Leader support for recovery: A multi‐level approach to employee psychological detachment from work

Sonnentag, Sabine; Kark, Ronit; Venz, Laura Article — Published Version Leader support for recovery: A multi‐level approach to employee psychological detachment from work Journal of Occupational and Organizational Psychology Provided in Cooperation with: John Wiley & Sons Suggested Citation: Sonnentag, Sabine; Kark, Ronit; Venz, Laura (2024) : Leader support for recovery: A multi‐level approach to employee psychological detachment from work, Journal of Occupational and Organizational Psychology, ISSN 2044-8325, Wiley Periodicals, Inc., Hoboken, NJ, Vol. 97, Iss. 4, pp. 1762-1788, https://doi.org/10.1111/joop.12538 This Version is available at: https://hdl.handle.net/10419/313778 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. http://creativecommons.org/licenses/by/4.0/ 1762 | J Occup Organ Psychol. 2024;97:1762–1788. wileyonlinelibrary.com/journal/joop Received: 15 January 2024 | Accepted: 14 July 2024 DOI: 10.1111/joop.12538 RESEARCH ARTICLE Leader support for recovery: A multilevel approach to employee psychological detachment from work Sabine Sonnentag1 | Ronit Kark2,3 | Laura Venz4 This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2024 The Author(s). Journal of Occupational and Organizational Psychology published by John Wiley & Sons Ltd on behalf of The British Psychological Society. 1Department of Psychology, University of Mannheim, Mannheim, Germany 2Department of Psychology, BarIlan University, Ramat Gan, Israel 3Exeter School of Business, Exeter University, Exeter, UK 4Institute for Management and Organization, Leuphana University Lüneburg, Lüneburg, Germany Correspondence Sabine Sonnentag, Department of Psychology, School of Social Sciences, University of Mannheim, A5, 6, C104, D68159 Mannheim, Germany. Email: [email protected] Funding information SHRM Foundation Abstract This research examines the role of leaders for employee recovery. We hypothesize that leader support for recovery (empathy for recovery, respect for boundaries, and role modelling) relates positively to employees' psychological detachment from work during nonwork time that, in turn, predicts wellbeing outcomes. We argue that leader support for recovery can only be effective when the leadermember exchange (LMX) relationship quality is sufficiently high. In a series of scaledevelopment and scalevalidation studies, we demonstrated the construct and content validity of a new measure of leader support for recovery. We tested our hypotheses with diary data collected from 152 employees. Respect for boundaries was positively related to employees' psychological detachment from work during nonwork time at the person level. Psychological detachment from work was positively related to low emotional exhaustion and a high morning recovery state, both at the person and the day level. LMX moderated the relationship between leader support for recovery (overall measure), empathy for recovery, and respect for boundaries on the one hand and psychological detachment on the other hand, such that the relationships became nonsignificant when LMX was lower. The study suggests that leaders in highquality relationships can contribute to employee recovery – a process that helps to maintain employee wellbeing. KEYWORDS leadership, multilevel methods, recovery, wellbeing | 1763 LEADER SUPPORT FOR RECOVERY BACKGROUND Recovery from daily work is crucial for employee health and wellbeing (Sonnentag et al., 2022; Steed et al., 2021). During stressful times, it is particularly important that employees mentally detach from work during afterwork hours and that they gain some mental distance from work so that they can start the next workday refreshed and with new energy (Clinton et al., 2017; Liu et al., 2021). Research has identified several factors that facilitate versus hinder psychological detachment from work, including specific leisure activities (Hahn et al., 2012; Liu et al., 2021), job demands (Germeys & De Gieter, 2018), and interpersonal job stressors (RodríguezMuñoz et al., 2017). Although there are a few exceptions (Bennett et al., 2016; Tement et al., 2020), the role of leadership for employee recovery in general and for employee psychological detachment in particular has been largely neglected. Leaders, however, might be very important for employee recovery and psychological detachment, respectively, because they influence employees' perspectives on work and on workhome boundaries (Kim & Beehr, 2020; Liao et al., 2016). More specifically, leaders might facilitate their employees' recovery by being mindful of signs of strain and depletion in their employees, acknowledging their recovery needs, and by reminding them about the importance of recovery. However, leaders also might undermine their employees' recovery by violating boundaries between work and nonwork life, for instance, by contacting them during leisure time. To better understand the role leaders play in employee recovery we, first, develop and validate a scale to measure leader support for recovery, and, second, suggest and test a model that introduces leader support for recovery as an important predictor of employee recovery (Figure 1). Specifically, we propose that leader support for recovery is positively related to followers' psychological detachment from work during afterwork hours and that psychological detachment, in turn, is associated with lower levels of exhaustion and higher levels of morning recovery states. We suppose that leader support for recovery may not be effective under all circumstances and examine leadermember exchange (LMX; Liden et al., 1997) as a moderator of the relationship between leader support for recovery and employee psychological detachment from work. To capture the substantial withinperson variation in recoveryrelated constructs and wellbeing (Podsakoff et al., 2019), we collect daily survey data. We include leader support for recovery and LMX as personlevel variables in the multilevel study design. We seek to make several contributions. First, our study adds to the recovery literature by examining the role of leadership for employee recovery. Until now, the vast majority of studies investigating predictors of recovery focused on job demands and job resources (e.g., job control; Bennett et al., 2018; Steed et al., 2021). We extend the predictor space by showing that also the leader, as an important aspect of the interpersonal work environment, plays a core role in facilitating employees' recovery processes. Second, we suggest leader support for recovery as a specific new multidimensional construct. As such, we highlight that leader support for recovery encompasses multiple facets that leaders need to consider when they want to support their followers' recovery processes. We build on earlier work by Bennett Practitioner points • Leaders play an important role in employees' recovery. • Employees can better psychologically detach from their work during nonwork time when leaders respect boundaries between work and nonwork life. Respecting boundaries should be a top priority for leaders. • Leader support for recovery cannot be effective in lowquality LMX relationships. Leaders should support their employees' recovery and strive to build good relationships with them. 1764 | SONNENTAG et al. et al. (2016) that demonstrated that leaders' expectations that employees work during leisure time can undermine employee recovery. However, we go beyond Bennett et al.'s approach and argue that leader support for recovery is more than just not expecting employees to show excessive work commitment during nonwork time. Rather, leader support for recovery includes behaviours that explicitly promote recovery, for instance by expressing empathy for recovery and acting as a role model. To be able to examine the multidimensional nature of leader support for recovery, we develop and validate short scales that capture various aspects of the construct. Third, we contribute to the broader literature on leadership and wellbeing (Harms et al., 2017; Montano et al., 2017) by extending the knowledge about the benefits of leadermember exchange (LMX). While Bennett et al. (2016) reported a direct negative association between high LMX and employee recovery, we demonstrate that a more nuanced perspective on LMX is needed. Although a high LMX might lead to follower overcommitment and neglect of recovery (Bennett et al., 2016; Harris & Kacmar, 2006), a high LMX is needed so that followers can trust that their leader expresses support for recovery out of a benevolent motive. CORE CONCEPTS We define leader support for recovery as a leader's behaviours that “encourage employees to recover during nonwork time” (Bennett et al., 2016, p. 16441). Building on this broad definition, we conceptualize leader's support for recovery as a multidimensional construct comprising empathy for recovery, respect for boundaries, and role modelling as three essential building blocks. We position our conceptualization of leader support for recovery within the broader literature on supervisor social support. Social support in general can be defined as the provision of “psychological or material resources … to a focal individual by partners in some form of social relationship” (Jolly et al., 2021, p. 292) and comprises emotional, informational, and instrumental support as its core dimensions (House et al., 1988; Jolly et al., 2021). Emotional support refers to the expression of empathy and communicates that the target person is cared for and appreciated (Cobb, 1976; Semmer et al., 2008). In the context of leader support for recovery, emotional support means that leaders show empathy for their employees' recovery needs. Informational support refers to giving advice and cognitive guidance and thereby helping to define and understand difficult situations (Cohen & Wills, 1985). One important way by which leaders can 1Although Bennett et al.'s definition is very suitable for capturing the essence of leader support for recovery as a broad multidimensional concept, Bennett et al.'s measure is comparably narrow and focusses on just one aspect of the broader concept (i.e., expectation to work during personal time). FIGURE 1 Conceptual model. | 1765 LEADER SUPPORT FOR RECOVERY informationally support employee recovery is by providing a personal example about how they integrate recovery into a busy life, that is, by role modelling. Generally, instrumental support includes financial help and the provision of other resources and tangible support (Cohen & Wills, 1985). It may also include actions that give “the recipient increased time for activities such as relaxation or entertainment” (Cohen & Wills, 1985, p. 313). Applied to leader support for recovery, the most important instrumental support may not be the provision of any material resource, but to refrain from intruding into employees' free time, that is, to respect boundaries between work and nonwork time. Empathy for recovery Empathy – the tendency to understand and experience internal states of other persons (Clark et al., 2019) – plays an important role in organizational life and is an important aspect of effective leadership (Burch et al., 2016). In line with a more general definition of leader empathy, we define leader empathy for recovery as a leader's “ability to accurately recognize and understand” (Meinecke & Kauffeld, 2019, p. 487) that employees need recovery. Empathy for recovery refers to the emotional aspects of leader social support. It means that leaders are attentive to their employees' recovery needs, that they acknowledge that recovery is important and that they let their employees know that they understand their recovery needs. Respect for boundaries Acknowledging employees' recovery needs is important, but it is not sufficient for effectively supporting employee recovery. To help employees recover, leaders need to respect the boundaries between employees' work and nonwork lives. Per definition, recovery is impeded when work intrudes into nonwork hours (Meijman & Mulder, 1998). These intrusions can take several forms, with completing workrelated tasks and being contacted for work purposes belonging to the most prominent intrusions (Butts et al., 2015; Ten Brummelhuis & Trougakos, 2014). Obviously, leaders are an important source of these intrusions, as they set expectations about task accomplishment and establish norms about jobrelated communication during nonwork hours. To enable recovery, leaders need to hold the boundaries around their subordinates' nonwork time. Typically, leaders who respect boundaries refrain from assigning tasks that need to be completed literally “overnight” and do not contact their employees during afterwork hours or on the weekend. Respect for boundaries is a core component of the instrumental aspect of leader social support. A leader's behaviour is most instrumental for employee recovery when the leader provides time and “peace of mind” by not intruding into employees' periods of rest. Bennett et al.'s (2016) measure of supervisory support for recovery captures this dimension of leader support for recovery as its main feature using items such as “I communicate with subordinates during nonwork time” and “The more productive employees work during nonwork time” (p. 1654; both items reversecoded). Role modelling Role modelling is an important strategy by which leaders can influence their employees (Brown & Treviño, 2014; Yaffe & Kark, 2011). In the recovery context, role modelling means that leaders set a positive example for recovery practices, by communicating about their own recovery needs and activities. For instance, when role modelling recovery, leaders tell their followers about how they integrate recovery into their daily lives, how they manage to take time for recovery, and which recovery activities they enjoy. Role modelling corresponds to the informational aspect of leader social support. According to social learning theory (Bandura, 1977), employees observe how leaders speak 1766 | SONNENTAG et al. about their personal recovery, extract information from their leaders' reports about recovery, and draw conclusions about the value of recovery and how they themselves could implement recovery into their lives. Thus, leaders' role modelling of recovery will have a positive impact on employees' recovery. Benefits of leader role modelling have been demonstrated in the workfamily literature: For instance, when supervisors role model workfamily balance, employees experience less workfamily conflict (Hammer et al., 2009). Leader support for recovery as a personlevel construct In line with earlier research (Bennett et al., 2016; Nie et al., 2021), we treat leader support for recovery as a personlevel variable. The degree to which leaders express empathy for recovery needs, respect employees' boundaries, and role model recovery differs from leader to leader. Of course, leaders must enact these behaviours on certain days, and the personlevel degree of support for recovery is the aggregate of leader behaviours occurring on individual days. We assume that this overall personlevel support for recovery is more relevant than what might happen on each individual day because leaders might not have contact with their employees every day (Barnes et al., 2015) and might not address the recovery topic every day. Most likely, leaders do not need to talk about recovery every day; for employees, it should be more important to know that their leader generally supports recovery. Leaders' support for recovery can also be effective on days when leaders and employees do not meet or do not talk about recovery. HYPOTHESES DEVELOPMENT Leader support for recovery and employee recovery at the person level Leader support for recovery will help employees detach from work during nonwork time. Psychological detachment from work during nonwork time (i.e., temporal mental disengagement from work; Sonnentag & Fritz, 2007) is a core recovery experience and, as such, plays a crucial role for employee wellbeing (Steed et al., 2021). Psychological detachment does not only mean avoiding work during nonwork hours but also refraining from workrelated thoughts. Such thoughts would keep the psychobiological system activated (Brosschot et al., 2006) and therefore impede restoration. Leader support for recovery will be positively related to employees' psychological detachment for several reasons. First, when leaders show empathy for recovery, they acknowledge that recovery is important. They give their employees the feeling that they understand their recovery needs and take them seriously. When employees experience that their recovery needs are taken seriously, they feel understood (Longmire & Harrison, 2018) so that they can see their own recovery needs as being legitimate. This experience, in turn, should