On moral grounds: Moral identity and moral disengagement in relation to military deployment

ABSTRACT

Following Blasi’s self-model of moral functioning, this exploratory study aims to gain insight into the construction of moral identity among military professionals experiencing daily moral dilemmas during deployment. Semi-structured interviews with 45 servicemen were content-coded and analyzed, exploring relationships between moral identity and verbalized moral disengagement. The results revealed three patterns, giving direction for further research. First, the analyses suggest that a higher moral awareness is associated with more justifications for one’s own behavior. Second, leaders showed more inclination toward conscious moral identity than their subordinates. Third, the number of moral dilemmas experienced during deployment were similar for servicemen of all ranks. Moreover, critical self-reflection and self-assessment were relatively underreported across all ranks.

What is the public significance of this article?—This study suggests that individuals showing a higher inclination toward a moral identity (i.e. making more expressions regarding the morality of their activities, either evaluating them as good or as bad), also try to distance themselves from situations and behavior they consider bad. This moral disengagement is found for both military leaders and subordinates. This study suggests that leaders utter more expressions that imply a moral identity, even though all ranks describe similar numbers of moral dilemmas.

Difficult decisions are at the heart of the military profession. After all, military operations demand decisions that have far-reaching (sometimes even lethal) consequences for servicemen themselves and for other parties involved (Seiler, Fischer, & Ooi, 2010). One reason for this is that servicemen are legally permitted to use violence, so they need to decide when and to what extent this is appropriate. Another reason for this is that they may witness the use of violence by others that demands a response (Drescher et al., 2011). Consequently, and when interacting with others, servicemen often face the question: what is the right thing to do? (cf. De Graaff, De Vries, Van Bijlevelt, & Giebels, 2017; Jennings & Hannah, 2011; Sparks & Siemens, 2014; Van Baarda & Verweij, 2006). Typically, such situations involve contradicting moral principles and/or a weighing of consequences for different parties involved. These situations usually are referred to as moral dilemmas (Kimhi & Kasher, 2015; Van Baarda & Verweij, 2006).

Military moral dilemmas

Moral dilemmas in military contexts do not present themselves with clear-cut answers as they consist of situations where values and loyalties cannot be lived up to simultaneously (e.g., Jennings & Hannah, 2001; Johnson, 2008). This can occur in combat- or life-threatening situations, but also in other peacekeeping- and daily life situations of military personnel. An example of the former would be an experience of Dutch military personnel when deployed in Afghanistan on a police-training mission. The Dutch soldiers did not approve of the treatment of captured insurgents by the local forces, which they considered inhumane. However, the local forces, all heavily armed and outnumbering the Dutch, did not appreciate the Dutch criticism.¹ This presented a clash between concerns for self-preservation on the one side and humaneness and compliance to the code of conduct on the other side. A typical example of the latter category concerns the often mentioned struggle of military health professionals between their duty to provide the best medical care and the goal of the swift and effective completion of the mission (Johnson, 2008).

Today’s growing complexity of military operations arguably creates more ambiguity and thus makes moral dilemmas an increasingly salient aspect of the military profession across ranks and arms of service (Broesder, Opden Buijs, Vogelaar, & Euwema, 2015; De Graaff, Schut, Verweij, Vermetten, & Giebels, 2016a). Indeed, a study by De Graaff et al. (2016a) showed that all servicemen, from low to high ranks, experience on a regular basis moral dilemmas during deployment, ranging from dilemmas related to their team-activities and the mission itself, to interactions with the local population, and to personal issues such as interactions with their home front. Furthermore, the salience of moral dilemmas is arguably fueled by (1) the rapid technological developments in weaponry leading to potentially more severe consequences of one’s actions, and (2) changes to the decision-making authority of deployed units, with the rise of a general tendency to decentralize decision-making. This provides more room for one’s own assessment of the situation and thus requires higher-level competencies (cf. Richardson, Verweij, & Winslow, 2004).

Moral dilemmas can shatter strong personal ideas about society and its values such as justice, humaneness and fairness, and thus strong ideas about life itself (cf. Folger, Cropanzano, & Goldman, 2005; Haidt, 2001). As such, inner conflicts may arise leading to a “[d]isruption in an individual’s confidence and expectations about one’s own or others’ motivation or capacity to behave in a just and ethical manner” (Drescher et al., 2011, p. 9), causing psychological, emotional and spiritual problems, referred to as moral injury (e.g., Maguen & Litz, 2012; Molendijk, Kramer, & Verweij, 2018; Shay, 2014).

With the stakes in military operations being high and the associated moral dilemmas potentially having major consequences for all parties involved, we consider it important to explore moral identity and the strategies that military personnel use – i.e. moral disengagement – to manage intrapersonal incongruences in moral dilemmas when in operational settings.

