Behaviors also Trickle Back: An Assessment of Customer Dysfunctional Behavior on Employees and Customers

Abstract

This study examined the trickle in, out, around and trickle back effect of dysfunctional customer behavior on employees and consequently employees’ incivility and service recovery efforts toward customers. Furthermore, this study has specifically tested the mediating effect of employee burnout to examine the trickle around and trickle back effect. To explore the multi-level trickle effect, this study has collected data from two sources, i.e., customers and employees. The data was analyzed with the help of AMOS. The results revealed that customer’s verbal aggression escalates employee’s burnout, which in turn affects employee’s incivility towards customers. However, the indirect paths from disproportionate customer demand toward service recovery efforts and employee’s incivility towards customers were found to be insignificant. This study addressed the existing gap in the literature by examining the trickle effect within and outside the boundaries of an organization. The results of this study laid down some useful managerial and theoretical implications.

1. Introduction

The negative interaction between customers and employees does not only create a contagious effect in a single episode, rather it has a larger impact on the upcoming events. The flow of positive or negative interactions from the transmitter has a significant effect on the behavior or attitude of the receiver. The receiver further transmits the positive or negative interaction towards the second receiver, which is termed as the trickle effect in the organizational psychology literature [1]. The asymmetric power structure causes the trickle-down effect in case of negative interaction where the victims cannot fight back. However, the victim finds some easier target to vent his/her aggression immediately or sometimes later [2]. If the harmed receiver does not find a soft target, he/she will vent aggression in the form of lowering the quality of interaction either by avoiding or minimizing the interaction [1]. The situation is different from reciprocity, where the recipient has the power to reciprocate the negative behavior in the form of fighting back by showing similar behavior [3]. In case of organizational hierarchy, the trickle effect usually travels from top–down, hence termed the trickle-down effect [1,4]. However, the flow of consequences can be more dynamic instead of only trickling down; it can travel up, in or out [2,5]. In such cases, the trickle effect crosses the organizational boundaries and flows in [6,7] or out [8] of an organizational setting. The trickle effect also occurs horizontally in the organizational hierarchy, when feeling, attitude or behavior affect people at the same level [2].

The extant literature shows that research has mainly focused on organizational level trickle effect (trickle-down or trickle out) [5] and ignores the trickle effect, which takes place between customers and employees of an organization. Customers affect organizational employees, especially, frontline employees who spend most of their time interacting with customers [9]. Hence, it is not surprising to observe the trickle effect taking place between the customers and employees of an organization [6]. Although infrequently reported [10,11], often customers harm employees by their words, actions, or gestures [12]. In service setups, frontline employees are more prone to such type of negative interactions [9,13]. Frontline employees are the face of the organization in a service set up as they set the mind of customers about the service quality [14,15,16,17]. They are expected to behave in-line with organizational norms [18], not only to build a positive image of the organization but also for their own physical and psychological resource gain. Such a strenuous situation affects employee perception about the job, profession and even about his/her efforts to recover [19,20]. Direct social influence has been widely discussed in past literature [21,22,23,24,25]. Nonetheless, how the feeling of one actor outside the organizational boundaries (customer) affects people in the organization (employee), which in turn, is passed on to other people outside the organization (customer, family), is still unexplored [5]. Very few studies discussed the counter behavior towards customers and coworkers [11,26]. Moreover, these studies relied upon single-source data [5], i.e., data for all the three levels of trickle effect were based on the perception of a single person. Wo [6] noted that the negative effects cross the limits of the direct or reciprocal relationship and spread to many levels. Trickle effect or indirect social influence encircles the flow of feeling, perception, attitude and behavior from source to transmitter and through him/her to other people within (down, up, horizontal) [1,2,27] or outside (in or out) the organization [6,28]. Dependence of these studies on single-source data created a gap in the literature and provides an opportunity to discover the phenomenon of trickle effect with multi-source data [5].

