Introduction
In the past decade, abusive supervision, that is, subordinates' perceptions of ‘the extent to which supervisors engage in the sustained display of hostile verbal and nonverbal behaviors, excluding physical contact’ (Tepper, Reference Tepper2000: 178), has been extensively investigated by practitioners and scholars in the organizational sciences (Harris, Kacmar, & Zivnuska, Reference Harris, Kacmar and Zivnuska2007; Harvey, Harris, Gillis, & Martinko, Reference Harvey, Harris, Gillis and Martinko2014). This destructive behavior of supervisors usually involves nonphysical actions like unfair demands and expectations from employees, disrespect in public, social isolation in the organization, exposing and exploiting subordinates' weaknesses, overburden, tight control over subordinates, threat, abusive language, insulting, and criticizing subordinates, name-calling, etc. (Tepper, Duffy, Henle, & Lambert, Reference Tepper, Duffy, Henle and Lambert2006a; Tepper, Moss, & Duffy, Reference Tepper, Moss and Duffy2011; Tepper, Uhl-Bien, Kohut, Rogelberg, Lockhart, & Ensley, Reference Tepper, Uhl-Bien, Kohut, Rogelberg, Lockhart and Ensley2006b). Some reasons for examining abusive supervision include its deleterious impact on individuals, as well as organizational outcomes and the increasing frequency at which such behavior is exhibited at the workplace (Khan, Moss, Quratulain, & Hameed, Reference Khan, Moss, Quratulain and Hameed2018; Walter, Lam, Van der Vegt, Huang, & Miao, Reference Walter, Lam, Van der Vegt, Huang and Miao2015). Therefore, scholars need to further investigate this escalating phenomenon to develop interventions that can help managers, practitioners, and organizations at reducing or eliminating such destructive supervisory behaviors (Tariq & Ding, Reference Tariq and Ding2018; Tariq & Weng, Reference Tariq and Weng2018; Tariq, Weng, Ilies, & Khan, Reference Tariq, Weng, Ilies and Khan2021).
While investigating the outcomes, existing research assumes a static approach of abusive supervision, i.e., some managers are abusive while others are not (e.g., see Burton, Hoobler, & Scheuer, Reference Burton, Hoobler and Scheuer2012; Eissa & Lester, Reference Eissa and Lester2017; Hoobler & Brass, Reference Hoobler and Brass2006; Nandkeolyar, Shaffer, Li, Ekkirala, & Bagger, Reference Nandkeolyar, Shaffer, Li, Ekkirala and Bagger2014; Tariq & Weng, Reference Tariq and Weng2018). Organizational scholars have neglected the potential within-person temporal variation approach of abusive supervision where it is measured in terms of occurrence rather than labeling a supervisor as abusive or nonabusive, i.e., some managers frequently engage in abusive supervisory behaviors, and in contrast, other managers do not (see Barnes, Lucianetti, Bhave, & Christian, Reference Barnes, Lucianetti, Bhave and Christian2015; Courtright, Gardner, Smith, McCormick, & Colbert, Reference Courtright, Gardner, Smith, McCormick and Colbert2016; Tariq & Ding, Reference Tariq and Ding2018 for exceptions). This study extends the latter stream of research and focuses on the within-person temporal variation rather than its static approach.
Up until now, organizational scholars have paid attention to investigating the numerous consequences of abusive supervision in the workplace (see Martinko, Harvey, Brees, & Mackey, Reference Martinko, Harvey, Brees and Mackey2013; Zhang & Liao, Reference Zhang and Liao2015). Nevertheless, the majority of research in this area has focused on work-domain outcomes of abusive supervision, such as employees' creativity (Liu, Liao, & Loi, Reference Liu, Liao and Loi2012; Liu, Zhang, Liao, Hao, & Mao, Reference Liu, Zhang, Liao, Hao and Mao2016), knowledge sharing (Wu & Lee, Reference Wu and Lee2016), workplace deviance (Lian, Brown, Ferris, Liang, Keeping, & Morrison, Reference Lian, Brown, Ferris, Liang, Keeping and Morrison2014), job performance (Harris, Kacmar, & Zivnuska, Reference Harris, Kacmar and Zivnuska2007; Tariq & Ding, Reference Tariq and Ding2018), and organizational citizenship behaviors (Zellars, Tepper, & Duffy, Reference Zellars, Tepper and Duffy2002). Limited research has been conducted to investigate the cross-domain outcomes of abusive supervision, that is, the potential deleterious influence of abusive supervision on an employee's personal and family life beyond the boundaries of an organization, for example, work-family conflict (Carlson, Ferguson, Hunter, & Whitten, Reference Carlson, Ferguson, Hunter and Whitten2012), family undermining (Hoobler & Brass, Reference Hoobler and Brass2006; Restubog, Scott, & Zagenczyk, Reference Restubog, Scott and Zagenczyk2011), and relationship conflict or tension with one's partner (Carlson, Ferguson, Perrewé, & Whitten, Reference Carlson, Ferguson, Perrewé and Whitten2011). Therefore, we focus on one common but heretofore neglected cross-domain outcome of abusive supervision: subordinates' overeating behavior. More specifically, in this research, we answer how and when abusive supervision leads to subordinates' overeating behavior.
