Do Ideals Differ and Matter? An Investigation of Authentic Leadership Ideals Among Norwegian Military Officers Across Generations – and Its Impact on Leadership Practice

Do Ideals Differ and Matter? An Investigation of Authentic Leadership Ideals Among Norwegian Military Officers Across Generations – and Its Impact on Leadership Practice Olav Kjellevold Olsen1,3, Rino Bandlitz Johansen2 and Sigurd William Hystad3 1 BI Norwegian Business School, NO 2 Norwegian Defense College, NO 3 University of Bergen, NO Corresponding author: Olav Kjellevold Olsen (olav.olsen@uib.no)


Introduction
The last decades' growing body of leadership research has focused on the importance of ideals, particularly on how followers' leader ideals in terms of prototypes influence their leaders' performance and, vice versa, how leaders' follower ideals influence followers (Junker & van Dick 2014). If and how the individual leaders' personal leadership ideals influence their own leadership performance has attracted limited attention. This lack of research is noteworthy because several organizational practices regard leadership ideals as important antecedents of effective leadership behavior. Hence, substantial efforts and resources are allocated to the coaching and socialization of leadership ideals. Additionally, formulations of such ideals become an integrated part of strategic leadership and numerous MBAs and leadership development programs (e.g., Bass & Bass 2008). The idea behind these practices seems to be that by nurturing the right ideals in leaders, their leadership behavior will subsequently improve. These assumptions could be questioned. Some researchers suggest that ideals are nothing more than a reflection of a person's self-image (Dunning, Perie & Story 1991;MacDonald, Sulsky & Brown 2008), questioning the relevance of external ideals as a unique predictor of leadership performance.
In this context, also the content and formation processes of leadership ideals represent an important research focus ). Related to the content of leadership ideals, an important question is whether or to what degree leadership ideals encompass a moral orientation (e.g., Olsen 2010). Notably, most contemporary leadership theories seem to emphasize this perspective (e.g., Bass & Bass 2008; SCANDINAVIAN JOURNAL OF MILITARY STUDIES Bass & Steidlmeier 1999). Contrary, leadership literature also suggests that such "soft" ethical ideals are misplaced in competitive contexts facing many contemporary organizations. Subsequently, morals is seen more as a form of impression management than an ideal to emulate "in real life" (Olsen 2010).
Furthermore, some studies show that leadership ideals may vary between leaders (e.g., Dorfman, Hanges & Brodbeck 2004), but few studies investigate why such differences occur -beyond the effects of cultural differences. Several studies find generational differences in values that may influence leadership preferences (e.g., Eisner 2005). Thus, one fruitful, but scarcely researched perspective explaining leadership ideals may be age -and more specifically, generational differences. This is particularly relevant given the relatively large variation in age that can be found among leaders in contemporary organizations (Cogin 2012). Nevertheless, the leadership literature including generational perspectives is still relatively scarce (Cogin 2012).
On this backdrop, we have several ambitions for this study: First, we need to describe the ideal leadership preferences of military officers. We have chosen a military context because it represents a particularly competitive and complex environment. The fact that errors in the military can be fatal may contribute to shaping leader ideals more clearly, giving less room for impression management than other settings (Hannah et al. 2009;Kolditz & Brazil 2005). On this basis, it is surprising that only few studies have investigated ideals of military officers. Second, we will investigate the centrality of authentic leadership (AL) in the leadership ideals of military officers. This is because AL is suggested to represent an ideal and morally oriented leadership practice, a "root construct", underlying all well-functioning leadership (Avolio et al. 2004). In addition, the validity of AL has recently been questioned due to retractions of some central publications and allegations of scientific malpractice -which in turn calls for further research related to validity (RetractionWatch 2016; Spoelstra, Butler & Delaney 2016). Third, we aim to investigate differences between younger and older officers in terms of leadership ideals, and to what degree these ideals have bearing on the leaders' own leadership behavior.

