Leading and Working From Home in Times of COVID-19: On the Perceived Changes in Leadership Behaviors.

Due to the COVID-19 crisis, managers and employees in many organizations suddenly are forced to work from home. Although working from home (WFH) is not a new phenomenon, it is new in its current scale and scope because of COVID-19. Against this background, we investigate the effect of WFH during the COVID-19 crisis on changes in leadership behaviors, and associated changes in perceived manager quality and productivity, at different hierarchical levels in organizations. Based on the literature, we develop two predictions in opposite directions. On the one hand, implementing WFH may force managers to show less direction and control and especially more delegation. On the other hand, research into the effects of exogenous shocks such as COVID-19, suggests that managers may become more controlling and delegate less. Consistent with the first prediction, we find that managers perceive they execute significantly less control and delegate more. Employees also perceive a significant decrease in control, however they perceive on average no change in delegation. Furthermore, and in line with the second prediction, employees of lower-level managers even report a significant decrease in delegation. Finally, our results show that increased delegation is associated with increased perceived productivity and higher manager quality. Together, these results suggest that in the context of the COVID-19 crisis, the effectiveness of WFH might be hampered by the fact that required changes in leadership behaviors, in particular in delegation, are difficult to realize in times of crisis.

The COVID-19 crisis strongly affects how and where managers and employees do their work. In most countries, governments decided in early 2020, almost overnight, that people should work from home as much as possible. Clearly, the concept of working from home (WFH) is not new. Since the 1970s the concept of “teleworking” or “telecommuting” has been around as a way to establish more flexibility in organizations (Haddon & Lewis, 1994). An Ipsos/Reuters poll showed that already in 2012, about 20% of employees in developed and developing countries worked remotely to some extent (Reaney, 2012). But what is new, is the scope and scale by which people work from home in 2020. A survey by Gartner (2020) indicates that across the globe, one-half of the companies had more than 80% of their employees WFH during the early stages of the COVID-19 pandemic. One of the major differences between WFH and “normal” working conditions is that managers have to manage their employees at a distance, which could have implications for their behaviors (Bonet & Salvador, 2017). Indeed, the Financial Times recently coined the term “arms-length managers” to describe the new reality for managers (Hill, 2020). In the current manuscript, we therefore investigate how COVID-19-induced WFH affects leadership behaviors in organizations.

In testing these effects, we depart from the notion that given the COVID-19 crisis, managers and employees do not only have to deal with the fact that they suddenly have to work from home, but that COVID-19 also simultaneously and unprecedentedly implies an exogenous external threat in terms of health or economic implications. The combination of these features turns the COVID-19 crisis into a truly unique, multifaceted crisis. As a result, we propose that the crisis may have different—in fact opposite—effects on leadership behaviors. On the one hand, WFH during the COVID-19 crisis implies by definition that managers and employees will be separated from each other in a physical sense and that managers cannot execute their role in the same way as they do in situations where they are in the same location as their employees. Consequently, one prediction is that in response to COVID-19-induced WFH, ceteris paribus managers will show less directive or controlling behaviors, and more delegation of responsibilities toward their employees.

On the other hand, following earlier findings on the effect of other crises such as the 2008 financial crisis, the COVID-19 crisis could also trigger a threat-rigidity response from managers. Managers and employees are not working from home because they suddenly decided that it was the preferred working arrangement, but it was forced upon them because of the COVID-19 crisis—an unprecedented health and economic crisis. Such situations of crises obviously affect both managers and employees because crises create uncertainty and stress, and thereby affect behaviors (Fritsche et al., 2011). Recent research into the effects of the 2008 financial crisis on leadership behavior (Stoker et al., 2019) shows, for instance, that the 2008 financial crisis went along with a significant increase in directive leadership across the board. This finding can be explained by the threat-rigidity hypothesis (Staw et al., 1981) which poses that individuals and organizations react to an external threat, such as a crisis, with actions that reflect rigidity. Research on neuropsychology in response to threat and uncertainty confirms this hypothesis. It shows that the anxiety that results from uncertainty and threat, leads to more careful, rigid, and halting approach behavior (Gray & McNaughton, 2000; McGregor et al., 2010). Consequently, an alternative second prediction is that this increased rigidity also happened in the context of the COVID-19 crisis, and thus may translate to the increased directive and controlling behaviors of managers, and less delegation of responsibilities.

