Followership at a distance: Follower adjustment to distal leadership during COVID-19.

Leader distance theory has received scant empirical attention in the extant literature; however, the "work from home" orders associated with COVID-19 have made this theory and its empirical findings highly relevant for organizations. Our study integrates leader distance theory and followership theory to understand how follower role beliefs affect follower effort, performance, and withdrawal under physical leader distance and varying conditions of leader interaction frequency. Using a three-wave survey methodology with 260 adults working remotely, our study finds that followers' levels of effort, performance, and withdrawal were contingent on leader interaction frequency. Specifically, followers with a coproduction role orientation, who see their role as more collaborative, reported higher levels of effort under conditions of high leader interaction. Furthermore, the indirect effect of coproduction on follower performance and withdrawal via effort was moderated by leader interaction frequency. The results for followers with passive role orientations, however, were in the opposite direction. These followers reported less effort when leader interaction was high, and the mediational chain predicting performance and withdrawal was contingent on leader interaction frequency. Our study contributes to the ongoing conversation about the positive and negative effects of leader distance and positions followership characteristics as important boundary conditions of distal leadership.

INTRODUCTION

The COVID‐19 pandemic created a global crisis for both large and small business alike (see Verma & Gustafsson, 2020, for a review). As the number of COVID‐19 cases grew worldwide, governments instituted “shelter in place” and “work from home” orders, and businesses slowed or completely halted operations, as increased numbers of employees began working remotely (Hickman & Robison, 2020). Popular press and scholarly articles called for leaders to better engage with their followers, suggesting that employees were struggling to adapt to remote work (Fournier et al., 2020; North et al., 2020), experiencing performance deficits and increases in emotional exhaustion (Rabow & Bullard, 2020), and struggling to understand how organizations were adapting to new operational expectations (Giritli Nygren & Olofsson, 2020; Verma & Gustafsson, 2020). Indeed, Time Magazine called the COVID‐19 outbreak “the world's largest ‘work from home’ experiment” (Banjo et al., 2020, p. 1), and another article stated that “proactive engagement from leaders is key” to maintaining follower motivation, morale, and productivity during virtual work (International Council on Active Aging, 2020, p. 1).

The leadership literature has largely assumed that leaders and followers need to be physically, socially, and interactionally close for leadership processes and outcomes to be effective (Antonakis & Atwater, 2002; Collinson, 2005; Napier & Ferris, 1993). For example, Antonakis and Atwater (2002, p. 12) note that the individualized consideration dimension of transformational leadership is defined by “behaviors that provide individualized and personalized attention to followers implying leader–follower closeness and intimacy.” Research posits that high levels of leader distance can adversely affect follower performance and well‐being because distal leaders have fewer opportunities to directly interact with, support, and develop followers (Avolio et al., 2004; Collinson, 2005; Podsakoff et al., 1996). Distal leaders are denied the closeness required for positive relationship development and thus may struggle to produce high‐quality exchanges and high levels of trust with their followers (Torres & Bligh, 2012). To date, research has primarily examined the outcomes of physical distance between leaders and followers (i.e., working in a different location or city; Avolio et al., 2004; Burrows et al., 1996; Fenlason & Beehr, 1994; Howell et al., 2005; Howell & Hall‐Merenda, 1999). Fewer studies have examined leader–follower distance resulting from varying interaction frequency in the context of high physical separation—the type of context created by the COVID‐19 pandemic.

Despite the assumption that followers suffer under distal leaders, Antonakis and Atwater (2002, p. 3) note that “leader effectiveness is contingent on matching the level of leader closeness that followers expect of the leader in various contexts.” In addition, research on leader distance has produced mixed results, with some followers reporting poorer performance under distal leaders, whereas others were unaffected (Howell et al., 2005; Shamir, 1995; Yagil, 1998). According to followership theory, not all followers share the same view of their role in the leadership process, and whereas some seek high levels of interaction with their leader, others suffice to remain removed from the leadership process (Uhl‐Bien et al., 2014). Followership role theory posits that individuals create and enact different constructions of the follower role in a hierarchical context (cf. Biddle, 1979; Katz & Kahn, 1966) and that behavior in these roles is largely a product of role beliefs and social constructions (Parker et al., 1997; Uhl‐Bien et al., 2014). Role beliefs drive expectations and behavior, and alignment between beliefs and behavior leads to higher self‐appraisals of performance (Biddle, 1979, 1986).

Research on followership role orientations confirms that individuals hold varying beliefs about the follower role, and expectations for how followers should interact with leaders (Carsten et al., 2010; Carsten et al., 2018). Whereas some believe that followers should be passive, deferent, and submissive (i.e., passive followership orientation), others believe that followers should engage with leaders in a collaborative way to advance leadership goals and objectives (i.e., coproduction orientation; Carsten et al., 2018). These differing role beliefs may have affected how followers adapted to the sudden changes in leader proximity or distance due to COVID‐19.

Followers with a coproduction orientation believe that they are required to interact and engage with leaders in a way that contributes to leadership processes, decisions, and outcomes. To stay motivated and feel as though they are productive in their role, it is possible that they would fare poorly under conditions of high physical distance and low perceived interaction frequency (Antonakis & Atwater, 2002; Oc & Bashshur, 2013). Without high levels of interaction and collaboration, these followers may feel alienated and unable to fulfill their role obligations, ultimately affecting perceived levels of effort and performance. Followers with a passive orientation, on the other hand, feel most productive when they are simply carrying out the leader's directives (Carsten et al., 2010). They do not require high levels of interaction because they believe that followers should not involve themselves in the leadership process. Moreover, Carsten et al. (2016) posit that too much interaction with their leaders may actually be frustrating and distracting for these followers. Thus, variations in followership role orientations may differentially affect how followers fare under conditions of leader distance.

Our study makes several important contributions to both leader distance and followership theory. First, we challenge the dominant assumption that followers need high levels of physical and interactional closeness in order to be effective. By examining individual differences in followership beliefs, we advance an understanding of the boundary conditions of leader distance and provide insight on why some followers need higher recurrent interaction with their leaders than others to be productive. In doing so, we add insight on the somewhat mixed results found in leader distance research (Antonakis & Atwater, 2002; Collinson, 2005).

