How workers' emotional dissonance explains the association between customers' relations, burnout and health in an Italian supermarket chain.

INTRODUCTION: Dealing with customers has increasingly became a risk factor that organisations should be aware of. Having to deal with difficult customers, pupils, patients it has been found to be the second most frequent reported risk factor in the recent 2019 ESENER survey (61%, up from 58%), just after repetitive hand or arm movements.
OBJECTIVES: The aim of the present study is to explore the relationship between negative relations with customers and its association with employees' burnout and general health.
METHODS: Data were collected using an anonymous, self-report questionnaire employed in an Italian company working in the large-scale retail sector. The sample was composed of 610 employees (70% women) working in 28 supermarkets.
RESULTS: Results showed that negative customer relations were associated to higher levels of emotional dissonance (b = .25; p = .000), which in turn was associated with higher levels of burnout (b = .72; p = .017) and consequently to higher levels of poor general health (b = .03; p = .000).
CONCLUSION: This study contributed to the current understanding of how negative relationship with clients could trigger a health impairment process which could lead to higher levels of burnout and poorer general health in employees. Findings suggest that organizations whose employees' work activities are characterized by a daily contact with customers should implement interventions such as training activities in order to enhance the employees' skills in dealing with custumers, for example on how to manage emotions at work.

Introduction

Dealing with customers has increasingly became a risk factor that organisations should be aware of. The recent Third European Survey of Enterprises on New and Emerging Risks highlighted that having to deal with difficult customers, pupils or patients is the second most frequent reported risk factor (61%, up from 58%), just after repetitive hand or arm movements (1). Moreover, in the previous survey (last detailed data available), 11% of the respondents of the European Union citizens answered that their jobs involved handling angry customers almost all the time and the percentage increased when considering the Italian sample (21%) (2). Despite research showed that it is not the mere interactions with customers that could impair the employees’ health (3), negative interactions such as dealing with angry or aggressive customers could lead to negative effects.

Among the in-role behaviors expected from people working with customers, there are treating customers with affability and kindness, smile and in general to display positive emotions and inhibit the negative ones, regardless of the customers’ actions (4, 5). In fact, working in service industries often means a daily interaction with customers, which sometimes could lead also to customer incivility, which is a significant problem for the service industry (6) and has been shown to be related to negative health outcomes on the employees’ such as psychological strain and emotional exhaustion (7, 8). Furthermore, negative relations with customers could enhance employees’ emotional dissonance, which refers to fake emotional displays or expressions (9) and involves the divergence between what employees feel and what they display (10). Specifically, emotional dissonance has been defined as the ‘discrepancy between the emotions felt and those expressed to meet organizational or professional display rules’ (11, p.379). It has been proven that the display rules with customers are fairly consistent across cultures (12).

A study conducted on service workers aimed to understand the positive and negative effects of emotions in organisations, showed that among different aspects of emotions (requirements to display positive emotions, the requirement to display negative emotions, the requirement to be sensitive to clients’ emotions, and emotional dissonance), it has been found that emotional dissonance was the stressful aspect of emotions at work (13). Emotional dissonance has been widely investigated in employees’ dealing with customers and the obtained results showed its association with negative health outcomes for employees, such as higher levels of strain and depression, decreased sense of work motivation and well-being, enhanced emotional exhaustion, burnout, and general health (3, 9, 14-17).

In a sample of service workers, it has also been found that emotional dissonance was significantly associated with a decreased sense of well-being, even after controlling for negative disposition (4). This means that at equal levels of negative affectivity, emotional dissonance is related to decreased well-being.

Furthermore, literature outlined the relevant role played by emotional dissonance in explaining the process which leads from work characteristics (such as negative relations with customers) and employees’ poor health. For example, Cheung and Tang found that emotional dissonance mediated the relationship between work characteristics (e.g. perceived display rules) and employees’ health measured as psychological health and somatic complaints (17). This mediating role has been confirmed in the literature in many studies especially considering burnout as a health outcome (e.g. 18, 19). Also, a recent study developed by Andela and Truchot demonstrated that emotional dissonance mediated the relationship between workload, patient’s suffering and burnout (20).