Ideals and authentic leadership in competitive and dangerous settings
An ideal can be defined as a principle or way of behaving that is of a very high standard, or something that is perfect, or the best possible. It revolves around the question, "If I could shape the world in my own way, how would I have it to be?" (Rescher 1987: 130). This goes beyond the limits of the practicable, but still stimulates our pursuit of desirable ends -providing meaning and guidance to our endeavors (Rescher 1987). This suggests that people are driven and motivated by ideals, as well as guided by them at the levels of personal decision. It is also worth noting that for the beholder, ideals may be implicitly as well as explicitly represented (Junker & van Dick 2014). Subsequently, individuals may evaluate and practice leadership through cognitive processes related to ideals that they are unaware of and cannot control for (Epitropaki et al. 2013).
AL can be viewed as an ideal-oriented leadership theory that aims to strengthen and supplement traditional leadership theories with an ethical basis (Luthans & Avolio 2003;May et al. 2003). Here, leaders' ability to behave transparently, to include followers in decision processes, and to self-regulate in a way that enables them to lead as a moral role model is considered of paramount importance (e.g., Olsen & Espevik 2017). Thus, authentic leaders' leadership is an expression of their true and real self (Shamir & Eilam 2005), and they merge moral principles with their leadership practice (Walumbwa et al. 2008). Some studies provide evidence for a relationship between AL and morals. Kiersch and Byrne (2015) found that followers attribute moral characteristics to authentic leaders, and Laschinger and Fida (2014) found that AL inspires moral behavior in followers. Others detected a positive relationship between leaders' emphasis on morals and productive leadership (Brown & Treviño 2006). In their study, Den Hartog et al. (1999) suggested that leaders' integrity and trustworthiness are related to positive leadership evaluations, and Olsen, Eid, and Larsson (2010) found that moral justice behavior under pressure is related to positive evaluations of leadership in a military context. Thus, in line with Nichols and Erakovich's (2013) findings from a U.S. student sample, military leaders' conceptions of ideal leadership could also be expected to encompass the dimensions of AL. Walumbwa et al. (2008) operationalize these dimensions as follows: (1) self-awareness, (2) balanced processing of information, encouraging critical perspectives, (3) relational transparency, open about own ideas and emotions, and (4) an internalized moral perspective.
There may be several arguments against such idealization of AL in a military setting. For example, followers' motivation to survive during combat and dangerous operations has been suggested to influence leadership preferences, so that tactical skills and competencies gain primacy over human-oriented behavior (Kolditz 2007;Sweeney, Thompson & Blanton 2009). It could also be argued that a competitive and demanding context like military operations may require a more manipulative leadership approach to get the job done. Along the same lines, Ledeen (1999) emphasizes manipulation, deceit, and destruction as vital leadership tools, and Ferris et al. (2007) suggested "strategic bullying" as an effective leadership approach. Notably, such claims have some empirical support. Rosenthal and Pittinsky (2006) found narcissism, in terms of arrogance, self-absorption, and hostility, an attribute of many powerful and successful leaders, while Deluga (2001) suggested a positive relation between Machiavellianism in U.S. presidents and their performance.
Nevertheless, we find the following relationship most likely: Hypothesis 1: AL -on all dimensions -represents an ideal form of leadership among military officers.

