. 11
( 18)


93 JUSTICE, Report on the Trial of Motor Accident Cases, 31.
94 Zuckerman, Civil Procedure, 846“58.
Trials and settlements 281

On the other hand, it may be thought undesirable to establish formal mechanisms
and procedures speci¬cally designed to put pressure on parties to settle for
amounts that may diverge signi¬cantly from those that would be awarded by a
court after trial.
Nor is it necessarily the case that early settlements save on administrative costs.
There is evidence that procedural changes in recent years have encouraged earlier
settlement of claims; but also that they have done this by requiring more to be spent
on investigating and preparing the claim at an earlier stage than hitherto (˜front-
loading™ of costs, it is called). The obvious question this raises is whether earlier set-
tlement is worth the cost. The answer must depend in part on who bene¬ts from
earlier settlement and who, if anyone, loses out. For instance, if it turned out that
earlier settlements disadvantaged claimants at the expense of their lawyers, we
might doubt the wisdom of spending more to encourage it. In fact we lack the
information that would be necessary to resolve such issues.
Most personal injury actions that end up in court go to the county court rather
than the High Court. A claim likely to be worth less than £50,000 cannot be started
in the High Court; and, in practice, ˜it would be unusual for a claim worth less than
£250,000 to begin in the High Court unless it [had] some special features of
di¬culty™.95 Claims begun in either court can be transferred to the other in appro-
priate circumstances. Claims with a likely value of £1,000 or less will normally be
dealt with by ˜small claims™ procedure.96 E¬ectively, this is a form of ADR, more akin
to arbitration than to traditional court proceedings. Procedure is relatively infor-
mal, and less adversarial than normal court procedure. Parties rarely have legal rep-
resentation (except in road-accident cases, which represent about a ¬fth of all small
claims, including non-personal injury claims); and normally the successful party
can recover only very limited costs. On the other hand, small claims are resolved on
the basis of the relevant legal rules, and there is a right of appeal from the small
claims (˜district™) judge to a circuit judge (i.e. a senior county court judge). Research
suggests that those who use the small claims procedure are generally happy with it;
but that like all court proceedings, it presents a forbidding prospect to the average

10.5 The time taken to achieve a settlement
Preparing for and conducting the trial of a personal injury claim can often be a
lengthy business; but the process of negotiation can take considerable time even in

95 P. Barrie, Personal Injury Law: Liability, Compensation and Procedure, 2nd edn (Oxford, 2005),
para. 37.05.
96 The lower limit for most types of case is £5,000. Because of the much lower limit for personal
injury claims, only about 15% of personal injury claims that get to trial in the county court are
dealt with as small claims. The reason given for the low limit is the complexity and di¬culty of
personal injury claims. The House of Commons Constitutional A¬airs Committee has recently
recommended that the limit be raised to £2,000“2,500: First Report, 2005“6, para. 53. On small
claims generally see J. Baldwin, Small Claims in the County Court in England and Wales: The
282 Chapter 10

cases that are settled.98 From the outset, it is important not to assume that time
spent on negotiating and litigating is necessarily time wasted, or automatically to
equate the passage of time with ˜delay™. These processes necessarily take time, and
in the absence of accepted and reasonably precise criteria, it is hard to de¬ne how
long is too long.
The Pearson Report provided a good deal of information about the time taken
to resolve claims, both in cases which are dealt with by the courts and in cases which
are settled.99 As to the former group, the Commission found that in 1974 the
average time between the date of the injury and the date of disposal of the claim
was 36 months.100 However, cases took much longer in the High Court (where all
serious claims were then heard) than in the county courts. In the High Court, the
average time between injury and disposal of the claim was 43 months in London
and 41 months in the provinces; in the county courts the ¬gure was 21 months.
A survey conducted for the Civil Justice Review,101 which sampled cases started in
1980“2 or tried in 1984, found that the average time from the incident giving rise
to the action to the trial was more than 5 years in the High Court and almost 3 years
in the county court. In 65% of cases studied, proceedings had not started within a
year of the incident, and in 19% of cases they were started more than 21„2 years after
the incident. Solicitors interviewed for the survey identi¬ed the main causes of
˜delay™ as waiting for the claimant™s medical condition to stabilize, waiting for
medical reports and waiting for trial.102 The ¬gures, however, conceal the wide vari-
ations between the shortest and the longest delays. The Cantley Working Party
found that among cases in which a High Court writ was issued in 1977, the average
time from injury to disposal was 45 months, and no fewer than 122 cases (out of a
total number of 5,701) were still outstanding 8 years after the injury.103
As for settled cases, according to the insurance survey conducted for the Pear-
son Commission, nearly half of all claims were disposed of within 12 months,
though it is plain that these must have been the less serious cases because they only
account for under a quarter of the total payments. By the end of 2 years from the
injury over 80% of claims had been settled; but nearly 5% took up to 4 years, and
nearly 1% (some 2,000 cases a year) were still unsettled after 5 years.104 These
¬ndings generally con¬rm the pattern found by other less comprehensive studies.105

Bargain Basement of Civil Justice? (Oxford, 1997).
97 J. Baldwin, ˜Litigants™ Experiences of Adjudication in the County Courts™ (1999) 18 CJQ 12.
98 Genn, Hard Bargaining, 100“8.
99 The distinction between litigated cases and settled cases is blurred in those cases in which the
settlement process itself involves the initiation of proceedings.
100 Pearson Report, vol. 2, table 129.
101 Cm 394, 1988.
102 The average waiting time from setting-down for trial until trial was 65 weeks in the High Court
in London, 54 weeks outside London and 2 months in the county court: Civil Justice Review,
para. 433.
103 Report of the Personal Injuries Litigation Procedure Working Party (Cmnd 7476, 1979) (the
Cantley Committee Report), Appendix E. See also Civil Justice Review, para. 414(ii).
104 Pearson Report, vol. 2, table 17.
Trials and settlements 283

Furthermore, the larger the settlement, the longer it takes to negotiate. The Pearson
Commission™s insurance survey found that 96% of payments under £500 were made
within 2 years, but only 42% of payments over £10,000 were made within this
time.106 And in the most serious cases of all, where payments over £25,000 were ulti-
mately made, only 27% were settled within 2 years. In exceptional circumstances,
settlements may take many years. One of the causes which led to the establishment
of the Pearson Royal Commission was the realization that many of the Thalidomide
cases were still unsettled after 10 years.
Delay in settling claims may be the result of ine¬ciency on either side or of
deliberate procrastination by insurers, who generally have less to lose by delay than
do claimants. An insurer has little incentive to keep the settlement process
moving107 except that which derives from the rule that interest is payable on
damages awards108 and from a desire to close ¬les. The only way the typical claimant
can put pressure on a dilatory insurer is to issue proceedings,109 the various stages
of which are subject to formal time limits (which may, however, be extendable by
the court).
The settlement process is also inherently time-consuming because of the
complex nature of the issues that arise in personal-injury actions and the
di¬culties that often attend the gathering of evidence.110 In cases of more serious
injury, a longer time must generally elapse before a ¬rm medical prognosis as to the
e¬ects of the injuries can be given; because tort settlements are normally in the
form of a once-for-all lump sum (even if the settlement is then structured), it is
often in the interests of both claimants and defendants to postpone ¬nal agreement
until the claimant™s medical condition has stabilized. Moreover, the larger the claim
the more prolonged and vigorous the haggling is inclined to be.111 The aim of the
insurer in the settlement process is to pay not what the claimant is legally entitled
to, but as little as possible. The smaller the claim, the more likely it is that the cost
of prolonged negotiation will outweigh the value of the claim, and this results in
small claims being paid more promptly. It is a sombre thought, and no credit to
the tort system, that the more serious a person™s injuries, the longer it takes for
the claim to be settled. If it were not for the social security system, which provides

105 See Winn Committee Report, Section IX and Appendix 13; ˜A New Breed of Settlor?™ [1967] New
LJ 198; Ison, The Forensic Lottery, Appendix C, 178“80; Harris 1984 Survey, 105“10.
106 Pearson Report, vol. 2, table 115; see also Ison, The Forensic Lottery, 179 and Law Com. No. 225,
How Much is Enough?, 70“1. The Harris 1984 Survey found a more direct correlation between
length of time and degree of residual disability than between time and size of award as such.
107 And, indeed, has reason to slow it down: T. Swanson, ˜A Review of the Civil Justice Review:
Economic Theories Behind the Delays in Tort Litigation™ [1990] Current Legal Problems 185,
108 Supreme Court Act 1981, s. 32A; but see Law Com. No. 56, para. 271.
109 Group actions are di¬erent. In such cases, lawyers have become very skilled at using media pub-
licity and other forms of pressure to encourage defendants and their insurers to make timely and
acceptable settlement o¬ers.
110 R. Dingwall, T. Durkin and W.L.F. Felstiner, ˜Delay in Tort Cases™ (1990) 9 Civil Justice Q. 353; R.
James, ˜The Causes and E¬ects of Delay in Personal Injury Claims™ (1985) 36 Northern Ireland
Legal Q. 222.
284 Chapter 10

many claimants with bene¬ts during the settlement process, the tort system would
probably have collapsed long ago. Compensation for lost earnings and for medical
expenses is needed when the earnings are lost or the expenses incurred, not 3 years
later. The e¬ects of delay on tort claimants are vividly portrayed in the following
During the litigation process the victim and the victim™s family face a long period of
¬nancial di¬culty during which debts accumulate, savings are reduced and legal costs
increase. These lengthy periods of ¬nancial hardship, in addition to physical distress,
a¬ect the degree of satisfaction with damages and attitudes to adequacy. The stress of
litigation and apprehension about the future may impede recovery, which has an
impact on the probability that accident victims will eventually return to the workforce.
Delay in receiving damages also in¬‚uences decisions about the use of money as acci-
dent victims seek to make up for the material comforts that they lacked in the years
while waiting for the case to be settled.

One of the main aims of the Woolf reforms was to encourage earlier settlement by
requiring lawyers to investigate and prepare claims at an earlier stage than formerly.
One result appears to have been an increase in the average cost of settling claims.
That earlier settlement is worth the cost is assumed rather than demonstrated.

