Instructions Using the Online Library, identify an article on an ethical issue in the workplace. This article should include a case study of an ethical incident by an organization. Create a case study

Risk Assessment of Workplace Violence:

Ethical and Procedural Issues

in Evaluating Employees

George E. Hargrave

ABSTRACT.

Mental health professionals are frequently called upon to

assess employees whose behaviors are seen as posing threats of work

place violence. This article discusses the ethical and procedural issues

involved in these evaluations. It addresses legal-ethical concerns associ-

ated with determining risk, the impact of assessment decisions on the

employee’s job, requirements of relevant employment law, and with

communicating results to others. Since workplace risk assessment fre-

quently involves the potential of targeted violence, the procedures are

discussed within the U.S. Secret Service framework for analyzing tar-

geted violence.

[Article copies available for a fee from The Haworth Document De-

livery Service: 1-800-342-9678. E-mail address: <[email protected]>

Website: <http://www.HaworthPress.com> ©

2001 by The Haworth Press, Inc.

KEYWORDS.

Workplace violence, risk assessment, ethical issues, em

ployee evaluation

George E. Hargrave, PhD, is affiliated with the Occupational Psychological Services

and is in independent practice as a forensic psychologist in cases involving employment

law. He is a consultant to employers and law firms throughout northern California. He is

a Diplomate in Forensic Psychology, ABPP, and provides workshops on assessment of

workplace violence for the American Academy of Forensic Psychology.

Address correspondence to: George E. Hargrave, PhD, 2740 Fulton Avenue, Suite

128, Sacramento, CA 95821.

The author wishes to express appreciation to Mary C. Hargrave, PhD, and Deirdre

Hiatt, PhD, for their review and comments on this manuscript.

Journal of Threat Assessment, Vol. 1(2) 2001



Workplace violence constitutes a growing problem in the United

States. Although such incidents are still statistically rare, their possibil

ity of occurrence is of major concern for employers. The federal Occu

pational Safety and Health Act (OSHA) and state counterparts require

that employers take steps to provide a safe workplace. Further, under

the doctrine of

respondeat superior

(let the master answer), an em

ployer can be held financially responsible for damage resulting from an

employee’s behavior, if there is evidence that the employer was negli

gent in such areas as hiring and retention practices, training, or control

of employees. Therefore, an employer that learns an employee has

threatened violence toward other co-workers or managers is required to

take action to ensure safety in the work environment.

Mental health professionals have a long history of providing services

to employers (e.g., employee assistance programs, fitness for duty and

workers’ compensation evaluations). It is, perhaps, only natural that they

have been called upon to consult in cases involving threats of workplace

violence. This new role, however, necessitates that such professionals

acquire additional knowledge and skills to be effective in their consulta-

tions.

Although the vast majority of workplace violence incidents involve

perpetrators external to the worksite, most mental health professionals

become involved in the relatively small number of cases where employ-

ees are the source of threat. These are also cases where the mental health

consultant may evaluate the identified employee. This article synthe-

sizes the current professional literature and addresses the primary ethi-

cal and procedural issues associated with evaluating employees when

determining risk of violence. It will focus on legal-ethical consider-

ations associated with determining risk, the impact of assessment deci-

sions on the employee’s job, requirements of relevant employment law,

and with communicating results to others. It will then review the proce

dural issues associated with conducting a comprehensive and defen

sible

evaluation. Finally, since workplace threats frequently involve

identified targets for potential violence, the assessment process will be

examined in the framework developed by the U. S. Secret Service for

evaluating targeted violence (Borum, Fein, Vossekuil, & Berglund,

1999; Fein, Vossekuil, & Holden, 1995).

ETHICAL-LEGAL ISSUES

Inherent in any risk assessment of violence is an element of predic

tion. As most clinicians are painfully aware, the professional literature

JOURNAL OF THREAT ASSESSMENT

is replete with caveats associated with this process. The early literature

(Monohan, 1981) portrayed efforts at violence prediction as dismal.

The “second generation” of studies (Lidz, Mulvey, & Gardner, 1993;

Menzies & Webster, 1995; Monahan & Steadman, 1994; Mossman,

1994; Otto, 1992), although more charitable toward clinicians’ abilities

to predict violence, were far from glowing in their support of the pro

cess. Mossman (1994), for example, summarized clinicians’ ability to

distinguish violent from nonviolent patients as having a “modest,

better-than-chance level of accuracy” (p. 790), noting that past behavior

was frequently a better predictor than clinical judgments or cross-vali

dated actuarial techniques.

