Investigating Configurations of Internal Corporate Social Responsibility for Work–Family Spillover: An Asymmetrical Approach in the Airline Industry

Abstract

Corporate social responsibility (CSR) and work–family interface have attracted considerable scientific interest; however, their relationship has not been considered yet. Drawing on the conservation of resources (COR) theory, this study fills this gap by examining the relationship between internal CSR and work to family spillover. While most previous CSR studies widely examined the net effect of a single CSR dimension, this study discovers configurations of five different internal CSR dimensions (internal dissemination, compensation, occupational health and safety, training, and legal employment) in explaining positive and negative work to family spillover. Using fuzzy-set qualitative comparative analysis (fsQCA), this study examines a primary database of 136 flight attendants working for Asian airline companies; results reveal that the combination of internal dissemination and compensation creates the most driving power in leading to positive spillover. The absence of internal dissemination, occupational health and safety, and legal employment leads to high negative spillover regardless of the presence of compensation. This study broadens the literature by linking internal CSR to employees’ perceptions of work–family spillover. FsQCA findings also make a methodological contribution to prior CSR research by indicating three configurations that explain positive and negative work–family spillover. Findings provide airline companies with practical guidelines that are useful to enhance positive spillover and reduce negative spillover from work to family domains among their flight attendants.

1. Introduction

Flight attendants have frequent interaction with a large number of passengers and spend the majority of their workday resolving their requirements and grievances (Karatepe and Bekteshi 2008; Tang et al. 2020). Due to long-distance flights and unscheduled or uncertain working timetable (Tang et al. 2020), flight attendants’ family life is interrupted, which prevents them from developing meaningful relationships with their partners and children (Tsaur et al. 2020). The emotional tiredness that results from flight attendants not being able to complete their nonwork-related responsibilities creates a vicious cycle that reduces their job satisfaction and increases depersonalization (Karatepe and Bekteshi 2008) and causes negative impacts on their service performance (Aboobaker and Edward 2019). There needs to be changes made to the working environment for flight attendants to better function in their professional work and personal lives.

Corporate social responsibility (CSR) refers to firms’ duties and obligations to their stakeholders, demonstrating management’s overall business supremacy (ISO 2010). Employees’ physical and psychological working environments are directly affected by CSR activities, which are regarded as internal CSR practices (Droppert and Bennett 2015). Internal CSR refers to activities and policies that consider and satisfy the physical and psychological well-being of employees (Espasandín-Bustelo et al. 2020; Sanusi and Johl 2020). Internal CSR has been found to promote various work-related attitudes, such as organizational commitment (Mory et al. 2015), and job satisfaction (Golob and Podnar 2021). However, organizational activities and practices cannot be manifested as CSR unless they fulfill the well-being of employees (Gond et al. 2017). Accordingly, prior scholars called for empirical research that broadens CSR’s roles, which has been limited to work-related outcomes (Rupp and Mallory 2015).

Employees’ quality of life is improved by CSR because it satisfies individual requirements, including social, health and safety, actualization, knowledge, and economic demands (Kim et al. 2018; 2020). CSR was found as an enabler of work engagement, career satisfaction, and voice behavior among flight attendants (Ilkhanizadeh and Karatepe 2017). However, research on the effects of CSR on employee well-being are still under investigation (Golob and Podnar 2021), especially in the airline industry (Chen and Tang 2021). In addition, enterprises in developing countries are more susceptible than their counterparts in developed countries due to their limitation in organizational resources and capabilities (Tsai et al. 2022), which may constrain the implementation of CSR guidelines and activities in developing ones (Luu 2020). As such, empirical research on CSR activities conducted in developing countries (e.g., Vietnam) has not been given much attention (Luu 2020; Thang and Fassin 2017).

Symmetrical thinking is examined via standard approaches (regression analysis, structural equation modeling), which only allows researchers to simply look at the net influence of distinct independent factors on a certain outcome (Misangyi et al. 2017). Indeed, prior CSR research has mostly examined the CSR–employee outcomes linkage through symmetrical approach (e.g., Espasandín-Bustelo et al. 2020; Golob and Podnar 2021; Ilkhanizadeh and Karatepe 2017; Kim et al. 2018; 2020; Luu 2020). Differentiating from prior CSR research, the present study adopts asymmetric thinking to generate various configurations of separate initiatives of internal CSR via fuzzy-set qualitative comparative analysis (hereafter fsQCA). FsQCA has been a powerful technological tool that analyzes the joint effects of two or more antecedent factors to enable an outcome, as well as investigating all possible interactions among antecedents to achieve more robust outcome, especially in the non-work domains (Chen and Tang 2021; Halme et al. 2020). Instead of focusing on a single quality alone, this approach explores the interaction of a variety of factors, allowing researchers to gain a more holistic view of organizational dynamics (Greckhamer et al. 2013).