Building on this notion, this study contributes to the literature by addressing the research call of Wo, Schminke [5] to study the “trickle around effects” that is the flow of single, negative or positive effect across the organizational boundaries. Secondly, this study addressed another research call from the same study by investigating whether the same action, the feeling, can trickle in multiple directions. Hence, this study investigates another dimension of the trickle effect in which the effect flows from originator outside the organizational boundaries (customer), affecting an employee within the organizational setting and then, as a consequence, affecting other customers. An important characteristic of the trickle effect is that the affected carrier spreads its effect with others while interacting with them, and the effect is indirect in nature [2]. Thirdly, the study tried to fill the gap in the literature by addressing the affective aspect of human interaction in the trickle effect model. According to [5], there is a dearth of research to study the effective construct in trickle effect literature. Lastly, this study takes into account the experience of female working employees, seeing their actions and interaction in the male dominated society of a developing country. Hence to address the above-mentioned research gaps, this study takes into account customer verbal aggression and disproportionate customer demands to affect employee psychological wellbeing (burnout), that in turn affect both other customers (in the form of employee incivility towards customers) and the frontline employees (service recovery efforts).

2. Literature Review

Experiences of negative feelings, attitudes or behaviors do not work in a vacuum; rather, they have a long-lasting impact on the transmitter along with the person who receives them and passively go to those who are not in the immediate circle of the originator due to social influence, individual’s actions, perceptions and attitudes or being influenced by others [29]. However, when the influence is directed on or through a third party, this is known as an indirect social influence [1]. In the latter case, the doer may or may not have a direct link with the third party, influenced by transmitter who is affected by the actions, feelings or attitudes of the primary doer [5]. This phenomenon spreads faster in case of negative behavior [2]. In an organizational setup, the feelings, attitudes or behaviors transmit from top to bottom due to power asymmetry [2,30]. In such cases, the recipient vents his/her aggression on someone else due to power disparity. They do so due to the actual or perceptual high cost of reciprocal direct aggression [31]. Another form of trickle effect is the negative effect, which transmits outside the organizational boundaries and vents on others like friends and family members or customers [32]. The transmission of negativity is stronger for people with hostile traits [33]. All of these trickle effects (trickle down, trickle in, trickle out, trickle around) have one common characteristic, either the originator or the end victim belongs to the organization. In a special case, the originator from outside the organizational boundaries (customer) affects people within the organization (employee). This indirect social influence brings the effect back to similar agent (other customers) [34]. We can term this effect as “trickle back effect”, a combination of trickle in, trickle out and trickle around effect. Frontline employees in a service setting face dysfunctional customer behavior [12], which affects their emotional wellbeing [24], and that in turn affects not only the employee’s attitude [35] but also customer’s wellbeing in the form of low-quality service recovery [36].

Deliberate violation of service interaction norms by inappropriately treating service employees is termed as dysfunctional customer behavior [10,37]. This behavior is synonymous with various names in literature such as “customers from hell” [38], jay customers [39], deviant customer behavior [40]. Such behavior is described as customer incivility (that is low-intensity rude behavior from the customer with ambiguous intent to harm service employees [41] and physical aggression (intended to physically harm employees [42]. In this study, we have specified dysfunctional customer behavior with the help of two dimensions, i.e., disproportionate customer demands and customer verbal aggression [43]. As these behaviors are directly linked to employee behavior alongside the customers, their illegitimate complaints may or may not harm employees for them to react towards customers [43].

Dysfunctional customer behavior leads toward burnout in the employees of the organization. Burnout is defined as a syndrome specific to people interaction [44] and considered as a prolonged reaction to perpetual interpersonal and emotional stress on the job [45]. It is further explained as a combination of emotion depletion (emotional exhaustion), distance and callous feeling about people (depersonalization) and doubts about one’s abilities in terms of accomplishment (personal accomplishment) [46].

Frontline employees in the service sector, e.g., flight attendants and railway crew, are frequently exposed to dysfunctional behavior [13,20,47,48] for longer periods. This type of interaction creates emotion depletion in frontline employees [49]. According to the conservation of resources (COR) theory [50], persistent exposure to such negative behavior and regular depletion of emotions ends up at the high level of stress [48]. The extant literature proves that dysfunctional customer behavior is related to depleted emotional states [48,51,52,53], hence creating burnout in employees [44]. This state of burnout further affects the performance of employees negatively [23,54]. Song and Liu [55] studied call center employees and found that verbal customer aggression is correlated with employee burnout. Subsequently, Karatepe [56] explored that motional dissonance escalates burnout and disengagement among frontline hotel employees in Ankara, Turkey. The literature further suggests that masking true emotions (either with the help of surface acting or deep acting) causes to deplete the energy and raise disengagement [57]. Undesirable customer behavior depletes employee emotional resources [58], which leave them with little or no resource and make them callous towards others [53,59]. These arguments provide a ground to conclude that customer aggression is a threat to employee emotional wellbeing [41] and a source of employee burnout. Based on the above discussion, we can hypothesize that,

Hypothesis 1 (H1). 