Generational differences in authentic leadership ideals
Differences in work-related attitudes, values, and behavior attributed to generational or age differences have gained increased interest in the research literature (Lyons & Kuron 2014;Sessa et al. 2007). A four-generation categorization is often employed, distinguishing between (1) "traditionalists", born prior to the end of World War II, (2) "baby boomers", born between the end of World War II and the early 1960s, (3) "generation X", born between the mid-1960s and the mid-1980s, and (4) "generation Y", born between the early 1980s and the late 1990s (Costanza et al. 2012). In the military, generations X and Y represent appropriate categorizations, as most of the officers are born between 1967 and 1997. Generational cohort theory contradicts a traditional belief that people mature and develop values, attitudes, and preferences as a function of age (McCrae et al. 1999;Lyons & Kuron 2014). The term generation is often used to describe genealogical kinship, defined as a group of individuals born within the same historical and socio-cultural context, facing the same formative experiences, and developing unifying commonalities as a result (Mannheim 1952). Thus, generational identity may predispose to a certain mode of thought and action, restrict the range of self-expression throughout the members' lives (e.g., Lyons & Kuron 2014), and stimulate common work-related expectations, expressed as psychological contracts (Dencker, Joshi & Martocchio 2008), which in cases of perceived violation may lead to negative emotional reactions, lack of commitment, and intention to quit.
The evidence of generational differences in work relevant attitudes and behavior is fragmented, but several studies show important differences that may influence leadership ideals. In terms of personality, some studies indicate that successive generations are more narcissistic and less empathic (Twenge & Campbell 2012)seemingly contrary to AL. On the other hand, others find an increase in both agreeableness (Andre et al. 2010) and social career interests (Bubany & Hansen 2011) in successive generations. Smola and Sutton (2002) report a decrease in the moral importance of work over time, while Hansen and Leuty (2012) find no differences in altruistic values between the generations. Others report a decrease in overall organizational commitment with successive generations (Costanza et al. 2012). The evidence related to generational differences in leadership ideals is very limited (Lyons & Kuron 2014), but in a study performed by Arsenault (2004), leadership behavior like loyalty, honesty, and caring is perceived as less important to successive generations. In the same vein, Sessa et al. (2007) find that younger generations of leaders are more self-focused and less considerate and consensual -seemingly contrary to AL ideals. In sum, this indicates that there may be differences between older and younger officers in terms of idealizing AL.

Hypothesis 2:
Older generation X officers view AL -on all dimensions -as a more ideal form of leadership than successive generation Y officers.

Ideals and leadership behavior
Following Rechter (1987), ideals represent a direction and motivation for actions and efforts -an ethos that one should strive for, representing how things ought to be. Several mechanisms may serve as explanation of a relationship between leaders' leadership ideals and behavior. According to the self-concordance model (Sheldon & Elliot 1999), goals consistent with individuals' core values or ideals are associated with enhanced and proactive goal striving -suggesting that leaders will be particularly motivated proactively to bring about results and leadership practices that concord with their own ideals. This relationship may be explained by social identity theory (e.g., Blasi 1984). Thus, leadership ideals, as previously shown in studies of moral ideals (Aquino & Reed 2002), may represent a part of one's self-concept and, more specifically, one's social identity, as an ideal that directs how one ought to behave as a leader. This leader identity may be more or less strong, and have more or less content derived from AL. Individual differences may occur in terms of both perceived centrality of leadership as a part of one's social identity and AL as an integrated part of this. Subsequently, the importance placed by an individual leader on AL norms like balanced processing, transparency, morality, and self-awareness or, alternatively, how important it is to be an authentic leader, may represent a motivational source of consistent and sustainable AL behavior, stimulated through self-regulatory mechanisms. One could argue that people with a high content of ethically oriented AL integrated in their ideals are likely to activate emotions of shame or guilt if challenged by temptations to deviate from or not proactively pursue these ideals, for example in terms of immoral and self-serving behavior at the expense of others (e.g., Shao, Aquino & Freeman 2008).
It is worth noting that several studies suggest that individuals have a tendency to adopt self-serving definitions of social concepts, like leadership, in concert with their own self-concept (Dunning, Perie & Story. 1991). Accordingly, social judgments serve a personal agenda in terms of preserving a positive self-concept by self-affirming, projecting own attributes onto definitions of social categories (MacDonald, Sulsky & Brown 2008). This was demonstrated by McElwee et al. (2001), showing an association between participants' selfevaluations and their descriptions of successful leaders -indicative of an egocentrism in defining leadership ideals. Others have suggested that self-assessments of leadership could shape leadership ideals. Following, the specific leadership ideals employed in a given situation may partially be determined by an individual's self-evaluation and less by an external ethos (Lord et al. 2001). This implies that in order to investigate the effects of ideals, the effects of self-evaluations should be controlled for in the model. On this backdrop, we suggest: Hypothesis 3: AL ideals predict AL behavior above the effects of self-evaluations.
Finally, it is also possible that older generation X officers are less prone to projecting their own self-evaluations into their ideal ratings of AL, compared to their younger generation Y colleagues. Longer and more challenging service experience may have provided them with a more secure basis for accepting personal shortcomings and a humbler view on own strengths and limitations -so that their ideals are more a representation of external ideals than self-serving biases. One could also expect a stronger association between AL ideals and behavior when these ideals are higher, as suggested for older generation X officers in Hypothesis 2 -due to a stronger influence on self-regulation mechanisms (e.g., Aquino & Reed 2002). On this basis, we suggest: Hypothesis 4: AL behavior of generation X officers is more influenced by AL ideals, controlling for the effects of self-evaluations, compared with generation Y officers.