There has been a greater concern in the general public about healthy eating habits as they are vital in developing a healthy lifestyle (Hesslink, Reference Hesslink2016; Liu, Song, Koopmann, Wang, Chang, & Shi, Reference Liu, Song, Koopmann, Wang, Chang and Shi2017). Extant literature in applied psychology and management has neglected to investigate an employee's eating behavior (see Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017 for an exception). In particular, the relationship between workplace stressor (i.e., abusive supervision; Nandkeolyar et al., Reference Nandkeolyar, Shaffer, Li, Ekkirala and Bagger2014; Tepper, Reference Tepper2000) and overeating behavior has been largely overlooked. Overeating behavior is considered to be present in anyone who is engaged in violation of eating in moderation (Herman, Polivy, & Leone, Reference Herman, Polivy, Leone and Mela2005; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017). Such behavior usually includes eating at unnecessary times or events (e.g., when an individual is taking late-night snacks) or consuming food when not needed (in this case an individual is suffering rather than being satisfied; Colles, Dixon, & O'brien, Reference Colles, Dixon and O'brien2007). We, from the perspective of self-regulation theory (SRT), propose two reasons based on subordinates' self-regulatory failure or impairment (Thau & Mitchell, Reference Thau and Mitchell2010) that answer the question of how abusive supervision leads to subordinates' overeating behavior.
Firstly, as stated in SRT, every individual has a single and limited pool of regulatory resources (Baumeister, Bratslavsky, Muraven, & Tice, Reference Baumeister, Bratslavsky, Muraven and Tice1998), available for constraining counter-normative or undesirable behavioral intentions (Beal, Weiss, Barros, & MacDermid, Reference Beal, Weiss, Barros and MacDermid2005). According to this theory, when an individual experiences a resource-demanding situation or those that require self-regulation (i.e., abusive supervision), he/she suffers from self-regulatory failure or impairment (Thau & Mitchell, Reference Thau and Mitchell2010; Wang, Liao, Zhan, & Shi, Reference Wang, Liao, Zhan and Shi2011), and has a decreased capacity to exert control over his/her counter-normative or undesirable behavioral intentions (Barnes et al., Reference Barnes, Lucianetti, Bhave and Christian2015; Muraven & Baumeister, Reference Muraven and Baumeister2000). It can thus be speculated, from the viewpoint of self-regulation impairment, that an employee experiencing a resource-demanding situation (i.e., abusive supervision) might find it challenging to resist the impulses or urges of consuming unhealthy food (referred here as employees' overeating behavior) and therefore lose control over his/her behavioral intentions toward overeating behavior.
Secondly, subordinates experience self-regulatory resource depletion and undesirable states (i.e., negative mood) when they encounter work-related stressors or negative events, such as workplace incivility (Meier & Gross, Reference Meier and Gross2015; Rusting & Nolen-Hoeksema, Reference Rusting and Nolen-Hoeksema1998). In line with this, it can be argued that abusive supervision, as a negative event or workplace stressor, (Nandkeolyar et al., Reference Nandkeolyar, Shaffer, Li, Ekkirala and Bagger2014) leads to subordinates' negative mood at the workplace due to self-regulatory resource depletion. Moreover, Tice and Bratslavsky (Reference Tice and Bratslavsky2000) and Tice, Bratslavsky, and Baumeister (Reference Tice, Bratslavsky and Baumeister2001) found that an individual, in the presence of a negative mood, is motivated to regulate it to attain immediate pleasure (e.g., overeating). Thus, from the perspective of self-regulation impairment, we propose that a subordinate's negative mood mediates the relationship between abusive supervision and their overeating behavior.
For a better understanding of the relationship between abusive supervision and subordinate's overeating behavior, we draw on job-recovery literature and introduce a contextual factor (i.e., subordinate's recovery experiences at the workplace) that can mitigate the effects of abusive supervision on his/her overeating behavior. On-job recovery experiences refer to the extent to which an employee perceives that the breaks at the workplace help him/her to regain the depleted self-regulatory resources (Demerouti, Bakker, Sonnentag, & Fullagar, Reference Demerouti, Bakker, Sonnentag and Fullagar2012; Sonnentag & Natter, Reference Sonnentag and Natter2004). An employee's on-the-job recovery experiences, such as socializing, napping, and relaxing, are related to a lower level of workplace negative mood and facilitate in reshaping his/her workplace behavior (Trougakos, Beal, Green, & Weiss, Reference Trougakos, Beal, Green and Weiss2008). Therefore, we propose that a subordinate's recovery experiences at the workplace can mitigate the effects of abusive supervision on his/her overeating behavior via negative mood.
This research integrates self-regulatory theory (Baumeister et al., Reference Baumeister, Bratslavsky, Muraven and Tice1998) with the job recovery literature (Mojza, Lorenz, Sonnentag, & Binnewies, Reference Mojza, Lorenz, Sonnentag and Binnewies2010; Sonnentag & Natter, Reference Sonnentag and Natter2004) to shed light on the consequences of abusive supervision on employees beyond the workplace. First, drawing from the perspective of self-regulatory resource impairment or depletion, we propose that employees' resources are likely to be depleted when they experience a resource-demanding situation, such as abusive supervision. Second, we introduce the negative mood of employees as the underlying mechanism of the maladaptive response of employees to abusive supervision in the form of overeating behavior. Finally, deriving from job-recovery literature, we argue that employees' recovery experiences at the workplace tend to mitigate the depleting effects of abusive supervision, such that they are less susceptible to the effects of abusive supervision on overeating behavior when they encounter a high level of recovery experiences at the workplace. A summary of our moderated mediation model is depicted in Figure 1.
Fig. 1. A proposed moderated mediation model.
