Customer citizenship as a reaction to hotel's fair treatment of staff: Service satisfaction as a mediator

Introduction

Customer citizenship behaviours are prevalent and beneficial helping behaviours that guests perform during the service encounter in organisations. They are highly regarded by managers because they go beyond the expected role of guests in the service encounter and, hence, exceed role minimums. Groth (2005: 11) defines customer citizenship as ‘voluntary and discretionary behaviours that are not required for the successful production and/or delivery of the service but that, in the aggregate, help the service organisation overall’. Examples of these behaviours include providing helpful customer service feedback or information when surveyed by the business and informing the business about the good service given by an individual employee. Prior research has generally endorsed customer citizenship behaviour as producing beneficial outcomes, such as creating positive outlooks on life, reducing stress in the service encounter exchange, or contributing to the service organisation's effective functioning (Lengnick-Hall et al., 2000).

Much of the recent literature in the hospitality industry has focused on the role guests can play in supporting the success of hotel service encounters. Specifically in hotel contexts, customer citizenship behaviour can be an important issue for hotels due to its prevalence and potential beneficial consequences. Examples of customer citizenship behaviour in the hospitality industry can include taking actions to reduce the hotel's costs or protect it from potential problems and expressing irritation with people who squander the hotel's resources (Zoghbi-Manrique-de-Lara et al., 2014). As guests use and dispose of products, services, and experiences during a long in situ presence in the hotel service encounter (Bettencourt, 1997; Groth, 2005; Halbesleben and Buckley, 2004), eliciting customer citizenship behaviour can indeed be important for hotels, due to the particular role guests play in supporting their success. Despite this, only a limited number of prior empirical studies on customer citizenship behaviour in the hospitality industry have examined its origins.

One approach that has received considerable attention as a way of explaining how hotel guests evaluate service providers' actions and react accordingly is the justice theory (Chebat and Slusarczyk, 2005; Dewitt et al., 2008; Karande et al., 2007; Karatepe, 2006; Kim et al., 2009). The justice theory has been applied in previous research to seek antecedents of customer citizenship behaviours. Results show that customers' perceptions of the (un)just treatment received by the hotel's representatives in the service encounter lead them to engage in citizenship behaviour (Yi and Gong, 2008; Zoghbi-Manrique-de-Lara et al., 2014). Although some prior justice research suggests that third parties can also make fairness judgments and react to the way others are treated (e.g. Skarlicki et al., 1999), the victim's perspective dominates this literature. There is, therefore, a lack of models that examines whether guests may react to a hotel in the form of citizenship behaviour based on the way they perceive management's treatment of staff, that is whether they perceive the staff as the target of the management's (mis)treatment. A question remains, therefore, about whether customers' perceptions of hotel management's behaviour towards staff are able to predict customer citizenship behaviour.

Perceived justice in service encounters is a multi-faceted concept rooted in the equity theory by Adams (1963), which has three dimensions: distributive, procedural, and interactional justice. In addition to being directed towards themselves, these three justice dimensions can be experienced by guests in the way the hotel management treats the hotel employees in the service encounter (Bowen et al., 2000; Solomon et al., 1985). In this case, the experiences of guests witnessing hotel staff (mis)treatment can take the form of distributive justice, or the degree to which they feel staff is fairly treated with respect to the outcome of the service encounter (e.g. guests witness that the hotel values the effort employees make or it is fair in assigning employees' workload); procedural justice, or the fairness in the policies and procedures applied to staff by the hotel (e.g. guests witness that the hotel takes actions that make it easier for employees to fulfil their duties or that it pays attention to employees' concerns); and interactional justice, or justice perceptions concerning the staff's interpersonal dealings with the representatives of the hospitality organisation (e.g. guests witness that the hotel is honest with its employees and treats them with kindness and consideration) (Blodgett et al., 1997; Tax et al., 1998; Voorhees and Brady, 2005).

In addition to the need for research on customers’ third-party reactions to justice towards staff in the form of citizenship behaviour, there is also a need to explain why this reaction might be possible. This paper invokes the unitarism theory derived from Alan Fox (1966) to justify this mediation. Unitarism emphasizes the shared interests of an organisation. The suggestion is made that if guests are motivated by either self-interest concerns or a moral imperative (Folger, 2001), they care about the hotel staff's (mis)treatment because, by threatening the staff's well-being, (in)justice towards staff threats the guests’ interest in having a satisfactory stay as well. Thus, satisfaction with hotel service may embody guests’ reasons for responding to (in)justice towards staff with citizenship, and the more guests perceive the management's treatment of staff as fair, the more the reactions of guests can follow a sort of ‘satisfaction mirror effect’ (e.g. Heskett et al., 1997), which has its origin in Alan Fox's (1966) unitarism. As a result, we postulate that (in)justice towards hotel staff will predict guests’ satisfaction with the hotel service, which, in turn, plays a mediating role in the relationship between (un)fair treatment of staff and customer citizenship behaviour.