A major obstacle in any prediction of violence is the very low

base-rate of such behavior. In risk assessment, the term base-rate refers

to the “known prevalence of a specified type of violent behavior within

a given population over a given time period” (Borum, 1996, p. 946). A

low base-rate inherently leads to an unacceptably high false positive

rate, or the incorrect prediction that a given trait (e.g., high potential for

violence) exists. Although most of the previously-noted caveats were

issued in response to research on clinicians’ questionable ability to pre-

dict violence in patients, the prediction problem is even more pro-

nounced in the assessment of workplace violence. In this latter context,

base-rates are unusually low. Data reported by Bulatao and VandenBos

(1996) for the period 1980 to 1992, reflected only 0.7 per 100,000

workers to be victims of homicide, and between 1,000 and 3,000 per

100,000 workers to be victims of non-fatal workplace violence. Since

most violence was committed by individuals who had no legitimate

business relationship with the worksite, the base-rate of actual employ

ees committing violence of any type is exceeding low. Given this limi-

tation, a profile of the typical violent employee is of limited value at

best.

A second, related problem involves the lack of data relating employ

ees’ characteristics to any actual expression of violence. Several authors

(e.g., Bensimon, 1994; Johnson, Meyer, & Feldmann, 1995; Resnick &

Kausch, 1995) have identified risk factors associated with workplace

violence; however, there is little research on how such factors combine

to produce violent behavior. The reasonable premise that a large num

ber of risk factors increases the likelihood of future violence (Resnick &

Kausch, 1995) must be balanced with the knowledge that this assump

tion does not have empirical support.

Although an employer is required to take action when confronted

with a potentially violent employee, the previous concerns raise the is

George E. Hargrave

sue of whether it is ethical to use an assessment process that lacks validity

to take action against an employee. Certainly evaluators’ conclusions

about workplace violence do not have the profound impact of restrict

ing another person’s interest in “life and liberty” that Ewing (1991,

p. 161) addressed in his sweeping indictment of predicting dangerous

ness in criminal settings. They do, however, have the potential of dam

aging the person’s ability to engage in the important life activity of

work. Adopting a model of risk assessment and management, as op

posed to violence prediction, has clearly been a step toward resolving

this ethical dilemma. The process of risk assessment must still consider

the impact on the employee as well as the safety and security of others.

Numerous articles have distinguished between the processes of vio-

lence prediction and risk assessment. Heilbrun (1997) noted the general

distinction that prediction involves a single decision that must be in-

formed by the most accurate available information, and after which

there is no continued jurisdiction, while risk assessment/management

involves judgments for which there is an ongoing legal jurisdiction and

where decisions may be made or modified at different times. In assess-

ing the threat of workplace violence, there is the dual focus on deter-

mining the degree of risk posed by the individual and on developing/

implementing policies and procedures to deal with threat and protect

the workplace. This is a dynamic process that typically involves ongo-

ing assessment and multiple interventions. The evaluator, instead of

giving a prediction of violence, provides an assessment of risk by iden-

tifying risk factors and specifying interventions to address them.

One of the most frequent referrals for risk assessment occurs when an

employee threatens violence toward a supervisor or co-worker(s). Such

situations clearly require that the evaluation determine the nature and

degree of risk for potential targets. Under these circumstances, how-

ever, the employee is also frequently evaluated for risk of violence in

the more general context of determining fitness for continued job duty.

This raises the next issue, namely that these assessments must be con

ducted within the framework of employment law, the most basic com

ponent of which is the 1994 Americans with Disabilities Act (ADA;

1994).

Mental health professionals typically become involved in risk assess

ments when the identified employee’s behavior is aberrant, or there is

knowledge that the employee has a history of psychiatric treatment. In

these cases, the individual’s condition may constitute a disability under

the provisions of the ADA and the Equal Employment Opportunity

Commission’s (EEOC, 1997)

Enforcement Guidance on the Americans

JOURNAL OF THREAT ASSESSMENT

with Disabilities Act and Psychiatric Disabilities.

