Volume 98, Issue 1 e12544
RESEARCH ARTICLE
Open Access

‘Distant but close’: Leadership behaviours, psychological distance, employee coping and effectiveness in remote work contexts

Anders Friis Marstand

Corresponding Author

Anders Friis Marstand

Durham University Business School, Durham, UK

Correspondence

Anders Friis Marstand, Durham University Business School, Mill Hill Lane, Durham DH1 3LB, UK.

Email: [email protected]

Contribution: Conceptualization, Methodology, Formal analysis, Project administration, Writing - original draft, ​Investigation, Writing - review & editing

Search for more papers by this author
Olga Epitropaki

Olga Epitropaki

Durham University Business School, Durham, UK

Contribution: Conceptualization, Writing - original draft, Methodology, Writing - review & editing

Search for more papers by this author
Ilias Kapoutsis

Ilias Kapoutsis

Athens University of Economics and Business, Athens, Greece

Contribution: ​Investigation, Writing - review & editing, Methodology, Formal analysis, Conceptualization

Search for more papers by this author
First published: 21 August 2024

Abstract

Drawing on construal level and conservation of resources theories, our paper focuses on the psychological distance employees experience from their manager in remote work contexts. We specifically examine the role of three leadership behaviours (initiating structure, consideration and vision communication) on employees' perceptions of psychological distance from their manager and the subsequent effects on employee task, emotion and avoidance coping and individual effectiveness outcomes. Using data from two independent studies (Study 1: a four-wave time-lagged online study of remotely working 338 participants; Study 2: a four-wave time-lagged study of 202 hybrid working professionals), we found that consideration and vision communication reduced employees' perceptions of psychological distance from their manager, while psychological distance decreased task coping. Support for a serial mediation model was also found, with consideration and vision communication indirectly influencing task performance and consideration indirectly influencing organizational citizenship behaviours and withdrawal behaviours via psychological distance and then via task coping. Our research results provide new insights into the role of leadership in remote work contexts and highlight the implications of psychological distance from the leader for employees' coping responses and individual effectiveness.

Practitioner points

  • Leaders are a key resource for employees in remote work contexts and can help them overcome the challenges of physical distance.
  • Employees experience being psychologically closer to their leader when leaders employ consideration and vision communication behaviours.
  • Leaders' consideration and vision communication behaviours are critical for helping employees engage in adaptive coping strategies and stay focused and productive in remote work contexts.

INTRODUCTION

Remote work has been on the rise and the recent pandemic accelerated the trend (e.g., Banjo et al., 2020). Studies have indicated that 42% of US workers were working from home full time during lockdown (vs. 2% pre-COVID-19), accounting for 60% of US economic activity (Bloom, 2020). At the same time, the potential impact of the pandemic on employee stress and anxiety was highlighted (American Psychological Association, 2020; Holmes et al., 2020; Trougakos et al., 2020). Post-pandemic the numbers of remote workers remain high and several organizations have opted for flexible models of working (Kossek et al., 2021), but reports indicate mixed effects of remote working on work outcomes. The increased flexibility and less time and money spent commuting may improve well-being. However, remote working has been shown to be stressful due to fatigue, greater isolation and work intensification (Cooper & Kurland, 2002; Mutebi & Hobbs, 2022; Van Zoonen & Sivunen, 2022). Work may be more flexible but after-hours work and workday span are increasing which may lead to digital overload and exhaustion (Microsoft, 2022). In such conditions, several questions for organizational leaders arise. How can they remain close to their teams when physical, in-person interaction is limited? How can they help employees cope with the challenges of remote work, remain focused on their tasks and achieve high levels of individual effectiveness? These are the questions that motivate our study.

Early scholarly work has generally considered physical distance to be an important barrier to leadership (e.g., Antonakis & Atwater, 2002; Bass, 1998; Podsakoff et al., 1984, 1996a). For example, Kerr and Jermier (1978) argued that physical distance neutralizes leadership behaviours and creates ‘circumstances in which effective leadership may be impossible’ (p. 396). In contrast, recent scholarly work has contended that communication technology can potentially help leaders overcome the challenges of physical distance (e.g., Antonakis & Atwater, 2002; Popper, 2013). Leaders can still be perceived as ‘present’ or ‘close’ in online interactions despite the physical distance.

We argue that in remote work contexts, employees' perceived psychological distance from the direct manager is an important variable to consider. Psychological distance is a key facet of physical distance in remote work (Leonardi et al., 2024) and refers to the subjective experience of distance between actors, in this case between leaders and followers (Lewandowski & Lisk, 2013). We draw from construal level theory (CLT; Liberman & Trope, 1998; Trope & Liberman, 2003, 2010), which assumes that objects, entities and people that are not present in the direct experience of a person are psychologically distant. Virtual teams and remote work contexts are considered highly relevant for CLT organizational applications as they ‘have become a common workplace phenomenon that can elicit psychological distance’ (Wiesenfeld et al., 2017, p. 389). There have also been calls for organizational research that utilizes CLT as a key theoretical framework for understanding leadership phenomena (e.g., Popper, 2013; Wiesenfeld et al., 2017).

We argue that in remote work contexts, the specific behaviours the leader exhibits during online work interactions can have important implications for employees' perceptions of their psychological distance from the leader. Prior conceptual application of CLT in the leadership domain (e.g., Berson et al., 2015; Shamir, 2013) has drawn attention to leaders' responses to both concrete, how questions, and high-level, why questions for psychological distance perceptions. Leadership behaviours that focus on both the how (i.e., concrete questions and short-term activities) and the why (i.e., high-level questions and future-oriented activities) can have important implications on how psychologically distant leaders are perceived to be (e.g., Shamir, 2013). Drawing from Yukl et al.'s (2002) tripartite taxonomy of leadership behaviours, we specifically focus on three core behaviours, namely initiating structure, consideration and vision communication. Initiating structure can help employees achieve proximal, concrete and short-term activities necessary for daily work functioning in remote working contexts and consideration involves showing continuous support and concern for employees' needs. Vision communication can further inspire employees towards longer term, distant needs, such as adapting to the wider challenges that VUCA (volatile, uncertain, complex and ambiguous) environments (Bennett & Lemoine, 2014) pose for organizational survival and functioning. By answering both how and why questions (Berson et al., 2015), these leadership behaviours can reduce the psychological distance employees experience from their manager.

We further argue that in work settings of increased physical distance, employees' psychological distance from the leader will be an important mediating mechanism of the relationship between leadership behaviours and employees' ability to cope with strain and sustain high levels of performance. Drawing from conservation of resources (COR) theory (Hobfoll, 1989), we posit that leaders are important contextual coping resources for employees in the potentially stressful remote working contexts (e.g., Costin et al., 2023; Leroy et al., 2021; Van Zoonen & Sivunen, 2022). We further suggest that psychological distance will function as a resource signal (Halbesleben et al., 2014) of the availability of the leader as a coping resource. High psychological distance from the leader signals limited resource availability in the context of remote work and further indicates to the individual that investment in active coping is less likely to have a positive effect. Thus, psychological distance will decrease active coping, such as task coping, and increase passive forms such as avoidance coping (Compas et al., 2001; Endler & Parker, 1994). Psychological distance and coping will further play a key role in employee individual effectiveness. Employees who perceive their leader as psychologically proximal and adopt active coping strategies will be more likely to stay focused on the tasks at hand and achieve high levels of performance in conditions of physical distancing (Brown et al., 2005). Given the call for research investigating how employees can be effective and perform in remote work contexts (Gajendran et al., 2015; Leonardi et al., 2024), we specifically focused on employee job performance as an outcome, which is composed of three behavioural sets, that is, task performance, organizational citizenship behaviours (OCB) and withdrawal behaviours (Campbell & Wiernik, 2015; Miner & Glomb, 2010; Rotundo & Sackett, 2002).

By integrating CLT (Trope & Liberman, 2003, 2010) and COR theory (Hobfoll, 1989) to examine the role of leadership and psychological distance for employee coping and task performance in remote work contexts, we aim to make the following contributions. First, we contribute to the literature of remote work (e.g., Bailey & Kurland, 2002; Gajendran & Harrison, 2007; Golden et al., 2008; Hill et al., 2003; Zhang et al., 2021) by focusing on psychological distance as a key facet of physical distance (Leonardi et al., 2024) and by explicitly addressing the role of leaders in follower outcomes in this context. We specifically investigate three core leadership behaviours, that is, initiating structure, consideration and vision communication (Yukl et al., 2002), as predictors of the psychological distance employees perceive from their manager due to their emphasis on concrete goals and abstract visions (Shamir, 2013). We examine psychological distance as an important mediating mechanism of the relationship between leader behaviours and employees' coping strategies in remote work contexts. We posit that perceived psychological distance signals limited availability of leader resources and discourages investment in active coping (Halbesleben et al., 2014). We investigate three types of coping and argue that psychological distance decreases task coping and increases emotion and avoidance coping. By examining a sequential mediation model, we highlight the role of psychological distance and coping strategies as important underlying mechanisms of the relationship between leadership behaviours and individual effectiveness outcomes such as task performance, OCB and withdrawal behaviours in remote work environments.