Before examining the predicted positive effects of justice towards staff on customer citizenship behaviour (H1) and satisfaction with hotel service (H2), we will first provide empirical evidence that all the variables used in this study are separate. We will then examine the expected mediating role of service satisfaction (H3). Finally, the paper will discuss theoretical and managerial implications of the findings.

Theoretical background and hypotheses

Dennis Organ (1988) defined organisational citizenship behaviour (OCB) as voluntary and discretionary individual employee behaviour that by exceeding role minimums is expected to promote overall organisational effectiveness. However, he later found that rather than an explicitly ‘extra-role’ behaviour (to use the Van Dyne et al. (1995) terminology), these behaviours may be OCB and still be perceived by employees as part of their ‘role’. Consequently, Organ (1997) later changed his definition to say that OCB is behaviour that supports the social and psychological context in which tasks are conducted. These shifts in the OCB conception have also included its multidimensional nature. Organ (1988) identified conscientiousness, sportsmanship, civic virtue, courtesy, and altruism as five well-established dimensions of OCB. A few years later, however, Williams and Anderson (1991) divided OCB into behaviours directed at individuals (OCB-I) and those directed at the organisation as a whole (OCB-O). An example of this interpersonal citizenship behaviour (OCB-I) would be assisting co-workers with their duties, while an example of citizenship directed at the organisation (OCB-O) would be defending the hotel when other employees criticise it.

Despite these important movements regarding the OCB concept and its facets, few studies have dealt with the boundaries of conventional OCB considering the possibility of its performance outside the figure of the employee. Thus, to date, only Zoghbi-Manrique-de-Lara et al. (2014) have empirically developed customer citizenship behaviour within hospitality contexts, where guests perform this helping behaviour during their stay at the hotel and directed at the hotel as a whole. Although not restricted to the hospitality industry, other scholars have also suggested the idea of OCBs performed by customers within the service encounter. They would include customer discretionary behaviour (Ford, 1995), customer voluntary performance (Bailey et al., 2001; Bettencourt, 1997), organisational citizenship on the part of customers (Lengnick-Hall et al., 2000), co-production (Gruen et al., 2000), customer extra-role behaviours (Aherne et al., 2005), and customer dedicational behaviours (Bendapudi and Berry, 1997; Kang and Hyun, 2012; Stanley and Markman, 1992). In turn, customer dedicational behaviours have been divided into different facets, such as enhancement, or customer actions leading to greater investment in the exchange (Bendapudi and Berry, 1997; Cross and Smith, 1995); cooperation, or customer actions that help to exceed what the firm can achieve by itself (Anderson and Narus, 1990); and advocacy, or customers' support for the service company (Bendapudi and Berry, 1997). Finally, Groth et al. (2004) have also identified different facets of customer citizenship behaviour that include: (a) providing the company with feedback that helps to improve it, (b) helping other customers, and (c) recommending the company to others.

Hotels as service providers (unlike ordinary staff or other guests, for instance) are likely to be perceived as ultimately being responsible for fairly treating staff and handling service during the guests' stay. Therefore, a measure of customer citizenship behaviour directed at the hotel performed by guests during their stay (Zoghbi-Manrique-de-Lara et al., 2014) will be the criterion variable in this study.

Prior research indicates that customers who have (un)favourable perceptions of justice are more inclined to perform various actions in the service encounter, such as displaying feelings of complaint satisfaction and loyalty (Karatepe, 2006) and loyalty-exit behaviour (Chebat and Slusarczyk, 2005). Hotel guests often act as third parties in hotel's service encounters, and they are quite likely to frequently become aware of a number of slights and disrespectful events towards staff by the hotel. Previous third-party literature suggested that an observer who witnesses (in)justice for others is likely to respond to this perceived justice in a manner similar to that of an actor–victim in the situation (e.g. Skarlicki et al., 1999). Because customers' perceived justice for the self is omnipresent in hospitality research (e.g. Karatepe and Shahriari, 2014; McCollough, 2000, for a review), and it has been found to be positively associated with customer citizenship behaviour (Yi and Gong, 2008), unfavourable justice perceptions for staff are also expected to be an antecedent of customer citizenship behaviour.