help them to respond to these needs by prioritizing recovery during nonwork periods. They know that they need not bother about work during afterwork hours, which makes detachment from work more likely. When leader empathy for recovery is low, however, employees cannot be sure that their recovery needs are legitimate and that they are “allowed” to detach from work. Accordingly, they will be less likely to detach. Second, respecting boundaries facilitates employees' psychological detachment from work. When leaders do not respect boundaries by, for instance, allocating urgent tasks shortly before the end of the workday or by contacting employees after work, they keep employees mentally connected to work and, consequently, employees' psychological detachment will be low. Importantly, not only leaders' actual boundaryviolating behaviours but also the possibility that the leader might cross the boundary and intrude into nonwork time may keep workrelated thoughts activated and impede psychological detachment. However, when employees know that their leaders respect the boundaries, employees can fully disconnect from work during nonwork time. Empirical findings from Bennett et al. (2016) support this assumption. When supervisors endorsed statements such as “I communicate with subordinates during | 1767 LEADER SUPPORT FOR RECOVERY nonwork time”, employees were less likely to belong to a leavingworkbehind recovery profile (i.e., a profile characterized by a high level of psychological detachment). Third, leader role modelling will help employees with psychological detachment from work. When leaders model putting attention on recovery, they convey that recovery is a desirable goal that will inspire employees to strive for recovery as well, for instance by refraining from jobrelated thoughts and activities during nonwork time. Moreover, when role modelling recovery, leaders may even tell how they achieve a state of recovery, transmitting helpful information to their employees, who in turn will learn how to implement these strategies themselves (Morgenroth et al., 2015). Pursuing recovery goals and knowing how to recover will make it more likely that employees achieve psychological detachment from work during nonwork time. Hypothesis 1. At the person level, leader support for recovery is positively related to psychological detachment from work. Leader support for recovery will not only be related to employees' psychological detachment from work but may be reflected in higher employee wellbeing, such as lower levels of exhaustion and higher levels of feeling recovered in the morning. Accordingly, we propose that leader support for recovery will have indirect effects on employee exhaustion and the morning state of being recovered via psychological detachment from work. Exhaustion is a stressrelated experience that “refers to feelings of being overextended and depleted of one's emotional and physical resources” (Maslach et al., 2001, p. 399). Psychological detachment from work during nonwork time should relate negatively to exhaustion. Employees who detach from work forget about their work demands during nonwork time. Thus, work is no longer mentally present for them, and their psychobiological systems that have been taxed during work are no longer activated and can return to their prestressor levels (Meijman & Mulder, 1998). As a consequence, exhaustion declines, resulting in lower exhaustion levels during afterwork hours. In line with this prediction, research found negative personlevel correlations between psychological detachment and exhaustion (Headrick et al., 2022). Taken together, because leader support for recovery should be positively related to psychological detachment from work and because psychological detachment from work should be negatively related to exhaustion, we expect a negative indirect effect from leader support for recovery to exhaustion via psychological detachment from work. Hypothesis 2a. At the person level, leader support for recovery is negatively related to exhaustion via psychological detachment from work. The state of being recovered is “an indicator of successful recovery resulting from activities or experiences pursued during previous leisure time” (Binnewies et al., 2009, p. 70). It results from the successful restoration of resources that had been depleted during work and expresses itself in a feeling of being refreshed and energetic. We propose that leader support for recovery will have a serial indirect effect on an employee's average morning state of being recovered, via psychological detachment from work and low exhaustion. When employees successfully detach from work and consequently experience lower exhaustion levels, they will benefit from the detachment process the next morning as well. Taken together, because leader support for recovery should be positively related to psychological detachment from work, psychological detachment from work should be negatively related to exhaustion, and exhaustion, in turn, should be negatively related to the morning state of being recovered, we expect a positive serial indirect effect from leader support for recovery to the morning state of being recovered via psychological detachment from work and exhaustion. Hypothesis 2b. At the person level, leader support for recovery is positively related to the morning state of being recovered via psychological detachment from work and bedtime exhaustion. 1768 | SONNENTAG et al. The moderating role of leadermemberexchange Research suggests that the relationship quality between leaders and their subordinates has an impact on how employees respond to their leader's behaviours (Greenbaum et al., 2018; Schriesheim et al., 1998). We propose that the relationship between leader and employee (leadermember exchange; LMX) moderates the association between leader support for recovery and employee psychological detachment from work during nonwork time. LMX is a leadership concept relying on core principles of social exchange (Dulebohn et al., 2012). According to the LMX approach, leaders develop distinct exchange relationships with their employees: Lowquality relationships focus on economic exchanges as prescribed in employment contracts. Highquality relationships go beyond these basic economic exchanges and include the exchange of a broader set of resources, such as “deep mutual affection and admiration” (Liden et al., 1997, p. 84). A highquality relationship will not only have a positive impact on employees' job performance and satisfaction (Dulebohn et al., 2012; Martin et al., 2016), but it will also help leader support for recovery to unfold its full potential. Employees in highquality LMX relationships trust their leaders, implying that they have “confidence in (the) character of (the) leader” (Dirks & Ferrin, 2002, p. 613). Applied to the context of recovery, they have confidence that the leader is sincere when expressing a positive view on recovery and when encouraging employees to take time for recovery and temporarily forget about work. This confidence implies that employees can allow themselves to take recovery seriously and to mentally detach from work when they are off the job. This trust in their leader being typical for employees in highquality LMX relationships lets employees “accept vulnerability” (Rousseau et al., 1998, p. 398) when demonstrating that they do not think about their job 24/7 because they can trust that their leader will accept – or even welcome – this temporal detachment from work. Therefore, the support the leader expresses for recovery will be more influential for the recovery processes of employees in highquality relationships. Employees in lowquality relationships, however, will not necessarily trust that their leaders have good intentions when emphasizing the importance of recovery. Leaders' attitudes about recovery will be less relevant for them because they feel more distance towards the leader. Consequently, for employees in lowquality relationships, their leader's support for recovery will have less impact. Accordingly, we suggest that the quality of the LMX relationship is a moderator in the relationship between leader support for recovery and psychological detachment from work. In other words, the quality of the LMX relationship needs to be reasonably high so that employees' detachment from work can benefit from leader support for recovery. Hypothesis 3. At the person level, LMX moderates the