An employee frequently facing deviant behavior in the form of disproportionate customer demands feels a higher level of burnout.

Hypothesis 2 (H2). 

An employee frequently facing deviant behavior in the form of customer verbal aggression feels a higher level of burnout.

As service employees cannot stay away from customers and their jobs [9], they provide required services at the cost of their emotional labor [60,61]. Such actions taken to recuperate the service and customer satisfaction is termed as service recovery, and employee’s ability to recover the service is measured as recovery effort [62]. These are corrective actions taken by employees to restore customer confidence as a result of service failure [63]. However, service recovery efforts leave the employee with little emotional and social resources, which spoil the quality and quantity of service delivery [45]. Tackling the dysfunctional customer by surface or deep acting depletes employee energy, and they are left with little energy to battle with the excessive and disproportionate demands of the customers, which affects their abilities to control and recover their performance accordingly [53,64]. In a sample of 170 call center employees in New Zealand, Rod and Ashill [65] found a negative link between emotional exhaustion and employee service recovery ability. Similarly, Essawy [66] and Sommovigo and Setti [36] found a negative association of burnout with service recovery. Many other studies, like Choi and Kim [23] and Karatepe and Choubtarash [64], in the service sector reported a negative relationship between burnout and service recovery to the same and other customers. Based on the above discussion, we can hypothesize that

Hypothesis 3 (H3). 

Burnout has a negative relationship with the service recovery efforts of frontline employees.

An employee with depleted emotions and heightened work demands feels disengaged not only about his/her job but also exhibits a distant behavior towards others. The distant behavior might be a coping strategy by an employee to halt further harm to the service interactions. In other cases, due to absence of cognitive resources and emotional depletion, the employee finds himself/herself unable to meet the service demands of quality and exhibition of normative behavior [7,26,45]. Eventually, the employee exhibits deviant behavior towards customers [11,67]. Usually, people reciprocate counterproductive work behavior towards the source [67,68], but due to power asymmetry and intensified emotions of the source, they find some soft target to vent on [2,69]. Based on the above discussion we can hypothesize that

Hypothesis 4 (H4). 

Employees who feel burnout exhibit incivility towards other customers.

Research talked about incivility spiral [70] or target similarity model [68], where the target of mistreatment reciprocates with the same behavior [71]. However, in the trickle effect, the negative behavior may or may not be reciprocated to the offender but be rather transferred to someone lower in the power hierarchy [30], which may be either family [72], a coworker [73] or even some other customer [11]. In the displaced aggression process [74], an employee with depleted emotion is unable to retaliate towards the source of mistreatment not only due to power asymmetry [75] but also due to heightened emotions of the offender. Interaction with dysfunctional customers raises the concern about interactional justice, depleted emotion and gives rise to distant behavior towards others [45]. However, the response comes in mild form incivility with no intention of harming the target [33]. In response to customer verbal aggression and disproportionate demands, an employee vents out his/her aggression on some other customers by showing a mild form of deviant behavior [76] and shows incivility that is ambiguous in terms of harm towards the target [33]. Empirical research also found the mediating role of burnout between the relationship of customer dysfunctional behavior and employee deviant behavior with customers [11,26].

According to the conservation of resources theory [77], employees lose their resources in negative service interactions with customer aggressions and disproportionate demands. Loss of emotional resources (emotional exhaustion) leads the employee to distanced behavior with people (depersonalization) [45]. The disengaged behavior further depletes the employee’s emotions and leads to resource loss spiral [78]. However, the power disparity causes the venting out the aggression in some mild form on others [74] who might have no connection with the matter [76]. Thus, the service employee does not follow organizational norms and expresses incivility with customers [11,26,71,79]. Empirical research has also found a tit-for-tat response from customer service employees; however, these studies have only considered reciprocal responses from the employee towards the same customers [80]. For example, Kim [26] found that employee burnout mediates the positive association between employee-customer incivility and employee incivility towards customers as well as a coworker. In another study, Kim and Qu [11] observed a significant indirect relationship between customer and employee incivility through burnout. Hence, we can hypothesize that

Hypothesis 5a (H5a). 

Burnout mediates the relationship between customer verbal aggression and frontline employee incivility towards customers.

Hypothesis 5b (H5b). 

Burnout mediates the relationship between disproportionate customer demands and frontline employee incivility towards customers.

