Abstract
The present study extends the body of research on turnover by examining the impact of abusive supervision on turnover in the context of entry-level hospitality employees, drawing on conservation of resources theory as an overarching theoretical lens. In addition, this research examined moderators in the abusive supervision–turnover relationship, specifically constituent attachment and employee age. With a sample of 980 restaurant front-of-house employees, data on abusive supervision, constituent attachment, and age were used to predict turnover over a 6-month period via logistic regression. Abusive supervision increased turnover among the sample overall. However, the effect was stronger for younger employees. In addition, constituent attachment increased the likelihood of turnover for younger employees who experienced greater abusive supervision. The present study serves to validate the adverse impact of abusive supervision on turnover along with factors that may strengthen or attenuate its impact.
The hospitality industry is currently the fifth largest employer in the United States (U.S. Bureau of Labor Statistics, 2015). The work in hotels and restaurants is fast-paced, highly social, and is often perceived as fun (Tews et al., 2013). Despite such benefits, turnover rates average approximately 75% annually (U.S. Bureau of Labor Statistics, 2019). The majority of the hospitality labor force is entry-level employees who voluntarily leave such jobs at high rates, creating a major management challenge. For example, such high turnover results in excessive replacement and training costs, reduced service quality, productivity losses, lower morale, and teamwork disruptions (Allen et al., 2010). On one hand, high turnover is somewhat inevitable because many jobs fall in the secondary labor market, which is characterized by part-time hours, low pay, limited benefits, and routine work (Bewley, 1995; Hudson, 2007). Many employees in the secondary labor market do not intend to remain in such work for a long-term career, but rather use such employment to gain experience and for supplemental income.
Abusive supervision is another factor that may be relevant in the turnover equation for entry-level hospitality employees. Managers have much greater control over how they interact with employees than they do over many other human resource management practices. Tepper (2000) defines abusive supervision as “the extent to which supervisors engage in the sustained display of hostile verbal and nonverbal behaviors” (p. 178). Such behaviors include belittling, breaking promises, displaying anger, making negative or rude comments, and giving an employee the silent treatment, among others (Tepper, 2000). Physical aggression, while certainly abusive, is not subsumed within the abusive supervision construct. Signaling its pervasiveness in the hospitality industry, abusive supervision is often portrayed among restaurant chefs (Bloisi & Hoel, 2008) and is a recurring theme in Anthony Bourdain’s (2007) Kitchen Confidential.
Perhaps surprisingly, only limited research has focused on abusive supervision and turnover. Tepper’s (2000) seminal article demonstrated the abusive supervision–turnover relationship, but additional studies have substantiated the link between abusive supervision and turnover intentions (e.g., Harvey et al., 2007; Mathieu & Babiak, 2016; Palanski et al., 2014; Pradhan & Jena, 2017; Tepper et al., 2009). More recently, Allen et al. (2016) demonstrated that abusive supervision adversely affects turnover more strongly when job embeddedness is low. Allen et al.’s findings suggest that the abusive supervision–turnover relationship may not be straightforward, signaling that attention should be paid to factors that amplify or attenuate the impact of abusive supervision on whether employees remain with or leave the organization.
To help advance turnover research and reduce turnover among entry-level hospitality employees, the present study seeks to further validate the relationship between abusive supervision and voluntary turnover and examine potential moderators of this relationship. Constituent attachment, the degree to which employees feel attached and have established bonds with their coworkers, is one moderator to be examined (Maertz & Campion, 2004). Previous research provides strong evidence that constituent attachment is significantly related to retention (e.g., Ellingson et al., 2016; Tews et al., 2014). Employee age is a second moderator to be examined. We argue that abusive supervision will more strongly affect turnover for younger employees. Because younger individuals may be less dependent on income and more sensitive to abusive supervision, they could be more apt to turnover when confronted with a supervisor’s maltreatment. Ultimately, we strive to increase our knowledge of turnover antecedents among entry-level hospitality employees, add to the nomological network of abusive supervision, and provide data-driven strategies to promote retention.
