Does Deployment to War Affect Public Service Motivation? A Panel

Morten Brænder
Lotte Bøgh Andersen
Aarhus University, Denmark
Does Deployment to War Affect Public Service
Motivation? A Panel Study of Soldiers Before
and After Their Service in Afghanistan
Morten Brænder is assistant professor
in the Department of Political Science
and Government at Aarhus University in
Denmark. He holds a master’s degree in the
history of ideas and comparative religion
and a doctorate in political science. In
his dissertation, “Justifying the Ultimate
Sacrifice—Civil and Military Religion in
Frontline Blogs,” he analyzed how American
soldiers deployed to Iraq justified their participation in the war. In his present research,
he studies how exposure to extreme situations affects motivation.
E-mail: [email protected]
Lotte Bøgh Andersen is professor
in the Department of Political Science and
Government at Aarhus University and the
Danish Institute for Local and Regional
Government Research. Her research
interests include motivation, behavior and
performance of public employees, leadership and the use of economic incentives
in the public sector, and the relationship
between professionals and service users.
She has published numerous articles about
public service motivation, professionals, and
performance in the public sector.
E-mail: [email protected]
Public Administration Review,
Vol. 73, Iss. 3, pp. 466–477. © 2013 by
The American Society for Public Administration.
DOI: 10.1111/puar.12046.
466
Exposure to the extreme stress of warfare may affect soldiers’ that public service motivation is important in many
perceptions of others and society. Using panel data from
different public (and private) contexts (Andersen and
two companies on a tour of duty to
Serritzlew 2012; Brewer 2008;
Afghanistan in 2011, this article
Vandenabeele 2009). But does
Testing the effect of exposure to it “go to war”? We test extreme
analyzes how different dimencombat on public service motisions of soldiers’ public service
cases to see how far we can
motivation are influenced by
stretch our knowledge in regard
vation may provide important
deployment to war. As expected,
insights about what public serv- to a particular field of research.
soldiers’ compassion decreased and
ice motivation actually means to Therefore, testing the effect of
commitment to the public interexposure to combat on public
the individual.
est increased, while self-sacrifice
service motivation may provide
did not change systematically.
important insights about what
Deployment to war was expected to affect inexperienced
public service motivation actually means to the indisoldiers more than their experienced colleagues, but this
vidual. Given that many countries deploy soldiers to
hypothesis was only partially satisfied. The key contribution areas of operation where they are exposed to combat,
of the article is the use of panel data and the examination
the societal relevance of studying soldiers is very high,
of motivational changes. Moreover, studying soldiers’ public as former soldiers become part of many public and
service motivation enables us to connect public administra- private organizations in their civil lives.
tion and military sociology and thereby to establish a better
understanding of motivation in extreme settings.
Furthermore, insights drawn from this study may be
applied to the study of other service providers working
under extreme conditions, such as publicly employed
utting one’s life at stake in the service of others
police officers or privately employed military contracand society constitutes the ultimate challenge
tors. Of course, there are limits to such comparisons.
for establishing and maintaining public service
First, deployed soldiers may be exposed to more permotivation. The effect of combat on motivation in
manent stress than police officers. Second, compared to
general has been studied for generations (Little 1964;
private contractors, soldiers are seen as conducting their
Marshall 1947; Moskos 1971; Stouffer, Lumsdaine,
service within tighter rules of engagement. Yet, keeping
and Lumsdaine 1949; Wong et al. 2003), but we
these points of hesitation in mind, the points of resemknow little about the public service motivation (Perry
blance are still obvious. Like soldiers, both police officers
and Wise 1990) of soldiers. And we know little about
and private contractors are expected to put themselves
how public service motivation is affected by expoin harm’s way and to inflict harm for the benefit of othsure to combat. This article begins to close that gap,
ers and society (Møller 2010, 185; Singer 2004, 522).
using unique panel data with Danish combat soldiers
Furthermore, other studies of groups serving under
before and after their deployment to the war zone in
extreme conditions have shown that peer identification
Afghanistan.
as a way of coping with exposure to stress is relevant for
others than soldiers (Sykes 2007; Le Scanff and Taugis
The contribution of the study is twofold: By testing
2002; Tekleab, Quigley, and Tesluk 2009). Accordingly,
this in an extreme setting, we may gain knowledge
we expect some of the basic social mechanisms revealed
about public service motivation in general. Moreover,
in this study to be more broadly applicable.
by using measures of public service motivation in a
study of soldiers, our hope is to make conclusions
This leads to the second point: the interdisciplinary
drawn from the study of soldier motivation more
contribution. We hope to facilitate the generalizability
broadly applicable. As for the first point, we know
P
Public Administration Review • May | June 2013
of insights gained from the study of soldier motivation, often
conducted in military sociology. Within the last decade, military
sociology has shown that soldiers enlist and serve for a whole range
of different motives, including making a difference for society
(Eighmey 2006). Yet, even though such motivating factors may
seem similar to those conceptualized as public service motivation,
they have not yet been tested as such. Additionally, using measures
of public service motivation in the study of soldiers is becoming
even more relevant as warfare broadens its focus from killing people
to winning the hearts and minds of the civil population, that is,
serving people by participating in humanitarian aid operations
(Bensahel, Oliker, and Peterson 2009).
Finally, this study is a panel study. It enables us to measure the
development of motivation as a result of deployment. In that regard,
it clearly contributes to the study of public service motivation:
what happens under extreme conditions? Moreover, although the
time dimension was pivotal in initial large-scale studies of soldier
motivation (Stouffer, Lumsdaine, and Lumsdaine 1949; Stouffer
et al. 1949), very few recent contributions examining the effects of
deployment have tested the same groups of soldiers both before and
after their tour of duty (Hosek, Kavanaugh, and Miller 2006; Wong
et al. 2003). In that respect, the panel structure of this study should
be of interest for researchers in the study of soldier motivation in
general.
Grant 2010, 691–92). Recent studies suggest that it can actually be
changed, even by nondramatic events such as labor market entry
(Kjeldsen 2012; Kjeldsen and Jacobsen 2012). Thus, we expect a
drastic event such as deployment to war to change public service
motivation substantially.
We see the public service motivation concept as multidimensional
(Kim 2011; Kim et al. 2013; Perry 1996; Vandenabeele 2008), and
we use three of Perry’s dimensions. Attraction to policy making is
not included in this study, for several reasons. First, it is less relevant
in regard to deployment to war than the other dimensions because
deployment to war is expected to primarily affect the altruistic and
service-oriented elements of public service motivation, given that
these elements can be seen as based on normative and affective
motives (Perry and Wise 1990). In contrast, attraction to policy
making can be seen as based on rational or instrumental motives,
which are hardly in themselves changed in a critical situation.
Second, several studies argue that attraction to policy making is
conceptually underdeveloped and/or statistically not satisfactorily
corroborated (for an overview, see Ritz 2011, 1129).
Third, other studies of public service motivation in Denmark have
found it difficult to measure this dimension reliably (Andersen,
Pallesen, and Pedersen 2011). In general, public service motivation
in Europe can have different contents and understandings than in
the United States (Vandenabeele 2007; Vandenabeele, Scheepers, and
Hondeghem 2006). But the three investigated dimensions—comOur research question is as follows: How does deployment to war
mitment to the public interest, self-sacrifice, and compassion—are
affect soldiers’ public service motivation, and is that effect moderfound in other Danish studies (e.g., Andersen
ated by earlier war experiences? This question
et al. 2012b), and we have made sure that the
is investigated by using a pre- and a postdeOur research question is as
dimensions meaningfully capture the original
ployment survey of 211 Danish soldiers. Both
follows: How does deployment
understanding, especially for compassion,
surveys were answered by 78 soldiers. We first
to war affect soldiers’ public
which originally focused on sympathetic
analyze the balanced panel of 78 soldiers and
feeling to people in distress. We did that by
then test the robustness of the findings by
service motivation, and is that
including questions about different groups in
including all respondents. We use a Danish
effect moderated by earlier war
the survey questions about compassion. We
translation of public service motivation
experiences?
asked about caring for the well-being of others
questions developed by Perry (1996). After
in general, of nonsoldiers, and of foreigners
presenting our theoretical expectations, data
(people who are not Danes). Our results (reported in table 2) demand methods are described in detail. We then present the findings
onstrate a high level of consistency between the soldiers’ perceptions
before ending the article with a discussion and a conclusion.