Moral identity

We contend that in all types of moral dilemmas during operations, servicemen are, due to the emphasis on training automated skills and drills, often inclined to act quickly and to reflect on their decisions and behavior only afterward. The element of the self that is responsible for translating considerations, judgments, principles and ideals into moral motivation, is referred to as moral identity (Aquino, Reed, Thau & Freeman, 2007; Blasi & Glodis, 1995; Hardy & Carlo, 2005). According to Aquino, Freeman, Reed, Lim, & Felps (2009) “a person’s moral identity is stored in memory as a complex knowledge structure consisting of moral values, goals, traits and behavioral scripts.” Blasi’s self-model of moral functioning, is an important framework for explaining the link between moral identity and behavior (Blasi, 1983; Walker, 2004) and entails not only centrality, but other aspects of moral identity as well that may explain the variance in outcome of ethical decision-making. The model integrates moral cognition and moral personality in order to explain behavior, in terms of Blasi: moral functioning (Walker, 2004). It consists of three sub-categories: 1) the moral self, 2) the individual’s sense of personal responsibility, and 3) self-consistency (Blasi, 1983; Walker, 2004). The first sub-category of moral self refers to the centrality of morality in considerations, including the salience of morality in one’s identity. Blasi argues that morality may have differing degrees of centrality in individuals’ lives: for some people, moral standards and values are well integrated in their daily activities, while such integration seems to be practically absent for others (Blasi, 1983; Walker, 2004). Sense of personal responsibility, the second sub-category, refers to the extent to which individuals consider themselves responsible for their own actions and the situation at hand. High-level feelings of responsibility can be categorized as proactive (having an intrinsic sense of responsibility) or reactive (a sense of being held accountable). Finally, self-consistency, refers to the reflection upon the integration of values and moral standards in daily activities. According to Blasi, self-consistency is a fundamental motive for functioning, that can be satisfied by congruence between judgment and action. That is, individuals evaluate their own behavior by comparing it with their personal moral compass in order to come to a sense of self-integrity (cf. Walker, 2004). These three subcategories jointly form one’s moral identity referring to a certain state of being regardless of whether the individual is considered a “good” person by third parties (see Table 1 for the operationalization of this construct).

Table 1.

Sensitizing concepts based on Bandura (1999) and Blasi (1983)

	Main category	Subcategory
Moral identity	Moral self
	Sense of personal responsibility	Pro-active responsibility/Accountability Low sense of moral worth
	Self-consistency
Moral disengagement	Reconstruction of own behavior or the incident	Moral justification Advantageous comparison Euphemistic labeling
	Reconstruction of own role in the incident	Displacement of responsibility Diffusion of responsibility
	Reconstruction of the parties involved	Dehumanization Blaming the victim

Moral disengagement

Behavior, however, does not only depend on stable personal traits. For example, Aquino et al. (2009) found that the centrality of an individual’s moral identity is not stable, but depends on situational and contextual factors as well. Other studies suggest similar findings as well. For example, (situational) features influence the outcome of the decision-making process, such as moral intensity (Jones, 1991), group dynamics (Fischer et al., 2011), social norms (Shotland & Straw, 1976), and emotions elicited in the situation (De Graaff et al., 2016a; Schut, de Graaff, & Verweij, 2015). As choosing between two “goods” or two “bads” is likely to create discomfort it may give rise to self-regulation processes, presumably to ultimately maintain a positive self-image (cf. Higgins, Bond, Klein, & Strauman, 1986). When individuals eventually display behavior that is not in line with their moral identity, they may justify this behavior by means of moral disengagement. This implies a psychological mechanism that permits an individual to selectively, either deliberately or unconsciously, reframe one’s own actions and to dissociate from them in order to isolate those actions from one’s personal standards of what is morally acceptable (Bandura, 1999). Bandura distinguishes three main categories of moral disengagement (see Table 1 for the operationalization of this construct). The first centers around the reconstruction of own behavior or the incident itself. This may involve masking what happened by using ambiguous language, or justifying actions by comparing personal behavior to inhumane actions carried out by other parties. The second category includes the reconstruction of one’s personal role in the incident, for example by arguing that one’s personal role is only minimal and one cannot be held accountable. The third category includes reconstruction of (the role of) the parties involved, such as blaming the victim or dehumanization.

Focus of this study

Despite the growing interest of previous studies in the domain of military moral dilemmas (e.g., Thompson & Jetly, 2014; Wead, 2015), most studies focused mainly on hypothetical moral dilemmas instead of mirroring the actual representations of the social reality servicemen actually encounter on a mission. That is, they often pose a rather obvious dilemma (for example: should I or should I not fire on a child soldier?) or focus on theoretical dilemmas that require split-second decisions such as the well-known footbridge and trolley-dilemma (e.g., Greene, Sommerville, Nystrom, Darley, & Cohen, 2001). The current study tries to fill this void by focusing on the less explicit daily moral issues as experienced by military personnel. This is important because moral dilemmas are arguably experienced by all ranks almost every day and do not only concern complex, tragic, or split-second decisions. Therefore, it is our goal to provide a first glance at patterns in moral identity and moral disengagement that are present in the psychological process of ethical decision-making in self-reported moral dilemmas. The exploratory and descriptive nature of this study offers a way ahead for future research and subsequently the practical implications thereof for the training of military personnel.