Literature review and hypotheses development
Employees at the workplace may encounter such events, situations, or behaviors that provoke undesirable emotional reactions in them and cause resource depletion or drain (see Barber, Taylor, Burton, & Bailey, Reference Barber, Taylor, Burton and Bailey2017; Barnes et al., Reference Barnes, Lucianetti, Bhave and Christian2015; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017; Shillamkwese, Tariq, Obaid, Weng, & Garavan, Reference Shillamkwese, Tariq, Obaid, Weng and Garavan2020). More specifically, concerning the leader–follower relationship, followers may face criticism, negative comments, or silent treatment from their immediate leaders (referred here as abusive supervision; Tepper, Reference Tepper2000, Reference Tepper2007). Organizational scholars have noted and defined abusive supervision as a work-related stressor (Nandkeolyar et al., Reference Nandkeolyar, Shaffer, Li, Ekkirala and Bagger2014) that can distract or sidetrack employees from performing their focal tasks. On one hand, such events may instigate tempting impulses or urges in followers as a response to such events (Tepper, Simon, & Park, Reference Tepper, Simon and Park2017; Thau & Mitchell, Reference Thau and Mitchell2010). On the other hand, to respond less impulsively, employees may attempt to regulate aversive emotional states, redirect their attention to focal tasks, and avoid the potential negative appraisal (Beal et al., Reference Beal, Weiss, Barros and MacDermid2005). Such measures can exhort employees to consume self-regulatory resources, defined as ‘the amount of mental capacity available to control and alter naturally occurring emotions, behaviors, and mental states’ (Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017: 1239). Congruent with self-regulatory theory (Baumeister, Reference Baumeister2003; Baumeister, Muraven, & Tice, Reference Baumeister, Muraven and Tice2000), when these limited and finite self-regulatory resources are excessively consumed to deal with abusive supervision, the employees may experience self-regulatory depletion or impairment. The self-regulation process plays an important role in maintaining employees' healthy eating habits and suppressing their overeating urges (i.e., unhealthy eating habits; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017; Tice & Bratslavsky, Reference Tice and Bratslavsky2000; Tice, Bratslavsky, & Baumeister, Reference Tice, Bratslavsky and Baumeister2001). Coping with abusive supervision can therefore elicit a maladaptive coping strategy of unhealthy eating (i.e., overeating behavior) in depleted subordinates.
To explain the underlying mechanism of the proposed relationship we draw on mood-regulation literature, which suggests that negative mood prompts a desire for mood-regulation as it is a natural tendency of human beings to avoid negative feelings or emotions and approach desirable feelings (Gross, Reference Gross1998; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017). Integrating this argument with the viewpoint of self-regulatory resource depletion implies that challenging and demanding situations at work, such as abusive supervision, prompt employees' negative workplace moods because they consider such situations as threatening (see Tepper, Simon, & Park, Reference Tepper, Simon and Park2017 for a qualitative review). Additionally, in the presence of negative mood, scholars (e.g., Tice, Bratslavsky, & Baumeister, Reference Tice, Bratslavsky and Baumeister2001) noted that individuals are motivated to achieve short-term goals (e.g., search for instant pleasure by overeating to cope with negative mood) rather than long-term goals (e.g., maintain healthy eating habits), and they may engage in overeating to alter or relieve such moods (e.g., ‘eating one's feelings;’ Canetti, Bachar, & Berry, Reference Canetti, Bachar and Berry2002). In doing so, employees may bring temporary comfort and find an escape from a stressful situation (Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017), but are unable to alleviate the central problem, that is, abusive supervision. Based on the aforementioned arguments, we reason that overeating behavior can be a maladaptive act or strategy adopted by depleted employees to cope with their negative mood caused by abusive supervision, and propose the following hypotheses:
Hypothesis 1: The subordinate's perceptions of abusive supervision positively influence his/her overeating behavior.
Hypothesis 2: The subordinate's negative mood mediates the relationship between his/her perceptions of abusive supervision and overeating behavior.
We previously argued that abusive supervision induces negative workplace moods in employees and leads to overeating behavior because abused employees consume their limited self-regulatory resources to cope with a demanding situation. Given the prominence of self-regulatory resources (Barnes et al., Reference Barnes, Lucianetti, Bhave and Christian2015; Beal et al., Reference Beal, Weiss, Barros and MacDermid2005; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017) in constraining undesirable emotions and behavioral intentions, the association between abusive supervision and subordinate overeating behavior may become weaker if subordinates have more self-regulatory resources. In this regard, job-recovery literature suggests that an apparent way to restore resources is to rest or take a break from those activities that caused self-regulatory resources depletion (Baumeister et al., Reference Baumeister, Bratslavsky, Muraven and Tice1998; Demerouti et al., Reference Demerouti, Bakker, Sonnentag and Fullagar2012). Workplace breaks may include anything from sleep to doing stretching exercises, going out for lunch, checking one's emails, or having a cup of tea. We, therefore, draw on job-recovery literature to introduce subordinate recovery experiences at the workplace as a natural way to replenish subordinates' self-regulatory resources.