of these different groups, indicating that compassion can be meaningfully said to cover sympathetic feeling to diverse types of people
Theory
in distress. Table 1 shows our understanding of the investigated
Our main expectation is that deployment to war affects the various
dimensions. They either can be seen as different facets of the public
public service motivation dimensions differently. In this section, we
service motivation construct (a formative measurement model),
first describe the public service motivation concept and the dimensions that we are testing; second, we discuss how deployment to war or they can be assumed to represent a single dimension describing the same underlying construct, where each measure is designed
is expected to affect these dimensions; and, finally, we propose two
to capture the construct in its entirety (a reflective measurement
hypotheses (each with testable implications).
model). Other studies (e.g., DeHart-Davis, Marlowe, and Pandey
2006; Vandenabeele 2009) have shown that the dimensions can have
Public service motivation has been defined in different ways
(Brewer and Selden 1998, 417; Perry and Wise 1990; Rainey and
Steinbauer 1999, 20; Vandenabeele 2007, 547). Here, we define it
Table 1 Investigated Public Service Motivation Dimensions
as “an individual’s orientation to delivering service to people with
Dimension
Understanding of Dimension
the purpose of doing good for others and society” (Hondeghem and
Commitment to the
Motivation to deliver public services to serve the relevant
Perry 2009, 6) because this definition emphasizes what we see as the
public interest
society, based on values and duty
central aspects of the concept: it is altruistic, it is individual, and it
Compassion
Emotionally (empathically) based motivation to do good
is linked to both service provision and doing good for society. Note,
for others in distress by improving public services
Self-sacrifice
The will to bypass one’s own needs to help others and
however, that the individual defines what “doing good” means. It
society by providing public services
can be discussed how stable public service motivation is (Wright and
Does Deployment to War Affect Public Service Motivation? A Panel Study of Soldiers Before and After Their Service in Afghanistan
467
different causes and consequences. Accordingly, we treat them separately in the analysis. Thus, we take into account the criticism raised
by Edwards (2011), for example, that when conceptually distinct
measures are channeled into a single construct, the resulting construct is conceptually ambiguous. We do not argue that a formative
model of public service motivation generally functions better than a
reflective one; we simply argue that establishing whether the effect of
deployment to war differs between the dimensions is interesting.
The expected variation between the dimensions is based on the
argumentation in Andersen et al. (2012b). They suggest that different public values (conceptions of the desirable for society) are differently related to commitment to the public interest, compassion,
and attraction to policy making and that public values can be seen
as explicating what “doing good” means. In contrast to the other
public service motivation dimensions, Andersen et al. (2012b) argue
that the self-sacrifice dimension can be seen as pure fuel behind
prosocial actions. This is in line with Kim and Vandenabeele (2010),
who argue that self-sacrifice can be seen as the basis on which the
other dimensions rest.
Three of the public value dimensions investigated in Andersen et al.
(2012b) are relevant for deployment to war: “the public at large,”
“professionalism,” and “user focus.” As discussed in more detail in
Andersen et al. (2012a), the first-mentioned dimension captures
values related to doing good for the whole society (such as accountability toward society in general and public insight and transparency). “Professionalism” captures values centered on compliance
with norms within an occupation, such as having independent
professional standards, while “user focus” concerns the incentive
of doing good for specific others (here, the direct users). Andersen
et al. (2012b) theoretically argue and empirically illustrate that commitment to the public interest is related to the “public at large” and
“professionalism” value dimensions, while compassion is expected
(and demonstrated) to be linked to “user focus.” Self-sacrifice is not
associated with any of the value dimensions.
This helps us formulate expectations as to the effect of deployment
to war on public service motivation dimensions. Given that compassion is both emotionally based (Kim and Vandenabeele 2010, 703)
and linked to sympathetic feeling to people in distress, it is expected
to be highly exposed to the stress of deployment, as dehumanization
can take place during deployment. Most combat soldiers serving
in Helmand Province, Afghanistan, experience actual combat. And
though it may not compare to battles in Hollywood blockbusters—in
this war, the enemy is rarely visible, and his most feared weapon is
the roadside bomb—the soldiers are all exposed to the extreme stress
posed by the combat situation, a stress that compares to no other situation. In order to transcend the societal and ethical norms that killing
is wrong (McMahan 2009), soldiers must establish a distance or an
“emotional withdrawal” (Grossman 1995, 160), enabling them to
perceive their enemies as inferior forms of life. This dehumanization is
a survival mechanism and may be a necessary aspect of war. As shown
by Guldbech (2009), Danish soldiers dehumanize their enemies. A
likely spillover from this emotional withdrawal is that the soldiers’
affective motivation to serve also decreases (Grossman 1995, 191).
We can distinguish among different groups benefiting from the
service delivered by the soldiers. First, as shown by several studies of
468
Public Administration Review • May | June 2013
military sociology, soldiers serve each other. Second, they serve civilians in the area of operation. And third, soldiers also serve for their
dear ones at home. What we define as the affective motivation to
serve others—that is, compassion—may regard both civilians on the
ground and those at home. Yet peer motivation is not compassion.
The relationship between the soldier and his buddies is a strictly
reciprocal. It is based on the mutual, in-group dependency experienced in extreme situations of combat (Shils and Janowitz 1948;
Stouffer, Lumsdaine, and Lumsdaine 1949, 135ff). Soldiers care for
each other for that reason. Compassion, however, indicates an asymmetrical, out-group relation. You only sympathize with people who
are worse off than yourself. And if that is the case, you can hardly
regard them as your peers.
For both theoretical and empirical reasons, we expect compassion
toward the civilian Afghan population to decrease as a result of deployment. Theoretically, we expect this to be the case because of the link
between compassion and values related to doing good for specific others (Andersen et al. 2012b) and to the sociopsychological mechanism
of dehumanization. This argument is supported by the soldiers’ own
statements. Thus, along with the pre- and postdeployment surveys
analyzed here, a number of qualitative interviews were conducted (22
before and 16 after deployment). Before their departure, the soldiers
stated that making a difference on the ground would be important,
although not crucial. After their return, most of them openly and
explicitly described their contempt for the Afghan civilians:
I think they are very, very bad persons, lots of them. They
sell out people for nothing, and they lie, and they steal, and
they.… Well, I just think they are disgusting. I simply don’t
get it. (Interview, private, Hussard guardsmen)
This lack of sympathy for the objects of one’s service is known in
military sociology (Moskos 1971, 231). However, we do not expect
this sympathy to decrease cumulatively in repeated deployments,
and accordingly, the effect of exposure to combat on compassion
is expected to be strongest for soldiers who are deployed to war for
the first time. Experienced soldiers will be more accustomed to the
situation (hardened by earlier deployment), and they may start the
investigated deployment with a lower level of compassion because of
the earlier deployment experience.
Another characteristic of deployment is that soldiers must be willing to sacrifice their lives for strangers, making it necessary for the
soldier to be part of a community in which those strangers are perceived as significant others. Thus, the military establishes a new primary group for the soldier to identify with (Ardant du Picq 1947,
110; Henriksen 2007; Newsome 2003; Shils and Janowitz 1948,
285 n. 7). Social psychology refers to this process as the deindividuation of group members (Barash and Webel 2009, 126). This means
that the values of professionalism become much more important
during deployment. Accordingly, we expect the public service motivation dimension linked to this value dimension—namely, commitment to the public interest—to increase during deployment.
The other value dimension linked to commitment to the public
interest, “the public at large,” may also be affected. There are arguments in favor of both decreased and increased importance of this
value dimension. On the one hand, a recent study of Danish combat
of the desirable, and therefore it is not as easily changed during
soldiers by Lyk-Jensen, Heidemann, and Glad (2012) shows that
critical situations, which alter peoples’ values. Our expectations are
abstract norms become less important during deployment. This sugexpressed in two hypotheses concerning the direct effect of deploygests that “the public at large” values decrease, leading to a decrease
ment and the moderation by earlier deployment, respectively.
in commitment to the public interest. On the other hand, there
are three reasons for expecting increased importance of “the public
Hypothesis 1: Deployment to war affects the public service
at large” values. First, individuals contributing to the realization of
motivation dimensions differently.
public values—such as democracy or justice—are likely to identify
with these values (Kim and Vandenabeele 2010, 703). Likewise,
H1a: Deployment to war affects compassion negatively.
Van der Dennen (2005, 82) argues that soldiers’ commitment to
the values of society can be a major motivation to engage in combat.