Second, we contribute to the leadership literature by addressing the ‘context deficit’ of leadership research (Johns, 2024). By examining leadership in the context of remote work, we specifically focus on physical distance as a discreet leadership context (Johns, 2006; Oc, 2018) and thus, offer a more nuanced, contextualized perspective of leadership phenomena. As Oc (2018) highlights, leadership research interest in the physical context is very recent and limited. As physical distance becomes a key contextual factor of work experiences, due to the prevalence of geographically dispersed teams and the intensification of remote work, examining key leadership processes in that specific context can offer novel and timely insights. We further contribute to research on leader distance (Antonakis & Atwater, 2002; Shamir, 2013) and answer the call for examining psychological distance and applying CLT theory in organizational research (Wiesenfeld et al., 2017). We especially highlight the importance of psychological distance from the manager in remote work contexts and further investigate specific leadership behaviours that can help employees feel closer to their manager, cope effectively and sustain high levels of performance.

Finally, our study has important practical implications, underscoring the crucial role leaders play in guiding employees towards adaptive coping strategies and maintaining their focus and productivity in remote and hybrid work settings. Even in physically distant organizational settings, leaders can still reduce psychological distance from their employees, help them cope and achieve positive individual and organizational outcomes. Given the anticipated continuation of the work-from-home and hybrid work trend (Brenan, 2020), our study provides leaders with valuable insights to effectively navigate and support employees in these evolving work environments.

THEORY AND HYPOTHESES

Remote work, leadership and psychological distance

Although remote work is not a new phenomenon, it was the COVID-19 pandemic that intensified it. Remote work (also known as telecommuting and work from home) is often offered as a flexible work arrangement to enable employees to manage work and family demands (Chong et al., 2020) but also as a way to reduce organizational costs (Conner, 2003). Post-pandemic, hybrid work (a form of remote work) has emerged as the ‘new normal’, typically involving an employee splitting their time between working at home and in the office. Past research on remote work has identified a paradox (Baruch, 2000; Ruth & Chaudhry, 2008; Zhang et al., 2021). On one hand, remote work can increase flexibility, perceived autonomy, job attitudes and performance (e.g., Gajendran & Harrison, 2007; Hill et al., 2003) but on the other hand, can lead to exhaustion, work overload, role ambiguity and professional isolation (Golden et al., 2008; Lundberg & Lindfors, 2002). It can also undermine the restorative functions of home (Hartig et al., 2007). Recent research has further pointed to an ‘interruptions landscape’ (e.g., intrusions, distractions and multi-tasking) associated with remote work during and post-pandemic with important implications for exhaustion and performance (Leroy et al., 2021). Declines in work belongingness and meaningfulness have also been reported (Afota et al., 2024).

Physical distance is a key component of remote work. In a recent review, Leonardi et al. (2024) conceptualized the physical distance of remote work as a multi-faceted construct including different types of distance such as psychological (subjective experience of distance), technological (discrepancies in technology-related features) and structural (organizational misalignments). They further argue that remote work arrangements introduce or exacerbate issues related to psychological distance. Specifically, they note that being physically apart prevents informal interactions that would normally help to develop connection and shared understandings. Employees working remotely often feel more socially isolated, as digital media cannot replicate the richness of in-person interaction, limiting the exchange of social information (Viererbl et al., 2022). Leonardi et al. (2024) further emphasize the importance of work resources from which people are distant from, namely, social (e.g., work relations), material (e.g., desks, equipment) and symbolic (e.g., formal dress code) resources. The role of managers is especially highlighted. Leaders can facilitate the access of remote workers to all three types of resources and help them overcome the challenges of physical distance.

In the leadership literature, the relationship between leadership and distance has attracted scholarly attention for almost 100 years (e.g., Bogardus, 1927; Park, 1924), but most existing scholarly work is conceptual rather than empirical (e.g., Antonakis & Atwater, 2002). Early work generally viewed leader physical distance as a barrier and claimed that it neutralizes leadership behaviours, that is, ‘make it effectively impossible for relationship and/or task-oriented leadership to make a difference’ (Kerr & Jermier, 1978, p. 395). Popper (2013) more recently argued that the discussion on leadership and distance is ‘complex and even obscure’ (p. 1) especially when considering followers' subjective assessment of distance, namely psychological distance.

In our paper, we focus on psychological distance as a key facet of physical distance in remote work. As Leonardi et al. (2024) indicate, psychological distance ‘… is dominant, although implicit, in much of the remote work research’ (p. 196) and can have important implications for people's perceptions of access to social, material and symbolic work resources. We draw from Trope and Liberman (2010) who defined psychological distance as ‘a subjective experience that something is close or far away from the self, here, and now’ (p. 440). It contains four dimensions: spatial (how distant in space the person perceives the target), temporal (how distant in time the person views the target), social distance (how socially distant the person views the target from the self) and hypotheticality (how close to reality the person views the target) (Bar-Anan et al., 2006). CLT further suggests that all types of distance are cognitively interrelated, and increasing the distance along one dimension, increases distance on others (Stephan et al., 2010).

We specifically focus on perceptions of psychological distance from the manager, who as a key resource holder, can play an important role in shaping employees' experiences of remote work. We argue that in remote working contexts, the behaviours a manager exhibits during virtual work interactions represent a primary mechanism for transmitting leadership (Eberly et al., 2013). A physically remote leader can still be perceived as psychologically proximal based on the specific behaviours the leader exhibits during online work interactions. Drawing from Yukl et al.'s (2002) taxonomy of three independent yet related leadership meta-categories (task-, relation- and change-oriented leadership), we focus on three core behaviours, representative of these meta-categories, that is, initiating structure, consideration and vision communication. Given the dearth of studies of leadership in remote work settings (Oc, 2018), we chose to go ‘back to the basics’ and focus on three ‘traditional’ leadership behaviours (Banks et al., 2018; Bormann & Rowold, 2018) that have been shown to be important in explaining variance in outcomes, such as follower attitudes and performance (DeRue et al., 2011; Judge et al., 2004).

We specifically argue that behaviours that emphasize proximal, concrete and short-term activities (initiating structure) and showing concern and kindness (consideration) can reduce psychological distance. Initiating structure focuses on managing daily complexity by clarifying expectations, prioritizing tasks, monitoring and providing feedback (Fleishman, 1973; Judge et al., 2004; Stogdill et al., 1962). Consideration behaviours describe the extent to which the ‘leader shows concern and respect for followers, looks out for their welfare, and expresses appreciation and support’ (Judge et al., 2004, p. 36). Prior meta-analyses (e.g., DeRue et al., 2011; Judge et al., 2004) have generally shown initiating structure and consideration to have moderately strong and positive relations with leadership outcomes (such as performance-related and attitudinal ones). Both initiating structure and consideration behaviours are likely to matter in the context of remote work. The structure and goal setting associated with initiating structure behaviours are important for establishing performance norms in online work interactions. Leader consideration further implies leader intimacy and can facilitate exchange of ideas and enhanced information flow (Avolio et al., 2014; Eberly et al., 2013). The leader's expression of concern can also instil positive emotions in followers that can decrease their perception of psychological distance from the leader (Van Boven et al., 2010).

Furthermore, in uncertain conditions and VUCA environments, leadership behaviours such as vision communication that address distal, long-term objectives, emphasize a common identity and offer a sense of meaning (e.g., Yagil, 1998) can also reduce psychological distance. Recent studies on virtual teams have shown visionary leadership behaviours to have positive effects in online work environments (Kahai, 2013). Joshi et al. (2009), for example, showed that inspirational leadership was strongly related to team members' commitment in geographically dispersed teams and Purvanova and Bono (2009) found vision communication to be strongly related to team performance in virtual teams. Based on the above discussion, we propose that initiating structure, consideration and vision communication behaviours can reduce the psychological distance employees experience from their manager in remote work contexts.

Hypothesis 1.(a) Initiating structure, (b) consideration and (c) vision communication are negatively related to psychological distance.

A sequential mediation model of leadership, psychological distance, coping and individual work effectiveness

We further argue that psychological distance can play an important role in employees' coping strategies and their subsequent work effectiveness in the potentially resource threatening remote work contexts (Leonardi et al., 2024). Coping is defined as ‘cognitive and behavioural efforts to manage specific external and/or internal demands that are appraised as taxing or exceeding the resources of the person’ (Lazarus & Folkman, 1984, p. 141). We view coping not as a stable trait as some prior research has examined it (e.g., Bolger, 1990; Murberg, 2009), but as a set of strategies individuals adopt to manage a stressor (Eatough & Chang, 2018). We focus on three types of coping: task, emotion and avoidance coping (Endler & Parker, 1994) that employees may employ to manage the taxing demands of remote work.

Task coping refers to purposeful, task-oriented efforts aimed at solving the problem, and maintaining concentration on the steps needed to fulfil task requirements (Brown et al., 2005). It is viewed as an active and adaptive strategy and it has been shown to have a beneficial direct effect on performance (Brown et al., 2005). In contrast, avoidance coping is generally viewed as a maladaptive strategy. Avoidance coping may involve seeking out other people (social diversion) or engaging in substitute tasks (distraction) (Endler & Parker, 1994). On the other hand, emotion coping has been viewed as both a maladaptive (Cheng, 2001) and an adaptive strategy (Baker & Berenbaum, 2007), depending to a great extent on the specific coping responses measured. Emotion-focused strategies such as positive reinterpretation and humour (Carver et al., 1989) can be adaptive whereas responses such as self-blame and self-criticism (Carver & Connor-Smith, 2010; Chang, 1998) are generally seen as maladaptive.