relationship between leader support for recovery and psychological detachment from work such that the positive relationship between leader support for recovery and psychological detachment breaks down (i.e., becomes nonsignificant) when LMX is low. Linking Hypotheses 2 and 3, we propose that the indirect relationship of leader support for recovery with exhaustion and the morning state of being recovered via psychological detachment depends on LMX. Employees in a highquality LMX relationship benefit from leader support for recovery in the sense that they show higher detachment from work and, in turn, will experience lower levels of exhaustion and higher levels of being recovered in the morning. Because they trust in their leader, they will feel encouraged that it is psychologically safe to detach from work, and this higher level of detachment will result in more favourable levels of exhaustion and a state of being recovered. Employees in a lowquality LMX relationship, however, will not be able to take advantage of their leader's support for recovery. Consequently, their level of psychological detachment from work will not be affected by leader support for recovery. In turn, it will also be less likely that leader support for recovery is reflected in low exhaustion and a high morning recovery state. We propose: | 1769 LEADER SUPPORT FOR RECOVERY Hypothesis 4. At the person level, we expect a conditional indirect effect such that the indirect effect between leader support for recovery and exhaustion and the morning state of being recovered via psychological detachment from work breaks down (i.e., becomes nonsignificant) when LMX is low. Daylevel processes: psychological detachment from work, bedtime exhaustion, and the morning state of being recovered In addition to the personlevel processes described in Hypotheses 1 to 4, recovery processes will be observed at the day level as well. Studies have shown that psychological detachment happening at the day level is associated with subsequent daylevel wellbeing outcomes (Liu et al., 2021; Sonnentag et al., 2008). We propose that daylevel psychological detachment from work is related to low bedtime exhaustion and, in turn, to a high morning state of being recovered. As described above, when detaching from work during afterhours, strain symptoms that have emerged during the workday are alleviated and exhaustion decreases. Not detaching from work, however, means to continue thinking about workrelated matters that will further draw on a person's energetic resources. Consequently, exhaustion levels will remain high. Indeed, daylevel studies demonstrated negative associations between psychological detachment from work and exhaustion (Bredehorst et al., 2023) and similar outcomes (Germeys & De Gieter, 2018). We seek to replicate these findings in our study: Hypothesis 5a. At the day level, psychological detachment from work is negatively related to bedtime exhaustion. We propose that daylevel psychological detachment from work should be associated with a high morning state of being recovered through low bedtime exhaustion. As argued with respect to Hypothesis 5a, psychological detachment from work should be negatively related to bedtime exhaustion. Low bedtime exhaustion, in turn, should contribute to the feeling of being refreshed in the next morning because it implies that one finishes the evening in a less depleted state that most probably translates into a relatively energetic state in the next morning. In contrast, high bedtime exhaustion reduces the likelihood of being fully recovered the next morning. Hypothesis 5b. At the day level, psychological detachment from work is positively related to the morning state of being recovered via low bedtime exhaustion. OVERVIEW OF STUDIES To assess leader support for recovery, we followed principles of scale development and validation (Colquitt et al., 2019; Hinkin, 1998; Lambert & Newman, 2023). Specifically, we developed measures for empathy for recovery, respect for boundaries, and role modelling within a broader research project on leaders' role for recovery that also aimed at examining leaders' role for taking vacations, encouraging breaks at work, and providing break infrastructure at the workplace. We started the scaledevelopment process by eliciting leaders' broader view on employee recovery in a qualitative study (Study 1). In Studies 2a, 2b, 3a, and 3b, we developed and tested the construct validity of the three specific scales assessing empathy for recovery, respect for boundaries, and role modelling, using a confirmatory factor analytic approach (Lambert & Newman, 2023). In Study 4, we tested content validity. Study 5 is our main study in which we tested our hypotheses.2 2This is the first publication from Studies 1, 2a, 2b, 3a, 4, and 5. Results from Study 3b focusing on a different research question and using different variables have been already published (Sonnentag & Schiffner, 2019). 1776 | SONNENTAG et al. detachment from work, exhaustion, and state of being recovered. Specifically, we modelled the three factors capturing leader support for recovery as a higherorder personlevel factor and LMX as a separate personlevel factor. At the day level, we specified the withinperson components of psychological detachment from work, exhaustion, and state of being recovered. The fit of this twolevel model was good, χ2 = 1233.671, df = 527,4 p < .001, CFI = .930, RMSEA = .036, SCF = .9872, AIC = 32,286.745, BIC = 32,972.316. This model fit the data better than alternative models such as an alternative (nonnested) model in which all four leadership variables loaded on one common personlevel factor, χ2 = 1444.395, df = 533, p < .001, CFI = .910, RMSEA = .040, AIC = 32,486.251, BIC = 33,142.015, and an alternative model in which all variables assessed in the daily surveys loaded on one common factor at the day and the person level, χ2 = 6034.064, df = 537,5 p < .001, CFI = .458, RMSEA = .098, SCF = .9796, AIC = 36,959.940, BIC = 37,595.832, SatorraBentler Δχ2 = 8104.388, Δdf = 10, p < .001. Analytic strategy Because our data had a multilevel structure (i.e., days nested within persons) and because we wanted to make full use of the available data (Newman, 2014), we analysed our data with a twolevel path model with Bayesian estimation, using Blimp 3.0.63 that allows for modelbased imputation of missing data (Enders et al., 2020). We tested Hypotheses 1 to 4 in the betweenperson part of the model (i.e., at the person level) and Hypotheses 5a and 5b in the withinperson part of the model (i.e., at the day level). Because the highorder factor model of leader support for recovery fit the data reasonably well, we tested our hypotheses (1) with an overall measure of leader support for recovery, for which we averaged scores on its three aspects (empathy, respect for boundaries, and role modelling) weighted by the respective factor loadings and (2) separately for the three aspects of leader support for recovery. To adequately test the daylevel mediation proposed in Hypotheses 5a and 5b, we modelled the relationships between psychological detachment, evening exhaustion, and the morning state of being recovered at the person and day level (Preacher et al., 2010). At the person level, we used manifest person means of the variables assessed at the day level. At the day level, we included the day of data collection as a control variable because participating in the study might have an impact on respondents' everyday behaviour (Gabriel et al., 2019). For this purpose, we added a continuous variable with study day coded from 1 to 10 to the analysis. When predicting bedtime exhaustion and morning state of being recovered at the day (i.e., withinperson) level, we controlled for bedtime exhaustion and the morning state of being recovered, respectively, from the previous day of data collection.6 At the day level, we specified paths involving the control variables as fixed and all other paths as random. For testing indirect effects, we relied on the product of coefficient approach (MacKinnon et al., 2002). Results We started our hypotheses tests with a maineffect model without any interaction terms. As the upper part of Table 2 shows, the overall measure of leader support for recovery was unrelated to psychological detachment from work. Consequently, the indirect effects from the overall measure of leader support for recovery to exhaustion via psychological detachment, median = −.041, SD = .032, 95% CI [−.121, 4We fixed two negative nonsignificant residual variances to 0. 