Policing: An International Journal of Police Strategies & Management

Burnout and violent victimization in police officers: a dual process model

Karoline Ellrich
Article information:
To cite this document:
Karoline Ellrich , (2016),"Burnout and violent victimization in police officers: a dual process model", Policing: An International
Journal of Police Strategies & Management, Vol. 39 Iss 4 pp. -
Permanent link to this document:
http://dx.doi.org/10.1108/PIJPSM-10-2015-0125
Downloaded on: 02 November 2016, At: 06:16 (PT)
References: this document contains references to 0 other documents.
To copy this document: [email protected]
The fulltext of this document has been downloaded 15 times since 2016*
Users who downloaded this article also downloaded:
(2016),"General strain and police misconduct: the role of organizational influence", Policing: An International Journal of
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

Police Strategies & Management, Vol. 39 Iss 4 pp. -

(2016),"Conflict and compatibility: perspectives of police officers with and without military service on the military model of
policing", Policing: An International Journal of Police Strategies & Management, Vol. 39 Iss 4 pp. -

Access to this document was granted through an Emerald subscription provided by emerald-srm:232579 []
For Authors
If you would like to write for this, or any other Emerald publication, then please use our Emerald for Authors service
information about how to choose which publication to write for and submission guidelines are available for all. Please
visit www.emeraldinsight.com/authors for more information.
About Emerald www.emeraldinsight.com
Emerald is a global publisher linking research and practice to the benefit of society. The company manages a portfolio of
more than 290 journals and over 2,350 books and book series volumes, as well as providing an extensive range of online
products and additional customer resources and services.
Emerald is both COUNTER 4 and TRANSFER compliant. The organization is a partner of the Committee on Publication
Ethics (COPE) and also works with Portico and the LOCKSS initiative for digital archive preservation.

*Related content and download information correct at time of download.

 Burnout and violent victimization in police officers: A dual process model

1. Introduction

Over the past four decades, burnout has been studied intensively—usually conceptualized as a
three-dimensional syndrome of emotional exhaustion, depersonalization, and diminished
personal accomplishment (Schaufeli et al., 2009). Overwhelming evidence has shown burnout
to have widespread detrimental effects on the individual, such as physical and psychological
complaints, substance abuse, and work–family conflicts, as well as on the organization in
terms of reduced job satisfaction, intentions to quit, impaired performance, and quality of
service (Alarcon, 2011; Schaufeli and Enzmann, 1998). However, consequences of burnout
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

such as aggressive or violent acts that could result from modified interactions between staff
and clients have largely been neglected. In reference to healthcare professions, a few scholars
have suggested that emotionally exhausted and cynical employees may be more vulnerable to
being threatened or even assaulted by their clients (e.g., Winstanley and Whittington, 2002, p.
311). A study of Bakker et al. (2000) yielded some evidence for this. Findings indicated that
general practitioners who developed emotional distance as a way of coping with their
exhaustion evoked demanding and threatening behavior in their patients. Such behavioral
effects of burnout may be particularly harmful for police staff who are confronted with
emotionally upset, hostile, and violence-prone civilians. Even minor changes in police
officers’ behavioral style associated with burnout, such as less attention, patience, or
emotional control (Schaufeli and Enzmann, 1998, p. 21), could easily increase the likelihood
of an escalation resulting in violence.

Yet, studies investigating the link between burnout and violent encounters are still rare. They
found emotional exhaustion and depersonalization, but not reduced personal accomplishment,
to be relevant predictors in this context (e.g., Burke and Mikkelsen, 2005). The current study
goes beyond prior research by exploring two different processes through which burnout may
be linked to victimization.

2. Theoretical background

Burnout as has been defined by Freudenberger (1974, p. 159) as “to fail, wear out, or become
exhausted by making excessive demands on energy, strength, or resources.” Most scholars
agree that burnout is a psychological response to prolonged work stress that results from a
mismatch between the person and the job (e.g., Schaufeli and Enzmann, 1998). According to
the most widely used conceptualization proposed by Maslach and colleagues (Maslach and
Jackson, 1981; Maslach et al., 2001) burnout consists of three dimensions: emotional
exhaustion, depersonalization, and reduced personal accomplishment. Emotional exhaustion
is characterized by a depletion or lack of energy. Workers typically report feelings of being
emotionally overextended, debilitated, worn out, or fatigued. They are no longer able to give
of themselves to others. Whereas emotional exhaustion can be understood as the individual
strain dimension of burnout, depersonalization represents the interpersonal dimension of this
 syndrome. Depersonalization refers to the development of cynical, negative, or callous
attitudes towards and feelings about recipients of one’s service. Such impersonal and
dehumanized perceptions of others are also displayed in workers’ behavior, such as treating
clients as objects rather than people or minimizing physical involvement (e.g., avoiding eye
contact; Maslach, 1976). Depersonalization is viewed as a dysfunctional attempt to cope with
exhaustion by distancing oneself emotionally and cognitively from one’s work. The third
component of burnout, reduced personal accomplishment, reflects the self-evaluation
dimension of burnout and is marked by negative feelings regarding one’s competence and
work achievements. Empirical evidence found the three burnout dimensions to have
somewhat distinct patterns of associations with other variables (Demerouti et al., 2001; Lee
and Ashforth, 1996): Emotional exhaustion, for example, was primarily related to job
stressors, whereas depersonalization showed stronger correlations with resources. Less
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

conclusive were the results in reference to reduced personal accomplishment.