Next to the fact that the change to WFH during COVID-19 may thus have opposing implications for leadership behaviors, based on earlier research (e.g., Fleenor et al., 2010), it is also conceivable that the perception of these effects on leadership behaviors differs between managers and employees. In fact, this is particularly likely, given the realities of WFH. Even though managers and employees may have different perceptions of leadership behaviors under normal situations, the potential for such divergence is radically larger if both are working from home and thus lack many of the channels and opportunities for seeing how the other is doing. This further amplifies the possibility that managers and employees may diverge in their perceptions of how COVID-19-induced WFH affects leadership. In fact, surveys conducted in the first months of the COVID-19 crisis among employees and employers suggest this to be the case, with employees feeling less supported than executives think (IBM, 2021).

To test these predictions, we conducted a study during the initial lockdown period (Spring 2020) using a survey among managers and employees in the Netherlands, to empirically test these two predictions against each other, by measuring the effects of leading and working from home during the COVID-19 crisis on the changes in leadership behaviors at different hierarchical levels in organizations, from both managers' and employees' perspectives. Specifically, and following the literature, we focus on two behaviors that are highly relevant both in the context of a crisis (Stoker et al., 2019) and in the context of WFH (Bailey & Kurland, 2002; Clear & Dickson, 2005). Namely, we investigate the changes in control and delegation. We study the perceptions of managers and employees (Ns = 404 and 561, respectively). Moreover, we explore whether these changes are related to the perceived quality of the manager and perceived productivity.

By doing so, our study makes several contributions. First, our study contributes to understanding the interactive effects of WFH and an external threat on changes in leadership behaviors. The study took place during the initial stages of the first lockdown in the Netherlands in April and May 2020, when most countries went into a government-mandated lock-down that forced most managers and employees to work from home. This situation created a research setting that is unprecedented and cannot be mimicked in the lab. Second, our findings add to the literature on organizational change and the literature into effects of large-scale, exogenous shocks or “disaster” literature more generally. Specifically, for WFH, the literature has been focused on the consequences of WFH in a context where organizations and managers intentionally and voluntarily chose to implement this type of organizational change (Bailey & Kurland, 2002). Our study is among the first to investigate the consequences of WFH when this type of organizational change is imposed upon managers and employees. In the field of disaster literature (Klomp & Valckx, 2014), looking at actual behavioral responses after exogenous shocks such as COVID-19 is rare, and our paper therefore tries to address this lack of knowledge.

Theoretical Background

WFH and Leadership Behaviors

Organizations started implementing WFH in the 1970s, and by now several studies have looked into the effects of WFH, or related concepts such as telecommuting, teleworking, and virtual work. Although most research has focused on employees (for an overview see, e.g., Makarius & Larson, 2017), Bloom et al. (2015) found that it is also becoming more and more common for managers to work from home occasionally: the share of managers in the United States, the United Kingdom, and Germany allowed to WFH during normal hours is almost 50%.

Managing from a distance has strong implications for employees and managers. Research shows that WFH, in general, has beneficial outcomes, both in terms of proximal variables such as perceived autonomy and lower work–family conflict, as well as distal variables such as job satisfaction, performance, turnover intention (Gajendran & Harrison, 2007), and productivity (Allen et al., 2014). Importantly, Allen et al. (2015) found that the relationship between managers and employees is not harmed by WFH. At the same time, it is also clear that there are downsides in WFH, both for employees and managers. Drawbacks for employees are that they have less social interaction, they have to maintain boundaries between work and nonwork (Ramarajan & Reid, 2013), and they often do not have space in their homes to attend to work (Bloom et al., 2015).

For managers, the main downside is (perceived) loss of control over their employees (Bailey & Kurland, 2002; Clear & Dickson, 2005). Based on a literature review, Bailey and Kurland (2002) state that managers are often reluctant to implement WFH, mainly because of this fear of losing control of employees (see also Kniffin et al., 2021). Indeed, research shows that managers of telecommuters often show greater task-oriented leadership than relation-oriented leadership (Madlock, 2012), whereas such task-oriented, close monitoring behavior is shown to be ineffective toward employees who work from home (Lautsch et al., 2009). Second, a main implication of WFH is that managers are expected to delegate responsibilities to their employees. Because of the fact that employees work from a distance, they are expected to get autonomy and responsibilities for topics such as working time, place, results, and methods (Contreras et al., 2020). Consequently, managers will have to share their power and delegate responsibilities to their employees (Cortellazzo et al., 2019).