Method

For the purpose of this study, we focused on experiences of Dutch military personnel contributing to peace operations around the world. At the time of the data collection the largest military operation the Netherlands contributed to was the International Security and Assistance Force (ISAF) mission Task Force Urzugan (TFU). We will shortly describe this operation to illustrate the type of missions our interviewees participated in.

The Dutch contribution to the ISAF mission TFU

After a conference of the United Nations and representatives of the Afghan people in Bonn (Germany) on December 5, 2001 it was decided to give the ISAF a central role in rebuilding Afghanistan. ISAF was formed by several European countries including the Netherlands. Its mission was to facilitate the reconstruction of Afghanistan and establish a safe, stable and secure (regional) environment by supporting the Afghan government authority, the Afghan National Security Forces and the Afghan National Police. As such, the mission was a peace mission in which reconstruction was the primary purpose. So, combat forces were present to secure the Dutch troops and other Dutch governmental employees (diplomats) that were giving humanitarian aid and contributed to humanitarian projects, not to enforce peace. Projects included training Afghan policemen, building schools and other governmental institutions, and helping to give access to working water irrigation, modern infrastructure and medical services for the local population. The so-called Dutch approach was developed by the cooperation of three ministerial departments: Defense (the ministry of defense), Development (the ministry of foreign trade and development cooperation) and Diplomacy (the ministry of foreign affairs). For the Dutch, the official start of the largest operation under ISAF was in August 2006, when they became responsible for the southern Afghan province Uruzgan, working together with the Australian forces (Task Force Uruzgan – TFU). The boots on the ground could largely be categorized in two types of troops: those responsible for security (combat forces) and those responsible for reconstruction tasks (members of the provincial reconstruction teams – PRT). Both types together formed a Smallest Unit of Action (SUA) when on patrol or during an assignment. The Dutch left Uruzgan in August 2010.

Participants

In the Netherlands, the target population of active military personnel with operational and deployment experience is a difficult (and rather small) group to reach for empirical studies. We, therefore, started off by asking the commander of the brigade that had contributed the largest operational deployment of Dutch forces up to that time (2009–2010), Task Force Uruzgan, for consent to approach his units for interviews. After having received consent, personnel who had been deployed in Task Force Uruzgan (Afghanistan) under ISAF command between 2006 and 2010) were invited to participate. We invited servicemen who were part of operational units (such as combat units and Provincial Reconstruction Teams), as those units were most likely to have experienced contact with the local population, other coalition partners, and high-risk situations. However, as this would only generate interview material regarding a specific mission of army and marine corps personnel (who were deployed in Uruzgan province in Afghanistan), we used snowball sampling to recruit additional participants from other similar missions as well (Noy, 2008). For this, we asked the participants from the original target group to ask their network to contact us if they were willing to be interviewed. In total, 60 servicemen from different branches of service volunteered to participate in the study. The criteria for inclusion were: (1) being in active military service at the time of the interview, (2) having had deployment experience abroad no longer ago than two years prior to the interview, (3) having experienced direct contact with the local population and/or coalition forces during a military deployment. The first two criteria were intended to create a research population with recent deployment experience that they could vividly remember. The third criterion would allow for a broad range of moral questions and dilemmas to occur (next to mission and home front issues). Ultimately, a total of 45 service members were interviewed. The remaining 15 individuals were excluded for either not meeting the study criteria (mostly for not having recent deployment experience) or for their unavailability at the time of the study.

We explained to all participants that the study aimed at gaining insight into how deployment experiences affected the daily life of individual service members. The participants were informed that the general research results would be used for educational and scientific purposes and for pre-deployment training programs, but they were not informed about the exact research goals. However, they were informed that their individual results would not lead to clinical assessments or treatments since this study was not focused on an assessment of their mental health. All interviewees were ensured confidentiality, meaning their commanding officers or other third parties would not be notified of their individual answers. Also, they were informed that they were free to end their participation at any time during the interview session.

Most of the participants had been deployed to Uruzgan in Afghanistan (87% of the participants). Others had been deployed in recent peace operations in Bosnia, Sudan, Somalia, Liberia, Angola and the Kunduz province in Afghanistan with similar operational tasks. The age of the participants ranged from 18–47 years (M = 28 years; SD = 8.3), and participants held ranks from private up to lieutenant-colonel.² As is the case in the general military population, the number of women participants was modest (N = 7, 15.5%; in the general population of the Dutch forces the percentage of female servicemen averages around 10% (Ministerie van Defensie, 2015).

Instrument

Interviews have proven to be an effective instrument, capturing a wide range of information regarding moral issues in a military context (cf. Benham Rennick, 2012; Nilsson, Sjöberg, Kallenberg, & Larsson, 2011). For the purpose of this study, data were collected by means of qualitative semi-structured interviews following a prepared interview guide (see Appendix A for a list of questions asked), which was pre-tested among a small number of veterans before using it in the main study. This interview guide ensured that all participants were asked similar questions by two behavioral scientists.

