Maslach and Jackson (1981) describe emotional exhaustion as individuals’ feeling of being emotionally tired due to their work. Maertz and Campion (2004) define it as a state of fatigue causing no enthusiasm to work, stress, and boredom with work and routine life. Employees who experience emotional exhaustion cannot properly fulfill their responsibilities and have poor performance. In addition, they feel psychologically and emotionally exhausted, i.e., they perceive their energy has run out and their emotional resources have been used up.

According to Bakker (2013), when employees are faced with excessive work demands and cannot manage the two-way conflict between work and family, they experience emotional exhaustion. This will lead to a poor job performance and unwanted results (De Cuyper et al., 2014). Moreover, Karatepe (2013) refers to emotional exhaustion as a feeling of fatigue that emerges when employees are subjected to excessive work demands and cannot handle roles and conflicts at work, in their family, or with colleagues in the company or organization where they work.

Netemeyer et al., (1996) categorize role conflicts at work into work–family conflict and family–work conflict. Work–family conflict generally occurs when work demands oblige employees to spend a large amount of time completing their tasks, thus interfering with their responsibilities as a family member (Chambel et al., 2017). Family–work conflict arises when family demands force individuals to devote more time to family matters, making it difficult for them to properly complete their tasks.

Furthermore, Kahn et al., (1964) explain the concept of work–family conflict using the role theory framework. According to them, the main determinant of individuals’ behavior is the behavior that is expected of them. In this regard, the role theory suggests that the expectations of each role can lead to inter-role conflict when there exists pressure to dominate the individuals’ time in order to satisfy all expectations because each role requires time, energy, and commitment. Hence, work–family conflict is a form of inter-role conflict at work and in the conflicting environment.

Inter-role conflict is further defined by Greenhaus and Beutell (1985) as the pressure that arises when individuals find it difficult to balance their responsibilities and roles as members of an organization and as family members. This pressure emerges when there is a conflict between individuals’ roles at work (or in their family) and other demands of family roles (or work roles) that also need to be fulfilled. Thus, this pressure increases, causing the effectiveness of one role to be hampered by other role activities (reciprocal role conflict). There are three types of work–family conflicts: time-based conflict, strain-based conflict, and behavior-based conflict. Time-based conflict occurs when role pressure makes individuals sacrifice their time devoted for one role to meet other role demands. Strain-based conflict arises when pressure from one role hinders the fulfillment of another role. Behavior-based conflict takes place when a specific behavior pattern of a role is incompatible with the expected behavior for the other roles.

A study by Frone et al., (1992) show that inter-role conflict between work and family or family and work is triggered by different antecedents and pressures: work and family conflict arises as a result of a job stressor and job involvement, while family and work conflict occurs as a result of a family stressor and family involvement.

For this study, the definition of “family” (in the Indonesian context) was taken from the Indonesian Marital Law No. 1/1974, which refers to family as an entity of communication and interaction between all parties in fulfilling their roles, such as those of a spouse, a parent with their children, or siblings (Wiratri, 2018).

Person–job fit is a form of person–environment fit. It is defined as the capacity of employees to find congruence between their abilities and the demands of the work they perform (demands–abilities fit) or between their desires and the attributes of their job (needs–supplies fit) (Edwards, 1991). Additionally, person–job fit is described as the degree of suitability between individuals and their work tasks. This definition includes compatibility between employees’ needs and the work equipment available to fulfill their job responsibilities, as well as suitability of their skills (Cable & DeRue, 2002; Kristof-Brown et al., 2005) and personality and abilities to meet work demands (Babakus et al., 2010; Donovan et al., 2004). For that reason, person–job fit plays an important role in the selection process: recruiters consider the level of compatibility between individuals’ knowledge, skills, and abilities and the requirements of the job to be the basis of the selection process (Karatepe & Karadas, 2016).