Few recent studies have used fsQCA to examine the complex relationship of CSR dimensions; however, these scholars have mostly examined CSR dimensions according to multiple stakeholders, such as consumer, environment, and society, rather than employee-oriented CSR (e.g., Chen and Khuangga 2021; Halme et al. 2020). Research on internal CSR has primarily relied on systematic analysis which only examines the net effect of a single internal CSR initiative, regardless of its various initiatives that have been presented in the literature (Espasandín-Bustelo et al. 2020; Luu 2020; Mory et al. 2015; Thang and Fassin 2017). To address the research gap, the present research takes an asymmetrical approach through which a specific outcome (e.g., work–family spillover) can be explained by the joint effects of two or more antecedents (e.g., internal CSR initiatives). Adopting a symmetrical approach allows us to determine the best combinations of internal CSR activities for enhancing the positive spillover and reducing the impact of work-related demands on home responsibilities among employees.

This study examines the relationships between five internal CSR dimensions including internal dissemination, compensation, occupational health and safety, training, and legal employment and flight attendants’ work to family spillover via asymmetric approach. This study makes at least four contributions. First, by documenting work to family spillover as an outcome of internal CSR, this study differentiates it from most previous CSR studies which have focused on work-related outcomes (e.g., Ilkhanizadeh and Karatepe 2017; Luu 2020; Mory et al. 2015; Sanusi and Johl 2020). The present research expands the few studies which have presented CSR’s nonwork-related outcomes, such as life satisfaction (Golob and Podnar 2021) and well-being (Kim et al. 2018). To our knowledge, this is the first research that has examined the relationship between internal CSR and work–family spillover. Second, differentiating from previous CSR studies which have predominantly used traditional methods to examine the effects of CSR in general and internal CSR in particular on employee-level outcomes, this study adopts fsQCA to develop configurations of different initiatives of internal CSR. Accordingly, this study makes a methodological contribution to CSR research by determining combinations of internal CSR activities that allow employees to use work-related resources for fulfilling their home-related tasks, and combinations that reduce the negative impact of work-related demands on home responsibilities. Third, this study extends literature by presenting internal CSR as a source that employees gain organizational resources to promote their positive spillover and decrease their negative spillover from work to family. Fourth, this study provides airline companies with practical guidelines in enhancing positive spillover and addressing negative spillover from work to family domains among flight attendants by implementing internal CSR initiatives.

The remaining sections of this paper are structured as follows: this study begins with theoretical foundation and a review of conceptual literature of CSR and work–family spillover. The research model with the corresponding propositions, followed by the data collection and sampling method is presented. The description of research results is then presented and interpreted. The paper concludes with implications as well as potential directions for further research.

2. Literature Review and Proposition Development

2.1. Conservation of Resources Theory

Conservation of Resources (COR) theory has been one of the most influential theories that explain work–family relationships and human well-being (Hobfoll 1989; Oren and Levin 2017). COR theory argues that individuals attempt to acquire, maintain, preserve, and promote the resources they value, and have a natural propensity to prevent resource loss (Hobfoll 1989). COR theory also states that different resources are simultaneously obtained for the accomplishment of a single goal (Halbesleben et al. 2014). Environmental conditions (e.g., supportive work–family environment) and individual characteristics (e.g., affectivity, work identity, personal health) have been viewed as resources in the COR theory (Wayne et al. 2004). In line with these discussions of COR theory, internal CSR can be classified as an organizational resource that individuals possess for their work and family roles.

According to COR theory, when individuals spend an excessive amount of resources (e.g., time, energy) in one position, their investment in another role decreases, impeding their capacity to satisfy the expectations of the latter role (Oren and Levin 2017). Following this theoretical logic, employees with sufficient resources from their organizations can gain positive spillover from their work to family domains; in contrast, they experience negative spillover if they lack the resources necessary to meet expectations in both the professional and personal spheres. Prior research has presented that flight attendants’ behaviors and attitudes has been substantially influenced by a number of organizational resources associated with CSR (Chen and Tang 2021; Ilkhanizadeh and Karatepe 2017; Tang et al. 2020). In accordance with COR theory (Hobfoll 1989), this study proposes that internal CSR initiatives serve as primary enablers of positive and negative processes of work to family spillover among flight attendants.

2.2. Corporate Social Responsibility

Internal and external CSR are the two types of CSR. Internal CSR includes policies and programs to improve working conditions, programs to maintain psychological and physical health, and training programs to equip employees with social skills (Espasandín-Bustelo et al. 2020; Mory et al. 2015; Rupp and Mallory 2015). External CSR aims to enhance external stakeholders’ well-being through promoting community and regional development, implementing philanthropy and sponsorship, and increasing social and environmental practices (Hameed et al. 2016). Regardless of the critical role of employees as the main stakeholder, studies on CSR have paid more attention to external stakeholders rather than considering aspects of employees (Rupp and Mallory 2015). Similarly, in the context of airline companies, research has indicated that CSR activities mostly emphasized firms’ efforts to address external issues or satisfy external stakeholders’ demands rather than employee-focused concerns (Cowper-Smith and De Grosbois 2011; Okumus et al. 2020).