H1b: Deployment to war affects commitment to the pubThe qualitative interviews revealed that experienced soldiers reflected
lic interest positively.
very much on their role in society—not, however, as a kind of flagwaving patriotism. Soldiers take their duty as soldiers very seriously.
H1c: Deployment to war does not affect self-sacrifice.
They regard it as a professional calling: Serving society and being
deployed whenever duty calls is a part of a soldier’s identity. In that
Hypothesis 2: Deployment to war affects public service motiview, the professional norms attributed to the officer corps in Samuel
vation more for soldiers without earlier deployment experiHuntington’s path-breaking study of The Soldier and the State (1957)
ence compared to soldiers with earlier deployment experience.
seem to apply more broadly to the experienced professional soldiers
studied here—privates, noncommissioned
H2a: Deployment to war affects compasofficers, and officers alike. Second, beyond
We argue that soldiers are likely
sion less negatively for soldiers with earlier
these individual factors, we argue that soldiers
to mutually reinforce each
deployment experience before deployare likely to mutually reinforce each other’s
motivation because of the power of socializaother’s motivation because of the ment compared to soldiers without this
experience.
tion in military units, which is amplified durpower of socialization in miliing deployment. Thus, the abstract values that
tary units, which is amplified
H2b: Deployment to war affects committhey may have before deployment are likely to
during deployment.
ment to the public interest less positively
be reinforced during their tour of duty.
for soldiers with earlier deployment experience before deployment compared to soldiers without this
The last reason we expect an increase in commitment to the public
experience.
interest during deployment to war is connected to the other two,
but it is also far more cynical: if there is no good reason to serve,
In the unbalanced random-effect regressions, we control for age and
then it must be invented. We expect the circumstance that soldiers
gender. As shown by Pandey and Stazyk (2008), several studies have
find less meaning in compassion, the service for concrete others,
found age to be positively associated with public service motivation.
to leave a void of meaninglessness that must be filled. One way to
This might confound the expected associations between age and
fill out the void is to pay heed to the abstract motives for serving,
the dimensions analyzed later. Likewise, because women tend to see
associated with commitment to the public interest. This argument,
compassion as more important than men (Camilleri 2007; DeHartthat in lieu of compassion soldiers chose to identify with the more
Davis, Marlowe, and Pandey 2006), and given the fact that more
abstract organizational motives of commitment to the public interwomen participated in the postdeployment survey, this might also
est, is in accordance with the findings of a recently published study
confound the results.
of public service motivation among Swiss public servants. Thus,
Giauque et al. (2012) argue that individuals with a high degree of
compassion find it harder to comply with organizational values than Methods
individuals with a high degree of commitment to the public interest. This section first describes the investigated case (the deployment of
Danish combat troops to Afghanistan) and then discusses the key
operationalizations.
Later, we will address the argument about abstract norms in greater
depth. For now, we argue that the three reasons for increased commitment to the public interest during deployment to war carry most Formally, Denmark still has compulsory service. Any young man
of age must attend a conscript examination, and, after having his
weight, making us expect an increase in the public service motivaphysical and mental abilities tested, he will have to draw a number.
tion dimension. As in the case of compassion, we expect the effect
The number indicates whether he must serve. Those men who are
of deployment to war on commitment to the public interest to be
not obliged to serve as well as women who would like to serve can
strongest for the inexperienced soldiers.
volunteer (given that they are physically and mentally fit for the
task).
Following Kim and Vandenabeele (2010, 704), we see the last
dimension, self-sacrifice, as the basis on which the other dimensions
This last point is important because, in reality, nobody is forced to
rest, or as a more constant psychological propensity (Wright and
serve today. Any individuals who explicitly declare before the conGrant 2010, 694). This is consistent with Andersen et al. (2012b),
script examination board their unwillingness to be deployed are
who find that self-sacrifice is not related to any of the investigated
discarded. Serving means putting yourself in harm’s way. Yet the
value dimensions. The basic argument is that self-sacrifice expresses
Danish Defense has no shortage of volunteers. On the contrary,
pure energy, the basic willingness to do good in any understanding
Does Deployment to War Affect Public Service Motivation? A Panel Study of Soldiers Before and After Their Service in Afghanistan
469
more young men and women wish to serve than the Danish
Defense can actually manage to train. In some settings, economic
motivation may explain why large numbers of young people
volunteer for dangerous tasks such as deployment to war. Yet this
hardly applies to Denmark. Granted, soldiers receive a postdeployment bonus, and upon their return, they seem to regard economic
motives as more important than they did before deployment (LykJensen, Heidemann, and Glad 2012). This is also the case for the
soldiers studied in this article (results not shown). Furthermore,
the fact that a tour of duty may be the first real job for some of the
young soldiers could explain why serving seems more profitable
than studying, for instance. Yet serving in the Danish military is
not a goldmine: there is no tax exemption for deployed soldiers,
and, compared to the bonuses disbursed in the private sector, the
salary and the postdeployment bonus merely make ends meet.
Accordingly, even though our study shows that economic motivation increases during deployment, it also shows that both before
and after deployment, all other motivational factors are seen as
more important by the soldiers. Like other countries deploying
large contingents to the wars in Iraq and Afghanistan, the Danish
Defense has been forced to stretch its personnel resources—not,
however, because it lacks volunteers but because it lacks qualified, trained manpower. Danish soldiers may differ from soldiers
in other countries, but such differences do not derive from the
recruitment practice, as Denmark in reality has an all-volunteer
force.
The independent variable in this study is not deployment in general
but, specifically, deployment to war. Soldiers themselves distinguish
between deployment to war (what they call “green missions,” named
after the color of military camouflage uniforms) and other types
of deployment (“blue missions,” in which soldiers operate under
the blue United Nations flag). On blue missions, a soldier may
face combat, but chances are, this will not happen; UN soldiers are
rarely considered part of the conflict. Tours to Iraq or Afghanistan
are considered green missions. Here, soldiers are often part of the
conflict and face actual combat.
Danish forces have been in Afghanistan since 2002, and since 2006,
the army has deployed soldiers to Helmand Province, an area that
is particularly important to the Taliban. In Helmand, some of the
hardest fighting since 2001 has taken place. In order to break down
boundaries between different personnel categories, almost everybody
goes on patrol outside the fence. Not everybody experiences actual
fighting, but the risk is there. The units patrolling are engaged in
skirmishes and suffer casualties.
What we call “deployment to war” is what the soldiers call “green
missions”—missions involving combat situations—which we expect
will challenge the individual’s views on others and society. When
controlling for war experience, the question is whether the respondents have been in Iraq or Afghanistan before. Respondents who have
been deployed to missions to the Balkans, Cyprus, or Lebanon, for
instance, are considered first-timers. The soldiers in this panel all
participated in the International Security Assistance Force (ISAF)
11 mission to Afghanistan. They belong to two companies, one of
the Hussar Guardsmen (Garderhusarregimentet) and another of
the Jutland Dragoons (Jyske Dragonregiment). They were deployed
from February to August 2011.
470
Public Administration Review • May | June 2013
The questionnaires were handed out as group enquêtes before
their departure and after their return. Before deployment, the total
number of soldiers in the two companies (an armored infantry
company and a motorized infantry company) was 244. In the
first survey, the response rate was 70 percent, and in the second,
it was 52 percent. A total of 78 soldiers answered both before and
after their deployment to Afghanistan, 89 answered only before
deployment, and 44 answered only after deployment. In total, 211
soldiers participated in the study. Later, we report the results of
fixed- and random-effects analyses for the balanced panel with 78
soldiers. Because Hausman tests indicate that fixed- and randomeffects models do not give different results for these soldiers, we
also present random-effects models with the unbalanced panel
of 211 respondents (where fixed-effect analyses cannot be used).
Reaching the same results here would strengthen the robustness of
our findings and enable us to include time-invariant variables such
as gender.
We find no reason to expect a bias in who answered the questionnaire both times or only once. First, the respondents handing in
questionnaires were from the same two companies. Not only are
they members of the most tightly closed social unit, the military
outfit, but also they filled out the questionnaires with the same
background of experiences during training and deployment.