Police officers belong to one of the first professional groups on which burnout research was
carried out (Maslach and Jackson, 1981). This may be explained by their high exposure to
multiple organizational and operational stressors (Brown and Campbell, 1990), making police
officers especially vulnerable to stress-related outcomes. Studies examining possible
consequences of burnout in police staff found negative consequences on individual well-
being, organization-related attitudes (e.g., job satisfaction), and in-role performance (Bakker
and Heuven, 2006; Burke and Mikkelsen, 2006). Effects of burnout that may influence police
officers’ actual behavior on duty and thus predict the outcome of a police–citizen encounter
have as of yet been largely neglected. Indeed, the few studies conducted on this topic found
emotional exhaustion and depersonalization to have an influence on escalating police–civilian
interactions (Euwema et al., 2004; Kop and Euwema, 2001). However, violent victimization
of police officers as a possible result of such conflicting encounters has not been considered
by these studies. With respect to healthcare workers, some scholars also suggested that
behavioral changes (e.g., lack of empathy, decrease in communication) due to emotional
exhaustion and depersonalization could make staff more vulnerable to client-related
aggression and assaults (e.g., Winstanley and Hales, 2015, p. 31; Winstanley and Whittington,
2002, p. 311). Gascon et al. (2013), for example, found empirical support for this by showing
strong positive associations between emotional exhaustion as well as depersonalization and
violence among healthcare workers. Furthermore, Bakker et al. (2000) reported that general
practitioners who developed emotional distance as a way of coping with their exhaustion
evoked demanding and threatening behavior in their patients.

Taken together, previous studies suggest that burnout—in terms of emotional exhaustion and
depersonalization—may be linked to violent victimization. The present study aims to
investigate whether such a relationship can also be found in police officers. Moreover, it goes
beyond earlier work by making assumptions regarding the mechanism underlying the
burnout–victimization link. Based on conceptual considerations of the burnout syndrome and
given empirical evidence, emotional exhaustion and depersonalization are proposed to be
associated with victimization via two different pathways.

2.1. Emotional exhaustion and victimization

 The first process through which burnout may increase police officers’ victimization risk can
be explained by a reduction in their physical and mental performance. There is a lot of
evidence showing people with higher burnout levels to work less efficiently at their jobs,
resulting in impaired in-role performance (e.g., Bakker and Heuven, 2006; Taris, 2006). The
strongest and most consistent positive relationship could be established for emotional
exhaustion. Since emotional exhaustion is characterized by a depletion of resources combined
with feelings of fatigue and overextension (Maslach and Jackson, 1981), the available mental
and physical capabilities may not be sufficient to perform one’s task adequately. A study
conducted by Veldhuizen et al. (2003), for example, found that emotionally exhausted
subjects had problems investing enough energy during the working day to maintain their
performance, which resulted in slower reactions and smaller numbers of correct responses on
a memory task. Compared to nonexhausted individuals they further complained more about
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

concentration problems, physical fatigue, and mental fatigue, and reported a higher need for
recovery. Additionally, burnout also defined by high scores of emotional exhaustion was
found to be associated with deficits in voluntary control over attention (van der Linden et al.,
2005), which is considered to be important for planning, problem solving, and inhibition of
inappropriate responses.

Sufficient mental and physical resources are essential for police officers’ performance,
especially in confrontations with difficult civilians. This is already reflected in the selection
process of new police recruits, who usually have to complete several tasks testing their mental
and physical abilities (Cox et al., 2014). When arriving at the scene, police officers usually do
not exactly know what to expect. Moreover, even situations that initially seem to be peaceful
can escalate, resulting in an attack on the police officer. In order to protect police officers
from being victimized, they are trained to exercise standard safety techniques and strategies.
In Germany, for example, an official policy of self-protection (cf. Internal Affairs Ministry of
Lower Saxony, 2002) exists, which informs officers how to behave during different types of
encounters with civilians. Reduced cognitive and physical energy due to high levels of
emotional exhaustion may hinder police officers from acting adequately (e.g., distancing from
subject), which in turn could fail to prevent a violent attack.

Hypothesis 1 posits that emotional exhaustion reduces police officers’ self-protecting

behavior, which heightens the risk of being victimized.

2.2. Depersonalization and victimization

In addition to the depletion of physical and mental power, burnout could also affect police
officers’ behavior in contacts with civilians by changing relevant work-related attitudes. Most
research examining the influence of burnout on attitudes has focused on organizationally
relevant aspects such as organizational commitment, job satisfaction, or intentions to quit
(e.g., Alacron, 2011). Studies on client-related perceptions suggested to be linked with
burnout—such as lessened cognitive empathy, stereotyping tendencies, and rigid and
schematic thinking (cf. Schaufeli and Enzmann, 1998, p. 21f.)—are rather rare. Indeed, the
depersonalization dimension in particular already refers to a negative shift at an interpersonal
 level. Changes in work-related attitudes that focus on recipients may trigger an inappropriate
behavior, which in turn fosters an escalation process. For police officers who are principally
allowed to use force, attitudes towards the use of violence against civilians might be of
particular importance here.

A few studies can be found that suggest a relationship between burnout and favorable
attitudes towards the use of violence in policing staff (e.g., Burke and Mikkelsen, 2005; Kop
and Euwema, 2001). Results supported such a link, showing a positive association between
burnout and this attitude. Lambert et al. (2010) additionally found that police officers with
higher burnout levels are also more likely to support punishment. The most consistent
predictor of positive attitudes towards violence or punishment was depersonalization (e.g.,
Burke and Mikkelsen, 2005; Lambert et al., 2010). Depersonalization is more than just
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

gaining emotional distance from one’s work by losing one’s concern towards civilians. It
describes a cynical and callous shift in police officers’ attitudes and feelings, making civilians
appear less human (e.g., Maslach, 1976). Perception of others as impersonal, worthless
objects combined with feelings of hostility and loss of empathy may lower one’s moral
threshold in dealing with aggression (see also Kop et al., 1999). In line with this, studies
found that cynical subjects, for example, are more likely to engage in unethical behaviors
(Andersson and Bateman, 1997).