When combined, these findings imply that at the level of leadership styles, two styles are especially relevant for WFH, namely directive leadership and empowering leadership (Lorinkova et al., 2013). Managers showing directive leadership are seen as giving clear and detailed directions to employees, exercising control, and expecting compliance with instructions (see, e.g., House, 1971; Kamphuis et al., 2011; Lorinkova et al., 2013; Somech, 2006). Empowering leadership can be defined as sharing or delegating power and allocating autonomy and responsibilities to employees (Cheonga et al., 2019). In order to be effective, managers are expected to lower their level of directive leadership and to increase their level of empowering leadership, because the behaviors that constitute these styles are important for the employees who work from home. Therefore, in this study, we examine the changes in these two leadership styles. We do so by investigating changes in two specific behaviors that can be seen as high representative of these two styles. We thus cast “exercising control” as a prime behavioral example of directive leadership, and “delegation” as a prime behavioral example of empowering leadership.

WFH often means that only employees telecommute, whereas in the current COVID-19 crisis both employees and managers work from home. Research indicates that this dual nature of WFH matters. Based on a large dataset of more than 11,000 respondents, Golden and Fromen (2011) conclude that employees with telecommuting managers responded less positively than those with managers in a traditional work mode, both in terms of work experiences (feedback, empowerment, professional development, and workload) and outcomes such as lower job satisfaction and higher turnover intentions. Basing themselves on social exchange theory (Blau, 1964; Homans, 1958), the authors suggest that these detrimental effects of telecommuting are caused by deteriorated relationships between managers and employees (Golden & Fromen, 2011).

More generally, when it comes to perceptions of managers' behaviors, such perceptions often differ between managers and employees (Lee & Carpenter, 2018). Managers tend to rate themselves higher on their own leadership behaviors than their employees (Fleenor et al., 2010). Especially in the context of an organizational change such as the implementation of WFH, this finding seems relevant. For the implementation of a concept that has comparable consequences for managers and employees, that is working in self-managing teams, Stoker and Van der Heijden (2001) report that managers have inflated perceptions of their own behaviors when it concerns the “desired” changes in leadership that fit with the new concept (see also Stoker, 2006). In a similar vein, it can be expected that in a WFH context where leaders are “expected” to show less control and more delegation, it is likely that managers will have an inflated perception of the decrease in controlling behavior and the increase the delegation of responsibilities.

WFH and Hierarchical Levels of Leaders

Leadership behaviors often differ across hierarchical levels of managers. This difference is primarily caused by physical distance, defined as “how far or how close followers are located from their leader” (Antonakis & Atwater, 2002, p. 684). Greater hierarchical differences in organizations often manifest in a greater physical distance, and in a lower frequency of direct interaction (Chun et al., 2009). There is a direct relationship between distance and hierarchy, such that it is more likely that managers at lower hierarchical levels have less physical distance with their employees than higher-level managers.

As a consequence, lower-level managers have more interactions with employees than higher-level managers, making it more likely that directive and more specifically controlling behavior is more prevalent among those managers. With respect to delegation, Jago and Vroom (1977) suggest a greater propensity for delegating behaviors at higher organizational levels. Empirical findings by Yukl and Fu (1999) indeed show that more delegation was used toward an employee who was, just like the manager him/herself, also a supervisor. Related, although not specifically testing the effect for different hierarchical levels, Haselhuhn et al. (2017) found that managers who feel powerless are less willing to delegate decision-making authority. Arguably, managers higher up the hierarchical ladder feel more powerful (Lammers et al., 2009), thereby suggesting that also the level of delegation will differ across hierarchical levels.

To conclude, without considering the current enforced move toward WFH, research strongly indicates that WFH affects both the possibility and the desirability for managers to show certain leadership behaviors. Because of the increased physical distance, managers can simply execute less control. Moreover, they are supposed to delegate more responsibilities. Crucially, we expect that changes in these leadership behaviors differ across hierarchical levels. Given the fact that for higher-level managers, the physical distance that goes along with WFH is already much more common than for lower-level managers, one can expect the changes in leadership behaviors to be the strongest for lower-level managers.

In our view, these conclusions about the relationship between WFH and leadership behaviors are relevant for WFH that was enforced upon organizations because of COVID-19 (see for a similar line of reasoning the position papers by Kniffin et al., 2021; Rudolph et al., 2020). As such, the tendency that WFH will affect leadership behaviors in the perception of both managers and employees seems relatively straightforward if one could zoom in or isolate the enforced WFH aspect of the COVID crisis. But this assumption would require that the implementation of WFH and the COVID crisis can empirically be separated, which is of course not possible, because of their simultaneous occurrence across the globe.