Despite the wealth of studies on leadership and stress (Harms et al., 2017) and well-being (e.g., Inceoglu et al., 2018), research explicitly focusing on leadership and coping strategies is surprisingly limited (e.g., Eatough & Chang, 2018). Drawing from COR theory (Hobfoll, 1989), we argue that leadership behaviours are important contextual coping resources (Halbesleben et al., 2014; ten Brummelhuis & Bakker, 2012). A key tenet of COR theory is that individuals strive to obtain, keep, foster and protect valuable resources and stress occurs when they perceive such resources to be threatened with loss (Hobfoll et al., 2018). Remote work has been identified as a potential stressor due to interruptions, isolation, digital presenteeism (Costin et al., 2023; Leroy et al., 2021; Nurmi, 2011; Song & Gao, 2020; Van Zoonen & Sivunen, 2022) and employees being distant from valued work resources (Leonardi et al., 2024). Prior research has stressed the role of social support as an important resource related to increased use of adaptive coping and decreased use of avoidance coping strategies (Holahan & Moos, 1987). Previous studies have highlighted the role of the leader as a resource and used COR theory to examine the role of transformational leadership (Braun & Peus, 2018; Stempel et al., 2023; Tafvelin et al., 2019), authentic leadership (Braun & Nieberle, 2017) and leadership competence (Mao et al., 2019) on key outcomes such as burnout and psychological safety. By enacting specific behaviours, leaders may provide important information about the abundance or scarcity of job and organizational resources to their employees. Leaders play a critical role in nurturing ‘resource caravans’, that is, bundles of work resources, and facilitating ‘resource caravan passageways’, that is, ‘the ecological conditions that either foster or block resource creation and sustenance’ in team and organizational settings (Hobfoll, 2011; Hobfoll et al., 2018, p. 106).

Leader supportive behaviours and consideration have been outlined as valuable contextual resources within the COR framework (Gorgievski et al., 2011; Halbesleben et al., 2014), and there are also studies that have proposed goal-focused and initiating structure behaviours to be important resources (Perry et al., 2010). By providing clear structure, concrete goals and objectives, initiating structure behaviours can reduce uncertainty and perceived resource threat, minimize the loss of employee cognitive and emotional resources and contribute to adaptive reactions to remote work-related stressors. Scholarly work has further pointed to the resource function of charismatic behaviours such as vision communication (e.g., Lepine et al., 2016). By helping employees envision an attractive future and frame a challenging situation positively, visionary leadership can also contribute to followers' adaptive coping responses.

Building on the above discussion, we further argue that psychological distance is an important mediating mechanism of the relationship between leadership and coping. In remote work contexts, the salience of loss or limited access to potential work resources, such as social, material and symbolic ones (Leonardi et al., 2024), can motivate individuals to seek external cues and signals regarding coping resource investment (Halbesleben et al., 2014). Psychological distance can function as an important resource signal of the availability of leader resources and can shape coping responses, suggesting that psychological distance mediates the relationship between leadership behaviours and coping strategies. As psychological distance from the leader is reduced due to the behaviours the leader exhibits in remote work contexts, the stress related to the salience of resource loss threat may be minimized and the value of investment in adaptive coping responses will become apparent. As a result, task coping will increase, and maladaptive emotion coping and avoidance coping will decrease. Prior CLT research has also shown that people are less likely to procrastinate performing a task when psychological distance is low (McCrea et al., 2008). Thus, employees who feel psychologically close to the leader will be more likely to remain focused on goal pursuit and concentrate on tasks communicated by the leader and less likely to self-blame or seek social diversion and distraction.

Furthermore, coping is important for job performance and individual effectiveness at work. Individual effectiveness has been conceptualized as ‘the tendency to contribute desirable inputs towards one's work role’ (Harrison et al., 2006, p. 309) and it consists of three behavioural sets, that is, task performance, OCB and withdrawal behaviours (Campbell & Wiernik, 2015; Miner & Glomb, 2010). In remote work settings, employees often lose access to important resources at work (Zhang et al., 2021) and experience challenges in managing interpersonal relationships with coworkers and coordinating complex tasks, hampering their task performance (Golden et al., 2008). Thus, understanding the role of leadership behaviours as an important resource for employees to perform their tasks is critical in the context of remote work. Effective coping enables individuals to resolve problems, stay focused on assigned goals and perform their tasks (Brown et al., 2005). Task coping has previously been linked with positive task performance (e.g., Casper et al., 2017) whereas maladaptive emotion coping and avoidance coping have been associated with negative outcomes (Baker & Berenbaum, 2007). If a negative event disrupts goal-directed behaviour and the individual instead seeks ways to strongly express emotion, self-blame or disassociate oneself from the event via diversion and distraction, performance is likely to suffer (Weiss & Cropanzano, 1996). Maladaptive emotion-coping strategies such as self-blame and wishful thinking remove the focus on the task at hand (Carver & Connor-Smith, 2010; Chang, 1998) and therefore, should reduce task performance. Engagement in avoidance coping strategies requires using resources (e.g., cognitive attention and time) on other activities that are not related to the work task and therefore, from a COR theory perspective, employees will further deplete their resources when using avoidance coping strategies eventually reducing their task performance (Nandkeolyar et al., 2014).

Taken together, the above arguments suggest sequential mediation. Initiating structure, consideration and vision communication behaviours are expected to decrease psychological distance by emphasizing concrete goals, attention to employee needs and abstract visions. Reduced psychological distance further signals availability of leader resources (such as social, material and symbolic) and will mediate the relationship between leadership behaviours and coping. Employees who perceive their leader to be psychologically proximal will invest in task coping, which will have a beneficial impact on their task performance. They will also be less likely to resort to maladaptive emotion-coping and avoidance coping strategies that can potentially hamper their task performance.

Hypothesis 2.Initiating structure, consideration and vision communication are positively and indirectly related to task performance via psychological distance, and then via (a) task coping, (b) emotion coping and (c) avoidance coping.

Employee coping may also influence the extent to which they will engage in extra-role performance behaviours, such as OCB, which refers to discretionary employee actions that benefit the organization (Organ, 1997; Podsakoff et al., 2000). It involves complying with the organization's norms and values, and going beyond one's job description to contribute to effective organizational functioning (LePine et al., 2002). We specifically focus on citizenship behaviours directed towards other individuals in the organization as remote working involves less social interaction with colleagues resulting in employees being less available to help other colleagues when support is suddenly needed (Gajendran et al., 2015). Employees' engagement in OCB when working remotely has also been shown to be lower than when working in the office (Smith et al., 2020). With leader behaviours playing an important role in connecting employees with the organization and reminding them of helping colleagues (Lautsch et al., 2009), leadership behaviours should arguably matter for employee OCB in remote contexts. Leader behaviours such as initiating structure, consideration and vision communication that reduce psychological distance, increase task coping and reduce emotional and avoidance coping will signal availability of leader resources and may increase discretionary behaviours such as OCB in remote work settings. Engaging in OCB requires expenditure of personal resources as individuals go over and above their job description requirements. Thus, employees who perceive high psychological distance from their manager will be reluctant to deplete personal resources by investing energy in discretionary behaviours. We further expect adaptive coping strategies such as task coping to be associated with high levels of OCB as task coping aimed at solving the problem will not only help employees perform their own tasks but also give them surplus to help colleagues (Brown et al., 2005). On the other hand, when pursuing maladaptive emotional strategies such as self-blaming, employees redirect their focus from the task to themselves and pay less attention to the needs of colleagues consequently reducing OCB. Similarly, when engaging in avoidance strategies, they concentrate on non-work-related tasks (Nandkeolyar et al., 2014), which limits the resources they can spend on colleagues eventually reducing OCB.

Hypothesis 3.Initiating structure, consideration and vision communication are positively and indirectly related to OCB via psychological distance, and then via (a) task coping, (b) emotion coping and (c) avoidance coping.

We further expect psychological distance and coping strategies to impact employee withdrawal behaviours (Spector et al., 2006), which represent another dimension of job performance (Rotundo & Sackett, 2002). Withdrawal behaviours involve removing oneself temporarily from a challenging work situation (Hanisch & Hulin, 1991). They are often viewed as individual attempts to get relief from work tasks without engaging in extreme behaviours such as quitting (Miner & Glomb, 2010). Withdrawal behaviours include tardiness, absenteeism and reducing the amount of time spent working for the organization (Spector et al., 2006). When working remotely, employees feel less connected to the organization and they are less visible to the leader, which increases the likelihood of employees withdrawing and spending time on other activities (Burbano & Chiles, 2022; Chong et al., 2020). Particularly when employees view the leader as being psychologically distant and the leader provides limited support, they feel more socially disconnected and are more likely to withdraw from work (Carsten et al., 2022). Past research has indicated that stressful situations may lead employees to engage in withdrawal behaviours (Welbourne & Sariol, 2017) and employees may withdraw from work to prevent further leakage of valuable resources. Reduced psychological distance from the leader signals availability of resources, and thus, employees will be less likely to withdraw to protect personal resources. High levels of task coping help employees focus on tasks and reduce withdrawal behaviours whereas emotion and avoidance coping that aim to enable employees to escape or limit their exposure to stressful situations (Nandkeolyar et al., 2014; Welbourne & Sariol, 2017) will be positively related to withdrawal behaviours.

Hypothesis 4.Initiating structure, consideration and vision communication are negatively and indirectly related to withdrawal behaviours via psychological distance, and then via (a) task coping, (b) emotion coping and (c) avoidance coping.