According to social-cognitive models (e.g., General Aggression Model, Anderson and

Bushman, 2002) a favorable attitude towards violence constitutes a relevant personal factor
that determines “individuals’ preparedness to aggress” (p. 35). This general preparedness to
aggress (along with situational variables) influences an individual’s present internal state
(including cognition, affect, and arousal), which, in turn, contributes to one’s actual
aggressive behavior in a given situation. Police officers with a more positive attitude towards
violence will have a greater accessibility for aggressive thoughts, emotions (e.g., anger), and
psychological or physical arousal in an encounter with a civilian. As such, they should be
more likely to respond to citizens in an irritating, hostile, or even aggressive manner (e.g.,
Kop et al., 1999). Such negative behavioral reactions can evoke feelings of provocation on the
part of the civilian, which increase the likelihood of a violent attack against the officer (cf.
Anderson and Bushman, 2002).

The second hypothesis suggests that depersonalization leads to a more positive attitude
towards violence, which increases police officers’ risk of being victimized.

3. Method

3.1. Procedure and sample

At the end of 2012 a paper-and-pencil survey was addressed to all uniformed patrol police
officers from five out of six police agencies in Lower Saxony, Germany. High-ranking police
officers doing mainly desk duties were excluded from the research because they do not have
regular contact with civilians. The questionnaire was distributed to the respondents using the
hierarchical structure of the police agencies. Ranking police officers within each
 organizational unit (starting from higher-level to lower-level units) were responsible for
administering the questionnaires to their patrol officers. Each participant was given an
envelope containing the questionnaire, a cover letter that emphasized the purpose of the study
and guaranteed anonymity and voluntariness of participation, and a second envelope
addressed to the research institute. In order to remind all patrol police officers to fill out the
questionnaire a nonpersonalized e-mail was sent to everyone three weeks later. Moreover, the
e-mail also expressed gratitude to those who had already completed the survey. No incentive
was provided for participation.

The questionnaire consists of two different parts. The first part asked for reports on three
police encounters—one each in cases of domestic violence, nondomestic violence, and control
of a suspected person. Previous research suggested these situations to be especially dangerous
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

for police officers (e.g., Ellis et al., 1993; Manzoni and Eisner, 2006). For each situation, only
the most recent police encounter was to be reported, in which the officer reached the location
first together with one colleague —irrespective of whether it was a call for backup or not.
Similar measurements have been used in previous research (e.g., Garner and Maxwell, 2002).
In reference to each encounter, participants were asked detailed questions regarding the
civilian (e.g., alcohol consumption), the situation (e.g., visibility conditions), and also their
own behavior (e.g., self-protecting behavior). Finally, it was assessed whether the encounter
resulted in violent victimization of the officer. In a second part of the questionnaire more
general characteristics describing the police officers and their profession, such as work-related
attitudes and burnout, were assessed.

A total of 1,931 patrol police officers participated in the survey, resulting in a response rate of
40.1% (base population: 4,814). The proportion of female police officers was 26.6%. The age
of the survey sample ranged from 21 to 62 years, with an average of 39.6 years. A comparison
with the total uniform patrol officer population in the five investigated police agencies
suggested good overall representation with slight overrepresentation of females and younger
police officers (24% female officers; mean age of 41.3 years). If all 1,931 participating police
officer would have filled out the questionnaire completely, information on 5,793 police
encounters would be available. Indeed, only a smaller group of 1,742 police officers who
reported 4,524 police encounters remained for the following analyses. Since most of the
missing responses referred to one of the three police encounters, an imputation strategy did
not seem adequate at this point. It is evident that police officers who failed to report a
particular encounter might not have experienced one recently. Thus, assigning data to these
persons would be problematic. Including solely participants with complete survey responses
seems to be the preferred strategy here. All results are presented based on the reduced sample
size (female officers: 26.8%; mean age: 39.1 years).

3.2. Analytic strategy

As a result of the measurement method, variables for each person are available at two levels:
Self-protecting behavior and victimization are assessed per subject for three police
encounters, resulting in maximal three distinct values (within-level variables); emotional
 exhaustion, depersonalization, and attitude toward violence are measured at the level of the
officer (between-level variables).

Due to the given data structure, multilevel structural equational modeling (SEM) was
performed to test the burnout–victimization model using Mplus 7 (Muthén and Muthén,
2010). Generally, SEM (Geiser, 2013; Kline, 2011) is a statistical modeling technique that
encompasses and extends several multivariate procedures such as factor analyses or
regression analyses. It can be used to test complex path models consisting of several
dependent and independent variables. More importantly, SEM is not restricted to observed (or
manifest) variables, and thus it also allows the inclusion of latent variables. Latent variables
represent a hypothetical construct that comprises multiple observed indicators.
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

3.3. Measurement

With the exception of victimization, the wording, mean, and standard deviation of all items of
the variables analyzed here are listed in the Appendix.

Within-level variables

Violent victimization: To assess violent victimization all police officers were asked, in
reference to each police encounter, whether they had experienced “physical violence with low
injury risk (e.g., pushing, resisting, struggling)” or “physical violence with higher injury risk
without using a weapon/dangerous object (e.g., beating, kicking, strangling).” Both questions
could be confirmed or denied. For the purpose of the present analyses, officers affirming at
least one of the two questions were classified as violent victims. Overall, in 28.2% of all
police encounters the police officer had been violently victimized in at least one of the two
forms.