In addition, based on Bowen (1986) and Vargo and Lusch's (2008: 2) suggestion that ‘the customer is always a co-producer […] who creates value’, and along with staff, forms part of the hotel's human resources, a social identity-based explanation is also provided to justify why justice towards staff is predicted to affect customer citizenship. Based on this approach, justice towards staff may be linked to customer citizenship behaviour because guests perform these behaviours to reciprocate to the hotel as if they were staff, with whom, as co-producers, they feel identified (Vargo and Lusch, 2008). For example, if a hotel has a sufficient reception staff, guests (in their check-in) can witness how receptionists are easily able to cope with their tasks. These events can be sensitive for guests who can feel distributive justice towards reception staff as their own, producing, as Lind and Tyler (1988: 191 (emphasis added)) suggested, ‘…both satisfaction and positive evaluations of the [hotel] organisation’. Guests’ feelings of identification with the reception staff performing the hotel service are likely to lead them to reciprocate to the hotel with increased citizenship behaviour, as if they were members of the hotel's workforce.

By contrast, events involving injustice for staff can become an embarrassment for guests, which can provoke responsibility attributions towards the hotel. As such, if guests witness receptionists being treated by the reception manager with unkindness and lack of consideration during the check-in, guests are likely to close ranks with the hotel staff and, ultimately, react to injustice towards staff with decreased customer citizenship behaviour.

Therefore, we postulate that guests will respond to observing (un)fair treatment of staff with customer citizenship behaviour (see Figure 1). Thus,

Hypothesis 1: Guests' favourable perceptions of justice towards staff will have positive effects on customer citizenship behaviour.

The evidence also indicates that the idea that perceptions of justice towards staff lead guests to perform customer citizenship behaviour is quite simplistic. Instead, mediators should underlie this relationship. Bove et al. (2009) found, for instance, that personal loyalty to the service worker serves as a partial mediator of the effects of perceived benevolence of the service worker and commitment to the service worker on customer citizenship. The prediction here is made that individual satisfaction with hotel service could act as a mediator that explains why perceptions of justice towards staff cause guests to find reasons to display customer citizenship behaviour. However, we must first predict that guests' favourable perceptions of justice towards staff will have positive effects on satisfaction with hotel service.

As stated in the preceding discussion, social exchange dynamics (Blau, 1964) between a service organisation and its guests may have similar consequences when guests act either as victims or observers of staff (un)fair treatment. Accordingly, providing a fair work environment for either guests or hotel staff is similarly essential for the hotel's success in providing a satisfactory service to guests (see also Liedtka et al., 1997; Normann and Ramirez, 1993; Schlesinger and Heskett, 1991; Silvestro and Cross, 2000). This idea can find justification in the unitarism theory. Unitarism is derived from the work of industrial sociologist Alan Fox (1966), who identified a perspective on employment that emphasizes the shared interests of an organisation. The ‘satisfaction mirror effect’ (Heskett et al., 1997) points to its origin in unitarism (Fox, 1966) and would support the belief that hotel employees’ satisfaction because of fairness is able to lead guests to experience satisfaction with the hotel service as well. As such, if workers experience job satisfaction when performing customer service jobs, their satisfaction will rub off on guests, who will also experience satisfaction in the service encounter. The specificities of the hospitality industry would reinforce this idea because hotel guests usually need the staff to be supported, given that their interest in obtaining satisfactory service is highly determined by ‘the strength of spirit’ of the staff.

Using these arguments as a guide, we postulate that if the hotel staff is treated fairly by their managers, guests are likely to feel a contagion with staff in ways that elicit satisfactory service experiences. As a result, perceived justice towards staff should predict guests' satisfaction with the hotel service (see Figure 1). Therefore,

Hypothesis 2: Guests' favourable perceptions of justice towards staff will have positive effects on satisfaction with hotel service.

Given the overlapping interests that guests and staff share within the hotel service encounter, the ‘satisfaction mirror effect’ (Heskett et al., 1997) can provide theoretical support for the idea that perceptions of justice towards staff lead guests to experience service satisfaction. However, this study also predicts the mediating role that service satisfaction may play in the relationship between justice towards staff and customer citizenship behaviour. The question is now why guests’ service satisfaction forms part of the underlying mechanisms that make it possible for fairness to ultimately lead guests to perform citizenship.

In terms of rational standards that apply directly to behaviour, fairness theories have identified two general reasons why guests respond to (in)justice: (a) self-interest concerns and (b) moral motivation. This paper first argues that, guided by a self-interest motivation, guests’ service satisfaction could explain why they perform citizenship after witnessing justice for staff. Based on social exchange theory (Blau, 1964), the argument is made that the predicted positive effects of guests’ favourable perceptions of justice towards staff on service satisfaction could lead guests to preserve this happy situation by displaying citizenship behaviour. Thus, guests are expected to use calculated reasoning when witnessing justice for staff in order to maintain the fruitful service encounter derived from the ‘shared or overlapping interests’ they share with staff (see also Bowen, 1986; Halbesleben and Buckley, 2004; Heskett et al., 1997). Using the exchange patterns, guests would positively reinforce the hotel's fair actions by displaying increased customer citizenship.