The ADA obviously

allows for an employer to take action if an employee’s behavior consti

tutes a “direct threat to the health or safety of the individual or others

that cannot be eliminated or reduced by reasonable accommodation”

(p. 19). The EEOC regulations explain, however, that “direct threat

means a significant risk of substantial harm to the health or safety of the

individual or others that cannot be eliminated or reduced by reasonable

accommodation” (p. 19). The significant risk must be “high and not just

slightly increased,” and the determination of threat “must be based on

an individualized assessment of the individual’s present ability to safely

perform the functions of the job, considering a reasonable medical judg

ment relying on the most current medical knowledge and/or the best

available objective evidence” (p. 19). To take personnel action against

an employee who poses a risk of violence, the employer “must identify

the specific behavior that would pose the direct threat” (p. 19). If the

employer’s action relies on a psychological evaluation, it places consid-

erable demands on the evaluator. In addition to being knowledgeable in

threat assessment, evaluators conducting these assessments must en-

sure that their procedures are consistent with standard practice for their

locale, be knowledgeable about essential duties of the employee’s job,

and be able to consult effectively with the employer on efforts to ac-

commodate the employee.

A fourth issue that arises in assessing risk of workplace violence is

that of confidentiality and informed consent. The evaluator’s role fre-

quently begins as a consultant to various members of the employer’s

risk management team, during which time there is typically much dis-

cussion of the case. Although hypothetical input from the mental health

consultant is valuable to this process, he/she must exercise caution in

commenting on the individual employee’s characteristics in the absence

of adequate data. The consultant should also assume from the outset that

he or she will be called upon to subsequently evaluate the employee.

Once this latter process is initiated, it is essential that all of the steps

necessary for an ethical forensic evaluation identified by such sources

as Melton, Petrila, Poythress, and Slobogin (1997) be followed. This

includes: (a) informing the employee of who initiated the evaluation;

(b) the professional’s role as an evaluator as opposed to a therapist;

the evaluation; (e) the foreseeable uses of evaluation data; (f) the pro

posed evaluation procedures; (g) the kinds of information that may re

quire disclosure to third parties; and (h) issues surrounding the legal

right to decline participation.

George E. Hargrave

Of particular significance is the issue of confidentiality. It is impor

tant for the evaluator to know how the jurisdiction’s statutes governing

medical records pertain to employment-related evaluations. For exam

ple, an employee undergoing such an assessment in California is con

sidered a

patient

under the statute governing the confidentiality of

medical records (

Pettus v. Cole,

1996). This exists in spite of the fact

that there is no

doctor-patient

treatment relationship. The release of any

clinical data to the employer can occur only with the specific written

consent of the employee. In the absence of this consent, the evaluator

can report back only the recommended action. Under all circumstances,

the evaluator must minimize intrusions into the employee’s life and

limit any feedback to that which is germane to the purpose of the evalu-

ation.

One of the exceptions to most states’ confidentiality laws is the men-

tal health professional’s duty to warn or protect potential victims of vio-

lence based on

Tarasoff v. Regents of the University of California,

(1976) and subsequent legal decisions. Whether this issue applies to

such professionals serving as evaluators depends on the context (Mel-

ton et al., 1997). Evaluators assessing an individual for the threat of

workplace violence must, therefore, know their jurisdictions’ legal

standards for the duty to warn or protect an identifiable third party when

they know, or should know, that the individual constitutes a significant

threat of violence. Resnick and Kausch (1995) note that in such situa-

tions the evaluator may need to take such protective steps as notifying

the employer or coworkers, contacting law enforcement, hospitalizing

the employee, or taking other reasonable actions.

PROCEDURAL ISSUES

Over the past twenty years, several authors have proposed proce

dures for assessing risk of violence. Monahan’s (1981) prototypic work

on violence prediction was important, because it shifted the assessment

process from variables internal to the subject to a focus on factors that

precipitated former violence; these included the context in which vio

lence occurred, the individual’s history of violence, base-rates of vio

lence for similar individuals, stressors in the individual’s life, the

similarity of current context to former contexts associated with vio

lence, and the availability of both means and victims for violence. More

recently, the U.S. Secret Service adopted a schema for assessing the

general risk of targeted violence (Borum et al., 1999; Fein & Vossekuil,

JOURNAL OF THREAT ASSESSMENT

1998; Fein et al., 1995). This latter approach does not rely on subject

profile data or threats as a threshold for risk; rather, it is based on three

principles and the responses to ten key assessment questions (Borum et

al., 1999). As noted previously, this framework appears to be well

suited for workplace risk assessments due to the potential for targeted

violence against others in this setting (e.g., supervisors, co-workers, hu

man resource personnel).