The Effect of a Crisis on Leadership Behaviors

Leaving aside the possible WFH ramifications for leadership behaviors, there is also the question of what a large-scale crisis such as the COVID-19 as such could imply for perceived changes in leadership behaviors (Bartsch et al., 2020). The COVID-19 crisis can be looked upon as an external threat to organizations from both a health and an economic perspective. In their seminal paper on the threat-rigidity hypothesis, Staw et al. (1981, p. 502) define a threat as “an environmental event that has impending negative consequences for the entity.” In this hypothesis, managerial responses to such threats are characterized by a restriction of information processing and constriction of control within the organization (Staw et al., 1981). For the 2008 financial crisis, Stoker et al. (2019) translated these managerial responses to leadership behaviors, and argue and show that this crisis led to an increase in directive leadership (see also Kamphuis et al., 2011). In addition, Feenstra et al. (2020) show that managers who perceive their position to be under threat, show less empowering leadership and delegate less responsibilities toward their employees in such situations (see also Mead & Maner, 2012).

In the threat-rigidity hypothesis, the environment is characterized by fundamental uncertainty as a result of a unique and unprecedented shock. Chattopadhyay et al. (2001) coin the latter as a “control-reducing threat,” where managers act in order to maintain control and to reduce the threat, by tightening their grip on the organization and their subordinates. Clearly, this description also applies to the COVID-19 crisis, which would imply that managers would show the same behaviors as were found in the context of other, similar threats. Therefore, it can be argued that the COVID-19 crisis could lead to a situation in which managers are inclined to increase their levels of control and decrease their delegation of responsibilities.

Overview of Research

As we described above, the combined WFH and COVID-19 shocks can possibly have opposing implications for perceived changes in leadership behaviors. Applied to the actual case of WFH following the outbreak of the COVID-19 crisis in the Spring of 2020, we explore the perceived changes in leadership behaviors according to managers and employees by investigating two leadership behaviors—exercising control and delegating responsibilities—that are exemplary behaviors for two leadership styles, namely directive and empowering leadership, respectively.

More specifically, for managers, we expect that WFH during the COVID-19 led to a change in their own perceived behaviors that is fully in line with what is expected from them in such a WFH context. Following the literature (Fleenor et al., 2010; Stoker & Van der Heijden, 2001), we argue that managers themselves will perceive that they show less controlling behavior and delegate more responsibilities toward their employees. Moreover, we propose that such an effect will be more pronounced for managers in lower management positions (Haselhuhn et al., 2017).

For the employees' perceptions of changes in leadership behaviors, the expected combined impact of WFH and the COVID-19 crisis is not as straightforward. On the one hand, following earlier research (Golden & Fromen, 2011), WFH would mean that managers will show less controlling behavior and more delegation. On the other hand, following literature on the impact of an external threat (Stoker et al., 2019), one could envision a change in even the opposite direction. Namely, such a crisis could lead to an increase in the exercise of control, and at least not an increase or maybe even a decrease in delegation. Moreover, since WFH has especially a strong effect on lower-level managers (Madlock, 2012), we also expect that the changes in perceived behaviors will be more pronounced for the employees of these lower-level managers.

Finally, we expect that changes in leadership behaviors relate to outcomes. More specifically, we will investigate whether changes in control and delegation are associated with the perceived quality of the manager, and perceived productivity. Based on the literature, it can be expected that an increase in exercising control will be negatively correlated with such outcomes (Lautsch et al., 2009), whereas an increase in delegation will be expected to be positively correlated with these variables.

Method

Sample and Procedure

We collected data among employees and managers in the Netherlands between mid-April and mid-May 2020. This time frame allowed us to measure respondents' reactions and perceptions of WFH, because our data collection started about 1 month after the (Dutch) government centrally requested WFH for all nonessential workers from March 16th onwards (Rijksoverheid, 2020). Respondents were approached via a website of a management platform (MT.nl). This online platform offers content to professionals and managers in the Netherlands. Calls to participate in the study were shared via the website, and via social media such as Twitter and LinkedIn. Respondents were offered the opportunity to participate in a gift-voucher raffle in exchange for their participation. We did not set an a priori sample size, but simply maximized participation within a fixed time frame.

We recorded 748 complete responses (316 men, 431 women, and 1 other). The mean age was 47.4 years. For our sample of respondents, 30.5% were in nonmanagement, 23.5% in lower management, 27.7% were in middle management, and 18.4% were in top management. Note that some of these respondents were self-employed and/or currently did not have a manager. In both cases, these participants were excluded from the analysis. In terms of education, 40% completed a bachelor’s degree in Applied Sciences, 52% received a master’s degree, and 8% a high school degree. In terms of the economic sector, 9% worked in manufacturing, 32% worked in trade and commercial or financial services, 46% worked in nonprofit organizations, and 13% in other sectors.

