Taken to the extreme: Transformational leadership, psychological capital, and follower health outcomes in extreme contexts

ABSTRACT

The purpose of this study is to examine the boundary conditions of transformational leadership, follower psychological capital, and their effects on follower mental health outcomes. Specifically, we utilize archival, multi-wave data from a military sample to examine whether the negative relationship between transformational leadership and adverse follower stress outcomes increases as the context shifts from a relatively safe environment to one in which follower lives are at risk. Additionally, psychological capital, a constellation of personal psychological resources, is also assessed to account for individual buffers against extreme stressors. Findings from the current study suggest that the negative relationship between transformational leadership and follower stress increases significantly when the context shifts to a high-risk, mortality-salient environment.

What is the public significance of this article?— This study advances the idea that positive leadership buffers against follower stress in extreme environments where lives are at risk. The researchers assessed leadership and stress in a U.S. Army organization before and during a combat deployment to Iraq. Positive leadership was more of a buffer against stress in a combat environment than it was in a safe environment.

The first two decades of the 21^st century were bookended by seismic events, the 9–11 terrorist attacks in 2001 and the global pandemic that exploded in 2020, each highlighting the importance of leadership in catastrophic situations. In both instances, organizational scholars reflected a broader interest in not only how leadership and perceptions about leadership change in such conditions (e.g., Bligh et al., 2004), but also how the well-being of followers is impacted by leadership when lives are at risk (e.g., Hu et al., 2020). A consistent theme throughout such literature is that leaders and leadership are especially important in extreme contexts, defined as “an environment where one or more extreme events are occurring or are likely to occur that may exceed the organization’s capacity to prevent and result in an extensive and intolerable magnitude of physical, psychological, or material consequences to – or in close physical or psycho-social proximity to – organization members” (Hannah et al., 2009, p. 898). Whereas a crisis is characterized by threats to high priority individual or organizational goals and bounded by a limited response time (Pearson & Clair, 1998), extreme contexts extend beyond the crisis in scope and intensity, examples of which include natural disasters, global pandemics, major industrial accidents, acts of terrorism, and, as in the case of the current study, warfare. In these extreme contexts, the proximal risk to life and the distal risk to well-being are always of paramount concern.

At the same time, while the leadership-stress linkages in extreme situations have been widely explored in popular press publications (e.g., Goodwin, 2018; Kolditz, 2010), the academic literature contains relatively few empirical tests of leadership in such contexts; most are retrospective qualitative studies (e.g., Dixon et al., 2017). Conversely, the effects of combat and trauma exposure on individuals are better understood and the empirical literature shows that combat veterans face a significantly higher risk for extreme occupational stress outcomes and other health risks, such as PTSD and mild traumatic brain injuries (Hoge & Castro, 2014), along with a range of stressors that may result in less debilitating, subclinical mental health problems (Korte et al., 2016).

Likewise, two highly relevant empirical contributions document how personal resources can ameliorate extreme occupational stress. First, using Conservation of Resources (COR; Hobfoll, 1989) as their theoretical framework, researchers studied U.S. military service members and found that resilience predicted fewer PTSD symptoms upon return from combat when mediated by perceived frequency of social connection (Britt et al., 2021). Second, researchers studied U.S. military service members during combat operations in Iraq and found strong negative relationships between resilience, anxiety, and depression (Schaubroeck et al., 2011). These studies compellingly illustrate how personal and social resources interact to buffer against severe occupational stress reactions, and both studies explicitly point to the potential for leadership to augment personal resources in buffering against occupational stress while in combat.