On the other hand, guests could also act out of a moral imperative (Folger, 2001). Cropanzano et al. (2003), Folger (1998, 2001), and Folger et al. (2005) refer to this moral behaviour in responding to (in)justice as a ‘deontic response’ (from the Greek, deon-: obligation), that is evolutionary-based reactions in the form of moral automatisms, emotions (e.g. anger), retributions to restore justice, and even irrationalities, at least in the short term. If justice towards staff is, for example, unfavourable, guests may not only feel ‘unhappiness contagion’, but they might also be morally outraged about the unjust way the hotel provides the service. As they can feel like ‘co-producers’ (Vargo and Lusch, 2008) who are close to staff, they feel that they share ‘a similar fate’ and emotional states with them (e.g. Zoghbi-Manrique-de-Lara et al., 2013). In this regard, Gaudine and Thorne (2001) state that emotions are especially present in moral decision-making, suggesting that service satisfaction can explain guests’ reactions to (un)fair treatment of staff based on the idea that this is the ‘right way to act’. Satisfaction with service may comprise, therefore, the feelings and emotions that ultimately lead guests to respond to (un)fair treatment of staff with citizenship behaviour. Thus,

Hypothesis 3: Guests' satisfaction with hotel service mediates the relationship between favourable justice towards staff and customer citizenship behaviour.

Method

Survey respondents

We surveyed guests over 18 years old who had stayed at least four nights at one of four 4-star hotels or three 5-star hotels in Gran Canaria. Satisfaction with hotel service and visible events of employee (mis)treatment are particularly important for hotels at this level. We observed an average occupancy of 80% for the 4-star hotels and 76% for the 5-star properties. With an average guest stay of 6.8 nights, we estimated a population size of 1,093,540 guests and established a sample of 384 guests for a 5% sampling error. As shown in Table 1, the sample includes guests who stayed in half-board (includes breakfast and dinner), full board (includes breakfast, lunch, and dinner), and all-inclusive packages.

OPEN IN VIEWER

Table 1.

Sample comparative descriptions.

Nationality Descriptions (total)	Germany (25%)	United Kingdom (16.2%)	Scandinavia a (30.6%)	Holland (4.7%)	Ireland (5.0%)	Spain (9.1%)	Others (9.4%)
All-inclusive (30.2%)	11.5%	53.7%	46%	14.3%	14.3%	16%	10%
Age -41 years (28.2%)	12%	36.5%	38.3%	12.5%	5.9%	40%	33.3%
Age + 55 (34.5%)	38.5%	32.7%	31.4%	43.8%	64.7%	23.3%	26.6%
Average stay (9.34 nights)	9.91	8.98	8.76	7.80	11.13	6.23	13.67
Male (57.7%)	62.2%	53.7%	55.2%	60%	62.5%	48%	58%
Female (42.3%)	37.8%	46.3%	44.8%	40%	37.5%	52%	42%
University degree or higher (54.5%)	51.3%	55.3%	60%	70%	33.3%	46.6%	53.8%

a

Scandinavian countries include Denmark, Sweden, Norway, and Finland.

Statistical analysis

The collected data were analysed using the statistical package for social sciences. Structural equation modelling (SEM) was used to assess the validity of the measures and test the hypothesised relationships through the statistical package AMOS 22.0. The mediation tests follow the approach of Anderson and Gerbing (1988), Baron and Kenny (1986), and Sobel (1982). We used planned confirmatory factor analysis (CFA) to test the construct validity of customer citizenship behaviour, along with measures of justice towards staff and satisfaction with hotel service. The fit indices used include goodness-of-fit (GFI), comparative-fit (CFI), normed-fit (NFI), Tucker–Lewis (TLI), incremental-fit (IFI), and root mean square error of approximation (RMSEA). Once the model was entered in AMOS 22, we checked whether the indexes provided a good fit (GFI ≥ .90, CFI ≥ 0.90, NFI ≥ .90, TLI ≥ .90, RMSEA ≤ .05), supporting the unidimensionality of the scales. We also employed a set of established procedures to check for the convergent validity and discriminant validity of our scales. Reliability was assessed by using the Cronbach's alpha. Results are shown in Table 2.

OPEN IN VIEWER

Table 2.

Results of confirmatory factor analysis.

