Psychosocial work environment, work engagement, and employee commitment: A moderated, mediation model

This paper contributes to the growing body of literature on the wellbeing of hospitality employees from a perspective of strategic human resource management. The role of high performance work systems (HPWS) in enhancing the affective commitment of hospitality employees is examined. The study found work engagement to mediate the relationships between HPWS, perceived organizational support, and affective commitment. Workplace bullying, a highly prevalent phenomenon in the hospitality sector, was found to mediate the relationship between HPWS and affective commitment, while psychosocial safety climate moderated this mediating impact. We will suggest the implications for managing psychosocial work hazards in hospitality organi-


Introduction
In the new global economy, hospitality is among the largest growing industries with a high labor-intensive and service-based operating environment in which attitudes and wellbeing of hospitality workers are a major concern in the organization's successful life (Hsu et al., 2019). Employees in the US hospitality sector experienced excessive workloads and long working hours, work pressure, time constraints, highly intense work conditions, inadequate training and protections for their health and safety, and are exposed to a high prevalence of workplace bullying (Ariza-Montes et al., 2017;Bentley et al., 2012;Page et al., 2018). Approximately one-third of the 90,000 complaints related to workplace bullying and harassment incidents that were reported to the US Equal Employment Opportunity Commission were from hospitality employees in 2015 (Golshan, 2017), resulting in adverse outcomes such as high turnover rates and low profitability (Jung and Yoon, 2018;Ram, 2018).
Workplace bullying, a feature of poor work environments, has been found to have negative health consequences (Ariza-Montes et al., 2017;Ram, 2018). It is critical for hospitality organizations to consider ways of effectively managing their human resources (HR) in order to attract and retain employees in the hospitality sector (Baum et al., 2016). Therefore, it is important to implement a strategic approach to HR management in order to improve the wellbeing of hospitality workers in the US (Murphy et al., 2018;Page et al., 2018). In this paper, we focus on the adoption of high-performance work systems (HPWS) to operationalize a strategic approach to human resource management (HRM), as it has been found to be effective in eliminating workplace bullying (e.g. Samnani and Singh, 2014).
Following social exchange theory (SET) (Cropanzano et al., 2017;Parzefall and Salin, 2010), we will argue that HPWS plays a key role, together with psychosocial safety climate (PSC) and perceived organizational support (POS), in preventing employee exposure to workplace bullying in the hospitality sector. This paper contributes to advancing knowledge in hospitality HRM in several ways. Firstly, we will provide much needed empirical evidence on the impact of HPWS on employee wellbeing (see the call by Guest, 2017) in the hospitality context. Our review showed few empirical studies examining the impact of HPWS on work outcomes and psychological wellbeing of hospitality employees (exception include Dhar, 2015;Karatepe, 2013a;Murphy et al., 2018). Secondly, the prevalence of workplace bullying has been associated with the dark side of organizational life (Bentley et al., 2012;Page et al., 2018). We will focus on the underlying mediating mechanism by examining the role of work engagement in strengthening the HPWS-affective commitment relationship. Finally, we will focus on the moderating effect of PSC in mitigating workplace bullying. Specifically, we argue that low levels of PSC will increase workplace bullying, while higher levels of PSC will enhance the protective influence of HPWS on employee attitudes and behaviors.

