Understanding and Managing Workplace Aggression

14
Understanding and Managing Workplace Aggression MARK J. MARTINKO SCOTT C. DOUGLAS PAUL HARVEY T here have been numerous reports of violence perpetrated by disgruntled members of organizations. Many of these events were perpetrated by troubled indivi- duals who blamed their colleagues for their personal problems. In one such case, a man named Michael McDermott shot and killed several coworkers at a consulting company in Massachusetts. He explained that he was angered by his company’s agreement to med- iate a tax dispute between himself and the Internal Revenue Service (IRS). Another example is Salvador Tapia, who killed six former coworkers after being terminated from his position at a supply warehouse. He spared some employees, explaining that he was only after those who deliberately wronged him. Other memorable examples include inci- dents at the U.S. Post Office, a day trader in Atlanta who turned on his colleagues, and, most recently, a student on an Indian reserva- tion in Minnesota who turned on his class- mates. Accompanying these well-publicized incidents has been a growing number of the- oretical and empirical works that attempt to explain workplace aggression. The purpose of this article is to describe and synthesize what has been done up to this point, so that both students and practitioners can develop a working model of the dynamics of aggression that can guide proactive measures to reduce it. FORMS AND DEFINITIONS OF AGGRESSION The most sensational form of aggression is physical violence. However, other forms of aggression – such as stealing, sabotage, gos- sip, incivility, and purposeful work slow- downs may also have detrimental consequences. In general, organizational aggression is any behavior that is intended to harm a person in the organization or the organization itself. Organizational aggression can be viewed as a continuum, with physical violence at the highest extreme graduating down to less severe forms (see Fig. 1). MODELS OF ORGANIZATIONAL AGGRESSION The early models of organizational aggres- sion took a sociological approach, emphasiz- ing organizational factors such as working conditions and policies. One of the more impressionistic models is the popcorn model by Robert Folger and Daniel Skarlicki, in which they likened organizational factors to hot oil and described organizational mem- bers as individual kernels. Their main point was that organizational conditions precipi- tate aggression, and that we are more likely to be able to predict the occurrence of aggres- sion by paying attention to organizational conditions (i.e., the pressure created by the heating oil) than trying to predict which person (i.e., kernel) will explode. The early research in this area generally supported these sociological models. It showed low- level but significant correlations between organizational factors such as rigid rules and procedures and incidents of aggression. Organizational Dynamics, Vol. 35, No. 2, pp. 117–130, 2006 ISSN 0090-2616/$ – see frontmatter ß 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.orgdyn.2006.03.002 www.organizational-dynamics.com 117

Transcript of Understanding and Managing Workplace Aggression

Page 1: Understanding and Managing Workplace Aggression

Understanding andManaging Workplace

AggressionMARK J. MARTINKO SCOTT C. DOUGLAS PAUL HARVEY

Organizational Dynamics, Vol. 35, No. 2, pp. 117–130, 2006 ISSN 0090-2616/$ – see frontmatter� 2006 Elsevier Inc. All rights reserved. doi:10.1016/j.orgdyn.2006.03.002www.organizational-dynamics.com

T here have been numerous reports ofviolence perpetrated by disgruntled

members of organizations. Many of theseevents were perpetrated by troubled indivi-duals who blamed their colleagues for theirpersonal problems. In one such case, a mannamed Michael McDermott shot and killedseveral coworkers at a consulting company inMassachusetts. He explained that he wasangered by his company’s agreement to med-iate a tax dispute between himself and theInternal Revenue Service (IRS). Anotherexample is Salvador Tapia, who killed sixformer coworkers after being terminated fromhis position at a supply warehouse. He sparedsome employees, explaining that he was onlyafter those who deliberately wronged him.

Other memorable examples include inci-dents at the U.S. Post Office, a day trader inAtlanta who turned on his colleagues, and,most recently, a student on an Indian reserva-tion in Minnesota who turned on his class-mates. Accompanying these well-publicizedincidents has been a growing number of the-oretical and empirical works that attempt toexplain workplace aggression. The purpose ofthis article is to describe and synthesize whathas been done up to this point, so that bothstudents and practitioners can develop aworking model of the dynamics of aggressionthat can guide proactive measures to reduce it.

FORMS AND DEFINITIONS OFAGGRESSION

The most sensational form of aggression isphysical violence. However, other forms of

aggression – such as stealing, sabotage, gos-sip, incivility, and purposeful work slow-downs – may also have detrimentalconsequences. In general, organizationalaggression is any behavior that is intendedto harm a person in the organization or theorganization itself. Organizational aggressioncan be viewed as a continuum, with physicalviolence at the highest extreme graduatingdown to less severe forms (see Fig. 1).

MODELS OFORGANIZATIONALAGGRESSION

The early models of organizational aggres-sion took a sociological approach, emphasiz-ing organizational factors such as workingconditions and policies. One of the moreimpressionistic models is the popcorn modelby Robert Folger and Daniel Skarlicki, inwhich they likened organizational factorsto hot oil and described organizational mem-bers as individual kernels. Their main pointwas that organizational conditions precipi-tate aggression, and that we are more likelyto be able to predict the occurrence of aggres-sion by paying attention to organizationalconditions (i.e., the pressure created by theheating oil) than trying to predict whichperson (i.e., kernel) will explode. The earlyresearch in this area generally supportedthese sociological models. It showed low-level but significant correlations betweenorganizational factors such as rigid rulesand procedures and incidents of aggression.