10.6 The amount of compensation
In Part Two we devoted a good deal of space to the rules of legal liability and the
principles on which damages for personal injuries are assessed by the courts. Now
that we have seen how settlements are arrived at in practice, it will be appreciated
that in cases which are settled, the legal principles laid down by the courts do not
by themselves determine how much, if anything, will be recovered. Naturally, the
law relating both to liability and the assessment of damages plays an important part
in in¬‚uencing the parties to a settlement. But the actual result is arrived at in a very
di¬erent way: the object of a settlement is not to arrive at the result which a judge
would probably arrive at according to the established legal principles.
In the ¬rst place, when a case is tried in court, the judge actually decides the
issues of liability and assessment of damages that are in dispute between the
parties. By contrast, when a case is settled by negotiation these things are never
¬nally determined. There is merely a greater or lesser probability that, if the case
were litigated, the judge would decide in this or in that way. Accordingly, in prac-
tice the damages are usually discounted to a greater or lesser extent according to
the parties™ estimates of the probabilities. If the claimant™s case is extremely strong,
it may be settled on the basis of full compensation; but where there is any doubt
about the facts, or where the relevant legal rules are uncertain or of uncertain
application, the damages which the parties estimate that a court would award if

111 Swanson thinks that the main problem in many cases is a lack of bargaining: ˜A Review of the
Trials and settlements 285

the claimant succeeded at trial are likely to be discounted by some percentage.
A settlement is a business bargain in which the claimant sells a claim to an insurer
for the best price on o¬er, which will be as little as the insurer can get away with.
The claimant will often be (reluctantly) willing to sacri¬ce some part of even a
good claim in order to reduce the risk (which can never be wholly eliminated) that
the action may fail wholly or in part.113 Furthermore, a settlement may be further
discounted if the claimant receives the money signi¬cantly sooner than if the case
had gone to trial.
Where the claim is less strong, or where there is the possibility of an adverse
¬nding on contributory negligence, the probable damages may be discounted quite
heavily. The settlement arrived at in the Thalidomide cases illustrates this proce-
dure very well: because of doubts about proving negligence against the defendants
and also about the validity of the claim as a matter of law,114 the claimants were
advised by their counsel to accept a settlement of 40% of ˜full compensation™, and
this was approved as a fair and reasonable settlement by the court. According to
Pearson Commission ¬ndings, over a quarter of cases settled with insurers are dis-
posed of on the basis of partial liability only, that is, on the basis that the claimant
must have been guilty of contributory negligence.115 In serious cases, in particular,
the e¬ect of this process of discounting may be that the compensation paid is inad-
equate to replace income lost, and to meet extra expenses incurred, as a result of
the injuries su¬ered. Typically, nothing can subsequently be done about this. The
inadequacy of the compensation is the price paid by the claimant for being relieved
of the risk of receiving even less or nothing at all if the case went to trial. In the case
of Thalidomide, however, both the government and the drug manufacturer subse-
quently (in 1996) gave additional funds (£7 million and £37 million respectively)
to the victims.
There are some grounds for believing that claimants tend to recover a larger pro-
portion of what they have ˜lost™ in cases of minor injury.116 The Pearson Report does
not provide statistics directly relevant to this issue, but it points in the same direc-
tion. For instance, it is clear from the Pearson ¬ndings that the proportion of the
settlement attributable to non-pecuniary loss is much higher in small claims.
Indeed, for claims of up to £5,000 (in 1973) over two-thirds, and in many minor
cases over 70%, was for non-pecuniary loss. For larger claims, the proportion

Civil Justice Review™, 190“1.
112 Law Com. No. 225, How Much is Enough?, xxi.
113 Research suggests that the most common reason why claimants accept o¬ers which they feel, or
come to feel, are inadequate is because they are advised to accept: Law Com. No. 225, How Much
is Enough?, 86. Other reasons are the amount of time already taken pursuing the claim, fear of
the costs of carrying on, and the e¬ect on their health.
114 At the time these cases were being settled there was doubt as to whether a duty of care could be
owed in respect of injuries su¬ered by an unborn child in the womb.
115 Pearson Report, vol. 2, table 117.
116 Conard, Automobile Accident Costs and Payments, table 6“14, 197; table 5“13, 179; ¬g. 5“11, 177.
The Osgoode Hall Study does not show such dramatic di¬erences between more and less serious
cases, though it shows serious under-compensation in fatal cases (ch. IV, table IV-6). The Harris
286 Chapter 10

attributable to non-pecuniary loss drops to around 50%.117 Since the maximum
amount normally awarded by a court for non-pecuniary loss is in the region of
£220,000 (in 2005), the proportion of the largest awards and settlements attribut-
able to non-pecuniary loss may be much lower even than these ¬gures suggest.
Sums paid in settlements for non-pecuniary loss may help compensate for the dis-
counts made on account of the risks of litigation and contributory negligence. The
Pearson ¬gures suggest that this happens to a greater extent in less serious cases.
Intuitively, one would expect small claims to be contested less vigorously by
insurers, both as to liability and as to quantum, than large ones because the smaller
the claim, the greater its nuisance value; or, in other words, the greater the likeli-
hood that the cost of contesting it will be greater than the amount of the claim.
There is no necessary relationship between the size of a claim and how di¬cult and
complex it is in legal terms. Large claims may be very simple and small claims very
complex. But the smaller the claim, the more likely that resisting it will not be
¬nancially worthwhile. Insurers do not like large claims and will investigate and
resist them with vigour.118 On the other hand, however weak the claimant™s case, it
may still pay the defendant to settle a small claim rather than to ¬ght. For reasons
of cost and administrative e¬ciency, insurers typically settle low-value claims on
the basis of rules-of-thumb “ such as that rear-end accidents are always the fault of
the following driver “ rather than a thorough investigation of the issue of fault.119
Such rules-of-thumb are not only simpler and cheaper to apply than the legal
concept of fault, but they also tend to be more favourable to claimants than strict
rules of law. If the claimant™s solicitor appreciates this, the claimant™s bargaining
position is improved. In practice, few claims are totally worthless so long as there is
at least some room for argument about fault. Because there are many more small
claims than large ones, the tort system has an inherent bias in favour of those with
low-value claims at the expense of the more seriously injured.
Group claims, once again, deserve special attention. In such cases solicitors may,
by making realistic threats of instituting proceedings in the USA, secure settlements
at mid-Atlantic rates. Even in cases with no US element, solicitors may secure for
group claimants sums in respect of non-pecuniary loss higher than prevailing rates
and even, in fatal cases, higher than the statutory maximum of £10,000 for bereave-
ment; or compensation beyond the limits laid down by international Conventions
dealing with air and sea disasters. In the case of very large group claims (such as the
claim against British Coal mentioned in n. 77) a non-judicial administrative
process may be established to assess compensation in individual cases. In a discus-
sion of the British Coal scheme in later related litigation120 the judge noted that
although compensation under the scheme was calculated according to common
law principles, the scheme was constructed in such a way that some awards would

1984 Survey found that there was no clear correlation between likelihood of obtaining damages
at all and degree of residual disability or amount of time o¬ work: 56“8.
117 Pearson Report, vol. 2, table 108.
118 Genn, Hard Bargaining, 69.
Trials and settlements 287

exceed and others would fall below what a court would award in individual cases.
Nevertheless, the scheme received judicial approval because the level of inaccuracy
in individual cases was more than o¬set by savings in administrative costs and in
court time. In fact, there were so many claims (more than 740,000) that it would
have been impossible for the court system to deal with even a small proportion of
The result of all this is that in cases which are disposed of by settlement “ and
this probably means about 99% of all claims “ the principles of law laid down by
the courts do not alone determine whether and how much compensation will be
paid. It is highly likely that the settlement process often leads to awards of com-
pensation higher (especially in low-value claims) or lower (especially in high-value
claims) than would be awarded in the courts. As Professor Conard and his col-
leagues at Michigan put it many years ago: ˜The statistics con¬rm what every lawyer
and adjustor knows “ the questions about negligence, proof, the defendant™s ability
to pay, and the client™s desire for an end of litigation, lead to compromises of claims
at levels which correspond to no theory of legal right.™121
These criticisms of the settlement process usually stimulate demands for reform
of that process, so as to improve the position of the claimant.122 But it must always
be borne in mind that the claimant with a tort claim is already in a very privileged
position compared with the great majority of the injured and the disabled. Thus, it
is possible to make criticisms of the settlement process not with a view to its
improvement but with a view to the abolition of the tort system of which it is a
major part.
119 R. Lewis, ˜Insurance and the Tort System™ (2005) 25 LS 85, 89.
120 AB v. British Coal Corporation [2004] EWHC 1372 at [325]“[328].
121 Conard, Automobile Accident Costs and Payments, 199.
122 E.g. Genn, Hard Bargaining, 169.
Trials and settlements 289
Part 4

Other compensation systems

First-party insurance

11.1 Types of ¬rst-party insurance
Under a ˜third-party™ or ˜liability™ insurance policy one person (the ˜¬rst party™ we
might say) is insured by the insurer (the ˜second party™) against the risk of being
held legally liable to another (the ˜third party™). Under a ˜¬rst-party™ or ˜loss™ insur-
ance policy the policy holder (the ¬rst party) is insured against the risk of su¬ering
loss speci¬ed in the policy by causes de¬ned therein. Nearly all accident risks can
be covered by ¬rst-party insurance of one kind or another. Life insurance, which
usually covers death from any cause except (in some cases) sane suicide, is by far
the most common form of ¬rst-party insurance against risks to the person (i.e.
death from personal injury and other causes). In 1996“7, 61% of UK households
were paying for some life insurance.1 A signi¬cant proportion of life insurance is
mortgage-related “ that is, it provides security against the death of the mortgagor.
Total UK premium income in 1997 for life insurance was about £28 billion.2 The
popularity of life insurance is partly attributable to the fact that it is used as a form
of investment, not merely as a protection against risks; but also partly to the fact
that premiums for this type of insurance are low relative to the bene¬ts provided.
One reason for this is that the administrative costs of life insurance are low: ˜. . . the
characteristics of death facilitate the administration of life assurance on a volun-
tary and individual basis, for death is inevitable, non-recurring, readily ascertain-
able when it occurs, is something which the assured is generally anxious to avoid as
long as possible, and is a matter of public record and inquiry™.3 There is little scope
for fraud or abuse4 on the part of those claiming life insurance bene¬ts.
Property insurance is very common in the form, for example, of household
contents and all risks cover. In 1996“7, 74% of UK households had some con-
tents insurance. Total UK premiums for property damage insurance in 1997
amounted to almost £8.3 billion.5 Motor insurance is also very big business. Gross

1 Association of British Insurers, Insurance: Facts, Figures and Trends 1998 (London, 1992), 9.
2 Ibid., 10.
3 Ison, The Forensic Lottery, 43.
4 Insurers call this ˜moral hazard™. Suicide is the only common form of abuse, but this is relatively
very rare.
5 Insurance: Facts, Figures and Trends 1998, 20.