Moreover, many studies investigated the relationship between burnout and other health outcomes such as psychological well-being and workplace phobic anxiety (21, 22). Specifically, in their review of the Job Demands-Resources Model, Bakker, Demerouti and Sanz-Vergel reported that burnout is related to health-related outcomes and in general physical health, specifically to depressive and anxiety disorders, mood disturbance, depressive symptoms and depression (23-27). Unfortunately, none of the studies mentioned before has been conducted on a sample of retail workers in supermarkets. Despite there are some studies that investigated burnout prevalence among retail workers in supermarkets (e.g. 28), there is a need to better understand the relationships existing between negative relations with customers and poor health. This is important also because, retail workers in supermarkets are victims of customers’ aggressions. A study conducted through surveys on grocery stores, call-centers and stores in general showed the perpetrators of customer aggression were both male and females (40.6% of the respondents), between 20 and 40 years old (45.2), mostly uneducated or less educated (39.8%) and with no difference on income (35.8%) (29). Thus, the aim of the present study was to examining the existing gaps in the literature concerning the mechanisms underlying the relationship between negative interactions with customer and employees’ poor health in a sample of retail workers which have frequent interactions with clients. Based on the existing theory and summarized research, the following hypothesis was formulated:

Negative customer relations are related to symptoms of poor health, through the subsequent mediation first of emotional dissonance (as a first mediator) and then of job burnout (as a second mediator).

The overall model is depicted in Figure 1.

Figure 1

The hypothesized serial mediation model

Methods

Sample and procedure

Data were collected using an anonymous, self-report questionnaire employed as a part of a psychosocial risk assessment project in one of the major Italian companies working in the large-scale retail sector. In accordance with the Human Resources department of the company,

The questionnaire enclosed a cover letter outlining the contents and purpose of the study and emphasized the anonymity of the participants and anonymity the confidentiality of information, in accordance with the guidelines for personal data processing defined by the Italian privacy law (law decree DL-196/2003). Therefore, participants informed consent was inherent in the survey’s completion and answers were provided in group sessions conducted by two components of the research team in order to answer questions about the research. Furthermore, the letter indicated that participants were entitled to withdraw at any time without giving explanations and the employer would not receive any information about the participants’ withdrawal. With regard to ethical standards for research, the current study complied with the latest version of the Declaration of Helsinki (World Medical Association, 2013).

The final sample was composed of 610 employees working in 28 supermarkets (average number of workers per supermarket 21.79, sd = 13.89), with a mean age of 40.63 years (sd = 8.26). Most of them were women (70%), were married (64.3%) and held a high school degree (96%). The majority of the sample had a permanent employment contract (91.3%) and worked full-time (64.1%). The mean job tenure was 13.72 years (sd = 9.61). A detailed sample description is reported in Table 1.

Table 1

Description of Study Participants

	Total sample (N= 610)
Gender: Men / Women	30% / 70%
Age: Mean (SD)	40.63 years (8.26)
Type of store: Supermarket / Minimarket	84.1% / 15.9%
Job: Cashier / Sale Assistant / Other	19.7% / 73.5% / 6.8%
Educational level:
Secondary School	46.2%
High School	51.3%
University degree	2.5%
Marital status:
Married	64.3%
Single	24.3%
Divorced	10.3%
Widower/Widow	1.1%
Work contract
Full time open-ended contract	62.4%
Part time open-ended contract	28.9%
Full time fixed term contract	1.8%
Part time fixed term contract	4.9%
Other contracts	2%
Job tenure, years: Mean (SD)	13.72 years (9.61)
Actual working hours: Mean (SD)	32.34 hours/week (7)

Measures

Negative customer relations. The perception of difficulties in managing the everyday relationship with customer was assessed using an ad-hoc measure that included three items (“I argue with customers”; “Interacting with customers is frustrating”; “I feel stressed by some customers”). These items were developed according to theoretical perspectives and empirical findings on customer-related social stressors (30), as well as the association between customer aggression and job-induced tension (4). Participants were asked to report how often they experience these situations on a 5-point frequency scale ranging from 1 = never to 5 = very often.

Emotional dissonance. The perceived discrepancy between the emotions actually experienced by workers and those displayed in order to fit the requirement of their role was measured using four items developed by Agervold (3). For instance, “You have to repress your true emotions to appear positive and friendly”. Each item was assessed on a 4-point frequency scale from 1 = never/rarely to 4 = almost daily.

Job burnout. Workers’ level of burnout was rated using the exhaustion subscale of the Maslach Burnout Inventory-General Survey (MBI-GS, 31). This scale consisted of five items, for instance “I feel emotionally drained from my work”. Each item was scored on a seven-point frequency scale ranging from 0 = never to 6 = every day. In the current study, Tthe mean level of emotional exhaustion was M= 2.61 (SD = 1.51). On average, the current sample reported a mild level of burnout, according to the three severity categories created by Ahola and colleagues (32): no (0–1.49), mild (1.50–3.49) and severe (3.50–6).

Poor health. The individual perception of poor health status was measured with the 12-item version of the General Health Questionnaire (GHQ-12; 33). This survey is one of the most renowned self-report measures of psychological strain and is aimed to identify minor psychological disorder in non-clinical populations. In particular, the symptoms covered by the GHQ-12 are the impaired ability to focus on tasks, the perception of being consistently under pressure, experiencing low mood and sad feelings, and the incidence of sleep problems due to worries. Respondents rated how often, during the last three months, they faced situations such as “been taking longer over the things you do?” (social dysfunction) and “having lost much sleep over worry?” (general dysphoria). These items were rated using a four-point frequency scale ranging from 0 to 3.