OVERVIEW OF THE STUDIES

We conducted two studies to test our sequential mediation propositions. The two studies were designed to examine the robustness of the findings in different contexts that differed in terms of remote and hybrid working. In Study 1, we collected data (N = 338) from employees in the United States who worked remotely during the first COVID-19 lockdown. In Study 2, data were collected from employees of a European country (N = 202) who worked both in the office and from home (i.e., hybrid) post-pandemic to constructively replicate the findings of the first study. We particularly studied these two remote work contexts to explore if our model would replicate across contexts. Due to the key difference of the two contexts—with one focusing on employees working remotely during the first COVID-19 lockdown, and the other on employees working in hybrid mode post-pandemic—we used the same design across the two studies in order to avoid confounding effects related to using different designs. To better capture the temporal order of the proposed relationships (Mitchell & Maxwell, 2013) and to reduce potential common method bias (Podsakoff et al., 2024), we used a four-wave time-lagged design with a 1-week lag between waves (Time 1 to Time 4). Before the first wave, participants completed a screening survey (Time 0) to ensure that respondents invited for the four waves matched our criteria for each study. The sequential mediation design, in which the predictor, mediator and outcome variables are measured at different time points, offers a more rigorous test of the hypothesized processes compared to cross-sectional mediation designs (Mitchell & Maxwell, 2013). Ethical approval for the studies was granted by the first author's university's institutional review board (IRB). Informed consent was obtained from respondents in both studies.

STUDY 1

Method

Sample and procedure

Data were collected in the United States during the first COVID-19 lockdown in 2020. Using an online screening survey, we recruited a sample of 1552 US-based Amazon Mechanical Turk (MTurk) workers at Time 0. We only requested that respondents worked in an organization and reported to a manager in the organization. Following Chmielewski and Kucker (2020), we used response validity indicators to remove low-quality data. We included respondents who were full-time employed, worked in an organization, reported to a manager in the organization and had worked in their job and with their manager for more than 1 month. Furthermore, due to the study's focus on remote work, we were only interested in including those who worked remotely. After excluding participants who stated that they were working in the office and those who indicated that they had on average zero minutes of online meetings with their manager per week over the past 2 months, the final sample at Time 0 was reduced to 697 respondents, who were invited to complete the Time 1 survey. For Time 1 to Time 4 we also used various attention, content and consistency checks and checked that respondents worked in the same organization and job and with the same manager at Time 0 to Time 4, which reduced the final sample at Time 4 to 338 who worked exclusively remotely (average response rate across Times 1–4 was 89.76%).

Forty per cent of the respondents were women. On average, respondents were 39.85 years old (SD = 10.71), had worked in their job for 6.62 years (SD = 5.90) and had worked with their manager for 4.19 years (SD = 3.80). The respondents worked in different functional areas including accounting and finance (14.8%), administration (8.6%), arts and design (3.3%), education (13.3%), engineering (5.3%), IT (19.8%), management (5.9%), marketing, sales and business development (10.4%), operations (9.5%) and other functional areas (9.2%).

Measures

Leadership behaviours, psychological distance and task performance items were rated on a seven-point scale ranging from 1 (strongly disagree) to 7 (strongly agree) and coping items were rated on a seven-point scale ranging from 1 (never) to 7 (always).

Initiating structure (Time 1)

We used six initiating structure items from the LBDQ-XII (Stogdill, 1963) that reflected task-focused behaviours, following Rosen et al. (2019). A sample item is ‘My manager schedules the work to be done’ (α = .90).

Consideration (Time 1)

We used four consideration items from the LBDQ-XII (Stogdill, 1963) that reflected supportive and caring behaviours, similarly to Johnson et al. (2012). One sample item is ‘My manager is friendly and approachable’ (α = .91).1

Vision communication (Time 1)

We used the five ‘articulating a vision’ items from Podsakoff et al. (1996b). One sample item is ‘My manager paints an interesting picture of the future for our work team’ (α = .95).

Psychological distance (Time 2)

We adapted spatial, temporal and social distance items from Lim et al. (2012) and developed four hypotheticality items based on CLT research (Bar-Anan et al., 2006; Trope & Liberman, 2003) to measure psychological distance (see Appendix 1 for items and confirmatory factor analyses (CFA)). Cronbach's alpha for the composite measure of psychological distance was .95.

Coping (Time 3)

We measured task, emotion and avoidance coping with the 24-item Coping Inventory for Stressful Situations (CISS) scale by Endler and Parker (1990). Sample items are ‘I schedule my time better’ (task coping, α = .91), ‘I blame myself for not knowing what to do’ (emotion coping, α = .90) and ‘I watch TV’ (avoidance coping, α = .80).

Task performance (Time 4)

We used the seven-task performance items (Williams & Anderson, 1991). A sample item is ‘I adequately complete assigned duties’ (α = .87).

Control variables

At Time 0, we controlled for respondents' experiences of workload changes over the last 2 months (less workload than before, the same workload or more workload than before) as excessive workload can affect stress amplifying the need to engage in coping strategies to manage the psychological distress (Kammeyer-Mueller et al., 2009). Following previous coping studies (e.g., Mawritz et al., 2014), we controlled for respondents' age, which was significantly correlated with coping and task performance. We also controlled for employees' tenure with their manager (measured in years and months at Time 0) as we examined employees' perceptions of their leader and their psychological distance to leader and as tenure with the manager is often used as a control variable in leadership studies (e.g., Palanski & Yammarino, 2011). As we examined employees' perceptions of their online interactions with their manager and their psychological distance to their manager during their online work interactions, we controlled for the amount of time (measured in hours and minutes) of online meetings employees had had with their manager per week over the past 2 months (measured at Time 0). Furthermore, at Time 4 we measured the amount of time of online meetings employees had had with their manager per week since completing the Time 0 survey and used this as a control variable of task performance Time 4.

We also collected data on some variables (e.g., gender and number of days respondents had been social distancing), which we did not eventually use as control variables as they were not significantly correlated with the outcomes (Becker, 2005).

Analytical strategy

First, we conducted confirmatory factor analyses (CFAs) and compared different measurement models to test the independence of the measures. We used item parcelling for coping and task performance to increase the sample size to indicator ratio. For each of these constructs we combined two or three items into three parcels using the factorial algorithm method (Landis et al., 2000; Rogers & Schmitt, 2004). For psychological distance we used four parcels that represented the four dimensions. We tested our hypotheses using structural equation modelling (SEM) and examined different structural models. All CFAs and SEMs were tested with maximum likelihood estimation in R using the Lavaan package (Rosseel, 2012). To estimate the sequential indirect effects, we used 10,000 bootstrap samples and 95% bias-corrected confidence intervals (CIs). We tested our models both with and without controls.

Results

Confirmatory factor analyses

We tested three different measurement models. In Model 1, items and parcels of each construct loaded onto their respective factor and factors were correlated. In Model 2, leadership behaviour items loaded together onto one factor, psychological distance parcels loaded on one factor, coping parcels together loaded on one factor, task performance parcels loaded onto another factor and factors were allowed to correlate. In Model 3, all items and parcels loaded on a single factor. The results suggest that Model 1 provided a good fit to the data, χ2(406) = 935.77, p < .001, comparative fit index (CFI) = .94, Tucker–Lewis index (TLI) = .93, root mean square error of approximation (RMSEA) = .06 and standardized root mean square residual (SRMR) = .06. On the contrary, Model 2 did not fit the data well, χ2(428) = 2682.47, p < .001, CFI = .73, TLI = .71, RMSEA = .13 and SRMR = .11, just as Model 3 did not provide a good fit to the data, χ2(434) = 4443.24, p < .001, CFI = .53, TLI = .49, RMSEA = .17 and SRMR = .14. Overall, Model 1 fitted the data much better than Model 2 (Δχ2 = 1746.70, Δdf = 22 and p < .001) and Model 3 (Δχ2 = 3507.47, Δdf = 28 and p < .001), which suggests that the scales reflect distinct yet related constructs.

Hypotheses testing

Descriptive statistics, correlations and reliabilities are presented in Table 1. Workload changes were excluded from subsequent analyses as the control variable was generally not significantly correlated with outcomes (Becker, 2005). To test our hypotheses, we examined three nested structural models. Model 1 is our hypothesized model, which is a fully mediated model. Model 2, which includes the same paths as proposed in the hypothesized model plus direct paths from leadership to task performance, builds on prior research that has shown task-, relation- and change-oriented leadership behaviours to directly affect job performance (DeRue et al., 2011). Finally, Model 3 is a partially mediated model, which includes all paths in the hypothesized model plus all direct paths from leadership behaviours and psychological distance to coping and task performance. This model also takes into account that leadership behaviours as resources may affect coping strategies and directly increase adaptive coping strategies and decrease maladaptive strategies (Ito & Brotheridge, 2003) and that psychological distance can directly affect job performance (Antonakis & Atwater, 2002; Leonardi et al., 2024). Although Model 1 provided a good fit to the data (χ2(419) = 965.97, p < .001, CFI = .94, TLI = .93, RMSEA = .06 and SRMR = .07), the fit was worse than for Model 3 (χ2(406) = 935.77, p < .001, CFI = .94, TLI = .93, RMSEA = .06 and SRMR = .06) as indicated by the significant chi-square difference test results (Δχ2 = 30.20, Δdf = 13 and p = .004). Model 2 (χ2(416) = 953.51, p < .001, CFI = .94, TLI = .93, RMSEA = .06 and SRMR = .06), on the other hand, provided a better fit to the data than Model 3 based on the non-significant chi-square difference test between Model 2 and Model 3 (Δχ2 = 17.74, Δdf = 10 and p = .060). Results of analyses with control variables followed the results of analyses without controls and did not lead to any different conclusions due to similar model fit indices and similar chi-square difference test results between Models 1 and 3 (Δχ2 = 27.91, Δdf = 13 and p = .009) and Models 2 and 3 (Δχ2 = 17.62, Δdf = 10 and p = .062). Unstandardized direct effects and standard errors for Model 2 without and with controls are presented in Figure 1. Results of the tests of the indirect effects in Model 2 without and with controls are shown in Table 2.