Masking one’s emotions (Emotional dissonance) not only results in immediate effects but leads to the employee feeling exhausted and negatively perceiving his/her abilities [81]. This can lead to the inability to recover the situation [65] or attempting to escape such situation by quitting [23,82]. This is because the employee loses all or most of his/her emotional resources in the masking process and is left with little resources to curb the demanding situation of service recovery [53,74]. Hence, the employee behaves more negatively by holding customer’s demands [83]. The literature further emphasizes that customer deviant behavior in the form disproportionate expectations and customer verbal aggression, which are associated with employee psychological wellbeing “burnout” and frontline employee’ burnout should be linked with the recovery service effort they made. Molino and Emanuel [61] and Zito and Emanuel [22] found that frontline employees’ emotional state is linked with the treatment they receive from customers and affects their wellbeing (burnout). Furthermore, a service employee can only provide quality services when he/she is treated with care. Moreover, Sommovigo and Setti [36] found the employee emotional state as an explanatory link between customer treatment and negative behavior towards customers in the form of poor services they receive. According to COR theory [50,78], when the interactions with customers engulf their resources, the employees avoid such interactions by reducing the quality of service, hence causing the customers to suffer. Interaction, that is considered as unjust in terms of treatment by customers is considered as a stimulus to deplete the energies [12,51] that make employees disengaged not only towards those customers but also to other people [45], and this disengagement takes the form of indirect reciprocation [6]. When the confrontation with such situation is persistent or comes from multiple actors, it leads to threatening the work environment [84]. Moreover, people link the situation with the context when they received such treatment before. If found congruent, they perceive the situation as similar to their previous experience [2]. Hence, the perception of negative interaction is stronger for those who have received such treatment before. However, the response to such behavior is not constant but rather different at various times. One reason for changing response may be due to the power disparity of frontline employees [2]. As a result, when employees receive unjust treatment either from supervisors or customers escalate their negative emotions [12,18], the balance requires depletion in performance [28]. On occasions, service employees bypass organizational norms and reply with the same tone as received from customers [41,70,85]. However, in the majority of the cases, due to asymmetrical power state of employees, they cannot retaliate due to organizational norms [18,86] but behave carelessly and exhibit less motivation for job performance or service recovery [28,45]. Moreover, when customers interact with the depersonalized employee, they perceive service negatively [59], and other customers also count this effect [61,87]. Based on the above arguments, we predict the following relationships:

Hypothesis 6a (H6a). 

Burnout mediates the relationship between customer verbal aggression and service recovery efforts of frontline employees.

Hypothesis 6b (H6b). 

Burnout mediates the relationship between disproportionate customer demands and service recovery efforts of frontline employees.

The research model is showed in Figure 1.

3. Methodology

3.1. Data Collection and Samples

The data was collected from 153 customer-employee dyads from beauty salons and the transportation industry. Beauty salon business is at a peak in Pakistan, and it is one of the most liked professions by female entrepreneurs [88,89]. Similarly, bus hostesses play a key role in maintaining the quality of services at luxury bus services in Pakistan [52]. Due to the high customer-employee interaction and its effect on overall service delivery, these service businesses provide an ample opportunity to study the customer dysfunctional behavior, employee burnout, employee service recovery efforts and employee incivility towards customers. Furthermore, some recent incidents of customer vandalism in these service businesses initiates a need to collect some empirical evidence from these businesses.

For data collection, 120 questionnaires were distributed among the bus hostesses of 10 companies through research enumerators using convenient sampling. In the meantime, the customers were approached after they received the services from their respective bus hostesses. Once the responses by customer and service providers matched, dyads were made. A total of 90 responses were received; however, 9 were discarded due to incomplete data, yielding 81 valid responses. In case of beauticians, the enumerators visited 25 beauty salons and collected data as mentioned above using convenient sampling. A total of 120 questionnaires were distributed, and 95 dyads were matched. However, after removing incomplete responses, 72 valid questionnaires were used for analysis. Since the dyad design was used for data collection from two sources, i.e., employees and customers, the data regarding customer dysfunctional behavior (disproportionate customer demands, customer verbal aggression) and feeling of burnout were collected from female beauticians and bus hostesses working in Pakistan. However, the data related to employees’ incivility toward customers and service recovery efforts of the employee were collected from customers who took services from the respective beauty salons and transportation companies. Besides, the data on service recovery efforts and incivility (in case it happened) of the frontline employee were collected from customers receiving service at the time of data collection to make employee-customer dyads.