Sample
This study included a sample of officer cadets (sub-lieutenants) at the Royal Norwegian Naval Academy (RNoNA), with an average of 1.6 years' previous military service before admission, and a sample of officers (average rank of major), with an average of 18 years' service experience, participating in a one-year executive leadership program at the Norwegian Defence University College (NDUC). A total of 141 officer cadets at the RNoNA and 91 officers at the NDUC were asked to participate. A total of 117 cadets at the RNoNA (a response rate of 83 %) at an average age of 21.9 years (SD = 2.61) and a total of 91 officers at the NDUC (a response rate of 100 %) at an average age of 42.6 years (SD = 5.80) accepted and completed the questionnaires.

Procedure
The Norwegian Navy Staff of Education, the Naval Academy ethics board, and the Norwegian Centre for Research Data (NSD) approved the study. Before data collection, we completed an informed consent procedure as well as an oral briefing. Here, the participants received written information about the main purpose of the study and were informed that participation was voluntary and that they could withdraw at any time. They were also assured that individual results would not be used in performance assessments at the RNoNA or NDUC. Data from the officer cadets at RNoNA was collected after they had finished eight months of extensive team-based leadership training, and data from the officers at NDUC was collected one week before the students completed their education programs. Using a Norwegian translation of the AL questionnaire (ALQ) (Walumbwa et al. 2008), the participants were instructed first to conduct a self-rating of their own leadership performance. In addition, the respondents were given instructions to visualize an ideal military leader, conducting direct leadership, and on this basis rate how closely each of the 16 statements in the ALQ measure represented their own view of ideal military leadership. By assessing already established items quantitatively, compared to for example a qualitative research interview, we expected to activate also implicit leadership schemas through recognition mechanisms -as previously suggested by Rest et al. (1999). Notably, during their training all respondents had rotated in roles as leaders and followers in fixed teams and participated in numerous joint exercises and group sessions. These activities had provided them with thorough knowledge of each other in both leadership roles and as team members. On this basis they were also asked to evaluate the leadership behavior of their peers (i.e., second source rating). The average number of peer raters was approximately seven for all groups. Data was manually transferred to SPSS 22.0 for statistical analyses, and all questionnaires were coded to achieve anonymity.

Measures
AL was measured by a Norwegian translation of the 16-item ALQ, measuring the first-order factors of transparency, self-awareness, balanced processing, and moral perspective (Walumbwa et al. 2008). Respondents rated their own and peer leader behavior statements on a scale of 0 (not at all) to 4 (frequently, if not always) on all of the four authentic leadership dimensions. Sample items included: the supervisor who "listens carefully to different points of view before coming to conclusions" (balanced processing), "makes difficult decisions based on high standards of ethical conduct" (internalized moral perspective), "seeks feedback to improve interactions with others" (awareness), and "encourages everyone to speak their mind" (transparency). Each subscale and the overall 16-item scale demonstrated adequate internal reliability with Cronbach's alpha ranging from 0.82 to 0.95. This measure of AL has also been established as a valid and reliable instrument in other studies (Gardner et al. 2011).

