of employees being made to work under stressful conditions.
Journal of Business Law
Psychiatric stress – an unacceptable cost to employers
Other related subjects:
Health and safety at work. Negligence
Employers’ liability; Guidelines; Health and Safety Executive; Negligence; Occupational
stress; Psychiatric harm
Management of Health and Safety at Work Regulations 1999 (SI 1999 3242) reg.3
Health and Safety at Work etc. Act 1974 s.2 (1)
Walker v Northumberland CC  1 All E.R. 737 (QBD)
Hartman v South Essex Mental Health and Community Care NHS Trust  EWCA Civ 6; 
I.C.R. 782 (CA (Civ Div))
Barber v Somerset CC  UKHL 13;  1 W.L.R. 1089 (HL)
Barber v Somerset CC  EWCA Civ 76;  2 All E.R. 1 (CA (Civ Div))
A phenomenon of the last decade has been the national concern with the problem of stress at the
workplace. The existence of stress in the workforce was not recognised until the final quarter of the
20th century; yet today it is officially recorded as a major cause of work-related ill health. Much
sickness absence is attributed to it. For this reason alone it is costly to employers. In addition if an
employee does suffer a stress-related psychiatric injury the employer may face a claim for damages.
Finally, the Health and Safety Executive (HSE) has made it clear that an employer who puts an
employee at risk of stress-related injury may face prosecution.
This article will trace how this situation arose. It will review the propositions of the Court of Appeal as
to when an employer can be liable for negligently causing an employee to suffer psychiatric stress
injury and the guidance of the HSE as to identifying stress when carrying out a risk assessment.
Finally, it will speculate as to whether observing these two sources of guidance will enable employers
to reduce the cost to them or to society as a whole of psychiatric stress-related illness.
The background to the problem
In the 1990s, employers’ liability for negligence began to develop in a different direction from the
general law of negligence. Over the course of 100 years the law developed so as to enable those
involved in sudden catastrophic occurrences to recover damages for psychiatric injury. The “nervous
shock” suffered often manifested itself in the form of what we should today recognise as
post-traumatic stress disorder (PTSD). However, in the early cases, from
Dulieu v White & Sons
until at least
Bourhill v Young,
the limited state of medical knowledge caused the victims
to rely on physical symptoms, such as a miscarriage.
Up until the final years of the 20th century, there was no remedy in the tort of negligence (or
otherwise) for psychiatric injury caused by stress in circumstances where the victim had not been
exposed to a shocking occurrence. The Law Commission’s report,
Liability for Psychiatric Illness,
published in December 1997 found no tradition of claimants recovering damages for psychiatric injury
caused by factors other than the shock of involvement in disaster. Thus it cited two cases in which
parents were unable to recover for the trauma they suffered as a result of watching a child physically
deteriorating over a period of time and eventually dying because of the defendant’s negligence.
The Law Commission’s report was published just at the point when employers’ liability drew away
from the general law on compensation for psychiatric injury. The report marked the “control
mechanisms” then recently put on the liability of a negligent defendant for nervous shock; namely
requirements that the claimant was either a primary victim at risk of physical injury
or a secondary
victim who witnessed a disaster involving a loved one.
The Commission also noted the first reported
of an employer incurring liability for stress, suffered over a period of time, as a result of work
“… [T]he social worker suffering from stress in
Walker v Northumberland County Council
required to show that his breakdown was caused by shock.”
Having considered the arguments for and against confining liability for psychiatric illness to situations
where that illness had been caused by shock, the Commission concluded that the shock requirement
should be abandoned. Admittedly, it was focusing on claims arising out of a single catastrophic event,
but it predicted that abandoning the shock test would not open the floodgates to litigation and it did
not believe that the courts would be unable to cope with the difficult issues of causation that might
arise. The proposal to abandon the shock test in claims following a single catastrophic event has not
been put into effect. The Commission failed to realise the extent to which
had, in the context
of employers’ liability, opened the floodgates to compensation for psychiatric injury without proof of
Developing awareness of work-related stress
At the time when the Health and Safety at Work Act 1974 was debated as a Bill before Parliament, a
passing reference was made to the incidence of work-related mental illness. It was claimed that 38
million working days were lost each year as a result of work-related mental illness.
No reference was
made to stress and there could clearly have been other causes, such as, for example, employment on
tasks where mercury was used.
The concept of stress in the context of personal injury did not
emerge until the 1960s.