3.2. Measurement of Variables

The questionnaire was designed by adopting the measures from previous studies. Disproportionate customer demands and customer verbal aggression were measured with four items each taken from [43]. Sample items are, “customers demanded special treatment” and “customers yelled at me”. Both the scales are sufficiently reliable with alpha value of (α = 0.95). Employee burnout was measured with the help of six items by using the short form of Bhanugopan and Fish [90] with as high reliability (α = 0.87). Sample item includes “I feel emotionally drained from my work”. Service recovery efforts of the employee were operationalized by customer rated three item scale from Jeong [91]. An example of the items is “I am satisfied with the employee solving the problem”. The scale of employee incivility towards customer was measured by the six-item scale of van Jaarsveld and Walker [71] with alpha value of (0.77). Sample item includes “(s)he was derogatory to customers”. The wording of items for customer rated scale was modified to fit customer rating. A five-point Likert scale from strongly disagree to strongly agree was used to measure all the items. The measurement items are given in Appendix A.

4. Results

To analyze the hypothesized model, including validity and reliability assessment, we used AMOS and process macro by Hayes [92].

Table 1. Descriptive statistics and correlations.

	Mean	SD	1	2	3	4	5	6	7	8	9
1. Age	2.86	0.704	1
2. M_stat	1.39	0.489	0.569 **	1
3. Edu	1.99	0.773	0.385 **	0.406 **	1
4. Exp	2.51	1.06	0.565 **	0.303 **	0.204 *	1
5. CVA	3.69	0.813	0.111	0.162 *	0.080	0.149	1
6. DCD	2.94	0.992	−0.124	−0.057	0.042	0.069	−0.072	1
7. EINC	2.77	1.14	0.180 *	0.137	0.111	0.012	0.338 **	−0.100	1
8. BO	3.64	1.00	0.273 **	0.135	0.052	0.143	0.507 **	−0.050	0.298 **	1
9. SRE	2.80	1.04	0.083	−0.027	−0.023	0.071	0.013	−0.250 **	0.103	−0.056	1

Notes: n = 153. CVA: Customer Verbal Aggression, DCD: Disproportionate Customer Demands, SRE: Service Recovery Efforts, BO: Burnout, EINC: Employee Incivility towards Customer. Marital status was coded as 1 = Single and 2 = Married. ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

exhibits correlational and descriptive statistics. After correlational analysis, we proceeded to psychometric properties assessment.

Table 2. Reliability and validity assessment.

	CR	AVE	1	2	3	4	5
1. DCD	0.846	0.580	0.761
2. CVA	0.790	0.488	−0.129	0.698
3. EINC	0.911	0.671	−0.135	0.390	0.819
4. SRE	0.858	0.668	0.292	−0.001	−0.095	0.817
5. BO	0.899	0.603	−0.062	0.605	0.315	0.072	0.777

Note: CVA: Customer Verbal Abuse, DCD: Disproportionate Customer Demands, SRE: Service Recovery Efforts, BO: Burnout, EINC: Employee Incivility towards Customer.

exhibits reliability and validity assessments. Composite reliability (CR) was used for reliability assessment. Average variance extracted (AVE) served the purpose of convergent validity [93] and the square root of AVE in comparison with inter construct correlation provided the evidence of discriminant validity [94]. For this purpose, we ran four types of models with one, two, three, four and five factor solution of the data [95]. Moreover, this assessment serves the purpose of common method variance analysis. As

Table 3. Model comparison.

Models	Chi Square (χ2)/df	CFI	TLI	SRMR	RMSEA
One Factor Model	1256.5/209	00.40	0.34	0.28	0.18
Two Factor Model	1165.2/188	0.46	0.40	0.28	0.17
Three Factor Model	727.9/206	0.60	0.54	0.19	0.13
Four Factor Model	521.3/203	0.82	0.80	0.16	0.10
* Five Factor Model	273.76/199	0.95	0.95	0.08	0.05

CFI: Comparative fit index, TLI: Tucker–Lewis index, SRMR: Standardized root mean square residual RMSEA: Root mean square error of approximation, (χ²)/df: Chi-square degree of freedom. * Model used in the study.

exhibits, one factor solution depicts a worse model fitness as compared to other models with a better fit and, ultimately, five factors model (χ² = 273.76, df = 199, CF = 0.95, TLI = 0.95, RMSEA = 0.05, RMR = 0.08) comes up with the best fit among the four models compared [96].