According to the World Business Council for Sustainable Development, CSR is defined as "the continuing commitment by businesses to behave ethically and contribute to economic development while improving the quality of life of the workforce and their families as well as of the local community and society at large” (WBCSD 2000). Companies known to be socially responsible businesses attempt to maximize profitable benefits and address non-work-related problems and challenges faced by stakeholders by employing different CSR initiatives. Scholars have largely examined the role of CSR in leveraging macro-level outcomes, such as firm reputation, firm performance, and environmental and social performance (Halme et al. 2020). In terms of micro-level outcomes, CSR is found to promote various work-related attitudes, such as organizational identification, organizational trust, and organizational commitment (Golob and Podnar 2021; Hameed et al. 2016; Hur et al. 2019; Mory et al. 2015) or work-related behaviors, such as proactive behavior, organizational citizenship behavior, strategic behavior, venture behavior, and innovative behavior (Hur et al. 2019; Luu 2020). Firms implement socially responsible activities as a modern approach to retain their workforce (Chen and Khuangga 2021). Organizational activities and practices cannot be manifested as CSR unless they can fulfill employees’ welfare (Gond et al. 2017). As a result, scholars have called for research that broadens CSR’s roles, which has been limited to work-oriented outcomes, into non-work benefits (Rupp and Mallory 2015).

Recent studies have indicated the positive effects of internal CSR on the life satisfaction and well-being of employees (Golob and Podnar 2021; Kim et al. 2018), which have suggested the importance of internal CSR in contributing to nonwork domains of employees. However, “internal CSR is a heterogeneous concept of which most researchers have only a vague idea” (Mory et al. 2015), which is a result of the lack of empirical studies investigating the influence of internal CSR on employees’ family lives (Barnett et al. 2020; Sanusi and Johl 2020). This gap in the literature calls for more research on CSR initiatives to explain non-work outcomes of internalized socially responsible initiatives rather than focusing on firms’ profitable benefits (Barnett et al. 2020).

2.3. Positive and Negative Work to Family Spillover

Positive work to family spillover reflects the extent to which gains from work domain (e.g., finance, values, moods, workplace resources, and skills) spill over to individuals’ family domain in ways that benefit the family domain (Grzywacz and Marks 2000). For instance, skills and knowledge experienced and learned through far-away-from-home travels provide individuals with practical lessons (e.g., time management skills) to teach their children how to become independent in pursuit of their dream career/ambitions. In contrast, negative work to family spillover occurs when demands from the work domain negatively impacts their performance of family-related responsibilities (Grzywacz and Marks 2000). For example, irregular and unscheduled work-related flights prevent flight attendants from taking care of their children.

Negative work to family spillover has been found as a predictor of negative outcomes, including a high level of job burnout, emotional dissonance (Giao et al. 2020), and high intention to leave the workplace among employees (Aboobaker and Edward 2019; Giao et al. 2020). In contrast, employees with high levels of positive work to family spillover are associated with positive outcomes, including a high level of genuine happiness, job satisfaction, and organizational commitment (Tang et al. 2020; Wayne et al. 2004) and lower intention to quit (Aboobaker and Edward 2019). Recent research demonstrated that enhancing positive spillover and reducing negative spillover from work to family domains simultaneously could increase employees’ job performance (Karatepe and Bekteshi 2008) as well as stimulate employees’ genuine happiness and well-being (Tang et al. 2020; Wayne et al. 2004). Therefore, researchers called for an integrated model which seeks synergies between work and family domain to not only avoid negative work to family spillover but improve positive one (Cinamon and Rich 2009; Weale et al. 2020).

Previous research has largely focused on establishing organizational-level determinants that can improve positive work to family spillover (Wayne et al. 2004) or buffer negative work to family spillover (Giao et al. 2020). However, it is much more important to explore organizational programs and policies designed to enhance positive spillover and reduce negative spillover between work and family domains at the same study (Cinamon and Rich 2009), particularly in the airline industry (Tang et al. 2020). In addition, recent studies identify some variables which simultaneously enhance the positive process and reduce the negative process of work to family spillover, such as perceived organizational support (Giao et al. 2020) and internal marketing (Tang et al. 2020). Regardless of the importance of socially responsible activities to employees’ family life (Guitián 2009), an extensive literature review shows no research conducted to explore the linkage between CSR and work–family spillover. Especially in Vietnam, work and family are equally important to employees because employees’ housework and professional work are inseparable and often interact with each other (Chen and Tang 2021; Phuong and Takahashi 2021). Indeed, prior research indicated that Vietnamese employees expect the inclusion of flexible working hours in employment contracts which allow them to spend time with their families (Phuong and Takahashi 2021). However, research on the effects of organizational levers on employees’ work–family spillover receives minor attention in Vietnam (Giao et al. 2020).