Statistics
One-sample t-tests were used to test if AL represented an ideal form of leadership among Norwegian military officers (H1). We defined a mean score of 3 on AL as the test parameter and considered any score above this value as indicative of the respondents considering this leadership as ideal. The choice of this cut-off was informed by Northouse (2016), who suggests that an average score equivalent to 3 or higher on an AL measure represents a high level. We started by testing the differences in means for the overall AL score. Follow-up tests with Bonferroni corrections (a/m) were then applied when testing for differences in the different dimensions of AL.
Independent t-tests were used to test our second hypothesis that older officers view AL as a more ideal form of leadership than successive generations. Here, older was defined as officers born between 1965 and 1979 (i.e., generation X) and younger as officers born between 1980 and 1997 (i.e., generation Y). The Bonferroni-adjusting approach described above was also used when testing for differences in the different dimensions of AL.
To test if AL ideals predicted AL behavior above the effects of self-evaluations (H3), we used OLS block-wise linear regression. In the first block, AL behavior was regressed on self-evaluated AL and age. AL as an ideal leadership form was then entered as a separate block to determine if, and to what extent, it could explain variations in peer-rated AL behavior. Finally, in block 3, an interaction-term between AL as an ideal and age was added to test our fourth and final hypothesis. Table 1 presents means, standard deviations, and correlations between self-rated and peer-rated AL, and self-rated AL and perceptions of AL as an ideal form of leadership.

Authentic Leadership as an Ideal
As can be seen from Table 1, the respondents on average reported scores on AL as an ideal above our test parameter of 3. Scores on the overall AL were statistically significantly higher than the test parameter (t[210] = 8.331, p < .001, Cohen's d = 0.56). Results from the one-sample t-tests of the individual AL dimensions further showed that, on average, all dimensions were rated as ideal (relevant statistics are summarized in Table 2).
Results from the independent t-tests comparing younger and older officers are shown in Table 3. Older officers perceived overall AL to be a more ideal leadership style (M = 3.32, SD = 0.39) than younger officers (M = 3.15, SD = 0.37), a difference that was statistically significant (M difference = 0.17, t[209] = 3.180, p = .002, Cohen's d = 0.45). Looking at the separate AL dimensions, balanced processing and moral perspective both reached statistical significance, with older officers rating the dimensions as more ideal (see Table 3). Judging by the Cohen's d displayed in Table 3, the differences in means can be said to be medium in magnitude for the overall AL score and the balanced processing subdimension, and large for the moral perspective subcomponent.
The interaction entered in block 3 revealed a statistically significant interaction between AL as an ideal and generation (B = 0.22, SE = 0.10, p = .03, 95% C.I. for B = 0.02; 0.42), explaining an additional 2 % of the variations in AL behavior (∆R 2 = 0.02, p = .03). This interaction is shown in Figure 1. Contrary to our Hypothesis 4, Figure 1 seems to suggest a negative relationship between AL behavior and AL as an ideal   among older officers, and a positive relationship between AL behavior and AL as an ideal among younger officers. Follow-up tests of the two single slopes showed that although statistically significantly different from each other, neither of the slopes was statistically significant by themselves (B = -0.08, SE = 0.09, p = .33, 95% C.I. for B = -0.25; 0.08 for older officers, and B = 0.13, SE = 0.07, p = .06, 95% C.I. for B = -0.01; 0.27 for younger officers).

Discussion
The main aim of the present study was to learn more about the content and relevance of leaders' leadership ideals, operationalized in terms of AL -and its variation among different generations of military officers. The starting point of the study was an observation and a critical reflection on the extensive focus on leadership theories and lectures embedded in leadership development programs in organizational and educational settings. Efforts seemingly set out to develop a more advanced understanding of leadership among the participants, which in turn is meant to "mature" leadership ideals and, subsequently, leadership practices. We are surprised by the limited amount of studies which address the content of leadership ideals among leaders -and how these relate to behavior. An obvious exception is the extensive cross-national studies of House et al. (2014), focusing on content, and the extensive body of research on implicit leadership and follower theories (Junker & van Dick 2014). But these studies rarely focus on how leaders' ideals influence their own performance. Instead they target how followers' leadership ideals influence their leaders. The current study started with an investigation of the content of leaders' leadership ideals. We operationalized this in terms of AL, given the ideal orientation of this increasingly popular theory, and because of a need for further validation of this theory following the retraction of several central studies -raising concerns regarding validity (RetractionWatch 2016). Our findings show, in concert with Hypothesis 1, that AL is highly embedded in leadership ideals among the officers in both age groups. This supports the claim that AL represents a fruitful approach to leadership in a competitive context like the military. Interpretations of our results consider that many of the respondents, particularly among the oldest, have extensive experience from war and international operations. Thus, our results indicate that morals and a human orientation have relevance -also in dangerous and highly competitive environments like armed conflicts. This may contradict Kolditz (2007), who have argued that the relevance of such human orientation is relatively marginal in these settings. It further contradicts scholars like Krantz (2006: 236), stating that "perhaps the sunny, idealized images of leaders who transform through inspiration, passion, and love function as a social defense against the darker more troubling realities of leadership". It is possible that the idealization of moral younger-and older generation officers. One standard deviation below and above the mean on authentic leadership as an ideal is used to plot the slopes.