Values in bold on diagonals show square root of AVE, and for discriminant validity, it should be above inter-construct correlations [94]

4.1. Common Method Variance

Although, data were collected from different sources, i.e., customer-employee dyads, still the chance of common method variance exists [67]. To address this issue, we utilized confirmatory factor analysis (CFA) with AMOS to analyze fit indices, i.e., comparative fit index (CFI), Tucker–Lewis index (TLI), standardized root mean square residual (SRMR), and root mean square error of approximation (RMSEA) [97]. In the first step, we ran one, two, three, four and five factor models. Our proposed five factor model, i.e., (customer verbal aggression, disproportionate customer demands, employee burnout, service recovery efforts and employee incivility) yielded good fit of the data as shown in Table 3 (χ² = 273.76, df = 199, CFI = 0.95, TLI = 0.95, SRMR = 0.08, RMSEA = 0.05). Comparison of the five factor model fit indices with one (χ² = 1256.5, df = 209, CFI = 0.40, TLI = 0.34, SRMR = 0.28, RMSEA = 0.18), two factor (χ² = 1165.2, df = 188, CFI = 0.46, TLI = 0.40, SRMR = 0.28, RMSEA = 0.17) three factor model (χ² = 727.9, df = 206, CFI = 0.60, TLI = 0.54, SRMR = 0.19, RMSEA = 0.13) and four factor models (χ2 = 521.3, df = 203, CFI = 0.82, TLI = 0.80, SRMR = 0.16, RMSEA = 0.10) indices confirmed that five factor model produced the best fit. This comparison reduced the potential of common method bias and suggests our measures as discrete.

4.2. Hypotheses Testing

Direct relationships were measured by regression analysis. The results of regression analysis including standard errors, t-values, and confidence interval are reported in

Table 4. Test of hypotheses.

Hypotheses	Relationship	Beta	SE	t-Value	Confidence Interval		R−Square	Supported/ Not Supported
					LL CI	UL CI
H1	DCD→BO	−0.014	0.07	−0.169	−0.15	0.12	0.257	Not supported
H2	CVA→BO	0.626	0.08	7.16	0.453	0.799		Supported
H3	BO→ SRP	−0.059	0.08	−0.695	−0.226	0.108	0.03	Not Supported
H4	BO→ EINC	0.340	0.08	3.84	0.165	0.514	0.09	Supported

.

Descriptive statistics in Table 1 shows no significant relationship between demographic variables (except age) and the constructs in the study. Hence, these were not included in the regression analysis. In the first instance, we checked for the direct relationships between customer verbal aggression and burnout (β = 0.626, se: 0.08 p < 0.05, [0.453; 0.799]). Results support (Table 3) significant positive association between the two variables. However, results show an insignificant relationship between disproportionate customer demands and employee burnout (β = −0.014, p > 0.05, [−0.15; 0.12]). Disproportionate customer demands and customer verbal aggression collectively explain 26% variance in burnout (R-square = 0.257). Furthermore, the results support a significant positive relationship between employee burnout and incivility towards customer (β = 0.34, se: 0.08 p < 0.05, [0.165; 0.514]). Subsequently, burnout explains almost 9% variance in employee incivility towards customers. Meanwhile, the result shows an insignificant relationship between employee burnout and customer-rated service recovery effort of employee (β = −0.059, se: 0.08 p > 0.05, [−0.226; 0.108]). Moreover, employee burnout explains only about 3% variance in service recovery efforts of the employee.

4.3. Mediation Analysis

This study used Process Macro (model four) developed by Hayes [92] with the bootstrapping technique to analyze the mediation hypotheses by comparing the model significance. This technique accessed mediation effect by observing the significance of direct and indirect paths from confidence intervals (CI). If the CI includes zero, the effect is considered as insignificant and vice versa. If the indirect path (from IV to mediator and from mediator to DV) is significant, it supports the mediation effect [92]. Furthermore, in the case of significant effect of direct and indirect paths, partial mediation occurs [98]. In our case, the indirect effect of customer verbal aggression on employee incivility towards customer via burnout is significant due to the absence of zero between upper and lower confidence intervals at 95% (see

Table 5. Test of mediation.