Thus, our goal in the present paper is to build upon these foundational works via an assessment of the nexus of personal resources, leadership style, and stress in an extreme context. We extend existing literature by focusing on a broader range of individual psychological resources and examining how the experience of occupational stress changes as people transition from a relatively low-risk and stable environment to a high-risk extreme context where mentally and physically challenging work is the norm and lives are persistently at risk. Answering the call by Inceoglu et al. (2018) for greater attention on how leader behaviors affect follower well-being, our intent here is to focus on follower well-being not as a means to other ends (i.e., attaining higher performance or greater retention), but rather as an ends unto itself given the chronic and devastating impact that extreme occupational stress reactions like PTSD can carry. To do so, we employ an archival, multi-wave dataset of U.S. service members with data collected prior to and during a 12-month combat deployment. Importantly, and in line with calls from Hannah et al. (2009) and Arnold (2017), our study offers a valuable boundary condition assessment of transformational leadership in an extreme environment, extending our understanding of leadership and well-being beyond the current boundary, the crisis. Likewise, we adopt the same theoretical framework as Britt et al. (2021) – the COR model – yet extend their theoretical contribution by integrating aspects of attachment theory and the job demands-resources model to account for the role of the leader in bolstering follower well-being as they work in an extreme context.

Theoretical framework

Resource-oriented models of stress such as conservation of resource theory (COR; Hobfoll, 1989; see also Halbesleben et al., 2014) suggest that individuals are motivated to build, retain, and protect psychological resources to be used to preserve well-being. Threats to resources are significant causes stress as core principle of the theory is that losing resources has a more powerful effect than gaining them, so individuals are highly motivated to employ resources as a means of guarding against such threats (Hobfoll et al., 2000). Resilience, or the ability to bounce back from adversity (Masten, 2001), is one such psychological resource, but there are many others including hardiness (Maddi, 2002), self-esteem (Baumeister et al., 2003), and coping skills (Butler et al., 2006). One such prominent psychological resources construct is an array of conceptually related traits (hope, resilience, optimism, and self-efficacy) that form a higher order construct labeled psychological capital (PsyCap; Luthans et al., 2007). PsyCap has been shown to guard against job stress, anxiety, and emotional exhaustion (Avey et al., 2011; Cheung et al., 2011) while promoting follower well-being (Avey et al., 2010). Among soldiers who recently experienced combat, PsyCap has been demonstrated to be negatively associated with mental health issues and behavioral manifestations of negative well-being, such as substance abuse (Krasikova et al., 2015) and depression (Schaubroeck et al., 2011). PsyCap will be particularly beneficial in both low- and high-stress contexts studied here because it is inherently positive and assists followers in overcoming the stressors they face: hope provides a pathway and motivation to succeed; resilience allows one to return to normalcy when facing setbacks; optimism allows one to cognitively reframe challenges as opportunities; and self-efficacy allows one to believe that they can perform tasks necessary to succeed in a given context (Luthans et al., 2007). Because it is a context-neutral psychological resource, we expect PsyCap to buffer against occupational stress in both contexts examined in this study, though as we describe later, its effectiveness at doing so may vary based on the nature of the context.

Hypothesis 1: PsyCap will be negatively related to extreme follower stress outcomes in both low-risk and extreme contexts.

We turn next to our primary focus, the leader-follower dynamic in an extreme context, and there are two theoretical frameworks particularly suited to describing the process followers go through to bolster their psychological resources. First, attachment theory (Bowlby, 1982; Harms, 2011) explains how relationship-oriented leaders can serve as important attachment figures who fulfill three key roles in stressful times: leaders bond with followers, help regulate their emotions, and assist in their adaptation to the context (Popper et al., 2000). Specifically, attachment theory suggests that most individuals readily interact with and explore their environments unless they meet threats that overwhelm their sense of safety, which in turn causes them to seek a “secure base” in relevant others, such as leaders and peers within the workplace (Harms, 2011). This is likely to be particularly true for military service members serving in combat who are unable to readily utilize family members for the social support that proximal attachment figures can provide. In the current study, we chose leaders as the focal attachment figure given our intent to examine the boundary conditions of how leader behavior influences follower well-being.

