consequences of their negligence; but contributorily neg-
ligent people do pay for the consequences of their contributory negligence. It is not
too much to say that the only signi¬cant group of people who are called upon to
bear the consequences of their negligence are accident victims themselves.50
This di¬erence between the e¬ect of a ¬nding of negligence and the e¬ect of a
¬nding of contributory negligence may have in¬‚uenced the courts in recognizing a
very important legal distinction between negligence and contributory negligence.
The test of negligence as applied to the conduct of claimants is more personalized
than the test of negligence applied to defendants. In other words, the courts are more
prepared to acquit claimants of negligence on grounds of their personal abilities and
characteristics (and so avoid the need to reduce their damages) than they are to
acquit defendants on such grounds (with the result that the claimant is deprived of
compensation). In particular, the age of the claimant is taken into account in deter-
mining contributory negligence. A young person is only expected to show the degree
of care which a person of that age should exercise, and the same may be true of an
elderly person. The importance of this in practice can be gauged from the fact that
young children and old people form a disproportionate number of the pedestrians
killed and seriously injured in road accidents. In 2004, for example, the recorded
casualty rate for pedestrians killed and seriously injured in road accidents was 9 per
100,000 for persons aged 30 to 39, but 23 per 100,000 for children aged 8“11, and 13
per 100,000 for those aged between 70 and 79.51 The Pearson Report recommended
that contributory negligence should not be available as a defence in road accident
cases where the injured person is a child under the age of 12.52 This would make very
little di¬erence to the practical position at present, though it is impossible to under-
stand why this proposal should have been limited to road accident cases.

50 The point is well made by A. Tunc, La Securit© Routiere (Paris, 1966), 31“7 and ˜The Twentieth
Century Development of the Law of Torts in France™ (1965) 14 International and Comparative Law
Q., 1089,1100“1.
51 Department for Transport, Road Casualties in Great Britain 2004, table 30.
52 Pearson Report, vol 1, para. 1077.
Fault as a basis of liability 57 Contributory negligence and family cases
At one time in the nineteenth century there was support for a doctrine of
˜identi¬cation™ under which one person might be so identi¬ed with the contribut-
ory negligence of another as to preclude the former from recovering damages even
though personally free from fault. For example, a child who was accompanying his
grandmother was severely injured at a station when his grandmother crossed the
lines and both were struck by a passing train. It was held that as the grandmother
had been found to be contributorily negligent the child could not recover.53 Some
such idea applies in claims under the Fatal Accidents Act 1976 where the claimant™s
damages must be reduced proportionately to any negligence on the part of the
deceased contributing to the death. Under a provision of the Congenital54
Disabilities (Civil Liability) Act 1976 the claim of a child for damages in respect of
injuries su¬ered before birth can be met with defences available to the defendant
against its mother. In general, however, the doctrine was thought unjust, and was
rejected in 1888 by the House of Lords which held that a person was not to be
a¬ected by the contributory negligence of another unless the former was legally
liable for that other™s acts; for example, if the latter was the former™s servant acting
in the course of employment.55
But there are cases in which the lack of some such doctrine produces strange
results in cases involving members of one family. Suppose that a person is injured
by the negligent driving of their spouse; the injured spouse can recover damages
from the other (in reality, from the insurer). The family as a whole will ˜gain™ from
the award of damages and the negligent spouse may well share in these gains.
Suppose, next, that one of the spouses is killed through the negligence of the other.
In Dodds v. Dodds 56 a man was killed in an accident caused by his wife™s negligent
driving. The wife could not obtain damages for the loss of her husband, but it was
held that their 81„2-year-old son was entitled to damages against his mother for
causing his father™s death, and he was awarded £17,000 “ paid, of course, by the
insurers. In such a case the bulk of the capital would probably be retained under
the control of the court until the child reached majority, but the income would
probably be paid to the mother for the maintenance and education of the child. So
in reality the negligent spouse would share the bene¬t of the award. On the other
hand, we may baulk at allowing a tortfeasor to bene¬t directly from an award of
damages in respect of the tort. In one case the question arose whether an injured
claimant could be awarded damages representing the value of care rendered
gratuitously to her by the tortfeasor (her partner). The House of Lords, having
decided that when such damages are awarded, they are ˜held on trust by the