Statistical Analysis

The study hypotheses were tested using a bootstrapping approach as outlined by Hayes (34). This method is suitable when sample sizes are relatively small because it produces a distribution using the observed data, from which statistical effects are estimated (35). In particular, the serial mediation postulated by the current study was tested using Model 6 in the SPSS macro PROCESS. To derive the direct and indirect effects, the model estimates all path coefficients at once. The bootstrapping method generated an estimate of the indirect effect, including a 95% confidence interval. When zero is not comprised in the 95% confidence interval, it can be assumed that the indirect effect of negative customer relations (i.e., the independent variable) on workers’ poor health (i.e., the criterion variable) is mediated first by greater levels of emotional dissonance (i.e., the first mediator) and, subsequently, by a higher degree of job burnout (i.e., the second mediator) among workers. Based on previous empirical results, we decided to control for the potential confounding effect of participants’ gender and age (e.g. 8, 32).

Results

Descriptive statistics

Table 2 reports the means, standard deviations, correlations, and internal consistencies of the study variables. All significant relationships between the variables were in the expected direction.

Table 2

Means, Standard Deviation, and Correlations Among the Study Variables

				r
	Range	M	SD	1	2	3	4	5	6
1. Gender (1= male)	0-1	.30	.46	--
2. Age	-	40.63	8.26	.03	--
3. Negative customer relations	1-5	2.28	.94	-.06	-.03	(.60)
4. Emotional dissonance	1-4	2.71	.88	-.09*	.07	.27***	(.76)
5. Job burnout	0-6	2.61	1.51	-.06	.07	.55***	.23***	(.85)
6. Poor health	0-3	.33	.45	-.17***	.04	.27***	.14**	.48***	(.86)

Notes. N= 610; *p < .05; **p < .01; ***p < .001. Cronbach’s alpha in brackets along the diagonal.

Model testing

The results of the hypothesized model are reported in Table 3. Concerning the direct effects, in our sample the occurrence of negative customer relations was positively related to emotional dissonance (included as first mediator), with b = .25 (p = .000), 95% CI = [.18; .32].

Table 3

Path Coefficients and Indirect Effects for Poor Health

	Path Coefficients						Indirect Effects
	To Emotional dissonance (ED)		To Job burnout (JB)		To Poor health (PH)
	b	SE	b	SE	b	SE
Gender (1= male)	-.14	.07	-.38	.56	-.14**	.03
Age	.01*	.01	.08*	.03	.01	.01
Negative customer relations (NR)	.25***	.04	4.25***	.28	.01	.02
Emotional dissonance (ED)			.72*	.30	.01	.02
Job burnout (JB)					.03***	.02
	R2 = .08***		R2 = .32***		R2 = .25***
							b	Boot SE	95% CI
Total							.13	.01	.01; .16
NR → ED → PH							.01	.01	-.01; .01
NR → JB → PH							.12	.01	.09; .15
NR → ED → JB → PH							.01	.01	.01; .02

Note: N = 610. * p < .05; ** p < .01; *** p < .001; 95% CI= 95% confidence interval using the bootstrap bias-corrected method using 5000 samples.

Furthermore, negative customer relations were associated with higher levels of job burnout (included as second mediator): b = 4.25 (p = .000), 95% CI = [.18; .32]. In a similar vein, also the perceived emotional dissonance reported a positive association with symptoms of job burnout, with b = .72 (p = .017), 95% CI = [.13; 1.32]. Concerning the model outcome, negative customer relations and emotional dissonance reported a non-significant association with poor health symptoms, whereas job burnout displayed a positive relationship: b = .03 (p = .000), 95% CI = [.02; .04]. According to the indirect effects reported in Table 3, emotional dissonance did not mediate the relationship between negative customer relations and poor health, whereas job burnout acted as a significant mediator between these variables. The obtained results indicate that emotional dissonance and job burnout sequentially mediated the relationship between the perceived occurrence of conflictual relations with customers, on the one hand, and symptoms of poor health among workers, on the other hand. This evidence provided support to Hypothesis 1: negative customer relations were associated with a higher perception of emotional dissonance and job burnout, which in turn was related to a greater perception of poor health symptoms. Among the control variables, gender reported a negative association with poor health, while workers’ age showed a significant association with emotional dissonance and burnout symptoms.