We used the constructs described in the introduction (see Table 1) as sensitizing concepts, enabling us to interpret the transcripts by organizing and identifying variations of the patterns found in the data (in terms of codes) and as such establishing relationships between these concepts (Lawrence & Tar, 2013; Richardson & Kramer, 2006). We made sense of the data on the basis of the content of the experiences, the context and (co)occurences of themes and ideas shared with us during each interview.

The contents of the transcripts of the in-depth interviews with active serving military personnel were then analyzed in order to gain insight in moral identity and moral disengagement in military deployment settings. Blasi’s self-model of moral functioning was used in order to code and interpret moral identity in the narratives. Bandura’s framework of moral disengagement was used as a starting point for the operationalization of cognitive reconstructions of the servicemen’s actions to make them acceptable (Bandura, 1986). Moral dilemmas were identified by participants describing their own morally challenging situations. We included situations that participants explicitly labeled as a moral dilemma, or were reported as situations that they considered difficult to cope with due to contradicting values and interests or situations leading to expected direct or collateral consequences. In addition to the so-called tragic dilemmas, everyday moral dilemmas were also drawn from the narratives and taken into account. Typical manifestations (markers) of moral dilemmas, moral identity and moral disengagement are presented in Table 2. These markers guided the coding of the transcripts of all narratives. For the data analysis, the total number of utterances of moral disengagement were counted per interview and used in the analyses. Thus, subcategories of moral disengagement (i.e. the individual mechanisms described by Bandura) and moral dilemmas (i.e. the categorization of everyday military moral dilemmas suggested by De Graaff et al., 2016a) were not addressed in the analyses separately but used as prompts.

Table 2.

Coding scheme based on Blasi (1983), Bandura (1999) and De Graaff et al. (2016a)

Construct	Markers/typical manifestations of the constructs in the interviews
Moral identity	Integrates the significance and salience of morality and values in an individual’s considerations (moral self). States feeling highly responsible or accountable in a specific situation for the consequences of actions (accountability). States having (no) moral obligation to act (high or low sense of moral worth). Refers to the reflection on integration of values in personal actions (self-consistency).
Moral disengagement	Makes reconstructions of own behavior or the incident itself. By: - justifying what happens, for example by saying that the goals justify the means - comparing the incident or own behavior with other situations that are considered worse, for example by saying that torture is permitted since the victim killed innocent children - using language that masks what happens, for example discussing collateral damage
	Makes reconstructions of own role in the incident. By: - stating not to be held responsible for what happened, for example by saying someone else gave an order - stating it is unclear who is responsible in the situation, for example by saying there were others present as well who could have intervened
	Makes reconstructions of the parties involved in the situation. By: - using language that dehumanizes individuals or groups of people, for example by referring to them as dogs instead of people - stating the victim has to blame himself for getting into this situation, for example by saying they started the terrorizing first.
Moral dilemmas	Refers to situations, at all levels of complexity and tragedy, in which required to make a decision when binding (personal) moral principles contradict. That is, perceived moral dilemmas that are cultural related, work related (such as regarding the mission or the unit) or personal (such as related to home front).

Procedure

Interview

One of the interviewers (the first author of this article) was continually present throughout all interview sessions; the second interviewer position was shared between two researchers. Both interviewers were affiliated with the Netherlands Ministry of Defense. Standard informed-consent procedure was followed.

A single interview session lasted approximately 50 minutes. The interviews were all digitally recorded and transcribed afterward. During the interview sessions, open-ended questions covering basic demographic information (such as age, rank and professional function) were asked and the participants’ experiences in the mission area were discussed. This was needed in order for the participants to construct internalized and evolving stories (narratives) in which they would address their moral identity. The participants were stimulated to share their experiences, so open-ended questions were used such as “Can you describe an example of a situation that was difficult to cope with during deployment?” This gave us the opportunity to elaborate on the respondent’s remarks. For example, we used follow-up questions addressing their moral identity with questions like: “Can you describe what it was that made this specific situation so difficult, what was at stake?” and, ’Can you describe a situation in which your values or norms were challenged?’ or, “Can you describe how you look back on your own behavior and the behavior of others at that time?” For all prompts, participants were also asked if they could reproduce more examples up to the point where they indicated there were no more.

Coding procedure

The recordings of the interview sessions were transcribed and systematically content analyzed by three raters, using a coding guide covering all constructs (see Table 2). Markers that highlighted key aspects of the construct, were drawn from the responses of the pilot study. First, two transcripts were independently labeled in terms of moral identity remarks and moral disengagement strategies by all three raters and jointly discussed afterward. This led to several categories and markers to be adjusted or narrowed down to more generally applicable codes. The resulting coding scheme was then used to let two coders independently code four interviews. Cohen’s Kappa was relatively low (.43), which appeared to be due to one rater leaving several narratives uncoded, whereas the other rater coded the narratives as moral dilemmas. Everyday dilemmas were left blank initially for the most part, suggesting the rater had primarily focused on the moral and intercultural dilemmas that were more severe or that ended tragically. It was decided that the rater who had left narratives blank needed to code these specific sections again (unaware of the labels used in the other rater’s codings). This was performed and led to a higher Kappa:.78, which reflects substantial inter-rater agreement. The two raters labeled the remaining interview transcriptions (Cohen’s Kappa displayed .78 for the remaining transcriptions as well). Disagreements about the labeling were resolved by discussion.