Second, the postdeployment survey was conducted immediately after deployment. Of the questionnaires in the survey after
deployment, 93 were handed in within two weeks after their
return, and the remaining surveys within six weeks. There are no
significant differences between these two groups in regard to the
results reported. When the second round of questionnaires was
conducted, one soldier had been killed in action, and only 22
out of the remaining 210 respondents had left the companies,
either because they were pursuing a civilian career or because they
had been transferred to another place of deployment within the
military. Active-duty soldiers not participating in the postdeployment survey were typically absent because they had other tasks to
attend to. Some were on the shooting range, some were on leave,
and others were handing in equipment. Thus, the dropout rates
were mainly caused by circumstances regarding the data collection,
as was the lower number of respondents to the postdeployment
questionnaire. Before deployment, the surveys were collected from
a whole company present at a time. After deployment, collecting
the questionnaires was up to the platoon and squad leaders, who
also had other things to attend to than improving the response
rate of this study.
The public service motivation dimensions are operationalized using
the Perry (1996) items translated to Danish. Table 2 presents a factor analysis of the applied public service motivation items rotated
with direct oblimin rotation to allow some covariation between
the public service motivation dimensions. The expected number
of factors (three) with eigenvalues of about 1 were extracted; the
loadings are high for the items expected to measure a given public
service motivation dimension and otherwise low. Most of the items
are standard public service motivation items, but we included three
compassion items concerning non-Danes, nonsoldiers, and people
with whom the respondent does not have a personal relationship
(table 2 contains the specific wording). All of the compassion items
follow the same pattern in the factor analysis (and show the same
Table 2. Principal axis factor analysis of items used to measure the public service motivation dimensions
Factor loadings
SS
Det er vigtigt for mig at bidrage til det fælles bedste
PSM30: Meaningful public service is very important to me
Det er meget vigtigt for mig, at de offentlige ydelser er i orden
It is very important to me that public services are adequate
Jeg ser helst, at offentligt ansatte gør det, der er bedst for samfundet som helhed, selvom det skulle gå ud over egne interesser
PSM34: I would prefer seeing public officials do what is best for the whole community, even if it harmed my interests.
Det er vigtigere at bidrage til samfundet som helhed end at hjælpe den enkelte
CPI4: Serving the public interest is more important than helping a single individual (Kim et al. 2011)
Jeg bliver følelsesmæssigt berørt, når jeg ser mennesker i nød
PSM4: It is difficult for me to contain my feelings when I see people in distress.
Jeg indlever mig i de vanskeligheder, andre står over for
COM3: I empathize with other people who face difficulties
Jeg tænker sjældent på andres velbefindende, hvis jeg ikke kender dem personligt
I seldom care about the well–being of others if I don’t know them personally (reversed in the index)
Jeg tænker sjældent på andres velbefindende, hvis ikke de er soldater
I seldom care about the well–being of others if they are not soldiers (reversed in the index)
Jeg tænker sjældent på andres velbefindende, hvis ikke de er danskere
I seldom care about the well–being of others if they are not Danes (reversed in the index)
Det er vigtigere for mig at gøre en forskel i forhold til samfundet end at opnå personlig vinding
PSM1: Making a difference in society means more to me than personal achievements.
Jeg er klar til at lide afsavn for samfundets skyld
PSM26: I am prepared to make sacrifices for the good of society.
Jeg sætter samfundsmæssige forpligtigelser over hensynet til mig selv
(PSM5): I put civic duty before self.
Jeg er villig til at risikere at skulle tilsidesætte mine personlige behov for samfundets skyld
SS4: I am willing to risk personal loss to help society
Man bør bidrage med mere til samfundet, end man modtager
PSM17: I feel people should give back to society more than they get from it
COM
CPI
– .051
– .152
.569
– .019
.035
.524
.059
.037
.520
.099
.151
.545
.174
– .567
– .002
.099
– .575
.141
– .041
.668
.012
.119
.666
.042
.087
.738
.075
.438
– .122
.278
.698
– .094
.124
.828
– .025
– .015
.764
– .085
.035
.577
.086
– .049
Note: There were three Eigen values above 1. Rotation is direct oblimin. Extraction: Principal axis. Cronbach’s alpha for items
shaded in the table: Commitment to the public interest (CPI): 0.62; Compassion (COM) 0.78 and Self-sacrifice (SS) 0.84.
differences between before and after deployment), indicating that
we are measuring a general compassion dimension rather than a
specific understanding of the concept related to the soldiers’ situation. Cronbach’s alpha is .78 for the items intended to measure
compassion, .62 for the items intended to measure commitment to
the public interest, and .84 for the items intended to measure selfsacrifice. Given that a threshold of .6 is acceptable for Cronbach’s
alpha if the number of items is low (Peterson 1994), all of these
tests indicate that the public service motivation dimensions are
both reliably and validly measured. Therefore, we proceed to form
three sum indices for each of the public service motivation dimensions scaled to go from 0 (minimum public service motivation) to
100 (maximum public service motivation). Exposure to combat, as
mentioned, is operationalized as deployment to war in Afghanistan,
simply by surveying the soldiers before and after the deployment.
Prior deployment experience, gender, and age are self-reported in
the survey before deployment. Age after deployment is calculated
as the age before deployment plus the time between measurements.
Means, standard deviations, and correlations are shown in tables A1
and A2 in the appendix.
Results
All of the analyses are regressions with the public service motivation
dimensions as dependent variables, deployment as the independent variable, and earlier deployment experience as the moderator.
Compassion, commitment to the public interest, and self-sacrifice
are analyzed in the mentioned sequence. Tables 3–5 present the
survey results from the balanced panel with the 78 soldiers who
replied twice. Among these 78 soldiers, 76 gave valid answers
concerning compassion both times they were surveyed, while the
corresponding numbers are 73 and 72 for commitment to the
public interest and self-sacrifice, respectively. Tables 6–8 show the
results for the unbalanced panel. Tables 3–5 show both fixed- and
random-effects regressions. In these analyses of the balanced panel,
the variable “prior deployment experience” is only included in the
random-effects models because the fixed effects capture this in the
fixed-effects regressions (i.e., it is time invariant). Analyzing the
unbalanced panel, we can only use random-effects regressions, as
fixed-effects regressions do not allow us to include respondents
who did not answer in both rounds. One benefit of random-effects
regressions is that we can include time-invariant variables such as
gender. Still, controls for specific characteristics of the individual
soldiers are more stringent in the fixed-effects analyses (models 3.1,
3.3, 4.1, 4.3, 5.1, and 5.3 in tables 3, 4, and 5).
Hypothesis 1a expected deployment to war to affect compassion
negatively, and table 3 and table 6 clearly support this expectation.
The results from the balanced panel (table 3) and the unbalanced
panel (table 6) show that the level of compassion decreases, on average, by 5 to 6 points on the scale between 0 and 100. We argue that
this is a substantially interesting effect because we are talking about
changes over time for the same people in a construct, which can be
seen as a stable trait or even a disposition that is difficult or slow to
change. The regressions of compassion also have a relatively higher
within R2 than the other public service motivation dimensions
(reflecting that more variance in the same soldier’s reported compassion is explained by the deployment than it is the case for commitment to the public interest and self-sacrifice). The effect is also
statistically significant on the .01 level. The result does not change
when we control for earlier deployment experience in model 3.3.
Does Deployment to War Affect Public Service Motivation? A Panel Study of Soldiers Before and After Their Service in Afghanistan
471
Table 3 Fixed-Effects (FE) and Random-Effects (RE) Regressions of Compassion
Regression coefficients (standard errors in parentheses), balanced panel
Table 6 Random-Effects Regressions of Compassion Regression coefficients
(standard errors in parentheses), unbalanced panel
Model 3.1 (FE) Model 3.2 (RE) Model 3.3 (FE) Model 3.4 (RE)
After deployment
After * prior
experience
Prior deployment
Experience
Constant
R2 Overall
R2 Between
R2 Within
sigma_e
sigma_u
rho
N
–5.724**
(1.93)
–5.724**
(1.93)
70.000***
(1.36)
.024
70.000***
(2.12)
.024
.105
11.898
16.466
.657
152
.105
11.898
14.154
.586
152
–5.100*
(2.39)
–1.823
(4.09)
70.000***
(1.37)
.020
.007
.107
11.962
16.508
.656
152
–5.100*
(2.39)
–1.823
(4.09)
3.800
(4.49)
68.700***
(2.62)
.030
.007
.107
11.962
14.188
.585
152
+p < .1; *p < .05; **p < .01; ***p < .001.