Work environment factors and workplace bullying
There are several examples of workplace bullying behaviors in the hospitality management literature. These include incivility (Han et al., 2016), verbal abuse (Kitterlin et al., 2016), physical abuse (Lyu et al., 2016), hostility (Mathisen et al., 2008), kitchen banter (Giousmpasoglou et al., 2018), and sexual harassment (Ineson et al., 2013). In this study, workplace bullying is defined as "a situation where one or several individuals perceive themselves to be on the receiving end of negative actions from one or more persons persistently over a period of time, in a situation where the targets have difficulty defending themselves against these actions" (Lutgen-Sandvik et al., 2007, 847). Workplace bullying includes person-related, work-related, and physically intimidating bullying behaviors (Einarsen et al., 2009).
Workplace bullying is a highly damaging psychosocial risk. It is found in hospitality organizations, which have a prevalence rate of between 11-20 % (Nielsen et al., 2010). Approximately 5.6 % of European hospitality employees (Ariza-Montes et al., 2017), 14 % of Canadian hospitality workers (Nieweler, 2015), and 30 % of US hospitality workers (Golshan, 2017) have experienced these behaviors. There is strong empirical support for the negative consequences of workplace bullying, such as poor psychological health and depression, stress, anxiety, and low levels of emotional wellbeing . Workplace bullying has also been found to impact negatively on job satisfaction, low affective commitment, and work performance in the general population (Lutgen-Sandvik et al., 2007;Nielsen and Einarsen, 2012). There is an increasing interest on its effect on the work attitudes and wellbeing of hospitality employees (Ariza-Montes et al., 2017;Hsu et al., 2019;Jung and Yoon, 2018).
HPWS includes a wide range of strategically oriented HR practises including comprehensive recruitment, rigorous selection, extensive training, performance management, competitive compensation, incentives and rewards, promotion, job security, flexible job design, employee involvement, and information sharing (Jiang et al., 2012). A large body of literature has shown that HPWS can lead to human capital and employee motivation along with financial and operational performance of organizations (see the meta-analytical review by Jiang et al., 2012). However, Woodrow and Guest (2014) argued that research into the effects of HR practises has neglected the role of HPWS in creating a safe environment to promote or protect employee wellbeing and minimize psychosocial risks and hazards like workplace bullying. This is another aim of the current study.
Social exchange can be developed by HPWS as these bundles of HR practises can be used to enhance the employer-employee exchange relationship. Employees view the implementation of HPWS through which they get economic and social resources, as well as their esteem and sense of personal accomplishment (Zhang et al., 2018). Such positive reciprocal relationships will lead to work engagement (Parzefall and Salin, 2010). HPWS has been shown to be a key indicator of management commitment to service quality in the hospitality sector (Murphy et al., 2018). HR practises have also been found to be effective in supporting hospitality employee engagement and retention (Zhang et al., 2013).
With its theoretical underpinning in SET, POS refers to employees' beliefs "concerning the extent to which the organization values their contributions and cares about their wellbeing" (Eisenberger et al., 1986, p. 501). Supporting this view, meta-analytical by Kurtessis et al. (2017) has found that POS to have a strong positive effect on individual and organizational outcomes, although there is only a limited literature on the role of POS in service-related businesses dealing with consumers in the sector (with the exception of Page et al., 2018). We argue that when employees perceive a positive and supportive work environment, they will produce positive work outcomes in response to the favorable treatment by their organization. Conversely, we predict that the presence of a negative workplace environment, where staff experience negative behaviors and excessive pressures from workplace bullying, will negatively affect organizational outcomes.
PSC is a facet of an organization's safety climate and is defined as the "policies, practises and procedures for worker psychological health and safety" (Dollard and Bakker, 2010, 580). It is conceptualized as support from senior management, prioritization of employee psychological health and safety over productivity goals, psychosocial health and safety communication, and involvement in psychological health and safety (Hall et al., 2010). There is empirical support for PSC in mitigating workplace bullying (e.g., Dollard et al., 2017;Law et al., 2011). Research also posits that job design and workplace environment are important antecedents of bullying, rather than interpersonal factors Fox and Cowan, 2015;Skogstad et al., 2011). The present study, therefore, utilizes PSC as a potential buffer of the influence of HPWS on workplace bullying.