3. Preposition Development

This study conceptualizes a broader responsibility of corporate activities that provide employees with various resources including internal dissemination, compensation, occupational health and safety, training, and legal employment (Chou et al. 2021; Diaz-Carrion et al. 2019; Liu et al. 2021; Newman et al. 2016). First, internal dissemination refers to the process of sharing information between management to employees (Chang et al. 2022). The disclosure of corporate information (e.g., CSR strategies) through which employees can re-examine their preferences in comparison with organizations’ values, goals, and culture (Chou et al. 2021). Second, compensation refers to “the monetary and nonmonetary rewards in a transparent, nondiscriminatory manner according to the employee’s performance” (Diaz-Carrion et al. 2019, p. 3). Kim et al. (2020) argued that the provision of a good compensation policy as an initiative of CSR, can satisfy employees’ material needs of family life. Third, occupational health and safety includes activities that consider both the physical and the psychological well-being of the workforce beyond the workplaces’ regulations (Diaz-Carrion et al. 2019). Fourth, the training program refers to objective, transparent, and nondiscriminatory actions (Mory et al. 2015) with an aim of providing employees a better sense-making or awareness of the organization’s CSR actions (Liu et al. 2021). Last, legal employment refers to companies’ obligation to abide by regulations and rules in employment, for instance fixed working hours according to the labor law (Newman et al. 2016). Kim et al. (2018) argued that the adoption of legislation as an initiative of socially responsible activities can serve to satisfy employees’ family life, which in turn promotes their overall life. Overall, we propose that five dimensions of internal CSR are important in contributing to employees’ family well-being.

Internal dissemination of CSR was found to encourage employees’ organizational identification (Chou et al. 2021), whereas legal employment does not serve the same function (Newman et al. 2016). Chen and Khuangga (2021) indicated that the fulfillment of a single dimension of internal CSR alone, whether economic or health aspect, and even their combination was not sufficient to induce job satisfaction and lower turnover intention among employees. Hossen et al. (2020) also found that employee empowerment and employment stability resulted in job satisfaction whereas fulfilling training or health and safety by itself cannot do the same function. The health and safety or training program fosters frontline employees’ affective commitment to their organization, however, internal dissemination by itself is inadequate to create the same outcome (Thang and Fassin 2017). Ilkhanizadeh and Karatepe (2017) indicated that the combination of four different CSR dimensions, including economy, legal, ethical, and philanthropy, stimulates job satisfaction and work engagement among flight attendants. Based on these discussions, we argue that a single internal CSR dimension is not sufficient on its own to generate a specific outcome.

Studies revealed that internal CSR dimensions create mixed impacts upon the positive and negative processes of work to family spillover (Karatepe and Bekteshi 2008; Mao et al. 2021). Relationships with supervisors and co-workers can only significantly increase positive spillover but cause no impact on negative spillover, whereas family-related feedback significantly eliminates the degree of negative spillover but creates no effect on positive spillover (Karatepe and Bekteshi 2008). In addition, employees’ perception of more work demands generated a greater degree of negative spillover but did not impact their positive spillover, whereas perceived more skill discretion at work predicted positive spillover but unrelated to negative spillover (Stoiko et al. 2016). Internal dissemination was portrayed as an effective organization-level tool to encourage employees’ attitudes and behaviors favorable to the organizations (Chang et al. 2022), which was presented as an effective instrument to decrease flight attendants’ negative spillover and enhance positive spillover (Tang et al. 2020). However, the regular communication caused negative impacts on employees’ abilities to meet their home responsibilities (Wang 2017). To implement an effective cross-domain communication channel (e.g., delivering information from workplace to home), training programs should be presented as an additional resource to minimize the negative effects of information dissemination (Wan et al. 2019). Accordingly, this study argues that the synergy between an individual’s work and the family domain results from compounding and combining these internal CSR aspects rather than a single dimension.

Previous studies have used COR theory as a theoretical foundation in explaining how perceived CSR influences employee reactions (e.g., Liu et al. 2021; Mao et al. 2021). For example, Mao et al. (2021) suggested that the distribution of CSR-related activities, such as training and accommodation provision were shown to stimulate positive moods such as efficacy, hope, resilience, and optimism among employees. Self-efficacy, positive, and energetic have been proposed as individuals’ psychological resources in increasing positive spillover and reducing negative spillover between work and family domains (Wayne et al. 2004). Prior research also found that the remaining CSR-related activities such as salary paid according to individual performance, voluntary pension insurance, social and health insurance, and annual vacation benefits enhance positive process and reduce the negative process of work–family spillover among flight attendants (Tang et al. 2020). Previous studies, based on COR theory, argued that positive work to family spillover occurs among employees who acquire sufficient resources whereas negative spillover is an outcome of those who cannot access these resources (Oren and Levin 2017). In other words, it is proposed that both the appearance and absence of a single internal CSR dimension is important to both positive and negative processes of work to family spillover. Based on these above discussions and COR theory, we formulate two propositions:

Proposition 1.

The presence of a single internal CSR dimension alone is not sufficient to result in a high level of positive work to family spillover.

Proposition 2.

The absence of a single internal CSR dimension alone is not sufficient to cause a high level of negative work to family spillover.

Figure 1 illustrates the conceptual model.