All participants completed the measures below. The data were collected as part of a combined project testing multiple independent research questions in one larger combined questionnaire (including loneliness and leadership ideals). These variables are not discussed here.

Measures

Self-rated Changes in Leadership Behaviors

As explained above, perceived changes in key aspects of the directive and empowering leadership were measured by focusing on two specific leadership behaviors. For a key aspect of directive leadership, we selected the exercise of control, by asking managers: “are you exercising less or more control towards your employees, since the start of the corona-crisis?” For a key aspect of empowering leadership, we selected delegation, and asked managers: “are you delegating less or more towards your employees, since the start of the corona-crisis?” Both items were self-developed. Respondents could answer on a 7-point scale between −3 (a lot less) and +3 (a lot more) with a neutral midpoint 0 (the same).

Changes in Leadership Behavior According to Employees

For employees, we measured the same two leadership behaviors. We asked employees: “is your manager exercising less or more control towards you, since the start of the corona-crisis?” and: “is your manager delegating less or more to you, since the start of the corona-crisis?” respectively. Both items were answered on the same 7-point scale between −3 (a lot less) and +3 (a lot more) with a neutral midpoint 0 (the same).

Leadership Position

Those respondents who currently held a management position (n = 404) indicated their own position, where they could choose between lower (n = 116), middle (n = 169), and higher management (n = 119). Those respondents who were being supervised (n = 561) indicated the position of their own manager on that same scale, between lower (n = 113), middle (n = 183), and higher management (n = 265). In total, 347 respondents indicated that they were managers themselves and that they currently had a manager to whom they reported. These respondents therefore answered both the questions with respect to the perceived changes in their own behaviors, and the questions about the perceived changes in their manager's behaviors.

Manager Quality

Respondents indicated how satisfied they were with the overall quality of their manager on a 7-point scale, between 1 = very unsatisfied and 7 = very satisfied.

Productivity Self-reported; Employees and Colleagues

Respondents indicated if their own productivity and that of their employees or colleagues had improved after the start of the COVID-19 crisis, on a 7-point scale, ranging from 1 (strongly disagree) to 7 (strongly agree).

Controls

To check for possible effects of biographical characteristics of the manager and the employee, we asked for their gender, age, tenure, and education, and percentage WFH. For the latter, respondents had to answer the question “To which degree is your job currently done remotely” (answering options: 0%, 25%, 50%, 75%, and 100%). Approximately 74.5% of the managers and 79.8% of the employees indicated that they worked 75% or more from home at the time of filling out the survey.

Results

Manager Perceptions of Changes in Leadership Behaviors

We first analyzed how managers perceive changes in their own leadership behaviors, see Table 1.

Table 1.

Means for the Full Sample of Managers and for Different Hierarchical Levels (Including ANOVA Results) for Changes in Perceived Leadership Behaviors in the COVID-19 Crisis According to Managers Themselves (Standard Errors Between Brackets).

	Full sample	Lower mgmt.	Middle mgmt.	Top mgmt.	ANOVA
N cell	404	116	169	119	F(2, 401)
Change in control	−0.30 (0.053)	−0.36 (0.106)	−0.28 (0.076)	−0.29 (0.101)	0.238, p = .789
Change in delegation	0.26 (0.049)	0.17 (0.097)	0.26 (0.071)	0.33 (0.096)	0.719, p = .488

Note. Score on a 7-point scale between −3 (a lot less) and +3 (a lot more) with a neutral midpoint 0 (the same). ANOVA = analysis of variance.

Managers themselves perceived on average a decrease in the degree of exercising control (M = −0.30, SD = 1.07; difference from the neutral midpoint, t = −5.731, p < .0001, 95% CIdif [−0.41, −0.20]). Managers also perceived an increase in their level of delegation (M = 0.26, SD = 1.00; difference from the neutral midpoint, t = 5.15, p < .0001, 95% CIdif [0.16; 0.35]). Both results are in line with our expectations. Contrary to our expectation, neither of the two effects differed significantly across management levels (see Table 1, F = 0.238, p = .789 for control, and F = 0.719, p = .488 for delegation).

Employee Perceptions of Changes in Leadership Behaviors

Next, we focused on the questions of whether and how employees perceive changes in the two leadership behaviors. The perceptions of employees are summarized in Table 2.

Table 2.