Analyses

After coding all narratives, we established per interview: a) the frequency of moral dilemmas, b) the total number of remarks about the elements of moral identity (per sub-category), and c) the total number of remarks regarding moral disengagement. The mean scores of a variable represent the average number of remarks made per participant for each category across the 45 interviews. In addition to the separate sub-categories of moral identity, a sum score of the three elements of Blasi’s model was computed (referred to as “overall moral identity score”).

Demographic variables

The number of participants in the different categories was not high enough to make useful and relevant distinctions in demographic variables such as age, gender and rank. We did, however, include the question of how many deployments the participants had experienced (i.e., a first deployment versus multiple deployments) as well as their positions in the chain of command (i.e., in a leadership or in a subordinate position). The first distinction was included because a first deployment may evoke more intense experiences or more elaborate reflections, while having had multiple deployments may make people more accustomed to the challenges they encounter. We considered leadership position to be a relevant variable to include because military leaders are subjected to ethics education more than their subordinates and their (sense of) responsibility may be higher due to their formal position.

Results

Descriptive data for our main constructs

All narratives were coded for the operationalization of the different variables. For each narrative, the times a variable (i.e., moral dilemma; moral identity; moral disengagement) was mentioned was determined. The means, standard deviations, and correlations are presented in Table 3. Table 3 shows that in total, 285 separate dilemma situations were described. On average, participants described 6 moral dilemmas per interview (M = 6.33; SD = 5.86). Also, each participant made almost 5 remarks referring to their moral identity (M = 4.76; SD = 3.94). With regard to moral disengagement, the participating servicemen made 2 of such utterances on average (M = 2.27; SD = 2.96).

Table 3.

Correlations between demographics, elements of moral identity and moral disengagement

	M	SD	1.	2.	3.	4.	5.	6.	7.	8.	9.	10.
(1) Leadership (1, 2)	1.62	0.49	-
(2) Number of deployments (1, 2)	1.59	.50	.38*	-
(3) Number of moral dilemmas	6.33	5.86	.19	.09	-
(4) Element 1 – Moral self	2.93	2.26	.52**	.22	.31*	-
(5) Element 2 – Sense of personal responsibility (general)	1.02	1.41	.58**	.27	.36*	.42**	-
(6) Element 2a – Sense of personal responsibility (proactive)	.67	1.11	.45**	.16	.27	.35*	.82**	-
(7) Element 2b – Sense of personal responsibility (reactive)	.36	.80	.40**	.31	.26	.26	.62**	.06	-
(8) Element 3 – Self-consistency	.80	1.24	.43**	.31	.39**	.42**	.54**	.30*	.53**	-
(9) Moral Identity score (element 1–3)	4.76	3.94	.64**	.32	.43**	.86**	.77**	.59**	.54**	.75**	-
(10) Low sense of moral worth	.60	.92	.19	.06	.40**	.37*	.41**	.29	.32*	.21	.43**
(11) Moral disengagement-score	2.27	2.96	.05	.07	.39**	.31*	.38*	.25	.32*	.18	.37*	.66**

n = 45

* = significance p < 0.05

** = significance p < 0.01

Leadership 1 = in a subordinate position, 2 = in a leadership position

Number of deployments 1 = deployed once, 2 = deployed more than once

An example of one of the 285 moral dilemmas described by the servicemen, is the following one, indicating cultural differences that caused a moral dilemma for a Dutch female officer serving in a UN mission abroad.

I was the only woman in the camp. Due to cultural differences I experienced frequent discussions with the coalition partners – from another ethnic background – concerning the things I can and cannot do. I understand that the coalition partners feel uncomfortable with me wearing sporting shorts and T-shirts when exercising. Although I outrank the coalition partners, I feel that my position as one of the few officers present is under strain. To me, the dilemma was how far I would go in my confirmation to the norms of the coalition partners’ culture regarding interactions with women and change my exercising routine although this would mean I would not be able to participate in all the sports events with my Dutch colleagues.

Another example of a moral dilemma was expressed by a captain. This captain considers the explicated mission goals to be ambiguous, hindering deciding the right course of action:

Our commanding officers told us our mission was to get back home in one piece, as there was too much political fuzz. I still believe that is a strange goal: why do we [the military] go to these areas just to get back safely? Then we’d better not go at all. In my opinion, we are there to establish safety and security, and that sometimes means people get hurt. I understand that we desire to get back all in one piece, but that can hardly be the primary goal of our military operation?! How is that supposed to help us in setting our operational goals and deciding what to do?

