Expectations of Leaders’ Mental Health

Abstract

Understanding the causes and consequences of varying mental health experiences in the workplace has gained significant research attention, yet little is known about the assumptions people hold about mental health at work, especially with regard to the expectations people may have of their leaders’ mental health. Given people tend to romanticize organizational leaders and have expectations regarding prototypical leader attributes, we consider whether people also hold expectations of leaders’ mental health. Drawing on implicit leadership theories, we propose that people will expect leaders experience better mental health compared to those occupying other organizational roles (e.g., subordinates). Using mixed methods, Study 1 (n = 85) showed that people expect that those in leadership roles enjoy higher well-being and experience less mental illness than those in non-leadership roles. Using vignettes in which an employee’s health was manipulated, Study 2 (n = 200) demonstrated that mental illness is incongruent with leadership prototypes. Using vignettes in which organizational role was manipulated, Study 3 (n = 104) showed that compared to subordinates, leaders are perceived as having more job resources and demands, but people expect that it is leaders’ greater access to organizational resources that facilitates their well-being and inhibits mental illness. These findings extend the occupational mental health and leadership literatures by identifying a novel attribute upon which leaders are evaluated. We conclude by considering the consequences of leader mental health expectations for organizational decision-makers, leaders, and employees aspiring to lead.

Understanding the important role of mental health in the workplace is by no means new; Chinoy (1955) and Kornhauser (1965) had already identified mental health issues and their consequences in the auto industry decades ago. Since then, research on the role, causes, and outcomes of mental health within the context of work has expanded substantially: Researchers have established that workplace characteristics (e.g., leadership, culture) affect employees’ mental health, mental health affects workplace behaviors (e.g., withdrawal, performance, interpersonal interactions), and unique mental health experiences emerge in the workplace (e.g., occupational stress, burnout) (see Kelloway et al., 2023 for a recent review). Such extensive research attention is well deserved; the prevalence of mental illness has been steadily increasing in recent decades, particularly within the working population (Mind Share Partners, 2021), even before the onset of the Covid-19 pandemic (Santomauro et al., 2021). This has resulted in several calls for action: In 2022, the United States (U.S.) Surgeon General released its first-ever report outlining the critical role of organizations in promoting and protecting the mental health of their employees, while the World Health Organization (WHO) recently released a set of recommendations to improve mental health specifically within the workplace (Staglin, 2023). Though research advancements and organizational initiatives broadly acknowledge the importance of mental health for workplace success, they do not consider whether all employees are viewed equally when it comes to their experience with their mental health.

Indeed, within the workplace context, we know little about the assumptions and expectations people hold regarding who may be considered most (or least) vulnerable to mental health challenges. This is an important limitation given (1) some people may be excluded from organizational accommodation programs designed to support mental health if they are expected to experience good mental health and (2) certain people may be excluded from work opportunities if they are expected to experience poor mental health. Thus, understanding who is deemed most or least vulnerable to mental health challenges at work may highlight underlying inequalities in the workplace while pointing to opportunities for intervention.

Though existing research has demonstrated that several individual difference factors (e.g., gender), personal histories (e.g., exposure to trauma), and behaviors (e.g., social withdrawal) inform who is deemed most vulnerable to mental health struggles (Furnham & Telford, 2012; Kelloway et al., 2023), work-related factors have not been explored to the same extent. We suggest one’s work role (i.e., the formal duties and responsibilities assigned by the organization) may be a particularly relevant characteristic that people draw upon to inform their mental health expectations of others. This is because work roles reflect a common social category by which people differentiate between others and differentiate others from themselves (Ashforth, 2001; Ashforth et al., 2000; Ashforth & Mael, 1989), and different work roles engender different prototypical expectations (Lord et al., 1984; Sy, 2010). One work role that may particularly engender unique mental health expectations is the leadership role.

Extent empirical evidence demonstrates that people hold unique expectations of what it means to be an organizational leader. Leader prototypes reflect people’s underlying beliefs regarding the attributes (e.g., intelligent, sensitive) and behaviors (e.g., charismatic, agentic) considered most typical of organizational leaders (Epitropaki et al., 2013; Lord et al., 1984). Research has established that people generally hold overarching, categorical expectations of their leaders (Epitropaki & Martin, 2004); for example, regardless of context, leaders are expected to be charismatic and sensitive (House et al., 2002). However, people also have context-specific leader prototypes (Hanges et al., 2000); for example, people expect leaders to be male (Braun et al., 2017), white (Petsko & Rosette, 2023), and happy (Trichas et al., 2017). We posit that people may also hold unique expectations regarding leaders’ mental health.

The purpose of this research is to examine whether people hold expectations regarding organizational leaders’ mental health. Informed by implicit leadership theories (ILTs) (Epitropaki et al., 2013), we predict that leadership roles will be associated with expectations of good mental health, because the characteristics and underlying assumptions of good mental health are more consistent with existing leader prototypes. Informed by the connectionist approach to leader prototypes (Foti et al., 2008), we further predict people will use contextual information regarding work roles (i.e., presence of organizational resources and demands) to inform their expectations, and more specifically, we hypothesize that people will view leaders as having greater access to organizational resources, resulting in expectations of better mental health.

We test our predictions across three experimental studies. In Study 1, we consider whether people associate different work roles (i.e., leader, subordinate, or employee) with different mental health experiences and establish that leaders are expected to experience higher well-being and less mental illness (depression and anxiety) than those in non-leadership roles. In Study 2, we test whether different health experiences (i.e., mental illness diagnosis, physical illness diagnosis, no health diagnosis) affect expectations of leadership suitability and establish that mental illness (but not physical illness) is viewed as incongruent with leader prototypes. In Study 3, we consider whether work role characteristics serve as contextual cues to inform leader mental health predictions, and we find that leadership roles are perceived as having greater access to organizational resources, which result in positive expectations of their mental health.

Our research offers three theoretical contributions. First, we contribute to research on mental health at work by demonstrating that people hold mental health expectations of others based on their organizational role. Second, we contribute to understanding people’s leader prototypes (Lord et al., 2020) by identifying a novel attribute (i.e., mental health) upon which leaders are perceived and evaluated. This contributes to the emerging connectionist perspective to ILTs (Foti et al., 2008), which suggest that when focused on salient contextual features, people can develop more domain-specific leader prototypes. Third, we contribute to the growing literature on leaders’ mental health (Barling & Cloutier, 2017); research has established that transitioning into leadership roles affects incumbents’ mental health (e.g., Fletcher & French, 2021), but no research has considered whether people hold expectations of the mental health leaders ought to experience, which may be relevant to whether employees emerge into leadership roles in the first instance. Results from this research also offer practical implications. Despite organizational efforts to destigmatize mental illness, covert discrimination in the workplace persists (Hastuti & Timming, 2021). One way in which mental illness stigmatization may manifest in the workplace is via individuals’ implicit categorization of those who are mentally healthy as leaders and those who experience mental illness as non-leaders. As such, we open up a discussion on how these social categorizations could engender organizational consequences for leaders, employees, and organizations.

Theoretical Background

We begin by conceptualizing mental health and briefly review the literature on lay perceptions of mental health, identifying both generally held beliefs about mental health and how mental health is viewed within the context of work. We then review the literature on ILTs and the connectionist approach to leadership theories and focus on explaining lay beliefs about leaders and the context-specific beliefs that have been identified. Next, we consider the overlap between beliefs about mental health and implicit beliefs about leaders to inform our hypotheses, suggesting that the characteristics of good mental health (i.e., well-being) are more consistent with leader prototypes, while the characteristics and assumptions of poor mental health (i.e., depression, anxiety) are incongruent with leader prototypes. Finally, we consider whether organizational characteristics associated with different work roles (i.e., resources and demands) may serve as additional contextual variables that people consider when informing their expectations of leaders’ mental health.

Mental Health: Conceptualization and Beliefs

Though there are several frameworks for conceptualizing mental health (e.g., see Kelloway et al., 2023), we conceptualize mental health as a multidimensional construct comprising positive (i.e., well-being) and negative (i.e., mental illness) aspects of mental health (Westerhof & Keyes, 2010). According to Westerhof and Keyes (2010), “mental health” reflects an optimal level of functioning and includes emotional (i.e., positive emotions), psychological (i.e., positive self-evaluations), and social (i.e., positive relations with others) dimensions of well-being. “Mental illness” is defined as the presence of sub- or clinical symptoms of internalized (i.e., negative feelings and thoughts experienced inwards, e.g., depression, anxiety) or externalized (e.g., negative thoughts and feelings exhibited outwards, e.g., substance abuse and aggression) disorders. This conceptualization implies that “mental health” and “mental illness” range along their own continuum and should therefore be studied separately (Keyes, 2005; 2006). To maintain clarity, we use “well-being” to refer to the positive dimension of mental health, “mental illness” to refer to the negative dimension of mental health, and “mental health” as the collective experience of both positive and negative health dimensions.