Mediating effects of work engagement and workplace bullying
Our first hypotheses relate to work engagement -a positive state of mind towards work -and employees' affective commitment to the organization as the ultimate outcome. Work engagement is considered to be a dimension of employee's wellbeing (Zhang et al., 2013). It is defined as "a positive, fulfilling, work-related state of mind that is characterized by vigor, dedication and absorption" (Schaufeli et al., 2002, p. 74). Affective commitment is considered the emotional bond between an employee and their organization (Meyer and Allen, 1991). These two constructs are seen to be discriminant and complementary components of work attachment (Hallberg and Schaufeli, 2006) in such a way that employees with high work engagement are unlikely to find it easy to separate from their work because of their investment of time and energy in work and self-identity with the work that they do, increasing the emotional bond with the organizations, leaders, and co-workers (Schaufeli and Bakker, 2004). Therefore, work engagement is a key predictor of affective commitment (Hakanen et al., 2006;Saks, 2006).
In line with SET, the implementation of HPWS emphasizing training, decision-making involvement, clear job description, appraisal, job security, incentives and rewards could positively lead to the wellbeing of hospitality employees as they reciprocrat the quality of exchange relationship with their employer (Murphy et al., 2018;Page et al., 2018). Karatepe (2013a) found work engagement to be a full mediator of the effects of such HR practices on job performance and extra-role customer service in a sample of frontline hotel employees and their managers. HPWS was also found to reduce frontline employees' intentions to leave (Karatepe, 2013b). Similarly, the implementation of HPWS was found to relate to work engagement and job satisfaction and reduce emotional exhaustion (Zhang et al., 2013), and high commitment and service innovative behaviors (Dhar, 2015). A more recent study (Karatepe et al., 2018) also found high involvement HR practises, a construct related to HPWS, to have significant positive effects on job engagement, job satisfaction, and affective commitment.
Drawing from SET, we argue here that when employees perceive that their wellbeing is supported through the implementation of HPWS, they have more trusting and better quality relationships with their employers, indicating the positive outcomes of HPWS on work engagement and affective commitment (Karatepe, 2013b;Karatepe et al., 2018). While work engagement has been found to predict affective commitment (Hakanen et al., 2006;Saks, 2006), there is strong evidence for work engagement to be a mediator of positive work outcomes (Karatepe et al., 2018;Scrima et al., 2014). We, therefore, predict: H1. The relationship between HPWS and affective commitment will be mediated by work engagement.
There is strong evidence linking POS with positive work outcomes from the perspective of SET. POS has been found to positively affect work engagement (Saks, 2006), wellbeing (Rhoades and Eisenberger, 2002;Parzefall and Salin, 2010), and affective commitment (Kurtessis et al., 2017). POS has been found to negatively influence leave intentions of hospitality employees (Cho et al., 2009) by predicting employees' affective commitment (El Akremi et al., 2014;Sharma and Dhar, 2016). There is a direct positive association between work engagement and affective commitment (Hakanen et al., 2006) and work engagement mediating this relationship (e.g., Karatepe et al., 2018). We therefore predict that: H2. The relationship between POS and affective commitment will be mediated by work engagement.
Workplace bullying is highly prevalent and damaging to the service industry and hospitality in particular (O'Driscoll et al., 2011;Ram, 2018). Page et al. (2018) noted the many studies that have demonstrated the impactful role of organizational factors on workplace bullying (see Einarsen et al., 2011;Notelaers et al., 2010;Skogstad et al., 2011). SET also posits that employees tend to seek balanced, reciprocal relationships with their employers and positive interpersonal relationships with others (Rhoades and Eisenberger, 2002). In the instance of a positive HPWS and POS, employees may perceive adequate support and fair treatment, thus negative behaviors like bullying are inhibited (Page et al., 2018;Tuckey et al., 2009), while a poor work environment is associated with the growth of bullying (Skogstad et al., 2011). We therefore posit that both HPWS and POS will be directly related to workplace bullying.
The literature demonstrates consistent and strong evidence showing that workplace bullying as a negative aspect of work environment directly results in various negative outcomes. For instance, employees experiencing bullying are likely to have low motivation (Bentley et al., 2012;Einarsen et al., 2011) and experience mental health problems, strains, low job satisfaction, and commitment to the organization (Nielsen and Einarsen, 2012). As argued above, work engagement is positively associated with affective commitment (e.g., Hakanen et al., 2006) and could play as a mediator for affective commitment (e.g., Scrima et al., 2014). We therefore hypothesize the following: H3. The relationship between workplace bullying and affective commitment is mediated by work engagement.
A recent study by Page et al. (2018) found workplace bullying to fully mediate the relationship between HPWS and hospitality employees' job satisfaction and leave intentions, in both cases negatively. In this instance, HPWS and POS can be seen as organizational resources that help employees to reduce negative behaviors like bullying. The presence of bullying in a workplace can be viewed by employees as organizational unfairness and a breach of psychological contract leading to negative work outcomes (Parzefall and Salin, 2010). Drawing from SET, it is contended that when employees perceive low levels of fairness of treatment and organizational care for their wellbeing, they will perceive the exposure of workplace bullying that results in a loss of job satisfaction, motivation to engage in their work, and commitment to the organization that is opposite to intention to leave (Nielsen and Einarsen, 2012). In line with previous arguments, we hypothesize: H4. The relationship between HPWS and work engagement is mediated by workplace bullying.