117

Page 2: Understanding and Managing Workplace Aggression

FIGURE 1 A CONTINUUM OF AGGRESSIVE BEHAVIORS IN ORGANIZATIONS

Anecdotal evidence also supports thisnotion, with notoriously rigid regulationsreceiving much of the blame for the episodesof aggression at the U.S. Postal Service in theearly 1990s.

More recent models, which include ourown work in attribution theory, describe howhuman factors interact with environmentalfactors. The basic notion of these more recentperspectives is that the ways individualsprocess information about their successesand failures (i.e., causal reasoning) is signifi-cantly related to whether or not they becomeaggressive. Thus, a person who is laid off andhas a pattern of making external attributionswill tend to blame the management (externalattribution) while another employee in thesame situation who tends to internalize failure

118 ORGANIZATIONAL DYNAMICS

FIGURE 2 A MODEL OF ORGANIZATIONAL

(2002), MARTINKO AND ZELLARS (1998(2002); MARTINKO, DOUGLAS, HARVEY

DOUGLAS, HARVEY, AN

will blame herself. Obviously, the personwho blames management is more likely tobecome aggressive than the person whoblames himself. Thus, our research demon-strates that understanding individual differ-ences significantly increases our ability topredict aggression.

A MODEL OFORGANIZATIONALAGGRESSION: OVERVIEW

A summary of the major individual andorganizational factors related to aggressionis provided in Fig. 2. The model is based onour own research as well as an integration ofwork that has been conducted over the last 15

AGGRESSION (ADAPTED FROM MARTINKO

); MARTINKO, GUNDLACH, AND DOUGLAS

AND GUNDLACH (2006); MARTINKO,D JOSEPH (2005))

Page 3: Understanding and Managing Workplace Aggression

years. A focal point of our model is that itemphasizes that the key to understandingaggression is to understand individuals’ cau-sal reasoning processes (i.e., their attribu-tions) for trigger events.

More specifically, the model depicts thechain of events that leads to aggressive beha-vior, starting with a trigger event such asbeing laid off. Both human and organiza-tional factors influence the frequency andintensity with which individuals experiencetrigger events. However, in our model, a keydeterminant of aggressive behavior is thebeliefs that people have about the causes ofthe trigger event. As we will describe in moredetail below, individuals who attribute thetrigger events to external, stable, intentional,and controllable causes are the most likely tobecome aggressive. On the other hand, peo-ple who make internal attributions andblame themselves are less likely to becomeaggressive.

As we will show later, causal-reasoningstyles can be measured and are identifiable,which enables us to anticipate aggressive-ness with more precision than in the past.Next, we will explain the variables in themodel emphasizing how this knowledgecan be used to manage aggression.

THE ORGANIZATIONALENVIRONMENT

Rigid Rules and Procedures

This factor has been identified in numer-ous analyses of aggressive incidents. Perhapsthe best example of rigidity comes from anincident that occurred at the U.S. Post Office.In this case, aggression was triggered whenan employee was refused permission to leavehis letter-sorting machine after his wife hadcalled and indicated that his house was onfire.

An obvious application is to reduce rigid-ity. However, a rule of thumb is that percep-tions of rigidity are guided by employees’perceptions that the policies of the organiza-tion violate norms for similar organizations.For example, a retailer that requires employ-

ees to carry transparent purses and does notallow employees to take normal breaks – ascompared with similar retailers – may beperceived as attacking the employees. Thus,a recommendation is to communicate infor-mation about the policies of similar organi-zations so that employees can establish arealistic baseline for their norms. In addition,management needs to make sure that when-ever rules and policies exceed industrynorms, they provide a solid rationale forthe deviations. For example, rigid policiesregarding theft might be justified by showingthat the organization averages 10% lessshrinkage than competitors, resulting in apay scale that is considerably better thanthe industry average.

Adverse Work Conditions

A significant amount of research has alsodemonstrated that there is a low-level corre-lation between aggression and adverse work-ing conditions such as high levels of noise,overcrowding, high or low temperatures,and threats to safety. As in the case of rigidrules and procedures, it would seem thatworking conditions are most likely to beassociated with aggression when they exceedthe norms for that type of work. Again, theapplication is for organizations to clearlycommunicate to employees that their work-ing conditions meet or exceed industry stan-dards, and, when conditions are harsh andcannot be improved, to provide reasonableexplanations for the current state of the workenvironment.

Aggressive Cultures

Research has generally confirmed thatthere are differences in the incidence ofaggression between cultures. A central tenetof Bandura’s social learning theory is thatpeople learn behaviors, and in particular,aggressive behaviors, by observing and mod-eling others.

Managers need to be aware of culture atthree levels. First, there is the social culture inwhich the organization is embedded. The

119

Page 4: Understanding and Managing Workplace Aggression

incidence of violent behaviors and the natureof violent behaviors are different in differentgeographical areas. Examples are high crimeareas throughout the country that may beboth urban and rural.

A second level of culture is occupational.Examples are professional sports such asfootball and basketball, but also other occu-pations such as ironworkers and police offi-cers, all of which are probably above thenorm in aggressive behaviors.