At the workplace, recovery experiences refer to the degree to which individuals perceive that the breaks they take help them to restore resources (Demerouti et al., Reference Demerouti, Bakker, Sonnentag and Fullagar2012; Sonnentag & Natter, Reference Sonnentag and Natter2004). Previous studies show that recovery experiences replenish self-regulatory resources and have a long-lasting effect on subordinates' emotions and behaviors at the workplace (e.g., see Demerouti et al., Reference Demerouti, Bakker, Sonnentag and Fullagar2012; Fritz & Sonnentag, Reference Fritz and Sonnentag2005, Reference Fritz and Sonnentag2006; Sonnentag, Reference Sonnentag2003; Trougakos et al., Reference Trougakos, Beal, Green and Weiss2008). While taking into account job-recovery experiences at the workplace, scholars have focused more on extensive breaks such as weekends (Fritz & Sonnentag, Reference Fritz and Sonnentag2005), vacations (Fritz & Sonnentag, Reference Fritz and Sonnentag2006), sabbaticals (Davidson et al., Reference Davidson, Eden, Westman, Cohen-Charash, Hammer, Kluger and Perrewé2010), and evenings (Sonnentag, Reference Sonnentag2003) rather than short breaks, e.g., relaxing, socializing, and napping that occurs during the workday (for exceptions see Demerouti et al., Reference Demerouti, Bakker, Sonnentag and Fullagar2012; Trougakos et al., Reference Trougakos, Beal, Green and Weiss2008). Under demanding and challenging situations, such as abusive supervision, subordinate recovery experiences can be a critical indicator of the replenishment or restoration of subordinates' self-regulatory capability. This is because taking breaks during the workday can restore the self-regulatory resources of employees (Barnes, Reference Barnes2012; Barnes, Wagner, & Ghumman, Reference Barnes, Wagner and Ghumman2012) and prepare them to effectively respond to potential upcoming demands of a workday (Barnes et al., Reference Barnes, Lucianetti, Bhave and Christian2015; Demerouti et al., Reference Demerouti, Bakker, Sonnentag and Fullagar2012; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017). Therefore, we propose and investigate the moderating role of recovery experiences of abused subordinates in attenuating the positive relationship of abusive supervision → subordinate negative mood → subordinate overeating behavior by helping them recover, replenish, or restore their depleted self-regulatory resources. We particularly focus on recovery experiences after short breaks on daily basis (i.e., relaxing, socializing, or napping), rather than extensive breaks (i.e., vacations, sabbaticals, or weekends), and their impact on replenishment or restoration of self-regulatory resources. The following hypotheses are thus proposed,
Hypothesis 3 (a): The subordinate's recovery experiences at work moderate the relationship between his/her perceptions of abusive supervision and negative mood, such that the relationship is weaker (stronger) when the subordinate's recovery experiences at work are higher (lower).
Hypothesis 3 (b): The subordinate's recovery experiences at work moderate the indirect relationship between his/her perceptions of abusive supervision and overeating behavior via his/her negative mood, such that the mediated relationship is weaker (stronger) when the subordinate's recovery experiences at work are higher (lower).
Method
As mentioned earlier, limited research (e.g., Barnes et al., Reference Barnes, Lucianetti, Bhave and Christian2015; Courtright et al., Reference Courtright, Gardner, Smith, McCormick and Colbert2016; Tariq & Ding, Reference Tariq and Ding2018) on abusive supervision has explored the potential within-person temporal variation approach of abusive supervision, whereby some managers frequently engage in abusive supervisory behavior while other managers do not frequently engage in such behavior. Moreover, negative mood, recovery experiences, and overeating behavior have often been measured through daily diary methods (e.g., see Derks & Bakker, Reference Derks and Bakker2014; Lanaj, Johnson, & Barnes, Reference Lanaj, Johnson and Barnes2014; Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017; Sonnentag, Binnewies, & Mojza, Reference Sonnentag, Binnewies and Mojza2008). Considering the aforementioned points, we measured our latent variables by using a daily diary approach where we asked participants to report abusive supervision, negative mood, overeating behavior, and recovery experiences for 10 consecutive working days. At Time 1 (i.e., at the end of the working day), we asked participants to rate abusive supervision, negative mood, and recovery experiences. At Time 2 (i.e., before going to bed on that working day), we asked participants to report their overeating behavior. Thus, we collected a single-source, multi-wave daily diary study to test our moderated mediation model.
We recruited research assistants and used a paper-and-pencil method to collect data from employees of a large service company headquartered in Islamabad, the capital city of Pakistan. The organizational scholars (e.g., see Khan et al., Reference Khan, Moss, Quratulain and Hameed2018) suggested that ‘abusive supervision most frequently occurs in high–power distance cultures… Pakistan, being high on power distance, appears to be a favorable context for studying abusive supervision’ (p. 2809). The research assistants contacted the Human Resource (HR) manager of the said company to invite employees to participate in our study. The research assistants then directly communicated the study's objectives to those employees who showed their consent and delivered each participant a package containing: (a) a letter explaining the instructions about the completion of daily surveys, (b) a general survey, (c) a daily booklet, and (d) return envelops. Following the instructions, participants first completed their respective general survey and were then asked to fill out their two daily questionnaires: (a) an afternoon questionnaire (to be completed after work when still being at work) and (b) a night questionnaire (to be completed before bedtime) for 10 consecutive working days.
The survey packages were delivered to 173 employees, out of which 159 were received back after completion. After detailed checking of the responses, the research assistants considered 115 responses valid for our final sample and excluded 44 responses due to the following reasons: (a) participants did not respond on all days (10 consecutive working days), (b) participants responded daily surveys at wrong times, (c) participants responded to Time 1 daily survey but didn't respond to Time 2 daily survey and vice versa. Therefore, we only considered those responses for our final sample in which the participants followed all the instructions.
Measures
Abusive supervision
To measure a subordinate's perceptions of abusive supervision on daily basis (i.e., within-person temporal variation rather than static approach), we adopted the 5-item scale developed by Johnson, Venus, Lanaj, Mao, and Chang (Reference Johnson, Venus, Lanaj, Mao and Chang2012) who specifically developed the daily level abusive supervision scale from the work of Tepper (Reference Tepper2000). We asked subordinates to rate ‘the frequency with which your supervisor engaged in each of the 5 behaviors today at work,’ using a 5-point Likert-type scale, ranging from 1 (not at all) to 5 (frequently). Sample items include, ‘Tells me I'm not capable,’ and ‘Makes negative comments about me to other.’
Negative mood
To measure a subordinate's negative mood at the workplace, we adopted the six-item scale from the work of Watson, Clark, and Tellegen (Reference Watson, Clark and Tellegen1988). We asked the participants to indicate the extent to which they agreed that each of the six items described their current mood that morning or afternoon at the workplace using a 5-point Likert-type scale, ranging from 1 (strongly disagree) to 5 (strongly agree). Sample items include, ‘Today I have felt distressed,’ and ‘Today I have felt upset.’