As explained earlier, the concept of person–job fit proposed by Edwards (1991) refers to the level of suitability of individuals with the work they perform from two angles: job demands and supplies. From the demand angle, we consider the extent of compatibility between employees’ abilities and work demands, while, from the supply angle, we examine the degree of congruence between their personality and work demands. Demand–ability fit requires a level of suitability between individuals’ skills, knowledge, time, and energy to carry out work tasks. If employees are encouraged to maximize their time and energy for successful job completion, role conflicts will arise (Edwards, 1996; Karatepe & Karadas, 2016).

The study by Karatepe and Karadas (2016) analyzes the relationship between person–job fit and work–family conflict and family–work conflict based on questionnaires applied to 282 FLEs at four-star and five-star hotels in Romania. The results show that person–job fit has a negative influence on family–work conflict and work–family conflict.

Another work by Afsar and Rehman (2017) investigates the influence of person–job fit on employees’ performance and reveals that it causes family–work conflict and work–family conflict. According to such authors, employees find it difficult to fulfill their responsibilities as family members, which results in them experiencing physical and emotional exhaustion in their efforts to reduce the conflicts between work and family. In addition, the authors find that person–job fit may have a direct impact on the urge to optimally complete work tasks, which encourages the emergence of work–family conflict. Furthermore, the study by Unal (2017) shows that person–job fit has a negative contribution on work–family conflict. Individuals who perceive a fair balance between getting things done and paying attention to how others feel and between the needs of others and their own needs and who create enough space for themselves will tend to miss activities with family because of the amount of time they have to spend on work responsibilities. Based on previous research regarding person–job fit and family–work conflict and the relevance of these phenomena, we propose the following hypotheses:

H1: Person–job fit is negatively related to Family–Work Conflict (FWC).

H2: Person–job fit is negatively related to Work–Family Conflict (WFC).

Moreover, Carlson et al., (2009) state that work–family conflict arises when there is not a balance between responsibilities and expectations from the role of individuals at work and their role as family members, causing them emotional exhaustion. This fatigue, in turn, burdens their mind and inhibits their motivation to achieve good performance. According to Pasewark and Viator (2006), business development increasingly triggers the possibility of family–work conflict among workforce, especially in occupations with a high level of customer/client contact. Also, the findings of their study report family–work conflict as a source of individuals’ dissatisfaction with their work. Job dissatisfaction occurs when employees feel that their work is not in line with their desires or values, leading to a conflict between two roles (family and work).

Besides showing that both family–work conflict and work–family conflict can drain employees' energy and resources, Jawahar et al., (2012) indicate that individuals who experience work–family conflicts will experience high emotional exhaustion and depersonalization.

Another study by Karatepe (2013) reveals that family–work conflict and work–family conflict significantly caused emotional exhaustion among managers and FLEs at hotels in Romania. In terms of family–work conflict, full-time hotel employees reported to feel that the place where they work is not yet family-friendly or support their family's needs when, for instance, they have to leave their children due to their job role. Some of them certainly sacrifice their time with family, which results in low morale and excessive stress at work. In addition, they feel unable to fulfill their family roles because they are being demanded to completely focus on meeting their work demands. This leads to emotional exhaustion among FLEs, which, in turn, results in stress at work.

A research conducted by Geraldes et al., (2019) with 2055 call center workers in Portugal shows that work–personal life conflicts, whether driven by time or job demands, positively influence burnout, both in the form of exhaustion and cynicism. Based on this, such authors propose a hypothesis that states that employees who experience family–work conflict and work–family conflict may be affected by emotional exhaustion. Therefore, we develop the following hypotheses:

H3: Family–work conflict is positively related to emotional exhaustion.

H4: Work–family conflict is positively related to emotional exhaustion.

Furthermore, Figure 1 below shows the research model that being explained above.

Figure 1. Research model Figure 1. Research model Figura 1. Modelo de investigación Source: Created by authors.

This empirical study was conducted with call center workers from financial institutions (banks and insurance firms) in Jakarta, Indonesia. We chose this city because most call centers in said country are based in Jakarta—the center of business and economy in Indonesia (CNN Indonesia, 2019). We employed judgmental sampling, which is a nonprobability sampling technique. Inclusion criteria included permanent employees with more than one year of service. For data collection, we used snowball sampling; team leaders received the questionnaire (both physically and electronically) and then helped disseminate it within the organization.