5We fixed the same residual variances to 0 as in the comparison model. 6When excluding all control variables from the analysis, results did not change. | 1777 LEADER SUPPORT FOR RECOVERY TABLE 2 Unstandardized coefficients from multilevel modelling including main effects only. Psychological detachment Bedtime exhaustion State of being recovered Median SD 2.5% 97.5% Median SD 2.5% 97.5% Median SD 2.5% 97.5% Person level Intercept 3.559 .077 3.406 3.708 2.777 .070 2.642 2.916 3.255 .066 3.128 3.387 Support for recovery (overall) .272 .139 −.004 .544 .053 .123 −.187 .295 .131 .106 −.078 .335 LMX −.140 .135 −.401 .127 −.282 .117 −.413 −.051 .012 .102 −.188 .216 Psychological detachment – – – – −.168 .075 −.317 −.024 .089 .065 −.041 .214 Bedtime exhaustion – – – – – – – – −.467 .072 −.607 −.323 Residual variance .819 .102 .649 1.050 .638 .078 .507 .810 .464 .057 .370 .591 Variance explained by coefficients 3.5% 10.7% 31.0% Day level Intercept 3.559 .088 3.382 3.729 2.753 .079 2.595 2.906 3.210 .078 3.054 3.361 Day of data collection .002 .010 −.019 .022 −.000 .008 −.017 .016 .011 .009 −.007 .028 Evening exhaustion (previous day) – – – – .043 .051 −.056 .144 – – – – State of being recovered (previous day) – – – – – – – – .023 .042 −.059 .105 Psychological detachment – – – – −.212 .041 −.298 −.136 .014 .049 −.084 .107 Bedtime exhaustion – – – – – – – – −.141 .053 −.248 −.038 Residual variance .638 .034 .576 .710 .402 .023 .358 .445 .426 .027 .377 .481 Variance explained by coefficients .0% 3.0% 1.4% Note: All relationships were tested in one overall model. 1778 | SONNENTAG et al. .003], and the morning state of being recovered via psychological detachment and exhaustion, median = .019, SD = .016, 95% CI [−.001, .059], were not significant. When computing models for the three separate aspects of leader support for recovery,7 we found that respect for boundaries was positively related to psychological detachment from work, median = .203, SD = .090, 95% CI [.025, .382]. The indirect negative effect from respect for boundaries to exhaustion via psychological detachment was significant, median = −.033, SD = .023, 95% CI [−.088, −.001]. The serial indirect positive effect from respect for boundaries to the morning state of being recovered via psychological detachment and exhaustion was significant, median = .015, SD = .011, 95% CI [.0004, .0435]. In contrast to respect for boundaries, both empathy and role modelling were unrelated to psychological detachment from work, and the indirect effects of empathy and role modelling, respectively, on exhaustion and the morning state of being recovered were nonsignificant. Taken together, we found partial support for Hypotheses 1 and 2: Results for respect for boundaries were in line with the hypotheses, but results for empathy, role modelling, and the overall measure were not. In the next step, we included the interaction term between leader support for recovery and LMX in the model (Table 3). The interaction term between the overall measure of leader support for recovery and LMX was a significant predictor of psychological detachment from work. Simple slope analysis showed that when LMX was high (+1 SD), leader support for recovery was positively related to psychological detachment from work, median = .488, SD = .170, 95% CI [.166, .826], but when LMX was low (−1 SD), leader support for recovery was not, median = .079, SD = .163, 95% CI [−.244, .398]. This pattern of findings is in line with Hypothesis 3. Figure 2 (Panel a) shows the simple slopes. The indirect negative effect from leader support for recovery to exhaustion via psychological detachment was significant when LMX was high (+1 SD), median = −.073, SD = .048, 95% CI [−.188, −.003], but nonsignificant when LMX was low (−1 SD), median = −.010, SD = .030, 95% CI [−.080, .045]. The serial indirect positive effect from leader support for recovery to the morning state of being recovered via psychological detachment and exhaustion was significant for high LMX, median = .034, SD = .023, 95% CI [.001, .091], and nonsignificant for low LMX, median = .004, SD = .014, 95% CI [−.021, .039]. Overall, the significant conditional indirect effects provide support for Hypothesis 4. In addition to this analysis relying on the overall measure of leader support for recovery, we ran the moderator analysis for the three separate measures as well.8 Testing Hypothesis 3 in separate analyses showed significant moderator effects for empathy, median = .190, SD = .088, 95% CI [.017, .363], and respect for boundaries, median = .226, SD = .085, 95% CI [.061, .394]. The interaction effect between role modelling and LMX was not significant, median = .175, SD = .095, 95% CI [−.016, .329]. Subsequent simple slope analysis showed that when LMX was high (+1 SD), empathy for recovery was positively related to psychological detachment from work, median = .378, SD = .143, 95% CI [.101, .664], but not when LMX was low (−1 SD), median = .058, SD = .137, 95% CI [−.212, .329]. Similarly, when LMX was high (+1 SD), respect for boundaries was positively related to psychological detachment from work, median = .464, SD = .134, 95% CI [.205, .732], but not when LMX was low (−1 SD), median = .082, SD = .102, 95% CI [−.118, .281]. Panels b and c in Figure 2 show the simple slopes. Tests of conditional indirect effects for empathy for recovery were not significant. Specifically, the indirect negative effect from empathy for recovery to exhaustion via psychological detachment was not significant, neither for high (+1 SD), median = −.049, SD = .037, 95% CI [−.138, .003], nor for low (−1 7Tables reporting main effects for the three separate aspects of leader support for recovery are available online in the Supporting Information (Tables S4–S6). 8Tables reporting the interaction effect for the three separate aspects of leader support for recovery are available online in the Supporting Information (Tables S7–S9). | 1779 LEADER SUPPORT FOR RECOVERY TABLE 3 Unstandardized coefficients from multilevel modelling including the interaction effect between leader support for recovery (overall) and LMX. Psychological detachment Bedtime exhaustion State of being recovered Median SD 2.5% 97.5% Median SD 2.5% 97.5% Median SD 2.5% 97.5% Person level Intercept 3.452 .090 3.276 3.629 2.783 .083 2.620 2.946 3.276 .076 3.128 3.423 Support for recovery (overall) (SUPREC) .285 .140 .016 .558 .052 .124 −.189 .230 .133 .105 −.075 .338 LMX −.073 .137 −.341 .193 −.284 .118 −.517 −.056 .001 .104 −.199 .209 SUPREC × LMX .243 .107 .035 .454 −.015 .097 −.205 .175 −.035 .080 −.195 .120 Psychological detachment – – – – −.162 .076 −.311 −.012 .086 .067 −.046 .218 Bedtime exhaustion – – – – – – – – −.468 .072 −.611 −.327 Residual variance .792 .097 .631 1.007 .642 .079 .510 .819 .465 .057 .372 .598 Variance explained by coefficients 7.9% 11.0% 31.1% Day level Intercept 3.560 .089 3.385 3.734 2.763 .081 2.596 2.917 3.206 .078 3.056 3.362 Day of data collection .002 .010 −.018 .022 .000 .008 −.016 .017 .011 .009 −.007 .029 Evening exhaustion (previous day) – – – – .040 .049 −.054 .137 – – – – State of being recovered (prev. day) – – – – – – – – .023 .042 −.058 .105 Psychological detachment – – – – −.212 .039 −.292 −.139 .015 .048 −.081 .106 Bedtime exhaustion – – – – – – – – −.139 .054 −.247 −.034 Residual variance .639 .034 .577 .710 .401 .023 .359 .449 .426 .026 .377 .479 Variance explained by coefficients .0% 3.0% 1.3% Note: All relationships were tested in one overall model. 