H5. The relationship between HPWS and affective commitment is mediated by workplace bullying.
H6. The relationship between POS and work engagement is mediated by workplace bullying.

H7.
The relationship between POS and affective commitment is mediated by workplace bullying.

Moderating effect of PSC
Utilizing a SET lens, PSC is proposed to be important moderator by reducing the presence of bullying in the HPWS-work engagement relationship. Together with HPWS, PSC increases common awareness of the organization's concerns and the provision of support and protection for employee wellbeing through policies and practises focusing on psychosocial health and safety. Additionally, PSC encourages the involvement and participation of all the organizational stakeholders in psychological health and safety protection programs (Dollard et al., 2017). In a high PSC environment, employees are able to use provided tools to freely communicate with managers about work-related issues (Dollard et al., 2017). As a result, employees feel safe and able to challenge bullying and cope with exposure by utilizing the supportive resources of PSC. We hypothesize: H8. PSC will moderate the indirect relationship of HPWS-workplace bullying through work engagement.
In summary, Fig. 1 presents the proposed moderated mediation model.

Participants and procedure
We distributed an online survey utilizing a two-wave research design (separated by a six-month interval in 2015) to a large sample of hospitality employees based in the USA with the assistance of Qualtrics. This design adopted in hospitality research (see Page et al., 2018; Xiong and King, 2018) enables a broader sampling frame and the potential of limiting social desirability with ensured anonymity and confidentiality by the third party provider.
Respondents must meet the inclusion criteria (that is, greater than 18 years old and working in an organization in the US hospitality sector). In Wave 1, we collected demographic data, HPWS, POS, PSC, workplace bullying, and work engagement data from a sample of 467 respondents. In Wave 2, we collected data on the dependent variable, affective commitment. The final sample comprised 203 US hospitality employees who completed the survey at two time points.
Procedural and statistical remedies check for common method variance (CMV) (Podsakoff et al., 2012). Harman's single factor test showed 11 factors revealing that none of the factors explained more than 50 % of the variance. Following Lindell and Whitney (2001), the difference of correlations of the latent variables with and without the marker variable was .05, indicating that the correlations between exogenous constructs and the endogenous variable could not be accounted for by the marker variable. These techniques are commony used in hospitality management research (see Min et al., 2016). The incorporation of moderation and mediation effects provided another check for CMV (Podsakoff et al., 2012).

HPWS
HPWS was measured with the 27-item scale from Sun et al. (2007) on a seven-point Likert scale from '1' means "strongly disagree" to '7' means "strongly agree". After removing four cross-loading items, exploratory factor analysis resulted in two factors (KMO = .944, 62.8 percent with eigenvalues greater than 1.0). CFA resulted in a single factor solution of 14 items (CMIN/df = 1.938, CFI = .979, TLI = .967, RMSEA = .068). Fit indices were above the cut-off values as suggested by Byrne (2016). Sample items include "great effort is taken to select the right person" (α = .96).

POS
POS was measured with the 8-item scale from Eisenberger et al. (1997). Two items were removed due to low factor loading. Respondents were asked to indicate their level of agreement on a sevenpoint Likert scale ('1' = strongly disagree to '7' = strongly agree). This scale had good model fit (CMIN/df = 1.765, CFI = .995, TLI = .989, RMSEA = .062). Sample items included "my organization strongly considers my goals and values" (α = .92).