Means for the Full Sample and for Different Hierarchical Levels (Including ANOVA Results) for Changes in Perceived Leadership Behaviors in the COVID-19 Crisis for Different Hierarchical Levels, According to Employees (Standard Errors Between Brackets).

	Full sample	Lower mgmt.	Middle mgmt.	Top mgmt.	ANOVA
N cell	561	113	183[a]	265	F(2, 558)
Change in control	−0.37 (0.05)	−0.43 (0.128)	−0.49 (0.082)	−0.26 (0.070)	2.204, p = .111, η2p = 0.008
Change in delegation	0.01 (0.049)	−0.29 (0.114)[a]	0.08 (0.082)b	0.10 (0.070)b	5.059, p = .007, η2p = 0.018

Note. Means with different subscripts differ at p < .05 (Tukey's HSD). Score on a 7-point scale between −3 (a lot less) and +3 (a lot more) with a neutral midpoint 0 (the same). ANOVA = analysis of variance.

Note that for delegation the N was 184.

Consistent with the above-mentioned findings on managers' perceptions, results showed that employees also perceived a significant decrease in control (M = −0.37, SD = 1.18; difference from neutral midpoint, t = −7.40, p < .0001, 95% CIdif [−0.47, −0.27]). Again, this decrease did not differ significantly across management levels (Table 2, F = 0.204, p = .111).

We found that employees did, however, not share the managers' perception of increased delegation. Employees perceived no significant change in delegation (M = 0.01, SD = 1.155; difference from the neutral midpoint, t = 0.292, p = .770, 95% CIdif [−0.08, 0.11]). Interestingly, Table 2 also shows that the results differ across management levels (F = 5.059, p = .007). In particular, those who have a lower level manager perceived a significant decrease in delegation (M = −0.29, SD = 1.21; difference from the neutral midpoint, t = −2.570, p = .011, 95% CIdif [−0.52; −0.07]), whereas those supervised at middle management levels saw no difference (M = 0.08, SD = 1.11; difference from the neutral midpoint, t = 1.00, p = .319, 95% CIdif [−0.08; 0.24]), which was also the case for employees reporting to top managers (M = 0.10, SD = 1.15; difference from the neutral midpoint, t = 1.392, p = .165, 95% CIdif [−0.04; 0.24]).

When testing whether the average perceived changes in the two leadership behaviors differ significantly between managers and employees, the independent samples t-test results (equal variances assumed) showed that there was no significant difference for control (t = 0.945, p = .345), but there was a significant difference for delegation (t = 3.506, p = .0001). Additional analyses showed no significant differences for subsamples that relate to gender, age, or percentage working from home, both for managers and employees.

Correlations Between Changes in Leadership Behaviors, and Manager Quality and Productivity

Finally, we investigated whether both managers' and employees' perceptions of changes in leadership behaviors are associated with outcome measures. For managers, we presented the Pearson correlations between perceived changes in their leadership behaviors with the perception of changes in their own productivity and their employees (see Table 3). Table 4 shows that for employees the Pearson correlations between perceived changes in control and delegation with manager quality and productivity.

Table 3.

Univariate Statistics and Pearson's Correlations for Manager Perceptions of Changes in Leadership Behaviors, and Productivity.

Variable	M	SD	1	2	3
(1) Change in control	−0.30	1.07	–
(2) Change in delegation	0.26	0.99	−.05	–
(3) Productivity self	4.04	1.65	.03	.17**	–
(4) Productivity employees	3.79	1.42	.01	.11*	.39**

Note. N = 404.

*p < .05, **p < .01.

Table 4.

Univariate Statistics and Pearson's Correlations for Follower Perceptions of Changes in Leadership Behaviors, Manager Quality, and Productivity.

Variable	M	SD	1	2	3	4
(1) Change in control	−0.37	1.18	–
(2) Change in delegation	0.01	1.16	.04	–
(3) Quality of manager	4.89	1.72	−.04	.27**	–
(4) Productivity self	3.92	1.66	.08	.18**	.01	–
(5) Productivity colleagues	3.66	1.33	.08	.13**	.02	.45**

Note. N = 561.

*p < .05, ** p < .01.

The main result for managers is that perceived changes in their own level of delegation correlated significantly and positively with the productivity of their employees (r = .11, p = .026) and with their own productivity (r = .168, p = .001), indicating that managers see an increase in productivity of themselves and their employees when they report to delegate more after the start of the COVID-19 crisis. For the perceived changes in control there were no significant correlations with these two productivity variables, see Table 3.