53 Waite v. North Eastern Railway (1858) El, Bl & El 719; 120 ER 679.
54 Note that ˜congenital™ in this context means ˜su¬ered as a result of events occurring before birth™,
not ˜attributable to genetic factors™.
55 The Bernina (1888) 13 App Cas 1; see too Oliver v. Birmingham & Midland Omnibus Co. [1933]
1 KB 35.
56 [1978] QB 543.
58 Chapter 2

claimant for the carer™, held that an award under this head could not be made where
the carer was the tortfeasor.57
On the assumption that the fault system is concerned with personal responsibil-
ity for harm, these results seem remarkable. On the other hand, if emphasis is put on
tort as a mechanism for providing compensation rather than for compensating on
the basis of responsibility, then the cases seem less strange (and the position under the
Fatal Accidents Act 1976 seems the one which is out of line). There is no particular
reason to deprive one family member of compensation on the ground that another
faulty one will indirectly bene¬t thereby, when the damages will be paid by an insurer
and when the real su¬erer, if damages are not awarded, will be the innocent victim. The assessment of contributory negligence
Even though the principle of reducing damages for contributory negligence may
appear to be based on a simple idea about personal responsibility, the principle
according to which the claimant™s damages are reduced is far from obvious or
straightforward. The claimant™s damages will be reduced having regard not just to
the degree of his or her fault “ whether that be slight or gross “ but according to the
degree of the claimant™s fault relative to that of the defendant. In addition, just as
the amount of compensation which a negligent defendant must pay bears no rela-
tion to the degree of his or her fault where he or she alone is to blame, so also the
amount of the loss which the claimant must bear when partly at fault depends not
just on the extent of that fault, but also on the extent of the loss itself. Let us con-
sider how all this works with a few illustrations.
First, a claimant who is 50% to blame for an accident in which they su¬er a loss
assessed at £10,000 will lose £5,000 as a result of their negligence. A claimant who
is a mere 10% to blame for an accident in which the loss is assessed at £100,000 will
lose £10,000 as a result of their negligence.
Secondly, a motorist who commits a trivial act of negligence and collides with a
defendant who was driving with gross negligence will be held perhaps 10% to
blame; but the motorist (guilty of the same trivial act of negligence) may be held
50% to blame if the defendant was no more negligent than the claimant. Yet the
claimant™s act of negligence is precisely the same in the two cases.
Thirdly, and similarly, a motorist who is driving with gross negligence will cer-
tainly be held very largely responsible and so recover very little if involved in an
accident partly due (say) to the negligence of a pedestrian who crosses the road in
front of the car; but if our motorist is fortunate enough to collide with another
grossly negligent driver they will probably recover 50% of the loss.
Fourthly, the last illustration shows that two motorists driving with gross negli-
gence will each recover 50% of their loss, assuming their negligence to be of a
similar degree. If three negligent motorists all collide simultaneously due to the
same degree of negligence, the responsibility of each for their own injuries will be