Workplace bullying
Workplace bullying was assessed by using the 22-item revised version of the Negative Acts Questionnaire (NAQ-R) constructed by Einarsen and his associates (2009). EFA resulted in a unidimensional factor (KMO = .959, 68.1 % with eigenvalues greater than 1.0). This finding contradicted the original scale developed in Europe where there were three dimensions (namely, person-related, work-related and physical intimidation bullying). One item was removed (item 22) and the unidimensional scale met the minimum cut-offs for goodness of fit (CMIN/df = 1.751, CFI = .975, TLI = .966, RMSEA = .061). Sample items included "being humiliated or ridiculed in connection with your work"; "being shouted at or being the target of spontaneous anger" (α = .98).

Work engagement
Work engagement was measured with the UWES-9 work engagement scale (Schaufeli and Bakker, 2003). This scale had a satisfactory goodness of fit (CMIN/df = 1.213, CFI = .996, TLI = .994, RMSEA = .033). Sample items included "when I get up in the morning, I feel like going to work" (α = .94).

PSC
PSC was measured by using the 12-item scale from Hall et al. (2010). This scale had a satisfactory goodness of fit (CMIN/df = 1.853, CF I = .990, TLI = .982, RMSEA = .065). Scale items included "psychological wellbeing of staff is a priority for this organization" and "my contributions to resolving occupational health and safety concerns in the organization are listened to" (α = .97).

Affective commitment
Affective commitment was measured with five items from Meyer and Allen (1991). This scale was collected six months after Time 1 data collection for independent and control variables. The scale met the minimum cut-off for goodness of fit (CMIN/df = 1.213, CFI = .996, TLI = .994, RMSEA = .033), indicating the convergent validity. Sample items included"I feel strong ties with my employer" (α = .95).

Control variables
We controlled for age, gender (male vs female), highest education (from high school to college education), race (re-coded as '1' White and '2' non-White), shift work (yes/no), type of hospitality organization and tenure.

Measurement model estimation
We used IBM SPSS v.25 and IBM AMOS v.25 to check scale validity and test the hypotheses. We followed the two-step approach suggested by Anderson and Gerbing (1988) to develop the hypothesized-six factor model. Fit indices for the six factor model were satisfactor (CMIN/ df = 1.569, CFI = .924, TLI = .918, RMSEA = .053, SRMR = .057). We conducted a Chi-square difference test to compare the fit indices of the hypothesized model with alternate models to check for discriminant validity (see Table 1). Result of the analysis indicated that the hypothesized model had the best fit. All standardized factor loadings (see Table 2) were statistically significant (p < .001) and above .60 (Hair et al., 2010). Table 3 shows that all the values of Average Variance Extracted (AVE) were above .50, confirming convergent validity. In addition to this, all correlation values among latent variables ranged from .08 to .80, and the square root of the AVE value for each construct was much larger than its correlation with any other construct (Fornell and Larcker, 1981). The heterotrait-monotrait ratio of correlations (HTMT) 2 values between the constructs were below .90 (Henseler et al., 2015). These results assured the discriminant validity of the six constructs. Finally, multicollinearity was not an issue as the tolerance values (ranged from .252 to .480) were greater than the threshold value of .10, and the VIF values ranged between 1.12 and 3.97 were below the threshold value of 10 (see Hair et al., 2010).