It was not until the late 1980s that the significance of work-related stress
injury became apparent, both in relation to physical and psychiatric injury. For example, while it was
recognised as early as 1911 that the use of pneumatic tools could cause secondary Raynaud’s
phenomenon (vibratory white finger), it was only in the 1980s that this was added to the list of
prescribed industrial diseases.
At about this time musculoskeletal problems associated with
computerised keyboard operation became the subject of litigation.
Mughal v Reuters
a High Court judge suggested that the physical symptoms from which a keyboard operator
allegedly suffered could be caused by a mental condition arising from his dislike of the work.
By the early 1990s, the HSE was commissioning wide-ranging research into the causes and
prevalence of work-related stress. Much of this work was led by Professor Tom Cox.
In 1995, the
HSE published guidance to employers
which it indicated the causes of stress at work,
its effect, employers’ responsibilities and the steps that employers might take to prevent stress-related
The HSE guidance provided what remains, almost word for word, the generally accepted definition of
“Stress is the reaction people have to excessive pressures or other types of demand placed upon
them. It arises when they worry that they can’t cope.”
It is clear from the second sentence that the guidance is concerned with psychiatric injury, but it
observes that such stress can result in a number of serious ill-health conditions, including thyroid
disorders, high blood pressure, heart disease, ulcers, depression and anxiety. The HSE therefore
accepted the close relationship between psychiatric and physical injury.
At the beginning of this century, the HSE believed that psychiatric injury was vying with
musculoskeletal injury as the major cause of sickness-related absence from the workplace. Statistics
showed musculoskeletal illness in Britain affected one million people a year and most of the figures
showed that it cost society £5.7 billion per annum.
The HSE further notes that some research
suggests a link between musculoskeletal disorders and psychosocial factors, including factors relating
to work and the work environment.
Of mental stress, the HSE stated:
• about one in five people say that they find their work either very or extremely stressful;
• over half a million people report experiencing work-related stress at a level they believe has actually
made them ill;
• each case of stress-related ill health leads to an average of 29 working days lost. A total of 13.4
million working days were lost to stress, depression and anxiety in 2001; and
• work-related stress costs society between £3.7 billion and £3.8 billion a year (1995/1996 prices).
The Labour Force Survey
Self-reported work-related illness in 2003/04
provides broadly similar
information and emphasises that the annual statistics remain fairly constant.
In fact it is, for a variety of reasons, extremely difficult to measure the amount of illness due to
work-related psychiatric stress. Several of these difficulties are
inferred in the statistics
given above. One difficulty is that stress illness can manifest itself in a variety of forms. For example,
in a recent case a publican claimed to have suffered a heart attack because of the pressures of trying
to make a success of a public house situated in a rough area. The claim failed because he was
unable to satisfy the Court that his employer was on notice that he was at risk but the Court did not
question whether psychiatric stress could cause a heart attack.
Indeed, researchers claim stress
and depression are just as likely to cause a heart attack as being fat or having high cholesterol.
second difficulty in trying to quantify the extent of the problem is that stress is subjective in its nature:
what one person perceives as challenging, another may find stressful. Thus in
Vahidi v Fairstead
House School Trust Ltd,
the Court of Appeal concluded that the stress illness suffered by a teacher
was caused solely by her attempts fundamentally to change her teaching methods. A further difficulty
is that close examination of the HSE’s statistics shows that they rely heavily on the recorded opinions
of those who consider themselves the victims of stress rather than on scientific evidence. Finally,
stress that manifests itself at the workplace may be attributable to the fragility of the worker as a result
of problems in their personal life, rather than an inherent problem with the work. A number of claims
have failed for this reason.
Nevertheless, it was opined in the House of Lords by Lord Steyn in
Johnson v Unisys Ltd
recent years there had been “an extraordinary intensification of work pressures”, so that, “[i]nevitably,
the incidence of psychiatric injury due to excessive stress has increased”.
One thing is clear, whether the ill health is real or imaginary, the fault of the victim’s employer or due
to an inability to cope with the harsh realities of 21st-century life, stress is extremely costly to
employers. When an employee is absent from work this clearly affects the productivity of the
workplace, but over and above this employees who are at work, but on the verge of succumbing to
stress are likely to be giving less than their maximum performance and may well impact on the
wellbeing and performance of their colleagues. However, this primary cost is perhaps relatively
insignificant compared with the cost of civil claims by those allegedly disabled by work-related stress.