Questionnaires were distributed ensuring confidentiality and anonymity. Some questions requested information about the demographic characteristics of the respondents (e.g., gender, age, marital status, number of children), and some others were developed to measure the research variables: person–job fit, family–work conflict, work–family conflict, and emotional exhaustion.

Of the 190 questionnaires distributed to the Call Center Officers (CCOs), only 154 or 81.1 % could be further processed because some respondents did not meet the inclusion criteria.

In terms of demographic characteristics, 87 respondents (56 %) were women, and the rest were men. As for age, 126 respondents (82 %) were 24–26 years old; 17 (11 %), 21–23 years old; and the remaining, 27–30 years old. The working tenure of 118 respondents (77 %) was 1–2 years, and that of the rest, 3–5 years. A total of 82 respondents (53 %) were single, and the remaining were married. Finally, 107 respondents (69 %) reported not to have children; 43 (28 %), 1–2 children; and the rest, 4 children.

For all constructs in this study, we employed scales that have been applied in previous studies. For instance, person–job fit was measured using three-item scales based on the study by Donovan et al., (2004). WFC and FWC were measured by means of a scale adjusted from the work by Netemeyer et al., (1996). These two latter variables consisted of five measurement items. Regarding emotional exhaustion, we used a scale adapted from the study by Maslach and Jackson (1981), and it included eight measurement items. All questions were measured using a 1–5 Likert scale, with scores ranging from 5 (strongly agree) to 1 (strongly disagree).

We analyzed the data via Confirmatory Factor Analysis (CFA) and Structural Equation Modeling (SEM). We tested the measurement and structural models following a two-step approach (Anderson & Gerbing, 1988). In the first step, we tested the observed variables using CFA in order to assess their convergence and discriminant validity (Anderson & Gerbing, 1988; Fornell & Larcker, 1981), as well as their Composite Reliability (CR) (Bagozzi & Yi, 1988). After testing the measurement model, we assessed the structural model with SEM.

We conducted the Goodness-Of-Fit (GOF) test with the results of the chi-square (χ2) statistic, Root Mean Square Error of Approximation (RMSEA), Standardized Root Mean Square Residual (SRMR), Comparative Fit Index (CFI), and Normed Fit Index (NFI). A model is considered to be a good fit if its chi-square to degree-of-freedom ratio is 3:1 or less, its RMSEA and SRMR are below 0.8, and its CFI and NFI are equal to 0.95 or higher (Hair et al., 2009). We calculated the input covariance matrix using LISREL 8.80 software.

Based on the results of the CFA, some items were reduced because they had a loading factor below 0. 50 (Hair et al., 1995; Igbaria et al., 1997): for instance, one question from the Work–Family Conflict and Family–Work Conflict variables, and three questions from the Emotional Exhaustion variable.

The five model fit factors showed good results (χ2 = 190.39, df = 97, χ2/df = 1.96; CFI = 0.95; NFI = 0.90; RMSEA = 0.079; SRMR = 0.087). All questions in this study had a loading factor above 0.50 (Hair et al., 1995; Igbaria et al., 1997). Table 1 shows that the Average Variance Extracted (AVE) of each latent variable was above 0.50, suggesting good convergent validity (Anderson & Gerbing, 1988; Fornell & Larcker, 1981). Likewise, the Composite Reliability (CR) of each latent variable was above 0.70, indicating the good reliability of the measuring instruments (Bagozzi & Yi, 1988).