TABLE 1. Means, standard deviations, correlations and reliabilities of variables (Study 1).
Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
1. Workload changes (T0) 1.78 0.70
2. Age (T0) 39.85 10.71 .06
3. Manager tenure in years (T0) 4.19 3.80 −.09 .39
4. Online time with manager (T0) 3.75 3.86 −.05 −.01 .05
5. Initiating structure (T1) 5.25 1.17 −.07 .04 .19 .16 (.90)
6. Consideration (T1) 5.36 1.36 .05 .03 .19 .11 .61 (.91)
7. Vision communication (T1) 4.99 1.40 .00 .03 .23 .17 .64 .77 (.95)
8. Overall PD (T2) 3.34 1.22 −.01 −.06 −.22 −.21 −.53 −.65 −.65 (.95)
9. PD—spatial (T2) 3.82 1.57 −.02 −.08 −.19 −.24 −.37 −.45 −.46 .87 (.92)
10. PD—temporal (T2) 3.20 1.28 .00 −.10 −.20 −.15 −.48 −.57 −.58 .87 .70 (.78)
11. PD—social (T2) 3.58 1.58 −.03 .02 −.16 −.19 −.48 −.68 −.66 .87 .63 .69 (.91)
12. PD—hypotheticality (T2) 2.79 1.21 .02 −.05 −.20 −.13 −.54 −.59 −.61 .86 .59 .70 .74 (.90)
13. Task coping (T3) 5.17 0.94 −.01 .16 .22 .14 .30 .30 .37 −.34 −.21 −.29 −.33 −.37 (.91)
14. Emotion coping (T3) 2.88 1.22 −.06 −.31 −.14 .02 −.04 −.11 −.14 .09 .01 .10 .13 .12 −.25 (.90)
15. Avoidance coping (T3) 4.20 0.93 −.11 −.11 .06 .12 .20 .16 .20 −.25 −.21 −.17 −.25 −.22 .27 .25 (.80)
16. Task performance (T4) 6.35 0.75 .03 .31 .20 −.04 .21 .18 .15 −.09 .01 −.13 −.04 −.19 .42 −.38 −.06 (.87)
17. Online time with manager (T4) 3.34 3.34 .03 −.01 .03 .38 .04 .01 .08 −.11 −.12 −.05 −.11 −.09 .10 .10 .13 −.12
  • Note: N = 338. Within rounding error, correlations greater than .106 in absolute magnitude are significant at p = .05; correlations greater than .139 in absolute magnitude are significant at p = .01 (both two-tailed test). Cronbach's alpha coefficients are reported on the diagonal. T0 = screening; T1 = Time 1; T2 = Time 2; T3 = Time 3; T4 = Time 4.
  • Abbreviation: PD, psychological distance.
Details are in the caption following the image
Results of testing the sequential mediation Model (Study 1). N = 338 individuals. T1 = Time 1; T2 = Time 2; T3 = Time 3; T4 = Time 4. Unstandardized path estimates are reported. Standard errors are presented in parentheses. For each path, numbers in the first and second row show results of analyses without and with controls, respectively. The dashed lines indicate non-significant relationships. *p < .05. **p < .01.
TABLE 2. Results of testing the sequential mediation Model 3 (Study 1).
Indirect effects B a SEa CIa B b SEb CIb
Initiating structure → PD → task coping → TP .022 .015 −.003, .059 .017 .014 −.006, .046
Initiating structure → PD → emotion coping → TP .005 .004 .000, .018 .003 .003 .000, .014
Initiating structure → PD → avoidance coping → TP −.005 .005 −.020, .000 −.003 .004 −.016, .001
Consideration → PD → task coping → TP .079 .025 .039, .137 .072 .023 .036, .128
Consideration → PD → emotion coping → TP .017 .010 .002, .044 .013 .009 .001, .038
Consideration → PD → avoidance coping → TP −.017 .011 −.048, −.003 −.014 .010 −.042, −.001
Vision communication → PD → task coping → TP .074 .027 .035, .140 .061 .024 .026, .119
Vision communication → PD → emotion coping → TP .016 .010 .001, .043 .011 .008 .001, .036
Vision communication → PD → avoidance coping → TP −.016 .011 −.047, −.002 −.012 .009 −.040, −.001
  • Note: N = 338 individuals. Unstandardized estimates (B), standard errors (SE) and bias-corrected confidence intervals (CI) are reported. CIs were estimated using 10,000 bootstrap samples. 95% CIs are reported for indirect effects. Column 1a (Column 2b) results are based on analyses without (with) control variables. We omitted estimates of control variables for brevity.
  • Abbreviations: PD, psychological distance; TP, task performance.

Hypothesis 1 stated that initiating structure, consideration and vision communication would be negatively related to psychological distance. As shown in Figure 1, Hypothesis 1a was not supported whereas Hypotheses 1b and 1c were supported. We further proposed that the three leadership behaviours would be positively and indirectly related to task performance via psychological distance, and then via coping strategies. Hypotheses 2a, 2b and 2c were not supported for initiating structure as the indirect effects on task performance via psychological distance, and then via coping were not significant. Hypotheses 2a and 2b were supported for consideration as the indirect effects on task performance via psychological distance, and then via task coping (estimate = .079, 95% CI [0.039, 0.137]) and via emotion coping (estimate = .017, 95% CI [0.002, 0.044]) were positive and significant providing support for sequential mediation. Similarly, we found support for Hypotheses 2a and 2b for vision communication as the indirect effects on task performance via psychological distance, and then via task coping (estimate = .074, 95% CI [0.035, 0.140]) and via emotion coping (estimate = .016, 95% CI [0.001, 0.043]) were positive and significant. While we proposed that the indirect effects of consideration and vision communication on task performance via psychological distance, and then via avoidance coping would be positive we found that they were negative and significant, meaning that Hypothesis 2c was not supported for consideration and vision communication. Overall, our results indicate that psychological distance together with task coping act as important mediating mechanisms that link consideration and vision communication to task performance.

STUDY 2

The purpose of our second study was to extend the results of our first study by examining the hypothesized relationships in a remote work context post-pandemic. Based on construal level theory and COR theory, we expect the previously presented pattern of relationships between leadership behaviours, psychological distance, coping strategies and individual effectiveness to replicate in a sample of employees with more flexible work patterns in a hybrid work context.

Method

Sample and procedure

Seventy-eight students from a European university were assigned to invite employees to participate in the study. Participants had to work in an organization, report to a manager and have some remote working in their current job. Students received class credit in exchange. Like in Study 1, respondents were invited to complete an initial online screening survey (Time 0) with demographic questions and then complete a questionnaire at four time points with a 1-week interval after completing the screening survey. We received 278 completed responses at Time 0 and 243 responses at Time 4 (average response rate across Times 1–4 was 96.70%). To participate, respondents had to provide their work email, name and contact details, including a mobile number. We informed participants that, for quality assurance, we might contact them to confirm their survey participation. All contacted participants confirmed their survey participation and verified their email addresses.

Our final sample included respondents who passed our attention check, indicated that they worked in the same organization and job and with the same manager at Time 0 and Time 4, and had worked in their job and with their manager for 1 month or more. We excluded those who only worked in the office and had no remote work. This reduced our final sample at Time 4 to 202 respondents, who were 40.32 years old (SD = 12.15), had worked in their job for 6.34 years (SD = 7.10) and had worked with their manager for 4.28 years (SD = 5.26). Of the 202 respondents, 58.4% were women. The respondents worked in different functional areas, including accounting and finance (28.2%), administration (2.0%), education (1.0%), engineering (2.5%), IT (10.9%), management (23.3%), marketing, sales and business development (4.5%), operations (1.0%) and other functional areas (26.7%).

Measures

We used the same measures and response scales as used in Study 1 to measure initiating structure (Time 1, α = .78), consideration (Time 1, α = .90), vision communication (Time 1, α = .91), psychological distance (Time 2, α = .92), task coping (Time 3, α = .78), emotion coping (Time 3, α = .85), avoidance coping (Time 3, α = .84) and task performance (Time 4, α = .83). Furthermore, we measured individual-targeted OCB (Time 4) with the seven-item scale by Williams and Anderson (1991). Items were rated on a seven-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). A sample item is ‘I help others who have been unwell’ (α = .82). Furthermore, we measured psychological withdrawal using eight items from Lehman and Simpson (1992). Items were rated on a seven-point scale ranging from 1 (never) to 7 (very often). A sample item is ‘Spend work time on personal matters’ (α = .80). We also measured three items (e.g., ‘I like the colour blue’) from the attitudes towards the colour blue scale (Miller & Chiodo, 2008) (Time 1, α = .79) to use these for a marker variable. Items were rated on a seven-point scale ranging from 1 (strongly disagree) to 7 (strongly agree). We followed the approach in Study 1 and controlled for respondents' experiences of workload changes over the past week (less workload than before, the same workload or more workload than before), respondent age, tenure with their manager and online time with their manager at Times 1 and 4. In Study 2, we also controlled for face-to-face time with the manager as employees working in hybrid mode could also meet with their manager in person.