	Factor loadings	Composite reliability	AVE
(F1) Justice towards staff
(Cronbach's alpha = .973)		.969	.707
In your view, the hotel …
X01 … seems to be honest with its employees	.889
X02 … shows concern for the rights of the staff as employees	.886
X03 … treats employees with kindness and consideration	.885
X04 … deals with the staff in a truthful manner	.875
X05 … takes actions that make it easier for employees to fulfil their duties	.865
X06 … allows the employees to make a complaint formally	.845
X07 … allows for employees' feedback about a decision made by the hotel management	.832
X08 … pays attention to employees' concerns	.830
X09 … is fair in appreciating employees' responsibilities	.826
X10 … values the effort employees make	.812
X11 … appreciates the effort employees make when they do the work well	.801
X12 … considers the stress employees cope with	.801
X13 … is fair in assigning the workload that employees have to cope with	.753
(F2) Satisfaction with hotel service
(Squared multiple correlation = .269; Cronbach's alpha = .952)		.908	.768
Y14 I am delighted with the services this hotel offers	.875
Y15 I am very satisfied with my stay at this hotel	.855
Y16 This hotel service exceeded my expectations	.833
(F3) Customer citizenship behaviour
(Squared multiple correlation = .203; Cronbach's alpha = .822)		.830	.552
Y17 I take actions to protect the hotel from potential problems	.827
Y18 I take actions to help reduce the costs of the hotel	.809
Y19 I support improvements in the hotel	.741
Y20 I show concern for the effective functioning of the hotel	.725

AVE refers to average variance extracted. SEM suggestions about modification indices include the following covariances: e1-e2, e3; e2-e3, e4; e5-e6, e7; e6-e8; e7-e8; e9-e10, e11, e12; e10-e11 and e12-e13.

CFA model fit without control variables: Cmin = 268.882; df = 143; p < .001; Cmin/df = 1.880; GFI = .93; CFI = .983; IFI = .984; TLI = .978; NFI = .966; RMSEA = .051.

We measured the variables as follows. For justice towards staff, we used a 13-item scale constructed by the authors, adapting scales from the literature on organisational justice (e.g. Moorman, 1991). It included aspects of distributive, procedural, and interactional justice. However, it is unlikely that guests acting as mere observers are in a position to make a clear distinction between these different aspects of employee (mis)treatment by the hotel. Therefore, a one-dimensional scale is expected. A four-item scale developed by Zoghbi-Manrique-de-Lara et al. (2014) was used to assess customer citizenship behaviour. We measured customer satisfaction with the hotel service using a three-item scale. Chi and Gursoy (2009) used a similar measure to assess hospitality companies' customer satisfaction.

Control variables

We included the following demographic variables as control variables because they could covary with the dependent and independent variables (e.g. Aquino et al., 2004): gender, age (five groups), a variable distinguishing self-employed persons from employed, a six-level graduated scale for educational level, and nationality. To control for potential home country differences, we gave this nationality variable a value of zero if the home country was Spain, and a value of one otherwise. The control variables were incorporated directly into the model as stand-alone variables (not as cause or effect indicators), allowing a structural path to all exogenous and endogenous factors within the structural, but not the measurement, portion of the model (Hancock and Mueller, 2006).

Results

Results suggest that customer citizenship behaviour has significant inter-correlations in the expected directions, as shown in Table 3. These results appear to be a good starting point to support the mediating role of service satisfaction in the link from justice towards staff to customer citizenship behaviour. CFA tested the fit of the hypothesised three-factor model and indicated that the variables were, in fact, three separate constructs. Calculating the average variance extracted (AVE) and Cronbach's alpha for each construct reveals that the former ranges from .707 to .552 (supporting convergent validity, per Bagozzi and Yi, 1988; Fornell and Larcker, 1981), and the latter ranges from .830 to .969, which is above the recommended alpha of .70 (Nunnally, 1978). As Table 2 shows in detail, according to SEM suggestions about modification indices, a few covariances among justice towards staff factor errors were entered. The indexes of the model provide an acceptable fit, with most statistics within the standards suggested by Hair et al. (2006): χ² = 428.550, df = 239, χ²/df = 1.79, GFI = .904, CFI = .969, IFI = .968, TLI = .962, NFI = .911, TLI = .932, and RMSEA = .050. Table 2 shows that composite reliability ranged from .822 to .973, which is greater than the standard of .6 (Hair et al., 2006). In addition to testing the convergent validity of the variables, as shown in Table 3, the results also support discriminant validity, as the square roots of the AVE values (on the main diagonal) were consistently greater than all the corresponding correlations (Fornell and Larcker, 1981).

OPEN IN VIEWER

Table 3.

Means, standard deviations, correlations, and reliabilities.