In addition to evidence-informed conceptualizations of mental health, lay conceptions of mental health have also been explored within the field of psychiatry for several decades (Star, 1955). There are three related but distinct approaches to understanding lay conceptions of mental health: (1) public attitudes, (2) lay theories, and (3) mental health literacy (Furnham & Telford, 2012). First, researchers have explored people’s general attitudes toward mental health experiences, including their beliefs about what people with mental illness (e.g., depression, schizophrenia) are like and how they ought to be treated. Insights from this approach have contributed to our understanding of mental illness stigma, pointing to mental illness being viewed with adverse emotional (fear, pity, and anger) and behavioral (attempts to socially distance) reactions (Schomerus et al., 2012; Walsh & Foster, 2021). Second, researchers have considered people’s lay theories about mental health, which assesses people’s underlying beliefs related to the causes, prevalence, consequences, and treatment of mental illness. This perspective contributes to our understanding of how people view mental health in general and why people stigmatize mental illness—particularly if they view disorders as emerging from controllable factors, as resulting in dangerous behavioral patterns, or as untreatable (Furnham & Telford, 2012). Third, researchers have explored public mental health literacy, which documents public knowledge, understanding, and recognition of mental disorders (Furnham & Telford, 2012).

Public attitudes and lay theories about mental health have been explored to a much lesser extent within the context of work but do show some consistent findings. For example, research demonstrates that mental illness is stigmatized in the workplace; hiring managers doubt the capabilities of employees with mental illness and consequently avoid hiring them (Brouwers, 2020; Janssens et al., 2021), while those suffering from mental illness believe that others look down upon them as a function of their illness, resulting in self-stigmatization and organizational withdrawal (Elraz, 2018). Beliefs about well-being in the workplace are generally less explored; however, employees exhibiting psychological capital (i.e., efficacy, hope, optimism, and resilience) are generally evaluated more positively, as organizational citizens and effective team members (see Newman et al., 2014 for a review).

In addition to generalized beliefs about mental health in the workplace (i.e., “mental illness results in poor employee capability”; “employees experiencing well-being are better team players”), we suggest people may believe that the degree to which others experience well-being and mental illness differs as a function of the work role. Indeed, extensive research has demonstrated that people develop different prototypical expectations of the attributes, characteristics, and behaviors of others based on their organizational role (e.g., leader, follower; Lord et al., 2020; Sy, 2010).

In the current study, we focus on understanding expectations of leaders’ mental health, concentrating on expectations related to their psychological well-being, depression, and anxiety. Psychological well-being reflects positive evaluations of oneself, one’s life, and one’s achievements and is composed of six factors, including experiencing high degrees of autonomy (i.e., ability to regulate one’s own behavior independent of social pressures), environmental mastery (i.e., ability to manage one’s environment and everyday affairs), personal growth (i.e., desire to continuously develop), positive relations with others (i.e., ability to positively engage with others and form meaningful relationships), purpose in life (i.e., a zest for life, strong goal orientation), and self-acceptance (i.e., positive self-evaluations) (Ryff & Keyes, 1995). Depression is characterized by two overarching symptoms: depressive affect (i.e., feelings of intense sadness, helplessness, and hopelessness) and anhedonia (i.e., feelings of disinterest and displeasure) (American Psychiatric Association, 2013). Anxiety is characterized by persistent and excessive worry and includes symptoms such as feelings of self-doubt and concern that something will go wrong (American Psychiatric Association, 2013).

We focus on these three indicators of mental health for three reasons. First, these indicators tap into both mental health continua (i.e., well-being and mental illness) and therefore provide a more complete assessment of mental health beliefs compared to if we just focused on one dimension (e.g., beliefs about leaders’ mental illness). Second, we focus on assessing perceptions of depression and anxiety as mental illness indicators specifically because they reflect the most common mental illnesses experienced in the workplace (Mind Share Partners, 2021), as a result of which they would be familiar to our sample of participants, and are therefore constructs upon which expectations/schemas may already exist. In addition, depression and anxiety are internalized mental illnesses, meaning negative feelings and thoughts are directed inwards and are therefore not always obvious to observers. Thus, people may be more likely to rely on their implicit theories surrounding these illnesses when developing expectations about the degree to which work roles experience these symptoms. Third, the characteristics associated with psychological well-being, depression, and anxiety range in their congruency with general leader prototypes and may therefore inform expectations, a topic we turn to next.

Leader Prototypes

As is the case with lay theories about mental health, people also hold implicit leadership theories (ILTs; Lord et al., 1984; 2020). Specifically, people possess cognitive schemas about organizational leaders (i.e., ILTs) that guide which attributes (e.g., traits, abilities, characteristics) and behaviors are considered most prototypical of leaders (Epitropaki & Martin, 2004; Junker & van Dick, 2014; Lord & Maher, 1991). Leadership schemas develop through early interactions with parents (Keller, 1999; 2003) and later through interactions with other leaders (e.g., coaches, teachers, supervisors). Once developed, ILTs are relatively impervious to disconfirming evidence from interactions with individual leaders (Epitropaki & Martin, 2005) implying their robustness through time and space.

Although scholars originally theorized that people held overarching ILTs that broadly categorized targets as leader versus non-leader (Eden & Leviatan, 1975; Rush et al., 1977), later research informed by categorization theory (Epitropaki & Martin, 2004; Lord et al., 1984; Offermann et al., 1994) demonstrated that leader schemas are comprised of multi-categorical prototypes; that is, observers categorize targets’ attributes along several dimensions and then consider whether their attributes are consistent with those of leaders. Across time (Epitropaki & Martin, 2004; Offermann et al., 1994; Offermann & Coats, 2018), research has shown that prototypical leadership attributes include sensitivity, intelligence, dedication, and dynamism. The more targets are perceived to match these attributes, the more likely they are to be considered a leader, with meaningful implications. Leaders perceived to match these prototypes have better leader–follower relationships (Epitropaki & Martin, 2005), receive more positive leadership evaluations (Porr & Fields, 2006; Hansbrough et al., 2021), and engender higher follower workplace attitudes (De Luque et al., 2008) and performance (Lord & Maher, 1991; Veestraeten et al., 2021). Taken together, people hold leader prototypes that serve as a top–down influence to determine who is viewed as a leader and who is not.

Extending information gathered from general ILTs, some researchers have adopted a connectionist approach to investigating ILTs (Foti et al., 2008). The connectionist approach suggests that information about a target’s leadership suitability is based on the simultaneous interpretation of existing leader prototypes and contextual information (e.g., the work environment in which this target is embedded). As such, when forming judgments about whether or not a target is “behaving like a leader” or “seems like a leader,” individuals subconsciously integrate salient bottom–up inputs (e.g., their observations of the work context) with their existing top–down inputs (pre-existing leadership prototypes). In essence, people’s leader categorizations are colored and shaped by both existing leader prototypes and salient contextual features in which targets are observed.

Several sources of contextual information shape leader schemas, including observers’ own attributes (e.g., leadership preferences), the work context (e.g., industry), and additional target attributes (e.g., gender) (e.g., Chiu et al., 2017; Foti et al., 2008; Lord et al., 2020; Sy & van Knippenberg, 2021; Trichas et al., 2017). The ways in which target attributes shape leader schemas are particularly relevant to the current study. For example, a target’s emotional expressions can shape whether the target is viewed as a leader, with positive emotional expressions being associated with higher leadership categorization compared to nervous emotional expressions (Trichas et al., 2017). As a second example, when managers are viewed as having more positive network ties and fewer negative network ties, they are viewed as possessing more social power and are in turn categorized as a leader (Chiu et al., 2017). Taken together, people categorize a target as a leader based on their existing leader prototypes and additional contextual information, demonstrating that leader schemas are informed by both top–down and bottom–up information. We therefore adopt this approach in considering whether an additional target attribute contributes to leader prototypes, namely, mental health.

Leader Mental Health Prototypes

Leader Well-Being Prototypes

We first consider how general (i.e., context-free) ILTs serve as top–down influences that guide and inform observers’ expectations of leaders’ well-being. Across two studies, Epitropaki and Martin (2004) found that leader prototypes are made up of four prototypical attributes: leaders are expected to be (1) sensitive, i.e., understanding, sincere, compassionate, warm, and sympathetic; (2) dynamic, i.e., bold, strong, in control, energetic and charismatic; (3) dedicated, i.e., motivated, unrelenting, and hardworking; and (4) intelligent, i.e., clever, knowledgeable, and intellectual. We posit that the central attributes, behaviors, and characteristics of all four prototypical leadership dimensions are more schematically consistent with the characteristics associated with higher levels of psychological well-being, and therefore, existing leader prototypes may serve as a top–down, social–cognitive framework that shapes the degree to which those in leadership roles are expected to experience psychological well-being.