Second, we advance the leader distance theory posited by Antonakis and Atwater (2002), by examining the relationships between followership beliefs and outcomes under conditions of varying interaction frequency. In doing so, we extend the theory beyond examining the traits and behaviors that make leaders effective distal actors and instead examine how follower traits and behaviors affect performance and withdrawal under leader distance. Thus, our study attempts to answer the call by Antonakis and Atwater (2002) to investigate the follower and contextual variables that affect the outcomes of distal leadership.

Third, we advance followership theory and specifically research on followership role orientations, by examining how follower role beliefs affect the ability to adapt and perform under conditions of leader physical distance and varying interaction frequency (Uhl‐Bien et al., 2014). In doing so, we highlight the importance of the leadership context in the continued study of followership behavior and effectiveness (Carsten et al., 2010; Carsten et al., 2018).

Finally, we offer advancements in research on the effects of the COVID‐19 pandemic on working adults and provide insight on the way that sudden changes in work routines differentially affected individual effort, performance, and work withdrawal. Previous research demonstrates that employees often suffer from physical and professional isolation, lapses in communication and trust, and reduced clarity of performance standards while working remotely (Raghuram et al., 2001; Raghuram et al., 2019). Few of these studies, however, have examined how interaction frequency with a leader affects important work outcomes (Gajendran & Joshi, 2012). Our study extends the research on virtual work to incorporate the important followership variables that affect how workers adapt under conditions of virtual interaction with distal leaders.

THEORETICAL BACKGROUND AND HYPOTHESES

Leadership and followership at a distance

The original theory of dyadic distance was advanced by Napier and Ferris (1993) who identified three dimensions of psychological, structural, and functional distance. Building on this original work, Antonakis and Atwater (2002) devised a more holistic theory of leader distance, noting that influence dynamics needed to create leadership are affected by the amount of physical distance, perceived social distance, and perceived interaction frequency between a leader and followers. Antonakis and Atwater (2002) note that these three dimensions are independent, measurable, and each describe one element of leader distance. The theory posits that variations of each type of leader distance impact follower outcomes such as performance evaluations, satisfaction, and withdrawal.

In the present study, we examine perceived interaction frequency under conditions of high physical distance. We excluded perceived social distance (e.g., power differentials or social status) because we were specifically interested in studying the distance dimensions that were directly and immediately impacted by the “work from home” orders associated with COVID‐19. Social distance, as a perception of distance based on rank and status, was of less interest given the restrictions of the pandemic. Physical distance is the actual physical proximity that exists between a leader and followers and has been studied with mixed results. Whereas some have found that physical distance enhances the perceptions of charisma (Shamir, 1995; Yagil, 1998), others have found that individuals and work units suffer when transformational leaders are physically distant (Howell et al., 2005). During the pandemic, remote work increased by 65%, forcing many employees to physically distance from their leaders for the first time in their careers (Hickman & Robison, 2020), and it remains unclear why some workers easily adapted to physical distance whereas others faced challenges (Banjo et al., 2020).

Perceived interaction frequency denotes the extent to which followers perceive a high or low level of interaction with their leader. Interaction can include exchange of information, directives, social support, or feedback. Higher levels of interaction frequency deepen the level of understanding between leaders and followers (Torres & Bligh, 2012), bringing leaders and followers together through frequent contact (Oc & Bashshur, 2013). In times of physical distance, leaders and followers may increase their interaction frequency to counteract the loss of physical closeness.

A number of leadership theories and empirical studies have included leader distance as a situational variable that impacts follower outcomes (see Collinson, 2005, for a review). For example, physical and interactional closeness allows for better collaboration, higher levels of creativity and insight, and better joint problem solving (Griffith et al., 2018; Hill et al., 2014; Howell et al., 2005). Conversely, followers of distal leaders have fewer opportunities for high‐quality exchanges, limiting the richness of their interactions, and subsequent perceptions of trust (Torres & Bligh, 2012). Although conventional leadership theories suggest, and some empirical research has found, that high levels of leader distance have adverse effects on followers (Burrows et al., 1996), Antonakis and Atwater (2002) note that the outcomes of leader distance on followers depend on the characteristics of leaders, followers, and the situation.

Followership theory provides a foundation for understanding the unique contribution followers, and following, make to leadership processes, and posits that individual follower characteristics and behaviors can significantly influence leadership outcomes (Uhl‐Bien et al., 2014). Followership theory has yet to be integrated with leader distance theory, and the only theory to acknowledge that some followers may function well without leader engagement is leadership substitutes theory (Kerr & Jermier, 1978). In this theory, Kerr and Jermier (1978) posit that followers with higher levels of knowledge, competence, commitment, and motivation may not need close supervision by their leaders and may engage in high levels of performance regardless of leader proximity. Although leadership substitutes have been studied quite extensively (see Podsakoff et al., 1996, for a review), the results are inconclusive and research rarely examines followers in the context of physical, social, or interactional distance. Building on the notion that followership characteristics may buffer the effects of leader distance, our study seeks to understand the effects that passive and coproduction role orientation have on follower effort, performance, and withdrawal in the wake of distance restrictions associated with COVID‐19.

Coproduction followers and distal leaders

Followership role orientations are defined as the beliefs that individuals have regarding the follower role, the behaviors that should be exhibited in the role, and the types of activities that make followers effective at interacting with leaders (cf. Parker et al., 1997; Uhl‐Bien et al., 2014). Two followership orientations have been studied in the empirical literature—coproduction orientation and passive orientation (Carsten et al., 2018). Coproduction followership orientation is the belief that followers enhance the leadership process by proactively contributing ideas and opinions or challenging the leader's assumptions when necessary (Carsten et al., 2010). It aligns with work on courageous followership (Chaleff, 2003) and exemplary followers (Kelley, 1992), by suggesting that leadership is a process coconstructed by leaders and followers working in a state of mutual influence and collaboration (Howell & Mendez, 2008; Shamir, 2007). Followers endorsing a coproduction orientation have been found to engage in higher levels of voice, constructive resistance, and proactive problem solving with their leaders (Carsten et al., 2010; Carsten et al., 2018; Carsten & Uhl‐Bien, 2012).