Results
A recent study using the same NAQ-R scale conducted in the hospitality sector in Australia (see Page et al., 2018) found that 2 HTMT shows the average of the correlations of indicators across constructs relative to the average of the correlations of indicators within the same constructs. This approach has been recently recommended as an alternative and superior approach for discriminant validity of constructs (Henseler et al., 2015).
approximately 20 % of the respondents reported experiencing bullying behaviors at least weekly. Our study found a similar situation in the US hospitality sector where 16.3 % reported that they experienced bullying behaviors at least weekly. There were six significant bullying acts that occurred at least weekly (ranging from 11.3-24.6 %): 'someone withholding information which affects your performance [work-related bullying]', 'being ordered to do work below your level of competence  Table 4 reports the descriptive statistics and correlations between variables. Surprisingly, shift workers reported less workplace bullying. Workers with higher education levels also reported less workplace bullying. There was no significant difference in workplace bullying between White and non-White respondents. Those who had worked longer in the organization and older workers also reported less workplace bullying. These demographic variables were then entered into the path analysis. As reported in Fig. 2, age and ethinicity were retained in the structural model as the other variables were removed due to a lack of significant associations. Non-White respondents reported more experience of workplace bullying behaviors. On the other hand, older respondents reported lower exposure to workplace bullying behaviors.
Our results provided empirical support for a moderated mediation model (CMIN/df = 1.881, CFI = .979, TLI = .964, RMSEA = .066, SRMR = .058). Interestingly, there was no direct association between HPWS and affective commitment. There was a direct association between HPWS and work engagement (β = .31, p < .001) and direct association of work engagement with affective commitment (β = .47, p < .001). Hypothesis 1 was supported as work engagement fully mediated the relationship between HPWS and affective commitment (effect = .18, SE = .07, 95 % CI [.08, .30]). Similarly, there was a direct association between POS and work engagement (β = .45, p < .001) and affective commitment (β = .38, p < .001) respectively. Hypothesis 2 was supported as work engagement partially mediated the POS -affective commitment relationship (effect = .22, SE = .04, 95 % CI [.15, .29]). Surprisingly, there was not support for Hypotheses 3, 4, 6 and 7. Hypothesis 5 was supported as workplace bullying was a full mediator of the relationship between HPWS and affective commitment (effect = .04, SE = .03, 95 %CI [.01, .09]). Hypothesis 8 was supported as PSC was found to moderate the full mediation of workplace bullying on the relationship between HPWS and affective commitment. As shown in Fig. 3, the moderation plot shows that when PSC was high, there was a lower level of workplace bullying with the presence of high levels of HPWS. When PSC and HPWS were low, the level of workplace bullying was consistently higher. Overall, this model explained 25 percent of variance in workplace bullying, 58 percent of variance in work engagement, and 67 percent of variance in affective commitment.

Discussion
This study contributes to the topic of HPWS and workplace bullying in the hospitality sector, a topic where scholars and practitioners have called for more research (Dhar, 2015;Page et al., 2018). Adopting a SET perspective (Cropanzano et al., 2017;Parzefall and Salin, 2010), our empirical study provided support for a moderated, mediation model which suggested that the influence of HPWS on affective commitment was fully mediated by workplace bullying and work engagement. As a moderator, PSC was found to positively moderate the negative impact of HPWS on workplace bullying, which subsequently leads to higher levels of affective commitment. Work engagement was found to act as a partial mediator of the relationship between POS and affective commitment of frontline hospitality employees.