292 Chapter 11

motor insurance premium income in the UK in 1997 amounted to some £7.6
Insurance against injury and illness takes a number of forms. Personal accident
insurance (PAI) is designed to cover short-term loss of earnings.7 Bene¬ts under
PAI policies are in the form of lump sums and sometimes periodical payments for
a ¬xed period. Critical illness insurance protects against the risk of conditions such
as cancer, heart attacks and stroke; bene¬ts are usually in the form of a lump-sum
payment. Income protection insurance (IPI) (formerly known as permanent8
health insurance) is designed to insure an income where the insured su¬ers long-
term sickness or disability. Private medical insurance (PMI) covers the cost of non-
NHS medical treatment (primarily hospital treatment and surgery). Both IPI and
PMI policies are taken out by individuals and by employers on behalf of employ-
ees through group policies. Long-term care insurance is designed to cover the cost
of care in the insured™s own home or in a nursing home made necessary by per-
manent illness or disability. It is marketed as a private supplement to care-related
social security bene¬ts and local authority care services.
Compared with that for life insurance, the market for insurance against injury
and ill-health is small. In 1997, UK premium income from IPI was around £389
million (representing some 1.3 million individual policies)9 and around £400
million from group policies. In 1997, UK premium income from PAI, IPI and PMI
together was around £3.3 billion. In 2001 only 15.4% of adults were covered by
PMI.10 Relatively few people insure themselves in the private market against dis-
ability which prevents them earning a living for weeks, months or even longer;11
nor against partial disability, which may deprive a person of some of the pleasures
of life, such as the ability to drive or dance or play sports, or which may in¬‚ict
chronic pain and discomfort.12 People whose whole living depends on some par-

6 Ibid.
7 Research conducted for the Law Commission in 1992“3 found that only one in ten recipients of
tort damages surveyed had received any payments under personal accident insurance policies:
Law Com. No. 225, How Much is Enough?, 155.
8 In the sense that ˜once an insurer has accepted a proposal, the terms of the policy are that renewal
cannot be refused until some pre-arranged expiry date, usually the retirement of the insured™:
P. Cor¬eld, ˜Private Insurance™ in Harris 1984 Survey, 222.
9 T. Burchardt and J. Hills, Private Welfare Insurance and Social Security (York, 1997), 13.
10 O¬ce for National Statistics, The Demand for Private Medical Insurance (2004). About equal
numbers of PMI policies are paid for by individuals and by employers respectively. Individuals
with high incomes are much more likely to be covered than individuals with low incomes; and
individuals in the top income percentile are very much more likely to be covered by employer-
purchased PMI than any other group. People in professional, managerial and technical occupa-
tions are more likely to be covered by PMI than people in unskilled occupations.
11 For a discussion, in the US context, of why insurance against disability is relatively rare see
K.S. Abraham and L. Liebman, ˜Private Insurance. Social Insurance and Tort Reform: Towards a
New Vision of Compensation for Illness and Injury™ (1993) 93 Columbia LR 75, 101“5.
12 Life insurance is much more common than disability insurance despite the fact that ˜at most ages
the risk of premature death is small by comparison with risks of illness or injury . . . The explan-
ation is perhaps to be sought in some peculiarity in attitudes to risk or more probably in imper-
fections in the private market for personal insurances against illness, injury and death™: Cor¬eld,
˜Private Insurance™, 224; see also 230“3.
First-party insurance 293

ticular ability sometimes have substantial personal accident insurance; for instance,
a musician may insure their hands, a footballer their legs, a ¬lm star their facial
appearance. Egon Ronay (the famous food critic) reportedly insured his taste buds!
Many comprehensive motor insurance policies contain an element of such insur-
ance, but the bene¬ts provided are usually very low.13 Similar cover is sometimes
included in other special ad hoc types of insurance, such as holiday insurance.
Personal accident insurance is often of short duration, being taken out to cover only
a particular event or risk; and many policies cover only a particular risk such as
being injured while playing sport.14 Accident insurance is an unattractive way of
insuring against income loss because the risk of accident is very low compared with
that of illness. What most people need is some form of life insurance to provide
bene¬ts on death plus some form of insurance against income loss, from whatever
cause, while they are alive.
The relative unpopularity of income protection insurance to provide cover
against income loss is due partly to the fact that it is quite expensive; partly to the
fact that bene¬ts under such policies are usually only payable for a relatively short
¬xed period;15 and partly to the fact that the risk of long-term income loss as a result
of ill-health is su¬ciently small that most people fail to appreciate it or are prepared
to run it. More importantly, perhaps, there are other sources of compensation for
income loss. The social security system provides bene¬ts for both short-term and
long-term incapacity for work. These bene¬ts are low, and many employees, espe-
cially the higher-paid, would be in serious ¬nancial di¬culties if social security
were the sole source of income replacement. Very many employed workers (as
opposed to self-employed workers) are covered by occupational sick pay schemes.
The Pearson Commission concluded that about half of those injured by accidents
who su¬er some loss of pay receive some sick pay from their employers, and that,
in aggregate, these sums totalled about £125 million per annum (in 1977 currency
values).16 In 1981 it was estimated that some 90% of employees participate in some
form of voluntary (or ˜occupational™) sick pay scheme.17 According to a 1985 survey
by the Social Security Policy Inspectorate,18 public employees and o¬ce and man-
agerial sta¬ are more likely to receive occupational sick pay than those working in
industrial, construction or transport ¬elds; larger ¬rms are more likely to have a
scheme than smaller ¬rms; and lower-paid employees are less likely to receive occu-
pational sick pay than well-paid ones.

13 Ibid., 224“5.
14 In Van Oppen v. Bedford Charity Trustees [1990] 1 WLR 235 it was held that a school has no duty
to take out insurance to cover children playing sport for the school or to advise parents of the
wisdom of purchasing such insurance. A specialized variant of accident insurance is insurance
against the risk of giving birth to a handicapped child.
15 The same is true of related forms of insurance such as mortgage protection insurance, which pro-
vides cover against the risk of not being able to keep up mortgage payments as a result of an unex-
pected reduction of income.
16 Pearson Report, vol. 1, paras. 137“41; see also Cor¬eld, ˜Private Insurance™.
17 See HC Debs, 6th Series, vol. 13, cols. 642“3 (23 November 1981).
18 Inquiry into Statutory Sick Pay (HMSO, 1985).
294 Chapter 11

There are wide variations in the conditions of entitlement and the amounts
payable under sick pay schemes, and there are also wide variations between di¬er-
ent industries. In some cases, full pay is granted; in others half pay; and in yet
others a ¬‚at-rate payment is made. Sick pay may be reduced to take account of
social security bene¬ts received by the employee. The 1985 Social Security Policy
Inspectorate survey found no employer which paid statutory sick pay19 in addition
to occupational sick pay when the latter was equivalent to the normal rate of pay.
Often a waiting period of a few days is prescribed before entitlement accrues. A
qualifying period of service with the employer is also often required. There are also
wide variations in the length of time for which sick pay is payable. Research done
for the Law Commission in 1992“3 found that fewer than one in ¬ve of the victims
surveyed received full pay for the entire time they were o¬ work as a result of their
injuries, and a small proportion received nothing.20 By contrast, in 67% of the fatal
cases surveyed, relatives received insurance money, thus con¬rming the popular-
ity of life insurance.21
A person who is so severely injured or disabled or sick that they are forced to
retire prematurely from employment may be entitled to a pension under an occu-
pational pension (or ˜superannuation™) scheme or may have a personal pension.
Some 55% of full-time male employees and some 56% of full-time female employ-
ees are members of an occupational pension scheme.22 Membership is more likely
the longer an employee has worked for the current employer, the higher the
employee™s wages, the larger the employer™s establishment and if the employee™s job
is non-manual.23 In both the public and the private sector, accrued pension rights
are normally una¬ected by premature retirement for sickness or ill-health, and are
payable as of right. Indeed, a large proportion of those who are members of such
schemes would get enhanced pension rights if they had to retire on grounds of ill-
health. In the public sector, 5 years™ service usually quali¬es the employee for such
bene¬ts, and most private schemes are at least as generous. Some schemes provide
bene¬ts that vary with the degree of incapacity, higher bene¬ts being payable in
cases of serious disability. Some schemes also provide bene¬ts for spouses of
employees who die in service. In addition to employees who belong to an occupa-
tional pension scheme, some 23% of men and 16% of women in full-time employ-
ment have a personal pension. Overall, some 68% of men and 65% of women in
full-time employment have some form of pension.
The main gap in private provision for insurance against income loss occurs
where entitlement under a sick pay scheme runs out before entitlement under a
pension scheme arises. It is here that income protection insurance might be

19 This is a social security bene¬t payable in the ¬rst instance by the employer, who may be able to
recover some of the amounts paid from the DWP: see
20 Law Com. No. 225, How Much is Enough?, 134“5.
21 Ibid., 242.
22 Social Trends 35, table 8.6 (2003“4).
23 M. Lunnon,˜New Earnings Survey data on occupational pension provision™ [1998] Labour Market
Trends 499.
First-party insurance 295

thought to have a role to play, but it has not really caught on even in this context;
although some employers do take out this form of insurance for their employees to
cover the gap.24
Not every risk protected against by the tort system is insurable in practice. For
instance, it is hardly possible to protect oneself by insurance against bereavement
as such. It is true that a person may insure their life for an amount which is much
greater than their income would warrant, so that if the person dies prematurely
their spouse or partner and children may to some extent be compensated in money
for the loss of the person as well as for loss of income; but few people are likely to
take such considerations into account in deciding how much life insurance to buy.
Moreover, the Industrial Assurance and Friendly Societies Act 1948 prohibits insur-
ance of the life of a child under the age of 10 which provides for any bene¬t other
than the return of the premiums. This Act is a reminder of the grisly fact that in
some circumstances life insurance on the life of a young child might be a tempta-
tion to homicide or at least neglect.
War risks are excluded from most insurance policies as a matter of course. Today
it has become customary for the State itself to take on the business of providing
war-damage insurance and of compensating those who su¬er personal injury, and
the dependants of those killed, as a result of enemy action. Other uninsurable risks
are also taken care of by special means. For instance, personal injury caused by
rioting would probably fall within the Criminal Injuries Compensation Scheme; in
the case of damage to property, compensation may be payable under the Riot
(Damages) Act 1886. Damage done by nuclear installations is the subject of a set of
legal rules and insurance practices designed to spread the cost of such risks as
widely as possible while retaining the compensation process in private hands.