4. Data Collection and Sample

A questionnaire for this current study was developed through a literature review, in which all items stemmed from empirical studies. Following the translation procedure recommended by Brislin (1980), the English questionnaire was then translated into the Vietnamese version and vice versa by two different academics with majors in tourism and hospitality management. A pilot study was conducted on 35 employees working in hospitality firms. Based on recommendations from two academics and a pilot sample, we modified the Vietnamese version to present the intention and spirit of the original version. Authors contacted the heads of the human resources department in airline companies to present the purpose of the study and ask for the permission of collecting data. After obtaining their approval, a total of 165 surveys were sent out to flight attendants. All flight attendants had to have worked for airlines currently operating in Asian countries for more than six months in order to ensure the certain quality level of rated items (Tang et al. 2020). To increase the response rate, respondents were informed that their individual responses would be kept anonymous as well as personally identifying information being excluded. Of the 165 distributed questionnaires, 136 valid questionnaires were received, yielding an overall response rate of 82%. Of the study participants, 92% were female and 62% were unmarried (38% married). The majority of respondents (45.5%) had organizational tenure of 1–3 years, followed by a group with 3–6 years (36%) and a group with more than 6 years of organization tenure (10.5%). Nearly half of participants (48%) were aged from 25 to 35, followed by 35–44 (32%) and 45–54 (12%).

5. Measures

5.1. Internal CSR

All items for internal CSR were adopted from prior studies (e.g., Espasandín-Bustelo et al. 2020; Chou et al. 2021; Diaz-Carrion et al. 2019; Luu 2020; Newman et al. 2016; Thang and Fassin 2017). Items were categorized into five dimensions, including (1) legal employment (four items), e.g., My working hours in my company do not exceed the maximum that the labor law permits; (2) compensation (three items), e.g., I am delighted with salary policies in my company; (3) internal dissemination (three items), e.g., Company holds frequency dialogues to communicate its organizational message well; (4) health and safety at work (four items); e.g., Company provides medical insurance for employees to minimize possible risks associated to their work; (5) training development (three items), e.g., Company provides employees with ongoing training to obtain necessary skills. All items were rated on a 7-Likert scale ranging from 1 “strongly disagree” to 7 “strongly agree”.

5.2. Work–Family Spillover

The study used the scale that was developed for the National Survey of Midlife Development in the United States (MIDUS) by Grzywacz and Marks (2000). The scale comprises of eight items, four to measure positive work to family spillover, e.g., Skills you use in the workplace are useful for household tasks; and four to measure negative spillover, e.g., Stress at work makes you irritable at home. This scale was validated with high reliability in prior research conducted in the airline industry (e.g., Tang et al. 2020). Participants were asked to rate all items on a scale ranging from 1 (“never”) to 7 (“frequently”).

6. Construct Validity and Common Method Variance

We conducted confirmatory factor analysis (CFA) to assess validity and reliability to validate the conceptual framework further. In this study, standardized factor loading ranged from 0.57 to 0.98, higher than the recommended threshold value of 0.5 (Nunnally and Bernstein 1994). This indicated that the study achieves reliability. Construct validity was examined in both convergence and divergence (see

Table 1. Correlation and validity among constructs.

Construct.	AVE	MSV	ASV	1	2	3	4	5	6	7
1.Internal dissemination	0.689	0.666	0.337	1
2. Compensation	0.600	0.130	0.075	0.24 **	1
3. Health and safety	0.740	0.666	0.350	0.74 **	0.27 **	1
4. Training	0.771	0.476	0.226	0.61 **	0.36 **	0.62	1
5. Legal employment	0.810	0.619	0.326	0.67 **	0.17	0.72	0.56 **	1
6. Positive spillover	0.526	0.373	0.159	0.36 **	0.11	0.48	0.24 **	0.57 **	1
7. Negative spillover	0.655	0.100	0.049	−0.28 **	−0.03	−0.17	−0.17 *	−0.24 **	−0.09	1

Notes: ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed).

). Convergent validity was assessed by average variance extracted (AVE) values (Hair et al. 2019). AVE values, ranging between 0.526 and 0.81, are higher than the recommended threshold of 0.5, indicating the convergent validity of constructs. In addition, the discriminant validity is confirmed because AVE values, ranging from 0.526 to 0.810, are higher than both maximum shared variance (MSV) (0.100–0.666) and average shared variance (ASV) (0.049–0.337) (Fornell and Larcker 1981). Furthermore, most of the correlation coefficients among theoretical constructs are consistently lower than the acceptance value of 0.7. Overall, the results indicated the constructs were reliable and valid for further analysis.

Since the database was collected from a single source simultaneously, the potential of Common Method Variance (CMV) is a serious issue. Harman’s single latent factor approach in which all construct items were loaded on a single latent factor to deal with CMV (Podsakoff et al. 2003). One factor yielded poorer fit indices (CFI = 0.55, TLI = 0.50, IFI = 0.55, RMSEA = 0.19) than the theoretical seven factor model (CFI = 0.91, TLI = 0.89, IFI = 0.92, RMSEA = 0.09). The Chi-squared difference also pointed out that the theoretical model yielded a significantly better fit than the one factor model (Δχ² = 802.28; Δdf = 20; p < 0.001), suggesting that a single factor did not explain the majority of the variance in this model. Hence, CMV is not a problematic issue in this study.