First, we argue that the prototypical expectation that leaders are dynamic (i.e., leaders are in control, strong, influential, charismatic; Epitropaki & Martin, 2004) may drive expectations that they also experience higher psychological autonomy (i.e., ability to regulate one’s own behavior) and environmental mastery (i.e., ability to manage one’s environment) given their conceptual overlap. The autonomy and environmental mastery dimensions of psychological well-being imply that individuals experience control over their own actions and can manage their immediate social environment and day-to-day affairs (Westerhof & Keyes, 2010). Leadership roles are linked to increased responsibilities over one’s own work and the work of others (Korman et al., 2021). Thus, we suggest observers may implicitly draw positive connections between leader prototypes/leadership characteristics and the autonomy and environmental mastery dimensions of psychological well-being.

Second, the dedication (i.e., leaders are hardworking, successful, highly motivated) and intelligence (i.e., leaders are intellectual, clever) dimensions of leader prototypes (Epitropaki & Martin, 2004) may drive predictions that leaders ought to experience greater purpose in life (i.e., strong goal orientation), personal growth (i.e., continuous development), and self-acceptance (i.e., positive self-evaluations). Purpose in life, personal growth, and self-acceptance collectively imply individuals’ approach to self-actualization (i.e., reaching one’s full potential; Maslow, 1970; Whitehead, 2017). Being hardworking and intelligent may be necessary precursors to such personal achievements. Moreover, given most organizational structures have some degree of hierarchy, with leadership titles at the top, observers may view those achieving leadership titles as indicative of growth and personal achievement, which might then extend into their expectations that leaders ought to experience higher psychological well-being. Indeed, as compared to those in subordinate positions, leadership roles are associated with greater status and power (Meindl et al., 1985), and such status and power itself may indicate to observers that leaders have achieved a sense of purpose, growth, and self-acceptance.

Third, we suggest that the sensitivity dimension of leader prototypes (i.e., being sincere, compassionate, and warm; Epitropaki & Martin, 2004) is consistent with the psychological well-being dimension of experiencing positive relations with others (Westerhof & Keyes, 2010). Indeed, leaders are expected to approach relationships with others in the workplace with social–interpersonal skill and concern (Epitropaki & Martin, 2004). In addition, leaders are expected to be central in positive social networks and disconnected from negative social ties (Chiu et al., 2017). As such, these underlying leader prototypes may drive expectations that those in leadership roles experience the psychological benefits of more and stronger social relationships with others, both within and possibly beyond the workplace.

Taken together then, we suggest that the collective leader prototypes that make up general ILTs set the socio-cognitive framework for how leaders’ psychological well-being will be perceived by others because people generate attributions that more easily align with their existing schemas (Fiske & Taylor, 1991; Foti et al., 2008) and because the underlying characteristics of psychological well-being are conceptually similar to the underlying attributes of leader prototypes. Thus, we predict:

• Hypothesis 1 (H1): Leadership roles will be positively related to perceptions of well-being.

Leader Mental Illness Prototypes

As with well-being, general leader prototypes may additionally serve as top–down influences that determine people’s expectations of leaders’ experience with mental illness. We propose that the prototypical expectations that leaders are charismatic and optimistic (i.e., dynamism dimension), in control and determined (i.e., dedicated dimension), and express empathy and warmth (i.e., sensitivity dimension) are inconsistent with the symptoms, behaviors, and assumptions associated with depression and anxiety. Indeed, depression is characterized by symptoms of low self-worth, disinterest, and withdrawal, while anxiety is characterized by self-doubt, excessive worrying, and feelings of being out of control.

Additionally, leadership roles are often prized and romanticized (Meindl et al., 1985) and viewed as roles in which occupants wield power and control (Korman et al., 2021). Yet, internalized illnesses generate stigmatized assumptions: those with depression are often perceived as having a lack of control over emotions while those with anxiety are viewed as having low ability to manage impulses (Connell et al., 2012). These stigmatized beliefs extend to the workplace, where mental illness is viewed as a weakness, an indication of low competencies, and a hindrance to organizational effectiveness (Brouwers, 2020; Janssens et al., 2021). As such, cognitive schemas surrounding leadership roles on the one hand and the symptoms and stigmatized beliefs about mental illness on the other hand are conceptually incongruent and should therefore be cognitively interpreted as a mismatch. Thus, we predict:

• Hypothesis 2 (H2): Leadership roles will be negatively related to perceptions of mental illness.

Contextual Inputs: Perceptions of Job Resources and Demands

In addition to general leadership prototypes, the connectionist approach to ILTs suggests people also consider multiple sources of contextual information to further refine and inform their leader categorizations (Adriasola & Lord, 2019; Foti et al., 2008; Junker & van Dick, 2014). One contextual cue that people may draw upon when forming expectations of work roles in general—and leaders’ mental health in particular—is their perception of job characteristics, namely, their access to job resources and exposure to job demands.

The job demands–resources model (Demerouti et al., 2001) suggests jobs are characterized by a broad set of demands (i.e., external stressors including physical, social, or organizational aspects of the job that require sustained effort) such as heavy workload or exposure to harassment. However, jobs also offer resources (i.e., physical, psychological, social, or organizational aspects of the job) that reduce job demands, facilitate goal achievement, and stimulate development, such as financial resources or decision-making discretion. The coexistence of resources and demands determines job strain in particular and mental health more generally. Put simply, jobs with greater resources and fewer demands should result in less strain and better mental health, while jobs with higher demands than resources should result in the opposite. We posit that people will perceive leaders as having greater access to job resources, which in turn would cue the expectation that they should experience better mental health.

In this research, we conceptualize job resources as the physical, social, and organizational factors that enable work-related goal attainment, including access to money (higher salary), job autonomy (i.e., discretion over one’s own work), and decision-making latitude (i.e., discretion over other’s work). Past research has supported the perspective that leadership roles are associated with greater organizational resources. For example, leaders have higher salaries (Barling & Weatherhead, 2016), job autonomy (e.g., Li et al., 2018), and decision-making latitude (Korman et al., 2021). We suggest that these job resources are visible and therefore recognized by observers. Leaders typically out-earn subordinates, and as more organizations adopt pay transparency policies, followers learn about leaders’ income. Given subordinates receive directives from leaders, it is clear that leaders have greater job control and decision-making authority than their subordinates. Since job resources indeed boost well-being and buffer against mental illness (Marmot, 2004; Sherman et al., 2012), we expect observers will perceive these additional resources as cues regarding leaders’ experience with mental health. We predict:

• Hypothesis 3 (H3): Leadership roles are indirectly and positively associated with expectations of well-being via perceptions of work role resources.

• Hypothesis 4 (H4): Leadership roles are indirectly and negatively associated with expectations of mental illness via perceptions of work role resources.

None of this suggests that leadership roles are without their own unique stressors or demands (Fletcher & French, 2021; Li et al., 2018). In this research, we conceptualize job demands as the physical, social, and organizational factors, including workload demands and interpersonal injustice, that deplete personal resources. Indeed, leadership roles have been associated with increased role demands (Li et al., 2018) and social mistreatment (Walsh et al., 2019) which in turn harms mental health. However, we argue that these demands are less visible to observers and thus have less impact on impressions of leaders’ well-being and mental illness. Indeed, leaders might hide any signs that they are struggling with the demands of their job for fear that they could be stigmatized as weak and evaluated negatively (Barling & Cloutier, 2017; Hastuti & Timming, 2021), and research does show that individuals in higher status work roles who self-disclose weakness do experience a “status penalty” in the form of lower evaluated social status (Gibson et al., 2018). Thus, even though leaders experience greater demands (Li et al., 2018), they may hide any visible signs that they are struggling with such demands, preserving the belief that leaders benefit from their additional resources. Given leadership roles are associated with increased organizational demands that may be less visible to observers, we do not form predictions regarding the role of job demands in informing expectations of leaders’ mental health. Instead, we choose to explore whether the relationship between leadership roles and mental health expectations is indirectly explained by perceptions of job demands.

Study Overview

We test our hypotheses using multiple methodological approaches. First, implementing an implicit association methodology (Uhlmann et al., 2012), we examine whether people associate “mental health at work” with organizational leadership roles (Study 1), testing H1 and H2. Second, using an experimental vignette approach, we examine whether varying health experiences are associated with leader suitability (Study 2), testing H2. Third, using a connectionist approach, we examine whether people directed to think about a leader draw on contextual cues, namely, established perceptions of job resources and demands, to inform their judgments about their mental health (Study 3), testing H3 and H4. ¹ See Figure 1 for an overview of our studies, predictions, and variables tested.