Peer-Rated Authentic Leadership
Authentic Leadership as an Ideal

Younger Generation Officers
Older Generation Officers -1 SD +1 SD integrity, openness, and inclusion found in the current study is colored by core elements of the Scandinavian culture, and less by the military operational context. According to Hofstede (1984), this culture stands out as highly feminine, emphasizing qualities like cooperation, consensus, and democratic decision-making processes. Conversely, high masculinity, typical of nations like the United States, the United Kingdom, and Germany, emphasizes competition and a high achievement orientation, ideally led by decisive and assertive leaders -seemingly in some contradiction to AL. Thus, future studies of AL as an ideal should include multicultural samples in order to better control for the impact of national culture, as demonstrated for example by Cogin (2012), as well as various work contexts and individual differences like gender, generation, and personality -to better understand the level of universality related to AL as an ideal. Furthermore, the results of this study have revealed generational differences related to leadership ideals -partially in line with Hypothesis 2. The younger officers rated an internalized moral perspective and subsequent moral integrity as less ideal than their older colleagues. This seems to concert with studies showing successive generations to be more selfish and less empathic (Arsenault 2004;Sessa et al. 2007;Twenge & Campbell 2012). On the other hand, it contradicts descriptions of generation X as self-oriented, independent, and outcome-oriented, even at the expense of rules and norms, in contrast to generation Y, who are described as having a strong sense of morals, social responsibility, and patriotism (Allen 2004;Eisner 2005). These inconsistencies between descriptions and findings may support the claim that research on generational differences may lack nuanced characteristics of generations due to a lack of empirical work beyond observation levels (e.g. Cogin 2012). Nevertheless, the differences in ideals may have practical implications. For example, according to Burke (2005), generational differences in terms of morals represent a major cause of conflicts, which in turn may hamper cooperation between older and younger colleagues -and subsequently have severe consequences in dangerous and complex operational settings.
More so, the results show a difference in the valuing of balanced processing of information. Older generation X officers appear to have more focus on inclusion of competing points of view as a basis for decision making. We find this surprising, as the older generation grew up in a time of more hierarchical and authoritarian structures at school and in society at large (Hellevik 2002). Generation X is often portrayed as individualistic and self-reliant, compared to generation Y, which is considered more sociable, more accustomed to being active in family decisions, and likely to expect to be involved in decisions at the workplace (Eisner 2005). It is possible that such experiences with hierarchies and authoritarian structures have made generation X more aware of the importance of more including and open cooperation (whereas the successive generation takes this more for granted). Alternatively, these differences could be seen more as a result of leader experience and maturing, reducing the relevance of generational influences as a theoretical explanation (Parry & Urwin 2011).
We underline that the results show a unified view on transparency and self-awareness. This may indicate that openness as well as the ability to understand how one influences other people are equally important in the leadership of young as well as older officers. It is also worth noting that both groups of officers value the ability to display true feelings and ideas as leaders, contrary to a traditional "stiff upper lip" approach emphasizing distanced and impersonal interaction.
Contrary to our expectations in Hypothesis 3, the results showed no effect of AL ideals on AL behavior after controlling for the influence of self-ratings in the sample as a whole. There is, however, a significant relationship between self-and others' ratings, suggesting that the officers' view of their own leadership performance is partially consistent with their colleagues' view hereof. This relationship could be explained by implicit leadership theories (e.g., Junker & van Dick 2014). Our study shows that AL represents an ideal form of leadership among the officers, which may transform into expectations toward their leader when in a follower position, which in turn influences the leader. On the other hand, it is surprising that ideals have no influence on behavior, above the effects of self-evaluations. This may imply that ideals are little more than a representation of the officers' own self-image, nurturing a self-serving purpose (Dunning, Perie & Story 1991). A relatively high correlation between self-and ideal ratings supports this notion, indicating that external leadership ideals, as a form of externalized ethos, have limited relevance as a source of leadership behavior. From this perspective, the relevance of the extensive teaching of leadership theories in leadership development and MBA programs could be questioned. A better track in such development programs might be to emphasize ILT and prototypes in the form of external expectations as a source of leadership behavior, and to emphasize recognition of these patterns and possibly change prototypes as development aims (e.g., Junker & van Dick 2014).
In line with Hypothesis 4, the results indicate that ideals influence generations differently. The leadership behavior of younger generation Y officers seems to be influenced positively (marginally, not significantly), while no relationship is found among the older generation X officers. We could speculate that this is the result of a more open and curious attitude in the successive generation, which may thus be more open to external influence (Hellevik 2002). It may also be, as suggested by Eisner (2005), that generation Y puts greater emphasis on tradition and community values, indicative of a stronger respect for external ideals, compared to more self-reliant and independent generation X officers, indicative of a reduced willingness to conform to external influences. And it is possible that the older officers to a larger extent have integrated the ideals into their AL practice through their long service experience, limiting the influence of the ideal on their present practice. The young officers may still have a longer way to go in terms of AL, combined with a possibly higher development focus due to their role as inexperienced cadets, and they may thus be more influenced by ideals. This is partially supported by the fact that the older officers show more AL compared to the younger officers, as perceived by their peers.
A difference between the generations in terms of ideal influences may also have practical relevance. In leadership development, this could imply that leadership interventions among older and more experienced leaders ought to reduce the theory focus aiming at stimulating sound leadership ideals (because they already have them, and they are not influenced by them), and that it should instead focus more on behavioral components and training of specific skills like conflict resolution and negotiation (e.g., Day et al. 2014), as theory and stimulation of ideals may be a more fruitful approach among younger leaders.