Table 4 Fixed-Effects (FE) and Random-Effects (RE) Regressions of Commitment
to the Public Interest Regression coefficients (standard errors in parentheses),
balanced panel
Model 4.1 (FE) Model 4.2 (RE) Model 4.3 (FE) Model 4.4 (RE)
After deployment
After * prior
experience
Prior deployment
Experience
Constant
R2 Overall
R2 Between
R2 Within
sigma_e
sigma_u
rho
N
4.110*
(1.78)
4.110*
(1.78)
77.825***
(1.26)
.021
77.825***
(1.64)
.021
.069
10.755
11.785
.546
146
.000
1.755
9.003
.412
146
5.452*
(2.22)
–3.769
(3.72)
77.825***
(1.26)
.029
.005
.082
1.753
11.753
.544
146
5.452*
(2.22)
–3.769
(3.72)
.092
(3.44)
77.793***
(2.05)
.029
.005
.082
1.753
9.070
.416
146
+p < .1; *p < .05; **p < .01; ***p < .001.
Table 5 Fixed-Effects (FE) and Random-Effects (RE) Regressions of Self-Sacrifice
Regression coefficients (standard errors in parentheses), balanced panel
Model 5.1 (FE) Model 5.2 (RE) Model 5.3 (FE) Model 5.4 (RE)
After deployment
After * prior
experience
Prior deployment
Experience
Constant
R2 Overall
R2 Between
R2 Within
sigma_e
sigma_u
rho
N
–1.667
(2.37)
58.056***
(1.68)
.002
.007
14.229
19.037
.642
144
–1.667
(2.37)
58.056***
(2.54)
.002
.000
.000
14.229
16.161
.563
144
–4.043
(2.92)
6.843
(4.95)
58.056***
(1.67)
.010
.004
.033
14.139
19.008
.644
144
–4.043
(2.92)
6.843
(4.95)
–1.004
(5.34)
58.404***
(3.15)
.010
.004
.033
14.139
16.318
.571
144
Gender
Age
Model 6.1
Model 6.2
Model 6.3
Model 6.4
–2.040
(8.36)
–1.343
(8.26)
–1.388
(8.32)
–1.385
(8.33)
.179
.251
.235
.242
(.31)
(.30)
–5.044**
(1.84)
(.38)
–5.038**
(1.84)
.283
(3.95)
61.631***
(7.70)
.004
.005
.105
11.898
15.676
.634
238
61.149***
(7.61)
.011
.001
.105
11.898
15.737
.636
238
61.467***
(8.84)
.011
.001
.105
11.898
15.808
.638
238
(.39)
–4.677*
(2.26)
.514
(4.04)
–1.058
(3.88)
61.205***
(8.90)
.011
.001
.107
11.962
15.824
.636
238
After deployment
Prior deployment
experience
After * prior experience
Constant
R2 Overall
R2 Between
R2 Within
sigma_e
sigma_u
rho
N
+p < 0.1; * p < .05; ** p < .01; *** p < .001.
Table 7 Random-Effects Regressions of Commitment to the Public Interest
Regression coefficients (standard errors in parentheses), unbalanced panel
Gender
Age
Model 7.1
Model 7.2
Model 7.3
Model 7.4
–.174
(5.75)
.365+
(.21)
–.644
(5.73)
.335
(.21)
3.128+
(1.63)
.263
(5.70)
.649*
(.27)
2.942+
(1.63)
–5.184+
(2.79)
70.711***
(5.37)
.011
.017
.069
1.755
9.524
.440
237
70.629***
(5.34)
.018
.012
.069
1.755
9.579
.442
237
64.647***
(6.21)
.037
.031
.069
1.755
9.455
.436
237
.274
(5.70)
.676*
(.27)
4.344*
(2.02)
–4.267
(2.90)
–4.020
(3.41)
63.666***
(6.26)
.041
.033
.081
1.753
9.492
.438
237
After deployment
Prior deployment
experience
After * prior
experience
Constant
R2 Overall
R2 Between
R2 Within
sigma_e
sigma_u
rho
N
+p < .1; *p < .05; **p < .01; ***p < .001.
is negative in all models in tables 3 and 6. If there is a difference,
the compassion of soldiers with earlier experience is reduced more
rather than less during deployment (this difference is not significant). Consequently, compassion decreases regardless of earlier
deployment. Importantly, the decrease in compassion is not only
directed toward Afghans and other foreigners—a comparison of the
soldiers’ answers before and after deployment for each compassion
item separately (see table 2 for wordings) demonstrates that compassion decreases systematically for all the operationalizations (analyses
not shown).
+p < .1; * p < .05; ** p < .01; *** p < .001.
Hypothesis 2a expected deployment to war to affect compassion
less negatively for soldiers who had earlier deployment experience
compared to soldiers without this experience. This is rejected, given
that the interaction between after deployment and prior experience
472
Public Administration Review • May | June 2013
Hypothesis 1b expected deployment to war to affect commitment to
the public interest positively, and this is supported in tables 4 and 7.
For the balanced panel, the effect is significant regardless of model
specification, but when models 7.2 and 7.3 (unbalanced panel) look
at the effect without including the moderation effect from earlier
deployment, the positive effect is only statistically significant at the
Table 8 Random-Effects Regressions of Self-Sacrifice Regression coefficients
(standard errors in parentheses), unbalanced panel
Gender
Age
Model 8.1
Model 8.2
Model 8.3
Model 8.4
–4.972
(8.33)
.759*
(.31)
–4.665
(8.35)
.782*
(.31)
–2.201
(2.20)
–4.059
(8.38)
.990*
(.39)
–2.328
(2.21)
–3.526
(4.04)
–4.055
(8.40)
.950*
(.39)
–4.286
(2.70)
–4.838
(4.18)
5.696
(4.60)
37.107***
(9.10)
.041
.045
.033
14.139
14.874
.525
231
After deployment
Prior deployment
experience
After * prior
experience
Constant
R2 Overall
R2 Between
R2 Within
sigma_e
sigma_u
rho
N
39.545***
(7.80)
.033
.041
.007
14.229
14.637
.514
231
39.582***
(7.81)
.035
.042
.007
14.229
14.705
.516
231
35.669***
(9.01)
.039
.047
.007
14.229
14.725
.517
231
+p < .1; * p < .05; ** p < .01; *** p < .001.
.1 level. Still, when model 7.4 introduces an interaction between
“after deployment” and earlier deployment experience, the effect of
deployment becomes significant for the reference category (soldiers
without earlier deployment). The interaction term has the expected
sign, and the coefficient has almost the same size as the coefficient
for “after deployment,” indicating that commitment to the public
interest does not change much for soldiers with earlier deployment experience. This is in accordance with hypothesis 2b, which
expected deployment to war to affect commitment to the public
interest less positively for soldiers who had earlier deployment experience before deployment compared to soldiers without this experience. The interaction is not statistically significant, probably because
of the unavoidable multicollinearity in models with cross-product
interaction terms. The increase in R2 between model 4.1 and model
4.3 and between model 7.3 and 7.4 also indicates that the interaction contributes to explaining commitment to the public interest.
Therefore, we conclude that hypothesis 1b is supported by the data,
while hypothesis 2b is partially supported. It should also be noted
that age has a positive effect on commitment to the public interest
when controlling for prior deployment experience (see table 7).
to war to affect public service motivation more for soldiers without
earlier deployment experience compared to soldiers with earlier
deployment experience, receives only partial support (from the test
of hypothesis 2b but not from the test of hypothesis 2a). It seems
to be the case for commitment to the public interest, but compassion does not decrease less for soldiers with earlier deployment
experience.
Discussion
In this section, we first discuss the theoretical implications of our
findings. Following deployment to war, compassion decreases
and commitment to the public interest increases. We evaluate the
strength and weaknesses of the study and discuss the need for future
research before the conclusion gives an overall answer to the research
question and addresses the practical implications of this study.
Deployment to war affects compassion negatively. Thus, our main
expectation concerning this affective dimension is confirmed.
However, the decrease in compassion does not happen exclusively
for first-timers. Being deployed seems to make both experienced and
inexperienced soldiers more ruthless. Thus demonstrated, the spillover effect of dehumanization seems to be more general than expected.
This does not mean that soldiers patrolling in Helmand Province
automatically lump civilian Afghans with the Taliban. Nevertheless,
engaging in warfare has a significant psychological impact on the
individual’s perceptions of others and society: seeing suffering does
not make it easier to identify with those who suffer. Actually, distancing oneself emotionally makes it easier to cope with the suffering.