Figure 1.

Study overview.

Note. “+” denotes predicted positive relationship; “−” denotes predicted negative relationship. H1: Hypothesis 1; H2: Hypothesis 2; H3: Hypothesis 3; H4: Hypothesis 4.

Study 1: Leaders’ Well-Being and Mental Illness Prototypes

The goal of Study 1 is to examine whether organizational leaders are cognitively associated with perceptions of higher well-being and lower mental illness. We use two methodological approaches to examine these predictions. First, to assess underlying associations between work roles and mental health, participants responded to open-ended questions that examined the associations they make with “mental health at work” and different work roles. Second, we randomly assigned participants to read short descriptions of one work role (a leader, subordinate, or employee) to evaluate the degree to which they believed a person in that role experiences well-being and mental illness.

Method

Participants

To recruit participants, we used snowball sampling (Goodman, 1961) with undergraduate students from a mid-sized Canadian university who were blind to the purpose of the study. Students were invited to a campus classroom and received a 15 min. tutorial on the purpose of snowball sampling techniques. They then received a brief script to email friends and family who were employed full-time. The script described the study as a “10-minute survey on workplace attitudes” and contained a link to a survey hosted on Qualtrics, an online survey tool. This resulted in the recruitment of 113 employees from the U.S. and Canada. Participants were excluded from the study if they failed attention checks (n = 11), failed manipulation checks (n = 6), or were unemployed (n = 11), resulting in a final sample of 85 employees (46 female; M age = 40.2 years, SD = 14.94). Students received course credit for their recruitment efforts, and survey participants were entered into a draw to win a prize.

Procedure

The online survey consisted of two parts. In Part 1, participants were asked two open-ended questions, namely, (1) “who comes to mind when thinking about ‘mental health’ in the workplace” and (2) “what comes to mind when thinking about ‘mental health’ in the workplace.” These questions are similar to an item generation approach in scale development (Epitropaki & Martin, 2004) and encourage participants to freely associate the construct of “mental health at work” with whoever and whatever comes to mind. Thus, this approach activates one construct (i.e., workplace mental health) and assesses what features, characteristics, and categories are associated with this construct. The term “mental health” was used so as to not bias participants into thinking about only well-being or mental illness, and the order of presentation of the questions was randomized. The space provided for written responses was limited to motivate implicit responding (Uhlmann et al., 2012), but there was no time limit for responding.

In Part 2, participants were randomly assigned to a “leader” (n = 29), “subordinate” (n = 26), or “employee” (n = 26) condition using a randomization tool available through Qualtrics. Each condition contained a brief description of the assigned work role. Participants assigned to the leader condition read, “Leaders may hold workplace roles in which they supervise, manage, or lead other people within a workplace setting.” Participants in the subordinate condition read, “Subordinates may hold workplace roles in which they are supervised, are managed, or are led by another person within a workplace setting,” while participants in the employee condition read, “Employees may hold different workplace roles within a workplace setting.” The “subordinate” and “employee” conditions were selected to directly contrast mental health beliefs about leadership vs. non-leadership roles. The “subordinate” label was used as it more directly implies a role without leadership responsibilities, and the “employee” label was used as it is neutral regarding hierarchy and therefore should not induce implicit followership schemas associated with the subordinate title (Sy, 2010). To strengthen the manipulation, participants were asked to reflect on the duties of the work role to which they were assigned and identify anyone they may know who holds this role, by listing initials, effectively drawing on both category-based and target-based representations to prime schemas of the role (Lord et al., 2020). Finally, participants evaluated the degree to which they expected the person in this work role to experience well-being and mental illness.

Part 2 Measures

Well-Being

Perceptions of psychological well-being was measured with four items (viz., optimism, emotional stability, happiness, and high self-esteem) consistent with Westerhof and Keyes’ (2010) conceptualization. Participants rated the frequency with which they believed leaders (subordinates/employees) experience each item on a weekly basis using a 1 (almost never) to 7 (very often) scale. Items were averaged (a = .70).

Mental Illness

Perceptions of internalized mental illness symptoms were measured with three items (viz., generalized anxiety, depression, chronic sadness). Participants rated these items using the same 7-point rating scale. Items were averaged (a = .73). Other items unrelated to mental health were also embedded in these scales to prevent hypothesis guessing, as were two attention checks (e.g., “to show you are paying attention, select ‘almost never’”).

Manipulation Check

To ensure participants were reflecting on the designated work role, at the end of the survey, they were asked to indicate which work role they had thought about (i.e., leader, subordinate, employee); only those who correctly identified the work role were retained for analyses. As an additional manipulation check, participants rated the degree to which they believed this work role had “income to meet needs” and “social status at work,” assessed on a 1 (low) to 7 (high) scale, with items averaged (r = .50). These items were included as they meaningfully differentiate between leader, employee, and subordinate work roles.

Control Variables

We controlled for participants’ current leadership status (“Do you currently supervise anyone in your job?”; 0 = no, 1 = yes) because those in leadership roles may draw on their own mental health to inform any predictions they have about leaders’ mental health. We also controlled for participants’ perceptions of the work role gender (“What gender did you imagine the [leader/employee/subordinate] to be?”; 0 = female, 1 = male), because leadership roles are associated with men while subordinate roles are associated with women (Braun et al., 2017; Schein et al., 1996), and internalized mental illnesses are stereotypically associated with women (Boysen et al., 2014); thus, the perceived target gender could alternatively explain any association found between work role and mental health expectations.

Analyses and Results

Part 1

We thematically coded participants’ open-ended responses using techniques described by Charmaz (2006). We were interested in identifying the work roles that participants listed in response to the question “Who comes to mind when you think about mental health in the workplace?” Of the 81 participants who responded to this question, 40 participants listed one work role, and 10 listed multiple work roles, resulting in 63 total listed work roles. The other participants did not list work roles, but rather industries or specific names of people, which we did not code. The listed work roles were separated into distinct codes, which we then aggregated into higher-order categories based on their similarity in workplace responsibilities. Three categories emerged: leadership roles (e.g., “management,” “leaders,” “supervisors”; n = 24), general work roles (e.g., “employees,” “coworkers,” “everyone”; n = 28), and subordinate work roles (e.g., “followers,” “entry workers,” “front line”; n = 10).

We next coded participants’ responses to the question “What comes to mind when you think about mental health in the workplace?” Responses were again reduced to specific codes, from which two distinct categories emerged: well-being (i.e., positively balanced mental health experiences; e.g., “stability,” “calmness,” “managing stress”; n = 35) and mental illness (i.e., negatively balanced mental health experiences; e.g., “stress,” “anxiety,” “burnout”; n = 27).

We used χ² tests to assess whether the work role(s) people listed were associated with indicators of well-being or mental illness in the workplace. Results showed that both categories of mental health perceptions (viz., well-being and mental illness) significantly differed by work role (viz., leader, employee, subordinate categories), χ² (2, 62) = 13.84, p = .001, Phi = .47 (see Table 1). ² Specifically, leader categories were significantly and positively associated with well-being in the workplace compared to subordinate and employee role categories, which were equally associated with thoughts of mental illness and well-being.

Table 1.

Implicit Association: χ² Results (Study 1, n = 62).

		What comes to mind when you think of mental health in the workplace?
		Positive mental health	Negative mental health	Total
Who comes to mind when you think of mental health in the workplace?	Leadership roles	20 (83%)	4 (17%)	24 (38%)
	General work roles	9 (32%)	19 (68%)	28 (45%)
	Subordinate roles	6 (60%)	4 (40%)	10 (16%)
	Total	35 (56.5%)	27 (43.5%)	62 (100%)

Part 2. Descriptive statistics and intercorrelations appear in Table 2.

Table 2.

Descriptive Statistics and Intercorrelations (Study 1, n = 81).

	M (SD)	1	2	3	4	5	6
1. Condition	1.06 (.84)	–
2. Job status	4.36 (1.62)	.445**	.50
3. Well-being	4.93 (1.07)	.305**	.349*	.70
4. Mental illness	3.33 (1.23)	−.200	.009	.225*	.73
5. Participant leader status	.53 (.50)	−.189	−.358**	−.164	−.020	–
6. Perceived gender	.52 (.50)	.170	−.039	−.045	.054	−.075	–

Note. Listwise deletion was applied. Alphas (and in the case of job status, the inter-item correlation) appear on the diagonal, italicized. Condition was coded 0 = subordinate, 1 = employee, 2 = leader role. Job status was rated on a 1 (low) to 7 (high) scale. Well-being and mental illness were rated on 1 (almost never) to 7 (very often) scales. Participant leader status was coded 0 = non-leader, 1 = leader. Perceptions of gender was coded 0 = female, 1 = male.