Overeating behavior
To measure a subordinate's daily overeating behavior, we adopted four-item scale from the work of Liu et al., (Reference Liu, Song, Koopmann, Wang, Chang and Shi2017) who developed daily level overeating behavior scale, which is appropriate for measuring overeating behavior among normal employee samples. We asked participants to rate their overeating behavior in the evening after work using a 5-point Likert-type scale, ranging from 1 (strongly disagree) to (strongly agree). These four items are ‘Today I ate too many junk foods after work,’ ‘Today I had too many unhealthy snacks after work,’ ‘Today I ate and drank excessively after work,’ and ‘Today I had too many late-night snacks before going to bed.’
Recovery experiences
To measure the daily recovery experiences of subordinates at the workplace, we followed the three-item scale from the work of Sonnentag (Reference Sonnentag2003) who specifically developed daily level recovery experiences at the workplace. We asked participants to respond to three items that linked feelings of recovery from short-term workplace activities or breaks using a 5-point Likert-type scale, ranging from 1 (not true at all) to 5 (very true). These three items are (a) ‘Because of these activities pursued today at workplace breaks, I feel recovered,’ (b) ‘Because of these activities pursued today at workplace breaks, I feel relaxed,’ and (c) ‘Because of these activities pursued today at workplace breaks, I was again full of energy.’
General survey
We used the general survey to assess the demographic information of our sample, for example, gender, age, education, tenure with supervisor, tenure with organization, and job experience. The participants were asked to record their responses to the general questionnaire before starting the daily diary sampling.
Results
Descriptive statistics
The intercorrelations, descriptive statistics (i.e., mean and standard deviations), and estimated reliabilities among the latent variables of our study are presented in Table 1, where we found preliminary support for our hypothesized relationships. We note that subordinates' perceptions of abusive supervision were positively correlated with their negative mood at the workplace (r = .54, p < .01) and overeating behavior (r = .46, p < .01). The subordinates' negative mood at the workplace was positively correlated with their overeating behavior (r = .34, p < .01). Finally, we found that the subordinates' recovery experiences were negatively correlated with their overeating behavior (r = −.16, p < .01).
Table 1. Intercorrelations, descriptive statistics, and estimated reliabilities among the latent variables
a Gender was coded 1 = male, 2 = female.
b Age was coded 1 = less than 25 years, 2 = 26–33 years, 3 = 34–41 years, 4 = 42–49 years, 5 = more than 49 years.
c Education was coded 1 = diploma, 2 = matriculation, 3 = undergraduate, 4 = graduate, 5 = postgraduate.
d Tenure with supervisor was coded 1 = less than 1 years, 2 = 1–2 years, 3 = 3–4 years, 4 = more than 4 years.
e Tenure with organization was coded 1 = less than 1 years, 2 = 1–2 years, 3 = 3–4 years, 4 = more than 4 years.
f Job experience was coded 1 = less than 1 years, 2 = 1–2 years, 3 = 3–4 years, 4 = more than 4 years.
Notes: N = 115 direct reports and 1150 daily ratings. Significance at: *p < .05; **p < .01; the figures in parentheses are α internal consistency reliabilities.
Confirmatory factor analysis
Confirmatory factor analysis using AMOS was conducted to confirm the factorial validity of the utilized measures. Byrne and Van de Vijver (Reference Byrne and Van de Vijver2010) and Schreiber, Nora, Stage, Barlow, and King (Reference Schreiber, Nora, Stage, Barlow and King2006) have recommended χ2/df, incremental fit index (IFI), comparative fit index (CFI), Tucker–Lewis index (TLI) and root mean square error of approximation (RMSEA) as the appropriate fit indices to assess the adequacy of a model. According to Hair, Black, Babin, and Anderson (Reference Hair, Black, Babin and Anderson2010), CFI, IFI, and TLI values above .90 and RMSEA scores below .08 represent the best model fit. The baseline four-factor model, i.e., abusive supervision, recovery experiences, negative moods, and overeating behavior, showed best fit to the data (CMIN/df = 2.67, CFI = .93, IFI = .93, TLI = .92, RMSEA = .05). The two alternative measurement models (Bentler & Bonett, Reference Bentler and Bonett1980) were compared and tested with the baseline model (see Table 2). In the first alternative model, recovery experiences and negative moods were combined into one factor, and the model was tested as a three-factor model. In the second alternative model, we loaded all constructs on a single factor, which showed a poor fit to the data. Thus, the baseline four-factor model was retained because of its best-fit indices over the two alternative models.
Table 2. Results of confirmatory factor analyses (CFA; model fit indices)
Notes: CFA, confirmatory factor analysis; incremental fit index; CFI, comparative fit index; RMSEA, root-mean square error of approximation; TLI, Tucker–Lewis index.
Analytical strategy
We followed the previous studies (e.g., Hongbo, Waqas, & Tariq, Reference Hongbo, Waqas and Tariq2019; Mawritz, Greenbaum, Butts, & Graham, Reference Mawritz, Greenbaum, Butts and Graham2017; Tariq, Weng, Garavan, Obaid, & Hassan, Reference Tariq, Weng, Garavan, Obaid and Hassan2020) to analyze our hypothesized moderated mediation model. Because of the within-person research design, hierarchical linear modeling (HLM; HLM version 6.08, Raudenbush, Reference Raudenbush2004) was used to test the hypothesized relationships. Initially, the intraclass correlations (ICC1) of dependent variables were calculated, and the results revealed significant between-individual variances in subordinate's negative moods (χ2 = 410.95; df = 108; p < .001; ICC = .20) and his/her overeating behavior (χ2 = 499.69; df = 108; p < .001; ICC = .25). Thus, it was appropriate to choose HLM as the analytical method for the present study. In addition, the main interest of this study was focused on within-person level (e.g., level 1) rather than between-person level (e.g., level 2), as suggested by prior research (Enders & Tofighi, Reference Enders and Tofighi2007); therefore, group-mean centering (e.g., individual-mean centering) was adopted for the independent variables (i.e., abusive supervision and subordinates recovery experiences) to rule out the potential between-person influence on dependent variables. Lastly, following the recommendation of Hayes (Reference Hayes2013) and Preacher, Rucker, and Hayes (Reference Preacher, Rucker and Hayes2007), we used moderated path analysis and bootstrapping to analyze the formal indirect effects (i.e., abusive supervision → negative mood → overeating behavior; mediation relationship) and conditional indirect effects of abusive supervision on overeating behavior via negative mood at the high/low values of recovery experiences.