The Present Study
A large body of research has demonstrated that abusive supervision has adverse consequences on individuals and organizations. Two recent pieces summarize the extant body of research to date: Mackey et al.’s (2017) meta-analysis and Tepper et al.’s (2017) narrative review. Mackey et al.’s (2017) meta-analysis demonstrated low-to-moderate associations of abusive supervision with affective commitment, counterproductive work behavior, depression, emotional exhaustion, job satisfaction, job tension, perceived organizational support, job performance, and work–family conflict. Similarly, Tepper et al.’s (2017) review concluded that the extant body of research has demonstrated that abusive supervision is associated with a range of adverse outcomes, including decreased job attitudes, decreased psychological health, and higher incidences of counterproductive work behavior and organizational withdrawal.
In extending the extant body of research in this area, we use conservation of resources (COR) theory as an overarching framework to explain the abusive supervision–turnover relationship. In previous research, COR has been used to study both abusive supervision (Whitman et al., 2014) and turnover (Lapointe et al., 2011; Wright & Cropanzano, 1998), but not in the same study. The underlying assumptions of COR are that individuals are motivated to acquire and protect valued resources, protect and enhance the self, and create environments that afford them success and pleasure (Hobfoll, 1989; Whitman et al., 2014). Actual or perceived resource loss may lead to increased psychological stress and a defensive position to conserve remaining resources. As abusive supervision represents a loss of supervisor support and the risk of loss of other resources, employees may leave to conserve remaining resources or to obtain better resources elsewhere. At the same time, constituent attachment and employee age could curb such turnover, because constituent attachment and age are potentially valuable resources. When employees are higher in constituent attachment, they could have access to greater social support to combat adverse abusive supervision effects. Moreover, older employees may be less sensitive to abusive supervision and possess better coping skills. Figure 1 presents a graphical presentation of our research model and corresponding hypotheses.
The Research Model
As seen through the lens of COR, abusive supervision could represent a number of actual and perceived resource losses that might lead to greater turnover. Arguably, the greatest resource loss is that of needed social support. Supervisors represent a major source of social support in the workplace that employees may draw on for emotional and instrumental support (Ng & Sorensen, 2008). When confronted with abusive supervision, employees may be less apt to seek out support from their supervisors as doing so may invite criticism and a potential loss of self-esteem. In the face of abusive supervision, employees could also perceive the loss of future valued resources, such as favorable shifts, favorable performance reviews, advancement opportunities, and pay raises. As employees who are the targets of abusive supervision likely perceive that they are devalued, employees may perceive a loss of expected gains with respect to the above aspects of employment. Furthermore, abusive supervision may adversely affect one’s reputation in the workplace (Harris et al., 2011). Because one aspect of abusive supervision involves belittling employees in front of others, peers may view employees less favorably, or employees could perceive a loss of status in the eyes of others. Employees may thus find it more difficult to develop cohesive workplace relationships. More holistically, abusive supervision could adversely affect an employee’s organization-based self-esteem, one’s beliefs regarding his or her capability, significance, and worth as a member within the organization (Pierce & Gardner, 2004). On the whole, individuals seek to achieve and maintain a healthy self-esteem and take measures to protect their self-esteem when their identity is threatened (Ito & Brotheridge, 2003). Employees could be more likely to turnover when subject to abusive supervision to replenish resource loss through employment elsewhere or through nonwork alternatives.
Given the negative consequences of abusive supervision and the host of associated costs, it is worth examining the factors that might mitigate the detrimental impact of such maltreatment. Constituent attachment could be instrumental in this regard. Constituent attachment results when individuals feel supported by their coworkers and experience high levels of workgroup cohesion (Ellingson et al., 2016). Such attachment is thought to be an important driver to enmesh individuals in organizations, thereby promoting retention (Maertz & Campion, 2004), and it has been found to be significant in reducing turnover among entry-level hospitality employees (Ellingson et al., 2016). Constituent attachment is valued because employees frequently work alongside one another in completing job responsibilities that require frequent interaction and teamwork. Furthermore, constituent attachment is important because high-quality coworker relationships may be perceived as less substitutable than other terms and conditions of employment. Whereas entry-level hospitality jobs may be easily substitutable with respect to compensation and benefits, high-quality coworker relationships may be less so (Steel, 2004).