1780 | SONNENTAG et al. SD) LMX, median = −.006, SD = .023, 95% CI [−.061, .036]. The serial indirect negative effect from empathy for recovery to the morning state of being recovered via psychological detachment and exhaustion was nonsignificant for both high, median = .022, SD = .017, 95% CI [−.002, .066] and low LMX, median = .003, SD = .011, 95% CI [−.017, .029]. Tests of conditional indirect effects for respect for boundaries were significant: The indirect negative effect from respect for boundaries to exhaustion via psychological detachment was significant when LMX was high (+1 SD), median = −.080, SD = .044, 95% CI [−.183, −.011], but not when LMX was low (−1 SD), median = −.013, SD = .021, 95% CI [−.063, .023]. The serial indirect negative effect from respect for boundaries to the morning state of being recovered via psychological detachment and exhaustion was positive and significant for high LMX, median = .037, SD = .022, 95% CI [.005, .089], and nonsignificant for low LMX, median = .006, SD = .010, 95% CI [−.011, .031]. For role modelling, none of the conditional indirect effects was significant. To sum up, our analyses provide support for Hypothesis 3 when examining empathy for recovery, respect for boundaries, and the overall measure of leader support for recovery. Results for respect for boundaries and the overall measure were in line with Hypothesis 4. For role modelling, however, neither Hypothesis 3 nor Hypothesis 4 received empirical support. Regarding the hypotheses specified for the day level, analysis showed that psychological detachment from work was negatively related to evening exhaustion, median = −.212, SD = .039, 95% CI [−.292, −.139]. Moreover, evening exhaustion was negatively related to the morning state of being recovered, median = −.139, SD = .054, 95% CI [−.247, −.034], and the indirect effect from psychological detachment to the morning state of being recovered via exhaustion was positive and significant, median = .029, SD = .013, 95% CI [.007, .057]. These findings are in line with Hypotheses 5a and 5b. GENERAL DISCUSSION In this paper, we examined the role of leader support for recovery for employees' psychological detachment from work and downstream exhaustion and state of being recovered. Based on a series of scaledevelopment and scalevalidation studies, we presented a short threedimensional measure of leader support for recovery that has good construct and content validity. In our main study (Study 5), we found that respect for boundaries but neither overall support for recovery nor empathy or role modelling predicted employees' psychological detachment from work. LMX moderated the relationship between leader support for recovery and psychological detachment: Only when LMX was high, overall leader support for recovery was positively associated with psychological detachment from work. When LMX was low, however, leader support for recovery was unrelated to psychological detachment. Looking at FIGURE 2 Interaction effects. | 1781 LEADER SUPPORT FOR RECOVERY the different aspects of leader support for recovery, the same interaction pattern became evident for leader empathy and respect for boundaries, but not for role modelling. Psychological detachment from work predicted low exhaustion and a high morning state of being recovered, both at the person and day level. The indirect effect of leader support for recovery to exhaustion via psychological detachment from work was significant when LMX was high but not when LMX was low. Similarly, the indirect effect of leader support for recovery to the morning state of being recovered via psychological detachment and low exhaustion was significant when LMX was high, but not when LMX was low. Theoretical and practical implications Our research offers important theoretical implications as it shows that a highquality relationship between leader and follower (i.e., high LMX) is crucial so that leader support for recovery can unfold its positive potential for employee recovery. A high level of overall leader support for recovery alone was unrelated to employee detachment from work during nonwork time. A high level of LMX, however, provides the necessary context within which leaders' support for employee recovery can become effective. Within highquality relationships, leaders' attitudes and communication about recovery are more likely to actually “reach” the employee and to be taken seriously, resulting in higher levels of psychological detachment. Within lowquality relationships, however, employees discount leaders' attitudes and communication messages so that leaders are less influential for employees' behaviour. A closer look at the three subdimensions of leader support for recovery brings more light into the role of LMX for employee recovery. We introduced leader support for recovery as a multidimensional concept comprising empathy for recovery, respect for boundaries, and role modelling. Our analyses showed that these three subdimensions – although highly correlated – show distinct result patterns in the prediction of employees' psychological detachment. These distinct patterns are particularly informative for understanding the role of LMX for employee recovery. Respect for boundaries was a powerful predictor of psychological detachment, irrespective of LMX. This finding implies that all employees benefit in terms of mentally detaching from work when their leader respects boundaries between work and nonwork time and does not disturb their employees during nonwork time. Empathy for recovery, however, did not show a significant direct association with psychological detachment from work, implying that – across all employees – empathy for recovery does not help to detach from work. The significant interaction effect between LMX and empathy for recovery suggests that a leader's empathy for recovery is only effective for employees who have a good relationship with their leader. Only when LMX is high, empathy for recovery can unfold its benefits because employees feel emotionally close to their leaders and can trust leaders' good intentions (Dulebohn et al., 2012). When LMX is low, however, empathy for recovery remains ineffective because when employees do not trust their leaders, employees may question if it is really in their own best interest to take time for recovery and fully detach from work. Another possible explanation for the interaction effect between LMX and overall leader support for recovery can be derived from leaders' way of framing goals. Kakkar (2019) suggested that in lowquality LMX relationships, leaders frame goals in a manner that primes a prevention focus among subordinates, who then focus on their obligations, expectations, and survival in the workplace. This concern by employees in a lowquality relationship may lead to overpreoccupation and overinvestment in work, which will result in less detachment and recovery after work hours, even when their managers verbalize support for recovery. Recently, Volmer et al. (2023) suggested that highquality LMX will be associated with state positive affect at work that in turn will be associated with higher levels of recovery from work during the evening. This line of research implies that when managers support recovery, a high LMX along with its associated positive affect may result in higher levels of recovery. This possibility should further be explored in future studies. With respect to the three subdimensions of leader support for recovery, respect for boundaries related directly to psychological detachment and showed the hypothesized interaction effect with 1782 | SONNENTAG et al. LMX. As explained above, respect for boundaries seems to be effective for all employees, even for lowLMX employees. Respect for boundaries might be so important because not only leaders' objective actual intrusions into their employees' nonwork time can harm psychological detachment, but just employees expecting that leaders might intrude impair psychological detachment. Therefore, even very rare actual intrusions might reduce employees' psychological detachment on evenings when no intrusion occurs. Our analysis further showed that empathy for recovery needs a highLMX environment to be effective in stimulating detachment. In such an environment, however, empathy for recovery is important because it signals that employees can take the liberty to detach from work. Finally, role modelling turned out to be ineffective for employee recovery as it showed neither a main effect nor a significant interaction effect with LMX in predicting psychological detachment. Although general role modelling of worklife balance is a helpful leadership behaviour with respect to workfamily conflict (Hammer et al., 2009) and employee workhome segmentation behaviour (Koch & Binnewies, 2015), in our research, role modelling of leisure and recovery activities did not help with psychological detachment, even not when LMX was high. This finding suggests that a leader who shows that they take time for leisure and recovery themselves does not help in employees' detachment from work. Maybe, a leader's role modelling encourages employees' engagement in leisure activities, but engagement in leisure activities may not translate into psychological detachment from work. Moreover, as the