Self-protecting behavior: Based on Schmalzl’s (2008) operationalization of the official

guidelines of self-protection behavior edited by the Internal Affairs Ministries in Germany
(e.g., Internal Affairs Ministry of Lower Saxony, 2002), a five-item scale was developed. For
each of the three police encounters participants were directed to report to what extent it was
possible for them to engage in these five behaviors (see the Appendix). The four-point answer
scale ranged from 1 “not at all” to 4 “fully.” A Cronbach’s α of .79 was calculated for the
present sample, indicating an acceptable internal consistency of the scale (see Bland and
Altman, 1997).

Between-level variables

Emotional exhaustion: The level of emotional exhaustion was measured by seven items (see
the Appendix) using the subscale “emotional exhaustion” of a revised German version of the
Maslach Burnout Inventory (MBI; Büssing and Glaser, 1998; Büssing and Perrar, 1992).1
With regard to other occupational groups (e.g., nurses), previous studies also applying this
instrument confirmed its validity and reliability (e.g., Büssing and Glaser, 2000). Respondents
1
This already unofficial revised version of the German MBI was provided by Dr. Jürgen Glaser with research
cooperation.
 rated the frequency of the experience of each item on a six-point scale (ranging from 1
“never” to 6 “very often”); the internal consistency of the scale is α = .91.

Depersonalisation: Analogue to emotional exhaustion, the depersonalization dimension was

assessed by four items of the corresponding subscale of the revised German version of the
MBI. As the items refer to changes in one’s relation towards clients, the wording was changed
to a police setting (see the Appendix). Again, the answer scale ranged from 1 “never” to 6
“very often.” Since Cronbach’s α is directly affected by the number of items (Nunnally,
1978), an internal consistency of Cronbach’s α = .65 may still be considered as sufficient,
although a need for improvement can be stated.

Attitude towards violence: Police officers’ favorable attitudes towards violence in interactions
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

with civilians was assessed via Kop and Euwemas’ (2001) scale of “attitude towards the use
of force” (p. 638), which has been successfully used in earlier police studies (e.g., Burke and
Mikkelsen, 2005). It consists of six items, which have to be answered on a five-point Likert
scale from 1 “completely disagree” to 5 “completely agree.” For the purpose of the current
study all items were translated into German by the author. To test the quality of the
translation, the German version of the scale was retranslated into English by a native speaker.
The reliability analysis of the scale resulted in a Cronbach’s α = .72, which indicates
acceptable internal consistency.

4. Results

As the dependent variable was binary coded (violent victimization: yes vs. no), binary logistic
multilevel SEM was applied to test the proposed burnout–victimization model.2 All
independent variables were modeled as latent factors with their indicator variables; violent
victimization is represented by a manifest measure. Figure 1 contains the results of the SEM
analyses.

XXX Figure 1 about here XXX

4.1. Test of the theoretical burnout–victimization model

Overall, the model determined an intraclass correlation (ICC), which indicates the percentage
of variance in the dependent variable that can be explained with between-level variables
(police officers), of 0.292. Thus, a maximum of 29.2% of the variance of violent victimization
refers to police officers’ attributes. This result generally supports the idea that victim-related
attributes are relevant to violent victimization, even though the most variance in violent
victimization is explained by within-level variables (police encounters).

2
An intercorrelation matrix of all measures will not be presented here because first, the constructs considered
refer to two different levels. Whereas burnout and attitude towards violence refer to the level of the person with
one value per subject, self-protecting behavior and victimization refer to the level of the police encounter with
three distinct values per subject. Second, all but one variable were specified as latent constructs; correlations
between latent variables at the same level are depicted in the SEM analyses in Figure 1.
 As shown in Table 1 (left column), the chi-square estimate became statistically significant at
p < .001, indicating a poor overall fit of the model to the data. Consistent with other
researchers (e.g., Manzoni and Eisner, 2006) less attention will be paid to this exact fit index.
Because of its sensitivity to sample size, even slight discrepancies between observed and
predicted covariance result in a significant chi-square score in large samples (Kline, 2011, p.
201). Rejection of the model should thus not be based solely on the result of the chi-square
test. Indeed, all other fit indices presented in Table 1 indicate an acceptable fit of the proposed
burnout–victimization model to the data. More in detail, the comparative-fit index (CFI) and
the Tucker-Lewis index (TLI), constituting incremental fit indices, show values greater than
.90, which is considered acceptable (e.g., Hox and Bechger, 1998) but somewhat lower than
what Hu and Bentler (1999) proposed to be a good model fit (e.g., CFI ≥. 95; see also Geiser,
2013). With reference to SRMRwithin = .056, SRMRbetween= .046 and RMSEA = .024 (SRMR
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

= standardized root mean square residual; RMSEA = root mean square error of
approximation), constituting absolute fit measures, the proposed model shows a fairly good
model fit (Hu and Bentler, 1999).

XXX Table 1 about here XXX

Further results of SEM analyses are visually depicted in Figure 1 showing standardized paths
and factor loadings. Consistent with general notation rules of path diagrams (see Geiser, 2013;
Hox and Bechger, 1998), rectangles represent observed (measured) variables, while latent
(unmeasured) factors are depicted by ellipses. To indicate direct paths in the model, single-
headed arrows are used; double-headed arrows indicate correlations.

In line with the reliability tests conducted before, the measurement models underlying the
latent constructs demonstrated a good fit of the data. All factor loadings for the observed
indicators are above .50 (see Kline, 2011), except for DP1 = .45 and AV4 = .36. The
relatively low factor loading of .36 may be due to the fact that this is the only recoded item of
the scale. However, this indicator was not deleted as it is part of a standardized instrument. It
should further be noted that in reference to the emotional exhaustion construct, a positive
correlation between the error variances of EE3 and EE4 (r = .47) was found, indicating that
both items additionally load on a second factor. This may be explained by the rather similar
wording of both items (see the Appendix) in combination with the fact that EE4 was directly
placed after EE3 in the questionnaire. Thus, the correlation of the error variance was allowed
in the SEM analyses. Although not explicitly formulated but in line with the burnout
conception (Maslach and Jackson, 1981), emotional exhaustion and depersonalization were
positively correlated with φ = .45 (p < .001).