Variables	M	SD	1	2	3	4	5	6	7	8
1. Gender	1.43	.49	–
2. Age	3.10	.92	−.227**	–
3. Nationality	3.13	1.92	−.027	−.056	–
4. Educational level	4.42	1.07	−.012	−.210**	.001	–
5. Employee/employer	1.75	.68	.233**	−.131*	.087	−.004	–
6. Justice towards staff	5.09	1.02	.049	.105	−.068	−.164**	−.090	(.841)
7. Service satisfaction	5.79	1.09	.051	.164*	−.064	−.110	.014	.506**	(.876)
8. Citizenship	5.29	1.08	.070	.179*	.081	−.268**	−.020	.418**	.478**	(.909)

We tested the resulting model using SEM. Figure 2 is a path diagram that shows the relationships between the observed variables (survey answers, in rectangles) and the unobserved latent variables (circles). The various fit indices used, also shown in Figure 2, reveal an acceptable fit of the model. The significant path from justice towards staff to customer citizenship behaviour supports Hypothesis 1 (B = .315; p = 002). To test Hypothesis 3, the three Baron and Kenny (1986) conditions for mediation were first inspected: (a) the independent variable (justice towards staff) has to predict the criterion variable (i.e. customer citizenship behaviour), (b) the proposed mediator (service satisfaction) has to be predicted by the independent variable (justice towards staff) and predict the criterion variable, and (c) the direct path between justice for staff and customer citizenship behaviour has to decrease (preferably to non-significance: full mediation) when the mediator is added. Baron and Kenny (1986: 1173) define a mediator (service satisfaction) as ‘a third variable, which represents the generative mechanism through which the focal independent variable [justice toward staff] is able to influence the dependent variable of interest [customer citizenship]’. According to standardised beta results in Table 4, conditions (a) and (b) are fulfilled. This pattern supports Hypothesis 2. The third condition (c) was also fulfilled but only for partial mediation. In summary, although justice towards staff was initially associated with customer citizenship behaviour (B = .315; p = .002), the direct path linking justice towards staff to customer citizenship behaviour decreases when service satisfaction is added (without reaching non-significance) (B = .178; p < .05), as suggested by Baron and Kenny to support partial mediation (see Table 4). OPEN IN VIEWER

OPEN IN VIEWER

Figure 2.

Accepted model of justice towards staff, satisfaction with hotel service, customer citizenship behaviour. We incorporated gender, age, self-employed/employee, educational level, and nationality differences as stand-alone control variables directly into the model (with a structural path to all exogenous and endogenous factors). We also entered covariance links between error residues of the justice towards staff variable following SEM suggestions about modification indices. They are: e1-e2, e3; e2-e3, e4; e5-e6, e7; e6-e8; e7-e8; e9-e10, e11, e12; e10-e11 and e12-e13. Note: N = 343. * p < .05; ** p < .01; *** p < .001. χ² [239, 343] = 428.550, df = 239, GFI = .904, IFI = .968, TLI = .962, CFI = .969, RMSEA = .050.

OPEN IN VIEWER

Table 4.

Chi-square and fit index comparison of the fully and partially mediated models of the effects of justice towards staff on customer citizenship behaviour through satisfaction with hotel service. ^a

Model	Betas	A direct path/s linking justice and citizenship?	χ2	df	GFI	CFI	RMSEA	PGFI	PNFI	AGFI
Justice→Citizenship	(B = .315; p = .002)	Main effect model	342.499	179	.910	.910	.057	.650	.738	.879
Justice→Satisfaction →Citizenship (Figure 2)	Justice→Satisfaction (B = .534; p < .001) Satisfaction→Citizenship (B = .351; p < .001)	Yes (B = .178; p < .05(+))	421.134	238	.912	.912	.047	.668	.752	.880
		No (**) ( χ d 2  = 7.41; dfd = 1; p < .001)	428.550	239	.904	.904	.050	.663	.750	.875

a

We incorporated gender, age, self-employed/employee, educational level, and nationality differences as stand-alone control variables (with a structural path to all exogenous and endogenous factors) directly into the model.

(+) Although in the saturated model justice towards staff has a significant direct path to citizenship, it is less significant.

(**) Considering the fit indices provided, this fully mediated model does not have a better fit when compared to the partially mediated model, given that the root-mean-square error of approximation (RMSEA) index is stronger, and goodness-of-fit (GFI), comparative-fit (CFI), parsimony goodness-of-fit index (PGFI), parsimony normed fit index (PNFI), and adjusted goodness of fit index (AGFI) parsimonious fit indices are weaker. However, chi-squared comparison results provide a significant difference.

In addition, a nested model comparison was then conducted using the sequential chi-square difference test. Following Anderson and Gerbing's (1988) recommendations, the hypothesised model (more constrained) was compared to the saturated alternative model (less constrained), in which a direct path from justice towards staff to customer citizenship behaviour was added. This latter model only represents a partially mediated model of the effects of justice towards staff on customer citizenship behaviour. The results of the nested model comparison once again confirm the mediating effect of satisfaction with hotel service on customer citizenship behaviour. In our analysis, we included the main effects between justice towards staff and customer citizenship behaviour (see Table 4). The fully mediated model – with no direct path linking justice towards staff and customer citizenship behaviour – was compared with the partially mediated model, which included a direct path linking justice towards staff with customer citizenship behaviour. The change in the chi-square test of the hypothesised model (χ² [239, 343] = 428.550), when compared with the saturated model (χ² [238, 343] = 421.134), is 7.41 for one degree of freedom, indicating a significant change (p < .001). However, according to the fit indices provided in Table 4, this fully mediated model does not show a better fit than the saturated model, given that the RMSEA index is stronger, and the GFI, CFI, parsimony GFI, parsimony NFI, and adjusted goodness of fit index parsimonious fit indices are weaker.