11.2 First-party insurance compared with tort liability
First-party insurance di¬ers from tort compensation in many important ways.
First, it is almost entirely optional. There are no legal requirements to insure one™s
life, person, earning capacity or property with a private insurer. Moreover, a person
who buys ¬rst-party insurance can choose the amount of coverage desired. It may
be thought that some degree of compulsion would be a good thing, even in the area
of damage to property. Insurance against ¬re damage to houses is, in many cases,
compulsory in practice because building societies and other mortgagees require the
property to be insured against such risks to protect their security. But mortgages
are eventually paid o¬, and insurance policies may then be allowed to lapse. Grave
hardship would obviously be caused to most people if their houses were destroyed
or seriously damaged by ¬re and they had no insurance cover. But the case for com-
pulsion is much less strong in relation to other property such as cars. Nobody who
runs a motor vehicle can be unaware of the requirement of compulsory third-party

24 Pearson Report, vol. 2, para. 144.
296 Chapter 11

liability insurance, and this means that the motorist is very likely to be made aware
of the choice between comprehensive cover and liability cover alone.
So far as insurance against lost income is concerned, there is an element of
compulsion in the sense that social security bene¬ts for income loss are funded
out of compulsory taxes and National Insurance contributions. Moreover, many
employees are required by their employers to belong to sick pay and pension
schemes as a term of their employment. But the State imposes no legal compul-
sion on individuals to insure privately against income loss, and relatively few people
do so.
A second major contrast between ¬rst-party insurance and the tort system is the
fact that in the case of ¬rst-party insurance the form of compensation normally
depends on what has been lost. Thus, loss of a capital asset (e.g. destruction of a
house or car) will be met by payment of a lump sum; whereas loss of income will
normally be met by periodical payments if the loss continues long enough to make
this desirable. There is, moreover, more scope for correcting mistakes in the case
of personal insurance than under the tort system. We have seen how tort dam-
ages must be assessed once-and-for-all, and how this may lead to over- or under-
compensation if things do not turn out as anticipated. Because an insurance policy
is a contract, it can contain provision for the reduction, for example, of compensa-
tion for loss of earnings if the insured™s earning capacity unexpectedly improves.
Again, if a person insures a valuable piece of jewellery against loss or theft and
payment is made under the policy, this usually has the e¬ect of transferring the legal
title to the jewellery from the owners to the insurers. If, therefore, it turns up again
later, the insurance company can claim it and the former owner cannot keep both
the insurance money and the article insured. In most cases, the article will be
o¬ered back to the owner on condition that the policy money is repaid.
A third point of di¬erence between tort compensation and ¬rst-party insurance
is that the latter often does not provide ˜full compensation™. The amount of coverage
is usually optional, but there are many types of insurance in which the standard
policy requires the insured to bear part of the loss either by imposing an upper limit
(or ceiling) on the bene¬ts payable or by requiring the insured to pay the ¬rst slice
of any claim (called an ˜excess™ or ˜deductible™). Such provisions are a very common
feature of comprehensive motor insurance policies, but they are also found, for
example, in householder™s policies (at least for some risks) and income protection
policies (which do not usually cover the ¬rst month or 3 or 6 months of lost income,
and normally limit the income loss insured against to three-quarters of the insured™s
normal earnings). Ceilings and deductibles are designed to reduce the risk of fraud-
ulent claims; to encourage the insured to take precautions against the occurrence of
the insured loss; and to reduce the cost of the insurance. The way ceilings achieve this
third aim is obvious. The basis of deductibles is that small claims are very much more
common than large ones, and the cost of processing them is high relative to their
value, so that by excluding them from the cover the cost of the insurance can be
signi¬cantly reduced. Under the tort system, the principle of full compensation rules
First-party insurance 297

out ceilings or thresholds;25 but this is not necessarily in the public interest. Money
spent on tort compensation is money not available for other purposes, and however
fair the full compensation principle may seem as between individual claimants and
defendants, it may seem less desirable when viewed in a wider social context.
A fourth major di¬erence between tort liability and ¬rst-party insurance, as we
have seen,26 is the fact that negligence on the part of the insured will often not a¬ect
a ¬rst-party insurance claim. However, an insurance policy may exclude from cov-
erage losses arising from speci¬ed events akin to contributory negligence “ for
example, driving while drunk in the case of a comprehensive motor policy. Delib-
erate acts that bring about the event insured against are not normally covered
because claims based on such conduct are usually fraudulent. Where there is no
fraud “ as often in the case of suicide under life policies “ liability is usually accep-
ted. Moreover, suicide is a cause of death re¬‚ected in the mortality tables on which
premiums are based.
Despite these di¬erences, there are certain similarities between tort liability and
¬rst-party insurance. A ¬rst-party insurance policy will give cover only against
certain de¬ned risks. Cover may be limited to events occurring in speci¬ed places
or at speci¬ed times or in speci¬ed ways. Such limitations may give rise to disputes
between insurer and insured as to whether the event on which a claim is based fell
within the risk or occurred in the circumstances de¬ned in the policy. Disputes of
this sort may turn on questions of interpretation of the words of the policy (exactly
what risks does it cover?) but they may also raise issues very similar to issues of
causation which arise in tort cases (how did the loss occur?).
In theory, there is no reason why non-pecuniary loss should not be recoverable
under a ¬rst-party insurance policy as it is under the tort system. But, with the
exception of life assurance, which may be for a sum much greater than the assured™s
earning capacity would warrant, non-pecuniary loss insurance is very uncommon.
Accident insurance policies, such as travel insurance and household insurance, may
provide limited cover for personal disabilities, but the sums o¬ered are usually
small and ¬xed in amount and can, anyway, be seen as designed to meet expenses
rather than to compensate for intangible loss. It is reasonable to suggest that there
is little public demand for insurance against pain and su¬ering or loss of amenity
as such.27 This is really not very surprising: monetary compensation is not a great

25 Liability insurance (even if compulsory) may impose a ceiling. For example, the sum insured
under the Employers™ Liability (Compulsory Insurance) Act 1969 must be £5 million per incident.
Any excess may be insured separately, perhaps with several insurers in layers; or it may be borne
by the insured. Similarly, a (compulsory) liability insurance policy may provide for an excess or
deductible. E.g., under the Road Tra¬c Act 1988 the ¬rst £300 of third-party property damage
liability does not have to be insured against, and the MIB will not pay the ¬rst £300 of any claim
(but the EU 5th Motor Insurance Directive of 2005 requires abolition of this deductible). In some
jurisdictions in the USA and Australia, lower and upper limits on various heads of damages have
been imposed by statute in an attempt to limit the cost of the tort system.
26 9.5.
27 But see S.P. Croley and J.D. Hanson, ˜The Non-Pecuniary Costs of Accidents: Pain and Su¬ering
Damages in Tort Law™ (1995) 108 Harvard LR 1785.
298 Chapter 11

deal of consolation for loss of amenity.28 Yet the tort system, backed up by liability
insurance, in practice compels people to buy this kind of insurance for each other
even though, given the choice, few buy it for themselves. This casts doubt on the
desirability of retaining damages for non-pecuniary loss in the tort system.
This line of argument could be extended. The system of tort liability with (com-
pulsory) liability insurance requires people to pay indirectly for bene¬ts (or levels
of bene¬ts) which they might not choose to buy under a system of (optional) ¬rst-
party insurance. Furthermore, these bene¬ts are only available if the relevant losses
are caused by tortious conduct of another, whereas bene¬ts under ¬rst-party insur-
ance policies are, with minor exceptions, usually payable regardless of how the
relevant loss was caused. In this way, the tort system infringes people™s freedom of
choice. Extending the argument in this way perhaps shows that it proves too much:
few people buy income-protection insurance and the sort of cover which is freely
available is limited; yet this provides no good argument for not compensating
people, via the tort system, for permanent loss of income, and no reason not to
require people by law to insure others against the risk of income loss by taking out
third-party liability insurance.29
Indeed, the wider argument has force only if we compare the tort system with
optional ¬rst-party insurance. But, of course, the National Insurance system is a
form of compulsory ¬rst-party insurance which gives people no choice whether to
insure or about the level of bene¬ts. There are some losses which we feel justi¬ed
in requiring people to insure against. On the other hand, levels of social security
bene¬ts are quite low, and we do not force people to buy greater cover in the private
market. In this light, the questions we must ask ourselves are, which losses are we
justi¬ed in requiring people to insure against, and what is the most e¬cient way of
securing such insurance? It may be that the tort system requires people to insure,
in e¬ect, against losses which they should not be required to cover; and it is cer-
tainly a very ine¬cient form of insurance. But the mere fact that some, or even
many, people would not voluntarily insure against a particular loss does not make
compulsory insurance against that loss necessarily undesirable.
Insurance policies in respect of property only provide compensation for the
¬nancial value of the property, although this might be de¬ned as the cost of replac-
ing the property rather than its sale value at the time it was lost, damaged or
destroyed. Although such ˜new-for-old™ cover increases the risk of fraudulent
claims, it is now very commonly available to private individuals under household
contents policies. Compensation for property damage in the tort system is also
basically calculated according to the market value of the property at the time of the
loss or destruction; but in practice, the sum awarded is often the cost of repairing
or replacing the property. Normally cost of repair or replacement will not be
awarded if this is greater than the value of the property at the time of the tort; but
28 See also S.A. Rea, ˜Economic Analysis of Fault and No-Fault Liability Systems™ (1986“7) 12
Canadian Business LJ 444, 451“2.
29 See J. Stapleton, ˜Tort, Insurance and Ideology™ (1995) 58 Modern LR 820, 833“7.
First-party insurance 299

in exceptional cases the claimant may be awarded more than the market value of
the property.30
Compensation for lost pro¬ts arising from damage to or destruction of pro¬t-
earning property is also available under both the tort system and via ¬rst-party
insurance, and the assessment of compensation would follow similar lines in the
two cases.
One ¬nal point by way of summary: the tort system is much more important as
a source of compensation for personal injury, illness and death than as a source of
compensation for damage to or destruction of property. This is because damage to
or destruction of valuable property, about which it would be worth litigating,
whether caused by a tort or not, is very likely to be the subject of a ¬rst-party insur-
ance policy. On the other hand, although people commonly insure against the risk
of death, and although private insurance against the risk of personal injury or
illness is available, many people have little or no private insurance cover against
many of the losses for which the tort system provides compensation.