Fundamentally, constituent attachment could attenuate the abusive supervision–turnover relationship because employees with greater constituent attachment likely have greater coworker social support resources on which to draw to combat the adverse effects of abusive supervision. When discussing COR, Hobfoll (1989) contended that workplace social support was particularly effective in combating resource loss when social interaction was an integral and nonstressful aspect of the job. Drawing on the cross-domain buffering hypothesis (Duffy et al., 2002), social support from one domain can buffer the negative consequences of adverse social relationships from another domain. On one front, supportive coworkers can be a sounding board and resource to help individuals devise strategies to cope with abusive supervision. Ito and Brotheridge (2003) identified seeking advice from others as one of the most effective responses to resource loss. Furthermore, supportive coworkers can help employees recuperate from a stressful situation and reduce emotional exhaustion and burnout (Sonnentag, 2003). After a negative encounter with a supervisor, supportive coworker interactions could help enhance an employee’s mood and sense of wellbeing. The resource gain associated with greater constituent attachment may thus compensate for the resource loss associated with abusive supervision. Accordingly, it is hypothesized:
A further issue to be explored is whether the impact of abusive supervision on turnover is contingent upon an employee’s age. Specifically, we argue that abusive supervision will more strongly relate to turnover for younger employees relative to older individuals. In general, younger employees are more likely to turnover than older ones (Posthuma & Campion, 2009). Younger employees are often still involved with formal education, and most do not intend to continue in entry-level work over the long term (Hudson, 2007). They are often employed in entry-level work to gain some initial work experience, but their career aspirations lie elsewhere (Newman, 2007; Siebert & Zubanov, 2009). Furthermore, younger individuals are often less dependent on work income due to fewer financial obligations and continued financial support from parents (Kirkpatrick Johnson, 2013). Newman (2007) colloquially refers to younger employees in entry-level employment in the hospitality industry as way-station workers: individuals employed in such work as a stopping-off point, making time until better things happen, and earning money until the next stage of their life begins. Given this reality, younger employees are less committed to their places of employment and may be more susceptible to negative forces in the workplace.
Abusive supervision could more strongly affect turnover for younger employees because they have fewer personal resources than older individuals that could be useful when responding to abusive supervision. From the perspective of personality, younger individuals tend to be lower in emotional stability than their older counterparts (Twenge, 2000; Williams et al., 2006), and lower levels of emotional stability reduce one’s ability to respond well to potentially stressful situations. Emotional stability, one of the Big Five personality dimensions, relates to the extent to which individuals are generally calm, relaxed, and free from worry (Costa & McCrae 1992). Those low in emotional stability are more likely to interpret potentially stressful situations as worse than they are and feel more negatively about such situations. Indeed, Zhang and Bednall (2016) found that younger individuals were more likely to perceive supervision as abusive, and thus younger employees may be more apt to “notice” abusive supervision.
In a related respect, younger individuals are more conflict averse (Alsop, 2008; Arnett, 2004), which can limit one’s effectiveness in dealing with challenging situations. When individuals are more conflict averse, they may be more likely to use passive strategies in responding to abusive supervision, such as avoiding feedback, avoiding the supervisor, and turnover (Whitman et al., 2014). Older individuals who are less conflict averse are more likely to use active strategies, such as discussing problems with supervisors, working harder, and seeking help and advice (Ito & Brotheridge, 2003). Older employees may also be more resistant to abusive supervision because they have had more exposure to it, given their greater age and experience. Through their greater experience with employment and life in general, older individuals likely have developed less sensitivity to maltreatment such as abusive supervision and have developed better strategies for coping. In this respect, older individuals could be less deterred by abusive supervision and less likely to leave.
Extending the preceding arguments, we contend that constituent attachment will be more influential in attenuating the abusive supervision–turnover relationship for younger employees than older ones. Among younger employees, constituent attachment is likely more valued, and younger employees may need to draw on social support more to combat the adverse effects of abusive supervision. One of the primary developmental tasks for younger individuals as they transition into full adulthood is developing a social network outside of their immediate families (Erikson, 1968). Interpersonal relationships through work provide an outlet for social contact, help individuals develop a sense of identity as an independent actor, and offer advice about work and career. In a related respect, friendships through work are likely more important among younger individuals because they have fewer family obligations and commitments than older adults. In support of the proposition that constituent attachment matters differently based on age, previous research found that it was a stronger predictor of turnover for younger individuals (Ellingson et al., 2016). Because they are less experienced in managing conflict, younger individuals may need to draw on social support more heavily to respond to abusive supervision. In contrast, older individuals may value attachment through work less, be able to draw more readily on other sources of social support, and be better equipped to manage conflict and respond actively to abusive supervision. Based on these arguments, we hypothesize:
Method
Overview
The study participants were 980 waitstaff in a restaurant setting. These individuals were from a restaurant group with locations in the Northeast and Midwest of the United States. All 2,056 waitstaff employed in the target restaurants were provided the opportunity to participate in the research. One thousand and twenty-nine employees (approximately half of those invited to participate) provided complete data on abusive supervision and constituent attachment through a paper-and-pencil survey. Turnover data were obtained 6 months after the survey through a report obtained from the organization’s corporate office. Forty-nine employees were excluded from the research as their decision to leave the company during the study period was not voluntary (i.e., they were fired). The survey responses and turnover data were matched through a confidential number for each employee. All waitstaff were part-time, typically working less than 40 hours per week. The final set of participants was 33% male, 23% non-White, and 27.25 years of age. These data were from a larger data set also used in Tews et al. (2019).