Finally, the Sobel test and Preacher et al.'s (2007) bootstrapping method were also used to examine the significance of the mediating role of satisfaction with hotel service (Hypothesis 3). The Sobel test shows whether the indirect effects of justice towards staff on customer citizenship behaviour via satisfaction with hotel service are different from zero. A z score larger than 1.96 supports the hypothesis about the indirect effect. The successive approximations calculation provides an estimate of the real indirect effect and its bias with a 95% confidence interval (CI). Both methods demonstrate (see Table 5) that the mediating role of satisfaction with hotel service in the relationship between justice towards staff and customer citizenship behaviour is significant (.0115 < CI_95% < .226, p < .001). The confirmation of justice towards staff's indirect effects on customer citizenship behaviour provided by the Sobel test also supports the mediation thesis. In summary, all of these patterns together provide support for Hypothesis 3 about partial mediation.

OPEN IN VIEWER

Table 5.

Indirect effects of justice for staff on customer citizenship behaviour through satisfaction service (Sobel's test results).

Indirect effect	Value	Std. error	Low 95% CI	Up 95% CI	Z(*)	Sig.
Justice → Satisfaction → Citizenship	.12960	.057344	.114623	.225827	2.26	.00015

Discussion

The purpose of this study was to test a scenario that seems likely to occur in service encounters: guests observing acts of justice towards staff decide to respond to the hotel organisation. The results indicate that justice towards staff leads guests to react in the form of increased customer citizenship behaviour, and that this link is partially explained by service satisfaction. Overall, this study offers several theoretical implications for service hospitality organisations, drawing on the way the surveyed equity context performed in predicting customer citizenship behaviour. The results of this study also suggest specific new courses of action for managers in hotel organisations. Finally, the paper reveals several avenues for future research.

As this study finds, fair treatment of employees by the hotel can evoke reactions that lie outside conventional immediate dyads, such as the employee–employer dyad, where the employees themselves react to their own fair treatment. In this study the ‘performer’ is, however, the customer, who does not react to his or her own fair or unfair treatment, but instead follows a third-party perspective. Moreover, this study also shows the importance of satisfaction with hotel service as a mediator that explains hotel guests’ reactions to justice towards staff. Therefore, this study has gone beyond previous research that modelled satisfaction with hotel service through a habitual sequence of relationships where customer satisfaction is usually posited as a predictor of customer citizenship behaviour (e.g. Aherne et al., 2005; Bettencourt, 1997; Groth, 2005; Gruen et al., 2000). As the findings show, satisfaction with hotel service may partially mediate the link between justice towards staff and guest citizenship.

In addition, in spite of the numerous studies on responses to perceived justice, few of them have involved full performance constructs. Instead, guests' attitudes and behavioural intentions (such as plans to return to the hotel, word of mouth intentions about the hotel, intentions to repeat the hotel purchase, to name a few) dominate this literature. However, our study takes another step by finding support for the connection between justice towards staff and guests' reported performance in the form of citizenship behaviour. This finding can be highly relevant for service encounter performance. Episodes of injustice towards staff by the hotel are not always merely diluted in guests' intentions, but instead they impact their full performance, and managers must make the staff aware of this. Furthermore, the mechanism this study has supported to show how and why staff treatment judgments may affect customer citizenship provides hotel managers with tools to try to control the degree to which guests support the effective functioning of their hotel. As the results suggest, this mechanism is based on eliciting a feeling of service satisfaction that leads to guests' citizenship behaviours, so that hotel managers should prevent this mechanism from being disabled. To this end, they should pay careful attention to any incident, via any route, that could decrease satisfaction with hotel service and neutralise the positive climate created by justice towards staff. Some researchers, for example, have found that guests show signs of reduced satisfaction with hotel service when perceiving a negative image and ambience (Jani and Han, 2014). Based on the study results, hotel management's success at controlling and discouraging these negative perceptions that can contribute to guests' unhappiness may have an immediate impact on hotel performance.