7. Fuzzy-Set Qualitative Comparative Analysis

This study explores the asymmetrical relationships among five CSR dimensions and two processes of work–family spillover by using fsQCA 3.0 software developed by Ragin (2017). This methodology has recently received increased attention in general management research (Chen and Khuangga 2021) in general and tourism research in particular (Chen and Tang 2021). This method has advantages over traditional analysis methods since it examines interdependence among different attributes rather than a single one, which provides a better understanding of organizational phenomena in a holistic manner (Greckhamer et al. 2013). Particularly, the fsQCA results explore not only the presence but a negation of a single variable, which influence on the other variables and on desired outcome (Ragin 2017). In addition, fsQCA was designed to work effectively with samples of relatively small to medium size (Kraus et al. 2018).

The application of fsQCA involves Pappas and Woodside’s (2021) three main steps including data calibration, the establishment of all possible configurations (the truth table), and obtaining the configurations. First, in order to calibrate the variables, three original values of 7.0, 4.0, and 1.0 from the 7-Likert scale were coded into three breakpoints of fuzzy-set calibration with 0.95 (full membership), 0.5 (crossover point), and 0.05 (full non-membership), respectively. Second, the truth table was generated to exhibit the possible paths of five CSR predictors grouped to explain the high levels of positive spillover and negative spillover. The frequency and consistency were used to refine the truth table. As our sample is 136, the frequency threshold was set at 1 or 2 (Fiss 2011; Ragin 2017). Moreover, the consistency values of 0.85 and 0.80, which are above the minimum recommended point of 0.75 (Pappas and Woodside 2021; Ragin 2017), were used to shorten the truth table of work to family spillover. A total of 13 configurations with 63 cases and 7 configurations with 57 cases for positive spillover and negative spillover, respectively, are listed in

Table 2. The truth table for a high level of work–family spillover.

Internal CSR					Number of Cases	Outcome	Raw Consistency
Outcome: Positive Work to Family Spillover
Internal Dissemination	Compensation	Health and Safety	Training	Legal Employment
1	1	1	1	1	2	1	0.8740
1	1	0	0	0	1	1	0.8696
1	1	1	1	0	1	1	0.8517
0	1	0	0	0	2	0	0.8444
1	1	0	1	0	1	0	0.8399
1	0	1	1	1	10	0	0.8264
0	1	0	1	0	03	0	0.8230
1	0	0	0	0	1	0	0.8001
0	0	1	0	0	1	0	0.7955
1	0	0	1	0	3	0	0.7773
0	0	1	1	0	4	0	0.7765
0	0	0	1	0	1	0	0.7395
0	0	0	0	0	33	0	0.5644
Outcome: Negative Work to Family Spillover
0	1	0	0	0	2	1	0.8230
0	1	0	1	0	3	1	0.8137
0	0	1	1	0	4	0	0.7981
1	0	0	1	0	3	0	0.7710
1	1	1	1	1	2	0	0.7419
1	0	1	1	1	10	0	0.6656
0	0	0	0	0	33	0	0.6481

.

Third, the outcomes of the fuzzy-set analysis for work to family spillover, based on the refined truth table, are produced. The fsQCA results are presented into three solutions: complex, parsimonious, and intermediate. The intermediate solution was found to constitute qualified subsets compared to two remaining ones (Ragin 2017). Using both parsimonious and intermediate solutions provides readers with an understanding of core and peripheral conditions. The core conditions are considered part of both parsimonious and intermediate solutions, whereas peripheral conditions are eliminated in the parsimonious solution and thus only appear in the intermediate solution (Ragin 2017). The presence of both core and peripheral conditions can better clarify the essential component of specific combinations of causal conditions (Fiss 2011). A black circle refers to the present condition, a white circle refers to an absent condition, and the size of the circles represents if the conditions are core (large circles) or peripheral (small circles), blank spaces indicate the causal factors are either present or absent.

8. Results of fsQCA

Three configurations for high levels of work–family spillover, in which the five CSR dimensions are predictors, and two processes of work to family spillover are two outcomes are presented in

Table 3. Configurations for positive and negative work to family spillover.

	Outcomes
	Positive Spillover		Negative Spillover
Antecedents	Path 1	Path 2	Path 3
Internal dissemination	●	●	○
Compensation	●	●	●
Health and Safety	•	°	○
Training	•	○
Legal employment		°	○
Index:
Consistency	0.817	0.869	0.790
Raw coverage	0.670	0.557	0.600
Unique coverage	0.147	0.034	0.600
Coverage solution	0.705		0.600
Consistency solution	0.801		0.790

Notes: a Frequency threshold = 1; Consistency threshold = 0.85. b Frequency threshold = 2; Consistency threshold = 0.80. Core conditions are represented by large circles ● (presence) and ○ (absence); Peripheral conditions are represented by small circles • (presence) and ° (absence); Blank cells represent “Don’t care” conditions, which means that the path always leads to the outcome without regarding to the levels of the “don’t care conditions”.