However, while serving in combat is certainly stressful, recent literature suggests that the ebb and flow of life in combat today is not altogether different than it was in World War I, “months of boredom punctuated by moments of terror” (Hancock & Krueger, 2010; Reinach, 1917). As outlined by Hannah et al. (2009), an individual’s experience of combat will vary widely based on many factors, including mission set, deployment location, and career field within the military. To be clear, not every deployed service member faces the most severe forms of occupational stress, which is echoed in several research reports conducted during combat operations in Iraq and Afghanistan. For example, the Mental Health Advisory Team 5 (US; 2008) notes that about half of approximately 2000 soldiers included in the study experienced traumatic combat events, including being attacked or ambushed, being shot at, or being adjacent to detonating Improvised Explosive Devices (IEDs) at some point during their 9-month combat deployment. Yet, importantly, those same participants also rated a host of chronic but rather mundane non-combat stressors (e.g., repetitive and boring work, not having the proper spare parts, lack of privacy, etc.) as negatively impacting their well-being and, even after controlling for combat experiences, those non-combat stressors were significant predictors of subsequent psychological health problems (Mental Health Advisory Team 5 (US), 2008). Thus, there is a task-oriented perspective that should be accounted for in the leader-follower dynamic within an extreme context. Here, the job demands-resources model (JDR) suggests that leaders can serve as important resources for followers across the range of stressful periods as they can clarify roles and responsibilities, remove hinderances, and provide instructional support to followers (Bakker & Demerouti, 2007) in addition to providing an avenue for obtaining actual physical resources for dealing with challenges (Mumford et al., 2007). Here too, prior research has consistently reported that leaders can serve as an important follower resource to sustain engagement when work demands are high (Breevaart & Bakker, 2018; Bakker & Demerouti, 2014). Thus far, we have intentionally referred to leaders in the abstract to make the broad case that leadership in this context influences follower well-being. We now integrate a theoretical perspective suggesting that a positive approach to leadership – specifically, transformational leadership – provides followers with essential support resources needed in sustained extreme contexts.

Transformational leadership

Despite some theorizing in this domain (e.g., Goodwin, 2018; Hannah et al., 2009; Kolditz, 2010), an open question remains as to the most effective form of leadership that can positively influence follower well-being in an extreme context. Prior research suggests that new genre approaches to leading, such as transformational leadership, are likely to be more beneficial to follower well-being in stressful contexts than others (for a review, see Arnold, 2017), and it is therefore plausible that such an approach would be particularly salient in an extreme context. Indeed, studies of follower stress in combat point out that transformational leaders “model ideal behavior, emphasize a collective sense of purpose, and stimulate followers to think about various ways they can improve group performance. They also exhibit individualized consideration, which in this context pertains to understanding which followers may not be responding in a resilient manner to traumatic exposures, and helping them reframe traumatizing incidents by emphasizing to the individual that he or she is not alone in his or her suffering (to lower loss appraisal), that coworkers are aiding him or her in maintaining a safe environment, and stimulating his or her thinking about duties in terms of challenge and potential for both individual and collective benefits” (Schaubroeck et al., 2011, p. 33). Thus, followers operating in extreme contexts will be better served with positive, relationship-oriented leadership that followers can call upon as a resource in times of significant stress.

Our narrowing to transformational leadership is in line with previous military research describing how supportive leadership within extreme contexts positively impacts the general mental health of followers exposed to combat stress, whereas destructive leadership exacerbates mental health problems, such as PTSD (McGurk et al., 2014). This finding is consistent with meta-analytic results showing a clear association between transformational leadership behaviors and lower stress and burnout in followers (Harms et al., 2017). Especially relevant to the current study, other research found that individuals seek out transformational leaders when faced with a significant challenge and during times of great uncertainty because they offer a positive vision of the future (Tepper et al., 2018). When required to perform highly stressful tasks, followers report lower negative affect and decreased threat appraisals when following transformational leaders (Lyons & Schneider, 2009), along with higher evaluations of organizational cohesiveness and commitment, self-efficacy, and appraisals of performance (Pillai & Williams, 2004), findings due in large part to followers’ feeling of connectedness with their leader. Finally, research employing a military sample showed that transformational leaders assist followers in cognitively reframing stressful tasks so that they see them as challenges worth undertaking rather than viewing them as hinderances, and this finding was attributed to follower perceptions of the transformational leader’s charisma (LePine et al., 2016).