*p < .05; **p < .001.

We began by conducting a series of manipulation checks to test the validity of the experimental manipulation using perceptions of job status. As expected, leaders were perceived as having higher job status than employees (M = 5.19, SD = 1.57 vs, M = 4.35, SD = 1.58, Cohen’s d = .53) and subordinates (M = 3.44, SD = 1.23, Cohen’s d = 1.24), all ts(78) > 2.12, ps < .04; and employees were perceived as having higher job status than subordinates, t = 2.21, p = .03, Cohen’s d = .64, supporting that participants were thinking of different work roles.

Next, a priori contrasts were used to assess whether work roles were differentially associated with expectations of well-being and mental illness, controlling for participant leadership status and perceptions of the target work role gender. As predicted in H1, leaders were expected to experience higher levels of well-being than employees (M = 5.35, SD = .73 vs. M = 4.83, SD = 1.17; Cohen’s d = .53; post hoc power = .62) and subordinates (M = 4.57, SD = 1.16; Cohen’s d = .81; post hoc power = .91); all ts(78) > .617, ps < .015. Leaders were also expected to experience less mental illness than employees (M = 2.97, SD = 1.23 vs. M = 3.53, SD = 1.13, Cohen’s d = .47; post hoc power = .53) and subordinates (M = 3.55, SD = 1.27; Cohen’s d = .46; post hoc power = .53), all ts(78) > −.634, ps < .033, thereby supporting H2. There were no differences in perceptions of well-being and mental illness between subordinate and employee conditions, all ts(78) < −.030, ps > .301. ³

Discussion

Across assessments, the results of Study 1 show that leaders are (1) more likely to be associated with positive mental health at work and (2) expected to experience greater psychological well-being and less depressive and anxious symptoms than those holding subordinate or employee roles. There were no differences in perceptions of well-being and mental illness between subordinates and employees, suggesting that it is the leadership role that influences perceptions of mental health, rather than a subordinate-labeling effect (Sy, 2010). These results, which emerged from two different methods, support H1 and H2. However, excluding participants who did not match eligibility criteria or follow instructions resulted in a small sample size and subsequent low power in analyses examining differences in expectations of mental illness (H2). We address these limitations in Study 2.

Study 2: Mental Illness and Leadership Prototypes

The goal of Study 2 was twofold. First, we aimed to replicate the findings from Study 1 showing that leaders are expected to experience less mental illness than non-leaders (H2), but with a larger sample to increase statistical power. Second, to avoid mono-operation bias (Shadish et al., 2002), we varied the experimental approach and investigated whether observers evaluate an employee with a mental illness as less leader-like via their perceptions of the target’s leadership attributes (Epitropaki & Martin, 2004) and potential to engage in constructive leadership behaviors (Rubin et al., 2005).

Method

Participants

Drawing upon the actual effect sizes established in Study 1 and using G*Power (effect size f = .25, a = .04, 1-β err prob = .80, n groups = 3), we determined that a sample of 159 was required. We oversampled to account for the expected loss of participants. Three hundred three students from a mid-sized Canadian university were recruited to participate in a 30 min online survey in exchange for course credit; 103 participants were excluded based on incomplete survey data (n = 18), failing attention checks and manipulation checks (n = 63), or repeating the survey multiple times (n = 12). This resulted in a final sample of 200 participants (56.5% female; M age = 19.65, SD = 1.43). All participants had previously been employed, and 51.5% reported leadership experience.

Procedure

Via Qualtrics’s embedded randomization tool, participants were randomly assigned to one of three experimental conditions in which they read a brief vignette giving basic information about an employee. The employee was described as a “well-trained, reliable worker, who is willing to put in the extra hours” across conditions. Thereafter, participants were randomly assigned to read that the employee was either (1) diagnosed with a mental illness of clinical depression (n = 65), (2) diagnosed with a physical illness of hypertension (n = 72), or (3) received news regarding an organizational change (control condition; n = 63).

The mental illness condition served as our experimental condition, with a diagnosis of clinical depression selected as it represents a prevalent internalized disorder and is well recognized by the general public. The “physical illness” condition was included to examine whether it is perceptions of mental illness, or any illness, that influences leadership evaluations, and hypertension was selected as it reflects a common physical diagnosis. The “no illness” condition was included as a control in which illness is not primed, but the employee undergoes a change (i.e., change in work location). After reading their respective vignettes, participants rated their perceptions of the employee’s leadership potential. See Appendix A for vignettes.

Measures

Leadership Suitability

We measured perceptions of leadership suitability using two scales. First, we drew upon Epitropaki and Martin’s (2004) 21-item ILT scale to measure perceived leadership attributes. Participants rated how characteristic each of the 21 traits was of the employee described in the vignette, using a 9-point scale (1: not at all characteristic; 9: extremely characteristic). The ILT scale comprises six dimensions, four of which represent prototypical leader attributes (total a = .90): sensitivity (3 items: understanding, sincere, helpful; a = .71), intelligence (4 items: intelligent, knowledgeable, educated, clever; a = .79), dedication (3 items: motivated, dedicated, hardworking; a = .83), and dynamism (3 items: energetic, strong, dynamic; a = .76). Two antitypical leader attributes (total a = .83): tyranny (6 items: domineering, pushy, manipulative, loud, conceited, selfish; a = .85) and masculinity (2 items: masculine, male; inter-item r = .64) were also measured. The recommendation to examine prototypical attributes as a single factor was followed (Epitropaki & Martin, 2004; 2005).

Second, we used a shortened version of Rubin et al.’s (2005) transformational leadership scale to assess participants’ perceptions of the target’s potential to enact high-quality leadership behaviors. Participants rated how likely the employee would be to engage in each dimension of transformational leadership if they were placed in a leadership role (i.e., “inspire others with his/her plans for the future,” “get the groups to work together for the same goal,” “lead by doing rather than simply telling,” “show respect for employees’ personal feelings”) rated on a 7-point scale (1: very unlikely; 7: very likely). Items were aggregated (a = .77).

Manipulation Check

To assess the validity of experimental manipulation, participants were asked, “What type of information did this person receive?” and selected from five categorical options: (1) diagnosed with depression, (2) diagnosed with anxiety, (3) diagnosed with high blood pressure, (4) a change in office location, or (5) I don’t know. Only those who selected the right condition were retained.

Control Variables

As in Study 1, we controlled for participants’ leadership role experience (“Have you ever been in an organizational leadership role in which you directly supervised others?”; 0 = no, 1 = yes) and their perception of the employee’s gender (“When reading the previous description, what gender did you imagine the person to be?” 0 = female; 1 = male; 3 = unspecified).

Results

Descriptive statistics and intercorrelations appear in Table 3.

Table 3.

Descriptive Statistics and Intercorrelations (Study 2, n = 200).

	M (SD)	1	2	3	4	5
1. Condition	1.01 (.80)	–
2. Prototypical leader attributes	6.37 (1.13)	−.209**	.90
3. Transformational leadership behaviors	4.75 (1.01)	−.242**	.624**	.77
4. Participant leader role experience	.52 (.50)	−.013	.055	.052	–
5. Perceived target gender	1.10 (.85)	.017	.142	.070	−.186*	–

Note. Listwise deletion was applied. Alphas appear on the diagonal and are italicized. Condition coded 0 = mental jllness, 1 = physical illness, 2 = no illness diagnosis. Prototypical and antitypical leader attributes rated on a 1 (not at all characteristic) to 9 (extremely characteristic) scale. Transformational leadership behaviors rated on a 1 (very unlikely) to 7 (very likely) scale. Participant past leader role experience was coded 0 = non-supervisor, 1 = supervisor, and perceptions of the work role gender was coded 0 = female, 1 = male, 3 = unspecified.

*p < .05; **p < .001.

A priori contrasts were used to assess whether illness diagnosis affected perceptions of prototypical leadership attributes and expectations for transformational leadership behaviors. We controlled for participant leadership role experience and their perception of the work role’s gender across analyses. As expected, an employee diagnosed with a mental illness was perceived as having less prototypical leadership attributes than the employee with no illness diagnosis (M = 5.96, SD = 1.11 vs. M = 6.55, SD = 1.04, respectively; Cohen’s d = .55; post hoc power = .87) or a physical illness diagnosis (M = 6.58, SD = 1.15; Cohen’s d = .55; post hoc power = .89); all ts(197) > −.596, ps < .001. ⁴

Similarly, the employee diagnosed with a mental illness was expected to engage in fewer transformational leadership behaviors than the employee with no illness diagnosis (M = 4.40, SD = 1.03 vs. M = 5.01, SD = .85, respectively; Cohen’s d = .65; post hoc power = .98) or a physical illness diagnosis (M = 4.85, SD = 1.05; Cohen’s d = .43; post hoc power = .80); all ts(197) > −.574, ps < .001. ⁵ Together, results offer additional support for H2, namely, that mental illness is negatively associated with leadership prototypes.