Test of formal mediation
Table 3 demonstrates the findings of the HLR analyses. The findings show that abusive supervision was positively correlated with subordinates' overeating behavior (r = .27, SE = .03, t = 8.82, p < .001, Model 6). Thus, we found support for our Hypothesis 1, that is, subordinates' perceptions of abusive supervision were positively related to their overeating behavior. Moreover, we found that abusive supervision was positively correlated with subordinates' negative mood at the workplace (r = .67, SE = .02, t = 24.30, p < .001, Model 2), and the negative mood was positively correlated with their overeating behavior (r = .48, SE = .02, t = 17.90, p < .001, Model 6).
Table 3. Results of (HLM) hierarchical linear modeling analyses
Notes : N = 115 direct reports and 1150 daily ratings; JTS, tenure with supervisor; JTO, tenure with organization; AS, abusive supervision; RE, recovery experiences; AS × RE, two-way interaction term of abusive supervision and recovery experiences; *p < .05; **p < .01; and ***p < .001.
Finally, the results of direct (r = .17, p < .001, LLCI = .12, ULCI = .23), indirect (r = .32, p < .01, LLCI = .27, ULCI = .38), and total (r = .50, p < .001, LLCI = .44, ULCI = .55) effects provide support for Hypothesis 2, that is, subordinates' negative mood mediates the relationship of their perceptions of abusive supervision and overeating behavior (Table 4).
Table 4. Results of direct, indirect, and total effects of abusive supervision on overeating behavior via negative mood
Notes : N = 115 direct reports and 1150 daily ratings; LLCI, lower level of the 95% confidence interval; ULCI, upper level of 95% confidence interval; **p < .01; and ***p < .001.
Test of the moderated mediation model
Table 3 also demonstrates the findings of the moderated mediation model. We found that the interaction term of abusive supervision and subordinates' recovery experiences (AS × RE) was significantly negative (r = −.10, SE = .04, t = −6.46, p < .05, Model 3). Therefore, we found support for Hypothesis 3 (a), that is, subordinates' recovery experiences at work moderate the relationship between their perceptions of abusive supervision and negative mood, such that the relationship is weaker (stronger) when subordinates' recovery experiences at work are higher (lower).
We followed the approach of Edwards and Lambert (Reference Edwards and Lambert2007) to plot the conditional effects of abusive supervision on the subordinates' negative mood at the values of recovery experiences. We plotted the first-stage moderation, and Figure 2 demonstrates that subordinates' recovery experiences at work moderate the relationship between their perceptions of abusive supervision and negative mood, such that the relationship is weaker (r = .43, t = 3.92, p < .001) when subordinates' recovery experiences at work are higher and stronger (r = .58, t = 12.96, p < .001) when the recovery experiences are lower.
Fig. 2. An interaction of abusive supervision and subordinate recovery experiences on subordinate negative mood.
Finally, Table 5 demonstrates the conditional direct and indirect effects of abusive supervision on the subordinates' overeating behavior at the values (−1 SD, Mean, and +1 SD) of recovery experiences. We found that the effect of abusive supervision on the subordinates' overeating behavior via negative mood at the workplace at the low value (−1 SD) of recovery experiences was significantly positive and weaker (r = .37, SE = .03, LLCI = .31, ULCI = .43). The effect of abusive supervision on the subordinates' overeating behavior via negative mood at the workplace at the mean value of recovery experiences was significantly positive (r = .29, SE = .02, LLCI = .25, ULCI = .34). The effect of abusive supervision on the subordinates' overeating behavior via negative mood at the workplace at the high value (+1 SD) of recovery experiences was significantly positive and weaker (r = .22, SE = .02, LLCI = .17, ULCI = .27). Thus, we found support for our Hypothesis 3 (b), that is, subordinates' recovery experiences moderate the indirect relationship between their perceptions of abusive supervision and overeating behavior through negative mood, such that the mediated relationship is weaker (stronger) when subordinates' recovery experiences at work are higher (lower).
Table 5. Results of conditional effects of abusive supervision on overeating behavior via negative mood at values of subordinate recovery experiences
Notes : N = 115 direct reports and 1150 daily ratings; LLCI, lower level of the 95% confidence interval; ULCI, upper level of 95% confidence interval.
Discussion
We conducted a single-source, multi-wave daily diary study to explore the consequences of abusive supervision. From the perspective of self-regulatory theory, we proposed and found support for the direct relationship between abusive supervision and the subordinate's overeating behavior (Hypothesis 1). Moreover, we hypothesized a moderated mediation model and found that a subordinate's negative mood mediates the relationship between abusive supervision and overeating behavior (Hypothesis 2), and a subordinate's recovery experiences at work mitigate the effect of self-regulatory resource depletion and weaken the relationship between abusive supervision and overeating behavior through negative mood (Hypotheses 3a and b). Our findings contribute to the management and organizational literature in several ways.