Conversely, several studies have documented that in settings where risks were present but non-life-threatening and the stakes were relatively low, transformational leadership had a positive, but more muted effect on employee stress (Liu et al., 2012), which was attributed to followers not having their personal resources overtaxed; in other words, followers relied on transformational leader less when the stakes were low, but leadership was still beneficial. Thus, we suspect that transformational leadership will benefit follower well-being in both low- and high-risk contexts, but more so in high-risk contexts where followers are most vulnerable because they will actively seek out the additional resources provided by transformational leaders.

Hypothesis 2. Transformational leadership will be negatively related to follower stress outcomes in both a low-risk and an extreme context, but this negative relationship will be stronger in an extreme context.

One of the primary goals of the present study is to understand the degree to which a transition from low- to high-risk contexts might cause a shift away from personal resources to those provided by others, namely transformational leaders. Integrating our review of PsyCap with transformational leadership, research indicates that despite the broad positive effects associated with PsyCap, not all individuals have inherently high levels of these characteristics. Consistent with COR theory, there is evidence that these psychological resources can be degraded by stressful situational factors. For instance, stressors such as unemployment uncertainty and significant work-family stress can reduce levels of PsyCap and, in turn, exacerbate general stress and erode meaning in life (Epitropaki, 2013; Liu et al., 2012). For many individuals, personal psychological resources such as PsyCap may be sufficient for dealing with low- to moderate-stress situations found in most work contexts, rendering the social support of transformational leaders and others less important for protecting well-being. However, under the extreme levels of stress associated with an active combat zone, when individuals are separated from other familiar sources of social support, even individuals with higher levels of PsyCap will likely rely more heavily on the leaders for support, encouragement, and direction. Further, this may be especially true for individuals working in organizations that already sustained losses, such as combat casualties as seen current study’s sample, given that this kind of trauma will likely drain their personal resources (Britt et al., 2021). Thus, we offer our final hypothesis:

Hypothesis 3: The negative relationship between transformational leadership and follower stress outcomes will be stronger for individuals with lower levels of PsyCap, but this moderation effect will be attenuated in an extreme context.

To test these hypotheses, we present an archival, multi-wave field study of U.S. military service members before and during exposure to an extreme context, combat operations in Iraq.

Method

Procedure and sample

As part of a larger field study of service members working within a combat zone, the lead author surveyed U.S. Army soldiers (T₁ = 423, T₂ = 340) at two time points approximately 180 days apart, once at home station and once amid serving in an extreme combat environment for over 90 days continuous days. It is important to note that the soldiers in the sample served in an all-male unit engaged in continuous “in sector” combat operations, with some members of the unit having been wounded or killed approximately 45 days prior to T₂, thereby meeting published criteria for service within an extreme environment (as compared to serving in a relatively safe “Green Zone” headquarters area; for additional details, see Hannah et al., 2009; Harms & Lester, 2012). The study was approved by the Institutional Review Boards of the United States Military Academy (DOD A20145) and the North Atlantic Regional Medical Command (DOD A10013), and all individuals providing data gave informed consent. Participation in this study was voluntary and no incentives or other forms of reward were offered. All study measures were administered at both Time 1 (T₁) and Time 2 (T₂) in their original language (English).