Results

Confirmatory factor analyses

For the measurement models, we followed the structure of the measurement models in Study 1 and included OCB and psychological withdrawal. In Model 1, items and parcels of constructs loaded onto their respective factor and factors were correlated. In Model 2, leadership behaviour items loaded together onto one factor, psychological distance parcels loaded on one factor, coping parcels together loaded on one factor, task performance and OCB parcels loaded onto another factor, psychological withdrawal parcels loaded on one factor and factors were allowed to correlate. In Model 3, all items and parcels loaded on a single factor. Model 1 provided a reasonable fit to the data, χ2(584) = 1165.6, p < .001, CFI = .87, TLI = .85, RMSEA = .07 and SRMR = .07. Model 2 did not fit the data well, χ2(619) = 2109.01, p < .001, CFI = .66, TLI = .63, RMSEA = .11 and SRMR = .10, just as Model 3 did not provide a good fit to the data, χ2(629) = 3351.67, p < .001, CFI = .38, TLI = .34, RMSEA = .15 and SRMR = .15. Overall, Model 1 fitted the data much better than Model 2 (Δχ2 = 943.41, Δdf = 35 and p < .001) and Model 3 (Δχ2 = 2186.07, Δdf = 45 and p < .001) suggesting that the scales reflect different yet related constructs.

Despite the time-lagged design, the constructs were measured by the same source and as such common method variance (CMV) could be a concern. To test for the presence of CMV, we used the CFA marker technique (Williams et al., 2010) and conducted CFAs in which we included our main variables and marker variable. The baseline model, which did not include factor loadings from the marker variable to the substantive variables (χ2(700) = 1304.16, p < .001) was compared with the CMV model, which included factor loadings from the marker variable to the substantive variables (χ2(699) = 1303.23, p < .001). The non-significant chi-square difference (Δχ2 = 0.93, Δdf = 1, ns) indicates that CMV is unlikely to be a biasing factor.

Hypotheses testing

Means, standard deviations, correlations and reliabilities are shown in Table 3. Following the approach used in Study 1, we included OCB and psychological withdrawal and tested three nested structural models. Model 1 provided a reasonable fit to the data (χ2(605) = 1211.66, p < .001, CFI = .86, TLI = .85, RMSEA = .07 and SRMR = .08), but the fit was worse than for Model 3 (χ2(584) = 1165.60, p < .001, CFI = .87, TLI = .85, RMSEA = .07 and SRMR = .07) which was also indicated by the significant chi-square difference test results (Δχ2 = 46.06, Δdf = 21 and p = .001). Model 2 (χ2(596) = 1185.50, p < .001, CFI = .87, TLI = .85, RMSEA = .07 and SRMR = .07), on the other hand, provided a better fit to the data than Model 3 based on the non-significant chi-square difference test between Model 2 and Model 3 (Δχ2 = 19.90, Δdf = 12 and p = .069). Results of analyses with control variables followed the results of analyses without controls and did not lead to any different conclusions due to similar model fit indices and similar chi-square difference test results between Models 1 and 3 (Δχ2 = 43.22, Δdf = 21 and p = .003) and Models 2 and 3 (Δχ2 = 21.00, Δdf = 12 and p = .050). Unstandardized direct effects and standard errors for Model 2 without and with controls are presented in Figure 2. Table 4 shows results of the tests of the indirect effects in Model 2 without and with controls. We also tested our structural Model 2 using the marker technique. Results of analyses without controls for the baseline model (χ2(712) =1323.89 and p < .001) compared with the CMV model (χ2(711) = 1323.27 and p < .001) showed that CMV is unlikely to be a biasing factor (Δχ2 = 0.62, Δdf = 1, ns). Overall, conclusions were the same with controls (Δχ2 = 0.64, Δdf = 1, ns).

TABLE 3. Means, standard deviations, correlations and reliabilities of variables (Study 2).
Variables M SD 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
1. Workload changes (T0) 2.24 0.54
2. Age (T0) 40.32 12.15 −.10
3. Manager tenure in years (T0) 4.28 5.26 −.13 .42
4. Online time with manager (T0) 3.48 3.99 .05 −.01 .00
5. F2F time with manager (T0) 3.32 5.41 .08 .03 .09 .43
6. Initiating structure (T1) 5.01 0.94 −.09 .07 .04 .06 .13 (.78)
7. Consideration (T1) 5.43 1.21 .03 −.07 −.03 .08 .11 .41 (.90)
8. Vision communication (T1) 4.96 1.18 .01 .03 −.01 .14 .10 .54 .68 (.91)
9. Blue colour (T1) 4.75 1.15 −.05 .04 .03 .00 −.04 .03 .03 .03 (.79)
10. Overall PD (T2) 2.98 0.91 −.02 .04 −.03 −.04 −.22 −.31 −.54 −.50 −.10 (.92)
11. PD—spatial (T2) 2.82 1.12 −.04 .05 −.01 −.04 −.21 −.20 −.35 −.33 .01 .82 (.90)
12. PD—temporal (T2) 2.88 0.98 .06 .02 .00 −.02 −.17 −.29 −.41 −.37 −.13 .83 .65 (.74)
13. PD—social (T2) 3.41 1.32 −.06 .02 −.06 −.09 −.20 −.22 −.58 −.45 −.16 .82 .50 .58 (.88)
14. PD—hypotheticality (T2) 2.91 1.03 .00 .05 −.03 .03 −.15 −.33 −.44 −.49 −.08 .83 .51 .60 .63 (.86)
15. Task coping (T3) 5.38 0.68 −.02 .09 .09 .02 .05 .06 .08 .07 .04 −.20 −.20 −.13 −.12 −.19 (.78)
16. Emotion coping (T3) 3.35 0.98 −.03 .01 −.07 .02 .01 .08 −.14 −.02 .14 .07 .10 .02 .04 .05 −.33 (.85)
17. Avoidance coping (T3) 3.22 1.05 .06 −.15 −.12 .03 .06 .05 .04 .06 .10 −.03 −.04 −.03 −.02 .00 −.14 .31 (.84)
18. Task performance (T4) 6.13 0.69 −.03 .09 .02 .14 .15 .07 .15 .12 .07 −.29 −.24 −.25 −.20 −.27 .44 −.26 −.01 (.83)
19. OCB (T4) 5.83 0.72 .03 .12 .05 .22 .20 −.07 .09 .15 −.07 −.13 −.17 −.07 −.13 −.05 .24 −.10 .05 .43 (.82)
20. Psychological withdrawal (T4) 2.52 0.85 −.07 −.25 −.18 −.12 −.09 −.17 −.16 −.15 .02 .26 .28 .18 .19 .20 −.32 .29 .10 −.41 −.21 (.80)
21. Online time with manager (T4) 7.98 7.32 .03 −.03 −.01 .49 .24 −.02 .10 .15 .05 −.02 −.03 .01 −.05 .02 .09 .03 .04 .08 .23 −.08
  • Note: N = 202. Within rounding error, correlations greater than .137 in absolute magnitude are significant at p = .05; correlations greater than .179 in absolute magnitude are significant at p = .01 (both two-tailed test). Cronbach's alpha coefficients are reported on the diagonal. T0 = screening; T1 = Time 1; T2 = Time 2; T3 = Time 3; T4 = Time 4.
  • Abbreviations: F2F, face-to-face; PD, psychological distance.
Details are in the caption following the image
Results of testing the sequential mediation Model (Study 2). N = 202 individuals. T1 = Time 1; T2 = Time 2; T3 = Time 3; T4 = Time 4. Unstandardized path estimates are reported. Standard errors are presented in parentheses. For each path, numbers in the first and second row show results of analyses without and with controls, respectively. The dashed lines indicate non-significant relationships. *p < .05. **p < .01.
TABLE 4. Results of testing the sequential mediation Model 3 (Study 2).
Indirect effects B a SEa CIa B b SEb CIb
Initiating structure → PD → task coping → TP .015 .023 −.014, .085 .012 .023 −.019, .078
Initiating structure → PD → emotion coping → TP .001 .004 −.001, .021 .001 .004 −.001, .022
Initiating structure → PD → avoidance coping → TP .001 .004 −.001, .020 .000 .004 −.002, .016
Consideration → PD → task coping → TP .066 .037 .014, .168 .061 .038 .007, .161
Consideration → PD → emotion coping → TP .005 .010 −.005, .044 .006 .011 −.005, .048
Consideration → PD → avoidance coping → TP .003 .009 −.007, .033 .001 .009 −.011, .031
Vision communication → PD → task coping → TP .021 .024 −.011, .097 .025 .026 −.005, .104
Vision communication → PD → emotion coping → TP .002 .004 −.001, .023 .003 .005 −.001, .029
Vision communication → PD → avoidance coping → TP .001 .004 −.002, .021 .001 .004 −.003, .019
Initiating structure → PD → task coping → OCB .009 .014 −.008, .053 .006 .013 −.009, .047
Initiating structure → PD → emotion coping → OCB .000 .002 −.001, .010 .000 .002 −.001, .015
Initiating structure → PD → avoidance coping → OCB .000 .003 −.001, .013 .000 .002 −.002, .009
Consideration → PD → task coping → OCB .037 .024 .008, .112 .031 .023 .003, .100
Consideration → PD → emotion coping → OCB .001 .006 −.006, .024 .002 .008 −.005, .036
Consideration → PD → avoidance coping → OCB .002 .006 −.005, .024 .001 .007 −.008, .021
Vision communication → PD → task coping → OCB .012 .015 −.005, .060 .013 .015 −.002, .069
Vision communication → PD → emotion coping → OCB .000 .003 −.001, .012 .001 .004 −.001, .020
Vision communication → PD → avoidance coping → OCB .001 .003 −.001, .014 .000 .003 −.003, .016
Initiating structure → PD → task coping → PW −.008 .012 −.045, .007 −.006 .011 −.042, .008
Initiating structure → PD → emotion coping → PW −.002 .006 −.030, .003 −.002 .007 −.033, .003
Initiating structure → PD → avoidance coping → PW .000 .002 −.008, .001 .000 .002 −.009, .002
Consideration → PD → task coping → PW −.034 .022 −.098, −.006 −.029 .022 −.096, −.003
Consideration → PD → emotion coping → PW −.009 .017 −.064, .009 −.011 .019 −.072, .010
Consideration → PD → avoidance coping → PW .000 .004 −.014, .005 −.001 .006 −.018, .007
Vision communication → PD → task coping → PW −.011 .013 −.060, .004 −.012 .013 −.064, .001
Vision communication → PD → emotion coping → PW −.003 .007 −.037, .002 −.005 .010 −.045, .003
Vision communication → PD → avoidance coping → PW .000 .002 −.009, .001 .000 .003 −.013, .002
  • Note: N = 202 individuals. Unstandardized estimates (B), standard errors (SE) and bias-corrected confidence intervals (CI) are reported. CIs were estimated using 10,000 bootstrap samples. 95% CIs are reported for indirect effects. Column 1a (Column 2b) results are based on analyses without (with) control variables. We omitted estimates of control variables for brevity.
  • Abbreviations: PD, psychological distance; PW, psychological withdrawal; TP, task performance.