Hypotheses	Paths	Co-Eff	SE	t-Value	p-Value	95% Confidence Interval		Supported/ Not Supported
						LL	UL
H5a	Indirect Effect CVA on EINC	0.122	0.06			0.011	0.252	Supported
	Direct Effect CVA on EINC	0.354	0.12	2.85	0.005	0.109	0.599
	Total Effect CVA on EINC	0.476	0.10	4.41	0.00	0.263	0.690
H5b	Indirect effect of DCD on EINC (axb)	−0.02	0.02	-	-	−0.078	0.033	Not Supported
	Direct effect DCD on EINC (c)	−0.09	0.09	−1.10	0.271	−0.276	0.078
	Total effect DCD on EINC	−0.116	0.09	−1.24	0.217	−0.301	0.069
H6a	Indirect effect of CVA on SRP (axb)	−0.05	0.06	-	-	−0.193	0.060	Not Supported
	Direct effect CVA on SRP (c)	0.073	0.12	0.598	0.551	−0.168	0.313
	Total effect CVA on SRP	0.017	0.10	0.164	0.84	−0.190	0.224
H6b	Indirect effect of DCD on SRP (axb)	−0.004	0.01			−0.014	0.027	Not Supported
	Direct effect DCD-SRP(c)	−0.268	0.08	−3.21	0.002	−0.433	−0.103
	Total effect DCD on SRP	−0.264	0.08	−3.17	0.002	−0.429	−0.100

). Moreover, the direct effect of customer verbal aggression on employee incivility towards the customer is also significant (having no zero in 95% CI), proving partial mediation. Besides this indirect path, all mediation paths were found to be insignificant.

5. Discussion

The trickle effect literature mainly examines the flow of perception, feeling and behavior down the organizational hierarchy or in a relationship where organizational agents (manager, or employee) are present. The investigation of trickle effect outside the boundaries of an organization and at the horizontal level is missing from the extant literature. Hence, the model presented in this study tried to discuss rarely investigated issue where the customer’s attitude and behavior trickle in, change the attitude of an organizational employee and turn back to customers in the form of employee behavior. The results supported three hypotheses while the rest of the hypotheses were not supported. According to the result, H2 is supported where customer’s dysfunctional behavior, in the form of customer verbal aggression, is significantly associated with the perception of employee wellbeing (burnout). Similarly, results supported H4 which stated that employee burnout leads to employee incivility towards customers. However, the results do not support H1, which states that disproportionate customer demands lead to employee burnout. Similarly, H3 is not supported, which states that burnout negatively leads to the service recovery effort. Among mediating hypotheses, H5a is supported, which states that burnout mediates the relationship between customer verbal aggression and service recovery efforts. However, this study does not support the mediating effect of burnout between disproportionate customer demands and service recovery efforts (H5b). Subsequently, burnout does not mediate the relationship between disproportionate customer demands and employee incivility towards customers (H6a). Likewise, burnout does not mediate the relationship between verbal customer aggression and employee incivility towards customers (H6b). Some of the results of this study are consistent with past research; however, some of them are anomalous with previous researches. For example, a positive association between customer verbal aggression and employee feelings of burnout is confirmed by previous findings [12,41,99,100]. However, the insignificant relationship between customer disproportionate demands with employee feeling of burnout is anomalous in relation to previous studies [23,24,35]. The reason might be the coping strategies or information processing route taken by service provider’s employees (cognitive or affective route) whereby the employees either think about customer demands (cognitive) or overly react (affective) to compensate the customers [5]. Direct association between employee burnout and employee incivility towards the customer is consistent with previous studies [11,26]. However, the insignificant relationship of burnout with service recovery efforts contradicts previous studies [23,36,67]. The mediating impact of employee burnout between the relationship of customer verbal aggression and employee incivility towards the customer is consistent with the previous literature [11]. However, the insignificant mediating role of burnout in other relationships (DCD-SRP, CVA-SRP and DCD-EINC) is quite surprising in light of previous research [35,36,101]. Customer verbal aggression might lead to employee incivility towards the customer in Pakistan due to gender discrimination of female employees and their instant reaction as protest or revenge in the male dominated society [102]. Subsequently, the insignificant relationship between deprotonate customer demands and employee burnout may be due to willful acceptance of marketing mantra, which advocates that customer is always right. Another reason for this happening is the individual power distance between customers and employees. It may also happen due to the gender disparity in a male dominated society like Pakistan where the female employees feel dominated by males. Lastly, it can happen due to the nature of the industry where the employees are working, e.g., the hospitality industry where the employees are expected to provide quality services even if the interaction is not fruitful.