For employees, we found comparable results. Table 4 shows that perceived changes in delegation were also significantly positively correlated with manager quality (r = .269, p = .0001) and the productivity as perceived by employees for themselves (r = .177, p = .0001) as well as their colleagues (r = .125, p = .003). Clearly, employees who have a manager who delegates more after the COVID-19 crisis, perceive themselves and their colleagues to be more productive. Again, perceived change in control was not significantly correlated with productivity. Also note that both for managers and employees, there was no significant correlation between the changes in the control and delegation, indicating that the two leadership behaviors are clearly distinct from each other.

To be clear, these are mere associations, and further research is needed to establish whether the reported changes in leadership actually drive changes in outcomes such as productivity. Interestingly, our results are in line with a recent large survey among U.S. workers who also report to be more productive after being enforced to WFH after the start of the COVID-19 crisis (Barrero et al., 2020).

General Discussion

Summary

We investigated whether COVID-19 and the associated need to work from home led to changes in leadership behaviors in a study conducted among Dutch managers and employees. Our results show first of all that managers' perceptions of the changes in their behaviors are in line with the expected or “desired” direction of the WFH concept. That is, managers perceived that they showed significantly less control and more delegation in the early stages of the COVID-19 pandemic (April and May 2020), irrespective of their management level. In contrast, although employees also perceive a significant decrease in the exercise of control by their manager, they do not observe a change in delegation. Moreover, those who are supervised at the lower-management level even perceive a decrease in delegation. Also, managers and employees' perceptions about the change in delegation significantly differed. Finally, our results suggest that when WFH, employees especially appreciate an increase in delegation, because such an increase is associated with higher self-reported productivity and productivity of colleagues, and higher perceived manager quality.

Theoretical and Practical Implications

Our results partly confirm our a priori predictions. In line with the WFH literature (e.g., Bailey & Kurland, 2002; Gajendran and Harrison, 2007; Kniffin et al., 2021), which argues for changes in leadership behaviors, we observe that managers, irrespective of their management level, perceive their behaviors to have changed in the—for WFH—“appropriate” direction. Specifically, managers perceive themselves to be less controlling and more delegating. Especially for the latter, it can be argued that in this context of WFH, delegating responsibilities to employees is a crucial skill or behavior that managers have to aspire. It is therefore likely, especially since employees do not perceive such an increase, that managers might be biased in their perception of these changes in this “desired” behavior. This finding is corroborated by the literature on self-enhancement (Fleenor et al., 2010), particularly in a context of organizational change (Stoker, 2006).

For employees, we proposed that the effect of the combination of WFH and the COVID-19 crisis as an external shock could lead to different and even opposing effects on perceived changes in leadership behaviors. We find that, in line with WFH literature, managers are perceived to be showing less control. However, employees do not perceive an expected increase in delegation, which can be seen as indicative of the fact that managers might find it difficult to show such leadership behaviors in times of a severe crisis, such as COVID-19. Moreover, employees from managers at lower-level management perceive even a significant decrease in delegation, implying that at this level, a threat-rigidity effect (Staw et al., 1981) overrides the WFH context. This result is even more relevant when confronted with our finding that an increase in delegation, as reported by employees, is positively associated with perceived manager quality and productivity during the COVID-19 crisis.

Our paper makes the following theoretical contributions. First, our results enrich the WFH literature. Following this line of research, WFH has positive outcomes for employees, such as perceived autonomy, lower work–family conflict, and productivity (Allen et al., 2014). For employees to be effective, delegation of responsibilities of managers is important (Cortellazzo et al., 2019). Also, in the context of the COVID-19 crisis, our results suggest that managers are instrumental in this respect, in particular via increases in their delegation, which is associated with perceived productivity. Therefore, our results suggest that even in a context where WFH is not a deliberate organizational choice, but instead is implemented in response to an external threat, employees and managers still experience the benefit of this key aspect of empowering leadership when they work from home.

However, the COVID-19 crisis thus also constitutes an external threat which implies that managers, at least according to their employees, do not delegate more than they did before the COVID-19 crisis. As reported previously, the well-known downside for managers in the context of WFH is the fear of losing control. This anxiety seems to be exacerbated by the external threat of the COVID-19 crisis (Kniffin et al., 2021; Rudolph et al., 2020), especially for lower-level managers. In this way, our study contributes to a better understanding of the consequences of an organizational change for leadership behaviors when the external environment is in a state of flux (Feenstra et al., 2020; Oreg & Berson, 2019).