The three fundamental principles underlying the Secret Service as

sessment approach involve the interaction between the subject and

his/her environment. The first principle assumes that targeted violence

results from an understandable process of a subject’s thinking and be

havior. The second principle states that violence emanates from the in-

teraction between the subject and past stressful situations, current

situation, and the targeted victim. The third principle views an identifi-

cation of the subject’s “attack-related” behaviors as critical to the inves-

tigation and resolution of the threat assessment (Borum et al., 1999).

The following discussion will examine the procedural issues for assess-

ing the threat of workplace violence in the framework of the ten ques-

tions outlined by Borum and his colleagues (1999).

Question 1: What motivated the subject to make statements or take

action, that caused him/her to come to attention?

When assessing the threat of workplace violence, the process of an-

swering this question is typically underway prior to the mental health

consultant’s arrival. An increasing number of companies and govern-

mental agencies have adopted procedures for dealing with workplace

violence, such as those modeled after guidelines published by the Inter

national Association of Chiefs of Police (1997), Littler, Mendelsen,

Fastiff, Tichy, and Mathiason (1996), and other sources. In such cases,

employers approach the investigation through risk management teams

and procedures they have in place. In the event that there is no organiza

tional framework for handling threats, the mental health consultant can

provide a valuable service in assisting the employer with this process.

This, of course, requires that the consultant have a sound knowledge of

issues the employer must attend to as well as the procedures for address

ing them.

Once the mental health consultant enters the assessment process, a

typical starting point is to review information that has been collected on

the threatening situation and to interview the individuals who provided

this information. The employee’s personnel records are generally an

George E. Hargrave

initial focus and provide such data as identifying information, work his

tory and performance ratings, grievances, and disciplinary actions. Risk

assessments differ from typical clinical evaluations by their emphasis

on data obtained from collateral sources. Input from co-workers and su

pervisors provide valuable information about the employee’s job status,

work and personal problems, aggressive displays, and statements of

threat. In some situations, concerned friends and/or family members are

willing to be interviewed; obviously, they are a valuable source in as

sessing the subject’s motivation. In addition to obtaining others’ per

ceptions of the reason for referral, it is important that interviews with

collateral sources focus on characteristics of the subject shown by the

professional literature to be associated with workplace violence (e.g.,

Bensimon, 1994; Johnson et al., 1995; Resnick & Kausch, 1995). This

includes information on the individual’s personal history, personality

characteristics, past incidents of aggression, ownership of weapons, and

substance abuse, as well as current factors that signal possible escalat-

ing violence (e.g., indications of increased problems outside of work,

increased expression of anger, drastic change in belief systems, de-

creased emotional control).

Question 2: What has the subject communicated to anyone

concerning his/her intentions?

This question can obviously be investigated during the interviews

noted in Question 1. When assessing threatening communications by

the subject, collateral sources should be queried about any unusual or

inappropriate ideas that indicate desperation, decompensation, and po

tential targets of violence.

If the data suggest that the employee poses a threat to others, he or

she should be set up for an immediate individual evaluation. In consul

tation with the evaluator, the employer must also implement an immedi

ate plan for protecting others. However, in many cases this initial

process does not indicate a need for immediate action and the employee

can be evaluated at a later date. This may occur when the problem was

resolved by some administrative action or when the employee entered

the hospital or other treatment. These latter cases, however, frequently

involve a risk assessment when the employee is cleared to return to

work.

Obviously, the subject’s motivation will be a focus of interviews

with the individual. When interviewing the employee, it is important to

use a structured format that contains the areas identified as relevant by

JOURNAL OF THREAT ASSESSMENT

the risk assessment literature (e.g., Heilbrun, 1998; Johnson et al., 1995;

Monahan, 1981; Resnick & Kausch, 1995). These will be covered in the

response to Question 5.