Data were screened for careless responding using the individual response variance index on the transformational leadership measure (IRV; Dunn et al., 2018). Specifically, soldiers indicating the same response to all 24 leadership items (IRV = 0) were flagged for careless responding. Missing values were then imputed using person mean values for each scale to preserve sample size and maximize statistical power. After screening, the samples sizes used for analysis were 371 and 317 at T₁ and T₂, respectively. Thus, careless responding rates of 12.3% and 6.8% were observed at T₁ and T₂, respectively, which is consistent with rates of carelessness reported in other research (see DeSimone & Harms, 2018; Meade & Craig, 2012). Additionally, certain hypothesis testing necessitated the creation of a combined dataset consisting of soldiers who completed all focal measures at both T₁ and T₂ (N = 227 before screening, N = 185 after screening).

Measures

Transformational leadership

Followers evaluated their leader’s transformational leadership using the 24-item Multifactor Leadership Questionnaire (MLQ; Bass & Avolio, 1989). Items were rated on a scale ranging from 0 (not at all) to 4 (frequently, if not always). Sample items include, “My leader … talks about his/her most important values and beliefs” (idealized influence), “ … talks optimistically about the future” (inspirational motivation). Coefficient alpha at time 1 (α_T1) was .94 with an average inter-item correlation ($\stackrel{‾}{{\mathit{r}}_{\mathit{i}{\mathit{j}}_{\mathit{T}1}}}$) of .51. α_T2 was .96, and $\stackrel{‾}{{\mathit{r}}_{\mathit{i}{\mathit{j}}_{\mathit{T}2}}}$ was .60.

Psychological capital

Psychological capital was measured using the 12-item version of the Psychological Capital Questionnaire (Luthans et al., 2008). Items were rated on a scale ranging from 1 (strongly disagree) to 6 (strongly agree). Sample items include, “I can think of many ways to reach my current personal goals” and “I always look on the bright side of things regarding my job.” α_T1 was .88, $\stackrel{‾}{{\mathit{r}}_{\mathit{i}{\mathit{j}}_{\mathit{T}1}}}$ was .38, .α_T2 was .89, and $\stackrel{‾}{{\mathit{r}}_{\mathit{i}{\mathit{j}}_{\mathit{T}2}}}$ was .40.

Occupational stress symptoms measure

Occupational stress symptoms were measured using the 11-item Mississippi Scale for Combat-Related Posttraumatic Stress Disorder (Keane et al., 1988). Items were rated on a scale from 1 (never) to 5 (extremely frequent). Sample items include, “I lose my cool and explode over minor everyday things” and “I feel like I cannot go on.” α_T1 was .79, $\stackrel{‾}{{\mathit{r}}_{\mathit{i}{\mathit{j}}_{\mathit{T}1}}}$ was .27, .α_T2 was .70, and $\stackrel{‾}{{\mathit{r}}_{\mathit{i}{\mathit{j}}_{\mathit{T}2}}}$ was .20.

Data analyses

Analyses were conducted in Microsoft Excel and SPSS (version 26). Excel was used for data cleaning, centering variables and computing interaction terms, and conducting the Fisher z transformations. SPSS was used for correlation, regression, and estimating internal consistency. Templates provided by Dawson (2022) were used to create Figure 1.

Figure 1.

The relationship between transformational leadership and post-traumatic stress disorder symptomology for various levels of psychological capital in a non-extreme situation (Time 1) and an extreme situation (Time 2).

Results

Hypothesis 1 predicted a negative relationship between psychological capital and PTSD in both ordinary and extreme contexts. As indicated in Table 1, the correlation between psychological capital and PTSD was −.38 (p < .05) at T₁ and −.33 (p < .05) at T₂. Additionally, as indicated by Model 1 in Table 2, the statistically significant, negative relationship between psychological capital and PTSD was present (T₁: B = −0.32, p < .05; T₂: B = −0.17, p < .05) even while holding constant the effects of transformational leadership. Taken together, these results provide support for Hypothesis 1.

Table 1.

Correlations between the focal variables and interaction term at Time 1 and Time 2.