In accordance with early writings on role theory (Biddle, 1986; Katz & Kahn, 1966; Selznick, 1957), individuals seek to engage in behaviors that align with their unique perceptions and role beliefs and perceive higher levels of performance and effectiveness when there is alignment between role beliefs and behavior (Biddle, 1986). Coproduction beliefs may motivate followers to collaborate and partner with leaders, ultimately exerting more work effort (Carsten et al., 2010). Work effort is defined as behaviors, both required and voluntary, that are beneficial to the organization (De Cooman et al., 2009). Although effort has been considered an important part of the motivation process, only recently have scholars differentiated between the two—calling motivation the cognitive antecedent to behavioral effort (De Cooman et al., 2009; Parsons, 1968).

Locke and Latham (1990) posited that effort is composed of three dimensions including direction, intensity, and persistence. Followers with a coproduction orientation exert effort in the direction of engaging with their leaders to help devise and meet work objectives (Carsten et al., 2010). These followers bring higher levels of intensity to understanding the direction their leader wants to go and independently thinking about and evaluating that direction to offer unanticipated insight or suggestions. Furthermore, followers with a coproduction orientation would be persistent in seeking opportunities to provide leaders with feedback, seeking new information, and offering their leader a divergent perspective. Exerting this type of effort would serve as a means of meeting the obligations and expectations dictated by their role beliefs (cf. Biddle, 1986).

Under conditions of physical leader distance, such as that produced by the COVID‐19 pandemic, coproduction followers may have relied on greater interaction frequency to maintain the high levels of effort and contribution that define their role beliefs. Research by Yagil (1998) found that close leaders are able to offer more individually tailored information and are perceived as more realistic and approachable. Moreover, close leaders are rated significantly more effective among followers who desired leader closeness. Followers with a coproduction orientation desire greater proximity to their leaders, and given the physical distance associated with COVID‐19, coproduction followers likely sought greater interaction to maintain higher levels of collaboration and partnership. Thus, when interaction frequency is high, we expect these followers to report greater amounts of work effort. However, when leader interaction is low, we would expect followers with a coproduction orientation to feel frustrated because they are not fulfilling their perceived role obligations. Without leader interaction, followers may experience declining levels of leader engagement and impact and subsequently perceive lower levels of work effort.

Leader–follower interaction frequency will moderate the relationship between coproduction orientation and follower effort such that high coproduction followers will exert more effort when leader interaction is high rather than low.

Passive followers and distal leaders

A passive followership orientation involves the belief that followers should remain silent and deferent and simply carry out the directives of leaders (Carsten et al., 2010; Carsten et al., 2018). It aligns with traditional views of followers as having less knowledge, competence, and motivation (Kelley, 1992) and requiring leader direction and influence in order to act (Collinson, 2006; Lipman‐Blumen, 2005). Followers with a passive orientation believe that leaders have more knowledge and insight than followers and that the follower role is best enacted by remaining detached from the leadership process. Research by Carsten et al. (2010) found that followers with a passive orientation reported that their role was mostly about executing the leader's directives and doing things “the leader's way” (p. 550). As a result, the effort exerted by followers with a passive orientation would be in the direction of receiving and carrying out orders. Their focus and intensity would be on completing tasks assigned by their leader. Furthermore, they would persist in their efforts to complete the task even amidst a global crisis such as a pandemic.

Followers with a passive orientation refrain from engaging with leaders on problem solving or decision‐making and believe that followers should not collaborate in the leadership process (Carsten et al., 2018). As a result, high levels of physical distance would not impede the follower's ability to meet their leader's directives and work objectives. Furthermore, high levels of interactional distance would only become a problem when it disrupted the delivery of direct orders or prevented the execution of those orders. Research by Howell and Hall‐Merenda (1999) found that followers fared well under conditions of sudden leader distance when their leader showed a more transactional leadership style. Reversing the lens, followers who expect to simply receive and carry out directives may perceive that their effort levels remain high under conditions of high physical distance and low interaction with their leaders.

Conversely, followers with a passive orientation may reduce their work efforts when a physically distanced leader seeks very high levels of interaction (Carsten et al., 2016). Under the COVID‐19 “work from home” order, leaders and followers alike would be without normal environmental cues regarding work priorities and levels of engagement, and some leaders may increase interaction with their followers as a means of participative problem solving, assessing work progress, or critically thinking about operational changes. High interaction frequency may be seen as a request for deeper levels of engagement or imply higher levels of responsibility that conflicts with passive role beliefs. As a result, high levels of interaction may incite frustration from followers who believe that they should spend their time executing orders rather than collaborating or engaging with their leader. Specifically, previous theorizing suggests that followers with a passive role orientation will experience conflict between their beliefs and behavior when asked to actively engage with a leader (Carsten et al., 2016). High levels of interaction may pose heavy demands on passive followers, and according to Schmitt et al. (2012), employees may decrease effort or “give up” when environmental demands cause negative affect. Negative affect, especially in times of crisis, has also been found to adversely affect cognitive functioning (Beal et al., 2005; James & Wooten, 2005; Yu et al., 2008). Thus, we posit that followers with a passive role orientation, who have high levels of interaction with their distal leader, will experience greater cognitive fatigue, weakening their intensity and persistence, and ultimately lowering their levels of effort.

Leader–follower interaction frequency will moderate the relationship between passive orientation and follower effort such that high passive followers will exert less effort when leader interaction is high rather than low.