With regard to the moral identity of servicemen during military deployment, the participants made a total of 132 remarks, for example, “I consider humaneness as the most important value in my life”, that refer to what Blasi refers to as their moral self, i.e., the significance and salience of morality issues for their considerations (M = 2.93; SD = 2.26). A total of 46 remarks were made regarding the second element of Blasi’s framework: sense of personal responsibility (M = 1.02; SD = 1.41).³ For example: “as their commander, I am responsible for their well-being, so I do not like it when I lack the time to plan a patrol mission meticulously”. The participants also made 27 remarks indicating a low sense of moral worth (M = .60; SD = .92). For example: “Who am I to have an opinion about that?” Finally, 36 remarks were made regarding the third element of Blasi’s self-model of moral functioning, self-consistency (M = .80; SD = 1.24). For example: “looking back on our actions, I wonder: am I what I have done? I have done things I would never do under normal circumstances.” Overall, we found that the three sub-categories of moral identity were positively associated (all r’s > .42, p < .01). Thus, the more participants were inclined to mention something in one category, the more they also referred to the other moral identity sub-categories (e.g., the moral self appears to be positively associated with a sense of personal responsibility).

A total of 102 remarks about moral disengagement were made. A Dutch infantry sergeant displays moral disengagement in the following example:

You need to understand that we all live in different worlds. The world at home, the world at work, the world during deployment. Those are all separate worlds, with their own and differing standards about right and wrong. You consider those standards as completely ‘true’ in that specific world, even when they are not ‘normal’ back home. So that’s why you do what you do …

Correlational analyses

The correlational analyses (see Table 3) show that there is a positive relationship between two elements of moral identity (i.e., moral self and sense of personal responsibility) and the moral disengagement-score. Thus, the more pronounced the moral side of one’s identity, the more reference to moral disengagement was made. Also, the number of moral dilemmas mentioned and remarks about a low sense of moral worth display a positive relationship with utterances of moral disengagement. This indicates that individuals who experience a higher number of moral dilemmas or consider themselves as being less able to act in a situation (i.e. having less agency), express more justifications for their actions (or lack thereof). Interestingly, a low sense of moral worth is positively related to the sense of responsibility in general and with accountability, but not with proactive responsibility. This suggests that an intrinsic willingness to take responsibility is not related to whether or not an individual believes that (s)he has a moral obligation to act, whilst for responsibility in general and for accountability this does seem to be the case.

Deployment experiences

Independent sample t-tests were conducted in order to examine the relationship between deployment experience (i.e., a first deployment (N = 15) versus multiple deployment (N = 22) experiences) and moral identity and moral disengagement (see Table 4). A (marginal) significant difference was found for overall moral identity (t(35) = − 2.00, p < .10). This can be attributed to differences in self-consistency (t(35) = − 1.90, p < .10) and a specific element of sense of personal responsibility namely accountability (t(35) = −2.27, p < .10), with servicemen who had multiple previous deployments mentioning these elements more often.

Table 4.

Means and SDs for main constructs in relation with deployment experience

	One deployment (n = 15)		Two or more deployments (n = 22)
	M	SD	M	SD	t
Moral identity score	3.53	3.31	6.14	4.22	−2.00*
Moral self	2.53	2.30	3.59	2.44	−1.32, ns
Sense of personal responsibility	.60	1.35	1.36	1.40	−1.65, ns
Accountability	.07	.26	.45	.74	−2.27*
Pro-active responsibility	.53	1.30	.91	1.07	−.96, ns
Self-consistency	.40	.74	1.18	1.47	−1.90*
Moral disengagement-score	2.13	3.23	2.59	3.20	−.43, ns
Low sense of moral worth	.67	.90	.77	1.02	−.33, ns
Moral dilemmas	6.00	5.94	7.14	6.65	−.53, ns

*significant difference at.1 level

Leaders and subordinates

Independent sample t-tests were conducted in order to address the relationship between being in a leadership position and expressions of moral dilemmas, moral identity elements and moral disengagement. No significant differences between leaders (N = 17) and subordinates (N = 28) were found for the number of moral dilemmas reported, the subcategory low sense of moral worth, and moral disengagement. For the three elements of moral identity described by Blasi (1983), the leaders scored significantly higher than their subordinates (see Table 5).

Table 5.

Means and SDs for main constructs in relation with leadership-position

	Yes (n = 17)		No (n = 28)
	M	SD	M	SD	t
Moral identity score	7.94	4.18	2.82	2.16	4.69***
Moral self	4.41	2.65	2.04	1.40	3.42**
Sense of personal responsibility	2.06	1.60	.39	.79	4.01***
Accountability	.76	1.09	.11	.42	2.38*
Pro-active responsibility	1.29	1.45	.29	.60	2.73*
Self-consistency	1.47	1.63	.39	.69	2.60*
Moral disengagement-score	2.47	3.22	2.14	2.84	.36, ns
Low sense of moral worth	.82	1.19	.46	.69	1.29, ns
Moral dilemmas	7.76	6.88	5.46	5.08	1.29, ns

* significant difference at.05 level

** significant difference at.01 level

*** significant difference at.001 level

Conclusion and discussion

To learn more about moral dilemmas of military personnel during deployment, the present study explored the construction of moral identity and moral disengagement in military operations. Semi-structured interviews with 45 service members were content-coded and analyzed. The results in this exploratory study can be broadly bundled into one of three indicative patterns. Further study is needed to determine whether these patterns would apply more consistently with a larger study population.