As shown in Figure 2, the direct effect of initiating structure and vision communication on psychological distance was not significant whereas the direct effect of consideration on psychological distance was negative and significant. Thus, Hypotheses 1a and 1c were not supported whereas Hypothesis 1b was supported. Hypotheses 2–4 were not supported for initiating structure as the indirect effects on task performance, OCB and psychological withdrawal via psychological distance, and then via coping were not significant. Similarly, Hypotheses 2–4 were not supported for vision communication as the indirect effects on task performance, OCB and psychological withdrawal via psychological distance, and then via coping were not significant. We found support for Hypotheses 2a and 3a for consideration as the indirect effects via psychological distance and then via task coping on task performance (estimate = .066, 95% CI [0.014, 0.168]) and OCB (estimate = .040, 95% CI [0.009, 0.117]) were positive and significant. Hypothesis 4a was also supported for consideration as the indirect effect of consideration on psychological withdrawal via psychological distance, and then via task coping (estimate = −.033, 95% CI [−0.097, −0.006]) was negative and significant. The remaining hypotheses for the indirect effects of consideration (i.e., Hypotheses 2b, 2c, 3b, 3c, 4b and 4c) were not supported. In summary, the indirect effect of consideration on the three outcomes (i.e., task performance, OCB and psychological withdrawal) was mediated by psychological distance and then by task coping.

OVERALL DISCUSSION

Our paper has aimed to offer new insights into remote work experiences and the role of leadership behaviours in helping managers overcome the challenges of being physically separated from their employees. Special emphasis has been placed on the role of perceived psychological distance for employee coping and individual effectiveness. The results of our first study focused on employees working exclusively remotely and showed that both consideration and vision communication reduced perceptions of psychological distance, whereas initiating structure did not have a significant effect. Psychological distance was further found to be negatively related to task coping and positively related to emotion and avoidance coping. Our results further supported our hypothesized sequential mediation model and showed that consideration and vision communication positively and indirectly influenced task performance via psychological distance and then via task coping. Interestingly, our results in the first study that was conducted during a COVID-19 lockdown showed that perceptions of low psychological distance from the leader were associated with increased avoidance coping. Employees who felt more psychologically close to the leader were more prone to seeking distraction activities such as watching TV and browsing social media sites. Earlier empirical research on coping responses during infectious disease outbreaks has also shown avoidance coping to be a common strategy in pandemics (e.g., Chew et al., 2020).

In the second study, we aimed to examine our model in a hybrid context post-pandemic and also offer a more holistic understanding of job performance implications in remote work contexts by including two additional dimensions of job performance, that is, OCB and withdrawal behaviours. Following findings from the first study, the indirect effects of task leadership behaviours were not significant. We replicated the same pattern of effects for consideration via psychological distance and then via task coping on task performance as those observed in the exclusively remote context and also found support for sequential mediation in relation to OCB and psychological withdrawal. The indirect effects of vision communication on task performance, OCB and psychological withdrawal via psychological distance, and then via coping strategies were not significant. This is an interesting finding. Leadership behaviours such as vision communication that address distal, long-term objectives, emphasize the big picture and offer a sense of direction and collective identity (e.g., Yagil, 1998) were crucial in exclusively remote contexts but not in hybrid contexts where employees had opportunities for physical, in-person interactions with their manager. Communication of the long-term vision and direction may alleviate uncertainty due to perceived lack of information in exclusively online environments, serve as a strong signal of the leader's resource availability, minimize psychological distance from the leader and enhance adaptive coping and individual effectiveness. This finding is in alignment with prior studies that showed visionary leadership behaviours to have positive effects in online work environments (Joshi et al., 2009; Kahai, 2013; Purvanova & Bono, 2009). In hybrid contexts, employees may not experience the same information uncertainty as they have access to additional cues from the leader during the in-person interactions (e.g., non-verbal and visual cues), and thus, the leader's emphasis on long-term goals may matter less for psychological distance. According to CLT research, people prefer to use more abstract messages when communicating with others who are distant, whereas they use more concrete messages when communicating with others who are near (Amit et al., 2013; Joshi et al., 2016). This aligns with our finding that abstract vision communication has a positive indirect effect on individual effectiveness when employees are distanced from their leader and work exclusively remotely.

Contributions to theory and practice

Our paper contributes to the recently burgeoning research on remote work (e.g., Bailey & Kurland, 2002; Chong et al., 2020; Gajendran & Harrison, 2007; Golden et al., 2008; Hill et al., 2003; Shockley et al., 2021; Zhang et al., 2021). Prior research has been limited and provided mixed evidence on the relation of remote work with key outcomes such as performance (Gifford, 2022). Given the expected growth of remote work in the future, a systematic research agenda and a closer look at the multi-faceted nature of physical distance in remote contexts is vital. We specifically focus on psychological distance as a key facet of physical distance (Leonardi et al., 2024) and examine the role of leaders in follower outcomes such as coping and individual effectiveness in this context. The existing but limited research on leadership in remote work contexts has highlighted the role of visionary leadership (Kahai, 2013; Purvanova & Bono, 2009), but a closer look at a broader set of leadership behaviours in that context is necessary for a more in-depth understanding.

We specifically investigate three core leadership behaviours, that is, initiating structure, consideration and vision communication (Yukl et al., 2002), as predictors of the psychological distance employees perceive from their manager and their indirect role on employee coping and in-role and extra-role performance and withdrawal. Consistent with prior research, we find vision communication to be an important leadership behaviour in a remote work context but further find support for the role of consideration behaviours. Leaders who show concern for their employees' needs and welfare, minimize psychological distance and overcome the challenges of physical distance by remaining accessible and psychologically close with their people.

We further contribute to research on leader distance (Antonakis & Atwater, 2002; Shamir, 2013) and address the call for more leadership focused CLT research (Wiesenfeld et al., 2017) by examining empirically psychological distance as an important outcome of leadership behaviours that emphasize concrete goals, concern and abstract visions in physically distant work environments. We also contribute to the literature on leadership and coping (Eatough & Chang, 2018). By integrating CLT (Trope & Liberman, 2003, 2010) with COR theory (Hobfoll, 1989), we argued that perceived psychological distance is an important resource signal regarding the availability of leader resources in remote working environments that can shape employees' coping responses and task performance. Our findings provided preliminary support for this claim. We found that consideration and vision communication leadership behaviours significantly reduced perceived psychological distance and psychological distance to be an important mediating mechanism of the relationship between leadership behaviours, task coping and job performance outcomes such as task performance, OCB and psychological withdrawal.

Furthermore, by focusing on remote work and addressing physical distance as a discreet leadership context (Oc, 2018), we provide a much needed contextualized perspective to leadership phenomena. Context is rarely explicitly addressed in leadership research and we could be missing a key opportunity to speak to grand challenges faced by organizations and societies (Johns, 2024). The dramatic rise of remote work post-pandemic reflects such a challenge and a fundamental shift in the way we work and to the meaning of ‘workplace’. Understanding the role of leaders in this context can help better equip organizations and individuals for the future world of work.

From a further practical standpoint, our study highlights the role of leaders in helping employees engage in adaptive coping strategies and stay focused and productive in the remote and hybrid environments. Despite the physical distance, technology allows leaders to remain psychologically close to their employees with beneficial outcomes for individuals and organizations. As the work-from-home and hybrid trend is predicted to continue (Brenan, 2020), our results offer leaders valuable insights. Supportive behaviours, like scheduling regular one-on-one and team meetings to discuss progress and challenges and address well-being concerns, along with visionary actions, such as clear and transparent communication of goals and objectives, can reduce perceptions of distance and enhance employee coping, remote collaboration and performance. Investing in leadership training in remote contexts can help leaders adapt to the changing landscape of work and offer them practical tools on how to lead from a distance.