First, the majority of research on abusive supervision has argued that some managers are abusive at the workplace while others are not (i.e., a static approach of abusive supervision, e.g., see Ahmad, Tariq, Weng, Shillamkwese, & Sohail, Reference Ahmad, Tariq, Weng, Shillamkwese and Sohail2019; Ahmed, Sumbal, Akhtar, & Tariq, Reference Ahmed, Sumbal, Akhtar and Tariq2021; Eissa & Lester, Reference Eissa and Lester2017; Mawritz et al., Reference Mawritz, Greenbaum, Butts and Graham2017; Yam, Fehr, Keng-Highberger, Klotz, & Reynolds, Reference Yam, Fehr, Keng-Highberger, Klotz and Reynolds2016). Researchers (e.g., see Barnes et al., Reference Barnes, Lucianetti, Bhave and Christian2015; Courtright et al., Reference Courtright, Gardner, Smith, McCormick and Colbert2016; Tariq & Ding, Reference Tariq and Ding2018) have argued that some managers frequently engage in abusive supervisory behavior, and in contrast, other managers do not (i.e., a within-person temporal variation approach of abusive supervision). We extend the latter line of inquiry by examining the daily relationships between abusive supervision, subordinates' negative mood, recovery experiences, and overeating behavior. Our findings add to this growing body of research by suggesting that organizational researchers should focus on a within-person temporal variation rather than a static approach to abusive supervision.
Second, while examining the consequences of abusive supervision, researchers have paid attention to exploring and investigating the work-related outcomes of abusive supervision (see Tepper, Simon, & Park, Reference Tepper, Simon and Park2017; Zhang & Liao, Reference Zhang and Liao2015 for reviews). Our study extended the outcome domain and explored the potential adverse effects of abusive supervision on employees' personal and family life beyond the boundaries of an organization. To do so, we focused on one common but overlooked cross-domain outcome of abusive supervision: subordinate's overeating behavior. By utilizing the self-regulatory resource impairment perspective, we found that facing abusive supervision at work depletes subordinates' resources, and they look for coping strategies, but when retaliation against supervisors is not an option, they turn to other maladaptive responses to cope. Being depleted from resources makes them prone to lose control over their behavioral intentions and thus they resort to overeating behaviors to cope with the stress from being abused.
Third, our study proposes the subordinate's negative mood at the workplace as an underlying mediating mechanism to explain the relationship between abusive supervision and subordinate's overeating behavior. Our findings extend the management literature by suggesting that an individual in the presence of a negative mood is motivated for mood regulation by searching for immediate pleasure (e.g., overeating).
Finally, we built our model further on the premise of job-recovery literature and proposed a moderated mediation model to elaborate the boundary conditions of the relationship between abusive supervision and subordinates' overeating behavior. The results of our study showed congruence with our propositions and we found that on-job recovery experiences mitigate the effect of abusive supervision on subordinates' overeating behavior via negative mood at the workplace.
Practical implications
Our research offers several important implications for managers and organizations alike. Firstly, in line with the detrimental consequences of abusive supervision on employees and organizations documented in research (see Tepper, Simon, and Park, Reference Tepper, Simon and Park2017), as well as its ineffectiveness in instrumentally increasing or boosting subordinates' job performance (e.g., Tariq & Weng, Reference Tariq and Weng2018; Walter et al., Reference Walter, Lam, Van der Vegt, Huang and Miao2015), our research also expounds the unfavorable outcomes of abusive supervision on subordinates. Organizations should therefore pay significant attention to limiting such behavior at the workplace by clearly communicating the aversive consequences instigated by abusive supervision to the managers. Moreover, by imparting a zero-tolerance policy against such destructive leadership styles, for example, giving punishments or demotions, organizations can make supervisors well aware of the consequences of being abusive toward subordinates and eventually be able to curb such behavior at the workplace.
Second, in contrast to the work-related consequences of abusive supervision, managers and organizations should understand the cross-domain negative outcomes of abusive supervision on employees' health and well-being. Our findings suggest that a failure in dealing with abusive supervisors might have possible long-term detrimental effects on the health and well-being of employees whereby, when abused, they may engage in overeating behaviors as a mood-altering strategy. As organizations strive toward creating healthy work environments for their employees, they should understand the effects stressors at work can have on employees' personal life. A possible strategy to deal with such issues would be to provide psychological support to employees and a safe space to report abusive supervision should they face it at work.
Finally, our findings suggest that it is worthwhile for organizations to provide sufficient recovery experiences at work to replenish the depleted self-regulatory resources of employees. Organizational scholars have reported that taking recovery breaks at work reduces fatigue and maintains the limited pool of employees' self-regulatory resources (Sonnentag & Binnewies, Reference Sonnentag and Binnewies2013; Sonnentag, Binnewies, & Mojza, Reference Sonnentag, Binnewies and Mojza2008; Trougakos & Hideg, Reference Trougakos, Hideg, Sonnentag, Perrewe and Ganster2009). To accomplish the goal of providing recovery experiences, organizations can provide napping pods, snack stations, and socialization breaks to employees. Such practices will ensure sufficient recovery experiences for employees and indirectly inform them that their well-being is also a workplace priority. Lastly, it is also plausible to argue that eating behavior at work (e.g., snacking at work) is itself a form of resource replenishment that could help abused employees to deal with their negative moods instigated by abusive supervision. Therefore, organization could provide more workplace eating avenues such as a snack cupboard for this purpose and fill it with healthy snacks, for example, fruits, nuts, and chopped vegetables.
Limitations and future directions
Despite the theoretical and managerial implications, our study has several limitations that need to be investigated and addressed by future studies. First, we used single-source to rate abusive supervision, negative mood, recovery experiences, and overeating behavior, which raises concern about the common method variance (CMV). Following the recommendations of Podsakoff, MacKenzie, and Podsakoff (Reference Podsakoff, MacKenzie and Podsakoff2012), we dealt with the CMV concern by conducting a multi-wave daily diary study (i.e., collected data two daily surveys for 10 consecutive working days), which is an effective method for reducing CMV (Lanaj, Johnson, & Barnes, Reference Lanaj, Johnson and Barnes2014). Despite the strength of a multi-wave daily diary study, we recommend future studies to conduct a multi-source study, for example, by incorporating supervisor and spouse ratings for a comprehensive understanding of abusive supervision and overeating behavior.