leader's life situation might differ largely from that of their employees (e.g., in terms of financial resources and leisure interests), employees might not identify sufficiently with their leader with respect to recovery processes, and this lack of identification may make role modelling ineffective (Morgenroth et al., 2015). During the past decade, researchers increasingly identified approaches that leaders may use to foster employee health and wellbeing, such as valuing health and being aware of potential health problems (Franke et al., 2014). Whereas this line of research has started to examine how healthpromoting leadership relates to other leadership constructs (Kaluza et al., 2021), our finding on the moderator effect of LMX may help to better understand the circumstances that are needed so that health promotion by leaders can be successful. Our findings can inspire this broader field of healthpromoting leadership to conceptualize specific healthpromoting behaviours, including leader support for recovery. We examined leader support for recovery as a personlevel variable. Accordingly, our findings imply that leader support for recovery can unfold its benefits – along with LMX – even on days when leaders and employees do not meet or when leaders do not express their support on a specific day. This result is important as it suggests that leaders do not need to show support for recovery every day – the employees' general perception that their leader is supportive of recovery does suffice. In terms of practical implications, our research showed that employees' psychological detachment from work during nonwork time is most likely when both overall leader support for recovery and LMX are high. With respect to leader support for recovery, particularly respect for boundaries is important. Thus, leaders may increase the likelihood that employees switch off from work during nonwork time by strictly respecting boundaries between work and nonwork life. First, leaders should refrain from contacting employees during the evening or on the weekend, for instance by avoiding phone calls or refraining from instant messaging. Second, leaders should allocate tasks and set deadlines in a way that employees do not need to spend afterhours on job tasks. In some companies, respecting boundaries may not be at the discretion of individual leaders but might require a change in company culture. Accordingly, top management should encourage managers to respect boundaries, implying that also leaders' boundaries are respected (Venz & Wöhrmann, 2023). Although leader empathy for recovery was not directly related to employees' psychological detachment from work, it predicted psychological detachment in interaction with LMX. Thus, leaders should be aware of employees' recovery needs and should be understanding when recovery is needed. Similar to general approaches to healthoriented leadership (Franke et al., 2014), leaders might become more aware of employees' recovery needs when they are more aware of and responsive towards their own recovery needs. Moreover, leaders might be more empathetic towards employees' recovery needs when they cultivate compassion within their work setting (Dutton et al., 2014). | 1783 LEADER SUPPORT FOR RECOVERY Having a good relationship with one's leader (i.e., high LMX) is important to take advantage of leader support for recovery. Thus, both leaders and employees may want to strive to improve their exchange relationship, for instance by focusing on similarity (Dulebohn et al., 2012), using delegation (Bauer & Green, 1996), and sharing personal experiences (Yaffe & Kark, 2011). Moreover, to establish a highquality LMX, leaders should provide especially new employees with guidance and advice surrounding their jobs and their roles within the organization (Sluss & Thompson, 2012). In terms of practical use of measures to assess leader support for recovery, our scales cover a broader construct space than does the Bennett et al. (2016) measure. Thus, when not only interested in respect for boundaries but also the leader's emotional underpinning of behaviours that should promote employee recovery, the assessment of empathy for recovery is important – also because results of this assessment could inform training interventions for leaders. Such interventions might not only emphasize respect for boundaries but incorporate a module that targets leader empathy. Regarding the respectforboundaries dimension, the Bennett et al. (2016) measure emphasizes leaders' views on highly performing employees and associated expectations, whereas our measure – in addition to its assessment of empathy and role modelling – more directly addresses leaders' behaviour of not intruding into employees' personal lives. Therefore, it could be seen as a measure that is conceptually closer to leader behaviour that deliberately supports employee recovery. Limitations and directions for future research Despite the new insights it provides, our research has some limitations. First, we assessed our variables with selfreported measures provided by employees. Obviously, this might raise concerns about common method bias (Podsakoff et al., 2003). Although we cannot fully rule out common method bias, our specific study design, including both personlevel and daylevel measures, reduces it. We assessed leader support for recovery and LMX with personlevel measures in the entrance survey and all other variables with daylevel measures, which we aggregated to the person level for personlevel analyses. This approach implies that the measures of leader support for recovery and LMX on the one hand and measures of the remaining variables on the other hand were assessed at different time points and with different temporal referents, making an overestimation of the empirical association less likely. Moreover, one of our core findings refers to the moderator effect of LMX – a result pattern that is less susceptible to common method bias (Siemsen et al., 2010). Second, related to measurement issues as well, we measured leader support for recovery from the employees' perspective, omitting the leaders' perspective. As often leaders and employees do not agree about leadership behaviour (Lee & Carpenter, 2018), it might be that leaders have a view on their support for recovery that differs from employees' perspective. However, we included leaders in Study 1 and used leaders' reports of critical recovery situations for developing our items. In a next step, it would be interesting to use our scales to assess leaders' views on how they support recovery – although employees' perception of leader support is probably the more relevant factor in driving employee recovery. Third, our research focused on the direct relation between leader support for recovery and employee detachment from work, neglecting possible underlying pathways. Thus, it remains unclear how an employee's perception that their leader supports recovery translates into psychological detachment from work. In addition to explicit communication about the need to recover, a leader might shape low availability and response expectations for afterwork hours that make it easier for employees to forget about work when being at home (Venz & Wöhrmann, 2023). In addition, leaders who support recovery possibly instil a recovery climate within their teams so that not only the leader endorses recovery during nonwork time, but that team members mutually reinforce a positive view on recovery. Future studies should address the mechanisms by which leader support for recovery helps detachment from work. Although role modelling of recovery was not a significant predictor of employee detachment from work in our research, future studies may want to continue to address this interesting facet of leader behaviour. For instance, it has been argued that leaders' engagement in serious leisure influences leader 1784 | SONNENTAG et al. performance (Bunea et al., 2023). Knowing that one's leader pursues a serious interest beyond work may inspire employees to explore a similar option for themselves, which, in turn, should help them to detach from work. In addition, it might be worthwhile to explore situational and personal circumstances under which leaders' recovery role modelling does play a role for employees' psychological detachment from work. We focused on how leader support for recovery relates to employee psychological detachment from work and downstream