A third level of culture is the various sub-cultures within organizations. Clearly the cul-ture of the U.S. Army, Walt Disney Co., andMicrosoft Corp. are different. However, wewould also expect differences within organi-zational cultures. In the U.S. Army, the GreenBerets undoubtedly exhibit more aggressivebehaviors than the part of the army devoted tologistics and support. Likewise, we feel quitesure that the lawyers for Microsoft are moreaggressive than the programmers.

Recognizing differences in culture, it isclear that policies and procedures designedto manage aggression need to be tailored tospecific cultures. This implies that policiesmay need to recognize differences both withinand across organizational cultures. Thus, thenotion that policies should be consistentacross an organization may be misguided.

HUMAN FACTORS

The environmental and organizational fac-tors described above do not affect all indivi-duals equally. Some people appear tobecome aggressive with little or no provoca-tion, whereas others appear to retain theircomposure regardless of organizational andenvironmental influences. Individual factorsthat appear to be related to the proclivity toact aggressively are reviewed below.

Gender

The research in this area is fairly unequi-vocal. Males are more likely to express overtaggression and be incarcerated for violentcrimes than females. For instance, note that

120 ORGANIZATIONAL DYNAMICS

all of the perpetrators described in the intro-duction were men. On the other hand,research indicates that gender differencesmay not be as great for less overt forms ofaggression, such as stealing and gossip.

This research suggests that incidents ofovert aggression could be managed by hiringdisproportionately more females than malesto staff particularly stressful positions. Forexample, at least in some organizations, itappears that there are a disproportionatenumber of females in stress-producing cus-tomer relations positions. Although we areunaware of research in this area, it may bethat females are less likely to react to custo-mers aggressively than males. As we writethis we are aware such hiring practiceswould not be politically correct, and mayalso be in violation of fair hiring practices.More about these conflicts will be said in thediscussion of managerial implicationstoward the end of this article.

Negative Affect ivity

Negative affectivity is a predisposition forindividuals to express dissatisfaction and toexperience negative psychological statessuch as stress, frustration, and anger. Whilethere has been very little research investigat-ing the relationship between negative affec-tivity and aggression, related constructs suchas depression have been cited in the mediaand in court documents as factors leading toaggression. An example is the 1989 work-place shooting in which Joseph Wesbecker,who had a history of depression, killed eightpeople at a printing plant.

Numerous studies have found that nega-tive affectivity is associated with job dissa-tisfaction and stress. One particularlyinteresting finding is that negative affectivityappears to account for a significant propor-tion of job satisfaction, suggesting that jobdissatisfaction may have more to do withindividual predispositions rather than jobcharacteristics. An obvious application ofthis finding is that hiring people who arealready positive in their affectivity can sig-nificantly increase overall job satisfaction. On

Page 5: Understanding and Managing Workplace Aggression

121

the other hand, the research suggests thateven the best efforts to increase job satisfac-tion by improving organizational conditionsmay fail to have marked effects on someindividuals. Assuming that stress and jobdissatisfaction are also related to aggressivebehavior, the research implies that aggres-sion could be reduced by screening for nega-tive affectivity during hiring.

Trait Anger

The literature distinguishes between twotypes of anger related to aggression: state andtrait anger. State anger is related to a person’sresponse to a specific incident. Trait anger isthe tendency to experience state anger overtime and in different situations. Some indi-viduals are higher on trait anger than others;our research indicates that these individualsare more likely to report and becomeinvolved in incidents of aggression. Anexample is Douglas Williams, who experi-enced chronic anger and took anger manage-ment classes to control his anger. He killedfive coworkers at a Lockheed Martin Corp.plant in 2003.

The same research suggests that indivi-duals high in trait anger are more likely thanothers to perceive situations as anger-pro-voking. An obvious implication of thisresearch is to screen employees for traitanger, particularly in work environmentsthat are high in the dimensions related toaggression discussed above.

Causal-Reasoning Styles

Causal reasoning concerns how peopleexplain the causes of their successes andfailures. Several different theories contributeto the notion of causal reasoning theory.Locus of control theory asserts that peoplehave consistent ways in which they accountfor their successes and failures. ‘‘Internals’’believe that both their successes and failuresare functions of their own abilities andefforts, whereas ‘‘externals’’ believe that out-side factors such as the environment andother people control their destiny.

Attribution theory is closely related to theconcept of locus of control. Attribution the-ory is less general in its approach and looks athow people’s attributions (i.e., beliefs) aboutthe causes of their successes and failuresaffect their emotions and behaviors in spe-cific situations. There are a variety of ways toclassify attributions, but the most accepteddimensions are locus of causality (internal orexternal), stability, controllability, and inten-tionality. We have postulated and foundsupport for the notion that individuals aremost likely to exhibit aggressive behaviorwhen they perceive that the cause of a nega-tive event is external, intentional, controlla-ble, and stable—such as a chronicallyabusive supervisor.

Although attribution theory generallyfocuses on the behaviors and emotions result-ing from specific attributions, it has also beenfound that people have attributional tenden-cies or styles that are related to their behaviorand emotions. More specifically, within thecontext of aggression, both theory andresearch have supported the notion of a hos-tile attribution style, which predisposes indi-viduals toward aggression. As suggestedabove, hostile attribution styles are character-ized by external, stable, controllable, andintentional attributions for failure and overlyoptimistic attributions (i.e., internal, stable,controllable, and intentional) for success.