Measures
Abusive supervision. Harris et al.’s (2011) six-item version of Tepper’s (2000) 15-item scale was employed herein. Among the items were: My supervisors give me the silent treatment and My supervisors are rude to me. A 5-point scale was used to assess each item with anchors ranging from 1 = strongly disagree to 5 = strongly agree (α = .84).
Constituent attachment. Four items were used to assess this construct (Tews et al., 2014). A sample item included as follows: I want to continue working with my coworkers here. A 5-point scale was used to assess each item with anchors ranging from 1 = strongly disagree to 5 = strongly agree (α = .85).
Age. It was measured in years based on the difference between the study date and the individual’s date of birth.
Voluntary turnover. Six months was used as the turnover window given the high turnover rates among entry-level hospitality employees. Three hundred and fifteen of the study participants were leavers, and 665 were stayers, yielding a voluntary turnover rate of 32% for the study period. The baseline group (0) was those who remained with the organization, and the contrast group (1) was those who left.
Prestudy tenure. The length of time an individual had been employed with the organization at the beginning of the study, was included as a control variable in the analyses. This variable was included as it has been found to decrease the likelihood of turnover (e.g., Tews et al., 2013; Tews et al., 2014), whereby individuals who have been employed with an organization longer are less likely to leave. In the present study, prestudy tenure was operationalized in years.
Discriminant Validity
Confirmatory factor analyses were conducted using Mplus 8 (Muthén & Muthén, 2017) to further validate the discriminant validity of the constituent attachment and abusive supervision scale items. Namely, a two-factor model was tested whereby each of the items was loaded onto its respective construct. Although the model possessed a statistically significant chi-square, χ2(34, n = 980) = 181.55, p < .01, the individual fit indices provided support for the model, comparative fit index = .96; Tucker–Lewis index = .95; root mean square error of approximation = .07, 90% confidence interval [.06, .08]; and standardized root mean square residual = .04. These results provide adequate support for the discriminant validity of the scales.
Analytic Strategy
Binary logistic regression was used to assess the impact of the independent variables on turnover. The independent variables were entered into the regression in three blocks, resulting in three models. Model 1 included the main effects for prestudy tenure, abusive supervision, age, and constituent attachment. Model 2 included the addition of the two-way interaction terms: abusive supervision × constituent attachment, abusive supervision × age, and constituent attachment × age. Finally, Model 3 added the abusive supervision × constituent attachment × age three-way interaction term. Prior to creating the interaction terms, the abusive supervision, age, and constituent attachment variables were mean centered to limit multicollinearity, which might otherwise bias the coefficients (Aiken & West, 1991). Since the proposed relationships were directional, tests of significance were one-tailed (McClave & Sincich, 2017).
Results
Table 1 presents the descriptive statistics and correlations among the study variables, prior to any transformations. Table 2 presents results of logistic regression analyses predicting turnover. The significant χ2 for each model indicates that at least one predictor in each model was nonzero. The Nagelkerke R2 for Models 1, 2, and 3 are .17, .17, and .18, respectively. The Δχ2 was nonsignificant from Model 1 to Model 2 (2.34, p > .05). However, the Δχ2 was significant from Model 2 to Model 3 (4.32, p < .05). Thus, the two-way interactions did not increase the variance explained, but the three-way interaction did. The main table entries in Table 2 are logistic coefficients, standard errors, and exponentiated logistic coefficients, exp(b). For exp(b), values exceeding 1 indicate an increased effect on turnover, and values below 1 indicate a decreased effect.