. Two parameters, including consistency and coverage, were required to assess the fit of the configurations in this study. Consistency indicates the degree to which a casual combination of similar composition leads to an outcome, while coverage refers to the number of cases for which a configuration is valid (Ragin 2017). Overall solution consistency values, ranging from 0.75 to 0.80, are greater than the threshold value of 0.75. This result indicates a high percentage of causal paths of similar composition, which led to the same outcome value. The set-theoretic consistency for each configuration was from 0.790 to 0.869 and for the overall solution from 0.790 to 0.801, with all values being above threshold of 0.75 (Pappas and Woodside 2021). All causal recipes have overall solution coverage of 0.705 for high level of positive work to family spillover, and 0.600 for high level of negative spillover, both being above 40%, indicating that the configurations explain a large proportion of participants with high levels of positive and negative work to family spillover. Moreover, unique coverage values of three configurations, ranging from 0.34 to 0.60, are greater than zero (Ragin 2017), implying that the proportion of cases can be explained uniquely by these configurations. Together, all combinations presented in Table 3 are sufficient to cause high levels of positive and negative work to family spillover (Ragin 2017). Internal dissemination and compensation are found as the two core present conditions for the increased levels of positive spillover, since they appear in both the parsimonious and intermediate solution (Fiss 2011). Internal dissemination, health and safety, and legal employment are shown as the core absent conditions to cause high levels of negative spillover, while training is found as the core absent one leading to high levels of positive spillover.

For high levels of positive work to family spillover, two configurations, Paths 1 and 2, present combinations for which internal dissemination and compensation must be present, depending on how these two conditions combine with others. Hence, P1 was supported. Specifically, Path 1 shows that a high-level combination of internal dissemination, compensation, health and safety, and training programs can lead to positive spillover regardless of the absence of legal employment. This configuration explains the largest part of the employee group cases with high positive spillover (67 per cent). Path 2 revealed that the high level of positive spillover can be achieved with a combination of compensation and internal dissemination regardless of the absence of three remaining conditions. This configuration explains 55 per cent of the employees with high positive spillover.

Additionally, the fsQCA findings showed a configuration that leads to high levels of negative work to family spillover. Path 3 indicates the absence of internal dissemination, health and safety programs, and legal employment causes high negative spillover if employees receive a high degree of compensation. This result supports P2. This configuration explains 60 per cent of the employees with high negative spillover. To visualize the fsQCA results, Figure 2 illustrates sufficient conditions for high positive work to family spillover (Paths 1 and 2) and high negative work to family spillover (Path 3).

9. Discussion

Drawing on COR theory and a sample of flight attendants working for airline companies in Asian countries, this study employs fsQCA to empirically examine the interactive and combinatory role of five different internal CSR initiatives including internal dissemination, compensation, occupational health and safety, training, and legal employment on positive and negative work–family spillover. This study discovered two configurations for high level of positive spillover and one configuration for high level of negative spillover. The results show that the two core conditions that best distinguish configurations that lead to a high level of positive spillover for flight attendants are compensation and internal dissemination. This finding is in line with previous findings of Kim et al. (2020) and Chen and Tang (2021) that the provision of a good compensation policy and internal communication across employees can satisfy employees’ material needs for their family life and enhance positive work to family spillover. Due to the higher raw coverage than the combination of compensation and internal dissemination, the coexistence of health and safety and training programs was found to cover more samples. This finding implies that the simultaneous implementation of all internal CSR dimensions without legal employment create the greatest effects for achieving positive spillover among flight attendants. In light of these results, our proposition (P1) that the presence of a single internal CSR dimension alone is not sufficient to result in a high level of positive work to family spillover is empirically supported.

Nevertheless, the implementation of the compensation system itself was found to generate high levels of negative work to family spillover due to the absence of other conditions (internal dissemination, compensation, and legal employment). This finding provides support for our proposition (P2) that the absence of a single internal CSR dimension alone is not sufficient to cause a high level of negative work to family spillover among flight attendants. The finding is in line with findings of Tang et al. (2020) that compensation alone was not found to prevent flight attendants against work–home interference. While training is found as a peripheral condition for high levels of positive spillover, this is not presented among those conditions explaining negative spillover. This finding is in line with Tang et al. (2020) who claim that training programs were designed to provide flight attendants with skills, which can be applied in the family domain but are not useful to reduce negative work to family spillover.

10. Theoretical and Practical Implications

Most previous studies on internal CSR have predominantly focused on the work-related consequences (Chen and Khuangga 2021; Chou et al. 2021; Hur et al. 2019; Mory et al. 2015) and studies conducted in Vietnamese context are not exceptional (Luu 2020; Thang and Fassin 2017). In response to previous scholars who have called for research that broadens CSR’s importance, which has been limited to work-oriented outcomes into non-work benefits (Barnett et al. 2020; Kim et al. 2018), this study links internal CSR to employee perceptions of spillover between their work and family roles. This study is the first of its kind in exploring a theoretical and empirical perspective of the concept of internal CSR initiatives and their influences on both positive and negative work to family spillover. Findings offer insights into which factors of the internal CSR concept significantly explain the high levels of positive process and negative process of work to family spillover.