Theoretical implications
Affective commitment is important for retaining employees as hospitality organizations are characterized by poor working conditions and lack of care for employee wellbeing (e.g., Ariza-Montes et al., 2017). This is particularly evident in the case of psychosocial work hazards such as workplace bullying (e.g., Page et al., 2018). The present study contributes to a small but emerging research area that has examined the role of the strategic approach to the management of hospitality employees in enhancing employees' affective commitment (Karatepe et al., 2018;Zhang et al., 2013).
Our findings highlight the importance of providing organizational resources such as HPWS and POS to enhance the affective commitment of frontline hospitality employees. Most hospitality studies tend to adopt a different theoretical perspective (e.g., Karatepe, 2013a,b) which do not lead to the incorporation of these two types of positive organizational resources. Although there is a near absence of empirical studies that have employed both HPWS and POS as potential antecedents of hospitality employee engagement (with the exception of Page et al., 2018), the current study corroborates previous research that has evidenced the powerful impact of POS on various aspects of wellbeing in the general management literature (e.g., Kurtessis et al., 2017), and the weaker and inconsistent positive impact of HPWS on wellbeing outcomes (Woodrow and Guest, 2014). In our study, as expected, the presence of positive and supportive work environment (i.e., POS) and HPWS components (such as selective staffing, extensive training, job security, clear goals and measures of performance appraisals, incentives and rewards) are both engendering in contributing towards the enhancement of affective commitment through work engagement among frontline hospitality employees. These findingsecho earlier research evidence for the positive impact of HPWS on wellbeing (Karatepe et al., 2018;Zhang et al., 2013) and commitment (Dhar, 2015;Karatepe, 2013b). The path loading of POS on work engagement (β = .46) was larger than the impact of HPWS (β = .31), suggesting that a positive, supportive environment has more influences in enhancing hospitality  employees' engagement where their work environment is poor (Ariza-Montes et al., 2017) and characterized by a high prevalence of workplace bulling (Ram, 2018;Page et al., 2018). This is consistent with the SET lens used in the current study.
We also contribute to a growing number of studies which adopted the NAQ-R scale as a research tool for examining workplace bullying (e.g., Einarsen et al., 2011;Page et al., 2018). As the US hospitality industry is typically characterized by poor work conditions (Kitterlin et al., 2016;Ram, 2018) and poor implementation of HR practices (Ariza-Montes et al., 2017), we found empirical support for a relatively high prevalence of workplace bullying in this industry, consistent with earlier research (Ram, 2018). Specifically, an average 16 % of the respondents reported multiple negative acts to occur frequently at their workplace. These were acts related to person-and work-related bullying in hospitality organizations. The most frequently reported bullying behaviors observed in the present US study (withholding information that affects performance; being ordered to do work below competence level; having your opinions ignored; spreading gossip and romours about you; excessive monitoring of your work) were different to those behaviors reported most commonly in a sample of Australian hospitality workers using the same measure (Page et al., 2018). These differences possibly reflect cultural and industry differences in bullying perceptions between the US and Australia . The bullying behaviors found in our study tend to reflect the cultural and managerial characteristics in the US hospitality sector which focus on individual performance, employment insecurity, and imbalanced power relations (Ariza-Montes et al., 2017;Ram, 2018).
Although workplace bullying in the hospitality industry could be argued to be "part of the job" (Ariza-Montes et al., 2017;Ram, 2018), we provide support to show that exposure to these toxic behaviors had a negative impact on affective commitment in a sector which is characterized by high turnover (Ram, 2018), and supports previous research that has evidenced the negative influence of bullying on affective commitment Page et al., 2018). This study further contributes to the hospitality management literature, especially on HPWS (Murphy et al., 2018) through the finding that a set of social exchanged-focused HR practices (see Sun et al., 2007) minimized the exposure of frontline employees to workplace bullying (Parzefall and Salin, 2010;Salin, 2003). This novel finding adds to our knowledge regarding the management of ill-treatment in the hospitality workplace by the implementation of HPWS components providing frontline hospitality employees the acquisition of technical and interpersonal skills, service delivery knowledge, and potentially lead to enhanced productivity (Karatepe, 2013b;Murphy et al., 2018). These practices may also reflect the perceived fairness of HPWS implemented amongst hospitality employees and in the influence of reciprocity from a SET view (Cropanzano et al., 2017;Parzefall and Salin, 2010), HPWS can mitigate the exposure to workplace bullying among hospitality employees.
The main contribution of our study is the moderating role of PSC in   mitigating the exposure of workplace bullying to enhance employees' affective commitment in the hospitality industry. While the impact of PSC in minimizing workplace bullying has been known in the psychology literature (e.g., Dollard et al., 2017), it has not been found in the hospitality literature. In our moderated-mediation model, PSC represents an organization's commitment to and prioritization of employees' psychosocial health and safety (Dollard and Bakker, 2010). We found that the positive impact of HPWS practices was moderated by the presence of a safe, psychosocial work climate (PSC), reducing the exposure of workplace bullying and increasing affective commitment (see H8). While we are unaware of other studies that have used PSC as a moderator, there is good evidence that shows PSC to be an effective tool to manage psychosocial risks such as workplace bullying (Dollard and Bakker, 2010). We note that this is the first study that has employed PSC in the hospitality context characterized by stressful working conditions -particularly customer interactions, together with high bullying rates. The findings indicate that to maximize the positive effects of HPWS on wellbeing of hospitality employees, the development of PSC plays an important role in strengthening a safe and positive environment where perceptions of stress and bullying are reduced, thereby promoting a high degree of reciprocal affective commitment behaviors from employees.