Moderated mediation effects on performance and withdrawal

Although effort has been positioned as a necessary factor in high levels of performance, it does not always lead to high performance reviews by managers (De Cooman et al., 2009). Moreover, research suggests that distal leaders have a difficult time differentiating individual levels of performance and focus more on group‐level performance outcomes (Yagil, 1998). Thus, to understand the moderating role of interaction frequency on the relationship between effort and performance, we chose to evaluate the followers' perception of self‐rated performance. Although manager‐rated performance is often preferable in organizational behavior research, measures of self‐rated performance may be appropriate when employees are overwhelmed with organizational or personal changes that negatively affect their perceptions of effort, control, and personal accomplishment (Demerouti et al., 2005). For example, a number of studies measure self‐rated, as opposed to other‐rated, performance during times of crisis or in high burnout contexts because it is the perception of the context's effect on one's own performance that is of primary concern (Aguiar‐Quintana et al., 2021).

Coproduction followers who are able to communicate frequently with their leader may be able to maintain direction, intensity, and persistence, fostering perceptions that they are performing effectively. Previous research suggests that the performance of a work unit suffered when transformational leaders, highly involved in interaction and collaboration, suddenly became distant (Howell et al., 2005; Howell & Hall‐Merenda, 1999). Following this research, it is likely that followers with a coproduction orientation, who also thrive on interaction and collaboration, would perceive their performance as negative under conditions of lower interaction with their leader. Indeed, Carsten et al. (2010) found that coproduction followers are highly proactive and engaged, and respondents suggested that high levels of effort are required to be effective as a follower. However, these followers may also require high levels of interaction with their leader to make them feel as though their effort is being exerted in the appropriate way and having the intended effects. They would require more validation of their effort to perceive that they are engaging in higher levels of performance, and under conditions of physical distance, this validation can only come through frequent interactions with their leaders. Thus, coproduction followers with high interaction frequency may have been able to engage in high levels of effort and performance (in alignment with their role beliefs) compared with coproduction followers with low interaction frequency.

Coproduction followers who experience low interaction frequency may also suffer greater work withdrawal than followers with higher interaction frequency. Work withdrawal is defined as behaviors individuals use to avoid specific elements of their work role while maintaining membership in the organization (Hanisch & Hulin, 1990). Work withdrawal behaviors include taking long breaks, being late for work, or finding reasons to avoid work and are associated with negative work attitudes and emotions (Rosse & Hulin, 1985; Weiss & Cropanzano, 1996). During COVID‐19, work withdrawal likely increased because employees were challenged to maintain work routines (Maurer, 2020). Followers with strong coproduction orientations who have little interaction with their leaders may suffer the most in terms of maintaining effort and avoiding withdrawal. Without regular leader interaction to support their role efforts or satisfy their role beliefs, coproduction followers may have experienced negative affect, confusion regarding how to execute their role, and negative attitudes about the work context. As a result, they may have decreased their effort and increased their work withdrawal behaviors. With low levels of interaction, coproduction followers may have struggled to maintain the proactive and integrative activities that bring meaning and significance to their role.

The positive indirect effect of coproduction orientation on performance via effort will be strongest for individuals who have a high degree of interaction with their leaders.

The negative indirect effect of coproduction orientation on withdrawal via effort will be strongest for individuals who have a high degree of interaction with their leaders.

In contrast to followers with coproduction orientations who believe their role should be collaborative and proactive, followers with a passive orientation believe that their role is best executed by following direct orders (Carsten et al., 2010). As such, they are likely to see their effort and performance in a highly positive way when they are able to execute on a leader's directives without high levels of engagement with their leader. Indeed, research suggests that followers of transactional leaders were able to maintain high levels of performance under sudden leader physical distance because transactional interactions (e.g., those focused on work transactions rather than empowerment or development) are more easily transferred to distal conditions (Howell et al., 2005; Howell & Hall‐Merenda, 1999). If passive followers are able to take clear direction and dedicate their intensity and persistence to carrying out orders, they will likely perceive that they are meeting their performance requirements in alignment with their role expectations (cf. Biddle, 1986). Conversely, followers with a passive role orientation who have a high level of interaction with their leaders may find that they are unable to clearly and effectively execute tasks. These higher levels of interaction may lead to role conflict and stronger perceptions that the increased interaction is detracting from their ability to effectively carry out orders (Rizzo et al., 1970).

Followers with a passive orientation who maintain high levels of interaction frequency with their leaders may also experience higher levels of withdrawal. Leaders who require higher levels of interaction, especially during a crisis, may show a lack of decisiveness, or incite uncertainty (Yukl et al., 2002). For passive followers, these leaders would create ambiguities that could impair performance and lead to negative affect (Brief & Weiss, 2002; Staw et al., 1981). Followers with a passive orientation rely on their leaders for clear and decisive direction and feel most satisfied when they can execute on a directive. High levels of interaction may impede followers' ability to direct effort in a way that aligns with their role beliefs, leading to the negative work attitudes and affect that are associated with work withdrawal (Hanisch & Hulin, 1990; Weiss & Cropanzano, 1996).

The negative indirect effect of passive orientation on performance via effort will be strongest for individuals who have a high degree of interaction with their leaders.

The positive indirect effect of passive orientation on withdrawal via effort will be strongest for individuals who have a high degree of interaction with their leaders.

METHOD

Participants and procedure

We recruited participants from Mechanical Turk (MTurk), a versatile online platform used to connect researchers with a diverse pool of participants (Litman et al., 2017). We specified inclusion criteria utilizing the customizable MTurk platform and adding inclusion questions within our Qualtrics online survey. Specifically, we invited MTurk participants who were over 19 years old, currently working at home, full time in the United States, and who had an MTurk approval rating of 90% or greater with at least 100 successfully completed tasks (Peer et al., 2014). In addition, we required that our participants be working for a supervisor, for pay, both prior to and after the “shelter in place” orders associated with COVID‐19. Participants who did not meet these requirements were taken to the end of the survey and were thanked for their participation. Data were collected in three waves with 2 weeks between each consecutive wave. Participants were compensated $1 for passing attention checks in each wave.