Discussion

Study 1 established that leaders are expected to experience higher well-being and lower mental illness than non-leader roles. Building on these results, Study 2 showed that an employee with a mental illness is viewed as less leader-like as indicated by lower perceived leadership attributes and expectations that the employee would engage in fewer constructive leadership behaviors. Confidence can be expressed in these results as a mental illness diagnosis was contrasted both with no illness and a physical illness diagnosis.

Study 3: The Mediating Role of Job Resources

While the first two studies tested the top–down influence of general ILTs on leader mental health prototypes, the role of contextual information in shaping these perceptions proposed in H3 and H4 remains untested. To further our application of a connectionist approach in examining context-specific ILTs, we now assess whether job resources serve as a contextual cue that relates to leaders being viewed as mentally healthier (i.e., higher well-being, H3, and lower mental illness, H4) than non-leaders. We also explore whether perceptions of job demands mediate the relationship between work roles and mental health expectations.

Method

Participants

Participants from the U.S. were recruited by Qualtrics, a recruitment service with access to a global panel of participants. Only employed individuals who passed all three attention checks (e.g., “to show you are paying attention, leave this response blank”) and completed more than 75% of the survey were retained, resulting in a final sample of 104 full-time employees (57% female; M age = 39.85 years, SD = 11.76) ⁶ .

Procedure

Via Qualtrics’s embedded randomization tool, participants were randomly assigned to one of two experimental conditions (“leader” [n = 52], “subordinate” [n = 51]) in which they read a brief vignette describing a target’s work role, skills, job duties, and performance. Participants in the leader condition read a similar opening vignette to Study 2; however, this time, the “employee” had a leadership title with leadership responsibilities. Participants in the subordinate condition read the same work role description described in Study 2; however, the “employee” label was changed to “subordinate” to better contrast perceptions of the work roles. No illness/change was described in these vignettes. After reading their respective vignettes, participants rated the degree to which they believed the target (a) had access to job resources, (b) experienced job demands, and (c) would experience well-being and mental illness in the future (see Appendix B for full vignettes).

Measures

Unless otherwise indicated, all measures were rated on a 1 (extremely unlikely) to 7 (extremely likely) scale. Items in all four measures are averaged.

Job Resources and Demands

To assess perceptions of job resources, participants indicated their perceptions of the work role’s current income (i.e., “have enough money to satisfy needs and desires”), job control (i.e., “has the ability to use discretion and skills at work”), and decision-making authority (i.e., “has a great deal of say and authority in making decisions for others at work”; a = .70). To assess perceptions of job demands, participants indicated their perceptions of the work role’s current job demands (i.e., “have to work very intensely and feel as though there are too many demands at work”) and experience of injustice (i.e., “at work, will feel cheated, disrespected, discouraged, and not justly rewarded for effort”; inter-item correlation = .48). These items were derived from Demerouti et al.’s (2001) conceptualization of job resources and demands.

Well-Being

Participants reported predictions for the leader’s or subordinate’s future (“three months from now”) psychological well-being. We used the highest loading item from each sub-dimension (autonomy, environmental mastery, personal growth, positive relations with others, purpose in life, self-acceptance; a = .84) of Ryff’s (1989) psychological well-being measure (e.g., “lead a purposeful and meaningful life [in the next three months]”). Evidence for the validity of shortened measures of well-being has been established (e.g., Diener et al., 2010).

Mental Illness

Expectations for future mental illness was assessed with five items adapted from the Diagnostic and Statistical Manual of Mental Disorders (4th ed.) (DSM-IV) (Ryff & Keyes, 1995) capturing depressive symptomology and diagnosis, anxious symptomology and diagnosis, and panic attacks (e.g., “experience symptoms of anxiety [in the next three months]”; a = .89).

Manipulation Check

To ensure the validity of the manipulation, participants first indicated which work role they reflected upon, selecting one of four options (i.e., leaders, managers, employees, or subordinates). Only participants who selected the correct work role based on their condition were retained for analyses. Participants also rated their perceptions of job status (two items: “how much status [power] do you feel this person has at work specifically, relative to others in their work organization; inter-item correlation = .77) using a 0 (very low) to 10 (very high) response scale.

Control Variables

As in Studies 1 and 2, we controlled for participants’ leadership role experience (“Do you currently supervise anyone in your work role?”; 0 = no, 1 = yes) and their perception of the work role’s gender (“When reading the previous description, what gender did you imagine [the leader/subordinate] to be?” 0 = female; 1 = male).

Data Analysis

To test our prediction that perceived access to job resources link the indirect effect of work role (leader = 1, subordinate = 0) on predictions of future well-being (H3) and mental illness (H4), we used PROCESS 4.0 for SPSS (Hayes, 2017; http://www.afhayes.com), drawing on a bootstrapped sample of 10,000, and determined significance based on whether confidence intervals (CIs) excluded 0. We used the same analytic approach to explore the indirect effects of job demands. Across analyses, we controlled for participants’ leadership role experience and perceptions of the target work role’s gender.

Results

Descriptive statistics and intercorrelations for all variables in Study 3 appear in Table 4. ⁷

Table 4.

Descriptive Statistics and Intercorrelations (Study 3, n = 104).

	M (SD)	1	2	3	4	5	6.	7.	8.
1. Condition	.50 (.50)	–
2. Job status	6.72 (1.94)	.398**	.77
3. Job resources	5.32 (1.20)	.333**	.338**	.70
4. Job demands	3.64 (1.41)	.200*	.042	−.226*	.46
5. Well-being	5.31 (1.00)	.268**	.261**	.787**	−.153	.84	–	–	–
6. Mental illness	3.35 (1.29)	.001	−.029	−.264**	.581**	-.291**	.89
7. Participant leader status	.56 (.50)	−.108	−.041	−.007	−.156	-.061	-.188	–
8. Perceived gender	.72 (.45)	.102	.063	−.018	.127	-.032	.149	-.009	–

Note. Listwise deletion was applied. Alphas (and in the case of job status and job demands, inter-item correlation) appear on the diagonal and are italicized. Condition is coded 0 = subordinate role, 1 = leader role. Job status was rated on a 0 (very low) to 10 (very high) scale. Job resources, well-being, and mental illness were rated on 1 (extremely unlikely) to 7 (extremely likely). Participant leader status was coded 0 = non-supervisor, 1 = supervisor, and perceptions of the work role (leader/subordinate) gender was coded 0 = female, 1 = male.

*p < .05; **p < .001.

The validity of the experimental manipulation was supported: using an independent sample t-test, the leader (M = 7.56, SD = 1.83, Cohen’s d = .86) was perceived as having significantly more job status than the subordinate condition (M = 6.11, SD = 1.85, t(101) = −3.970, p < .001, Cohen’s d = 1.24). Further, compared to the subordinate condition, the leader condition was perceived as having greater access to job resources (subordinate M = 4.92, SD = 1.19; leader M = 5.71, SD = 1.08; t(101) =  −3.55, p < .001, Cohen’s d = 1.14) and more job demands (subordinate M = 3.35, SD = 1.22; leader M = 3.91, SD = 1.54; t(101) = −2.05, p = .043; Cohen’s d = 1.39).

We next examined whether perceptions of job resources explained the relationship between work role and future mental health expectations. Support emerged for H3 and H4. After controlling for leader experience and perceptions of gender, leaders were perceived to have greater job resources than subordinates, resulting in expectations of higher well-being (indirect effect: b = .526; CI: [.215, .898]; direct effect: b = −.005; CI: [−.269, .260]) and lower mental illness (indirect effect: b = −.238; CI: [−.520, −.042]; direct effect: b = .137; CI: [−.388, .662]). See Table 5.

Table 5.

Predicting Work Role Well-Being and Mental Illness through Perceptions of Job Resources (Study 3, n = 103).