The relationship between hostile attribu-tion style and aggression has been well docu-mented in the psychological literature.Anecdotally, a case in which a Xerox Corp.employee named Byran Uyesugi killed sevencoworkers provides an illustration of this rela-tionship. Although Uyesugi did not give aclear motive for the killings, witness testi-mony suggested that he had a deeplyingrained tendency to blame external factorsfor his problems. This suggests that Uyesugipossessed a hostile attribution style, whichmay have predisposed him toward aggressivebehavior.

Our own research has demonstrated sig-nificant differences in organizational aggres-sion related to attribution styles. In onestudy, we compared a sample of MBA stu-

Page 6: Understanding and Managing Workplace Aggression

dents with a sample of prisoners who hadbeen convicted of violent crimes and foundthat the prisoners fit the hostile attributionpattern much more closely than the MBAs. Inanother study, we found that the incidents ofaggression in both a transportation companyand a school system were related to external,controllable, stable, and intentional attribu-tions for negative outcomes. Thus, people’sattribution patterns appear to be a key vari-able explaining differences in individualreactions to environmental events.

THE DYNAMICS OF THEPROCESS: CAUSALREASONING, EMOTIONS, ANDAGGRESSION

Predicting aggression in organizations isanalogous to predicting the perfect storm.As in the case of traumatic weather eventssuch as tornados and hurricanes, althoughwe have a reasonable understanding of thefactors that contribute to the phenomena, itoftentimes fails to emerge when it is antici-pated, and experts are oftentimes surprisedwhen it does appear. In a 1999 RadissonHotel shooting, for instance, the perpetrator,an employee named Silvio Izquierdo-Leyva,gave no emotional or behavioral clues that hepossessed any violent tendencies. ‘‘We hadno problems with him . . . he just snapped,’’was one coworker’s observation.

The social and physical dynamics sur-rounding Hurricane Katrina provide a vividillustration of the difficulties in predicting thedynamics of both physical and social systems.Despite detailed analyses of the forecastedpath of Hurricane Katrina, the hurricane’slanding apparently surprised authorities,who didn’t order an evacuation of NewOrleans until the day before the storm. Simi-larly, although in hindsight experts canexplain the aggressive behaviors that accom-panied the aftermath of Katrina, those whowere in authority apparently did not antici-pate the size and scope of the problem.

In this section, we make a rudimentaryattempt to explain the dynamics of organiza-

122 ORGANIZATIONAL DYNAMICS

tional aggression. However, it is important todifferentiate between instrumental and orga-nizational aggression before we begin ourexplanations. As indicated in the introduc-tion, organizational aggression is any actintended to harm an organization or its mem-bers. Instrumental aggression, on the otherhand, is aggressive behavior that is used toacquire resources. Thus, the motives foraggression on the part of a bank robberare directed toward acquiring the bank’smoney while in the case of organizationalaggression, the behavior is retaliatory, result-ing from an actual or perceived injury orinjustice.

Trigger Event

According to our model, organizationalaggression is triggered by negative outcomesthat are perceived to be unjustified, such as alayoff or reprimand by a supervisor. Threemajor types of injustice have been identified:procedural (policies were not followed), dis-tributive (the outcome is unfair given theamount of effort that was expended or ascompared to the rewards received by others),and interactional (the communication of theoutcomes was disrespectful and devaluedthe recipient). Thus, a person who does notreceive an expected promotion could feelinjured by the organization because proce-dures were violated, their efforts were notfairly rewarded, or because they felt deva-lued and disrespected by the way the out-come was communicated.

Organizational and IndividualFactors

As our model indicates, the degree towhich injustice is perceived will be influ-enced by both the characteristics of the orga-nization and the traits of the individual. If theorganization already has a culture of aggres-sion, rigid rules and policies, and stressfulworking conditions – and the employee has adifficult task along with a hostile attributionstyle, negative affectivity, and trait anger – itis more likely that the individual will per-

Page 7: Understanding and Managing Workplace Aggression

ceive a negative outcome as unjust and bemotivated to make a hostile attribution.

The dimensions of attributions have beenfound to influence both emotions and beha-vior. In particular the locus of causalitydimension has been found to affect emotions.When people make an internal attribution fora negative outcome they are likely to blamethemselves and feel badly. On the other hand,external attributions direct negative emotionsoutward. If the person believes that the causeof a denied promotion is an unfair supervisor,the person is likely to feel anger directedtoward the source of injustice. While therehas been less research on dimensions besideslocus of causality, the research that has beendone, along with the theoretical literature,indicates that if the person also believes thatthe cause of the negative outcome was inten-tional, controllable and not likely to change(i.e., stable), they are also likely to feel angertowards the perceived cause.

Emotions and Aggression

The final stage before aggression is emo-tions. Our model and research, as well theresearch evolving in psychology, indicatesthat behavior is often driven by emotions.More specifically, in the case of aggressivebehavior, anger almost always precedesorganizational aggression and may be thebest predictor. Thus, even though all of theindividual characteristics and organizationalcharacteristics described above may be pre-sent, unless the person perceives injusticeand becomes angry about it, aggression isunlikely. Emotions and the causal reasoningthat leads to the emotions appear to be cri-tical factors in understanding when aggres-sion is most likely to occur.