Managerial implications
As noted in previous research in the hospitality and tourism sectors (Bentley et al., 2012;Ram, 2018), there is a clear need to manage the risk of exposure to workplace bullying behaviors as it usually associated with organizational exit (D'Cruz and Noronha, 2010). Given the already high levels of labor turnover across the hospitality sector globally, this makes intervention to manage the psychosocial environment a priority HR concern. The HPWS bundles in the current study incorporate selective staffing, training and internal mobility, performance orientation, job design, and clear communication. Collectively, these bundles of HR practises reflect an emphasis of high performance and participative management in hospitality organizations. Our study has provided more evidence of the potential for HPWS to enhance both engagement and commitment of sector employees. We recommend the adoption of a strategic HR approach to be accompanied by organizational efforts to promote a positive work environment that is operationalized through POS and effective management of the psychosocial safety climate.
Workplace bullying has consistently been associated with unwanted outcomes for individuals and organizations. It appears to weaken or even reverse the impact of organizational efforts to enhance worker engagement and commitment. While the hospitality sector is characterized by a number of factors that increase the likelihood of interpersonal bullying behaviors from managers, co-workers, and customers (see Roper and Menten, 2017), is clear from extant research that bullying is not a simple interpersonal phenomenon . We argue that workplace bullying must be viewed within the hospitality industry as a symptom of psychosocial work hazards.
Organizational-level interventions for managing and responding to workplace bullying could be established. PSC can, therefore, be operationalized as an important, effective, and potent organizational intervention to reduce the negative impacts of bullying in the HPWS-affective commitment relationship in the hospitality context as it is characterized by stressful working conditions and high psychological and emotional demands associating with customer service. As the current study is one of the first in hospitality management to measure PSC in the US hospitality context, we note that the instrument can assist management to create, achieve, and manage a psychologically healthy work conditions in order to reduce the experience of stress and bullying, and ultimately promote a high degree of reciprocal commitment behaviors from hospitality employees.

Limitations and future studies
Considerably more research attention is needed within the hospitality context, looking at sector employee wellbeing and its role in enhancing commitment and retention. Indeed, further studies should look to see if the relationships observed in the present study are found  in other national settings (Jacobson et al., 2014) within the hospitality (Ram, 2018) and the wider visitor economy (Page et al., 2018). However, the prevalence of workplace bullying in the present study is similar to that found in Australia (see Page et al., 2018) who used the same instrument to measure workplace bullying. Specifically, research should continue to explore the conditions under which strategic HR approaches contribute to, as well as negatively affect, sector employee wellbeing, especially in light of the fact that HPWS and similar strategic HR approaches are increasingly applied in hospitality organizations to enhance performance. A second focus of future work in this field should be to examine the influence of PSC on the work environment, as either an antecedent, mediator or moderator, and its role in perceptions of bullying and other psychosocial factors such as stress, harassment and fatigue amongst sector staff.
While our data collection was over two points in time, there is a risk it could potentially be affected by common method bias. CMV was checked by undertaking prescribed procedural and post-hoc statistical checks to ensure the risk of common method variance was minimized (Podsakoff et al., 2012). Certainly, future research should seek to employee longitudinal data collection that is able to examine the influence of study variables on wellbeing outcomes over the longer-term. Despite this limitation, and the need for further research expressed above, the study does provide a starting point towards understanding the organizational conditions that promote sector wellbeing and, in particular, employee commitment in a sector that struggles globally with the retention of workers.

Conclusion
This research was concerned with measures to enhance affective commitment in the hospitality sector. The study focused on the role of HPWS, work environment factors (such as PSC and POS) and workplace bullying, in enhancing affective commitment of hospitality employees. The major findings were that work engagement mediated the relationship between HPWS and affective commitment. Similarly, work engagement also mediated between POS and affective commitment. Furthermore, workplace bullying negatively mediated the association between HPWS and affective commitment, suggesting that even engaged employees have difficulty committing to the organization where they perceive a toxic work environment. This finding was in line with the small body of previous research on HPWS and workplace bullying in the hospitality management literature. The study also found that PSC accentuated the negative effect of HPWS on workplace bullying-affective commitment relationship, suggesting a potential solution for hospitality organizations to manage psychosocial risks in a sector characterized by negative workplace events.