Descriptive Statistics and Correlations Among Variables
Note: n = 980. Correlations with turnover are point-biserial. Significance levels reflect one-tailed tests.
p < .05. **p < .01.
Logistic Regression Predicting Voluntary Turnover
Note: n = 980. For the exponentiated regression coefficients, Exp(b), values exceeding 1 indicate an increase in voluntary turnover, while values below 1 indicate a decrease. Significance levels reflect one-tailed tests.
p < .05. **p < .01.
Hypothesis 10, which indicated that abusive supervision would be unrelated to employee turnover, can be rejected in favor of Hypothesis 1a, which indicated a positive effect. The logistic coefficient of .22 for abusive supervision in Model 1 was significant at the .05 level, along with the coefficient of .20 in Model 2. Holding constant the other variables in the model, a one-unit increase in abusive supervision increased the odds of turnover on average by a multiplicative factor of 1.25 in Model 1 and 1.22 in Model 2.
Hypothesis 20 cannot be rejected in favor of Hypothesis 2a, which indicated that constituent attachment would attenuate the abusive supervision–turnover relationship. The coefficients for the abusive supervision × constituent attachment interaction term were nonsignificant in both Model 2 and Model 3.
Hypothesis 30, which indicated that the relationship between abusive supervision and employee turnover would be the same for younger employees and older employees, can partially be rejected in favor of Hypothesis 3a, which indicated that the positive relationship between abusive supervision and employee turnover would be stronger for younger employees. The significant logistic coefficient of −.04 (p < .05) for the abusive supervision-age interaction term in Model 3 suggests that abusive supervision is more problematic for younger employees. Figure 2 depicts this interaction. The relationship between abusive supervision and turnover is plotted separately for employees at the 20th and 80th age percentiles. Among younger employees, the probability of turnover increases from low to high levels of abusive supervision. However, among older employees, the probability of turnover does not increase from low to high abusive supervision. Moreover, when abusive supervision is high, the probability of turnover for younger employees is greater than for older employees.
Interaction Between Abusive Supervision and Age in Predicting Turnover
Hypothesis 40 cannot be rejected in favor of Hypothesis 4a, which suggested a three-way interaction among abusive supervision, constituent attachment, and age. This alternative hypothesis proposed that among younger employees, constituent attachment would more strongly reduce the impact of abusive supervision on turnover. The three-way abusive supervision × constituent attachment × age interaction term of −.05 in Model 3 was significant at the .05 level, but the sign was not consistent with the hypothesized direction. Figure 3 depicts this interaction. Like Figure 2, the effect of abusive supervision on turnover can be seen separately for younger and older employees, but in separate panels for low and high constituent attachment at the 20th and 80th percentiles. When constituent attachment is low, the relationship between abusive supervision and turnover is low and is similar for younger and older employees. However, when constituent attachment is high, the relationship between abusive supervision and turnover is greater. Furthermore, when constituent attachment is high, the rate of turnover increases for younger employees and decreases for older employees as abusive supervision increases.
Interaction of Abusive Supervision, Constituent Attachment, and Age in Predicting Turnover
Discussion
For hospitality enterprises, high rates of turnover are a perennial problem, placing a large burden on managers and other employees and compromising the guest experience. While such turnover is often attributed to a number of poor HR practices (Ellingson et al., 2016), the present study highlights that social influences, such as abusive supervision, should be considered in the turnover equation. By examining turnover as an outcome of abusive supervision, along with moderators of the abusive supervision–turnover relationship, the present study contributes to the general management literature overall in addition to turnover research that is specific to entry-level hospitality employees.
Drawing on previous research (Harris et al., 2011; Whitman et al., 2014), the present study used COR theory as an overarching framework to support the hypothesized relationship between abusive supervision and turnover, along with the proposed moderating relationships. We had argued that abusive supervision would lead to a loss of perceived or actual supervisor support, as well as reduced organization-based self-esteem, loss of status and reputation, and lowered expectations of future favorable workplace conditions. Such losses of current and anticipated resources would then lead to greater turnover. As hypothesized, abusive supervision did increase the likelihood of turnover.