	Mean	sd	Transformational Leadership	Psychological Capital	Interaction	PTSD
Mean			2.66	4.35	0.28	2.20
sd			0.86	0.84	0.79	0.74
Transformational Leadership	2.25	1.05	1	.39*	−.12*	−.17*
Psychological Capital	4.18	0.95	.27*	1	−.15*	−.38*
Interaction	0.27	1.14	−.07	−.21*	1	.17*
PTSD	2.11	0.57	−.24*	−.33*	.04	1

*p < .05.

sd = standard deviation, PTSD = Post-traumatic stress disorder symptomology. Time 1 correlations are above the diagonal (means and standard deviations on the top rows), N = 371. Time 2 correlations are below the diagonal (means and standard deviations on the left two columns), N = 317.

Table 2.

Hierarchical regression models predicting PTSD using transformational leadership, psychological capital, and their interaction.

	T1, N = 371		T2, N = 317		Combined Sample, N = 370 (185 at T1, 185 at T2)
Predictor	Model 1	Model 2	Model 1	Model 2	Model 1	Model 2	Model 3
Intercept	2.20*	2.18*	2.11*	2.10*	2.17*	2.17*	2.15*
Transformational Leadership (TL)	−0.03	−0.02	−0.09*	−0.09*	−0.07	−0.07	−0.07
Psychological Capital (PC)	−0.32*	−0.31*	−0.17*	−0.17*	−0.23*	−0.23*	−0.24*
TLxPC		0.11*		−0.02		−0.02	0.22*
TLxPCxTime							−0.32*
R2	.14*	.16*	.13*	.13*	.13*	.13*	.18*
ΔR2		.02*		.00		.00	.05*

*p < .05

Regression weights are unstandardized. Time was dummy coded (T₁ = 0, T₂ = 1).

Hypotheses 2 predicted that the correlation between transformational leadership and PTSD would be weaker in the home station environment and stronger in the extreme context. Correlation matrices for the focal variables and interaction term are provided in Table 1. The relationship between transformational leadership and PTSD symptomology was statistically significant and negative at both T₁ (r = −.17) and T₂ (r = −.24), with the Time 2 relationship being stronger, as predicted.¹ These results indicate transformational leadership is negatively and statistically correlated with PTSD in both home station and extreme contexts, but the relationship is stronger in the extreme context, providing some evidence in support of Hypotheses 2.²

Hypothesis 3 examined the three-way interaction between transformational leadership, psychological capital, and context for predicting PTSD, specifically predicting that the interaction between transformational leadership and psychological capital would be attenuated in an extreme context. To test this hypothesis, we first examined the interactive effect of transformational leadership and psychological capital at T₁ and T₂ separately using hierarchical regression models. Results are displayed in Table 2 (T₁ and T₂ columns) and depicted in Figure 1. Of note, the negative relationship between transformational leadership and PTSD was statistically significant in the extreme context (T₂), but not in the home station environment (T₁), lending further support to Hypotheses 1 and 2. Consistent with Hypothesis 3, results indicate a small but statistically significant interaction effect in the home station environment, but no statistically significant interaction effect in the extreme context. In other words, low levels of PsyCap were associated with a stronger relationship between transformational leadership and PTSD symptomology in the extreme context, though the interaction effect was small (B = 0.10, p < .05). On the other hand, the relationship between transformational leadership and PTSD symptomology was not statistically influenced by levels of PsyCap in the home station environment (B = −0.02, ns).

To statistically explore the three-way interaction between transformational leadership, PsyCap, and environmental context, we used the combined sample to examine the interaction between time (i.e., context) and the interaction between transformational leadership and psychological capital in individuals who provided data at both T₁ and T₂. Results are displayed in Table 2 (“Combined Sample” columns) but should be interpreted with some caution as soldiers are each represented twice in this sample (once for T₁ and again for T₂), which violates the independence assumption of regression and independence of residuals. Consequently, we are including it as a purely exploratory supplemental analysis to further examine Hypothesis 3. Further, we do not recommend interpreting any regression weights other than the three-way interaction, and we adopted a more conservative Type I error cutoff (α = .01) for the three-way interaction effect to account for the inflated sample size.³ Results indicate a statistical difference in the interaction between transformational leadership and psychological capital across contexts (B = −0.32, ΔR² = .05, p < .01), demonstrating statistical support for Hypothesis 3.