30 See further P. Cane, Tort Law and Economic Interests, 2nd edn (Oxford, 1996), 92“3.

Compensation for criminal injuries

12.1 Tort claims
Victims of criminal o¬ences causing personal injury can claim in tort against
the criminal, but in the great majority of cases the criminal would not be worth
suing.1 In its Third Report, the Criminal Injuries Compensation Board (CICB)
(now the Criminal Injuries Compensation Authority (CICA)) reported that it had
found the number of cases in which o¬enders would be worth suing to be seven-
teen, i.e. 0.7% of the cases which had by then come before the Board.2 By the time
of the Seventh Report the number had dropped to 6 out of nearly 6,000 cases, i.e.
about 0.1%.3 Just occasionally, tort liability may be established against someone
other than the o¬ender personally. For instance, the o¬ender™s employer may be
vicariously liable in cases of fraud. In cases of personal injury, such liability will not
often be established because criminal assaults and similarly violent conduct will not
often be regarded as within the course of the o¬ender™s employment (so as to
render the employer vicariously liable), but it may be in some cases.4 Personal tort
liability may occasionally be imposed on the victim™s employer, for example, where
it is alleged that the employer had exposed the victim to unnecessary risk of
criminal attack, for example, by sending a young woman to the bank to collect
1 A bizarre exception was a case in which two victims of sexual assaults brought a successful tort
action against their attacker (W v. Meah [1986] 1 All ER 935) who had himself been awarded tort
damages on the basis that his criminal activities were a result of his having been negligently
injured in a car accident (Meah v. McCreamer (No. 1) [1985] 1 All ER 367); the assailant then
unsuccessfully sued the driver to recover the amounts awarded to the two victims (Meah v.
McCreamer (No. 2) [1986] 1 All ER 943). Recourse to tort law by victims of violent crimes has
apparently increased in recent years. Most controversial are cases in which the person sued has
been prosecuted and acquitted, or where the prosecuting authorities have decided not to proceed
because of lack of evidence. The most notorious example is that of O.J. Simpson; but there have
been several successful tort claims in such circumstances in the UK. Critics argue that allowing
tort claims in such cases deprives the alleged o¬ender of the protections of the criminal law,
notably the higher burden of proof (˜beyond reasonable doubt™, not ˜on the balance of probabili-
ties™, as in civil law). The force of the argument lies in the fact that as a result of the victim™s success
in the civil action, the defendant may be ˜branded as a criminal™ in the public mind (see e.g. ˜Man
Branded as a Killer in Civil Ruling™, The Times, 25 March 1998).
2 Para. 21; CICB Fifth Report (Cmnd 4179, 1969), para. 11.
3 Para. 17; see also CICB Seventeenth Report (Cmnd 8401, 1981), paras. 53“4.
4 P.S. Atiyah, Vicarious Liability in the Law of Torts (London, 1967), esp. 262¬; F.A. Trindade and
P. Cane, The Law of Torts in Australia, 3rd edn (Melbourne, 1999), 736“9.

Compensation for criminal injuries 301

wages for a company™s employees.5 As a general rule, the police cannot be sued
by victims for allegedly negligent failure to apprehend criminals or to prevent
crime.6 On the other hand, prison authorities who negligently allow a criminal to
escape may incur liability for injury or loss resulting from crimes committed by the
From its inception until 1994, awards under the Criminal Injuries Compensation
Scheme (CICS) were assessed according to principles similar to the rules for assess-
ing damages for personal injury in tort cases.8 Now, however (as we will see later),
the CICS is what has been called an ˜enhanced tari¬™ Scheme. For a number of
reasons, awards under this ˜new™ enhanced-tari¬ Scheme may, in certain cases, be
less than could be obtained by suing in tort.9 This may encourage tort claims in some
situations where, hitherto, such action may not have been worthwhile.

12.2 Compensation orders
Under the Powers of Criminal Courts Act 200010 a magistrates™ court may award up
to £5,000 compensation for any personal injury, loss or damage resulting from an
o¬ence,11 and other criminal courts may award an unlimited sum.12 Compensation
orders are designed to be a summary remedy for use in simple cases. It has been
stressed that the victim of a criminal assault has a tort remedy by way of civil pro-
ceedings,13 and should be left to pursue that remedy in serious or complicated

5 Williams v. Grimshaw (1967) 3 KIR 610; Houghton v. Hackney BC (1961) 3 KIR 615; Charlton v.
Forest Printing Ink Co. [1978] IRLR 559.
6 Hill v. Chief Constable of West Yorkshire [1989] AC 53.
7 Home O¬ce v. Dorset Yacht Co. [1970] AC 1004.
8 As at October 2004, 425 claims made under the ˜old scheme™ were outstanding: Criminal Injuries
Compensation Appeals Panel, Annual Report, 2003“4, 9.
9 The maximum award under the new CICS is £500,000. In 2001 an almost completely paralysed
youth was awarded £4.75 million under the old Scheme. Earlier in 2001 an award of about £50,000
under the new Scheme to a young teacher, Sarah Potts, who was very seriously injured while pro-
tecting students from an attacker, caused considerable controversy: ˜Why the rules on compensa-
tion should be changed™, The Times, 20 February 2001.
10 Ss. 130“4. For discussion of the relationship between ˜reparation™ and punishment in the crim-
inal justice system see L. Zedner, ˜Reparation and Retribution: Are they Reconcilable? (1994) 57
Modern LR 228; T. Campbell, ˜Compensation as Punishment™ (1984) 7 U. of New South Wales
LR 338.
11 Orders can be made in favour of persons other than the direct victim of the crime.
12 See generally D. Miers, Compensation for Criminal Injuries (London, 1990), chs. 8“11; D. Moxon,
J.M. Corkery and C. Hedderman, Developments in the Use of Compensation Orders in Magistrates
Courts since 1988, Home O¬ce Research Study No. 126 (London, 1992); D. Tucker ˜Victims™
rights? “ wrong™ [1991] New LJ 192; C. Flood-Page and A. Mackie, Sentencing Practice: An
Examination of Decisions in Magistrates™ Courts and the Crown Court in the Mid-1990™s, Home
O¬ce Research Study 180 (1998). There is no lower limit on compensation orders, whereas there
is a lower limit on payments under the CICS.
13 Somewhat controversially, civil liability is not a precondition of the making of a compensation
order: Miers, Compensation for Criminal Injuries, 200“5; M. Wasik, ˜Compensation Orders and
Civil Liability™ (1985) 48 Modern LR 707. If the victim brings a civil action against the o¬ender
after a compensation order has been made, and the civil court holds that the injury, loss or damage
resulting from the o¬ence was less than the criminal court took it to be, the compensation order
may be discharged or the amount payable under it reduced.
302 Chapter 12

cases.14 Long-term orders for compensation, which may take years to pay o¬, are
generally discouraged.15
In 1988“9 over 100,000 orders were made by magistrates™ courts, and in 1989
more than 11,000 orders were made by Crown Courts. Orders were made by
magistrates™ courts in 39% of cases in which an order could have been made;16 and
in 17% of cases by Crown Courts. Looking just at o¬ences against the person,
orders were made in 46% of cases by magistrates™ courts and 28% of cases by Crown
Courts. The use of compensation orders has declined considerably since 1990. In
1990, 29% of defendants sentenced for an indictable o¬ence in a magistrates™ court
were ordered to pay compensation, but in 1996 the corresponding ¬gure was
19% and in 2003, 15%. In 1990, 21% of those convicted of a violent o¬ence in the
Crown Court were ordered to pay compensation, but in 1996 only 12%, and in
2003 only about 7%.17 Even when an order is made, enforcing it can be di¬cult.
This is a reason why compensation orders are rarely awarded in conjunction with
a custodial sentence.18 According to the Home O¬ce, only 1% of o¬enders sen-
tenced to immediate custody are ordered to pay compensation.19 Compensation
orders are not commonly made in conjunction with a ¬ne because, it is said,˜crimes
where there is likely to be a direct victim are far more likely to be punished with a
community sentence than a ¬ne™.20
The court is required to give reasons for not making a compensation order
(although in practice, this requirement is very often not complied with). The main
reasons that have been given are that the o¬ender lacked the means to pay com-
pensation,21 that the court lacked relevant information,22 that compensation was
not sought and that the relationship between the victim and the o¬ender made an

14 R v. Daly [1974] 1 All ER 290.
15 Ibid. Compensation orders for large sums may be made in exceptional cases. In 1996 an order
for some £160,000 was made against a thief who stole rare books from Oxford college libraries.
The court was told that the defendant had liquid assets of nearly £200,000: The Times,
17 February 1996.
16 Orders cannot be made in cases in which compensation would be available under an insurance
policy required by the Road Tra¬c Act 1988 or from the MIB.
17 It is not clear why the use of compensation orders has declined so much. It may be that fewer
o¬enders are being convicted, and of those who are being convicted, a greater proportion is impe-
cunious. (I owe these points to Professor Andrew Ashworth.)
18 M. Ogden, ˜Compensation Orders in Cases of Violence™ [1985] Criminal LR 500. Another reason
may be that to impose a custodial sentence and a compensation order may be thought unduly
severe: Home O¬ce Consultation Paper, Compensation and Support for Victims of Crime (2004),
12. In 2003, compensation orders were made in some 120,000 cases of indictable and summary
o¬ences (not counting summary motoring o¬ences). The compensation order was the sole or
main penalty in about only 8,000 of these.
19 Home O¬ce Consultation Paper, Compensation and Support for Victims of Crime (2004), 11.
20 Ibid., 12.
21 There is a statutory obligation to take account of the means of the o¬ender; whereas in the law of
tort, of course, this is irrelevant. The court is required by statute to give priority to a compensa-
tion order over a ¬ne, and a compensation order can be made on its own. But it appears not
uncommon for compensation to be reduced because of lack of means at the same time as a ¬ne
is imposed in addition.
22 Ss. 95 and 96 of the Courts Act 2003 (obligation to provide information about resources) may
help to overcome this problem.
Compensation for criminal injuries 303

order inappropriate because, for example, they were members of the same family.23
In relation to personal injuries, another important reason is lack of information
about the nature of the injuries24 and di¬culties in assessing appropriate amounts
of damages. There are also considerable problems in enforcing compensation
orders and collecting ¬nes.25
Both ¬nancial and non-pecuniary loss may be the subject of compensation. In
addition to compensation for the injuries themselves, compensation for distress
and anxiety arising from the o¬ence can be awarded.26 In practice, however, com-
pensation orders are less likely to be made in cases where the victim was not phys-
ically hurt. It seems that in many cases, orders in respect of personal injury fall far
short of full compensation. In 2003, the average amount of compensation ordered
by the Crown Court in cases of violence against the person was £608. Of 2,604
orders made in such cases, only 267 were for more than £1,000. In magistrates™
courts, the average compensation in cases of violence against the person was £245;
and of 7,748 orders made in such cases, only 190 were for more than £1,000.27
Awards under the CICS are reduced to take account of the proceeds of any com-
pensation order, damages award or settlement (2001 Scheme, para. 48);28 and recip-
ients of awards are required to repay to the CICA any such amounts received after
the CICS award is made (2001 Scheme, para. 49). In 2002“3 the CICA recovered
£209,000 from the proceeds of civil actions and £522,000 from the proceeds of
compensation orders in respect of awards made under the new (enhanced tari¬)
Scheme, which in that year totalled £160.3 million. Thus, criminals provided less
than 0.5% of the compensation paid out under the CICS in that year. In an attempt
to increase this ¬gure, in 2004, the Criminal Injuries Compensation Act 1995 was
amended to give the CICA the power to recover directly from convicted criminals
compensation paid to victims under the CICS.29