Limitations and future research directions

Despite its contributions, our study also has limitations. Although the time-lagged design of our studies is known for reducing CMV, the relationships between variables could still be inflated (Podsakoff et al., 2024). Using the CFA marker technique (Williams et al., 2010), we found that CMV was unlikely to be a biasing factor. Furthermore, we hypothesized that followers experiencing less psychological distance demonstrate increased task coping and, subsequently, superior performance in remote work contexts. However, a reverse pattern of effects could also be plausible (e.g., those with high performance perceive lower psychological distance). While our studies employed a sequential mediation design with temporal separation between the measurement of predictors, mediators and outcomes, which offers advantages over cross-sectional designs by incorporating temporal precedence (Mitchell & Maxwell, 2013; Podsakoff et al., 2024), we cannot definitively rule out reversed causal or reciprocal effects. To better explore these relationships and strengthen causal assumptions, future research should employ a full longitudinal approach by measuring all main variables at all time points. We must also acknowledge that we could not measure all possible variables and alternative explanatory mechanisms of the relationship between leadership, coping and task performance. For example, other relational constructs such as trust (Halbesleben & Wheeler, 2015; Jarvenpaa & Leidner, 1999) and identification processes (Connaughton & Daly, 2004; van Knippenberg & Hogg, 2003) may play a role in a remote work context. Previously examined mediators of the relationship between leadership behaviours and task performance, such as role ambiguity, role conflict, justice and leader–member exchange (LMX) (Gottfredson & Aguinis, 2017) can be included in future studies. In addition to leadership, other contextual resources, both work-related, such as perceived autonomy (Gajendran & Harrison, 2007) and family-related such as spousal support (Halbesleben, 2010), can be examined in future studies.

Our research focused solely on psychological distance as the key mechanism through which leadership behaviour relates to followers' coping and performance. Nonetheless, hybrid work arrangements allow followers to physically interact with their leaders, which may help leaders bridge psychological gaps (Leonardi et al., 2024), modifying these leadership effects. Although we accounted for face time with the manager in Study 2 and found it having negligible impact, future research could explore, in addition to psychological distance, other types of physical distance in remote work contexts such as structural and technological distance. This will further support the robustness of our findings.

Another potential limitation is that we did not measure the specific types of online media participants used in the interactions with their manager. Prior studies (e.g., Purvanova & Bono, 2009) suggest that the rich and synchronous nature of online work interactions allows for a more proximal leadership experience. Thus, media richness can be studied as a possible moderator in future research. Furthermore, as one reviewer suggested, employees in different occupations may have different needs for leaders engaging in certain leadership behaviours in remote work contexts, suggesting that future research could adopt this more fine-grained approach to the study of leadership and psychological distance.

In conclusion, our paper showed that leaders can remain close to their employees while working remotely, help them cope and perform effectively. Leaders who employed consideration and vision communication behaviours were perceived as more psychologically proximal to their employees in exclusively remote environments, whereas consideration was found to be more important for reducing psychological distance in hybrid contexts. This experience of psychological proximity to the leader was further found to increase adaptive coping and improve job performance outcomes.

AUTHOR CONTRIBUTIONS

Anders Friis Marstand: Conceptualization; methodology; formal analysis; project administration; writing – original draft; investigation; writing – review and editing. Olga Epitropaki: Conceptualization; writing – original draft; methodology; writing – review and editing. Ilias Kapoutsis: Investigation; writing – review and editing; methodology; formal analysis; conceptualization.

FUNDING INFORMATION

This project was supported by a grant from Durham University awarded to Olga Epitropaki and a grant from Birkbeck, University of London awarded to Anders Friis Marstand.

CONFLICT OF INTEREST STATEMENT

The authors have no conflict of interest to declare.

APPENDIX 1

We adapted the four spatial, three temporal and three social distance items from Lim et al. (2012) and developed four hypotheticality items to measure psychological distance to the manager. Each item started with ‘During my online work interactions with my manager’. Items from Lim et al. (2012) were generally adapted by changing the tense and using ‘we’ instead of ‘the group’. For example, ‘I felt our group was spatially close’ was changed to ‘I feel we are spatially close’. For spatial distance, we used these items: ‘I feel I am in the same place as he/she is’, ‘I feel we are spatially close’, ‘I feel we respond to each other closely’ and ‘I feel he/she interacts in the same place as I am’. For temporal distance, we used ‘I feel I am interacting simultaneously with him/her’, ‘I feel we are not temporally distant’ and ‘He/she gives quick responses to my actions’. For social distance, we used ‘I feel it is easy to become friends with him/her’, ‘I feel socially close to him/her’ and ‘I feel he/she is a socially important person for me’. For hypotheticality, we developed four items, which reflected that hypotheticality refers to ‘how likely is the target event to happen, or how close it is to reality’ (Bar-Anan et al., 2006, p. 609) and the notion of ‘certain vs. uncertain and real vs. hypothetical events’ (Trope & Liberman, 2003, p. 417). The four hypotheticality items are ‘I feel my work plans become more real’, ‘My job tasks seem clear’, ‘I get a better picture of what will happen at work’ and ‘I am certain about what I need to do at work’.

We examined the psychological distance dimensions using CFA analyses in Study 1. The four-factor model with correlated factors showed a reasonable fit to the data (χ2(71) = 402.65, p < .001, CFI = .92, TLI = .89, RMSEA = .12 and SRMR = .07). It provided a better fit to the data (Δχ2 = 354.50, Δdf = 5 and p < .001) than a two-factor model in which spatial and temporal distance items loaded on one factor and social distance and hypotheticality items loaded on another factor (χ2(76) = 757.15, p < .001, CFI = .83, TLI = .80, RMSEA = .16 and SRMR = .09). Similarly, the four-factor model provided a better fit to the data (Δχ2 = 518.09, Δdf = 5 and p < .001) than another two-factor model in which spatial and social distance items loaded on one factor and temporal distance and hypotheticality items loaded on another factor (χ2(76) = 920.74, p < .001, CFI = .79, TLI = .75, RMSEA = .18 and SRMR = .09). Finally, the four-factor model also fitted the data better than the one-factor model (χ2(77) = 1087.12, p < .001, CFI = .75, TLI = .70, RMSEA = .20 and SRMR = .09), which was supported by the significant chi-square difference test (Δχ2 = 684.47, Δdf = 6 and p < .001). The standardized item loadings from the four-factor model with correlated factors averaged .86, .74, .88 and .84 for spatial distance, temporal distance, social distance and hypotheticality, respectively, which indicated that the items were good indicators of their intended factors. Overall, this suggests that the four dimensions are distinct yet related. Furthermore, Cronbach's alpha coefficients are all acceptable: spatial distance (α = .92), temporal distance (α = .78), social distance (α = .91) and hypotheticality (α = .90).

We also examined the psychological distance dimensions using CFA analyses in Study 2. The results for Study 2 showed that the four-factor model with correlated factors (χ2(71) = 214.19, p < .001, CFI = .92, TLI = .90, RMSEA = .10 and SRMR = .06) provided a better fit to the data (Δχ2 = 184.26, Δdf = 5 and p < .001) than a two-factor model in which spatial and temporal distance items loaded on one factor and social distance and hypotheticality items loaded on another factor (χ2(76) = 398.45, p < .001, CFI = .83, TLI = .79, RMSEA = .15 and SRMR = .09). Similarly, the four-factor model provided a better fit to the data (Δχ2 = 354.97, Δdf = 5 and p < .001) than another two-factor model in which spatial and social distance items loaded on one factor and temporal distance and hypotheticality items loaded on another factor (χ2(76) = 569.16, p < .001, CFI = .73, TLI = .68, RMSEA = .18 and SRMR = .11). Finally, the four-factor model also fitted the data better than the one-factor model (χ2(77) = 647.52, p < .001, CFI = .69, TLI = .63, RMSEA = .19 and SRMR = .10), which was supported by the significant chi-square difference test (Δχ2 = 433.33, Δdf = 6 and p < .001). The standardized item loadings from the four-factor model with correlated factors averaged .84, .71, .84 and .78 for spatial distance, temporal distance, social distance and hypotheticality, respectively, which indicated that the items were good indicators of their intended factors. Overall, these results underscore the robustness of the four-dimensional structure of psychological distance and suggest that the four dimensions are distinct yet related. Furthermore, Cronbach's alpha coefficients were all acceptable: spatial distance (α = .90), temporal distance (α = .74), social distance (α = .88) and hypotheticality (α = .86).

DATA AVAILABILITY STATEMENT

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

  • 1 Following recommendations for scale adaptations (e.g., Heggestad et al., 2019) and as we used shortened scales for initiating structure and consideration, we conducted an additional study to examine part-whole correlations. We collected data from 211 full-time employed individuals via Prolific. Of these, 208 (47.1% females; 21.93 years of experience) had a direct manager whom they reported to at work. We compared the 10-item Initiating Structure scale with the shortened six-item version, which produced a very strong part-whole correlation (r = .96, p < .001). Furthermore, we compared the 10-item Consideration scale with the shortened four-item version, which also produced a very strong part-whole correlation (r = .94, p < .001).