In summary, the results indicate that psychological capital and transformational leadership are both negatively related to PTSD in both contexts, with a stronger negative relationship between transformational leadership and PTSD at T₂. Transformational leadership and psychological capital interact to predict PTSD at T₁, but not at T₂. Finally, when considering only soldiers present in both contexts, results indicate a statistically stronger interaction effect at T₁ (home station environment) than T₂ (in extremis environment).

Discussion

Although the argument has been made for the universality of the positive effects of transformational leadership (e.g., Bass, 1997), there are others who have asserted that truly great leadership can only really be demonstrated in times of crisis (Bryman, 1993). The present study sought to extend beyond this claim and boundary by integrating and building on several theoretical frameworks – conservation of resources theory, attachment theory, and the job demands-resources model – and utilizing a unique sample to evaluate whether the threat level of environments might strengthen the negative relationship between transformational leadership and an extreme follower occupational health outcome. Specifically, we argued that under relatively ordinary, low-threat circumstances, the well-being of individuals will be largely accounted for by their personal psychological resources such as PsyCap. That is, under ordinary conditions, individuals with high levels of psychological resources rely less on emotional and social support from their leaders. However, when under extreme, high-threat circumstances, individuals are more likely to seek security, resources, and reassurances from their leaders, regardless of their personal resources. Our results showed exactly this pattern of results: Transformational leadership mattered most when followers were in high-threat, mortality-salient environments.

Theoretical implications

Three fundamental theoretical frameworks served as the foundation for our study and guided our assessment of the nexus of personal resources, leadership style, and stress in an extreme context. In line with Britt et al. (2021), we relied on COR theory to establish that the resilience serves as a buffer against PTSD, but we extended their theoretical contribution by including the higher-order constellation of PsyCap resources, which adds hope, optimism, and self-efficacy. We also called upon attachment theory and the JDR model to establish that a transformational leadership style known for incorporating aspects of both relationship-oriented and task-oriented behaviors would buffer against stress in low- and high-risk contexts, while also providing evidence that they theory has greater utility in the high-risk context.

The present results add to the increasingly complex story surrounding the relationship between contextualized leadership and follower well-being. Specifically, this study suggests that the threat level of environments can be a key moderator of this relationship. This potentially helps to explain why prior meta-analytic reviews of the relationship between leadership behaviors and follower well-being have demonstrated a fairly high degree of effect size heterogeneity (Harms et al., 2017).

Transformational leadership specifically has recently received a great deal of scrutiny, with some claiming that this and other new genre approaches to leadership are overly positive, too ideologic, and overstate the importance of a leader within an organization (Alvesson & Einola, 2019; Credé et al., 2019). While the theory underpinning transformational leadership may have its limitations, the current study nonetheless shows that the theory still has significant value in explaining how followers cope in particularly harsh environments. That said, the present study provided evidence that suggests that the assumption that the effects of transformational leadership are largely uniform across individuals and contexts (Bass, 1997) is not supported. Rather, the context of leadership, whether in terms of the surrounding environment or in terms of the characteristics of the followers, matters.

Finally, supporting calls from Arnold (2017) to explore the boundary conditions of transformational leadership and employee well-being, the current study extended our understanding of the utility of transformational leadership beyond the current contextual boundary, the crisis. This is a particularly important theoretical contribution as the number of people working in extreme contexts is non-trivial and hard to ignore; over 9 million people in the U.S. alone serve in the military, public safety, and public health sectors (U.S. Bureau of Labor Statistics, 2021; U.S. Department of Defense, 2021). The present study directly addresses and extends this call by demonstrating that PsyCap can not only serve as supplementary factor to leadership in the development of PTSD, but can also serve as a more readily-available alternative resource to draw upon in lower stakes situations.