Limitations and future research
This study utilized a simple cross-sectional design with only one measuring point. According to Lyons and Kuron (2014), studies involving the effect of generations should ideally utilize time-lagged designs, preferably over several years, to control for the impact of other variables. As such, a cross-sectional design comparing two cohorts can only speculate in generational differences as predictor. However, the stability of selection criteria in a military context over time and a similar form of basic training and primary socialization for both groups may still support the use of generation theory as an explanation of ideal differences in the current study.
In order to better generalize the findings, the study could have included cultural differences as control variable, as well as a multinational sample. In this way, the argument that AL is viewed as an ideal in a competitive military setting would have stood stronger. Including different organizations and professions related to dangerous and competitive work contexts could as such have increased the generalizability of the study (Lyons & Kuron 2014).
The participants had to evaluate AL in an isolated manner, without comparisons to alternative leadership styles. This could have led participants to provide overly positive evaluations due to lack of trade-offs. More so, one could question the relevance of AL as operationalization of leadership ideals, representing a root construct, which indicates that many aspects of well-functioning leadership have been left out of the investigation. Here, both a scale that includes more negatively loaded items and perspectives, like laissez-faire leadership or more active bullying behavior (e.g., Einarsen, Aasland & Skogstad 2007), and more constructive approaches like instrumental leadership (Antonakis & House 2014) could have broadened the range and nuanced the content of ideals. In the same vein, the positive wording of the items embedded in the AL measure may have caused a possible activation of impression management influences, and as such common method biases (Podsakoff et al. 2003). Controlling for impression management could be useful in future studies. Our study lacks information about variation in operational experiences among the respondents. By including a measure of combat exposure, future studies could better identify leadership preferences among people who have faced life-threatening situations, representing a better design for investigating the validity of morally oriented leadership under extreme conditions.