We do not know whether the decrease in compassion is permanent,
but our analysis indicates that it is not, given that compassion does
not differ significantly between experienced soldiers and first-timers
(models 6.3. and 6.4). The finding that experienced soldiers were
just as affected as their inexperienced colleagues may be related
to circumstances on this particular tour of duty. It may have been
more difficult for soldiers to identify emotionally with the objects
of service; it may have been more difficult to see the specific result
of their service. Yet, given that controls for which company the
soldiers belonged to (not shown) did not alter any of these results,
these findings are not related to differences between the military
units. Therefore, the results suggest that compassion decreases for all
soldiers during deployment.
Hypothesis 1c expected that deployment to war does not affect selfsacrifice, and tables 5 and 8 indicate that this expectation is correct,
given that we find no statistically significant effects. Deployment
does not affect self-sacrifice with or without control for prior
deployment experience, and this applies to both the balanced and
the unbalanced panel. In line with other studies, table 8 indicates
that self-sacrifice significantly increases with increasing age.
Given our results concerning the implications of hypothesis 1 (hypothesis 1a, 1b, and
1c), hypothesis 1 is supported. Deployment
to war affects the public service motivation
dimensions differently: compassion decreases
and commitment to the public interest
increases during the first deployment. Selfsacrifice is not affected by deployment to war.
Hypothesis 2, which expected deployment
As for commitment to the public interest, our expectation that
deployment increases the individual’s resort to more abstract motives
is met, albeit only partially. We expected an increase in the level of
commitment to the public interest for several reasons. First, soldiers
internalize the values of the organization that they serve (related to
the public at large), values that are positively associated with commitment to the public interest. Second, soldiers mutually reinforce
the values of each other, and deployment
also increases the salience of professionalism
Deployment to war affects
values, which are positively related to committhe public service motivation
ment to the public interest. Third, you cannot
dimensions differently: compas- serve without a purpose. If dehumanization
sion decreases and commitment takes place and soldiers therefore cannot find
motivation in serving specific others, they are
to the public interest increases
expected to increase their motivation to serve
during the first deployment.
society. We cannot exclude the possibility
Does Deployment to War Affect Public Service Motivation? A Panel Study of Soldiers Before and After Their Service in Afghanistan
473
that disappointment with the civil population, the concrete objects
of service, entails a need to establish and collectively confirm more
abstract objects, such as those implied in the idea of making a difference for society as a whole. If there is no purpose for sacrificing your
life, you must make one up yourself.
As mentioned earlier, Lyk-Jensen, Heidemann, and Glad (2012), in
a recent study of soldier motivation, reach conclusions that may be
seen as conflicting with our findings. We find that the motivation
to serve concrete others, operationalized as compassion, decreases
as a result of deployment, whereas the more abstract motivation
to serve society as a whole, commitment to the public interest,
seems to increase. Lyk-Jensen, Heidemann, and Glad, on the other
hand, find that both the soldiers’ concrete motivation and their
abstract motivation decrease following deployment. Their results
are especially relevant in this regard, as their group of respondents
also included the soldiers from ISAF 11 studied in this article. Yet
what is mainly emphasized by this difference is that there is more
to soldier motivation than just public service motivation, and vice
versa. First, the surveys used in this article also included questions
regarding soldier motivation aligning with those traditionally in
focus in military sociology. When focusing on these questions
only, our results do not differ substantially from those reported
by Lyk-Jensen, Heidemann, and Glad. We also find that both the
(abstract) patriotic motivation and the (concrete) motivation to
help civilians in the area of operation decrease. Second, if what we
find cannot be seen as a difference between abstract and concrete
reasons, then what is it? In that regard, it is worth noticing that the
four items used in the index measuring commitment to the public
interest all mention either the “public” or “community.” They all
emphasize the importance of the collective whole without specifying the properties or boundaries of this collective whole. On the
contrary, the abstract motivating factors affected differently all
specify the characteristics of the social unit for which they serve as
“my country” or “Denmark.” Abstract as they may be, they are still
more specific than the commitment to the public interest items,
and, like the concrete questions, they seem to decrease.
Moreover, as mentioned concerning our expectations, there are a
number of substantial similarities between our findings in regard
to this extreme case of soldiers going to war and the observations
made in a recent large-scale study of the relationship between public
service motivation and work satisfaction among Swiss public servants. Similar to Giauque et al. (2012), we study the public service
motivation dimensions separately and find that these dimensions
change in opposite directions. Last but not least, we also argue that
this is closely associated with the encounter between individual and
organizational norms and expectations.
and the exposure to the organizational realities (i.e., deployment to
war) observed in our study. Deployment to war causes soldiers’ compassion to decrease and their commitment to the public interest to
increase. We argue that this is a result of the realities on the ground,
causing the concrete individual aspirations of doing good for others
to be replaced by an identification with more abstract values, which
can be fulfilled more easily in the specific organizational context,
leading to a lower degree of resignation.
The primary strength of this study is that it analyzes panel data.
Whereas studies using cross-sectional data can clarify what characterizes soldier motivation, this approach enables us to examine how
soldier motivation actually develops. Furthermore, it is unique in
a military sociological context because it analyzes soldier motivation by means of the public service motivation measures. As for the
still-increasing number of studies analyzing public service motivation within the public administration literature, this study has the
advantage of focusing on a group of public service providers that is
seldom taken into consideration—a group for whom self-sacrifice
does not mean giving up social or economic benefits but putting
life at stake, and hence a group whose will to serve constitutes an
extreme case that has the potential for improving our understanding
of what motivation actually is.
Of course, this study also has its weaknesses. First, we may be able
to say how the public service motivation dimensions were affected
by deployment to war for these particular soldiers, deployed
between February and August 2011, but we cannot be sure that we
would make the same observations had we studied another cohort.
Likewise, it is a rather small panel with 211 individual respondents,
with only 78 who answered the questionnaire both before and
after their deployment. The small N limits the number of control
variables that we can meaningfully include in the analyses. On the
other hand, most of the expected effects prove significant, even with
this limited number of observations. Moreover, we have conducted
a number of robustness analyses and tried to include exposure to
actual combat exposure and to excitement motivation in all regressions (not shown). Neither altered the results reported here.
As stated earlier, studying public service motivation in regard to
soldier motivation is important because it extends the study of this
type of motivation to a group of public employees placed in an
extreme situation, and by means of the public service motivation
measures, we can compare soldiers with other groups. Yet it still
remains to be analyzed how important public service motivation is
for the individual soldier compared to other types of motivation. The
R2 estimates in our models also indicate that soldiers’ public service
motivation is affected by other variables besides deployment to war.
Age, gender, experience, and deployment to war make a difference in
regard to soldiers’ public service motivation, but there is still much
Granted, whereas Giauque et al. regard job satisfaction as a produnexplained variation among soldiers. This suggests that the insights
uct of these public service motivation dimensions (2012, 179–80),
from military sociology are still important and can be fruitfully
we study commitment to the public interest and compassion as
combined with public service motivation. This
dependent variables, which is possible because
may enable us to better understand the general
we have a panel study. Nevertheless, the
Deployment to war causes
decrease in compassion, as well as the different
association between organizational fit and
soldiers’ compassion to decrease effects deployment to war seemed to have on
the specific public service motivation dimencommitment to the public interest and other
sions shown by Giauque et al. contribute to
and their commitment to the
measures of abstract motivation. Furthermore,
explaining the close relationship between the
public interest to increase.
such an interdisciplinary approach may
specific public service motivation dimensions
474
Public Administration Review • May | June 2013
provide us with valuable information about the key variable in
military sociology: peer motivation. Thus, the fact that compassion
decreases during deployment for all specifications of “others” (not
only nonsoldiers or foreigners) seems to indicate that peer solidarity during deployment does not lead to specific distancing from
other groups but nevertheless coincides with a general emotional
withdrawal.
In sum, three aspects need to be studied further. First, using panel
data has enabled us to analyze how the public service motivation dimensions studied here are affected by the hardships of war,
but a larger panel would have been preferable. Second, studying
the effect of several deployments remains a challenge. This was
especially clear in regard to the unexpected finding concerning the
effect of employment on compassion for all soldiers (and not only
inexperienced soldiers). In order to state whether this is attributable to task-specific conditions on this particular deployment
or whether it reflects a more general picture, we need to study
different deployments. Last but not least, in order to analyze the
mechanisms behind the effect of deployment on public service
motivation—that is, to what degree war affects the soldier cognitively and emotionally—we need in-depth qualitative studies of the
processes that soldiers go through. By means of our study, we can
identify the effects, while in-depth studies can clarify how and why
these effects take place.
expose their employees to, as this might change their public service
motivation.