Existing CSR literature that has largely employed traditional methods which only examine internal CSR as whether a single-dimensional construct (Golob and Podnar 2021; Hur et al. 2019; Kim et al. 2018; 2020) or its specific dimension (Espasandín-Bustelo et al. 2020; Chou et al. 2021; Luu 2020; Newman et al. 2016; Thang and Fassin 2017) on shaping employees’ behavioral and attitudinal outcomes. However, these studies downplayed the interdependencies and combinatory role of different internal CSR practices. Similarly, few studies have examined the interplays among different resources in explaining the positive and negative spillover from a resource conservation perspective (Oren and Levin 2017). Hence, this study adopted asymmetric analysis to group these socially responsible initiatives into various combinations for achieving a mutually beneficial interaction between employees’ professional work and family life.

Moreover, the CSR initiatives adopted by national airlines are limited to published reports or shared documents on national airlines’ websites, which do not provide detailed information relating to relevant CSR activities adopted to address specific goals (Cowper-Smith and De Grosbois 2011; Okumus et al. 2020). Using primary data collected from flight attendants, our findings reveal the effectiveness of combining different internal CSR dimensions and their relevance to work–family spillover. In addition to the importance of compensation policies (e.g., flexible compensation plan in which employees can decide which parts of their salary are monetary and which are social benefits), in order to enhance negative work to family spillover among flight attendants, airline companies should not ignore other CSR dimensions, such as the establishment of communication channels for delivering organizations’ CSR values towards frontline employees in the process of implementing internal CSR programs. For airline companies that do not comply with the employee/labor laws and occupational regulations, training programs are necessary besides providing access to the healthcare system (e.g., analyses and monitors the health and safety risks that are associated with work, offering health and social insurance). In addition, to prevent flight attendants from suffering a high level of negative work to family spillover, human resource departments should focus on two main mechanisms. First, providing training programs and seminars in which cabin managers or chief pursers serve as role models through sharing their personal experience dealing with work and family-related problems that are beneficial to flight attendants. Hence, airline companies should carry out supervisory training sessions where cabin managers’ skills and ways of assisting or providing support for flight attendants’ personal and family lives can be learned, upgraded, and evaluated (Kelly et al. 2014). Moreover, the interactive training or seminars should be planned and organized to allow both cabin managers and flight attendants to discuss innovative working practices and processes to increase their control over their work time (Kelly et al. 2014). The second is subsidizing health insurance contributions for flight attendants who raise dependent family members (e.g., children and retired parents).

Our findings acknowledge that positive and negative processes of work to family spillover are two distinct concepts that are contributed to by unique organization-level resources. Indeed, different solutions with varied priorities on each component can influence two processes of work–family spillover. This finding is in line with prior studies, which found that positive spillover and negative spillover have distinct antecedents (Weale et al. 2020). Accordingly, airline companies should better design distinctive internal CSR activities to facilitate the positive process and reduce the negative process of work–family spillover. In the present study, internal dissemination and compensation are two core components to enhance positive spillover. However, the fulfillment of compensation leads increased negative spillover. The failure of compensation occurs due to the absence of other internal CSR dimensions including internal dissemination, health and safety, and legal employment. Communicating of values and making-sense of socially responsible activities to employees and financial rewards (e.g., performance-based compensation) given for socially responsible behavior allow flight attendants to enhance their positive spillover, whereas legal frameworks of employment (e.g., working hours according to employment contract, business operation according to rules and methods) are required to reduce negative spillover from their work to family domain.

11. Limitations and Future Research Directions

This study is subject to several limitations that provide fruitful directions for future research. First, the collection of data at a single point in time may limit the causal conclusion of present research findings. The designs of longitudinal research or cross-lagged research are recommended to further explicate the cause–effect relationship between internal CSR and two processes of work–family spillover. Second, the data were only collected from flight attendants working for airline companies based in Asian countries, which has limited the generalizability of research findings. To enhance the generalizability of the research findings to diverse populations, further examination of the present model should be conducted in other tourism and hospitality industries such as casinos. Prior research indicated that individuals have higher expectations of raising successful children in Asian countries rather than European ones (Wuyts et al. 2015), which means that parents in Asian countries should spend more resources for their children’s development. As such, it is worthwhile to design research in different countries, in terms of Asian countries vs. European ones. Third, this study indicates that internal CSR initiatives are organizational resources that enable two processes of work–family spillover. However, employees’ abilities to handle their daily life are contributed by their personal resources (Wayne et al. 2004). In the future, researchers could expand the current model by including personal resources (e.g., emotional intelligence, self-efficacy, and personal identity) which jointly interact with internal CSR resources to explain work–family spillover.

12. Conclusions

CSR has become a critical strategy for airline companies to strengthen their corporate image and reputation but its effects on flight attendants, especially their life outside of the workplace have been overlooked. Drawing on COR theory, this study employs fsQCA to examine asymmetrical relationships among five internal CSR dimensions (including internal dissemination, compensation, occupational health and safety, training, and legal employment) and two processes of work to family spillover among flight attendants. Differentiating prior CSR research that has predominantly examined its net effect, our findings reveal two configurations of positive spillover and one configuration of negative spillover. A range of important theoretical and practical implications arise from these fsQCA findings, offering critical research agendas for future investigations.