Several applications emerge from the dis-cussion above. Although it may seem some-what simplistic, anger is a major cue thatmanagement must address to reduce inci-dents of aggression. Second, anger is pre-ceded by perceived injustice that isattributed to stable, controllable, externaland intentional causes such as abusive man-agement practices. As a result, anger may be

attenuated if management addresses the cau-sal reasoning of the perceived victims andcan give a credible explanation that con-vinces the victim that the perceived injurywill not continue (unstable), was not control-lable, and was not intended by management.In the event that management was the legit-imate cause of the negative outcome, angermay still be mitigated by convincing thevictim that the outcome was not intentionalor controllable, and will not happen again(unstable). Finally, in order for such a strat-egy to work, it is imperative that the lines ofcommunication are open both ways. Man-agement must communicate, but it must alsodevelop channels for feedback and listen towhat employees are saying.

REDUCING AND MITIGATINGAGGRESSION

In the discussion above, we emphasizedresearch and theory that support our modelof organizational aggression. Along the way,we also pointed out some of the more obviousimplications. In this part of the paper we focuson the proactive and defensive strategies man-agements can use to reduce aggression.

Developing an Action Plan

Based on our model and theory, the follow-ing steps are recommended for developingstrategies to mitigate organizational aggres-sion: (1) select and empower a task force; (2)meet and identify the basic strategy, needsand issues; (3) collect data and informationboth inside and outside of the organization; (4)develop an action plan; (5) present the actionplan to management; (6) modify and revisethe plan if necessary; (7) implement, and (8)feedback and follow-up. We will now discussthese steps in more detail.

Task force identification. The process ofdeveloping an action plan for reducing andmanaging aggression is probably best carriedout with representation from all parts of theorganization. Broad representation has a

123

Page 8: Understanding and Managing Workplace Aggression

number of advantages. It demonstrates thatmanagement is serious about managingaggression, increases the probability thatproblems will be identified, and facilitatesboth the communication and support of theprogram once implemented.

In general, the task force should includethe major internal stakeholders in the orga-nization including respected members ofsenior management, a cross section ofemployees, representation from each of themajor functional areas of the organization,and a facilitator with the expertise needed toassist the group.

Identification of process, needs, and issue-s. Here, the group identifies the type of dataand information that will be needed, the size,nature and scope of the intervention, andestablishes roles and responsibilities withregard to data collection.

Data and information collection. This stepshould start with basic questions and infor-mation about each of the organizational fac-tors identified in the model of aggressiondepicted in Fig. 1. Each of these areas andthe considerations that should be addressedare discussed below.

Organizational policies and practices. Acomparative analysis of the rigidity andreasonableness of the policies and practicesin the organization compared to otherorganizations should be conducted. Thus,for example, policies meant to deter theftin retail organizations should be comparedwith similar policies in similar organizations.Examples of questions are as follows: If thiscompany requires transparent purses, doother similar retailers have the samerequirement? If so, what is the shrinkagerate in companies that have, versus do nothave, this policy? Are the break times andworking hours in this company similar to, ordifferent from, those of our competitors?

In addition to the general audit of organi-zational policies, there should also be anexamination of the policies that have beendeveloped with respect to aggression. Some

124 ORGANIZATIONAL DYNAMICS

important questions to consider are whetheror not there are specific policies; whether thepolicies define aggression in a way thatenables managers and other organizationalmembers to identify and report aggressiveincidents; whether policies on aggression arecommunicated to the employees; whetherthere is a clear communication point forreporting incidents of aggression; andwhether there are plans in place to deal withdifferent types of incidents.

Finally, organizations and managementsshould be aware that the Occupational Healthand Safety Act of 1974 requires organizationsto have in place policies and procedures toensure the health, safety and welfare of theiremployees. Many experts today believe thataggressive behavior needs to be disciplinedimmediately to keep it from escalating andencourage policies that indicate zero tolerancefor harassment, bullying, intimidation, orthreats. The U.S. Post Office is often cited asan example of the effectiveness of zero-toler-ance policies that are credited for significantreductions in aggressive behaviors. Whensuch a policy is implemented, it is recom-mended that all employees are given a writtencopy of the policy and be required to sign astatement indicating that they have read andwill comply with the policies.

Working conditions. Some potential ques-tions with respect to working conditionsare: Is the safety record of this companyabove or below the industry average? Arepressures for sales and production in thiscompany higher or lower than competitors?Do some jobs in the company have morestressful working conditions than others?Are management policies and proceduresthe same, or do they differ in different partsof the organization? Are there low cost waysto improve working conditions and percep-tions of working conditions? Are there anyproblems with temperature control, cleanli-ness, or working hours?

Culture. With respect to culture, relevantquestions are: Is there a culture ofaggression in this organization or any part

Page 9: Understanding and Managing Workplace Aggression

of the organization? Do prospectiveemployees come from cultures that deviatefrom the norm in terms of aggressiveness? Arethere any policies, practices, or rituals thatencourage aggression—such as initiations,formal or informal? Are there proceduresthat encourage reports of aggressive orunethical behavior (e.g., steroid use)? Arethere cases of abusive supervision? Is therea history of whistle blowing in theorganization and how are whistle-blowerstypically treated (e.g., Jose Conseco)?

