Violence in the workplace has received growing attention, in part due to the increase in litigation following incidents in the workplace. Although classic cases of a disaffected worker shooting his fellow employees or his employer are rare, data show a rise in many other kinds of violent incidents in the workplace. The definition of workplace violence most commonly focuses on physical attacks but has been expanded of late to include non-physical attacks, such as harassment and threats. Government agencies have responded with regulations and guidelines for minimising the occurrence of such incidents, and many private corporations and businesses have likewise attempted to implement policies and training programmes that address violence in the workplace issues. However, most companies lack expertise in confronting problematic employees, and many have begun to look to outside mental health providers, as distinct from their own human resource departments, to deal with the problem. This paper looks at the current research on workplace violence in the USA and offers suggestions concerning the roles that mental health professionals with forensic expertise can play in this expanding field.

DEFINITION OF WORKPLACE VIOLENCE

Although there is no standard definition of workplace violence, one common, if loose, description is “any work or work-environment problem that negatively affects employee production and safety”. This includes on-site problems as well as off-site conditions and occurrences (i.e. stalking, telephone harassment and other confrontations). The federal Occupational Safety and Health Administration (OSHA, 1996) defines occupational injury as “an injury which results in death, significant medical treatment, or loss of consciousness which results in loss of work, restriction in ability to work, termination, or transfer”. Some state definitions include “emotional harm” with signs or symptoms that may be “immediate or delayed” (State of California, 1993, p. 39). Although traditional concerns have focused on physical assault, workplace violence is today seen to also include verbal threats and intimidation, as well as sexual harassment.

Government entities as well as privately held companies are increasingly being criticised for their lack of appropriate policies and procedures for addressing violence in the workplace. Some companies have looked internally to their human resources departments or to their employee assistance programmes (EAPs) for the provision of psychological consultation and treatment of those who are believed to be potentially violent. The role of the EAPs and human resources departments in assessing and reporting ‘fitness for duty’ has been growing (Reference McLean, Ellis and MossMcLean et al, 1984). However, their capacity to assess objectively any potentially disruptive employees and to deal with them is often constrained by concerns of double agency (i.e. the company may have a vested interest in moving even limited-risk employees out of the work force, and employees are correspondingly reluctant to respond to questions or to participate in review proceedings). As a result, companies have become more active in seeking outside speciality consultation for assessment of potential violence.

Although the so-called general-duty clause of the federal Occupational Safety and Health Act of 1970 required policies, procedures and training programmes specific to violence in the workplace, as of 1994 less than 20% of businesses reported compliance (Bureau of Labor Statistics, 1994). The 1970 act mandates that all employers provide a place of work that is free of identified hazards that can cause serious physical harm or death to an employee, including circumstances where fellow employees constitute the ‘hazard’. Whether or not the employer has actual knowledge of such hazards, the company can be held liable for damages resulting from the actions of an employee.

STATISTICS ON WORKPLACE VIOLENCE

In the 1980s homicide was the third leading cause of work-related death in the USA (Reference Castillo and JenkinsCastillo & Jenkins, 1994), with about 13% of workplace deaths being reported as homicides (Reference Bell, Stout and BenderBell et al, 1990). During the early 1990s an average of 20 workers were murdered while at work each week in the USA and an estimated one million workers were physically assaulted annually. Data from the Bureau of Labor Statistics (1994) indicate that 1063 work-related homicides occurred in 1993. Overall, workplace homicide has now become the second leading cause of occupational fatalities for men and is the number one cause of death for women (Bureau of Labor Statistics, 1994). Estimates for 1995 are that over two million workers were physically attacked, six million workers were threatened and 16 million workers were harassed (Bureau of Labor Statistics, 1995). With regard to non-fatal assaults, women experienced 55% of the work-related assaults that involved loss of work time, although women comprise 45% of the national workforce. A large-scale survey of working women suggests that 50% of women in the workforce will be sexually harassed at least once during their career, potentially resulting in significant psychological and other health-related consequences (Reference FitzgeraldFitzgerald, 1993). Recent surveys have found that a majority of victims of harassment do not report the incident (58%) and about half of the threats to employees are inadequately reported (43%) (Northwestern National Life, 1993).