In the first survey, participants completed measures on their followership role orientation, interaction frequency with their leader, and demographic variables. A total of 721 participants who met the inclusion criteria and passed attention checks were retained for Survey 1. These 721 participants were sent invitations to participate in wave two, which asked questions regarding their work effort and affect. A total of 640 participants cleared all the attention checks and were retained (88% response rate). The third survey asked those 640 participants to report on their performance, withdrawal, and adjustment to virtual work. After accounting for attention checks, our final sample size consisted of 260 participants (36% response rate). Participants were mostly female (56%) with an average age of 39 years, an average tenure of 7.69 years in the organization, and 5.18 years working with their supervisor. Approximately 48.9% of the sample had bachelor's degrees, 19.9% master's degrees, 15.8% high school diplomas, 7.9% associates degrees, and 6.4% professional and doctoral degrees. Seventy‐four percent of respondents were White, 7.1% Hispanic, 6.4% Black, and 9.8% Asian, and 3.4% placed themselves in “other” category. Finally, participants spanned from various industries such as education (18.4%), healthcare (8.3%), government (6%), retail (5.6%), finance (4.9%), insurance (3.8%), and entertainment (3.8%), and 50% reported their industry as “other” or declined to report.

Measures

Coproduction and passive role orientation were measured with Carsten et al.'s (2018) nine‐item scale. The five items that measure coproduction orientation (α = .80) ask participants their beliefs regarding whether followers should collaborate with leaders to enhance the leadership process (i.e., “Followers should be on the lookout for suggestions they can offer to superiors” and “Followers should be proactive in thinking about things that could go wrong”). The four items that measure passive orientation (α = .79) gauge beliefs that followers should be highly deferent and simply follow orders (i.e., “Followers do not have to take responsibility for thinking about how things get done”). Both role orientations are measured on a 1 (strongly disagree) to 6 (strongly agree) scale and have shown to demonstrate good reliability and construct validity (Carsten et al., 2018).

Interaction frequency was measured with a single item asking followers to indicate the number of hours per day, on average, that they spent interacting with their leaders via phone, email, internet conferencing, and so forth. Research on single items indicates that they can yield valid ratings for these types of questions (Carsten & Uhl‐Bien, 2012).

Follower effort was measured with De Cooman et al.'s (2009) 10‐item work effort scale (α = .92). Individuals were asked to report on a number of work effort behaviors (“I put a lot of energy into the tasks I commence”) on a scale from 1 (strongly disagree) to 5 (strongly agree).

Follower self‐rated performance was measured with Williams and Anderson's (1991) seven‐item in‐role performance scale (α = .85). Participants reported the extent to which they agree with statements assessing performance on required duties and responsibilities (“I adequately completed assigned duties”) on a 5‐point Likert scale (strongly disagree to strongly agree). Measures of self‐rated performance have been found to correlate highly with manager‐rated performance (Behrman & Perreault, 1982) and are considered appropriate in contexts where the employee's perception, rather than the manager's perception, is theoretically relevant (Demerouti et al., 2005).

Work withdrawal was measured with Lehman and Simpson's (1992) nine‐item measure of psychological work withdrawal behaviors (α = .83). Participants rated items such as “spent time on non‐work activities during work hours” on a 5‐point frequency ranging from never (1) to always (5).

Inattentive responding was assessed with three items at each time point. Items such as “The previous page asked you to rate beliefs about …” were measured with a multiple choice format such as “(a) leaders (b) organizations (c) followers (d) soccer.” One response option was scored as correct, and other options were scored as incorrect.

Control variables were included in each of the three surveys to allow us to isolate the effects of followership role orientation and leader interaction frequency on effort, performance, and withdrawal. Given the volatility of the COVID‐19 pandemic and the fact that many employees were experiencing challenges associated with work–family balance, virtual adjustment, and isolation, we collected and controlled for both demographic and substantive variables that may have also related to self‐rated effort, performance, and withdrawal. On the substantive side, we used Watson et al.'s (1988) measure to assess positive (α = .92) and negative state affect (α = .94) given previous research to suggest that affect is strongly related to both performance and withdrawal (Iverson & Deery, 2001; Kaplan et al., 2009). We also included Raghuram et al.'s (2001) scale on virtual adjustment (α = .84) including items such as “Since I started working virtually, I have been able to balance my job and personal life.” On the demographic side, we controlled for employee gender, employee age, employee hours worked per week, employee organizational tenure, and tenure with the leader. These variables were included to account for the way the pandemic differentially affected citizens of different ages and genders (Milliken et al., 2020), as well as variables that could alternatively explain performance and withdrawal outcomes (Ng & Feldman, 2010).

RESULTS

Analytic strategy

We first tested our measurement model to ensure that all scale items loaded on the appropriate factors. The results of our measurement model test confirmed that the five‐factor model showed acceptable fit with the data (Hu & Bentler, 1999): χ 2(550) = 2684.19, comparative fit index (CFI) = .93, incremental fit index (IFI) = .93, root mean square error of approximation (RMSEA) = .08, standard root mean square residual (SRMR) = .06. We tested hypotheses with hierarchical regression and moderated mediation using the PROCESS MACRO (Hayes, 2013). We utilized Model 7 of the process macro allowing us to investigate whether the indirect effects of effort are conditional on the levels of interaction frequency with the leader. Given our study hypotheses, we tested for moderated mediation by examining whether indirect effects were significant at higher levels of the moderator. Utilizing the moderated mediation index advanced by Hayes (2015), we conducted bootstrapping analysis with 10,000 resamples to produce 95% confidence intervals (CIs), which allow for a test of significance of conditional indirect effects (Hayes, 2013). This method avoids power and normality issues associated with the traditional Sobel test (Rucker et al., 2011).

Means, standard deviations, and intercorrelations among study variables appear in Table 1. Hypothesis 1 stated that followers with a coproduction orientation would report higher levels of effort under conditions of high interaction frequency. This hypothesis was supported by the positive and significant interaction term, which resulted from our model test (β = .14, p < .01). Similarly, Hypothesis 2 stated that followers with a passive role orientation would report less effort under conditions of high interaction frequency with their leader. This hypothesis was also supported by a negative and significant interaction term (β = −.12, p < .01).