Second, we conducted the study in a large service company headquartered in Pakistan, which may limit the generalizability of our findings to Western countries as significant differences exist in eating behaviors between Western and Eastern countries (Liu et al., Reference Liu, Song, Koopmann, Wang, Chang and Shi2017). Moreover, Kim, Haines, Siega-Riz, and Popkin (Reference Kim, Haines, Siega-Riz and Popkin2003) found significant differences among specific aspects of eating behaviors in different countries, for example, Chinese people tend to be better at moderating food intakes as they consume more vegetables, bean products, and whole grains in comparison to the people of the USA. Therefore, while investigating abusive supervision-overeating behavior, we encourage researchers to conduct comparative studies between Eastern and Western samples.
Third, while attempting to study eating behaviors within the organizational context, our study only focused on investigating the relationship between abusive supervision and employees' overeating behavior and ignored the possible influence of employees' overeating behaviors on their short-term (e.g., job performance and job satisfaction) and long-term outcomes (e.g., health and well-being). The underdeveloped line of inquiry testing the influence of eating behaviors on employees' outcomes can be of interest to organizational researchers and we encourage them to explore how eating behaviors at home and work might affect employees' emotions, attitudes, and behaviors at work.
Fourth, we proposed and tested the moderated mediation model that implicitly suggests that an abused employees' overeating behavior is the maladaptive response to abusive supervision. Therefore, the causal direction of our moderated mediation model's findings could be an important limitation of our study. Our research design does not allow us to test such causal inferences. For example, it is plausible to argue that eating behavior at work (e.g., snacking at work) is itself a form of resource replenishment that could help abused employees to deal with their negative moods instigated by abusive supervision. Therefore, we call for further studies to use cross-lagged panel designs to test such causal inferences.
Conclusion
While there exists an abundance of research on the detrimental consequences of abusive supervision, we believe that our study stands in contrast as we investigate the adverse outcomes of abusive supervision beyond the workplace and in the personal life of subordinates. Employees may often be victims of abusive supervision, but retaliation against the abusive supervisor is rarely an option for them, and they may resort to maladaptive strategies that may, in turn, prove harmful rather than beneficial. We integrated the self-regulatory theory with job-recovery literature and elaborated that when depleted of resources after facing abuse from supervisors, subordinates may experience a negative mood. To remedy this situation, and rid themselves of the negative mood, they may look for immediate pleasure in the form of overeating. Our study further suggested that to alleviate the negative consequences of abusive supervision on subordinates' overeating behavior via negative mood, recovery experiences in the form of socializing, napping, and relaxing should be made available to subordinates. Such experiences will not only replenish their depleted resources but also refrain them from engaging in maladaptive behaviors after facing abusive supervision.
Acknowledgement
An earlier version of this manuscript was presented at the 81th Annual Meeting of the Academy of Management, A virtual experience.
Financial support
This research is supported by the Deanship of Scientific Research (DSR) at the King Fahd University of Petroleum and Minerals (KFUPM), Dhahran, KSA under Project SR201016, is gratefully acknowledged.
Hussain Tariq, PhD, is an assistant professor at KFUPM Business School, King Fahd University of Petroleum & Minerals (KFUPM), Dhahran, KSA. He earned a PhD in Organizational Behavior from the School of Management at the University of Science and Technology of China (USTC), Hefei, China. Before joining KFUPM, he held positions as an Assistant Professor of Management at NUST Business School (NBS), National University of Sciences and Technology (NUST), and as an Associate Member of the Centre for Cross-Cultural Management at the University of Wollongong, Australia. His research focuses on the root of destructive leader behaviors, structures, and practices that enhance teamwork and collective motivations in organizations. He has published his work in several reputed management journals such as International Journal of Hospitality Management, Applied Psychology, Journal of Knowledge Management, Personnel Review, Employee Relations, Journal of Service Theory and Practice, and Business Ethics: A European Review amongst others.
Asfia Obaid, PhD, is an associate professor in management and Head of Department at NUST Business School, National University of Sciences and Technology, Islamabad, Pakistan. She received her PhD in Business Administration from Manchester Business School, University of Manchester, UK. She has made contribution to knowledge through various avenues that include delivering training sessions, consulting with organizations on industry-specific problems, publishing of research articles, conducting research grant projects. Her core areas of interest include strategic HR, compensation, and performance management.
Muhammad Burhan, PhD, is a senior lecturer in HRM at the Department of Management, Huddersfield Business School, University of Huddersfield, UK. Before joining the University of Huddersfield, he held position as a lecturer in HRM at Coventry University, Business School, UK. He received his postgraduate degree in strategic management from the University of Bradford, UK and a PhD in strategic human resource management from the University of Huddersfield, UK. Muhammad has previously worked at Middlesex University, London and the University of Huddersfield. His research interests center on contextual HRM, leadership, people analytics, and sustainability, primarily in SMEs.
Muhammad Subhan has done Master in Human Resource Management from NUST Business School, National University of Sciences & Technology, Islamabad, Pakistan. His research focuses on leadership in particular an abusive supervision and its impact on employees’ behavior and attitude within the organizations and also beyond organizational boundaries.
Sumbal Babar is a PhD candidate in business administration at the NUST Business School, National University of Sciences & Technology, Islamabad, Pakistan. Her research focuses on the HR process approach, in particular the impact of employees' perceptions, understanding, and attributions of HR on their behaviors.