The consequences of violence in the workplace can be devastating beyond the immediate injuries. Not only may aggressive acts cause damage to company property and lead to increased insurance costs, legal costs, tarnished company image and diminished employee morale, but it may also lead to loss of confidence in the company, which in turn results in lost business. The exposure of employees to workplace violence can result in loss of worker productivity and increased employee turnover, with associated costs and diversion of management resources. Estimates suggest that workplace violence results in annual costs exceeding 55 million dollars in lost wages alone (Reference BachmanBachman, 1994). Escalating worker compensation premiums further drain company revenues.

LEGAL IMPLICATIONS

The law has kept pace with the reported empirical developments regarding incidence and types of workplace violence. The contribution that law can make to curbing workplace violence is potentially significant. The targets of law that seek to reduce workplace violence include employers and employees, inside and outside perpetrators and even victims or potential victims.

Civil law is the primary regulator of conduct in this area and employers constitute the primary target. The theory is that civil liability or the threat of it will move employers to improve the safety of the workplace. This expectation appears to be well shared by trial judges and juries in the USA, who have shown an inclination to hand down incentive-creating, if not punitive, verdicts that would catch the attention of any risk-conscious employer (occasionally on facts or theories that do not readily justify the outcomes).

Liability through tort law (the law of negligence) is the main civil law avenue for motivating employers. There are two distinct strands. The first seeks to ground liability on, and thus move employers to correct, deficiencies in the security of the physical premises. The threat at which this is aimed is primarily violence by outside, unknown perpetrators. The case outcomes here show a marked propensity to find liability on the part of the employer. At times they are based on theory that is distinctly disincentive-creating, such as where employers are held to account for having voluntarily undertaken steps to improve security but these steps (in hindsight) are deemed insufficient. Liability would not have been found had the employer done nothing (Martin v. McDonald's Corporation, 1991; Vaughn v. Granite City Steel, 1991).

The other strand is aimed at known perpetrators (i.e. employees) over whom the employer is assumed to have control but which control he or she allegedly failed to exercise properly. There are several overlapping components to this liability theory, some or all of which may be tried in one case to see which one will ‘stick’ (i.e. yield the desired, maximum result for the plaintiff): negligent hiring, negligent supervision, negligent retention and even negligent referral. The last theory, in particular, puts employers in a legal ‘damned-if-they-do/damned-if-they-don’t' situation, because recommendations that say too much about the referred employee's violence potential and those that say too little can both lead to liability when things go wrong (Doe v. Methacton School District, 1995).

There is also agency law at the victim/plaintiff's disposal where recovery is sought against the employer for violence done. Its advantage from the plaintiff's perspective is that it obviates the need to prove negligence on the employer's part, because the employer's responsibility for the employee's misconduct is automatic and vicarious — a function of the employment relationship as such. On the other hand, there is a limitation on this theory (not always surmountable even in American law) in that the misconduct must have occurred “in the course of employment” (Lisa M. v. Henry Mayo Newhall Hospital, 1995).

Contract law provides another avenue, although it tends to be further removed than negligence or agency. Employment contracts do not typically ensure the employee's safety from violence, and much less do they give such assurances to third-party victims. However, American legal ingenuity is not readily stymied by mere factual realities. A safety-assuring contract can be found to be implied (i.e. in company handbooks or manuals that provide safety-related regulations or otherwise articulated company policies or guidelines that can be seen to have some bearing on health and safety). Again, the disincentive-creating aspect of liability premised on such implied “ contracts” should be readily apparent (Foley v. Interactive Data Corp., 1988).

Civil Rights laws of various origins and aims exist to compensate for many emotional injuries that may be suffered on the job due to discrimination or harassment. Employer liability, usually based on the direct fact that the employer ‘did’ the discriminating or harassing or, alternatively, that he or she knew or should have known that others were ‘doing it’, has been found in some cases not to require such cumbersome proof. Knowledge in these cases is simply imputed to the employer by virtue of the status of employer, and responsibility is simply assumed under this plaintiff-friendly legal conceptualisation known as “strict liability” (Miller v. Bank of America, 1979). Victims may bring civil rights actions for misdeeds allegedly based on race, religion, gender, age, nationality or disability — the latter opening up the doors to some interesting types of claims by ‘victims’ with, for example, major weight or personality problems, self-induced substance addictions and even mental illnesses that formerly might have made the very fact of employment unthinkable.