The generalizability of the findings is increased by the fact that it is
compassion in general that decreases, and not only identification
with the foreigners (here, the Afghan population) whom soldiers
meet during their deployment. This observation is especially important when we turn to the central question raised by our findings:
what can be done to promote the positive effect of deployment
on commitment to the public interest and to reduce the negative
impact on compassion? Before addressing that question, allow us to
remark that a lot is already being done to avoid emotional withdrawal. The Danish Defense specifically seeks to prevent cultural
distancing based on racial and ethnic differences, for instance, by
ensuring that most patrols are done together with Afghan National
Army soldiers.
These initiatives may explain why the decrease in compassion is general and not concentrated on compassion related to foreigners. Had
it been simply a cultural distancing, compassion would have been
delimited to Danes; had it simply been organizational, it would
have been delimited to soldiers in general. However, the decrease
happened approximately equally for all items measuring compassion. As these remarks illustrate, we recognize the dilemma caused
by putting forward crude policy recommendations with regard to
avoiding emotional withdrawal: closer social cohesion between the
soldiers during deployment might prevent the general decrease in
compassion by fulfilling human needs for relatedness (Gagné and
Deci 2005), but this might also increase the cultural distance to the
population, highlighting the soldiers as an in-group and others as
an out-group. On the other hand, encouraging close relationships
between Danish soldiers and the Afghan population during deployment would probably be both too dangerous and show little effect,
given that forced intimacy seldom works anyway.
Conclusion
The research question was how deployment to war affects soldiers’
public service motivation and whether earlier deployment experience moderates this effect. We found that deployment to the war
zone in Afghanistan affected Danish soldiers’ public service motivation differently for the different dimensions: compassion decreased,
and commitment to the public interest increased during the first
deployment. The soldiers’ level of self-sacrifice was not significantly
changed. Concerning the moderation from earlier deployment
With the purpose of preventing some of the impact that killing in
experience, we found that deployment to war seemed to affect
war has on soldiers, Guldbech (2009) argues that Danish soldiers’
commitment to the public interest more for soldiers without earlier
predeployment training should focus more on ethics and on reassurdeployment experience compared to soldiers with earlier deploying the soldiers that society benefits from their service. The recomment experience. In contrast, compassion actually seemed to
decrease a little more for soldiers with earlier deployment experience mendations following from our observations, that the decrease
compared to soldiers without this experience. None of these interac- in compassion should be prevented and that the increase in commitment to the public interest should be furthered, are consistent
tion effects, however, was statistically significant. The main finding
with those of Guldbech. Yet initiatives after
is thus that soldiers become more normatively
repatriation should not be ignored. Civiland less affectively motivated during deployThe main finding is thus that
military cooperation is very important during
ment. This is a contribution to the discussion
international operations, but civil-military
soldiers become more normaabout the stability of public service motivarelations can also be important when soldiers
tion (Wright and Grant 2010, 691–92) and
tively and less affectively motireturn home. More generally, the emotional
suggests public service motivation can change
vated during deployment.
detachment of deployed soldiers shown in this
substantially when an individual is exposed
study raises important questions regarding
to a dramatic event such as deployment to
the lasting impact of deployment after repatriation and how a lasting
war. We have investigated public service motivation changes in an
negative effect can be avoided. More interaction between soldiers
extreme setting, but the findings are relevant for many other situand civilians and more openness concerning military affairs, for
ations because of their important implications for how we study
example, could contribute to soldiers’ reintegration after deployand even use public service motivation. The finding that public
ment in terms of reestablishing their affectively based motivation
service motivation is a dynamic state can help us understand other
to contribute to society. Distancing oneself from the sufferings of
questions (such as whether higher levels of public service motivaothers and embracing abstract motives may be an important survival
tion found among public employees are attributable to attractionselection-attrition or socialization and adaptation mechanisms), and mechanism during deployment, but if the result is lasting emotional
withdrawal, it may not only affect the soldiers’ perceptions of others
it also suggests that managers should be careful about what they
Does Deployment to War Affect Public Service Motivation? A Panel Study of Soldiers Before and After Their Service in Afghanistan
475
Appendix
Table A1 Descriptive Statistics for All Soldiers (soldiers answering in both rounds in parentheses)
Variable
Values
Obs.
Mean
SD
Min.
Max.
Commitment to the public interest (CPI)
Sum index of item shaded for CPI in table 2, going theoretically from 0 (low CPI)
to 100 (high CPI)
291
(146)
80.02
(79.88)
14.78
(14.12)
6.25
(25)
100
(100)
Compassion (COM)
Sum index of item shaded for COM in table 2, going theoretically from 0 (low
COM) to 100 (high COM)
292
(152)
64.69
(67.14)
19.94
(18.65)
10
(10)
100
(100)
Self-sacrifice (SS)
Sum index of item shaded for SS in table 2, going theoretically from 0 (low SS) to
100 (high SS)
283
(144)
58.11
(57.22)
20.89
(21.47)
5
(5)
100
(100)
Gender
0 = man
1 = woman
297
(156)
.03
(.05)
.17
(.22)
0
(0)
1
(1)
Age
Age in years
243
(156)
24.72
(24.71)
4.95
(5.22)
19
(19)
42
(42)
After deployment
0 = Surveyed before investigated deployment
1 = Surveyed after
297
(156)
.42
(.50)
.50
(.50)
0
(0)
1
(1)
Prior deployment experience
0 = Without prior deployment experience
1 = Have prior deployment experience
245
(156)
.34
(.33)
.47
(.47)
0
(0)
1
(1)
Table A2 Bivariate Correlations
Commitment to the public interest
Compassion
Self-sacrifice
Gender
Age
After deployment
Commitment to the public interest
1.0000
Compassion
.0793
Self-sacrifice
.4734
.2699
1.0000
Gender
.0119
–.0387
–.0610
Age
.1144
.0512
.1637
.0368
1.0000
After deployment
.0854
–.0708
–.0389
.0539
.0482
1.0000
Prior deployment experience
–.0347
.0301
.0483
.0869
.6457
.0121
1.0000
and society but also their chances of reintegrating with others and
into society.
References
Andersen, Lotte Bøgh, Torben Beck Jørgensen, Anne Mette Kjeldsen, Lene Holm
Pedersen, and Karsten Vrangbæk. 2012a. Public Value Dimensions: Developing
and Testing a Multidimensional Classification. International Journal of Public
Administration 35(11): 715–28.
———. 2012b. Public Values and Public Service Motivation: Conceptual and
Empirical Relationships. American Review of Public Administration. Published
electronically on March 28. doi: 10.1177/0275074012440031.
Andersen, Lotte Bøgh, Thomas Pallesen, and Lene Holm Pedersen. 2011. Does
Ownership Matter for Professionals’ Public Service Motivation? Review of Public
Personnel Administration 31(1): 10–27.
Andersen, Lotte Bøgh, and Søren Serritzlew. 2012. Does Public Service
Motivation Affect the Behavior of Professionals? International Journal of Public
Administration 35(1): 19–29.
Ardant du Picq, Charles. 1947. Battle Studies: Ancient and Modern. Harrisburg, PA:
Military Service Publishing.
Barash, David P., and Charles P. Webel. 2009. Peace and Conflict Studies. 2nd ed.
London: Sage Publications.
Bensahel, Nora, Olga Oliker, and Heather Peterson. 2009. Improving Capacity
for Stabilization and Reconstruction Operations. Santa Monica, CA: RAND
Corporation.
Brewer, Gene, A. 2008. Employee and Organizational Performance. In Motivation
in Public Management: The Call of Public Service, edited by James L. Perry and
Annie Hondeghem, 136–56. Oxford, UK: Oxford University Press.
476
Public Administration Review • May | June 2013
Prior deployment
1.0000
1.0000
Brewer, Gene A., and Sally Coleman Selden. 1998. Whistle Blowers in the Federal
Civil Service: New Evidence of the Public Service Ethic. Journal of Public
Administration Research and Theory 8(3): 413–39.
Camilleri, Emanuel. 2007. Antecedents Affecting Public Service Motivation.
Personnel Review 36(3): 356–77.
DeHart-Davis, Leisha, Justin Marlowe, and Sanjay K. Pandey. 2006. Gender
Dimensions of Public Service Motivation. Public Administration Review 66(6):
873–87.
Edwards, Jeffrey R. 2011. The Fallacy of Formative Measurement. Organizational
Research Methods 14(2): 370–88.