TABLE 1

Correlations and descriptive statistics

	Mean	SD	1	2	3	4	5	6	7	8	9	10	11	12	13	14
1. Age	39.01	11.08	‐
2. Gender	1.64	1.03	−.01	‐
3. Tenure in org (years)	7.69	6.99	.50**	−.06	‐
4. Hours/week	39.54	8.47	.05	−.05	.28**	‐
5. Tenure with leader	5.18	4.67	.40**	−.00	.60**	.10	‐
6. Virtual adjustment	3.57	0.85	−.06	−.03	−.02	−.04	−.00	(.84)
7. Positive affect	3.83	1.06	.18**	.08	.20**	.06	.20**	.10	(.92)
8. Negative affect	1.93	0.91	−.18**	−.06	−.16**	.09	−.17**	−.06	−.44**	(.94)
9. Coproduction role orientation	4.56	0.68	.01	−.17**	.16**	.12	.11	.13*	.30**	−.05	(.80)
10. Passive role orientation	2.54	0.87	−.16*	.16*	−.19**	−.11	−.16**	.03	−.24**	.07	−.66**	(.79)
11. Interactions with leader	2.80	0.88	−.01	.19**	−.00	.11	.04	.02	.16*	−.02	.07	−.02	‐
12. Effort	4.34	0.55	.18**	−.17**	.16**	.01	.18**	−.05	.42**	−.38**	.39**	−.34**	.04	(.92)
13. Self‐rated performance	4.44	0.51	.16*	.01	.18**	.17**	.08	.14*	.25**	−.19**	.17**	−.21**	.05	.31**	(.85)
14. Work withdrawal	2.09	0.63	−.20**	.12	−.15*	.06	−.19**	−.09	−.31**	.45**	−.15*	.18**	−.13*	−.42**	−.30**	(.83)

Note: N = 260. Coefficient alphas for reflective (but not formative) measures are represented on the diagonal.

Significant at p < .05.

Significant at p < .01.

To test our moderated mediation hypotheses (Hypotheses 3–6), we investigated whether the mediating effects of follower effort are contingent on level of interaction frequency by computing effort at high and low levels of interaction frequency (Hayes, 2013). We controlled for six demographic variables and three substantive variables; however, only the three substantive variables (positive and negative affect, and virtual adjustment) produced significant coefficients in predicting effort, performance, and withdrawal. As revealed in Table 2, the positive indirect effect of coproduction role orientation on self‐reported performance through effort was weaker for low (.03; 95% CI: .001 to .08) and stronger for high (.09; 95% CI: .03 to .16) interaction frequency with the leader, thus supporting Hypothesis 3. A plot showing the conditional indirect effects of coproduction on self‐rated performance is presented in Figure 1. The indirect effect of coproduction role orientation on work withdrawal was also moderated by effort. Table 2 shows that the negative indirect effect was stronger for individuals reporting higher (−.11; 95% CI: −.21 to −.05) than lower (−.04; 95% CI: −.10 to −.001) interaction frequency, supporting Hypothesis 4 (see Figure 2).

TABLE 2

Conditional indirect effects of coproduction role orientation on outcomes via effort, moderated by interaction with leader

Model predicting performance	Coefficient	SE	LLCI	ULCI	R 2
Mediator variable model: Effort					.38***
Constant	5.13***	.69	3.77	6.49
CRO	−.15	.15	−.43	.14
Interaction with the leader (LI)	−.67**	.23	−1.12	−.21
CRO × LI	.14**	.05	.05	.24
Dependent variable model: Self‐rated performance					.17***
Constant	2.45***	.38	1.70	3.21
Effort	.23**	.07	.11	.37
CRO	.01	.05	−.09	.10
Index of moderated mediation	.04	.02	.01	.08
Indirect effect
Low interaction	.03*	.02	.00	.08
Medium interaction	.06*	.02	.02	.11
High interaction	.09*	.04	.03	.16

Note: N = 260.

Abbreviations: CRO, coproduction work orientation; LLCI, bootstrap lower level confidence interval; SE, bootstrap standard error; ULCI, bootstrap upper level confidence interval.

p < .05.

p < .01.

p < .001

FIGURE 1

Conditional indirect effect of coproduction role orientation on self‐rated performance at values of the moderator. CI, confidence interval

FIGURE 2

Conditional indirect effect of coproduction role orientation on work withdrawal at values of the moderator. CI, confidence interval

Hypotheses 5 and 6 predicted a conditional indirect effect between passive role orientation and performance and withdrawal, respectively. We again included all nine control variables in the model; however, only the three substantive variables (positive and negative affect, and virtual adjustment) produced significant coefficients in predicting effort, performance, and withdrawal. The negative indirect effect of passive role orientation on performance via effort was stronger for high (−.05; 95% CI: −.12 to −.02) and null for low (−.01; 95% CI: −.04 to .02) interaction frequency with the leader, supporting Hypothesis 5 (see Table 3). This conditional negative indirect effect is depicted in Figure 3. Finally, the positive indirect effect between passive role orientation and work withdrawal via effort was stronger for high (.07; 95% CI: .03 to .14) and null for low (.01; 95% CI: −.02 to .06) interaction frequency with the leader, supporting Hypothesis 6 (see Table 3 and Figure 4).

TABLE 3

Conditional indirect effects of passive role orientation on outcomes via effort, moderated by interaction with leader

Model predicting performance	Coefficient	SE	LLCI	ULCI	R 2
Mediator variable model: Effort					.35***
Constant	3.77***	.42	2.93	4.60
PRO	.19	.19	.12	−.04
Interaction with the leader (LI)	.32**	.11	.11	.51
PRO × LI	−.12**	.03	−.20	−.04
Dependent variable model: Self‐rated performance					.18***
Constant	2.75***	.42	1.92	2.6
Effort	.22***	.06	.09	.34
PRO	−.05	.04	−.13	.02
Index of moderated mediation	−.03	.02	−.07	−.01
Indirect effect
Low interaction	−.01	.01	−.04	.02
Medium interaction	−.03*	.01	−.07	−.01
High interaction	−.05*	.03	−.12	−.02

Note: N = 260.