57 See further 6.2.3.
Fault as a basis of liability 59

assessed as if the negligence of the other two was that of a single defendant.58 Thus
each will recover 50% of their loss. Contributory negligence and the fault system
Finally, and perhaps most seriously, the combination of the contributory negli-
gence principle with the ˜no liability without fault™ principle produces the result
that a person injured without fault on the part of anyone receives no tort compen-
sation, whereas a person who may be very largely to blame for his or her own
injuries can receive some tort compensation. Suppose that two workers are working
side by side in a factory and are both severely injured, su¬ering losses assessed at,
say, £100,000 each. Worker A is injured by gross negligence on their own part and
slight negligence on the part of a fellow worker; worker A will recover perhaps 20%
of the loss, i.e. £20,000, not of course from the fellow worker but from the employer
or its insurers. Worker B, on the other hand, is injured entirely by ˜accident™.
Worker B will receive not a penny in tort compensation.
When we look at other compensation systems, we will see that there are many situ-
ations in which the aim of compensating victims of injury and damage is so para-
mount that it is thought unjusti¬able to reduce the compensation because of fault
on the part of the victim. Life insurance or ¬re insurance or comprehensive motor
insurance would not be such attractive propositions if they did not provide protec-
tion against the risk of negligence on the part of the victim. The point is that however
attractive the idea that no one should incur tort liability in the absence of fault on
their part, the proposition that no one should receive compensation except for loss
or damage attributable to the fault of another may seem much less attractive.
Nobody knows quite what is the quantitative e¬ect of the law of contributory
negligence. The Harris Survey found that in 26% of the cases studied in which tort
damages were obtained, there was some reduction explicitly on the ground of con-
tributory negligence.59 In the survey of insurance claims handled in November 1973
conducted for the Pearson Commission, it was found that 26% of claims settled
were disposed of on the basis of partial liability;60 but this included cases settled
without any payment at all. The cases settled with a partial admission of liability
comprised about 31% of the number of cases in which some payment was made.
None of these ¬gures, however, tells us anything about the size of the discount
made on account of contributory negligence. In an Australian survey it was found
that the claimants who were found contributorily negligent lost 39% of their dam-
ages, but this does not tell us what proportion this bears to all tort recoveries.61 In
Scandinavia, where apportionment is permitted much as it is in England, it has been

58 Fitzgerald v. Lane [1989] AC 328.
59 For the results of the 1984 survey see Harris 1984 Survey, 91“2.
60 Pearson Report vol. 2, table 117.
61 Australian Committee Report, paras.130“1. The proportion of successful actions in which a
reduction for contributory negligence was made was found to vary from 10 to 28% (the ¬gures
correlated jurisdictions and whether or not the claimant su¬ered permanent disability).
60 Chapter 2

estimated that abolition of the defence would increase the cost of motor insurance
by at least 7.5%.62 The usefulness of the doctrine
Does the doctrine of contributory negligence serve any useful purpose? From one
point of view the answer must be ˜no™, at least in the law relating to personal injuries.
Since tortfeasors are almost invariably insured against liability, the doctrine is not
needed to spare an individual defendant the injustice of being made to compensate
an injured person who was partly to blame for his or her own injuries. It operates, in
e¬ect, as a penal device: the contributorily negligent claimant is punished by being
deprived of some of the compensation to which they would otherwise be entitled.
Penal laws are usually justi¬ed on the grounds of their deterrent value, but it is very
doubtful if the doctrine of contributory negligence has any deterrent value in per-
sonal injury cases. It is true (as has been argued)63 that fairness may still seem to
demand that if the claimant complains of the defendant™s negligence, the former
must be prepared to bring their own conduct onto the scales. The answer to this point
surely lies in the e¬ect of current insurance practice. When this is taken into account,
ignoring the claimant™s carelessness surely seems less unjust in the typical case where
the claimant is uninsured and the defendant insured. The negligent defendant will
not pay for their negligence, while the negligent claimant typically will pay for their
own negligence, if the damages payable by the defendant are reduced for contribu-
tory negligence.
There is, however, a pragmatic argument that may favour the retention of the
doctrine of contributory negligence in personal injury cases. If attention is con¬ned
exclusively to the tort system, the case for abolition appears to remain strong. But
when the whole scene is surveyed, the case weakens. For it then becomes apparent
that a claimant who recovers any tort damages is in a sense very fortunate compared
with most other victims of accident and disease. As we saw in chapter 1, we are here
talking of some 6.5% of accident victims, and a very much smaller proportion of
those who su¬er illness or disabilities from other causes. The ¬nancial provision
made for this very small proportion of the disabled and injured is already generous
by comparison with what is available to the others. It would seem wrong to improve
it still further, even by abolishing doctrines unjust in themselves.
Certainly, so long as the tort system retains anything like its present structure, it
would be undesirable to abrogate the defence of contributory negligence in relation
to property damage. If the doctrine were swept away altogether it would mean that
in many road accidents in which two motorists cause damage to their vehicles by
their combined negligence, each would be entitled to claim in full from the other.
Such a result might be acceptable in a personal injury claim, but if applied to all cases