4.2. Hypotheses

All proposed paths in the model (Figure 1) were significant and in the expected direction.
More in detail, police officers reporting higher levels of emotional exhaustion were less able
to enact self-protecting behavior in police encounters (ß = -.40, p < .001). The coefficient of
the path from self-protecting behavior to violent victimization was negative (ß = -.58, p <
.001), indicating that poor performance increases the risk of being violently attacked during
 police–citizen interaction. Thus, Hypothesis 1 suggesting a relation between emotional
exhaustion and victimization via reduced self-protecting behavior can be confirmed.

Consistent with the first part of Hypothesis 2, participants who are more depersonalized also
hold a more favorable attitude towards the use of violence in interactions with civilians, which
is represented in the positive path coefficient of ß = .48 (p < .001). A positive attitude towards
violence, in turn, is significantly positively related with violent victimization. With a ß = .15
(p < .01), it is considerably lower compared to all other standardized coefficients.
Nevertheless, results also support the proposed link between depersonalization and
victimization in line with Hypothesis 2.
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

4.3. Test of the final burnout–victimization model

To test whether the inclusion of additional regression paths between the model variables
results in a better fit of the theoretical model, five alternative SEMs were conducted. As can
be seen in Table 2, each of the five performed logistic multilevel SEMs contains only one
additional path. Table 2 also presents the chi-square value of the alternative models as well as
the changes in chi-square compared to the theoretical model. With reference to the chi-square
distribution, a positive difference of χ² > 3.841 indicates a significant improvement of the
model fit at p < .05, χ² > 6.635 at p < .01, and of χ² > 10.828 at p < .001. Inspection of the
output revealed that only the inclusion of the direct path between emotional exhaustion and
violent victimization resulted in a significant positive change of the chi-square value (14.855,
p < .001).

XXX Table 2 about here XXX

The fit statistics of this final model, which includes the additional pathway, can also be found
in Table 1 (second column). All indices remain identical or show only marginal changes, with
the exception of the significantly lower chi-square score. The paths and the respective
standardized regression coefficients of the final model are also depicted in Figure 1. It should
be noted that all factor loadings of the latent variables are identical to those of the theoretical
model. The coefficient of the path of emotional exhaustion on violent victimization, which
was added, was positive and significant with ß = .16 (p < .01). Emotional exhaustion thus
increases police officers’ risk of being attacked not only by a reduction in self-protecting
behavior as suggested in Hypothesis 1—it also affects victimization directly. Additionally, a
slight reduction in the magnitudes of the standardized coefficients can be stated for the
relationship between attitude toward violence and victimization (ß = .15 to ß = .10) and to a
lesser extent for the link between emotional exhaustion and self-protecting behavior (ß = -.42
to ß = -.40). None of the remaining coefficients differ from the theoretical model.
 5. Discussion

The current study aimed to provide more insight into the link between burnout and
victimization—a topic still suffering scarce attention in research.

Analyses first provided empirical evidence for the general notion that police officers are not
merely passive recipients of violent circumstances; police–citizen encounters leading to
violence are rather the result of an interactive process in which both parties interpret the
situation and decide how to react to each other (Alpert et al., 2004). Police officers with
elevated levels of burnout were found to have a higher risk of being assaulted during their
most recent encounters compared to their less burned-out counterparts. This finding may also
be relevant for other professionals with similar job profiles such as correctional officers or
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

security guards.

Overall, data supported the proposed burnout–victimization model, suggesting that emotional
exhaustion and depersonalization are linked with violent victimization via different pathways:
In line with Hypothesis 1, emotionally exhausted police officers reported decreased enactment
of self-protecting behavior during an encounter, which in turn heightened their risk of being
attacked.

As expected in Hypothesis 2, depersonalization was positively linked with violent

victimization by enhancing police officers’ favorable attitude toward violence.

Yet, less evident is the association between officers’ attitude toward violence and violent
victimization. Although the path coefficient was significant, it was quite weak. Based on
social-cognitive models of aggression (e.g., Anderson and Bushman, 2002), police officers
holding more favorable attitudes towards violence should be more likely to engage in violent
behavior, which in turn increases their risk of being attacked. Since police officers’ actual use
of force is suggested to be a mediating factor here, the weak association might be explained
by the fact that the link between police officers’ attitude towards violence and violent
behavior is indeed rather small. Aggressive behavior is considered to be the result of the
convergence of multiple personal and situational factors (id.). Police officers’ attitude towards
violence is only one indicator of their preparedness to aggress. Moreover, situational factors
and attributes of the civilian (e.g., impaired by alcohol) seem to be more important predictors
of police officers’ violent behavior (e.g., Garner and Maxwell, 2002). Furthermore, more
general attitudes toward the use of force against civilians may have less impact on police
officers’ behavior during a single encounter, because they had to learn how to control
themselves in such situations. This may be especially true in Germany, where police officers
are constrained by their organization as well as society to use only minimal force against
civilians. Burke and Mikkelsen (2005), for example, pointed out that stronger correlations
between burnout and use of force have to be expected in countries with higher rates of police
violence.