Abbreviations: CRO, coproduction work orientation; LLCI, bootstrap lower level confidence interval; PRO, passive role orientation; SE, bootstrap standard error; ULCI, bootstrap upper level confidence interval.

p < .05.

p < .01.

p < .001

FIGURE 3

Conditional indirect effect of passive role orientation on self‐rated performance at values of the moderator. CI, confidence interval

FIGURE 4

Conditional indirect effect of passive role orientation on work withdrawal at values of the moderator. CI, confidence interval

DISCUSSION

Leader distance theory suggests that physical distance and perceived interaction frequency between leaders and followers will affect the leadership process by limiting opportunities for influence, support, and collaboration (Antonakis & Atwater, 2002; Collinson, 2005). Previous research, however, has found mixed results with regard to how followers fare under conditions of leader distance (Avolio et al., 2004; Griffith et al., 2018; Howell et al., 2005; Howell & Hall‐Merenda, 1999; Podsakoff et al., 1996; Shamir, 2005; Torres & Bligh, 2012), suggesting that followership variables may be important boundary conditions of leader distance theory. To our knowledge, our study is the first to integrate followership theory with leader distance theory in an effort to understand how individual follower beliefs affect follower effort, performance, and withdrawal under conditions of leader physical distance and varying levels of interaction frequency.

Our study found that followers with a coproduction orientation engaged in more effort under conditions of high leader interaction frequency, and the indirect effect via effort on both performance and withdrawal was contingent on leader interaction. This result suggests that followers with a coproduction orientation engaged in more effort and had stronger effort to performance links when they had high levels of interaction with their leaders. Furthermore, coproduction followers demonstrated a negative effort to withdrawal link under conditions of high interaction frequency with their leaders.

For followers with a passive role orientation, however, relationships were found in the opposite direction. Specifically, followers with a passive role orientation reported less effort under conditions of high interaction frequency with their leaders. Furthermore, the negative indirect effect on performance via effort was stronger for passive followers reporting high levels of interaction frequency with their leaders. This mediational chain was also significant when predicting withdrawal under conditions of high interaction frequency. Taken together, our results suggest that coproduction followers engaged in more effort and reported greater performance and less withdrawal under high levels of leader interaction, whereas passive followers reported less effort, worse performance, and greater withdrawal under conditions of high leader interaction. Our study provides an important extension for both leader distance theory and followership role theory and offers insight to leaders and organizations regarding how to manage remote workers in the height of the COVID‐19 pandemic.

Theoretical implications

Leadership theory and research has largely assumed that leaders and followers need to be in close physical proximity in order to allow for the proper levels of influence, support, development, and collaboration (Antonakis & Atwater, 2002; Collinson, 2005; Napier & Ferris, 1993). Previous research on distal leadership, however, has found inconsistent results with some followers experiencing performance deficits under distal leaders whereas others were unaffected (Avolio et al., 2004; Howell et al., 2005; Howell & Hall‐Merenda, 1999; Podsakoff et al., 1996). As noted by Antonakis and Atwater (2002), the inconsistent results may indicate that followers and context play important roles in understanding the positive or negative effects of leader distance.

Our study contributes to the leadership and followership literature by suggesting that followers vary in their role beliefs and expectations for interaction under conditions of distal leadership. Building followership into current theories on leadership processes is in keeping with calls for a more “balanced” approach to leadership (Shamir, 2007) and one that seeks to understand how follower characteristics impact leader and follower outcomes (Uhl‐Bien et al., 2014). Specifically, our study finds that followers differ in their role orientations and that some followers believe that the role involves active collaboration whereas others believe that the role is best executed by simply following orders and abstaining from high levels of interaction. Understanding variance in followership role orientations challenges that notion that followers are an “undifferentiated mass or collective” (Collinson, 2006, p. 179) or that followers are “passive recipients of a leader's influence” (Shamir, 2007, p. x). Conversely, our results demonstrate that followers perceive their role in different ways, and followership role orientations have implications for how leaders should engage with followers under conditions of leader distance. Future research could extend this contribution further by studying how both leadership style and followership role orientations affect outcomes of distal leadership.

Our study contributes to leader distance theory by integrating followership variables as important boundary conditions on the effects of leader distance (Antonakis & Atwater, 2002). Leaders and followers who are physically distant cannot rely on context or norms to understand expectations, priorities, or levels of engagement. Although the leader distance literature has focused on the characteristics and behaviors that leaders need to adapt to high levels of distance (Yagil, 1998), research has yet to understand the followership characteristics that promote follower adaptation. Kerr and Jermier (1978) suggest that follower competence may serve as a substitute for leadership, allowing followers to remain effective in the absence of their leader. Our findings extend this early work by suggesting that beliefs about the follower role, and how to enact the follower role, also influence follower effectiveness under conditions of distal leadership. Future research could further advance the theory by studying follower role beliefs under perceived social distance in addition to physical separation and varying levels of interaction frequency.

The finding that passive and coproduction followers reported different levels of effort, performance, and withdrawal under distal leadership contributes to our understanding of followership role theory (Uhl‐Bien et al., 2014). For example, the finding that followers with a coproduction orientation reported greater effort, greater performance, and less withdrawal under conditions of high interaction frequency suggests that some followers require high levels of collaboration with their leaders in order to feel effective in their role (cf. Biddle, 1979, 1986). It is possible that higher levels of interaction allowed these followers to align their effort with their role beliefs regarding collaboration (Carsten et al., 2010; Carsten et al., 2018), engaging with their leaders in meaningful ways, and thus leading to more positive perceptions of performance. Similarly, higher levels of interaction, and greater alignment of effort with role beliefs, may have also helped to keep these followers engaged throughout the COVID‐19 crisis, ultimately leading to less behavioral withdrawal. Given that employees experienced many challenges associated with the pandemic (Fournier et al., 2020; North et al., 2020), these results are important in that they allow us to understand the conditions under which certain followers thrive.