12.3 Other sources of compensation
Victims of crime may obtain compensation from other compensation systems.
Victims of personal injuries may bene¬t from the social security system, or occa-
sionally from personal accident insurance; in fatal cases a spouse or partner may

23 In 2002, a Crown Court judge reportedly made but then, in the face of public protest, withdrew,
a compensation order in favour of an alleged drug dealer who had been assaulted by the father of
one of his supposed clients: The Times, 17 May 2002.
24 See R v. Horsham Justices, ex parte Richards [1985] 2 All ER 1114.
25 Home O¬ce Consultation Paper, Compensation and Support for Victims of Crime, 12“13;
Compensation and Support for Victims of Crime: Summary of Responses to Home O¬ce
Consultation Paper (2004), 6“8.
26 Bond v. Chief Constable of Kent [1983] 1 All ER 456. Compensation for bereavement can be
awarded to persons entitled to such damages under the Fatal Accidents Act 1976 for an amount
up to the statutory maximum (currently £10,000).
27 These ¬gures come from Criminal Statistics for England and Wales 2003.
28 See 12.4.1 for the history of the CICS.
29 Ss. 7A“D, inserted by Domestic Violence, Crime and Victims Act 2004, s. 59.
304 Chapter 12

receive the proceeds of a life insurance policy or bereavement bene¬ts under social
security. Loss of property from criminal activity is quite likely to be covered by
private insurance. Occasionally, in cases of fraud, the government may provide
compensation, as after the collapse of the Barlow Clowes investment bank follow-
ing a recommendation of the Parliamentary Commissioner for Administration
(the Parliamentary Ombudsman).

12.4 Criminal injuries compensation scheme
12.4.1 Justi¬cations for the Scheme
Until the Criminal Injuries Compensation Scheme30 was set up in 1964,31 most
victims of violent criminal attack received little ¬nancial support other than through
the social security system. Many regarded this as inadequate, partly because of public
sympathy for the occasional victim of an especially vicious crime, and partly because
the social security system o¬ers no compensation for disability as such, except to
those injured at work. Nevertheless, the idea of selecting yet another group of injured
and disabled people for special treatment is not easily defensible, and it is hard to
believe that the government would have regarded the Scheme with much favour if it
had not been of relatively small quantitative signi¬cance, and therefore cheap com-
pared with most other claims for compensation which are made on the State.32
The discussions which led up to the Scheme reveal an extraordinary intellectual
confusion. The Home O¬ce Working Party which ¬rst looked into the matter, for
instance, rejected the idea that the State should assume the burden of compensat-
ing victims of criminal violence because it had a duty to protect citizens; this was a
˜fallacious and dangerous doctrine™, because the State could not possibly protect its
citizens from attack at all times and all places, and because, in any event, if there
was such a duty it would be impossible to con¬ne it to personal injury as opposed
to damage to property.33 The Committee went on to say that they could ¬nd no
˜constitutional or social principle on which State compensation could be justi¬ed™,
but they nevertheless thought that compensation could be based on the more ˜prac-
tical™ ground that ˜although the welfare state helps the victims of many kinds of mis-
fortune, it does nothing for the victims of crimes of violence, as such™.34 Accordingly,
although there was no ˜bounden duty™ to compensate these victims, such compen-
sation could be justi¬ed.

30 See generally D. Miers, State Compensation for Criminal Injuries (London, 1997).
31 The Scheme was last revised in 2001. It is worth noting that there have been six major reviews of
the scheme “ in 1973, 1983, 1991, 1999, 2004 and 2005. This is a good indication of its problem-
atic nature. Concerning the 2005 review, see the postscript to this chapter (p. 327).
32 P. Rock, Helping Victims of Crime (Oxford, 1990), 78, 84“5.
33 Cmnd 1406, 1961, para. 17. In 1985 it was said that ˜in present economic circumstances a central
fund to make public provision for victims of property crime is not practicable™: First Report of the
Home A¬airs Committee on Compensation and Support for Victims of Crime, HC 43 (1984“5); and
government Reply (Cmnd 9457, 1985), para. 19.
34 Ibid., para. 18.
Compensation for criminal injuries 305

The Committee never really came to grips with the crucial issue, which is not
whether victims of criminal violence ought to be compensated by the State, but
whether there are any grounds for giving such victims ¬nancial support over and
above social security bene¬ts available to others. The Committee did point out that
the Welfare State did nothing for the victims of crimes of violence ˜as such™. But why
this should matter, provided it does something for them? The working party
perhaps thought that social security bene¬ts were too low. If this is so, the right
solution is to increase bene¬ts across the board, not to provide extra bene¬ts for
particular groups of needy people at the expense of the generality.
A number of arguments for special treatment for victims of violent crime were
listed by a JUSTICE Report in 1962.35
• First, criminal injuries are analogous to war injuries, and the State has always accepted
some obligation to provide for those injured in war and their dependants.36 But there
is an obvious di¬erence between people who are requested or required by the govern-
ment to risk life and limb in war and ordinary victims of crime. At all events, bene¬ts
payable in respect of injury, illness of death attributable to military service are much
more generous than those available under the CICS in its current form.37
• Secondly, the State discourages citizens from carrying weapons for self-protection. But
this ignores the fact that citizens gain more than they lose from living in a generally
weapon-less society.
• Thirdly, citizens are under a moral duty to assist the State, for example, by going
to the assistance of a police o¬cer e¬ecting an arrest or suppressing violence, and
they may be deterred by the absence of compensation for injury su¬ered while doing
so. But why should altruism be any more rewarded when it takes this form rather than,
for example, that of trying to rescue a child from a burning house or a climber trapped
on a mountain?
• Fourthly, neglect of the interests of victims of crime makes it more di¬cult to adopt
enlightened penal policies because every demand for better treatment for criminals is
met by complaints that society is looking after the criminal better than the victim.38
This argument is based on political expediency, not on principle. On the other hand,
it may be said that while greater use of non-custodial sentences, for example, can be
justi¬ed in the interests of society as a whole as well as of o¬enders, it does involve risks;
and that those who su¬er as a result should be compensated by society. The general
principle underlying this approach, namely that individuals who su¬er extraordinary
loss as a result of the execution of government policy should be compensated, is highly

35 JUSTICE, Report on Compensation for Victims of Crimes of Violence (London, 1962).
36 An analogy was also drawn with riot damage which is, in some cases, the subject of State com-
pensation under the Riot (Damages) Act 1886.
37 The compensation scheme for members of the armed forces who su¬er injury, illness and death
attributable to military service has recently been overhauled: Ministry of Defence, A Framework
Document for the Armed Services Compensation Scheme for Injury, Illness and Death Due to Service,
(March 2004). For background and an account of the scheme that has been replaced see N.J. Wikeley,
Wikeley, Ogus and Barendt™s The Law of Social Security, 5th edn (London, 2002), ch. 21.
38 This was an important strand in the thinking of early proponents of victim compensation: Rock,
Helping Victims of Crime, ch. II.
306 Chapter 12

controversial and could only be accepted in this context if it could be shown that
victims of penal policy were particularly deserving.
• Fifthly, since the State prohibits the victim from ˜taking the law into their own hands
to obtain redress™, it should provide some e¬ective alternative. This argument proves
too much because ˜taking the law into one™s own hands™ is, with few exceptions, pro-
hibited generally and not just to victims of crime. The State accepts no general obliga-
tion to make good the lack of a defendant worth suing. Perhaps a more convincing
argument is that the risk of injury from criminal attack is so remote that it would be
impractical to insure against that risk and no other. There may be, therefore, something
to be said for the view that the whole population should be regarded as insured by the
State, and the compensation paid for by taxation.39 But this still fails to explain why this
kind of injury should be singled out for special treatment by the State: it is true of very
many victims of personal injury that they are not insured.
• Sixthly, o¬enders are often imprisoned for long periods, thus depriving the victim of
any chance of e¬ective redress. This argument is misleading. In 2003, for example, of
some 38,000 persons convicted of indictable o¬ences against the person in England
and Wales, some 11,500 received an immediate custodial sentence, while some 4,000
were ¬ned and another 4,000 received an absolute or conditional discharge. Some
10,300 persons convicted of o¬ences of violence against the person were made subject
to a compensation order in favour of their victim.40

Another line of argument seeks to set victim compensation in a wider context: the
o¬ender, it is said, owes a public duty to the State to submit to punishment and a
private duty to the victim to pay compensation.41 This private duty may give rise to
liability in tort, for instance, or to the making of a compensation order. However,
for various reasons, many victims have no realistic prospect of being properly com-
pensated by the o¬ender, and the State owes a residual obligation to victims to ¬ll
this compensation gap. This obligation of the State ˜forms part of the State™s general
duties of enforcing the criminal law and protecting individual rights™.42 In answer,