On a more fundamental level, the fact that WFH was implemented in organizations overnight as well as involuntary, makes for an unprecedented organizational change which is completely at odds with the conditions under which WFH has been studied so far (Kniffin et al., 2021; Rudolph et al., 2020). Organizational changes, such as the implementation of WFH, are mostly based on a set of basic change management processes and principles (Stouten et al., 2018) that do not include enforced and unplanned organizational change such as currently is the case. For instance, under “normal” conditions, the success of WFH depends on managers' trust in employees (Kaplan et al., 2018), the possibility of alternative output controls (Groen et al., 2018), and even new monitoring techniques (Bhave et al., 2020). These conditions are clearly not (automatically) met in the COVID-19 situation, when the implementation of WFH in early 2020 was involuntary and abrupt. Therefore, WFH in times of COVID-19 asks for a radical rethink of and consequently novel research into such change situations.

In terms of practical implications, there are a few suggestions that might be helpful for practitioners who have to deal with WFH. First, our results indicate that if there is one leadership behavior that needs to be developed in a WFH context, it is delegation. Crucially, our results show that during the COVID-19 crisis, the perception of changes in delegation differs significantly between managers and employees. Our results serve as a reminder that, on average, manager self-perceptions, like all self-perceptions, suffer from self-enhancement (Dunning et al., 2004), especially when it concerns behaviors that are essential to the new concept, in our case the concept of WFH. Such self-enhancement has implications for managers and practitioners such as HR professionals. For managers, it is recommended to regularly ask for feedback on behaviors from employees and peers. For HR professionals, such differences in self–other ratings can be used in training and development settings, where managers not only learn how to develop new or different leadership behaviors, but also learn how to collect and use feedback (Stoker & Van der Heijden, 2001).

Secondly, it should be taken into account that the alleged positive effects of WFH on (perceived) manager–employee relations are counterbalanced by the impact of the external threat of COVID-19, which reinforces the attention that organizations should pay to the question of whether the warranted increase in delegation of responsibilities actually occurs. In line with the findings of Feenstra et al. (2020), our results seem to confirm that managers might find it difficult to share their power when they are under threat. Therefore, if organizations expect their managers, who also WFH in the context of the COVID-19 crisis, to delegate responsibilities to their employees, it is important to pay attention to feelings of insecurity or instability of these managers, which can be done, for example, by the HR department.

Thirdly, both managers and employees currently have to work from home without the supporting context that needs to go along with this organizational change. Our results not only point at the importance of thinking about necessary activities to train or develop leadership behaviors in the direction that supports WFH, but also at the relevance of providing employees and managers with proper instruments to work from home in an effective manner. Following insights from the WFH and virtual teams' literature, it seems especially important to equip managers with different types of communication technologies to connect and engage with their employees (Crisp & Jarvenpaa, 2013). Consequently, managers should have the knowledge, skills, and abilities to work effectively with such technologies. Moreover, it requires organizations to provide their managers and employees with the technological means, both in terms of hardware and specific software to be able to WFH (Belzunegui-Eraso & Erro-Garcés, 2020).

Limitations and Future Research

One obvious limitation of our findings is that they were based on self-perceptions—which may differ from objective changes in leadership behaviors. Future research may wish to remedy this limitation. At the same time, it will be rather difficult to mimic the simultaneous changes of the two real-world shocks that are central to our paper, WHF and COVID-19, using a research design that focuses on actual behaviors only. A second limitation is that one would like to disentangle the effects on perceptions of behaviors due to the WFH change, from those that stem from the COVID-19 crisis. For example, one possible effect could be that managers find it difficult to delegate responsibilities at the start of this COVID-19 crisis (when we collected our data), but might become more experienced in showing such behavior later on, for example, in the second half of 2020, when WFH became more normal, and the threat of the COVID-19 crisis became less. Future research should look into this matter, by comparing leadership responses at the beginning of the COVID-19 crisis, as we do here, with responses later on. It would also be interesting to investigate the separate effects of the COVID-19 crisis and WFH if over time, hopefully, the threat of the crisis wears off and organizations are left (only) with the WFH legacy of the COVID-19 pandemic. In the latter case, WFH would again become more of a deliberate organizational choice and one could investigate whether or not the effects that our study finds for the WFH change in the wake of the initial lockdown decision persist. Research by Barrero et al. (2020), for instance, finds that U.S. workers who face “mandatory” WFH because of the COVID-19 crisis, would prefer to continue to work significantly more from home once the COVID-19 crisis has been resolved. Finally, our sample consists of 748 highly educated Dutch employees. Future research should also look into the relationship between WFH on the aftermath of the COVID-19 crisis for different institutional settings and samples.