Practical implications

The wars in Afghanistan and Iraq served as reminders that extreme occupational stress and trauma can lead to chronic, debilitating stress reactions that include anxiety, depression, PTSD, and suicidality in those who serve in extreme contexts, especially for those who do so repeatedly and for extended periods of time (Mental Health Advisory Team 5 (US), 2008). While organizations such as the U.S. Army continue to expend significant resources on developing and bolstering personal protective factors (such as resilience), less focus has been afforded to educating and training leaders on how to lead in ways to ameliorate extreme occupational stress (Cornum & Lester, 2012). Thus, organizations habitually working in extreme contexts should consider augmenting their personal resource development practices with leadership development focused on relationship- and task-oriented styles, such as transformational leadership. On that note, however, our study suggests that under low-risk/low-stress conditions, most followers likely have the personal resources to address and overcome most occupational stressors. Therefore, leadership training interventions may want to clarify that leaders do not have to be “on” at all times, but instead that it is essential for leaders to provide support when subordinates may feel overwhelmed in extreme contexts.

Beyond leadership, the present results also suggest that leaders may want to take a more long-term view of how they can be supportive of their followers. For example, rather than trying to play the role of the heroic leader when situations become chaotic and resources are stretched, it may be more prudent to plan for extreme contexts, train for them, and when possible, work toward preventing them altogether. Similarly, the present study suggests that investing in subordinates to build their own personal and psychological resources (i.e., PsyCap) for coping with threats through training could prove beneficial, particularly under ordinary (i.e., non-extreme) circumstances. Specifically, we would argue that, just as a supportive parent raises their children to be self-sufficient and to leave the home when they reach adulthood, one marker of an effective leader is the desire to instill in their followers the skills and self-confidence needed to bolster their self-reliance in stressful situations.

Limitations

No study is perfect, and the present archival study is no exception. One potential criticism of the present study is that it utilized transformational leadership to assess positive leadership behaviors. However, transformational leadership remains a widely used model for understanding the relationship between leadership, stress, and well-being (Harms et al., 2017). Further, we believe that transformational leadership measures still broadly reflect positive, value-laden, and follower-centered leadership that can prove beneficial for helping followers cope with extreme stress. Nonetheless, we believe that replicating and further exploring these effects with other models of leadership and different methods of data collection would prove valuable as part of a broader effort to advance theorizing and practice pertaining to the relationship between leadership behaviors and follower well-being.

A second limitation of this study is the potential for common method bias (CMB; Podsakoff et al., 2003). The separation of measures by time may mitigate common time variance. However, the reliance on self-report measurement and survey methodology for all variables included in our model yield the potential for other forms of CMB. As this was an archival study, we were unable to employ any methodological remedies for CMB, so we instead caution readers to interpret these results with the caveat that variance due to a common source and common measurement method may have influenced the results of this study.

Another potential criticism would be that the present results are being tested in a unique sample (the U.S. Army) and in a unique setting (active military deployment and combat). Though the number of people working in extreme contexts is on the rise, we would agree (and hope it is the case) that most workers will never face such risk in their workplace. However, the military has a long history of providing large-scale samples and exceptional studies from which organizational scholars have derived important truths about the nature of work (e.g., Project A; Campbell, 1990). Further, we offer that this study is suggestive that other high-fidelity occupations, such as first-responders and healthcare professionals, likely work in contexts where leadership may be particularly impactful due to the prevalence of mortality risk. Future research may want to pay particular attention to this effect as a potential confound or moderator.

Conclusion

Leadership matters. And when it comes to the relationship between leadership and well-being, leadership matters most when the psychological resources of followers are insufficient to deal with the challenges and dangers they face. These results not only enable a better understanding of leadership in extreme conditions, but we hope they will facilitate future research examining when, why, and how leaders can facilitate follower well-being in various contexts.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

The data that support the findings of this study are available from the corresponding author, Paul B. Lester, upon reasonable request.