Based on additional analyses, one direct path from emotional exhaustion to violent
victimization was integrated into the final model. Thus, police officers with higher levels of
emotional exhaustion were more likely to report an attack during their last police encounter,
which cannot be explained by reduced self-protecting behavior. Police officers suffering from
 lack of energy are perhaps less motivated to engage in socially competent behavior, such as
perspective taking, communication, affect regulation, or empathy (cf. Rose-Krasnor, 1997).
A study by Kop and Euwema (2001), for example, showed exhausted police officers using
more verbal force in interactions with civilians, which in turn can provoke an attack. Thus,
integrating measures of police officers’ own aggressive and violent behavior in future studies
could enhance our understanding of police–citizen interactions that result in violent
victimization.

As the results show, high levels of emotional exhaustion should warrant more attention in
police organizations. For example, organizations should concentrate on individual-oriented
cognitive-behavioral stress management education and multimodal interventions aimed to
enhance persons’ coping skills (see Schaufeli and Enzmann, 1998) as well as managerial
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

interventions improving communication structures, feelings of being rewarded for one’s

effort, social support, etc. Emotional exhaustion, in fact, correlates more strongly with
common job-related stressors resulting from the organization itself, such as work overload or
rule inflexibility than to client-related stressors (e.g., Gaines and Jermier, 1983; Schaufeli and
Enzmann, 1998).

5.1. Limitations

This study has some limitations that should finally be addressed. A first point concerns the
cross-sectional design of the study. Such data are commonly used to examine structural
models that indicate a temporal ordering. Nevertheless, data collected at one point in time are
not suitable to prove causal relationships. Indeed, work-related victimization (e.g., bullying)
and burnout have been suggested to be cyclical, with victimization experiences also
increasing vulnerability to burnout (Wistanley and Hales, 2015; Winstanley and Whittington,
2002). Yet, these studies are based on cross-sectional data, too. To provide more insight into
the dynamic of the victimization–burnout process, prospective studies using longitudinal data
are necessary. Second, since self-reports of the police officers were used exclusively, common
method biases may partly explain the correlations observed between the variables (for details
see Podsakoff et al., 2003). A third limitation that should be stressed is the lack of attention
paid to further variables such as aggressive and violent behavior of the police officer or the
influence of the second police officers who were present during the encounters. Moreover,
attitudes of the civilians should be integrated in order to rule out third variable problems. For
example, it can be suggested that a drunken civilian may account for the rather strong relation
between self-protecting behavior and violent victimization. Some of the problems mentioned
previously can be addressed in future studies by using multimethod designs, including self-
reports from different sources as well as observational data (e.g., Euwema et al., 2004).
Observational data can not only provide more insight into the dynamics of the interaction, but
also allow more objective measures. In Germany, use of force by police officers, for example,
is still is a very sensitive subject, making it extremely difficult to assess such information via
self-reports. Fourth, generalizability of the study findings have to be questioned in reference
to other police groups (e.g., Gaines and Jermier, 1983) as well as police officers in other
countries because of varying policing cultures (e.g., Crow et al., 2004).
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)
 References

Alarcon, G. M. (2011), “A meta-analysis of burnout with job demands, resources, and

attitudes”, Journal of Vocational Behavior, Vol. 79 No. 2, pp. 549-562.
Alpert, G. P., Dunham, R. G. and MacDonald, J. M. (2004), “Interactive police-citizen
encounters that result in force”, Police Quarterly, Vol. 7 No. 4, pp. 475-488.
Anderson, C. A. and Bushman, B. J. (2002), “Human Aggression”, Annual Review of
Psychology, Vol. 53 No. 1, pp. 237-51.
Andersson, L. M. and Bateman, T. S. (1997), “Cynicism in the workplace: some causes and
effects”, Journal of Organizational Behavior, Vol. 18 No. 5, pp. 449-469.
Bakker, A. B. and Heuven, E. (2006), “Emotional dissonance, burnout, and in-role
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

performance among nurses and police officers”, International Journal of Stress Management,
Vol. 13 No. 4, pp. 423-440.
Bakker, A. B., Schaufeli, W. B., Sixma, H. J., Bosveld, W. and van Dierendonck, D. (2000),
“Patient demands, lack of reciprocity, and burnout: A five-year longitudinal study among
general practitioners”, Journal of Organizational Behavior, Vol. 21 No. 4, pp. 425-441.
Bland, J. M. and Altman, D. G. (1997), “Statistics notes: Cronbach's alpha”, British Medical
Journal, Vol. 314, p. 572.
Burke, R. J. and Mikkelsen, A. (2005), “Burnout, job stress and attitudes towards the use of
force by Norwegian police officers”, Policing: An International Journal of Police Strategies
& Management, Vol. 28 No. 2, pp. 269-278.
Burke, R. J. and Mikkelsen, A. (2006), “Burnout among Norwegian police officers: Potential
antecedents and consequences”, International Journal of Stress Management, Vol. 13 No. 1,
pp. 64-83.
Büssing, A. and Glaser, J. (2000), “Four-stage process model of the core factors of burnout:
the role of work stressors and work-related resources”, Work & Stress: An International
Journal of Work, Health & Organisations, Vol. 14 No. 4, pp. 329-346.
Büssing, A. and Glaser, J. (1998), Managerial Stress und Burnout. A Collaborative
International Study (CISMS). Technische Universität, Lehrstuhl für Psychologie, München.
Büssing, A. and Perrar, K. M. (1992), “Die Messung von Burnout. Untersuchung einer
deutschen Fassung des Maslach Burnout Inventory (MBI-D)“ [Measuring burnout: A study of
a German version of the Maslach Burnout Inventory], Diagnostica,Vol. 38 No. 4, pp. 328-
353.
Brown, J. M. and Campbell, E. A. (1990), “Sources of occupational stress in the police”,
Work & Stress, Vol. 4 No. 4, pp. 305-318.
Cox, S. M., McCamey, W. P. and Scaramella, G. L. (Eds.) (2014), Introduction to Policing,
Sage, Thousand Oaks, CA.
Crow, M.S., Shelley, T. O., Bedard, L. E. and Gertz, M. (2004), “Czech police officers. An
exploratory study of police attitudes in an emerging democracy”, Policing: An International
Journal of Police Strategies & Management, Vol. 27 No. 4, pp. 592–614.
 Demerouti, E., Bakker, A. B., Nachreiner, F. and Schaufeli, W. B. (2001), “The job demands-
resources model of burnout”, Journal of Applied Psychology, Vol. 86 No. 3, pp. 499-512
Ellis, D., Chaoi, A. and Blaus, C. (1993), “Injuries to police officers attending domestic
disturbances: An empirical study”, Canadian Journal of Criminology, Vol. 35 No. 2, pp. 149-
168.
Euwema, M. C., Kop, N. and Bakker, A. B. (2004), “The behaviour of police officers in
conflict situations: How burnout and reduced dominance contribute to better outcomes”, Work
& Stress, Vol. 18 No. 1, pp. 23-38.
Freudenberger, H. J. (1974), “Staff burn‐out”, Journal of Social Issues, Vol. 30 No.1, pp. 159-
165.
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