62 J. Hellner, ˜Tort Liability and Liability Insurance™ in F. Schmidt ed., Scandinavian Studies in Law
1962 (Stockholm, 1962), 131, 159 n. 9.
63 See Schwartz, ˜Contributory and Comparative Negligence™.
Fault as a basis of liability 61

of damage to vehicles, it would result in a considerable and wasteful recourse to tort
liability and liability insurance, rather than to the personal accident insurance of the
vehicle owner. Given that property damage-only accidents are six or seven times as
frequent as personal injury accidents,64 this would undoubtedly increase the cost of
motor insurance by adding to the administrative cost.

2.5.2 Volenti non ¬t injuria
The defence of volenti non ¬t injuria is also sometimes referred to as the defence of
˜(voluntary) assumption of risk™. It has been associated with at least three types of
case. In some cases it is indistinguishable from the defence of contributory negli-
gence, except that it is used to deny liability altogether rather than as a ground for
apportioning damages. In other cases the claimant agrees not to sue the defendant
for any injury as a result of tortious conduct of the latter. In yet other cases a person
is taken to have consented to the defendant acting in accordance with a standard of
conduct lower than that normally required by the law. Failure to distinguish clearly
between these three types of case has caused much confusion. Volenti and agreement not to sue
Agreements not to sue for damages for death or personal injury caused by negli-
gent conduct may take the form of a clause (called an ˜exclusion clause™) in a con-
tract to which the claimant is a party, or of a written notice by which the claimant
has expressly or impliedly agreed to be bound.65 Such clauses or notices are (by
virtue of s. 2 of the Unfair Contract Terms Act 1977) ine¬ective in any case where
the defendant™s liability arose in the course of carrying on a business. Nor can a
defence of volenti be founded solely on such an agreement. Where the liability arose
out of the sale or hire-purchase of goods, an exclusion clause will be ine¬ective
against a claimant who did not acquire the goods in the course of a business (a ˜con-
sumer™) whether or not they were supplied by the defendant in the course of a busi-
ness. These provisions signal the importance we place on compensating people for
personal injury and death caused by negligence, and they recognize the fact that
individual consumers often have no choice whether or not to agree to exclusion
clauses, even if they are aware of their existence.
Where a passenger in a motor vehicle su¬ers personal injury as a result of negli-
gence of the driver, the passenger will not normally be bound by any agreement or
understanding with the driver that the passenger will not sue the driver.66 This is so
regardless of whether the agreement is in the form of a contract term, or is based on
a written notice or arose in some other way. The purpose of this provision is to

64 No precise ¬gures for property damage-only accidents are available because there is no obligation
to report these; this is an estimate of the Road Research Laboratory: The Cost of Road Accidents in
Great Britain LR 79 (HMSO, 1967).
65 The most common example is where a landowner puts up a prominent notice addressed to
persons entering their land.
66 Road Tra¬c Act 1988, s. 149(2).
62 Chapter 2