The first pattern from the correlational results suggests that a higher awareness of moral challenges in the situation (i.e., a stronger moral identity and a higher awareness of moral challenges) is associated with more justifications for one’s own behavior. Thus, when moral identity is more prominent and the number of moral dilemmas experienced is high, the participants express more rather than fewer mechanisms of moral disengagement. Earlier research showed that individuals who act in line with their values feel better about themselves (Hitlin, 2007). Similarly, for this study, it is possible that the more one considers oneself to be a morally responsible person, the more necessary it seems to correct violations of this self-image when having to deal with double bind issues. The double bind reflects an incongruity between two (or more) self-guides; self-directive standards for being (e.g., Higgins et al., 1986; Visser, 2003). An example of a moral dilemma overtly concerning a double bind is the earlier described situation in which the captain wonders about the explicated (in itself ambiguous) mission goals and thus about what the appropriate line of action is. Thus, the more one perceives moral dilemmas to exist, the more it seems necessary to alleviate the tension resulting from choosing only one value or option, but disregarding another one.

The second pattern we found suggests that leaders (officers as well as NCOs) are more inclined to develop conscious moral identities than their subordinates. They score significantly higher on moral self, meaning the salience and significance of moral issues in their considerations is higher. This finding indicates that the participants are quite capable of describing values and principles that are important to them and that they are able to describe situations where their (core) values were challenged. This can be explained by the fact that they receive more extensive ethics education than their subordinates. Also, their sense of proactive personal responsibility is significantly higher, which arguably follows from the duties connected to their role. Moreover, leaders score higher on self-consistency. This might be due to the fact that they are aware of the role model function they hold for many of the troops. These results should be interpreted cautiously, however. Since leaders receive more extensive training in moral reflection than their subordinates, it may well be true that they are simply better in verbalizing moral issues than their subordinates. Further research on this matter is therefore required. Interestingly, no significant differences between leaders and subordinates (or: non-leaders) were found in the number of moral dilemmas described, the number of utterances of moral disengagement and sense of moral worth. The fact that leaders display a similar (low) sense of moral worth as their followers is notable. After all, the leaders are assumed to be in a position to make a difference and to initiate change, in their own team, their own organization and in interaction with local populations and coalition partners. Future research should aim to address this issue throughout all ranks, including the higher-ranking (staff) officers (e.g., battalion or brigade commanders).

Finally, when looking more closely at the number of utterances, a third pattern becomes clear. Regardless of their position as leader or subordinate, or whether they had more than one deployment, participants described an average of six moral dilemmas. The fact that servicemen are able to recognize morally difficult situations does not necessarily reflect a high level of moral competence or responsibility, since moral competence implies more than just recognizing the challenge as such (cf. Park & Peterson, 2006). Critical self-reflection and self-assessment (measured in self-consistency) are relatively underreported even though moral identity is salient in the servicemen regardless of their rank. These results may hint at the strong urge of individuals to maintain a positive self-image (cf. Hitlin, 2007): when individuals are well aware of the moral dimensions of the situations they find themselves in, they may also be more sensitive to their own roles in those situations. So, when their actions run counter to their personal beliefs of being a “good” person they tend to use response strategies like moral disengagement in order to maintain their sense of “good” self. As such, the use of moral disengagement strategies may indicate a possible drawback of fostering a salient moral identity. In line with this observation, remarks indicating self-consistency were not made very often. It appears that participants are able to describe the moral issue and what they believe to be at stake, but they seem to be less prone to reflect upon their actions and thoughts. Addressing critical self-reflection is therefore an interesting avenue for future research in the military context.

Practical implications

In pre-deployment training Dutch units receive dilemma training and cultural-awareness training. This is not mandatory, however, and commanders may decide that there is no time or necessity for the unit to undergo that training. Also, there is no specific dilemma training for so-called “individual missions” in which a military member is deployed on his or her own, and not as part of a unit. So, ethics training is therefore not a standard element of pre-deployment training. Preparing for all kinds of moral dilemmas might mitigate possible negative side-effects (wrong-doings) that are “approved” by moral disengagement. One of the patterns identified in this study suggests that scoring high on the self-model of moral functioning is not sufficient: it seems to promote rather than hinder the cognitive process of moral disengagement. It is therefore important that servicemen receive additional education on ethical decision-making and related processes such as moral disengagement. Recent studies investigating the relationship between moral reasoning and mindfulness indicate that intentionally paying attention enhances awareness of the experience and contributes to the moral reasoning process (e.g., Ruedy & Schweitzer, 2010). Intentional contemplation or reflection is therefore relevant in order to make sense of the moral dimensions in a situation. When this reflexivity is indeed minimal for the servicemen, military organizations should integrate this aspect more extensively in ethics training programs, in order to stimulate meticulous decision-making.