exhaustion and state of being recovered. Future research might want to examine how leader support for recovery affects leaders' own recovery. Although an emphasis on recovery might also benefit leaders, less positive scenarios are possible as well. For instance, by protecting their employees from overwork and ensuring that employees get enough recovery, not all tasks might get completed and, consequently, leaders might cover up for any unfinished tasks, which, in turn, reduces leaders' own recovery (Syrek et al., 2017). Similarly, when respecting employees' boundaries, work processes might become more effortful and cumbersome from the leaders' perspective. CONCLUSION With everincreasing demands on employees, it is important that employees detach from work and get enough recovery during nonwork time so that they can start their workdays refreshed. Our research shows that leaders can support psychological detachment from work, particularly by showing empathy for recovery and respecting boundaries. Importantly, leader support for recovery works best in high LMX relationships. Overall, this research contributes to the accumulating evidence that leaders play a core role in protecting employee health and wellbeing. AUTHOR CONTRIBUTIONS Sabine Sonnentag: Conceptualization; funding acquisition; writing – original draft; formal analysis; data curation; methodology. Ronit Kark: Conceptualization; funding acquisition; writing – review and editing; methodology; formal analysis; data curation; methodology. Laura Venz: Investigation; writing – review and editing; data curation; methodology. ACKNOWLEDGEMENTS This research was supported by a SHRM Foundation grant to the first two authors. This grant is gratefully acknowledged. We thank Tamir Rubin, Yarden Keinan, and Catarina Schiffner for their help in collecting and coding the data. Open Access funding enabled and organized by Projekt DEAL. FUNDING INFORMATION This study was funded by a grant from the SHRM Foundation. However, the interpretations, conclusions and recommendations are those of the authors and do not necessarily represent the views of the SHRM Foundation. CONFLICT OF INTEREST STATEMENT The authors do not have any conflict of interest. DATA AVAILABILITY STATEMENT The data that support the findings of this study are available on request from the corresponding author. The data are not publicly available due to privacy or ethical restrictions. ORCID Sabine Sonnentag https://orcid.org/0000-0002-9464-4653 Ronit Kark https://orcid.org/0000-0003-0506-9683 Laura Venz https://orcid.org/0000-0003-0642-988X | 1785 LEADER SUPPORT FOR RECOVERY REFERENCES Anderson, J. C., & Gerbing, D. W. (1991). Predicting the performance of measures in a confirmatory factor analysis with a pretest assessment of their substantive validities. Journal of Applied Psycholog y, 76(5), 732–740. https:// doi. org/ 10. 1037/ 00219010. 76.5. 732 AnismanRazin, M., Kark, R., & Ashforth, B. E. (2023). “Doing distance”: The role of managers' enactment of psychological distance in leader–follower relationships. Group & Organization Management, 105960112311582. https:// doi. org/ 10. 1177/ 10596 01123 1158264 Bandura, A. (1977). Social learning theory. Prentice Hall. Barnes, C. M., Lucianetti, L., Bhave, D. P., & Christian, M. S. (2015). You wouldn't like me when I'm sleepy': Leaders' sleep, daily abusive supervision, and work unit engagement. Academy of Management Journal, 58(5), 1419–1437. https:// doi. org/ 10. 5465/ amj. 2013. 1063 Bauer, T. N., & Green, S. G. (1996). Development of leadermember exchange: A longitudinal test. Academy of Management Journal, 39(6), 1538–1567. https:// doi. org/ 10. 5465/ 257068 Bennett, A. A., Bakker, A. B., & Field, J. G. (2018). Recovery from workrelated effort: A metaanalysis. Journal of Organizational Behavior, 39(3), 262–275. https:// doi. org/ 10. 1002/ job. 2217 Bennett, A. A., Gabriel, A. S., Calderwood, C., Dahling, J. J., & Trougakos, J. P. (2016). Better together? Examining profiles of employee recovery experiences. Journal of Applied Psychology, 101(12), 1635–1654. https:// doi. org/ 10. 1037/ apl00 00157 Binnewies, C., Sonnentag, S., & Mojza, E. J. (2009). Daily performance at work: Feeling recovered in the morning as a predictor of daylevel job performance. Journal of Organizational Behavior, 30(1), 67–93. https:// doi. org/ 10. 1002/ job. 541 Bredehorst, J., Krautter, K., Meuris, J., & Jachimowicz, J. M. (2023). The challenge of maintaining passion for work over time: A daily perspective on passion and emotional exhaustion. Organization Science, 35, 364–386. https:// doi. org/ 10. 1287/ orsc. 2023. 1673 Brosschot, J. F., Gerin, W., & Thayer, J. F. (2006). The perseverative cognition hypothesis: A review of worry, prolonged stressrelated activation, and health. Journal of Psychosomatic Research, 60(2), 113–124. https:// doi. org/ 10. 1016/j. jpsyc hores. 2005. 06. 074 Brown, M. E., & Treviño, L. K. (2014). Do role models matter? An investigation of role modeling as an antecedent of perceived ethical leadership. Journal of Business Ethics, 122(4), 587–598. https:// doi. org/ 10. 1007/ s1055 101317690 Bunea, E., Kark, R., & Hammond, M. (2023). “Leisureship”: Impact of pursuing serious leisure on leaders' performance. Human Resource Management Review, 33(2), 100950. https:// doi. org/ 10. 1016/j. hrmr. 2022. 100950 Burch, G. F., Bennett, A. A., Humphrey, R. H., Batchelor, J. H., & Cairo, A. H. (2016). Unraveling the complexities of empathy research: A multilevel model of empathy in organizations. In Emotions and organizational governance (Vol. 12, pp. 169–189). Emerald Group Publishing Limited. Butts, M. M., Becker, W. J., & Boswell, W. R. (2015). Hot buttons and time sinks: The effects of electronic communication during nonwork time on emotions and worknonwork conflict. Academy of Management Journal, 58(3), 763–788. ht t ps:// doi . org/ 10. 5465/ amj. 2014. 0170 Clark, M. A., Robertson, M. M., & Young, S. (2019). “I feel your pain”: A critical review of organizational research on empathy. Journal of Organizational Behavior, 40(2), 166–192. https:// doi. org/ 10. 1002/ job. 2348 Clinton, M. E., Conway, N., & Sturges, J. (2017). It's tough hangingup a call': The relationships between calling and work hours, psychological detachment, sleep quality, and morning vigor. Journal of Occupational Health Psycholog y, 22(1), 28–39. ht t ps:// doi. org/ 10. 1037/ ocp00 00025 Cobb, S. (1976). Social support as a moderator of life stress. Psychosomatic Medicine, 38(5), 300–314. https:// doi. org/ 10. 1097/ 00006 84219760 900000003 Cohen, S., & Wills, T. A. (1985). Stress, social support, and the buffering hypothesis. Psychological Bulletin, 98(2), 310–357. ht t ps:// doi. org/ 10. 1037/ 00332909. 98.2. 310 Colquitt, J. A., Sabey, T. B., Rodell, J. B., & Hill, E. T. (2019). Content validation guidelines: Evaluation criteria for definitional correspondence and definitional distinctiveness. Journal of Applied Psycholog y, 104(10), 1243–1265. https:// doi. org/ 10. 1037/ apl00 00406 Dirks, K. T., & Ferrin, D. L. (2002). Trust in leadership: Metaanalytic findings and implications for research and practice. Journal of Applied Psycholog y, 87(4), 611–628. https:// doi. org/ 10. 1037// 00219010. 87.4. 611 Dulebohn, J. H., Bommer, W. H., Liden, R. C., Brouer, R. L., & Ferris, G. R. (2012). A metaanalysis of antecedents and consequences of leadermember exchange: Integrating the past with an eye toward the future. Journal of Management, 38(6), 1715–1759. https:// doi. org/ 10. 1177/ 01492 06311 415280 Dutton, J. E., Workman, K. M., & Hardin, A. E. (2014). Compassion at work. The Annual Review of Organizational Psycholog y and Organizational Behavior, 1(1), 277–304. https:// doi. org/ 10. 1146/ annur evorgps ych03141 3091221 Enders, C. K., Du, H., & Keller, B. T. (2020). A modelbased imputation procedure for multilevel regression models with random coefficients, interaction effects, and nonlinear terms. Psychological Methods, 25(1), 88–112. https:// doi. org/ 10. 1037/ met00 00228 Flanagan, J. C. (1954). The critical incident technique. Psychological Bulletin, 51(4), 327–358. https:// doi. org/ 10. 1037/ h0061470 Franke, F., Felfe, J., & Pundt, A. (2014). The impact of healthoriented leadership on follower health: Development and test of a new instrument measuring healthpromoting leadership. German Journal of Human Resource Management, 28(1–2), 139–161. https:// doi. org/ 10. 1177/ 23970 02214 0280010