By exploring the customer dysfunctional behavior and its trickle in, out and around effect, this study has made some useful contributions to the discipline of human resources and marketing management. Secondly, this study has filled a gap in the literature by responding to the research call of Wo and Schminke [5] to explore the trickle effect in multiple directions, from customers to employees and then back to customers. This study is one of its kind in the developing society’s service context, which addressed the multilevel trickle effect of behaviors by utilizing the multi-rater data. Hence, this study extends the theoretical rigor of trickle effect in the service sector of a developing country where the female frontline employees interact with customers.

5.1. Theoretical Implications

The findings of this study provide several important theoretical implications. First, this study is the first attempt to measure the impact of customer deviant behavior on other customers involved in the deviant interaction. As customer–employee interaction does not occur in a vacuum, the interaction crosses the boundaries of the direct social cap [5]. Research mainly focuses on the interaction between employee and focal customer while ignoring other customers outside the deviant interaction frame [87]. Hence, the result of this study is based on the effect of direct interaction on a third party (other customers). It highlights the importance of managing the direct interaction but also its effect on other customers (trickle effect). Secondly, the study corroborates previous studies that highlight customer misbehavior (customer aggression) as the strongest determinant of employee burnout [11,100]. Thirdly, significant indirect relationship CVA-EINC and insignificant CVA-SRP through BO confirm the proposition by Wo [5] that employees use two paths for assessment of customer interactions: one through the cognitive process and the other through the effective process, and that the reaction of the affective path is more severe and instant as compared to reaction through the cognitive path. Moreover, female employees react instantly after finding something like discrimination or a challenge to their gender role [102]. Fourthly, data for the study were collected from bus hostesses and female beauticians and their customer/passengers in Pakistan. The nation is still comparatively marginally represented in hospitality literature, especially the female population. Hence, this empirical study frames the emotional state of women in the service sector of a male dominated society of a developing country.

5.2. Practical Implications

This study offers some practical implications for the management of service organizations in general and women staffed organizations in particular. Firstly, this study spotlights the impact of customer deviant behavior on the employee, which is rarely reported and addressed in developing countries. Hence, the organizations should address these issues and device some policies to secure the employees against such interactions. Second, as the study found, employees respond instantly to intentional forms of deviant customer behavior. Hence, management should create awareness among employees, through mentoring and training, about such occurrence to mentally prepare them for such incidents [100]. Third, at the customer end, management should try to find out reasons why their customer behaves in such a way to operate remedial actions. Although the indirect link between disproportionate customer demands and employee reaction is insignificant, management should clarify rules and organizational policies not only for the employee but for the customers as well, through social media or instruction boards in waiting rooms or informative videos during services. In the future, female employees might raise their perception of gender discrimination [102]. Lastly, keeping in view the low participation of females in the national economy, this study highlights the importance of training of employees in services sectors that will enhance their skills and competencies. This will help them provide exceptional services to customers, resulting in fewer complaints and less burnout [11].

5.3. Limitation and Future Research Directions

Although this study makes some significant contributions to the existing literature of organizational behavior, like other human efforts, it is not free of limitations. Though the data were collected from two sources, the cross-sectional nature of data restricts us to infer causal relationships. Future research may use the temporal or multi-wave longitudinal study to analyze the change in effect as a result of the change in behavior over time. Secondly, the generalization of results outside this research should be used cautiously due to the emphasis on bus hostesses and beauticians only. This might raise concerns about the generalizability of results over other sectors [11]. Besides, future studies can relate to certain sectors only, e.g., bus services or beauty salons separately to understand some other constructs and interactions that are specific to these industries, to have more accurate results to infer generalizability. Moreover, future studies may use multi-sector and multi-cultural data to check and verify the results to generalize it across all services sectors. Individual striving for resource conservation, positive psychological states (i.e., psychological capital) or expectations about services (psychological contract) will buffer [103] or intensify [104] action by and reaction to both employees and customers. Future studies should consider the individual or multiple boundary conditions to see the interaction effect on the above relationships. Future studies may use single or multiple moderators to explore these boundary conditions. Observation by other customers of customer incivility and employee response also paves the way for how customers think to react in service interactions and may come up as social support from customer’s side [87]. Future studies should also incorporate multi-rater assessment of incivility and response of employees to see the real spread of trickle effect in various directions. Future studies may also check for the transfer of trickle effect from work to family. It would also be an interesting study to see the interacting effect of genders of employees as well as the customer to see how employees perceive the deviant behavior from same and opposite genders. Lastly, the data were collected using non-probability sampling technique, i.e., convenient sampling, which could be considered as another limitation of this study.