Given that customer citizenship behaviour had only been found to be significantly associated with guests' perceived justice for the self (e.g. Yi and Gong, 2008), the present study is the first one to support this link based on a third-party perspective of perceived justice. These findings provide hotels with the main strategy for strengthening healthy employee reciprocations in the form of customer citizenship behaviour, which is, obviously to prevent unfavourable perceptions of justice towards staff. In the success of this preventive strategy, the visibility of events of justice towards staff seems to be critical (e.g. verbal disrespect towards staff, inconsiderate manners, unfair overload in the front office work, being unappreciative of the effort employees make when they do the work well, ignoring the stress employees cope with, to name a few), so that service areas where staff interact daily make up a dangerous window that should receive particular attention. Managers must discuss this issue with the immediate supervisors of staff. No episode of mistreatment of staff is innocuous, and even less so if it is also visible.

Addressing questions raised but not responded to in this study could certainly be a basis for future research. Since it varies across organisations, the impact that customer citizenship behaviour can have on the effective functioning of a hotel determines the interest in eliciting these behaviours within hotels. As discussed earlier, several arguments support some benefits of customer citizenship behaviour in the hotel organisations but there are many more. One is based on the expected favourable perceptions of quality service that customer citizenship behaviour may produce in other customers, which is a potential subject for future research. In addition, the guests’ physical proximity to the employee-victim is a factor that is likely to lead to higher identification with this employee-victim and can play an important role in exacerbating the results presented here. The example mentioned earlier about unfairly treated receptionists at the reception desk during the guests’ ‘check in’ could serve as an illustration of this possibility. Future research can focus on the impact of fair or unfair treatment towards other collectives (such as other customers, suppliers, etc.) on customer citizenship behaviour.

Other questions not responded in this study have to do with the nature and potential impact of customer citizenship, compared to conventional citizenship (OCB), on the effective functioning of the hotel industry, which is another possible topic for future research. As discussed earlier, in 1997, approaching Borman and Motowidlo's (1993) concept of contextual performance, Organ (1997: 95) changed his definition of conventional citizenship (OCB) to say that it is ‘performance that supports the social and psychological environment in which task performance takes place’. This really opens up a very interesting question because the ‘context’ can become a different sort of operating environment when we consider guests’ citizenship activities. What is the hotel's true operating context? Is it the space created by the service encounter that allows the exchange among its three actors, i.e. contact staff, guest, and hotel organisation (Bateson, 1985)? Does the impact of customer citizenship and conventional OCB on hotel performance significantly differ? Are both able to be defined as contextual performance that ultimately supports the technical core of a hotel?

As the technical core of hospitality organisations seems to deal with providing guests with high-quality service, the potential impact of customer citizenship and conventional OCB on the effective functioning of the hospitality industry could be postulated as being similar. The conventional form of contextual performance or extra-role performance (to use the Van Dyne et al. (1995) terminology) includes staff activities endorsing and protecting the hotel when something or somebody puts it at risk, and when directed at individuals, helping and cooperating with co-workers and customers (Brief and Motowidlo, 1986; George and Brief, 1992). Acting across the organisational, social, and psychological context of the hotel, guests who perform customer citizenship should support the technical core of a hotel as well. Therefore, customer citizenship seems to reinforce the impact of conventional OCB on hotel effectiveness, thus playing a complementary citizenship-based role in hospitality organisations. Due to customer citizenship, guests support improvements or show concern for the effective functioning of the hotel by engaging in contextual performance activities (Borman and Motowidlo, 1997) that lubricate the hotel's organisational, social, and psychological context, ultimately supporting its technical core. However, this discussion only points to a new previously unexplored avenue for future research.

Finally, we acknowledge that this study has several weaknesses. First, we used a cross-sectional methodology, increasing the likelihood that the study could suffer from mono-method/source bias. As Stone-Romero and Rosopa (2008: 349) suggest, there is nothing wrong with talking about an assumed causal model in an article. However, they also state that it must be clearly acknowledged ‘that (a) a number of other causal models are also consistent with the same pattern of covariances, and (b) their findings do not provide a valid basis for making causal inferences about relations between or among variables’. Moreover, the employees in the study have certain job conditions and norms that are often inherent to the peculiarities of workers in the hospitality sector. Consequently, the performance of the constructs used in our study, as well as their implications, could vary in other institutions in other industry contexts.

In conclusion, the lack of attention to third-party reactions to (in)justice in studies in the hospitality industry has yielded patterns that do not sufficiently explain the role of guests' perceptions of (in)justice towards staff in their responses to (in)justice, particularly those that help the success of a hotel. Guests in a hotel often act as uninvolved third parties who witness justice-related events in other guests' service encounters. If they perceive signs of justice towards staff, the results suggest that they are likely to take action in the form of customer citizenship. Moreover, the results support the mediating role of satisfaction with hotel service in this relationship. This enables managers to communicate to guests that, by engaging in fair treatment of the staff, they are making an effort to provide them with a satisfactory stay in the hotel and, thus, are deserving of their help.