In addition to demonstrating that abusive supervision significantly increased turnover, there was evidence that abusive supervision was more strongly related to turnover for younger employees relative to older ones. It was argued beforehand that abusive supervision might be particularly problematic for younger individuals since their employment is more transient and tentative. Younger individuals have fewer constraints binding them to organizations, and they may be more likely to leave an organization when confronted with maltreatment. Furthermore, younger individuals tend to be lower in emotional stability (Twenge, 2000; Williams et al., 2006) and more conflict averse (Alsop, 2008; Arnett, 2004). They may thus have greater challenges combating the potential resource losses associated with abusive supervision.
One of the unexpected findings was the nature of the interaction between abusive supervision and constituent attachment for younger employees. Counter to expectations, the results indicated that younger individuals higher in constituent attachment were more likely to turnover as abusive supervision increased. Greater constituent attachment could lead to greater turnover in two respects. Because those with greater constituent attachment likely receive more coworker support, their coworkers might encourage them to leave an abusive supervisor and provide the social support necessary to cope with the risk and uncertainty of leaving one’s job. In addition, those with greater constituent attachment could be leaving since their coworkers, to whom they are closely attached, are also leaving. Research finds employees frequently make decisions whether to stay or leave based on coworkers’ actions (Rubenstein et al., 2018). In a positive light, constituent attachment had a main effect that reduced levels of turnover, consistent with previous research (Ellingson et al., 2016; Tews et al., 2014). Unlike previous research (Ellingson et al., 2016), the results from this study suggest that constituent attachment may be beneficial for promoting retention of employees irrespective of age, even though it did not attenuate the effects of abusive supervision on turnover.
Given that abusive supervision was found to be positively related to turnover, especially among younger employees, abusive supervision should be minimized and appropriately remedied in the workplace. Fundamentally, organizations must first recognize that abusive supervision has a detrimental impact. Then organizations should minimize the drivers of abusive supervision suggested in previous research, such as organizational norms that support hostility (Mawritz et al., 2014), role models of abusive supervision (Liu et al., 2012), and supervisor role overload and frustration (Eissa & Lester, 2017). Furthermore, supervisor training could address how to manage potential conflicts with employees in a constructive, nonaggressive, and supportive manner. If employees perceive abusive supervision, tools to help them communicate their concerns would also be valuable.
The generalizability of this research should be considered in light of its limitations, as with all research. One limitation is that the sample was drawn from a single U.S. company and was primarily Caucasian. Additional research would thus be valuable to further examine the impact of abusive supervision on turnover with other samples of employees in the United States and in other cultures to extend the present study and the work of Allen et al. (2016). While common in turnover research, an additional limitation of this study is that the turnover criterion used in this research treated all voluntary turnover as equal. Individuals may voluntarily leave their organizations for reasons that are more controllable (e.g., leaving for a similar job with similar terms and conditions of employment) or for reasons that are less under the organization’s control (e.g., relocation or attending school). Abusive supervision may likely have a greater impact on turnover that is more controllable, but it would also be valuable to assess its magnitude on factors that are believed to be less controllable.
In addition to the opportunities articulated above, several avenues for future research are worth pursuing. One opportunity is examining the impact of abusive supervision on turnover relative to supervisor support, as research on both abusive supervision and supervisor support on turnover is limited. In addition to analyzing direct effects, it also could be valuable to examine how supervisor support and abusive supervision interact in influencing turnover. Furthermore, it could be worthwhile to examine mediators in the abusive supervision–turnover relationship, such as employee engagement (Saks, 2006). In addition, future research should examine the potential moderating effect of gender on the abusive supervision–turnover relationship. It is possible that abusive supervision will have a greater effect on turnover for females than males. This relationship is credible given that Burton and Hoobler (2006) found that abusive supervision had a more pronounced negative effect on females than males. A final suggestion is analyzing the relationship between abusive supervision and turnover among employees in jobs in the primary labor market that are characterized by full-time hours, higher compensation, and better benefits, where turnover rates are generally lower (Hudson, 2007). Such research would test whether abusive supervision is as problematic when the terms and conditions are more favorable.
Conclusion
Research on abusive supervision appears to have grown at an exponential rate since Tepper’s seminal work on abusive supervision in 2000. Despite the relatively large body of research on abusive supervision in the workplace, research on abusive supervision and turnover has been quite limited relative to research on turnover intentions. The present study has helped fill this gap in the literature by examining the abusive supervision–turnover relationship along with moderators in the hospitality industry, where turnover rates are especially high and problematic. We hope that this study has increased our knowledge of turnover and abusive supervision in a substantive way that will stimulate future research in this area.