Limitations

Even though the present study proposes interesting avenues for further research, some limitations need to be addressed. For example, we could only address the verbal reactions of service members. Our material is therefore dependent on the verbal and reflexive capacities of the participants and their willingness to spontaneously share their thoughts and feelings. According to Haidt (2001) individuals only start to deliberate on intuitive judgments when asked to explain the causes of their judgments and actions. Haidt refers to this as the post hoc problem (Haidt, 2001). If that is correct, it would appear that moral considerations are not conveyed spontaneously without unprompted reflection. Moreover, since different modes of personal reflexivity have been proposed (cf. Caetano, 2015), it is also possible that the servicemen did reflect more upon their behavior than what they shared spontaneously with the interviewers. This might be an alternative explanation for the differences found in moral identity between the participants in a leadership position and a subordinate position since leaders have received more extensive training that helps them to reflect and consequently verbalize their ideas and feelings.

Furthermore, a widely-acknowledged issue in interpreting interview data is the degree to which symbolic phenomena are recognized by the researchers when coding the material (cf. Crittenden & Hill, 1971). Despite the fact that the raters of the material were well informed about the military context and had both engaged in participant observations in differing deployment areas, this could have played a role. Although we believe our approach to have been suitable for the exploratory nature of this study, using multi methods in future research may confirm or refute our findings.

Inherent to the exploratory nature of the current study and the fact that only the experiences of military personnel who voluntary chose to sign up to be interviewed were examined, some caution is warranted in the interpretations of the findings. Although this study provides a first glance at the moral identity of military personnel’s moral disengagement, further studies are required to support the validity of the current findings across missions, the pre-mission work-up training and expectations, the size of the deployed unit, the relationship with other echelon entities, and a host of other factors that could influence responses toward the moral dilemmas encountered in the mission.

In addition, the results might also be functionally biased. In our study, we focused on military personnel who have duties outside the perimeter, such as patrol missions, humanitarian assignments, combat support and convoys. For this reason, most of our participants were army or marine corps personnel. In view of technological advances, the moral issues of, for example, air force flight personnel as well as on-base medical teams are relevant to address as well. Moreover, we have not addressed the moral dilemmas and moral identity of military personnel who remain on camp, such as analysts and staff personnel.

Finally, a cultural bias may have occurred since only Dutch military personnel took part in this study. The extent to which our observations are generalizable to other countries’ troops, such as the U.S., whose operations differ enormously, remains an open question. We therefore suggest that future research devote attention to other branches of service as well as other nationalities when addressing moral identity. Another avenue for future research would be to conduct longitudinal research and investigate whether moral identity perception is stable, or shifts over time due to (traumatic) life events or other factors. We suggest using a more quantitative method and broad research sample to further elaborate on this theme.

Conclusion

Ideally, moral identity serves individuals in their ethical decision-making, preventing them from making ethical violations. However, this exploratory study appears to have revealed that although the servicemen consider significance and salience of their moral identity and moral dilemmas at the heart of their profession, their critical self-reflection and self-assessment seems to be somewhat underdeveloped. Such critical reflection may also prove to be valuable in tackling moral disengagement. Counterintuitively, in this specific exploratory sample, our findings suggest that a more strongly developed moral identity serves to initiate more moral disengagement processes, rather than inhibit them, presumably to alleviate the discomfort caused by frequent double-bind situations. Consequently, further research should be conducted, investigating how military personnel deal with moral dilemmas in terms of activating their moral identities and verbalized moral disengagement.

Appendix A.

Interview questions (translated in English from Dutch)

Whether the questions are asked explicitly and the order in which the questions are asked depend on the reactions of the participant.

General questions demographic information

1. Can you introduce yourself?

   1. What is your name,
   2. age,
   3. your rank,
   4. and when were you deployed?

Ice-breaker questions

1. Can you tell us something about that deployment?

   1. What country were you deployed to?
   2. What was the aim of the mission?
   3. Was this an individual deployment, or were you deployed with your unit?
2. Can you describe your activities during this deployment?

   1. What were your tasks?
   2. What did a “regular” day look like?

Moral issues

1. Can you describe a situation in which your values or norms were challenged?

2. Did you experience situations that you consider “not normal”? For example, because you saw behavior that you consider unethical?

   1. Can you elaborate on this situation?
   2. What made it unethical to you?
   3. What effect did this situation have on your (emotional) state?
   4. Would you say ethics is important to you?
   5. What are important values to you as a military?
3. Can you describe an example of a situation that was difficult to cope with during deployment?

   1. Can you describe what it was that made this specific situation so difficult, what was at stake?
   2. Can you describe how you look back on your own behavior and the behavior of others at that time?
4. Did you confront situations in which you had to make a difficult decision, or that you perceived to be a dilemma?

   1. Can you explain what made this situation difficult?
   2. What made it a dilemma?
   3. What was at stake in this situation?
   4. What did you do?
   5. Why did you do this? What thoughts and emotions did you have?
   6. Afterward, what did you think of and feel about the situation and your actions?

Acknowldgement

We would like to thank the Netherlands Ministry of Defence for allowing us to conduct this study. We note for the record that the views expressed in this article do not necessarily reflect the views of the Ministry. Finally, we would like to thank Derek Suchard, Marco van Bommel, Michelle Schut, Ria Wolkorte and Eva van Baarle for their comments on previous versions of this article.

Disclosure statement

No potential conflict of interest was reported by the authors.