Gaines, J. and Jermier, J. M. (1983), “Emotional exhaustion in a high stress organization”

Academy of Management Journal, Vol. 26 No. 4, pp. 567-586.
Garner, J. H. and Maxwell, C. D. (2002), Understanding the use of force by and against the
police in six jurisdictions, final report, National Institute of Justice, Washington, DC,
available under: https://www.ncjrs.gov/App/AbstractDB/AbstractDBDetails.aspx?id=196694
[28.10.2015].
Gascon, S., Leiter, M. P., Andrés, E., Santed, M. A., Pereira, J. P., Cunha, M. J., Albesa, A.,
… and Martínez‐Jarreta, B. (2013), “The role of aggressions suffered by healthcare workers
as predictors of burnout”, Journal of Clinical Nursing, Vol. 22 No.21-22, pp. 3120-3129.
Geiser, C. (2013). Data Analysis with Mplus, The Guilford Press, New York.
Hox, J. J. and Bechger, T. M. (1998), “An introduction to structural equation modelling”,
Family Science Review, Vol. 11, pp. 354-373.
Hu, L. T. and Bentler, P. M. (1999), “Cutoff criteria for fit indexes in covariance structure
analysis: Conventional criteria versus new alternatives”, Structural equation modeling: A
multidisciplinary Journal, Vol. 6 No. 1, pp. 1-55.
Internal Affairs Ministry of Lower Saxony (Ed.) (2002). Leitfaden (LF) 371 zur
Eigensicherung im Polizeidienst [guidelines of self-protection in policing], Bartels & Wernitz,
Berlin.
Johnson, R. (2011), “Predicting officer physical assaults at domestic assault calls”, Journal of
Family Violence, Vol. 26 Vol. 3, pp. 163-169.
Kline, R. B. (2011), Principles and Practice of Structural Equation Modeling, The Guilford
Press, New York, NY.
Kop, N. and Euwema, M. C. (2001), “Occupational stress and the use of force by Dutch
police officers”, Criminal Justice and Behavior, Vol. 28 No. 5, pp. 631-652.
Kop, N., Euwema, M. and Schaufeli, W. (1999), “Burnout, job stress and violent behaviour
among Dutch police officers”, Work & Stress, Vol. 13 No. 4, pp. 326-340.
Lambert, E. G., Hogan, N. L., Altheimer, I., Jiang, S. and Stevenson, M. T. (2010), “The
relationship between burnout and support for punishment and treatment: A preliminary
 examination”, International Journal of Offender Therapy and Comparative Criminology, Vol.
54 No. 6, pp. 1004-1022.
Lee, R. T. and Ashforth, B. E. (1996), “A meta-analytic examination of the correlated of the
three dimensions of job burnout”, Journal of Applied Psychology, Vol. 81 No. 2, pp. 123-133.
Manzoni, P. and Eisner, M. (2006), “Violence between the police and the public. Influences
of work-related stress, job satisfaction, burnout, and situational factors”, Criminal Justice and
Behavior, Vol. 33 No. 5, pp. 613-645.
Maslach, C. (1976), “Burned-out”, Human Behavior, Vol.5 No. 9, 16-22.
Maslach, C. and Jackson, S. E. (1981), “The measurement of experienced burnout”, Journal
of Occupational Behavior, Vol. 2 No. 2, pp. 99-113.
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