Table 1. Validity and reliability of measurement indicators/instruments/items Table 1. Validity and reliability of measurement indicators/instruments/items Tabla 1. Validez y confiabilidad de los elementos/indicadores/instrumentos de medición

Latent variable	Indicator		Standardized loading factor	Composite Reliability (CR)	Average Variance Extracted (AVE)
Person–Job Fit	PJF1: My skills and abilities are very compatible with the demands of my job.		0.86	0.815	0.604
	PJF2: My personality is very compatible with the requirements of my job.		0.88
	PJF3: There is a good fit between my job and myself.		0.55
Work–Family Conflict	WFC1: The demands of my job interfere with my family life.		0.79	0.824	0.542
	WFC2: My working hours make it difficult for me to fulfill my responsibilities as a family member.		0.62
	WFC3: I am unable to fulfill my responsibilities at home due to the demands of my job.		0.73
	WFC5: I usually have to cancel my attendance to family events due to the demands of my job.		0.79
Family–Work Conflict	FWC1: The demands of my family or partner interfere with my work-related activities.		0.77	0.805	0.509
	FWC2: I sometimes have to leave my work because of family matters.		0.63
	FWC3: The requests of my family or partner interfere my work-related activities .		0.76
	FWC4: Some family matters interfere with my job responsibilities, such as starting to work on time, completing daily tasks, and working overtime.		0.68
Emotional Exhaustion	EE1: My job makes me feel emotionally exhausted.		0.50	0.841	0.527
	EE2: I always feel tired when I wake up in the morning because I have to go back to work.		0.89
	EE3: Working all day long makes me feel very tired.		0.57
	EE4: I feel frustrated with my work.		0.91
	EE5: I feel like on the edge.		0.67

Source: Created by the authors.

Table 2 shows the correlation matrix, along with the mean and standard deviation of each construct used in this study. As for the demographic variables, working tenure shows a significant negative correlation with emotional exhaustion.

Table 2. Summary statistics and correlations of the observed variables Table 2. Summary statistics and correlations of the observed variables Tabla 2. Resumen de estadísticas y correlaciones de las variables observadas

	Variable	1	2	3	4	5	6	7	8
1	Person–Job Fit	-
2	Work–Family Conflict	-0.627**	-
3	Family–Work Conflict	-0.462**	0.737**	-
4	Emotional Exhaustion	-0.072**	0.253**	0.216**	-
5	Gender	0.014	-0.060	-0.062	0.022	-
6	Marital Status	-0.080	0.040	-0.006	0.060	-0.009	-
7	No. of Children	0.146	-0.098	-0.032	0.062	0.005	-0.676**	-
8	Working Tenure	-0.083	-0.051	-0.075	-0.206*	-0.103	-0.251**	-0.178*	-
	Mean	4.78	4.35	4.30	4.36	1.56	1.53	1.33	1.23
	Standard Deviation	1.53	0.49	0.54	0.47	0.49	0.50	0.52	0.42

Source: Created by the authors. Notes:

The composite scores of each construct were calculated by averaging the respective item scores. * p < 0.05, ** p < 0.01 (one-tailed test).

Based on Figure 2 the structural model test yielded good results (χ2 = 169.65, df = 99, χ2/df = 1.71; CFI = 0.96; NFI = 0.91; RMSEA = 0.068; SRMR = 0.064). According to such results, person–job fit has a negative impact on both family–work conflict (β = -0.88, t = -8.93) and work–family conflict (β = -1.33, t = -7.57), which supports hypotheses 1 and 2. Additionally, only family–work conflict shows a positive impact on emotional exhaustion (β = 0.61, t = 1.67), while work–family conflict has no effect on emotional exhaustion among call center workers (β = -0.32, t = -0.96).

Figure 2. Results of the Structural Model Test Figure 2. Results of the Structural Model Test Figura 2. Resultados de la prueba del modelo estructural Source: Authors’ elaboration (2019).

This study had two purposes. First, it aimed to examine the influence of person–job fit (as one of the individual stressors of role conflict) on call center workers in terms of both family–work conflict and work–family conflict. Second, it sought to determine whether role conflict (family–work conflict and work–family conflict) experienced by call center workers and emotional exhaustion while working are correlated. To analyze the relationship between the variables under study, questionnaires were distributed among call center workers in the financial services industry (banks and insurance companies) in Indonesia. This research yielded several interesting findings.