prevent passengers being deprived of the advantages of the third-party liability
insurance which users of motor vehicles must take out. Such insurance also covers
liability for property damage (up to ‚¬ 1 million),67 and agreements to exclude liabil-
ity for such damage are also ine¬ective.
In other contexts, however, the position with regard to negligent damage to goods
is quite di¬erent. Damage to property does not so urgently cry out for compensation
as death and personal injury; goods of signi¬cant value are often insured by their
owner against the risk of damage; and signi¬cant property-damage claims are much
more likely to be made by businesses (who are more able to look after their own inter-
ests) than by individuals. For these reasons, it is not so important to regulate the
exclusion of tort liability for damage to property. Under the Unfair Contract Terms
Act 1977, clauses which exclude liability for property damage can be e¬ective if they
are ˜reasonable™, except where the clause is contained in a contract for the sale or hire-
purchase of goods to a consumer, in which case it will be ine¬ective. At least in rela-
tion to contracts between business concerns, the availability of insurance will be an
important factor in deciding the issue of reasonableness.68 It is not clear, however, to
what extent the insurance factor will be held relevant in other cases. For example, will
the courts hold that car-parking companies may reasonably exempt themselves from
liability for negligent damage to cars on the ground that the owners can insure them-
selves and that many do so? There is a good deal to be said for the view that such an
exclusion would not be unreasonable (at least if it did not extend to the ¬rst slice of
damage, which is typically not recoverable under property insurance policies); other-
wise prudent (insured) owners would be paying for damage to less prudent (unin-
sured) owners.
So far we have been discussing cases where the defendant has committed a tort
and where the e¬ect of the relevant agreement is to protect the defendant from
liability. By contrast, sometimes the e¬ect of the relevant agreement or consent by
the injured person is to prevent conduct that would otherwise be tortious from
amounting to a legal wrong. Examples are agreement to allow a person on to one™s
land, so preventing the entry from being a trespass; and agreement to bodily
contact, or even to being hit, as in sports, where the agreement prevents the conduct
amounting to assault. In this type of case it is usually said that the defendant™s
defence is one of consent to trespass or assault rather than assumption of the risk
of trespass or assault; but the two defences are clearly related, and they have the
same legal e¬ect of depriving the claimant of a cause of action. Volenti and contributory negligence
It is sometimes argued that a person who voluntarily does something that presents
a risk of being injured by negligent conduct (such as taking a ride in a car driven
by a drunk, or testing an explosive device without taking shelter) should not be
67 This provision was introduced in order to comply with EC law, and it was a retrograde step.
68 Photo Production Ltd v. Securicor Transport Ltd [1980] AC 827; George Mitchell Ltd v. Finney Lock
Seeds Ltd [1983] 2 AC 803.
Fault as a basis of liability 63

allowed to recover damages for any resulting personal injuries because that person
has willingly or voluntarily ˜assumed the risk™ of being injured. This involves serious
confusion of thought. In the ¬rst place, taking a risk will entail a willingness that
the risk should occur only if the person taking it knew the nature and extent of the
risk. So, for the defence of assumption of risk to succeed, the defendant must ¬rst
prove that the claimant knew of the risk in some detail. But this is not enough69
because willingly taking a known risk may be the result of a choice between evils,
and not of indi¬erence as to whether the risk materializes, or of a desire that it
should materialize, still less of an intention to abandon the right to sue for damages
should the risk occur. A pedestrian who crosses the road certainly incurs a known
risk of being injured by a driver™s negligence, and is willing to be injured in the sense
of preferring to incur the risk rather than to stay permanently on one side of the
road. But willingness to run this risk is no justi¬cation for barring an action against
a negligent motorist who runs the pedestrian down.
Something more is needed to justify refusing damages to an injured claimant.
Perhaps the additional factor is that the risk must be a very great one, either in the
sense of very likely to materialize or in the sense that any resulting injury is likely
to be very great, or both. In some circumstances, however “ for example, where
people in distress or danger are being rescued “ the taking of great risks is felt to be
justi¬ed, and a person who takes such a risk to e¬ect a rescue will not be denied
damages on the ground of assumption of risk. It seems then that the only remain-
ing possibility is to treat the defence as con¬ned to those cases in which the claimant
ran a risk that was unjusti¬ed or unreasonable in the circumstances. If this is correct,
the only di¬erence between the defence of assumption of risk and that of contrib-
utory negligence is that in the former case the claimant must actually have known
of the risk whereas in the latter case it is enough that the claimant knew or ought to
have known of the risk. On this basis, in any case in which a defence of assumption
of risk would be available, a defence of contributory negligence would also be avail-
able. If this is so, why do courts ever allow a defence of volenti to succeed and deny
the claimant any damages at all when the defence of contributory negligence allows
the court to reduce the damages awarded to the claimant by such proportion as it
thinks ¬t? In the great majority of cases this discretion enables the court to achieve
what it regards as a just solution, and consequently a defence of volenti rarely suc-
ceeds in this sort of case today.
There are two reasons why the defence survives. The ¬rst is that although the dis-
cretion to reduce a claimant™s damages for contributory negligence is wide, it does
not allow the court to award nothing70 “ which it may want to do if the claimant
acted in a grossly negligent way. The second reason is that damages are apportioned
for contributory negligence according to the relative degree of fault of the two
parties, and cases occur in which the claimant has acted very unreasonably, so that