The overt costs of legal fees incurred and damages awarded may well be but a small part of the true
cost. They may be overshadowed by the effect of litigation on the efficient operation of the
organisation. The true cost, which must include the impact of litigation on the use of human
resources, is likely to spread far beyond the boundaries of the HR department. To arrive at some idea
of the significance of civil liability it is necessary to trace the development of case law.
The development of civil liability
Until relatively recently claims against employers by employees for personal injury have been based
either on the tort of negligence, or on breach of statutory duty, and it is through the tort of negligence
that liability for work-related psychiatric injury has developed. There have recently been attempts to
litigate for occupational stress through other branches of the civil law, but only liability for negligence
will be considered in depth here.
Rather surprisingly, although there is now a long history of successful claims for nervous shock, there
are relatively few reported examples of employers incurring liability for nervous shock suffered by their
employees. Interestingly, by the 1990s these cases were all ones in which the claimant was a
secondary victim shocked by witnessing injury to others. Of these the only significant successful claim
Dooley v Cammell Laird & Co Ltd,
heard at Liverpool Assizes. The claimant crane driver was
lent by the first defendant, his employer, to the second defendant to enable the delivery of equipment
to the hold of a ship where the second defendant’s men were working. Due to a defective sling the
load crashed into the hold. The claimant suffered nervous shock because he feared that workmen
may have been injured. He claimed against both defendants alleging negligence and against his
employer for breach of statutory duty. Donovan J. found the employers were not negligent but were
liable to the claimant for breach of duties imposed on them, as occupiers of the shipbuilding yard, by
the Shipbuilding Regulations 1931.
His Lordship noted that the Regulations were intended to protect
employed persons against the risk of bodily injury and commented:
“No suggestion is made that the words ‘bodily injury’ … exclude injury to nerves, which after all are a
part of the body.”
Alcock v Chief Constable of South Yorkshire,
Lord Oliver referred to
as one of a group of
“… [w]here the negligent act of the defendant has put the plaintiff in the position of being, or of
thinking that he is about to be or has been, the
involuntary cause of another’s death or
injury and the illness complained of stems from the shock to the plaintiff of the consciousness of this
In December 1998 the House of Lords unanimously agreed that employers’ liability was only an
aspect of the general law of tort and employees who were secondary victims enjoyed no special
preference in nervous shock claims against their employer.
remains good law, it
can only be because of the exceptional circumstances identified in
by Lord Oliver.
This brief account of employers’ liability for nervous shock suffered by their employees demonstrates
how little authority there was, when the case of
Walker v Northumberland CC
was heard in the
Newcastle District Registry by Colman J. in November 1994, that the employer’s implied duty of
reasonable care extended beyond care for the physical safety of the employee.
The facts of the case were that Mr Walker suffered a breakdown due to the heavy workload he had to
carry as an area social services officer employed by Northumberland County Council. Following his
return to work the employer did nothing to lessen his burden and he fell ill a second time, this time so
seriously that he was dismissed. In defending the negligence claim the employer conceded it had a
duty to exercise care to protect Mr Walker from risks that were reasonably foreseeable and only
contested whether Mr Walker’s illness was foreseeable. Colman J. therefore touched but briefly on
the duty issue stating:
“There has been little judicial authority on the extent to which an employer owes to his employees a
duty not to cause them psychiatric damage by the volume or character of the work which the
employees are required to perform. It is clear that an employer has a duty to provide his employee
with a reasonably safe system of work… Whereas the law on the extent of this duty has developed
almost exclusively in cases involving physical injury to the employee as distinct from injury to his
mental health, there is no logical reason why risk of psychiatric damage should be excluded from the
scope of an employer’s duty of care…”
The remainder of the judgment is largely directed towards whether the illness was foreseeable. He
cited from the judgment of Dillon L.J. in
Petch v Customs and Excise Commissioners,
unsuccessful claim, as support for the view that an
employer would be liable if an
employee suffered a breakdown in circumstances where the workload was so heavy as to make a
breakdown foreseeable. In the view of Colman J. Mr Walker’s second illness was foreseeable and he
therefore gave judgment for the claimant. He rejected the employer’s argument that the financial
position of the Council prevented it from alleviating Mr Walker’s workload after he had recovered from
his first illness.