Eighmey, John. 2006. Why Do Youth Enlist? Identification of Underlying Themes.
Armed Forces and Society 32(2): 307–28.
Gagné, Marylène, and Edward L. Deci. 2005. Self-Determination Theory and Work
Motivation. Journal of Organizational Behavior 26(4): 331–62.
Giauque, David, Adrian Ritz, Frédéric Varone, and Simon Anderfuhren-Biget. 2012.
Resigned but Satisfied: The Negative Impact of Public Service Motivation and
Red Tape on Work Satisfaction. Public Administration 90(1): 175–93.
Grossman, Dave. 1995. On Killing: The Psychological Cost of Learning to Kill in War
and Society. Boston: Little, Brown.
Guldbech, Kim W. H. 2009. “Soldaten og de etiske overvejelser—ud og hjem” [The
Soldier and Ethical Considerations—Out and Home]. Master’s thesis, Royal
Danish Defense College.
Henriksen, Rune. 2007. Warriors in Combat—What Makes People Actively Fight in
Combat? Journal of Strategic Studies 30(2): 187–223.
Hondeghem, Annie, and James L. Perry. 2009. EGPA Symposium on Public
Service Motivation and Performance: Introduction. International Review of
Administrative Sciences 75(1): 5–9.
Hosek, James, Jennifer Kavanagh, and Laura Miller. 2006. How Deployments Affect
Service Members. Santa Monica, CA: RAND Corporation.
Huntington, Samuel P. 1957. The Soldier and the State: The Theory and Politics of CivilMilitary Relations. Cambridge, MA: Belknap Press of Harvard University Press.
Kim, Sangmook. 2011. Testing a Revised Measure of Public Service Motivation:
Reflective versus Formative Specification. Journal of Public Administration
Research and Theory 21(3): 521–46.
Kim, Sangmook, and Wouter Vandenabeele. 2010. A Strategy for Building Public
Service Motivation Research Internationally. Public Administration Review 70(5):
701–9.
Kim, Sangmook, Wouter Vandenabeele, Bradley E. Wright, Lotte Bøgh Andersen,
Francesco Paolo Cerase, Robert K. Christensen, Céline Desmarais, Maria
Koumenta, Peter Leisink, Bangcheng Liu, Jolanta Palidauskaite, Lene Holm
Pedersen, James L. Perry, Adrian Ritz, Jeannette Taylor, and Paola De Vivo.
2013. Investigating the Structure and Meaning of Public Service Motivation
across Populations: Developing an International Instrument and Addressing
Issues of Measurement Invariance. Journal of Public Administration Research and
Theory 23(1): 79–102.
Kjeldsen, Anne Mette. 2012. Dynamics of Public Service Motivation: A Panel
Study of Attraction, Selection and Socialization: Effects in the Production
and Regulation of Danish Social Services. Paper presented at the 16th Annual
Conference of the International Research Society for Public Management,
Rome, Italy, April 11–14.
Kjeldsen, Anne Mette, and Christian Bøtcher Jacobsen. 2012. Public Service
Motivation and Employment Sector: Attraction or Socialization? Journal of
Public Administration Research and Theory. Published electronically on October
29. doi: 10.1093/jopart/mus039.
Le Scanff, Christine, and Jaqueline Taugis. 2002. Stress Management for Police
Special Forces. Journal of Applied Sport Psychology 14(4): 330–43.
Little, Roger W. 1964. Buddy Relations and Combat Performance. In The New
Military: Changing Patterns of Organization, edited by Morris Janowitz,
195–224. New York: Russell Sage Foundation.
Lyk-Jensen, Stéphanie Vincent, Julie Heidemann, and Ane Glad. 2012. Soldater—før
og efter udsendelse, en analyse af motivation, økonomiske forhold og kriminalitet
[Soldiers Before and After Deployment: An Analysis of Motivation, Economic
Conditions, and Crime]. Copenhagen, Denmark: Danish National Centre for
Social Research.
Marshall, S. L. A. 1947. Men against Fire: The Problem of Battle Command in Future
War. New York: William Morrow.
McMahan, Jeff. 2009. Killing in War. Oxford, UK: Oxford University Press.
Møller, Lars R. 2010. Vi slår ihjel, og lever med det [We Kill and Live with It].
Copenhagen, Denmark: Informations Forlag.
Moskos, Charles C. 1971. Vietnam: Why Men Fight. In The Anti-American
Generation, edited by Edgar Z. Friedenberg, 217–37. Chicago: Aldine.
Newsome, Bruce. 2003. The Myth of Intrinsic Combat Motivation. Journal of
Strategic Studies 26(4): 24–46.
Pandey, Sanjay K., and Edmund C. Stazyk. 2008. Antecedents and Correlates of
Public Service Motivation. In Motivation in Public Management: The Call of
Public Service, edited by James L. Perry and Anne Hondeghem, 101–17. Oxford,
UK: Oxford University Press.
Perry, James L. 1996. Measuring Public Service Motivation: An Assessment of
Construct Reliability and Validity. Journal of Public Administration Research and
Theory 6(1): 5–22.
Perry, James L., and Lois Recascino Wise. 1990. The Motivational Bases of Public
Service. Public Administration Review 50(3): 367–73.
Peterson, Robert A. 1994. A Meta-Analysis of Cronbach’s Coefficient Alpha. Journal
of Consumer Research 21(2): 381–91.
Rainey, Hal G., and Paula Steinbauer. 1999. Galloping Elephants: Developing
Elements of a Theory of Effective Government Organizations. Journal of Public
Administration Research and Theory 9(1): 1–32.
Ritz, Adrian. 2011. Attraction to Public Policy-Making: A Qualitative Inquiry into
Improvements in PSM Measurement. Public Administration 89(3): 1128–47.
Shils, Edward, and Morris Janowitz. 1948. Cohesion and Disintegration in the
Wehrmacht in World War II. Public Opinion Quarterly 12(2): 280–315.
Singer, Peter W. 2004. War, Profits, and the Vacuum of Law: Privatized Military
Firms and International Law. Columbia Journal of International Law 42: 521–49.
Stouffer, Samuel A., Marion Lumsdaine, and Arthur Lumsdaine. 1949. The American
Soldier: Combat and Its Aftermath. Princeton, NJ: Princeton University Press.
Stouffer, Samuel A., Edward Suchman, Leland DeVinney, Shirley Star, and Robin
Williams, Jr. 1949. The American Soldier: Adjustment during Army Life.
Princeton, NJ: Princeton University Press.
Sykes, Gresham M. 2007. The Society of Captives: A Study of a Maximum Security
Prison. Princeton, NJ: Princeton University Press.
Tekleab, Amanuel G., Narda R. Quigley, and Paul E. Tesluk. 2009. A Longitudinal
Study of Team Conflict, Conflict Management, Cohesion, and Team
Effectiveness. Group and Organization Management 34(2): 170–205.
Van der Dennen, Johan M. G. 2005. Combat Motivation. Peace Review 17(1):
81–89.
Vandenabeele, Wouter. 2007. Toward a Theory of Public Service Motivation: An
Institutional Approach. Public Management Review 9(4): 545–56.
———. 2008. Development of a Public Service Motivation Measurement Scale:
Corroboration and Extending Perry’s Measurement Instrument. International
Public Management Journal 11(1): 143–67.
———. 2009. The Mediating Effect of Job Satisfaction and Organizational
Commitment on Self-Reported Performance: More Robust Evidence of the
PSM–Performance Relationship. International Review of Administrative Sciences
75(1): 11–34.
Vandenabeele, Wouter, Sarah Scheepers, and Annie Hondeghem. 2006. Public
Service Motivation in an International Comparative Perspective: The UK and
Germany. Public Policy and Administration 21(1): 13–32.
Wong, Leonard, Thomas Kolditz, Raymond Millen, and Terrence Potter. 2003. Why
They Fight: Combat Motivation in the Iraq War. Carlisle, PA: U.S. Army War
College and Strategic Studies Institute.
Wright, Bradley E., and Adam M. Grant. 2010. Unanswered Questions about Public
Service Motivation: Designing Research to Address Key Issues of Emergence and
Effects. Public Administration Review 70(5): 691–700.
Does Deployment to War Affect Public Service Motivation? A Panel Study of Soldiers Before and After Their Service in Afghanistan
477
Copyright of Public Administration Review is the property of Wiley-Blackwell and its content may not be
copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written
permission. However, users may print, download, or email articles for individual use.