69 Smith v. Baker [1891] AC 325.
70 Pitts v. Hunt [1991] 1 QB 24.
64 Chapter 2

the court wishes to award little or nothing, but in which the defendant has been
equally or even more negligent. For instance, a passenger who consents to be driven
(or ¬‚own)71 by a person in a heavily drunken state is doing something very foolish
indeed, but at the same time the court could hardly assess the defendant™s share of
responsibility as less than that of the claimant; so the court™s power to reduce the
damages the claimant will receive is limited in this kind of case.72 The same is true
of cases of joint negligent action, for example, where two workers together do
something very negligent and injuries occur to one or both. It is in such cases that
courts may be attracted to the defence of volenti.73
This is hardly satisfactory. The truth is that recourse to the defence of volenti in
cases where the claimant has been injured partly by their own fault and partly by
the fault of the defendant, ¬‚ies in the face of the provisions of the apportionment
legislation and of the case-law which has grown up around it. The main reason for
these di¬culties is undoubtedly the prevalence of liability insurance. It may be
thought, for instance, that as between a drunken driver and a willing passenger, the
main responsibility for injuries to the passenger should rest on the driver. But as
between the willing passenger and some third party who will actually pay any com-
pensation “ an insurance company, or an employer vicariously liable, or the State “
the passenger™s responsibility may be thought to be great enough to justify award-
ing little or no compensation. Volenti and standard of care
The classic example of the type of case we are concerned with here is that in which
the claimant is injured while watching some sporting event “ for example, by a ¬‚ying
puck at an ice hockey match;74 or by a horse at a show-jumping contest;75 or a by car
at a race track.76 The question in such cases is whether a person should be allowed to
complain of conduct which would or might be negligent in a di¬erent place, or in a
di¬erent context. Whether the defendant in such cases has failed to take reasonable
care depends in part on whether the claimant is a willing spectator: for example,
a spectator at a cricket match may be said to have accepted the risk of being hit and
injured by a six, but a pedestrian on the street outside the playing ¬eld surely has not.
Putting the matter another way, a driver who races round a race track at 100 m.p.h.
in front of willing spectators is not driving negligently just because of the speed of
the car; but it would be negligent to drive at the same speed on a public road.
Whether the consent of the claimant should be allowed to a¬ect the standard of
care required turns on ideas of personal responsibility. For example, it has been
held that a woman who willingly goes to a jeweller to have her ears pierced cannot

71 Morris v. Murray [1991] 2 QB 6.
72 Owens v. Brimmell [1977] QB 859; also Gregory v. Kelly [1978] RTR 426.
73 E.g. ICI v. Shatwell [1965] AC 656.
74 Murray v. Haringey Arena [1951] 2 KB 529.
75 Wooldridge v. Sumner [1963] 2 QB 43; see also Rootes v. Shelton (1967) 116 CLR 383.
76 Hall v. Brooklands Auto Racing Club [1933] 1 KB 205.
Fault as a basis of liability 65