No reference is made in the judgment to absence of authority in the general law of negligence for
liability for psychiatric illness caused other than by shock. However, his Lordship did note that Mr
Walker’s first illness was “attributable to the impact on his personality of his work”
predicted that claims for illness due to work overload “… will often give rise to extremely difficult
evidential problems of foreseeability and causation”.
Inspired by this decision, supported by trade unions, and encouraged by the media, there were many
stress-related claims. In 1998 a survey established that 459 cases were in progress.
The press both
highlighted the incidence of litigation and noted individual cases.
Many claims led to out-of-court
Some cases were successfully fought in county courts, but not formally reported.
It was not until 2002 that the Court of Appeal, being required in
Sutherland v Hatton
to review four
county court decisions granting compensation to employees, took the opportunity to thoroughly
review employers’ liability in negligence for causing psychiatric injury to employees. Hale L.J.,
delivering the judgment, set out 16 propositions for the guidance of judges trying such cases.
First, she accepted that the existence of the duty of care in respect of psychiatric injury could be taken
for granted where the employee was a primary victim
: such claims might be brought in either tort or
contract. She placed the
case in the latter category.
Moreover, in claims arising from the
stress of doing the work the employee is required to do the claimant is a primary victim, and the
special control mechanisms that apply where the claimant is a secondary victim do not apply.
Most of the remaining propositions concern the factors that will determine whether the conduct of the
employer had been negligent. Proposition number
two affirms the burden on an employee
alleging personal injury caused by an employer’s negligent conduct:
“The threshold question is whether this kind of harm to this particular employee was reasonably
Her Ladyship’s third proposition is that foreseeability depends upon what the employer knows (or
ought reasonably to know) about the individual employee. She endorsed the established principle that
in employers’ liability the content of the employer’s duty may be higher in situations where the
employer knows of the vulnerability of the employee.
Noting that it is harder to foresee mental
disorder than physical injury, she said that an employer is usually entitled to assume that the
employee can withstand the normal pressures of the job unless it knows of some particular problem
or vulnerability. In her fourth proposition she asserted that there are no occupations which should be
regarded as intrinsically dangerous to mental health.
The fifth proposition addresses factors likely to be relevant in answering the threshold question of
foreseeability. They include:
• The nature and extent of the work done by the employee. Is the workload much more than is normal
for the particular job? Is the work particularly intellectually or emotionally demanding for this
employee? Are the demands which are being made of this employee unreasonable when compared
with the demands made of others in the same or comparable jobs? Or are there signs that others
doing this job are suffering harmful levels of stress? Is there an abnormal level of sickness or
absenteeism in the same job or the same department?
• Signs from the employee of impending harm to health. Do they have a particular problem or
vulnerability? Have they already suffered from illness attributable to stress at work? Have there
recently been frequent or prolonged absences which are uncharacteristic of them? Is there reason to
think that these are attributable to stress at work, for example because of complaints or warnings from
them or others?
These two paragraphs address both the particular and the general. Paragraph (a) initially focuses on
the particular employee in the particular job, but goes on to consider indicia that the claimant may not
be especially vulnerable but manifesting a problem common to the work group. Presumably if the
whole group is at risk the evidence of the problem should be more apparent and the claimant’s illness
the more readily foreseeable. Paragraph (b) outlines circumstances that put the employer on notice
that the particular employee is at risk.
Propositions 6-13 qualify the employer’s duty. Proposition 6 provides that the employer is generally
entitled to take at face value what it is told by an employee. It does not generally have to make
searching enquiries of the employee or their
medical advisers. Moreover (proposition 7), in
order to trigger a duty to take steps, the indications of impending harm to health must be plain enough
for any reasonable employer to realise that it should do something about it. Nevertheless (proposition
8), the employer is only negligent if it fails to take steps which are reasonable in the circumstances,
bearing inmind the magnitude of the risk of harm occurring, the gravity of the harm which may occur,
the costs and practicability of preventing it, and the justifications for running the risk. The size and
scope of the employer’s operation, its resources and the demands it faces are relevant in deciding
what is reasonable; these include the interests of other employees and the need to treat them fairly,
for example in any distribution of duties (proposition 9). The employer can only reasonably be
expected to take steps which are likely to do some good (proposition 10). An employer with a
confidential advice service is unlikely to be in breach of duty (proposition 11). If the only reasonable
and effective step would have been to dismiss or demote the employee, the employer will not be in
breach of duty in allowing a willing employee to continue in the job (proposition 12). In all cases, it is
necessary to identify the steps which the employer both could and should have taken before finding it
in breach of its duty of care (proposition 13).