Discussion

By noticing negative customer relations as an increasing risk factor in the workplaces, the main aim of this study was to analyse how this psychosocial risk factor could determine employees’ psychological health. Using a serial mediation approach, results of the present study showed that negative relationships with customers is related to higher levels of emotional dissonance which in turn is related to higher levels of burnout and in turn poor health of employees.

These results are in line with previous studies that demonstrated how customer incivility is related to reduced employee performance, increased turnover and stress (36). Overall, results presented in this study are in line with the Job Demands-Resources model, which posit that job demands (as negative customer relationship) could trigger what has been defined as the health impairment process, which consist in a deterioration of employees’ health in terms for example of increased burnout levels (37).

Moreover, according to previous studies which showed that emotional dissonance could act as a mediating variable between psychosocial risk factors and employees’ psychological health (e.g. 18, 19), results of this study highlighted that negative relations with customers was related to emotional dissonance, which in turn was related to decreased psychological health, suggesting the importance to consider both emotional dissonance and burnout in explaining reduced levels of employees health due to negative relationships with customers. These results add knowledge to the relationship between an increasing risk factor in organisations (negative relations with customers) and employees’ health, responding to the growing awareness that negative relations between customers and employees could have on employees’ psychological well-being (6, 38, 39).

Among the control variables, gender reported a negative association with poor health, thus suggesting that women reported a greater occurrence of poor health symptoms. No differences have been found between males and females concerning burnout, despite previous studies showed that females tend to report higher levels of burnout (e.g., 40). Furthermore, in the current sample workers’ age showed a significant association with burnout symptoms, as suggested by previous studies (41).

This study has also some limitations that need to be acknowledged. First of all, the study is cross-sectional, thus no causal interpretations could be inferred. Furthermore, the variables were investigated only through self-reported measured enhancing the occurrence of potential common method biases. Moreover, Cronbach alpha value of one variable (negative relations with customer) was not so high, despite still acceptable as according to Nunnally’s recommendation, scales with item consistencies higher than .60 can be used for research purposes (42). Moreover, in this study we considered only one psychosocial risk factor (negative relations with colleagues), while other psychosocial risk factors could have a relevant importance in the occurrence of negative health outcomes such as burnout. Despite the authors did not distinguish among different work sectors or positions, the meta-analysis of Lee and Ashforth found that the most important job demands able to predict burnout were role ambiguity, role conflict, role stress, stressful events, workload and role pressure (43).

Lastly, all the participants worked in supermarkets, thus generalisation of these results to other types of jobs that involved the relations with customers should be cautious. More information about the processes that link negative relationships with customers and health could be gained through other methodologies such as qualitative focus groups. This technique could help to understand for example which dynamics in the relationships with clients can trigger higher levels of emotional distance or burnout.

The present study showed how negative relations with customers are related to employees’ burnout and health and contributed to the literature by highlighting the importance of considering both emotional dissonance and burnout in investigating the relationship between a job stressor such as negative relationships with customers and poor workers’ health. To the best of our knowledge there are currently no studies that analysed the processes linking relationships with customers, emotional dissonance, burnout and health in a sample of Italian retail workers in supermarkets.

Thus, employers aiming to enhance employees’ psychological health should implement activities able to control these two risk factors. For example, despite it is impossible to prevent or eliminate aggressive clients, it could be useful for organizations to develop interventions in order to enhance team reflexivity and revaluation. Andela and Truchot found that these were able to buffer the negative effects on employees’ burnout of emotional dissonance (20). Moreover, organizations whose employees’ work activities are characterized by a daily contact with customers should implement interventions such as training activities in order to enhance the employees’ skills in dealing with customers, for example on how to manage emotions at work. Recent studies showed that also mindfulness-based interventions seem to be useful in reducing burnout levels. In a randomised study conducted on teachers showed that workers attending the mindfulness-based training showed lower levels of occupational stress and burnout at post-program and follow-up, than did those in the control condition (44). Despite training is sometimes considered by employers as an extensive and cost-consuming activities, there are studies with promising results that are testing also other type of interventions such as meditation apps (45). Implement interventions able to decrease employees’ burnout is also crucial as many studies demonstrated how burnout is also related to employees’ absenteeism (e.g. 46). We recommend occupational physicians to not underestimate the potential harmful effect of negative relations with customers and to consider also this psychosocial risk factors when dealing with workers which shows burnout symptoms. In the case of work-related stress assessment, occupational physicians should not treat this psychosocial risk factor as unchangeable but as a factor that can be taken under control by companies by providing in their monitoring plans interventions (for example the training programs mentioned above) in order to help workers manage their emotions and reduce the emotional dissonance and consequently show lower symptoms of burnout and higher levels of general health.

In conclusion, this study tried to add knowledge concerning the processes related to employees’ burnout and health in the service sector, considering the growing awareness of negative customer relation as an emerging psychosocial risk factor.