39 A.M. Linden, Report of the Osgoode Hall Study on Compensation for Victims of Crime (Toronto,
1968), 3. This approach led some to advocate that the State should simply meet judgments
obtained by victims of crimes of violence against o¬enders, but the working party pointed out
some formidable objections to this: Cmnd 1406, paras. 136“40.
40 These ¬gures are derived from Criminal Statistics for England and Wales 2003. Ironically, however,
the argument is more ¬rmly based now than in the past because of increased use of custodial sen-
tences (in 1990, e.g., of nearly 52,600 persons convicted of indictable o¬ences against the person,
only about 5,800 received an immediate custodial sentence, and over 19,000 were ¬ned: Criminal
Statistics: England and Wales 1990 (Cm 1935, 1992), table 7.2) and a sharp decline in the use of
compensation orders (from 27,000 in 1990).
41 For a view that payments to the victims of crime should not be seen as compensation but as
a recognition that wrong has been done and that crime causes its victims mental su¬ering, see
J. Shapland, ˜Victims, the Criminal Justice System and Compensation™ (1984) 24 British J. of
Criminology 131. Under this approach, the compensation function would be left to tort law and
the social security system. Shapland would not limit this approach to crimes of violence (135) or
to intentional crimes (145), and it is not clear why it should be restricted to criminal wrongs, since
victims of non-criminal tortious wrongs (especially those who su¬er serious injuries) no doubt
often also su¬er the sort of mental su¬ering experienced by victims of crime.
42 A. Ashworth, ˜Punishment and Compensation: Victims, O¬enders and the State™ (1986) 6 Oxford
J. Legal Studies 86.
Compensation for criminal injuries 307

it might be argued that the State has analogous duties at least to people injured by
non-criminal but still unlawful conduct (such as tortious negligence), and that
victims of crime deserve no better treatment than these others.
In the parliamentary debates which preceded the establishment of the Scheme
Lord Shawcross did not feel it necessary or useful to attempt to justify the case for
compensation ˜by an elaborate theoretical or philosophical speculation as to why
the State should intervene in a matter of this kind™.43 It was enough to rely on the
˜public instinct™. Lord Dilhorne, speaking on behalf of the government, rejected the
notion that the State was ˜liable™, but regarded the Scheme as an extension of our
˜welfare system™.44 Since nobody suggested that the State was legally liable, this was
tantamount to asserting that the State was not morally liable; but nevertheless Lord
Dilhorne thought that the ˜welfare system™ should be extended to cover it. Others
justi¬ed the Scheme by asserting that society was ˜responsible™ for much crime
because of the laxity of its moral code and the failure to bring up its young as good
citizens rather than criminals.45 Only Lady Wootton protested that the attempt
to assess people™s needs by reference to fault was ˜illogical and uncivilized™.46 The
debate in the Commons followed much the same lines though with even less
attempt to justify the Scheme, and with universal self-congratulation on a good job
well done.
One particular piece of sophistry which this intellectual confusion produced
was the suggestion that the compensation should be awarded ex gratia, not as of
right; and this suggestion was eventually incorporated in the Scheme propounded
by the government.47 Although, in a technical sense, payments under the Scheme
were ex gratia because the Scheme was not authorized by or embodied in a statute,
in reality the CICB had no discretion to refuse claims which fell within the terms
of the Scheme; and the payment of compensation “ though not legally enforce-
able “ followed automatically once the Board had determined that it should be
awarded.48 So it was held, soon after the establishment of the Scheme, that decisions
of the Board could be challenged in judicial review proceedings on the ground that
they did not comply with the terms of the Scheme.49 The new Scheme has a statu-
tory basis in the Criminal Injuries Compensation Act 1995.
In 1978 a Working Party50 appointed to review the operation of the CICS
reiterated the views of the Working Party of 1964 that the public feels a sense of
responsibility for, and sympathy with, the innocent victim, and ˜it is right™ that this
feeling should ¬nd practical expression in the provision of compensation by the
43 HL Debs, 5th series, vol. 245, col. 263.
44 Ibid., vol. 257, col. 1353.
45 Ibid., vol. 245, cols. 267, 269.
46 Ibid., vol. 257, col. 1381.
47 See para. 4 of the 1990 version of the Scheme (which we will call ˜the old Scheme™) (Miers, State
Compensation for Criminal Injuries, 243).
48 Review of the Criminal Injuries Compensation Scheme: Report of an Interdepartmental Working
Party (HMSO, 1978), para. 5.1.
49 R v. CICB, ex parte Lain [1967] 2 QB 864.
50 See n. 48.
308 Chapter 12

community. The Pearson Commission, after referring to some of the above criti-
cisms,51 simply asserted that compensation for criminal injuries was ˜morally
justi¬ed as in some measure salving the nation™s conscience at its inability to pre-
serve law and order™, and insisted that ˜it is right that there should be reasonable pro-
vision for the victims of crime™.52 At that time, awards under the CICS were
calculated in much the same way as tort damages; but the Pearson Commission
made no attempt to explain why ˜reasonable provision™ should mean provision
according to scales of tort damages. It could be argued with some justi¬cation that
levels of tort damages are not ˜reasonable provision™ but exceptionally generous
provision. And it is odd that the Pearson Commission should have used the argu-
ment that compensation for the victims of crime ˜salves the nation™s conscience at
its inability to preserve law and order™, for this argument had always been repudi-
ated by governments and previous working parties. The Commission™s only rec-
ommendation was that the Scheme should be looked at again in light of the
proposals on damages in tort.
The upshot of all this is that the Scheme was in fact based on little more than
feelings of sympathy for innocent victims of crime who are unable to obtain redress
from the o¬ender.53 As one writer has said: 54
Society is seen to recognise and sympathise with the innocent victim™s su¬ering and
this serves to rea¬rm that the victim™s faith, and that of the general public, in society
and its institutions has not been misplaced . . . [the] role [of the CICS] is to symbolize
social solidarity with the victim of violence.

In 1984 the government appointed another working party to make recommenda-
tions for putting the Scheme on a statutory footing.55 Appropriate provisions were
included in the Criminal Justice Act 1988,56 but they were never brought into oper-
ation.57 However, in 1994 the government, worried by the ever-increasing cost to
the public purse of criminal injuries compensation and by the time taken to process

51 As made in the second edition of this book.
52 Pearson Report, vol. 1, paras. 1588, 1591.
53 In 2001 the CICA was reported to have ˜agreed to consider™ claims from relatives who saw the
events of 11 September on television and received farewell phone calls from victims. Criticising
the authority, Libby Purves (The Times, 20 November 2001) described the authority™s remit as
being ˜to compensate victims for the failure of the authorities to protect them from violent crime™.
In reply, the Chief Executive of the CICA (The Times, 21 November 2001) described its job as
being ˜to provide a tangible token of public sympathy to people in Great Britain who have su¬ered
clear injury as a result of a crime of violence™ even if the crime took place on the other side of the
Atlantic. (Under the CICS, the injury must have been su¬ered in Great Britain, but the crime need
not have occurred there.) ˜Successive governments have recognized that the public feel . . . sym-
pathy with the innocent victim of a crime of violence™: Home O¬ce, Compensation for Victims of
Violent Crime: Possible Changes to the Criminal Injuries Compensation Scheme (1999), para. 11.
54 P. Du¬, ˜Criminal Injuries Compensation: the Symbolic Dimension™ [1995] Juridical Review 102.
55 Report of an Interdepartmental Working Party, Criminal Injuries Compensation: a Statutory
Scheme (HMSO, 1986).
56 P. Du¬, ˜Criminal Injuries Compensation: The Scope of the New Scheme™ (1989) 52 Modern
LR 518.
57 HC Debs, 6th series, vol. 163, col. 410.
Compensation for criminal injuries 309

applications for compensation, altered the basis for calculation of CICS awards.58
Instead of being analogous to tort damages, awards were now to be made accord-
ing to a tari¬ of listed injuries. In more serious cases, at least, the level of awards
under the tari¬ was considerably lower than under the tort basis of assessment. This
change was greeted by loud protests and, when challenged in the courts, was held
to be illegal by the House of Lords.59 As a result, applications dealt with under the
illegal scheme had to be reassessed according to the rules of the old CICS. In res-
ponse to the House of Lords™ decision, the government secured the passage of the
Criminal Injuries Compensation Act 1995, and under its provisions the old CICS
was replaced by a new tari¬ scheme. The 1996 CICS (called by the government an
˜enhanced tari¬ scheme™) was more generous than the 1994 tari¬ scheme; but as we
will see, it departs from principles of tort damages in some important respects. The
1996 version of the CICS was replaced by a new version in 2001 (which we will call
the ˜2001 Scheme™), which increased the levels of awards generally and for multiple
injuries in particular, and extended eligibility for awards in fatal cases to same-sex
partners. So far as the rules of entitlement to compensation are concerned, the new
Scheme di¬ers from the old in only minor respects. The discussion which follows
focuses on the new CICS, as there are now relatively few outstanding claims made
under the old Scheme.

12.4.2 The scope of the Scheme
The Scheme provides for the payment of compensation to persons who have
One or more personal injuries61 directly attributable to (a) a crime of violence (includ-
ing arson, ¬re-raising or an act of poisoning); or (b) an o¬ence of trespass on the
railway; or (c) the apprehension or attempted apprehension of an o¬ender or sus-
pected o¬ender, the prevention or attempted prevention of an o¬ence, or the giving of
help to any constable who is engaged in any such activity.

The Scheme also provides for payments to the dependants of those killed in the
same fashion. Whereas a compensation order can be made only against a convicted

58 The changes were set out in a Home O¬ce White Paper, Compensation Victims of Violent Crime:
Changes to the Criminal Injuries Compensation Scheme (Cm 2434, 1993).
59 R v. Secretary of State for the Home Department, ex parte Fire Brigades™ Union [1995] AC 513.
60 2001 Scheme, para. 8. The text of CICS and the Guide to the Criminal Injuries Compensation
Scheme (hereafter referred to as the ˜Guide™) are available on the CICA website.
61 ˜. . . personal injury includes physical injury (including fatal injury), mental injury (that is, tem-
porary mental anxiety, medically veri¬ed, or a disabling mental illness con¬rmed by psychiatric
diagnosis) and disease (that is, a medically recognized illness or condition™: 2001 Scheme, para. 9.
Being conceived and born with a congenital defect as a result of an act of incestuous sexual inter-
course is not a personal injury within the meaning of the CICS: P™s Curator Bonis v. Criminal
Injuries Compensation Board (1996) (available on the website of the Criminal Injuries
Compensation Appeal Panel (CICAP): www.cica.gov.uk). (Being born is not actionable damage at
common law, either.) However, where a woman becomes pregnant as a result of rape and an award
is made to her in respect of non-consensual vaginal intercourse, an additional amount is also
payable to her in respect of each child born alive which she intends to keep: 2001 Scheme, para. 27.
310 Chapter 12

o¬ender, compensation can be given under the CICS even if the o¬ender is never
convicted.62 Indeed, compensation may be paid even in cases where the o¬ender is
acquitted because the burden of proof of guilt applicable under the CICS is the civil
burden ˜on the balance of probabilities™, not the criminal burden ˜beyond reason-
able doubt™ (2001 Scheme, para. 20). Crime of violence
Confused thinking about the justi¬cation for the Scheme has, not surprisingly, gen-
erated di¬culty in de¬ning its scope. Originally, the Scheme did not use the words
˜crime of violence™. It was believed that it would be su¬cient to refer to injuries
˜directly attributable to™ an o¬ence, on the assumption that unless the o¬ence was
one of violence, it would not lead to personal injury. Of course, this is not so, as
the express extension of the Scheme to arson and poisoning shows. For instance,
o¬ences under industrial safety or food and drug safety legislation can easily lead
to personal injuries, but these kinds of o¬ences were not meant to fall within the
Scheme. Therefore, both the Board63 and a Divisional Court of the Queen™s Bench
Division64 have held them to fall outside the Scheme. The term ˜crime of violence™
is nowhere de¬ned. There is, for instance, no exhaustive list of crimes that qualify
as crimes of violence for the purposes of the CICS. On the other hand, the Scheme
itself classi¬es arson, ¬re-raising and poisoning as crimes of violence; and while
most of the items in the tari¬ of awards available under the Scheme are cast in terms
of injuries, a few (relating to physical and sexual abuse) refer to speci¬c criminal
conduct. But in general, whether conduct amounts to a crime of violence is to be
decided on a case-by-case basis by reference to the purposes of the CICS (whatever
they may be).
Conduct will count as a crime of violence only if the relevant injuries were
in¬‚icted intentionally or recklessly (in the technical sense in which these words are
used in the criminal law). Behaviour which is merely careless, even if grossly so,
does not constitute a crime of violence. For example, in one case, a man was injured
by ¬‚ying glass when a snowball was thrown at his kitchen window.65 The applica-
tion for compensation was rejected on the basis that there was no evidence that the
snowball had been thrown with the intention of causing damage. In fact, it
appeared that children were in the habit of playing near the house, and it was likely
that the snowball had been thrown in the course of a game.
Particular di¬culty has arisen in relation to suicide. In R v. Criminal Injuries
Compensation Board, ex parte Clowes,66 a policeman was injured by a gas explosion
in a house where he had been investigating a gas leak. The occupier of the house