Once information is collected from both collateral sources and the in

dividual, it can be

organized in a framework that assesses individual/ per

sonal variables (e.g., substance abuse, violence history, psychopathology,

impulse control, weapons access, etc.), situational factors (e.g., job prob

lems, conflicts with co-workers, other stressors in the individual’s life),

and the interaction between these two sets of variables (Barling, 1996;

Heilbrun, 1998). Overall, the data should provide an understanding of

the employee’s motivation for the threatening behavior.

Question 3: Has the subject shown an interest in targeted violence,

perpetrators of targeted violence, weapons, extremist groups,

or murder?

Borum and his colleagues (1999) noted that, “some perpetrators of

targeted violence show an unusual interest in acts similar to the one they

are planning,” and that the individual’s affiliation with extremist groups

may provide “permission” for such an act (p. 332). If there is evidence

of extremist group membership, the assessment focus should include

such areas as those identified by Pynchon and Borum (1999), including

distorted perceptions (e.g., prejudices), tendencies to dehumanize non-

group members, a diminished sense of accountability or self-awareness,

support of violence as a legitimate means to accomplish an objective,

and willingness to violate group norms toward more deviant behavior.

Although some authors have listed fascination with workplace vio

lence as a warning sign (e.g., Hatcher & Reynolds, 1996), how this in-

terest correlates with actual violence is unknown. If present, such

interests should be assessed both in discussions with collateral sources

and during the interview(s) with the individual. These inquiries should

also include questions about any individuals the employee perceives as

enemies. An expressed interest in weapons is an important assessment

area, since it likely correlates with possession of weapons. Given the

lethality of guns and their prevalence in workplace violence (Johnson,

Meyer, & Feldmann, 1995), a weapons assessment should focus on

such factors as the number and types of guns owned, any recent acquisi

tions, whether the individual legally or illegally carries a gun, any re

cent renewed interest in guns, any history of brandishing a gun, and any

incidents of using a gun or other weapon in a conflict. Information ob

tained in response to this question provides additional data as to the in

George E. Hargrave

dividual’s motivation for violence as well as an assessment of potential

targets and the means to carry out violence.

Question 4: Has the subject engaged in attack-related behavior,

including any menacing, harassing, and/or stalking-type behavior?

Responses given to questions about the subject’s interest in targeted

violence, weapons, and extremist groups naturally lead into an assess

ment of the individual’s history of violence and intimidation. Clearly,

evidence of past aggression and violence is the primary predictor of

subsequent violence, particularly if the circumstances of the past and

future incidents are similar. Thus, any evaluation should focus on deter-

mining frequency and severity of such acts, circumstances under which

they occurred, the contribution of drugs and/or alcohol to the process,

whether the acts were planned or impulsive, and the subject’s thoughts

and feelings associated with each incident. In circumstances where the

behavior was associated with actual violence, it is also important to as-

sess the types and severity of acts, the role of weapons, and the specific

contexts under which violence occurred. This information should be

sought from both the individual and collateral sources. The interview

focused on current concerns should include any indications of specific

targets and the individual’s access to weapons.

One particular concern is stalking behavior, which is defined as “the

willful, malicious, and repeated following of another that threatens his

or her safety” (Meloy & Gothard, 1995, p. 258). The act of following

that does not rise to this level is generally referred to as

obsessional fol-

lowing

(Meloy, 1996). In the data reported by Johnson and colleagues

(1995), stalking or hostage-taking occurred in approximately one-third

of workplace violence cases. Data compiled by Hargrave (1999, 2000)

comparably found stalking or obsessive following to have occurred in

approximately one-third of cases where an actual threat of violence was

found. Given this prevalence of stalking or obsessional following

among individuals who perpetrate workplace violence, it becomes an

important area to assess. A relatively frequent threat of workplace vio

lence involves the stalking of employees by intimate or formerly inti

mate others. In such instances, it may become necessary to protect the

targeted employee and co-workers without being able to evaluate the

stalker. This typically requires that the evaluator interview the victim to

assess the relationship between victim and stalker. In addition, any ob

tainable information from the victim or others should be used to assess

the suspect’s history of general and domestic violence, psychiatric and

JOURNAL OF THREAT ASSESSMENT

Top of Form

/ 21

Bottom of Form