Outcome: job resources
Variables	B	SE	t	LLCI	ULCI
Constant	4.957	.281	17.649	4.399	5.515
Condition	.814	.234	3.474	.349	1.280
Participant leader status	.068	.235	.290	−.398	.534
Perceived gender	−.140	.260	−.539	−.655	.375
Outcome: well-being
Variables	B	SE	t	LLCI	ULCI
Constant	1.973	.310	6.371	1.358	2.588
Condition	−.005	.133	−.035	−.269	.260
Job resources	.646	.055	11.82	.537	.754
Participant leader status	−.134	.126	−1.069	−.384	.115
Perceived gender	−.040	.139	−.286	−.316	.237
Outcome: mental illness
Variables	B	SE	t	LLCI	ULCI
Constant	4.813	.616	7.819	3.591	6.035
Condition	.137	.265	.518	−.388	.662
Job resources	−.293	.109	2.696	−.508	−.077
Participant leader status	−.445	.250	−1.779	−.941	.052
Perceived gender	.389	.277	1.405	−.160	.938
Indirect effects
		Effect	Boot SE	LLCI	ULCI
Work role expectations of well-being through job resources		.526	.176	.215	.898
Work role expectations of mental illness through job resources		−.238	.124	−.520	−.042

Notes. LLCI: lower limit confidence interval; ULCI: upper limit confidence interval; SE: standard error.

Finally, we explored whether perceptions of job demands explained the relationship between work role and future mental health expectations. Using the same control variables, perceptions of job demands did not explain the indirect relationship between work role and well-being (indirect effect: b = −.072; CI: [−.201, .024]; direct effect: b = .593; CI: [.202, .985]) or mental illness (indirect effect: b = .275; CI: [−.030, .601]; direct effect: b = −.387; CI: [−.815, .040]). ⁸

Discussion

Results from Study 3 identify one contextual cue that may explain why leaders are expected to have higher well-being and less mental illness than non-leaders. Specifically, leaders are expected to have greater access to job resources, which signals to observers that leaders might also experience better mental health. Moreover, though leaders are also expected to experience higher job demands, results showed that these perceptions did not drive expectations of leaders’ future mental health experiences, suggesting that people may over-rely on the perceived benefits of organizational resources in facilitating mental health and may discount the costs of job demands.

General Discussion

The importance of mental health in the workplace is garnering increased research and socio-political attention. Major organizational bodies (e.g., the WHO) and political decision-makers (e.g., President Biden’s State of the Union address) continue to make calls to action in light of the ongoing “global mental health crisis,” pointing to organizations as particular places for intervention (Staglin, 2023). Yet, organizational mental health accommodation programs tend to target employees deemed most vulnerable, while those in leadership roles are often neglected (Barling & Cloutier, 2017). We suggest one reason for such neglect may be the expectation that certain work roles are deemed “mentally healthier” than others. The goal of our research was to examine whether people hold positive mental health prototypes for organizational leaders.

Results from Study 1 showed that people do indeed expect leaders to manifest greater well-being and less mental illness than those in non-leadership roles (supporting H1 and H2). Results from Study 2 showed that mental illness is inconsistent with leader prototypes (supporting H2), and Study 3 demonstrated that people draw upon contextual cues, namely, job resources, to inform these perceptions (supporting H3 and H4).

These results should be interpreted in light of previous findings; though people believe leaders ought to experience better mental health than other work roles, leaders in fact experience both higher demands and resources that respectively hinder and benefit their mental health, essentially cancelling one another out, and resulting in leaders experiencing similar degrees of mental health issues as those in non-leadership roles (Debus et al., 2019; Fletcher & French, 2021; Li et al., 2018). Recent evidence further suggests that those in top echelon leadership roles do experience mental health disorders and, as a function of those disorders, end their tenure earlier than those without mental illness (Keloharju et al., 2023). Taken together, our results suggest people have biased expectations of the quality of mental health that their organizational leaders possess, such that they believe leaders to be mentally healthier than they are in reality.

Theoretical Implications

Results from this research offer theoretical contributions to the workplace mental health literature, ILT, and research on leaders’ mental health. First, we contribute to the occupational mental health literature. To date, most research examining beliefs about mental health in the workplace have focused on how mental illness is broadly stigmatized, resulting in discriminatory hiring practices and withdrawal (e.g., Elraz, 2018; Hastuti & Timming, 2021; Janssens et al., 2021). However, we extend this literature by suggesting that our expectations of others’ mental health at work may not be universal, but rather may differ based on a target’s organizational role and responsibilities. Indeed, people develop differing expectations of others based on organizational titles (Lord et al., 2020; Sy, 2010); however, no research to date has explored whether work roles engender different mental health expectations. We find that leadership roles may be particularly vulnerable to biased views of mental well-being. Intriguingly, these findings may also explain why mental illness continues to be stigmatized within organizations despite efforts to reduce such views. Indeed, one implicit way in which stigma against mental illness may manifest is by categorizing those with mental illness as unfit to lead, while those exhibiting well-being as most suitable for leadership roles.

Second, our findings extend our understanding about the context-specific prototypes people hold about leaders. Beyond general leader prototypes, research is increasingly adopting a connectionist approach (Foti et al., 2008) to identify the specific attributes people expect to see in their leaders. That is, in addition to general leader prototypes, researchers have identified that people hold, for example, specific expectations regarding the gender (Braun et al., 2017), race (Petsko & Rosette, 2023), emotions (e.g., Sy & van Knippenberg, 2021), and social networks (e.g., Chiu et al., 2017) expected of leaders. Our research adopts this lens and finds that people also hold mental health expectations of leaders, suggesting that targets exhibiting well-being may be more likely to be categorized as leaders, while those with mental illness are not.

Moreover, we demonstrate that expectations of leaders’ mental health may also emerge as a function of work role characteristics. That is, we suggest observers may use information regarding work role characteristics, for example, leaders’ access to organizational resources, to justify implicitly why those in leadership roles should experience better mental health. Intriguingly, though leadership roles were associated with greater perceptions of both job resources and demands compared to those in subordinate roles, only organizational resources influenced future mental health expectations. One potential explanation for this finding is that observers are prioritizing information consistent with their top–down leader prototypes. That is, the expectation that leaders are generally powerful and in control (Epitropaki & Martin, 2005) is schematically aligned with the characteristics of well-being (environmental mastery, personal growth) and the positive benefits of organizational resources (e.g., greater decision discretion and autonomy). Incorporating the belief that leaders may also have additional organizational demands does not necessarily fit as nicely in this narrative, and thus, its influence on mental health expectations may be implicitly downplayed. We therefore contribute to explaining why implicit leader prototypes may be so robust through time (Epitropaki & Martin, 2005), despite the presence of disconfirming evidence.

Third, our findings contribute to the growing body of research focused on understanding leaders’ mental health. To date, research on leaders’ well-being has received considerably less attention than research on subordinate well-being (Barling & Cloutier, 2017). In their review of research on leaders’ mental health, Barling and Cloutier (2017) proposed that one reason for this lack of attention is because leaders are generally expected to be mentally healthier, and so, any research on their well-being would be considered redundant. Our results therefore directly test and support this proposition. However, since their review, research has demonstrated that leaders are just as subject to mental illness as non-leaders (Debus et al., 2019; Fletcher & French, 2021; Li et al., 2018). This suggests people’s expectations of leaders’ mental health do not align with leader’s lived experience and further justify the need to study and understand leaders’ unique experience with their mental health. As such, a natural extension of our results would be to consider the consequences of leader mental health expectations, which we discuss in greater detail within our future directions.

Organizational Implications

The results of our research can inform those (1) appointing leadership roles, (2) seeking leadership opportunities, and (3) in leadership roles. First, leadership prototypes influence hiring and promotion decisions (Foti et al., 2012; Hansbrough et al., 2015), such that hiring committees are more likely to select applicants whose attributes match their leader prototypes. If people involved in hiring decisions believe that leaders ought to be mentally healthy, indicators of well-being and mental illness may influence leadership selection decisions. Although many people choose to actively conceal mental health struggles at work (Bril-Barniv et al., 2017), recent research shows more people are disclosing mental health status to organizations (Hastuti & Timming, 2021). In an inclusive work culture, disclosure has benefits (e.g., more flexible work arrangements, more organizational support). However, disclosure may simultaneously influence those involved in leadership selection, biasing them against applicants disclosing mental illness.

Second, people with mental illnesses may not apply for leadership roles if they see their illness as incongruent with leadership prototypes. Individuals with traits incongruent to those of leadership prototypes are less likely to apply for leadership roles (DeRue et al., 2015; Epitropaki et al., 2017; Kwok et al., 2018), and people with mental illness avoid job opportunities when they perceive stigma (Elraz., 2018). Seeing oneself as non-leader-like has consequences itself, such as low positive affect (Hoption et al., 2012) and engaging in counterproductive workplace behaviors (Fine et al., 2016).

Given these implications, those responsible for leadership selection should receive training designed to recognize biases regarding mental health, including how biases influence decision-making. Informing decision-makers in this way has been shown to mitigate bias against the hiring of pregnant women job applicants (Morgan et al., 2013) and female leader applicants (Leicht et al., 2014). At the same time, people involved in leader recruitment should be made aware that some qualified applicants may choose not to apply if they believe existing mental illnesses could handicap them. Recent research showing that opt-out mechanisms, whereby all qualified candidates are considered for a position, may help overcome such biases by mitigating the unwillingness to apply for the position (Erkal et al., 2021).