complain if the conditions under which the procedure is done are less hygienic than
those which exist at a doctor™s surgery.77 Or take the case of a woman who applied
for a job knowing that she was allergic to a substance which the employers used,
but not anticipating the extent of the risk she thereby faced. The employers also
knew of the allergic condition, and the issue was whether the employers were under
a duty not to employ the woman, assuming no precautions were possible. The court
said, ˜No™.78 The result would be di¬erent, however, if an employer subjected its
existing employees to a new risk: in such a case the fact that the employees ˜will-
ingly™ went on working for the employer would not allow it to argue that they had
in some sense accepted the risk.
The e¬ect of a decision that the claimant willingly accepted a lower standard of
care than might otherwise be expected will normally be to acquit the defendant of
negligence and to deprive the claimant of any damages. Looking at these cases
from another point of view, it may sometimes (but not always) be possible to argue
that what the claimant has done is to take less than reasonable care for their own
safety. When this is so, the better approach might be to hold that claimant con-
tributorily negligent and to reduce the damages awarded rather than to deny com-
pensation entirely.

2.5.3 Illegality
This defence is not of much importance in practice: a court which wants to penal-
ize an injured person for being in breach of the criminal law at the time the injuries
were su¬ered will usually be able to do so by allowing a defence of contributory
negligence79 or volenti to succeed. It does have a role in road accident cases where
the court thinks that the claimant should recover nothing.80 This result cannot be
achieved under the apportionment legislation, and the volenti defence is not nor-
mally available in road accident cases.81
The defence is of theoretical interest because it raises in acute form the question
of the proper role of tort law “ is it to compensate the injured, or to give e¬ect to
judgments about fault by compensating injured persons in appropriate circum-
stances, or to deter culpable conduct? The situation is made more complicated by
worries about the extent to which the courts would lower their prestige and credibil-
ity if they were to ˜help™ criminals by awarding them damages. A number of basic
problems have troubled the courts. The ¬rst concerns the extent to which the civil
law ought to be used as an adjunct to or reinforcement for the criminal law. The
argument that it should be so used is a two-edged sword where both the claimant
and the defendant were acting illegally at the time of the injuries. To deny compen-
sation might deter the claimant from criminal activity in the future, but relieving the

77 Phillips v. William Whitely [1938] 1 All ER 566.
78 Withers v. Perry Chain Co. [1961] 1 WLR 1314.
79 E.g. Revill v. Newbery [1996] QB 567.
80 Pitts v. Hunt [1991] 1 QB 24.
81 Road Tra¬c Act 1988, s. 149(3).
66 Chapter 2

defendant of liability for tortious conduct could hardly have a deterrent e¬ect.
A second point is that much modern ˜criminal™ legislation is in fact only regulatory;
it is very often designed to co-ordinate human behaviour for the sake of e¬ciency
or to set safety standards to protect people from their own carelessness or stupidity.
Little, if any, stigma will attend breach by injured persons of many such laws; and so
while deterrence of breach by means, for example, of a ¬ne may be desirable, the
unpredictable and usually much more serious sanction of the denial of a civil
remedy may seem an unnecessary and unduly harsh sanction. Thus, in a number of
cases the question of whether a plea of illegality should succeed has been said to
depend in part on whether the ˜public conscience™ would be ˜a¬ronted™ or the ˜ordin-
ary person shocked™ if the claimant were allowed to recover.82
A third point relates to the allocation of resources: if a choice has to be made
between allowing the claimant to recover from the defendant™s insurer or, on the
contrary, leaving the claimant in the position of needing to rely on social security
bene¬ts, it is by no means obvious that any good purpose is served by denying
recovery against the defendant.
A fourth problem concerns the relationship between the illegal act and the
injuries. In the formulation adopted above the issue was put in terms of whether
the claimant was acting illegally at the time the injuries were su¬ered. But the
defence is unlikely to succeed unless the fact that the claimant was acting illegally
was in some fairly strong sense a cause of the injuries. The basic principle appears
to be that the fact that the claimant was acting illegally at the time the injuries were
su¬ered provides no answer to a claim for damages. If the rule were otherwise,
many people who su¬er personal injuries on the road or at work would recover no
tort damages because breach of some tra¬c or safety regulation by the injured
party is a common contributory cause of injuries. It is only in rather extreme cases
that the courts have thought it right to relieve a negligent defendant of liability in
order to express disapproval of illegal conduct on the part of the claimant.
Where the claimant and the defendant are jointly involved and co-operating in
illegal activity, one approach is to bar the claimant from recovery only if the nexus
between the act of negligence and the illegal activity is such that the standard of care
owed in the particular circumstances could only be determined by taking into
account the illegal nature of the activity in which the parties were engaged.83 If a
thief is injured when a companion plants explosives in an allegedly negligent way to
blow a safe, the court will not inquire into whether the burglar alarm had sounded
or whether the police were on their way or whether the furtive nature of the occa-
sion made it inappropriate to apply to the defendant a standard of care which would
be appropriate to a lawful activity. The reason for this approach appears to be one