There is also the possibility of augmenting employer accountability through regulatory law, such as that administered by the federal OSHA (29 U.S.C. § 654 (a) (1)). This type of law makes the employer directly responsible for the safety of the workplace and although it is primarily aimed at physical conditions or hazards, can be interpreted to include incidents of violence or harassment. The federal law appears to be limited to conditions that “are causing or are likely to cause death or serious physical harm”. However, agency guidelines, which are as binding on employers as the law itself, may be promulgated to incorporate lesser harms, including conceivably pure emotional ones. Also, some states have OSHA-style regulations in place whose application may be considerably broader than the federal provisions.

Affirmative obligations on individuals who have control over others to protect potential third-party victims, such as those imposed via the famous Tarasoff case on the therapists whose patients threaten harm, may arguably be extended to employers who (as emphasised by the negligence and agency laws discussed earlier) are presumed to have substantial, if not equal, control over their employees (Tarasoff v. Regents of the University of California, 1976). There is at least one known case (from Texas) where this theory was attempted. Although the plaintiff failed, the lack of success stemmed from facts deemed inhospitable to the theory rather than perceived inhospitability of the theory itself (Peek v. Equipment Services, Inc., 1995). Given a ‘laboratory’ of 50 states in the USA where these sorts of theories can be tried out, it can be anticipated that in the near future there will be a plaintiff who will succeed on a claim that the employer failed to exercise an affirmative duty to warn or protect him or her from the predictable violence of an employee.

Finally, the criminal law can be brought into play in situations where the employer's conduct (or omitted action) is so egregious as to pass from civil negligence or recklessness to where it is justifiable to impute criminal intent, punishable by the criminal laws of the state where the violation took place. Agents of the OSHA entities already have the power, occasionally exercised, to recommend such criminal prosecutions in appropriate cases. There is, however, no reason why use of the criminal law should be confined to the initiatives of such regulatory middle-men.

Criminal laws against stalking and related forms of harassment exist today in each of America's 50 states. Their aim is not to inculpate the arguably accountable employer, but to deter and punish directly the (potential) victimiser or perpetrator. The effectiveness of these laws has been questioned by many a commentator, but it is clear that some of the doubt is a function of the fact that the laws' successes are not known: where there is no incident, there is no “story” (not even an abstract entry in the ledger of a police desk-clerk or a faceless statistic in the annals kept by an academically concerned criminologist or sociologist).

As a last legal note, it deserves to be recognised that negligence law can be applied against the victim of violence as well as against those who may be viewed as accountable for the conduct of the perpetrator. This is under the doctrine of contributory negligence, which may reduce or defeat the claim of the injured plaintiff because he or she was partially or mostly at fault in provoking the incident. Doctrines that withhold or diminish compensation for the injured tend not to be popular in practice, but the underlying theory is defensible: it aims to deter conduct that generates injuries and injury claims. There is no principled reason to exempt would-be victims from this well-established legal calculus, even if it may seem hardhearted to apply it to actual victims.

THE ROLE OF MENTAL HEALTH PROFESSIONALS

The mental health professional should have a clear understanding of his or her role and the reason for the consultation — whether it is a pre-employment screening, a fitness-for-duty evaluation of an already hired individual or an emergency threat assessment. The clinician's method should always include an interview with the employee, an interview with supervisors and a review of pertinent files (i.e. medical/psychiatric reports, police reports and personnel files). A thorough inquiry into the individual's psychiatric history, as well as possible substance abuse, arrest and prior employment record, should be part of any workplace-related evaluation. Personality characteristics such as coping strategies, frustration tolerance, quantity and quality of personal stressors and past episodes of violence also should be addressed. Often, psychological testing can assist with the evaluation of relevant personality dynamics. In some cases, screening for organic brain disease should be considered.