Human factors audit. The purpose here is toidentify human factors related to aggression.We recognize that there are privacy, legal,and ethical constraints on both acquiring andusing information about human factors.However, considering the OccupationalHealth and Safety Act of 1974, there are legaland moral implications that arise from thefailure to manage conditions that create adangerous and hostile work environments.We will try to strike a balance between theseethical and legal concerns.

The first step is to observe the frequencyand intensity of the four factors depicted inFig. 1. Although the degree to which thefactors can be observed varies, measurementprocedures are available for all four factors.Obviously gender is directly observable, butthere are also a number of psychometricscales that measure masculine and femininedispositions as well as androgyny. Thesescales have been used in the study of leader-ship and, in general, measure the degree towhich a person is goal-oriented, relationship-oriented, or is able to integrate both goals andrelationships in their leadership styles. Thesetypes of scales might be particularly relevantif there is a culture of abusive supervisionpromulgating rigid rules and procedures.

The psychological traits described in Fig. 1can be observed directly and measured withpsychometric instruments. More specifically,a number of instruments have been devel-oped to measure attribution style, and havebeen found to be reasonably reliable andvalid. In addition, in our own work withattribution style, we found close to 90%

agreement between our judgments of a per-son’s attribution style formed during struc-tured interviews and the scores thatinterviewees attained on our psychometricmeasure. Thus, attribution style can be mea-sured with psychometric tests, but it can alsobe identified by trained interviewers.

Several scales have been developed forassessing negative affectivity. Although weare unaware of attempts by interviewers tojudge affectivity, we would expect that, as inthe case of attribution style, trained inter-viewers could make informed judgments.

Finally, trait anger is often measured bythe STAXI inventory. In addition, state angermay also be measured through reports ofaggressive incidents from personnel records.An audit of personnel records may be able toidentify areas or subunits of the organizationwhere expressions of anger are most preva-lent.

Action planning. This process typicallystarts with a one- or two-day workshop inwhich the data from the prior phases is usedto establish action plans. The first morningtypically begins with a review of the data.This part of the process can be facilitated byproviding summaries of the data prior to theworkshop and asking the participants tobring to the meeting a list of five to eightkey issues. The first part of the meeting thenfocuses on consensus building with regard tothe key issues with group-processing techni-ques such as nominal group decision mak-ing. This step ends with a list of the keyissues.

The second step is the establishment ofgroups to address the issues identified. Thisis usually done on a voluntary basis, with thenotion that major stakeholders associatedwith each issue elect to join the groups work-ing with the problems that are most relevant.

The action planning groups are thentasked to: (1) clearly define the problem;(2) identify who owns the problem; (3) pro-pose a solution for the problem; and (4)propose a management structure for leader-ship and follow-through. Since the problemsand interventions are ideally driven by the

125

Page 10: Understanding and Managing Workplace Aggression

data, they will vary among organizations.However, there are a number of interven-tions that have been discussed in the litera-ture that should be considered by allorganizations. These include policies onaggression, selection procedures, workrestructuring, employee assistance pro-grams, attributional counseling, and beha-vioral interventions. Since discussion ofthese interventions is available elsewhereand in our own work, we will provide abbre-viated descriptions of these strategies here.

Policies on aggression. As stated above, theHealth and Safety Work Act of 1974 man-dates that employers implement policies toensure their employees’ safety and welfare atwork. Thus, at the very least all organizationsshould develop a policy prohibiting aggres-sion, and that policy should be clearlycommunicated to all employees. Because ofpotential litigation, it is prudent to providecopies of the policy to all employees and toobtain their signatures indicating that theyhave read, understand, and intend to complywith the policy. As discussed above, zerotolerance policies appear to be quite effective.

Selection procedures. Commonly acceptedprocedures designed to mitigate aggressionare background checks, screening for drugand alcohol abuse, reviews of driving records,and inquiries about any gaps in employmentthat could indicate incarceration. It is alsoprudent to verify information regardingeducation, employment, credit, and militaryservice.

Organizations should also considerscreening candidates with respect to traitanger, negative affectivity, and attributionstyle. As described above, there are reason-ably reliable measures available. However,the choice to use these instruments presentsa significant dilemma. More specifically,since the validity of these tests can be ques-tioned, the use of such testing could be inter-preted as potentially violating both theAmericans with Disability Act and TitleVII of the Civil Rights Act. On the otherhand, as more evidence supporting the use

126 ORGANIZATIONAL DYNAMICS

of these variables is developed, the failure toconsider these types of variables could alsobe considered negligent under the Healthand Safety Act of 1974. At present our ownbias is to develop information about thesevariables in the employment interview pro-cess rather than using psychometric testing.As suggested above, we have found consid-erable agreement on the assessments of attri-butions obtained during interviews andthose that were the result of psychometricassessments.