Her Ladyship then returned to a matter touched on in proposition 2, namely that the claimant has to
show that the injury is attributable to stress at work. In proposition 14 she held that the employee
must show that the employer’s breach of duty caused or materially contributed to the harm suffered.
Where the harm suffered has more than one cause, the employer should only pay for that proportion
of the harm which is attributable to its wrongdoing, but it was for the defendant to raise the question of
apportionment (proposition 15). The assessment of damages should take account of any pre-existing
disorder or vulnerability and of the chance that the claimant would have succumbed to a
stress-related disorder in any event (proposition 16).
These propositions were implicitly accepted by the House of Lords in
Barber v Somerset CC.
then several cases have been appealed to the Court of Appeal to be received with some surprise,
even irritation. Thus in
Hartman v South Essex Mental Health and Community Care NHS Trust,
which six cases were appealed, Scott Baker L.J. in delivering the judgment noted that despite the
decisions of the Court of Appeal in
and the House of Lords in
judges were finding
difficulty in applying the appropriate principles in claims
arising from stress at work. Some
cases were fought over many days at a cost of time and money disproportionate to the value of the
claim. He also noted that in
Lord Walker warned that while Hale L.J.’s judgment gave useful
“… [I]t must be read as that, and not as having anything like statutory force. Every case will depend
on its own facts…”
Vahidi v Fairstead House School Trust Ltd,
the Court of Appeal regretted that
mediation had not taken place since the courts had settled many of the principles in stress at work
It appears from the litigation that the difficulty is not with the principles governing the determination of
whether conduct is negligent but, as Scott Baker L.J. commented in
in applying them to the
particular facts of a case. This is hardly surprising insofar as so many of the propositions concern
foreseeability and how the employer should respond to what is foreseeable: it has long been
recognised that individual judges have different perceptions about what is foreseeable, so that
negligence litigation is a forensic lottery.
In practice, this leads to the Court of Appeal evaluating the
trial judge’s treatment of the evidence. However, the decision of the House of Lords in
demonstrates the dangers of the Court of Appeal substituting its own judgment for that of the trial
judge. The majority of their Lordships supported the view, expressed by Lord Walker at , that the
trial judge had had the advantage of seeing and hearing the witnesses, and in this borderline case the
judge was entitled to come to the conclusions that he did. There was insufficient reason for the Court
of Appeal to set aside his findings. In his dissenting opinion Lord Scott disagreed thus:
“… [T]he question, in my respectful opinion, is not whether the trial judge ‘was entitled’ on the material
before him to come to the conclusions he came to. The question is whether his conclusions were
correct. The Court of Appeal thought they were not. A statement that the trail judge was entitled to
come to the conclusions leaves that question unanswered. The conclusions are judgmental. They are
not findings of primary fact; they are not conclusions as to how some judicial discretion should be
exercised. They constitute the judgment of the trial judge based upon his findings of primary fact. Why
cannot the Court of Appeal substitute their own judgment if it thinks the trial judge’s judgment was
wrong? (paragraph 11).”
It is little comfort to an employer to know that Lord Walker considered that the well-known statement
of Swanwick J. in
Stokes v Guest Keen and Nettleford (Bolts & Nuts) Ltd
remains the best
statement of general principle, “… the overall test is still the conduct of the reasonable and prudent
The defendant employer can only speculate whether a court will consider it has acted in a
reasonable and prudent manner. The number of appeal cases in which the decision of the trial judge
has been overruled, in effect on the appeal court’s view of the evidence, would justify an employer
concluding it was playing a game of roulette. In
three of the four decisions were overruled
to the advantage of the employer and in
the appeal court ruled in favour of the employer in
four of the six appeals.
What is interesting about
is that two of the appeals throw doubt upon Hale L.J.’s eleventh
proposition concerning the value to an employer of having a confidential advice service. In
itself the Court of Appeal held that confidential information held by the employer’s occupational health
department could not be deemed to be within the employer’s knowledge for the purposes of
determining what was foreseeable. In this particular case the confidentiality of the service was in the
employer’s favour. In
the Home Office unsuccessfully relied on the fact that it had
a care team to support prison officers; but the Court may have believed the claimant’s contention that
this system did not operate well. More recently in delivering his judgment in
Daw v Intel Corp (UK) Ltd