PRE-EMPLOYMENT SCREENING EVALUATIONS

The employer is expected — by custom if not by law — to perform an adequate screening of a potential employee via a limited background investigation, which should include telephone interviews with previous employers and a check of personal references. Managers and supervisors typically conduct the application interview and address the applicant's general competency, work skills, interpersonal skills and fundamental aptitude with regard to the appointed position. Background checks are constrained in scope today by the Equal Employment Opportunity Commission's requirement that inquiries must be associated with an authentic qualification for employment (Reference Capozzoli and McVeyCapozzoli & McVey, 1995). Criminal records can be obtained for minimal processing fees through the local court-houses. When interviewing candidates with a history of mental illness, the examiner should be aware that the federal Americans With Disabilities Act of 1994, as well as the statutes of most states (Reference GoldbergGoldberg, 1994), prohibit discrimination against a potential employee on the basis of broadly defined mental and physical disabilities. Prior to the evaluation, the candidate must be made aware that the screening is a precondition of employment and that it will most likely influence the hiring decision.

The pre-employment evaluation includes a thorough clinical interview with pertinent historical data, which should be provided by the employer prior to the evaluation, if available. During the clinical interview previous relationships with employers and family should be explored, as well as any criminal history. Simply asking the prospective employee how he or she feels about certain past employers may yield a wealth of information. The informed clinician takes into consideration any sudden changes in the evaluee's mood or personality, and questions any inconsistent information given. Following the interview, psychological testing may be requested when issues surface regarding personality difficulties, interpersonal functioning, impulsivity and drug/alcohol misuse.

FITNESS-FOR-DUTY EVALUATIONS

A fitness-for-duty (FFD) evaluation differs from pre-employment screening in that the person being evaluated is already an employee but has done or undergone something that negatively affects performance at work. The forensic professional needs to understand the expectations of a given work-related position and the employee's usual duties and responsibilities in order to make informed recommendations or decisions. If impairment is identified, the evaluation must address interventions that are indicated before an employee can return to work. Certain accommodations may need to be made to permit the employee to return to work in spite of his or her difficulties or limitations. The FFD evaluation encompasses risk management, behavioural health care, elements of labour law and internal disciplinary procedures.

A review of the employee's work-related history and an investigation of any recent incidents help in the assessment. If there is any history of mental illness, adherence with medication should be assessed. It may turn out that the employee's performance was unremarkable until the employee chose to discontinue psychotropic medication. Evidence of cognitive decline or other neurological problems may require the use of a neurologist or neuropsychologist. Situational factors, such as a change from day to night shift, marital conflict, company downsizing or harassment by another employee, can also cause decreased work performance.

THREAT ASSESSMENT

Threat assessment is an examination to identify and manage potentially violent-individuals in the workplace. The purpose of the assessment is to keep a potentially dangerous situation from materialising into actual mayhem. The media has promoted the idea that violent behaviour is the result of an individual ‘snapping’. This is rarely the case. Those who commit violence tend to escalate their grievances and demands over a period of time before performing any acts of aggression.

The role of the clinician is to aid the referring company in distinguishing a real threat from an employee ‘blowing off steam’. May individuals utter threats without true intention to carry them out or they lack the psychological or physical capacity to commit a violent act. However, an individual who demonstrates impulsivity, displays weapons, designates a potential target (i.e. the boss or another employee) and feels justified in committing violence should be seen as a significant risk. Violence is a process and does not occur in a vacuum; it is usually the result of interactions between the perpetrating individual, the triggering environment and the lack of intervention by the referring company.

The clinician should begin by assessing the most recent act or verbalisation of violence, in as much detail as possible, as well as considering accounts of the nature and frequency of past violent acts committed by the employee. The evaluee should be asked to describe thoughts and feelings before, during and after previous violent incidents. A poor prognostic sign is the employee's incapacity to identify these thoughts and feelings.

VIOLENT PERSON PROFILE

Because violence occurs in even the most affluent work environments, the inference is that precipitants lie primarily with the perpetrating worker. The major types of individuals who become violent in the workplace are ‘disgruntled’ employees, customers or clients, stalkers, and random offenders. Bensimon (Reference Bensimon1994) found that disgruntled workplace killers were generally in their 30s to 40s and were or believed they were about to be sacked or made redundant. These individuals also believe that they have been treated poorly and have not been respected. They often file grievances and are frequent complainers.

Stalking is a behaviour and not a psychiatric diagnosis, and often involves the search for either revenge or rapprochement after the cessation of a relationship. Because a good number of violent incidents in the workplace have involved stalkers of employees, the employer's concern should extend to rebuffed partners of employees who could bring their wrath to the work site. There are observable signals among victims as well as perpetrators. Pathe & Mullen (Reference Pathé and Mullen1997) stated that 53% of stalking victims in their study reported a decrease or cessation of work or school attendance. The clinician must interview the potential victim and examine the perpetrator's history of violence, including threats. If evidence of a mental illness is evident, the clinician may be able to hospitalise the stalker involuntarily as an alternative to assisting in having criminal charges brought.