The argument that those with mental illness should be excluded from leadership roles can be discounted. People with mental illness (1) benefit from employment opportunities (Silván-Ferrero et al., 2022), (2) can actively cope with mental illness given appropriate treatment and organizational support, and (3) precisely because of their experience with mental illness, may develop coping strategies and attributes (e.g., empathy) that promote high quality leadership behaviors (Ghaemi, 2011). Further, mental illness affects mostly everyone, with reports suggesting that almost half of the working population suffer mental illness symptoms (Mind Share Partners, 2021). To discount such a large population from leadership selection would certainly restrict the leadership pool of potentially talented and suitable individuals.

Finally, people in leadership roles may be directly affected by leader mental health schemas. Leaders may feel as though they do not have the same access to mental health accommodation programs given they ought to experience better mental health. Further, leaders may become motivated to hide any signs of mental illness, including avoiding organizational resources supporting well-being at work, afraid that displaying indication of mental illness could result in negative evaluations (Hansbrough et al., 2015). Just as the “think manager, think male” phenomenon persists (Ferguson, 2018; Koenig et al., 2011), “think leader, think well-being” stereotypes may perpetuate the negative associations with mental illness at work.

Strengths

Several strengths inherent in the current research allow for greater confidence in the validity of the findings. First, to ensure construct validity, we conceptualized leader mental health along two continua (i.e., well-being and mental illness) and treated these conceptualizations as separate throughout the studies. Operationalizations of well-being and mental illness matched their conceptualization, and we varied how they were operationalized across samples to reduce mono-operation bias (Shadish et al., 2002). Second, to enhance internal validity and generalizability, we drew upon diverse samples and methods, replicating support for our hypotheses. For example, we used samples drawn from employees, students, and panel data, adopting both grounded and experimental approaches. Third, the use of experimental methods enhances the ability to make causal inferences, increasing the internal validity of our results.

Limitations

Despite these strengths, this research has limitations. First, methodological limitations could limit the interpretation of our results. For example, statistical conclusion validity (Shadish et al., 2002) emerging from the small sample and low power in Study 1 could limit any interpretations. However, we addressed this issue in Studies 2 and 3, which used larger samples with sufficient power to test our hypotheses. In addition, we were limited to testing indirect effects using cross-sectional data in Study 3. However, our hypotheses were theoretically driven, and more importantly, the independent variable was manipulated experimentally (Pirlott & MacKinnon, 2016). Moreover, the outcomes (i.e., expectations of well-being and mental illness) were couched as predictions for future events. As a result, inferences about the mediating function of job resources remain plausible (Spector, 2019). A second potential limitation concerns the operationalization of mental health expectations and the leadership role. We only operationalized three indicators of mental health, namely, psychological well-being, depression, and anxiety. As such, whether people have differing expectations of other mental health indicators (e.g., more severe mental illness disorders such as schizophrenia; externalized disorders such as substance abuse) remains untested. Nonetheless, by operationalizing three different indicators, concerns about mono-operation bias are reduced. In addition, leadership roles were operationalized at the lower levels of the organizational hierarchy, and future research should assess whether the current findings replicate for leadership roles such as the chief executive officer (CEO).

Future Directions

We offer two broad suggestions for future research. First, future research should expand the operationalizations of mental health and leader role. Across all three studies, mental illness was operationalized as internalized disorders (e.g., depression, anxiety), and future research could consider whether people also hold leadership prototypes related to externalized disorders (e.g., substance abuse, conduct disorder) and whether prototypes vary by illness severity (e.g., depressive symptoms vs. schizophrenia). Similarly, the level of leadership was not specified in Studies 1 and 3, and future research could investigate whether expectations of leaders’ mental health replicate across different levels of leadership (e.g., middle vs. top management; Sy & van Knippenberg, 2021).

Second, it is critical to consider the many consequences of leader mental health expectations for leaders, their followers, and organizations. First, future research should assess whether leader mental health expectations affect leaders’ health behaviors, including their likelihood of (a) disclosing mental illness and (b) accessing organizational mental health accommodation programs. Second, research might consider whether followers’ mental health experiences influence their willingness to pursue leadership roles. That is, individuals with a history of mental illness might perceive their mental illness as limiting their potential access to organizational leadership roles as a function of leader mental health stereotypes. This may ultimately increase their own self-selection out of leadership opportunities. Third, researchers should consider the organizational consequences that might emerge if leaders disconfirm role expectations by disclosing mental illness. On the one hand, existing research on ILTs would suggest that such disclosure would result in negative follower attitudes and workplace behaviors (Hansbrough et al., 2021; De Luque et al., 2008; Veestraeten et al., 2021) that could impede organizational functioning. On the other hand, leaders’ disclosure of mental illness could help to destigmatize beliefs surrounding mental illness, enabling others to disclose struggles (Hastuti & Timming, 2021), thereby creating a psychologically safe workplace. Taken together, results from our research could serve as a springboard for future research testing the consequences of leader mental health beliefs.

Conclusion

In her book on presidential leadership and illness, McDermott (2008) wrote, “Powerful leaders are not exempt from illness by virtue of their position or its influence.” Despite this, people romanticize leaders (Collinson et al., 2018; Meindl et al., 1985), expecting them to enjoy greater well-being and an absence of mental illness. However, as previous research has established, there is no clear evidence that leaders enjoy meaningfully better mental health than non-leaders (Debus et al., 2019; Fletcher & French, 2021; Li et al., 2018). Given these inconsistencies, it is time to acknowledge that leaders’ mental health matters, possibly more than people may expect.

Author Biographies

Anika E. Cloutier is an assistant professor in the Faculty of Management at Dalhousie University. Her research focuses on understanding predictors (e.g., mental health, aspects of one's home-life) of leader emergence and leadership behaviors.

Julian Barling is a distinguished university professor at Queen's and the Borden Chair of Leadership in the Smith School of Business. He has written a number of books, including the recent Brave new workplace: Designing productive, healthy and safe organizations and The science of leadership: Lessons from research for organizational leaders.

Appendix A: Study 2 Vignettes

Vignette Description Prior to Manipulation

The following is a description of employee X [name removed to maintain confidentiality]. Employee X is a worker for a mid-sized organization within the marketing sector and works with approximately 20 employees under the leadership of one manager. All of the employees working in Employee X’s department have a good working knowledge of marketing principles as demonstrated by their prior and current work experience. Employee X has worked for this organization for 12 years, and has been described as well-trained, reliable worker, who is willing to put in extra hours. Employee X’s current position requires some solitary work and some interaction with people inside and outside of the organization.

Mental Illness Condition:

Employee X has been feeling down for some time (e.g., experiencing sadness), and sought out help from a physician. This physician diagnosed Employee X with clinical depression, a risk factor for other illnesses. Employee X was surprised by this diagnosis but is following the treatment prescribed by the physician.

Physical Illness Condition:

Employee X has been feeling unwell for some time (e.g., experiencing dizziness), and sought out help from a physician. This physician diagnosed Employee X with clinical levels of high blood pressure, a risk factor for heart disease. Employee X was surprised by this diagnosis but is following the treatment prescribed by the physician.

Control Condition:

Employee X was informed that the organization will be moving office locations. This new location will be about the same distance from home (5-minute difference), the cost of parking is not expected to change but the location is the opposite direction of where this leader used to work – from the West end of the city to the East end.

Appendix B: Study 3 Vignettes

Vignette Describing Leader Role:

The following is a description of Leader X [name removed to maintain confidentiality]. Leader X leads a mid-sized organization within the marketing sector and is directly responsible for approximately 20 employees. All of the employees working in Leader X’s department have a good working knowledge of marketing principles as demonstrated by their prior and current work experience. Leader X has worked for this organization for 12 years, and has been described as a well-trained, reliable leader, who is willing to put in extra hours. Leader X’s current position requires some solitary work, leadership responsibilities, and some interaction with people inside and outside of the organization.

Vignette Describing Subordinate Role:

The following is a description of subordinate X [name removed to maintain confidentiality]. Subordinate X is a worker for a mid-sized organization within the marketing sector and works with approximately 20 employees under the leadership of one manager. All of the employees working in subordinate X’s department have a good working knowledge of marketing principles as demonstrated by their prior and current work experience. Subordinate X has worked for this organization for 12 years, and has been described as well-trained, reliable worker, who is willing to put in extra hours. Subordinate X’s current position requires some solitary work and some interaction with people inside and outside of the organization.