82 E.g. Kirkham v. Chief Constable of Manchester [1990] 2 QB 283; Saunders v. Edwards [1987] 1 WLR
1116; Pitts v. Hunt [1991] 1 QB 24, 45“6 per Beldam LJ; but note the reservations of Dillon LJ at
56. See also Rance v. Mid-Downs Health Authority [1991] 1 QB 487 (public policy would not allow
P to sue for loss of a chance to have an illegal abortion).
83 Ashton v. Turner [1980] 3 All ER 890; Pitts v. Hunt [1991] 1 QB 24 per Balcombe LJ.
Fault as a basis of liability 67

of ˜public policy™, but it is not clear what the policy is: it may be that if the courts were
to engage in such inquiries this would lower the respect felt for the courts; or some
vague feeling that if things go wrong in the course of criminal activities, even if by
the negligence of one of the criminals, the criminals deserve everything they get. To
this extent the compensatory aim of the law is subordinated to other values. This is
not altogether surprising because even when no-fault systems of motor accident
compensation are adopted, there is often much dispute as to whether persons
involved in criminal activities should be entitled to claim. The alleviation of need
and su¬ering regardless of fault is clearly an important part of our morality, but it
is unlikely that all elements of personal responsibility will ever be eliminated from
popular views about the proper way to deal with non-criminal injuries.
So far we have been discussing cases in which the claimant™s illegal conduct was
not a consequence of the defendant™s alleged tort. Suppose prison authorities neg-
ligently fail to prevent a person in their custody from committing suicide?84 Should
the person™s dependants be allowed to recover from the prison authorities? Or
suppose that as a result of injuries received in a car accident, a man™s personality
changes, he commits rape and is imprisoned?85 Should he be allowed to recover
from the negligent driver for loss su¬ered as a result of his crimes? Or suppose hos-
pital authorities discharge a mentally ill person who then commits manslaughter.86
Should the person be allowed to recover damages from the hospital authorities for
loss su¬ered as a result of his crime?
In this type of case, the question the courts ask is whether ˜ordinary people™
would be shocked and a¬ronted if damages were awarded, not whether the
claimant™s conduct was technically illegal. In the suicide case, the action was
allowed, but not in the manslaughter case. Damages were awarded to the rapist, but
this result was heavily criticised in the manslaughter case. At all events, the cases
demonstrate the importance of notions of personal responsibility in traditional
tort law. On the other hand, they might also be thought to provide evidence of the
role of liability insurance in extending the frontiers of tort liability.87 It is highly
unlikely that any of these actions would have been brought if the defendant in each
case had either not been insured or a public authority whose liabilities are under-
written by the taxpayer.

84 Kirkham v. Chief Constable of Manchester [1990] 2 QB 283 (P of unsound mind); Reeves v.
Commissioner of Police of the Metropolis [1998] 2 WLR 401 (P of sound mind).
85 Meah v. McCreamer (No. 1) [1985] 1 All ER 367. In Meah v. McCreamer (No. 2) [1986] 1 All ER
943 the court refused the rapist damages representing compensation he had been ordered to pay
his victims, but on the ground that the loss was too remote and not on the ground of illegality.
86 Clunis v. Camden and Islington Health Authority [1998] 3 All ER 180.
87 See further 9.8.

The Scope of the Tort of Negligence


. 3
( 18)