Perpetrators of violence usually have a traceable history of disputes, interpersonal conflicts and failures. They tend to have a history of threatening or assaulting others, combined with extremely poor coping skills. They also tend to have little or no family support and have difficulty accepting criticism from others, particularly co-workers. They see themselves as victims of injustice and may see others as persecuting them. Perpetrators are also known to externalise blame and hold grudges against co-workers and supervisors. Individuals diagnosed with borderline and/or antisocial personality disorders have an increased likelihood for becoming violent.

One of the best predictors of future violence is a history of violence (Reference Klassen and O'ConnerKlassen & O'Conner, 1988), with an increase in potential with each prior act of aggression. Other research has shown aggression to be a relatively stable characteristic (Reference Huesmann, Eron and LefkowitzHuesmann et al, 1984). External factors such as recent termination or lay-off may cause the perpetrator to act aggressively and/or violently. Other factors that increase the potential for violence include alcohol or drug misuse and the availability of guns. Alcohol misuse has been linked to an increased likelihood that the user will ‘misread’ situations, and has been shown to increase impulsivity.

Early studies suggested that the mentally ill were no more violent than others in the community (Reference BrownBrown, 1985). However, more recent studies report that mental illness increases the likelihood of violence (Reference Swanson, Monahan and SteadmanSwanson, 1994). In psychotic disorders, increased risk may be a function of paranoid delusions, feeling threatened or thought insertion. Feldman & Johnson (Reference Feldmann and Johnson1994), however, demonstrated that most perpetrators of workplace violence are not so markedly disturbed as to be diagnosed with a psychiatric disorder, but personality disturbances were noted with qualities that parallel criminal populations. Low intelligence and neuropsychological impairment also are factors that increase the likelihood of violence (Reference MonahanMonahan, 1992), as well as low socio-economic status (Reference Swanson, Holzer and GanjuSwanson et al, 1990). The combination of substance misuse with other risk factors significantly increases the likelihood of violent acting out.

NEW PERSPECTIVES ON VIOLENCE IN THE WORKPLACE

Over the past few years workplace violence has attracted growing interest. With this growing interest come multiple opportunities for mental health professionals to demonstrate their knowledge and understanding of human behaviour and thereby aid in assessment, prevention and intervention of workplace violence. Although employee violence in the workplace is a relatively rare occurrence, the consequences far outweigh the cost of implementing policies, training programmes and procedures to prevent violent incidents from happening. Although Monahan (Reference Monahan1984) argued that no one psychological measure is highly predictive of future violent behaviour, we as experts in human behaviour have special capacities to predict violence (Reference MossmanMossmann, 1994). The discrimination of a violent individual from a non-violent one requires the clinician to identify a certain threshold where employees registering above the threshold will be deemed and dealt with as dangerous. The skilled and well-informed mental health professional who takes into consideration all relevant variables will decrease the likelihood of obtaining false-negative and false-positive errors. Currently, clinical psychometric measures are inadequate, suggesting that a greater emphasis should be placed on a comprehensive evaluation, including demographic data, work-related stressors, history of violence, coping skills, as well as measures related to impulsivity and antisocial behaviours. Modestly accurate algorithms for classifying violent recidivism have been shown to be useful (Reference Webster, Harris and RiceWebster et al, 1994). Models that address workplace variables and personal variables are in their infancy, but hold much promise. Research has demonstrated the importance of the interaction between personal and environmental factors in predicting violent behaviour. As mental health professionals, we can contribute to the development of forensic evaluations with regard to workplace assessments and interventions. Despite certain limitations, trained mental health professionals possess expertise in identifying individuals and circumstances that can lead to workplace violence. They also have expertise in ascertaining which individuals require treatment for mental health disorders, substance use disorders and conflictual personality disorders. The role of the mental health professional is to communicate risk assessments to the employer and recommend appropriate risk-reducing strategies that the company can comply with. By using the clinician, the company in turn can decrease the risk of legal liability and contribute to the well-being of its business culture as a whole.

Clinical Implications and Limitations