Communication. Communication is a criticalelement in any intervention designed toreduce aggressive behavior. In particular,communication regarding trigger eventssuch as layoffs is critical. For example, if aperson is laid off, it makes sense that theprobability of aggression will be lessened ifthe individual receives an explanation of theprocess and the considerations (e.g., pastperformance as well as economic andbusiness conditions) that were made in thedecision. One of the worst examples of poorcommunication was experienced by one ofthe authors, who worked in a manufacturingplant when the labor force was reduced from6,000 to 4,000 employees within a six-monthperiod. Rather than explaining what wasgoing on, management decreased comm-unication and took the draconian measureof waiting until several Friday the 13ths toannounce the layoffs. There were numerousacts of vandalism as well as at least onephysical altercation that could potentiallyhave been avoided if management hadtaken the time and effort to explain therationale for their actions.

The research on organizational justicedescribed earlier suggests guidelines forcommunicating management decisions. Spe-cifically, our knowledge of the three forms ofjustice suggests that people are less likely tofeel victimized if managers communicatedecisions regarding negative outcomes (i.e.,potential trigger events) clearly and withempathy (interactional justice) and includean explanation of how the decision was made(procedural justice) as well as why the out-

Page 11: Understanding and Managing Workplace Aggression

comes are fair in terms of employees’ inputs(distributive justice).

Work design. There has been a considerableamount of research on work design, butreducing aggression has not been the focus.However, since attributions associated withaggression are often associated with external,stable, and controllable attributions, workdesign that allows more self-control andautonomy should be considered as poten-tially effective in reducing hostile attribu-tions and their consequences. In addition,organizations should consider alleviatingconditions surrounding the job that suggestthat the workers are powerless—such asinadequate parking, poor restroom facilities,filthy working conditions, poor ventilationand temperature control, and poor lighting.

Organizational culture. Another strategy forreducing aggression is to facilitate a culturethat discourages aggression. Strategies couldinclude but are not limited to vision and goalstatements stressing respect and cooperation,reward procedures that punish aggression,and the physical presence of security cameras,alarms, and security guards. Supervisorypractices and policies should also beconsidered to make sure that there is not aculture of abusive supervision or incivility.

Employee assistance programs. All organi-zations should have an EmployeeAssistance Program (EAP) designed to helpemployees address personal problems thatmay be affecting their job performance suchas alcohol and drug abuse, marital problems,and financial problems. The program can takethe form of having an onsite coordinator.However, even very small organizationscan implement EAPs by helping managersidentify potential problems and encouragingemployees to consult professionals whoseservices are covered in their organization’shealth insurance program. EAPs can beimportant during layoffs and other impor-tant trigger events so employees do notreact with aggression directed towardsthemselves or others.

Attribution counseling. The goal of attributioncounseling is to help the employeesrealistically analyze the causes of triggerevents so they can develop constructiveattributions. It appears that this type ofcounseling might have been helpful in theByran Uyesugi case mentioned above, in thatit may have reduced his apparent tendencytoward hostile attributions and the asso-ciated aggressive behaviors.

Although we are not encouraging man-agers to become amateur counselors, theycan intervene by providing clear communi-cation about the causes of negative outcomes.Similarly, when managers notice a pattern ofhostile attributions, they can intervene byproviding the information needed and byaddressing problems when they are legiti-mate.

Behavioral interventions. Organizations alsoneed to have in place contingency plans inthe event that aggressive acts occur. It isbeyond the scope of this article to provideadvice on law enforcement procedures thatmay have legal implications. However,common sense suggests that proceduresshould be developed for identifying when alegitimate threat occurs and who will beresponsible for addressing threats. Werecommend that organizations consult withtheir own security people, local law enfor-cement, and the guidelines that have beensuggested by the Occupational Safety andHealth Administration (OSHA). The com-prehensive set of guidelines that OSHA hasdeveloped for addressing workplaceaggression is very helpful. In addition,OSHA also provides guidance regardingrecord keeping, developing incidencereports, and conducting surveys to assesspotential threats.

In concluding our discussion of the action-planning phase, we emphasize that the pro-cess of developing the action plans is data-driven, so that it responds to the needs of theorganization. We also underscore that parti-cipative processes with broad representationfrom all parts of the organization are advan-tageous in understanding the nature of

127

Page 12: Understanding and Managing Workplace Aggression

existing problems, developing the actionplans, and in ultimately implementing theplans.

Management approval. In this step, theaction plans are presented to senior manage-ment. Ideally, senior management attendsthe last session of the planning phase wherethe task forces propose their plans. At thispoint, senior management can adopt theproposals or modify the plans. In most casesit is advisable to provide approval in princi-ple and then for senior management to makethe final decisions considering resources,legal considerations, and the viability ofthe proposed programs.

Implementation. It is highly recommendedthat implementation of the plans is a lineresponsibility. Provided that line managershave had active opportunities for inputthroughout the process, commitment to theprograms should be relatively easy toachieve.

Feedback and follow-up. Someone in theorganization, frequently the person who isresponsible for managing compliance withOSHA, should assume the responsibility formaintaining incidence records and serve asthe focal point for identifying whether theorganization is achieving the goals of theaction plans. At the very least, a report tosenior management every six months shouldbe required. If problems remain, task forcesmay be reinstated or formed to analyze theproblems and develop solutions.