Maslach, C., Schaufeli, W. B. and Leiter, M. P. (2001), “Job Burnout”, Annual Review of
Psychology, Vol. 52 No. 1, pp. 397-422.
Muthén, L. K. and Muthén, B. O. (2010), Mplus User’s Guide, Muthén & Muthén, Los
Angeles, C.A.
Podsakoff, P. M., MacKenzie, S. B., Lee, J. Y. and Podsakoff, N. P. (2003), “Common
method biases in behavioral research: A critical review of the literature and recommended
remedies”, Journal of Applied Psychology, Vol. 88 No. 5, pp. 879-903.
Rose-Krasnor, L. (1997), “The Nature of Social Competence: A Theoretical Review”, Social
Development, Vol. 6 No. 1, pp. 111-135.
Schaufeli, W. B. and Enzmann, D. (1998), The burnout companion to study and practice: A
critical analysis, Taylor & Francis, London, UK.
Schaufeli, W. B., Leiter, M. P. and Maslach, C. (2009), “Burnout: 35 years of research and
practice”, Career Development International, Vol. 14 No. 3, pp. 204-220.
Schmalzl, H-P. (2008), Einsatzkompetenz [Operation Competence], Verlag für
Polizeiwissenschaft, Frankfurt am Main.
Taris, T. W. (2006), “Is there a relationship between burnout and objective performance? A
critical review of 16 studies”, Work & Stress, Vol. 20 No. 4, pp. 316-334.
Van der Linden, D., Keijsers, G. P., Eling, P. and Schaijk, R. V. (2005), “Work stress and
attentional difficulties: An initial study on burnout and cognitive failures”, Work & Stress,
Vol. 19 No. 1, pp. 23-36.
Veldhuizen, I. J., Gaillard, A. W. and De Vries, J. (2003), “The influence of mental fatigue on
facial EMG activity during a simulated workday”, Biological Psychology, Vol. 63 No. 1, pp.
59-78.
Winstanley, S. and Hales, L. (2015), “A preliminary study of burnout in residential social
workers experiencing workplace aggression: Might it be cyclical?” British Journal of Social
Work, Vol. 45, pp. 24-33.
Winstanley, S. and Whittington, R. (2002), “Anxiety, burnout and coping styles in general
hospital staff exposed to workplace aggression: A cyclical model of burnout and vulnerability
to aggression”, Work & Stress, Vol. 16 No. 4, pp. 302-315.
 Table 1: Fit statistics of the Burnout–Victimization Models

theoretical model final model

ICC .292 .292
χ²M 834.114 819.259
dfM 234 233
p < .001 < .001
CFI .923 .925
TLI .911 .913
RMSEA .024 .024
SRMR .056 within/.046 between .056 within/.043 between
N police encounters 4.524 4.524
N persons 1.742 1.742
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

Note. ICC = Intraclass Correlation, CFI = Comparative-Fit Index, TLI = Tucker-Lewis Index, RMSEA = Root
Mean Square Error of Approximation, SRMR = Standardized Root Mean Square Residual.

Table 2: Chi-square values of alternative models (changes in chi-square shown are significant
at p < .001)

chi-square (df) change in chi-square

theoretical model 834.114 (234) -
+ DP
Vic 837.786 (233) -a
+ DP
SB 590.219 (233) -a
+ EE
AV 843.052 (233) -a
+ EE
Vic 819.259 (233) 14.855
+ AV
SB 923.952 (233) -a
Note. DP = depersonalization, EE = emotional exhaustion, Vic = victimization, SB = self-protecting behavior,
AV = attitude towards violence; a as chi-square indicates an impairment of the model fit, no difference score was
computed.
 Figure 1: Burnout–victimization model (shown: standardized coefficients, all paths are
significant at p < .01)

AV1 AV2 AV3 AV4 AV5 AV6

.58 .62 .56

.54 .36
.45 .63
DP1

DP2 .73
depersonalization attitude towards
DP3 violence
.51 .48
DP4
.63
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

EE1 .45 between level

.74
EE2
.15/.10a
.85
EE3
.47 .75 emotional
EE4
.82 exhaustion
EE5
.67
EE6 .77

EE7 .65
.16
-.42/-.40a

self-protecting violent victimization

behavior -.58

.54 .66 .71 .49 .64

within level
SB1 SB2 SB3 SB4 SB5

Note. Dotted arrow = additional pathway of the final model. a coefficients that changed by including the
“emotional exhaustion
violent victimization” path in the model; all other coefficients remain constant.
 Appendix: Wording and descriptive statistics of the items

Mean SD

Burnout: Emotional exhaustiona (n = 1,742) 3.13 1.28

EE1 I feel frustrated by my job. 2.98 1.38
EE2 I feel burned out from my job. 3.21 1.25
EE3 I feel worn-out at the end of a workday. 2.98 1.29
EE4 I am emotionally exhausted from my job. 3.27 1.36
EE5 I feel fatigued again when I get up in the morning and have to face 2.07 1.27
another day on the job.
EE6 I feel that I can’t go on any longer. 2.61 1.19
EE7 I feel that I'm working too hard on my job. 3.13 1.28
Downloaded by United Arab Emirates University At 06:16 02 November 2016 (PT)

Burnout: Depersonalizationa (n = 1,742)

DP1 I don’t really care what will happen to some civilians. 3.00 1.26
DP2 Since I took this job, I have become more callous towards civilians. 2.94 1.29
DP3 The way I treat civilians makes me feel uncomfortable. 2.06 0.89
DP4 I think I treat some civilians as if they were impersonal ‘objects’. 1.93 0.99

Attitude towards violenceb (n = 1,742)

AV1 When a suspected person behaves annoyingly, use of force is 2.40 1.14
adequate.
AV2 A police officer is regularly in situations where it is more adequate to 2.16 1.07
use physical force than to keep on talking to a person.
AV3 Some people can only be brought to reason the hard, physical way. 3.27 1.32
AV4 A tough, physical approach should be used less on the street. (-) 2.86 1.10
AV5 Sometimes forceful police actions are highly educational for 2.57 1.19
civilians.
AV6 Police officers should act more often using force. 1.62 0.89

Self-protecting behaviora (n = 4,524): To what extent was it possible to:

SB1 coordinate your actions with your colleague (e.g., via eye contact) 3.01 0.83
during the encounter?
SB2 provide backup during the encounter? 2.80 0.91
SB3 keep civilians at a sufficient distance during the encounter? 2.54 0.98
SB4 explain police procedures (e.g., use of force) before implementing 2.91 1.00
them?
SB5 direct the interaction with the civilian? 3.22 0.77

Note. a Items were translated by the author . b The English translation was obtained by Kop and Euwema (2001);
(-) reverse coded item