62 2001 Scheme, para. 10.
63 CICB First Report (Cmnd 2782, 1965), para. 14(3).
64 R v. CICB, ex parte Webb [1986] QB 184.
65 CICB Sixteenth Report (Cmnd 8081, 1980), para. 17. See also e.g. Twenty-First Report (Cmnd 9684,
1985), para. 20; Twenty-Fifth Report (Cm 900, 1989), paras. 16.3“4.
66 [1977] 1 WLR 1353.
Compensation for criminal injuries 311

had committed suicide by breaking o¬ the end of the gas pipe and allowing gas
to escape. A Divisional Court (by majority) held that the o¬ence of unlawfully
damaging the property of another67 was not a crime of violence because it was
concerned with damage to property, not with violence to the person. But the court
also held that the o¬ence of unlawfully damaging property with the intention
thereby of endangering life, or recklessly as to whether life would thereby be endan-
gered,68 could constitute a crime of violence because although violence to the
person was not part of the de¬nition of the o¬ence, it could have personal injury
as one of its consequences. However, when the Board reconsidered the case, it came
to the conclusion that the deceased had not acted intentionally or recklessly to
cause injury to another, and that the only thing present in his mind was a desire to
take his own life.
A signi¬cant number of applications to the Board are made by train drivers who
su¬er anxiety and depression after their trains strike people who have deliberately
gone onto the tracks, often to commit suicide. This constitutes the o¬ence of
unlawfully endangering the safety of railway passengers.69 In R v. Criminal Injuries
Compensation Board, ex parte Webb,70 the Court of Appeal held that this was not a
crime of violence; that it was for the Board to decide whether particular crimes were
crimes of violence; and that the words ˜crime of violence™ ought to be given the
meaning the reasonable and literate person would give them. The fact that the com-
mission of a crime carries with it the danger of violent injury to another (in such a
case as this, e.g. by derailment of the train) does not make it a crime of violence.
Although most crimes of violence involve the in¬‚iction or threat of force, some
might not. It is the nature of a crime, not its consequences, which determines
whether it is a ˜crime of violence™. Although Clowes was not overruled or even dis-
approved in Webb, it is perhaps unlikely that the o¬ence held to be a crime of vio-
lence in Clowes would be so classi¬ed following to the approach in Webb. It has been
said that Webb reduces the scope of the Scheme ˜to something closer to what was
envisaged by its creators™71 by concentrating on crimes which involve as one of their
ingredients the intentional or reckless in¬‚iction of violence to the person.
As a result of pressure exerted by the train drivers™ unions, the CICS was
amended to include such cases. Why should train drivers bene¬t more than other
innocent objects of public sympathy such as the victims of train crashes or ¬res in
public places? This extension of the Scheme is as unprincipled as the whole idea of
compensating the victims of violent crime.72 Moreover, despite the fact that the
67 Criminal Damage Act 1971, s. 1(1).
68 Ibid., s. 1(2).
69 O¬ences Against the Person Act 1861, s. 34. In 2002“3 trespass and suicide on railway tracks
caused 256 deaths and 137 cases of injury.
70 [1987] QB 74 (CA).
71 P. Du¬, ˜Criminal Injuries Compensation and “Violent” Crime™ [1987] Criminal LR 219, 228.
72 A proposal by the government to remove railway suicides from the Scheme was abandoned in
the face of opposition from the railway industry and unions: Home O¬ce, ˜Compensation
and Support for Victims of Crime™: Summary of Responses to a Home O¬ce Consultation Paper
(2004), 18.
312 Chapter 12

extension was primarily motivated by the problem of railway suicides, it is framed
in terms of the o¬ence of ˜trespass on a railway™. Thus it is apt to cover cases where,
for instance, people deliberately go onto the tracks but with no intention of taking
their own life. Even more anomalously, it could cover a case where a person places
a plank of wood on the track in order to derail a train.73 The formulation was
chosen in preference to the o¬ence of unlawfully endangering the safety of railway
passengers because the latter o¬ence can be committed by merely negligent con-
duct,74 whereas the o¬ence of trespass can, perhaps, only be committed intention-
ally or recklessly.75
An indication of the anomalous lack of principle involved in including such cases
in the CICS is that the extension was prompted primarily by cases in which the train
driver su¬ers only mental injury. Under the 2001 Scheme (para. 9), compensation
for mental injury standing alone is not payable unless the applicant ˜was put in rea-
sonable fear of immediate physical harm to his own person™, or had a ˜close rela-
tionship of love and a¬ection™ with a person who su¬ered physical injury or death,
or was ˜the non-consenting victim of a sexual o¬ence™, or was a railway employee
who ˜either witnessed and was present on the occasion when another person sus-
tained physical (including fatal) injury directly attributable to the o¬ence of tres-
pass on a railway, or was closely involved in its immediate aftermath™. It will be noted
that taken alone, the ¬rst two categories cover a narrower area than the common law
rules of tort liability for mental injury. At common law, people directly involved in
accidents and disasters, either as participants or rescuers, may recover in their own
right for mental injury standing alone; but under the CICS, this is true only in rela-
tion to injury resulting from trespass to the railway. Moreover, whereas a person who
˜accidentally™ su¬ers injury while engaged in an activity ˜directed to containing,
limiting or remedying the consequences of crime™ may recover compensation under
the CICS only if they were ˜taking an exceptional risk which was justi¬ed in all the
circumstances™,76 this limitation does not apply to cases of trespass to the railway.
In one of its reports the Board said that the job of de¬ning the term ˜crime of
violence™ would tax ˜the ingenuity of the participants in a moot™.77 This is partly, no
doubt, because the restriction of the Scheme to crimes of violence is impossible to
explain even in terms of its avowed ˜justi¬cations™: it is not only victims of violent
crime who may attract public sympathy. For instance, victims of large-scale frauds,
such as those that gave rise to the Barlow Clowes and BCCI bank collapses, may also
arouse such feelings. Nevertheless, the European Convention on the Compensation
of Victims of Violent Crimes and the EU Directive on compensation to crime
victims are also limited to crimes of violence.78 There is constant pressure for the

73 Miers, State Compensation for Criminal Injuries, 92.
74 Ibid., 91.
75 But the tort of trespass to land can be committed negligently and, indeed, purely innocently.
76 2001 Scheme, para. 12.
77 CICB Twentieth Report (Cmnd 9399, 1984), para. 17.
78 Council of Europe, European Treaty Series, No. 116; EU Council Directive 2004/80/EC respec-
Compensation for criminal injuries 313

recognition of particular crimes as falling within the Scheme. For example, it has
held that child abduction can be a crime of violence even if the child is not physi-
cally hurt.79 On the other hand, it has been held that bigamy is not a crime of vio-
lence, even if it involves procurement of sexual intercourse by deception and in¬‚icts
mental harm.80 It has also been held that consent of the victim may (but will not
necessarily) prevent conduct being a crime of violence even if the criminal could
not have pleaded consent as a defence to a criminal prosecution in respect of the
conduct.81 Such an ad hoc approach is hardly a satisfactory way of de¬ning the
limits of a compensation scheme.
Injuries caused by children and mentally unstable persons cause considerable
di¬culty. Unsoundness of mind can excuse a person from criminal responsibility;
and in England children under 10 (8 in Scotland) are treated as incapable of com-
mitting crimes. In such cases it cannot easily be said that the victim™s injuries were
attributable to a crime. The Scheme provides that compensation can be paid even
though the o¬ender could not be convicted of the relevant o¬ence because of age
or insanity.82 According to the Guide to the old Scheme, this provision requires the
Authority to decide whether, if the o¬ender had not been immune from conviction
on one of the stated grounds, the conduct in question would ˜as a matter of fact . . .
have amounted to a crime of violence if committed by™ a person not labouring
under the relevant disability.83 The trouble is that this question will usually be
impossible to answer. Conduct will amount to a crime of violence only if the per-
petrator had the requisite criminal intent. Some children below the age of criminal
responsibility may be capable of forming such an intent, but the very reason why
the immunity exists is because many young children are not so capable; and even
if a particular child could form the requisite intent, it may be very di¬cult to prove
that they had done so.84 In the case of insanity, if a person is legally insane, they are,
by de¬nition, incapable of forming the requisite intent. Nevertheless, the Court of
Appeal has said that the CICS ˜would be a most defective scheme if anyone injured
by a mentally unbalanced person could not be paid compensation™.85 It is the lack
of a reasoned justi¬cation for the Scheme that has generated this problem. Why
should compensation be paid to a person intentionally injured, but not to a person
accidentally (that is, neither intentionally nor recklessly) injured, by a young child
when the child is not criminally responsible in either case? Either both should be
compensated, or neither.

79 Independent, 8 March 1993.
80 Gray v. Criminal Injuries Compensation Board (1998) (available on the CICAP website).
81 R v. Criminal Injuries Compensation Appeals Panel, ex parte Wade (2000) (available on the CICAP
82 ˜or diplomatic immunity™: 2001 Scheme, para. 10. ˜Insanity™ is a legal term with a technical
83 Guide to the old Scheme, para. 7.
84 The Guide to the old Scheme dealt separately with cases of children playing dangerous games, as
does the Guide to the 2001 Scheme, 26.
85 R v. Criminal Injuries Compensation Board, ex parte Webb [1987] QB 74, 80.
314 Chapter 12 Accidental injuries
Despite all that has just been said, it is possible in certain cases to recover com-
pensation under the CICS for accidental injuries (that is, injuries not caused


. 11
( 18)