Followers with a passive role orientation responded very differently to high levels of leader interaction. These followers reported less effort when leader interaction frequency was high and reported worse performance and higher levels of withdrawal. These results challenge scholarly and popular press articles suggesting that leaders should maintain high levels of engagement with followers during the pandemic (Fournier et al., 2020; International Council on Active Aging, 2020; Rabow & Bullard, 2020). Instead, they align with followership theory and recent empirical investigations suggesting that followers are highly different in their perceptions of the follower role, as well as their behavioral expectations and their views about interacting with leaders (Carsten et al., 2010; Carsten et al., 2018; Shamir, 2007; Uhl‐Bien et al., 2014). Our results align with followership role theory, suggesting that followers who see their role as largely passive and deferent may have negative reactions to leaders who try to engage them on a high level (Carsten et al., 2016). It also aligns with research suggesting that during the pandemic, employees adapted positively to independent work that did not require high levels of interaction or collaboration (North et al., 2020). For followers with a passive role orientation, working independently, without being asked to collaborate, is distinctive of their role beliefs.

Finally, our research contributes to the growing literature on the effects of virtual work during COVID‐19. During the early stages of the pandemic and work from home orders, there was a concern about employee productivity due to employee simultaneous involvement in work and parental roles, as well as isolation from coworkers. Yet, productivity estimates were mixed, with some employees experiencing greater work output and others experiencing less. Our research suggests that productivity likely remained high among passive followers who experienced less interaction with their leaders and coproduction followers who had more interaction.

Practical implications

As organizations continue to grapple with the effects of the pandemic, our results may offer some help to leaders who are struggling to understand varying levels of follower engagement and effectiveness. For example, leaders who have increased interaction and engagement with followers may experience mixed results with some followers maintaining high levels of effort and performance, and others declining. Leaders may benefit from understanding their followers' role orientation, their beliefs about whether they should collaborate and engage with their leaders, to modify their approach to individual followers. Instead of applying a “one size fits all” approach with their direct reports, leaders should attempt to tailor their approach to their followers' style and role expectations.

Leaders would also benefit from understanding that their level of interaction can have indirect effects on performance and withdrawal. Followers who simply want to follow orders and avoid lengthy discussions or collaboration may experience conflict between their role beliefs and behaviors when interaction frequency is high. Given that work withdrawal can lead to negative organizational performance and turnover (e.g., Bakker et al., 2004; Wright & Bonett, 1997; Wright & Cropanzano, 1998), both leaders and followers could benefit from understanding how their role beliefs, and expected behavior, might lead to greater levels of withdrawal.

Our results also have implications for organizations attempting to bring employees back to a traditional office setting as the pandemic subsides. For example, followers with a passive role orientation may have enjoyed working at home with minimal interaction with their leaders. These followers may have found that they were more effective in their role when they were not asked to collaborate or engage in the leadership process. Conversely, followers with a coproduction orientation may have enjoyed higher levels of interaction with their leaders, which may change as the dyad resumes close physical proximity. If leaders increased their interaction with these followers as a result of being physically distant, it would help them to understand that the increased interaction has positive effects on effort and performance. Organizations may benefit from surveying employees about their perceived productivity while working from home and gauge follower interest in returning to work as a method of instituting a staggered “return to work” schedule.

We recommend that both leaders and followers receive training on followership role orientations in an effort to gain self‐awareness about their own role beliefs and the behaviors they seek to engage in. Such a development program could provide dyads with a better understanding of different leader and follower styles and expectations for behavior. Increased understanding and appreciation for the differences in followership beliefs may aid leaders and followers who seek to work more effectively as dyad partners.

Limitations and future directions

Although our study has many strengths and is the first to examine followership role beliefs in the wake of leader distance, our results should be interpreted cautiously until they have been replicated in a variety of settings and with multiple methodologies. Our research is correlational in nature, which precludes us from drawing causal conclusions. In addition, given that we used only subjective measures, we cannot draw conclusions about objective work outcomes. Future research could serve to improve upon some of the measures used in this study. For example, gathering more objective ratings of performance, or ratings from multiple sources, could allow researchers to understand differences in subjective and objective ratings of effort and performance.

Additionally, our data represent working adults in the United States, and it is possible that cultural variables play a role in whether a follower is able to adapt to remote work in times of crisis. It is possible, for example, that countries with lower power distance may have fewer followers with a passive role orientation and thus may experience more incidents of followers requiring interaction and collaboration. Moreover, countries with higher power distance may experience greater discrepancies of leader–follower interaction under conditions of high leader physical distance. Future research could examine our model across cultural contexts to understand how cultural variables relate to follower role orientations, effort, performance, and withdrawal under conditions of leader distance.

Our research examined variations in interactional frequency under conditions of high leader physical distance. Future studies could also examine the influence of perceived social distance with respect to role orientations. Specifically perceived differences in status, rank, social standing, and power could be relevant to understanding how follower role orientations affect effort, performance, and withdrawal. For example, it is possible that high levels of perceived social distance affect how followers interact with leaders and the extent to which they collaborate and engage with leaders.

Finally, although our investigation moves closer to achieving the “balanced view” of leadership called for by Shamir (2007) by identifying followers are important agents in the leadership process, we concur with Shamir's caution that we not err the other way by becoming too follower centered. Therefore, we offer our study and results as “the other half” of the leader‐centered view of leader distance (Shamir, 2007). Although our study moves us closer to understanding the integration of leadership and followership theory, future research could incorporate leadership variables to assess the interactions of leadership and followership styles in understanding the outcomes of close versus distal leadership.

CONCLUSION

By integrating leadership distance theory and followership theory (Antonakis & Atwater, 2002; Collinson, 2005; Uhl‐Bien et al., 2014), our study extends theory and research on the effects of leader distance. Our results challenge the traditional perspective that leader closeness is always better and positions followers as an important boundary condition for understating distal leadership. By studying followers as a central aspect of distal leadership, we highlight the importance of followership in leadership and pave new avenues for future empirical and theoretical development.