For instance, our results indicate that person–job fit negatively affects role conflict among call center workers (a type of FLEs). In other words, from the demand angle, the more appropriate their level of skills, knowledge, and abilities to perform their work duties, the less likely they will experience role conflict (family–work conflict and work–family conflict). From the supply angle, if call center workers perceive that their job is compatible or in line with their own personality, they will feel more energetic and comfortable in carrying out their work routines (Edwards, 1991; Kristof et al., 2005). These findings reinforce those of a previous study conducted by Karatepe and Karadas (2016) with FLEs at hotels in Romania.

Moreover, we only found a relationship between family–work conflict and emotional exhaustion among call center workers. This is consistent with the findings of Karatepe (2013) and Glaser and Hecht (2013). To maintain an excellent customer service in accordance with Indonesian regulations, most call centers operate 24 hours a day. Thus, call center agents must work effectively on shifts at the convenient time for customers. This includes night shifts, weekends, and public holidays to get a head start on heavy customers’ complaints. These shifts can sometimes prevent employees from spending quality time with their family.

Call center workers experience family–work conflict in multiple forms. According to Greenhaus and Beutell (1985), there are three forms of family–work conflict: time-based conflict, strain-based conflict, and behavior-based conflict. Regarding time-based conflict, a 24/7 shift schedule does not allow call center workers to attend important family events and reduces their quality time with family (with their children, spouse, or parents) due to their work demands. This will, in turn, cause them emotional strain because they are unable to meet their family demands as a family member. In terms of strain-based conflict, the emotional fatigue experienced by call center workers during their working hours makes it difficult for them to satisfy their roles of their nonwork life. With respect to behavior-based conflict, emotional fatigue may impact their family life at home, which might be incompatible with their behavioral expectation as a family member.

Work–family conflict was proven not to have a significant effect on the emotional exhaustion experienced by call center workers. Even though this contradicts the results reported by previous studies (Karatepe, 2013; Karatepe et al., 2010), it supports those of a longitudinal study conducted by Richter et al., (2015) into the effect of work–family conflict on emotional exhaustion. Work–family conflict becomes a normal consequence of performing one’s job (Richter et al., 2015), as call center workers have already accepted it as a job’s risk. Additionally, due to these employees’ awareness of the characteristics of their work as an emotional labor, they are expected to always show positive emotions and suppress the negative ones when dealing with customers (Babakus et al., 2010). This awareness makes call center workers accept the risk of their job which demands them to always express positive emotions and suppress the negative ones to properly meet their job demands.

Further studies should consider several aspects. First, this study did not take into account gender (Karatepe, 2013) and marital status as control variables to test the influence of family–work conflict on emotional exhaustion. Second, we recommend future research to test the direct effect of person–job fit on emotional exhaustion (Mengenci, 2014; Mulki et al., 2006). Third, the sample size of this study is small, as the number of samples obtained was limited to specific categories in the financial services industry. In addition to banking and insurance firms, such industry in Indonesia includes finance companies and stock exchanges. Therefore, future studies into call centers in the financial services industry could cover all categories or types of financial businesses in Indonesia. Fourth, we suggest further research to investigate the negative impact of emotional exhaustion on both the organization and employees, such as in terms of job satisfaction (Mulki et al., 2006), job performance, job embeddedness (Karatepe, 2013), and turnover intention (Ducharme et al., 2007; Knudsen et al., 2009; Shih-Tse Wang, 2014).

The results of this study have several managerial implications. First, the level of compatibility between individuals (both in terms of personality and skills) and their job (or person–job fit) can prevent role conflict (family–work conflict and work–family conflict) among call center workers in the financial services industry. Therefore, to provide a quality customer service, call center managers (in the financial industry) must assess candidates’ personality and skills during the selection process. Otherwise, they will experience difficulties regarding employee retention (Babakus et al., 2010; Karatepe & Karadas, 2016).

Second, since family–work conflict plays an important role in emotional exhaustion among call center workers, companies need to create a family-friendly work environment (e.g., health insurance for children and permissions so that employees can receive their children’s report cards) to reduce family strains and, thus, encourage them to give their best at work.