CONCLUSION

Our purpose in writing this article was toprovide a solid theoretical and empiricalfoundation for developing strategies to miti-

128 ORGANIZATIONAL DYNAMICS

gate organizational aggression. As describedin our model, organizational aggressionoften emerges from a combination of bothhuman and organizational factors. Althoughwe cannot predict aggressive behavior withcertainty, we can reduce the probability ofaggressive acts by actively managing envir-onmental and human factors. Because everyorganization is different, we emphasize theneed to do a thorough audit before develop-ing programs to mitigate aggression. Oncethe status of factors within a particular orga-nization is known, organizations can developaction plans to mitigate aggression. Theseaction plans typically address organizationalfactors by shaping the culture within theorganization, structuring jobs to reduceanger leading to aggression, and reducingthe incidence of abusive supervision andoverly rigid policies. The human side ofthe equation can be managed through selec-tion processes, policies on aggression, clearcommunications regarding trigger events,EAPs, attributional counseling, and carefulintervention when aggressive incidents dooccur.

In conclusion, although we know that itis almost impossible to predict specific acts ofaggression, we have accumulated knowl-edge and experience that allow us to reducethe probabilities of aggressive incidents.Given this knowledge, failure to managefactors that are known to be associated withaggression can be considered negligent andorganizations may be found to be criminallyliable when acts of violence are perpetrated.It is our hope that this article will be a usefulstarting point for organizations that want totake proactive steps to reduce the probabilityof aggression in their workplaces.

Page 13: Understanding and Managing Workplace Aggression

SELECTED BIBLIOGRAPHY

Descriptions of organizational aggressioncome from CNN and can be found athttp://www.cnn.com. They include: ‘‘Hotelworker kills 5 in shooting spree’’ (30 Decem-ber 1990); ‘‘ Shooting suspect pleads notguilty to office killings’’ (27 December2000); ‘‘Seven die in Chicago warehouseshooting’’ (28 August 2003); ‘‘Medical Jour-nal Apologizes to Prozac Maker’’ (28 January2005); ‘‘Plant shooter in anger managementclasses’’ (10 July 2003); and ‘‘Suspect in Hon-olulu shooting spree faces first-degree mur-der charges’’ (3 November 1999).

The sociological approach to organiza-tional aggression is described by R. Folgerand D. P. Skarlicki, ‘‘A Popcorn Metaphorfor Employee Aggression,’’ In R. W. Griffin,A. O’Leary-Kelly, and J. M. Collins (Eds.)Dysfunctional Behavior in Organizations: Violentand Deviant Behavior (Stanford, CT: JAI Press,1998), pp. 43–82. Other work demonstratesthat human factors contribute to our abilityto understand aggression and includes M. J.Martinko, M. J. Gundlach, and S. C. Douglas,‘‘Toward an Integrative Theory of Counter-

productive Workplace Behavior: A CausalReasoning Perspective,’’ International Journalof Selection and Assessment, 2002, 10(1/2), 36–50; and S. C. Douglas and M. J. Martinko,‘‘Exploring the Role of Individual Differencesin the Prediction of Workplace Aggression,’’Journal of Applied Psychology, 2001, 86(4), 547–559.

An article by H. Bensimon, ‘‘Violence inthe Workplace,’’ in Training and Develop-ment Journal, 1994, 1, 27–32 describes howrigid bureaucracy precipitated many of theincidents of aggression in the U.S. PostOffice. A discussion of the effects of work-ing conditions is provided by R. A. Baron,‘‘The Physical Environment of Work Set-tings: Effects on Task Performance, Inter-personal Relations, and Job Satisfaction,’’ InB. M. Staw and L. L. Cummings (Eds.),Research in Organizational Behavior (Green-wich, CT: JAI Press, 1994), Vol. 16, pp. 1–46.Bandura’s work describes how culturesaffect aggression through modeling. SeeA. Bandura, Aggression: A Social LearningAnalysis (Englewood Cliffs, NJ: Prentice-Hall, 1973).

Mark J. Martinko (Ph.D., University of Nebraska) is the Bank of AmericaProfessor of Management at Florida State University. His researchfocuses on attribution theory and leadership and he has authored or co-authored seven books and more than 80 articles and book chapters. He isa past president and fellow of the Southern Management Association aswell as the president of Performance Associates, which specializes inexecutive/leadership coaching and development. His most recent booksare Thinking Like a Winner: A Guide to High Performance Leadership (2002)and Attribution Theory in the Organizational Sciences: Theoretical andEmpirical Contributions (2004). (Tel.: +1 850 893 2786; fax: +1 850 6447843; e-mail: [email protected]).

Scott Douglas (Ph.D., Florida State University) is an Associate Professorof Management at The University of Montana. He is the author andcoauthor of several articles and book chapters on counterproductive work

129

Page 14: Understanding and Managing Workplace Aggression

130 ORG

behaviors, and is currently active in executive development and consultingin the area of managing workplace aggression and violence (Tel.: +1 406243 5695; fax: +1 406 243 2086; email: [email protected]).

Paul Harvey is a Ph.D. student in organizational behavior at Florida StateUniversity. His research interests include attribution theory, emotions,perceptions, and ethics. He has an MBA from SUNY-Binghamton andreceived his Bachelor’s degree in finance from the University of Connecti-cut. He has industry experience working in manufacturing finance andpersonnel recruiting (Tel.: +1 850 644 2038; fax: +1 850 644 7843; e-mail:[email protected]).

ANIZATIONAL DYNAMICS