Does it pay off to offer family-friendly practices? Exploring the missing links to performance

Family-friendly practices (FFP) are viewed as an important tool to achieve equal opportunities and a supportive culture, with potential benefits for the organizational image and its employer branding. To date, however, there is an imperfect understanding of whether FFP affect firm performance and what are the mechanism of the possible association. This study tests a comprehensive model to investigate whether FFP affect firm performance and analyzes the role played by work–life balance (WLB) as a mediator between the provision of FFP and employee attitudes and firm performance. Using a sample of 724 public and private Portuguese companies and Partial Least Squares (PLS) regression, the results show that the availability of FFP indeed has a positive influence on WLB, organizational commitment, and job satisfaction, but it does not impact family satisfaction. Moreover, WLB mediates the relationship between FFP and firm performance. It is noteworthy that men and women do not perceive the impact of these practices the same way, with FFP having a greater impact on female job satisfaction than on their male counterparts. In addition, participants with lower levels of education and lower levels of annual income perceive a stronger impact of FFP on organizational commitment and WLB than the remaining participants. Men and participants with lower levels of education also perceive a stronger impact of WLB on family satisfaction.


Introduction
Family-friendly practices (FFP) have become increasingly widespread in the current context of organizations (Beauregard & Henry, 2009;Butts et al., 2013;Heikkinen et al., 2021;Kersley 2006;Masterson et al., 2021). These practices aim to answer the challenges that result from workplace demographic trends, such as increased female labor-market participation and the growing numbers of dual-earner couples and single parents in the workforce (Berg et al., 2003;Bodkin & Fleming, 2021;Butts et al., 2013;Poelmans & Sahibzada, 2004;Wood et al., 2003). However, to date, it is unclear whether firms benefit from FFP, and more empirical studies on the effects of these practices are necessary (Beauregard & Henry, 2009;Masterson et al., 2021;Nayak & Pandey, 2022;Schoen & Rost, 2021). On the other hand, some authors have noted that the process by which the provision of family-friendly practices affects employee attitudes and firm performance is still an under-researched topic (Allen, 2001;Beauregard & Henry, 2009;Chen et al., 2018;Kelly et al., 2008). In particular, there has been a limited discussion on the role played by work-life balance (WLB) as a mediator between the provision of FFP and employee attitudes and firm performance (Beauregard & Henry, 2009;Eby et al., 2005;Ferdous et al., 2021;Kelly et al., 2008).
The main aim of this study is to contribute to current knowledge on the effects of FFP provision and the mechanisms through which this provision affects employee attitudes and firm performance. More specifically, we aim to address two gaps in the literature. First, while previous studies have tended to focus either on employee attitudes or firm performance, we intend to test a comprehensive model including both types of effect. Second, we investigate not only direct effects on employee attitudes and firm performance but also indirect effects that allow for clarification of the mediating role of WLB.
With these objectives in mind, the ensuing sections are structured as follows. First, we present a literature review on FFP and their effects on WLB, employee attitudes, and firm performance. Based on this review, we propose the research model and the research hypotheses. Subsequently, we present the method used in the empirical study and the results obtained. Following the discussion of the results, we identify the theoretical and managerial contributions as well as the main limitations, and we offer some suggestions for future research.
The main theoretical approaches guiding the development of our research model are signaling theory (Spence, 1973) and social exchange theory (Blau, 1964). According to signaling theory, individuals use observable characteristics of organizations to derive inferences about unobservable characteristics. Several authors (Beauregard & Henry, 2009;Butts et al., 2013;Casper & Harris, 2008;Chen et al., 2018;Grover & Crooker, 1995;Onken-Menke et al., 2018) have noted that, in accordance with this theory, the provision of FFP may be interpreted as a sign of organizational concern for its employees or as evidence of a generally favorable treatment of employees. On the other hand, the norm of reciprocity is a core aspect of social exchange theory (Cropanzano & Mitchell, 2005). When perceiving organizational concern or favorable treatment, employees may feel obliged to reciprocate with positive attitudes and behaviors that are beneficial for the organization (Beauregard & Henry, 2009;Butts et al., 2013;Chen et al., 2018;Hornung & Glaser, 2010;Muse et al., 2008;Ngo et al., 2009;Talukder, 2019).
Using structural equations modelling, we test a comprehensive model of the effects of FFP provision employing a sample of 724 firms in Portugal. Given that these effects depend on the individual circumstances of employees (Masterson et al., 2021), we will also conduct multigroup analysis to test for significant differences in the relationships under study. The results indicate that WLB does have a mediating role in the relationship between FFP and firm performance and that gender and education level need to be considered when analyzing the effects of FFP on employee attitudes. The study, thus, contributes to an unveiling of the direct and indirect links between FFP and firm performance, and also serves to identify individual characteristics that HRM managers need to consider when setting up FFP policies.

Family-friendly practices
FFP are generally defined as practices that personally assist and integrate employees' work and family roles (Frye & Breaugh, 2004). These practices are often viewed as an important tool to achieve equal opportunities and a supportive culture, as well as to promote a non-discriminatory workplace between men and women (de Hoop et al., 2018;Wise & Bond, 2003). By offering these policies, a company promotes a caring and positive image of itself and could, in consequence, become more appealing to candidates (Casper & Buffardi, 2004;Gray, 2002;Lambert, 2000).
Studies on the effects of FFP can generally be grouped under those operationalizing FFP as a bundle of practices and those focusing on one of two categories: flexible work arrangements (including flexible schedules and telework) and dependent care assistance (including financial assistance and on-site child care) (Masterson et al., 2021). Several authors have argued that FFP tend to have a stronger positive effect on employees when presented in bundles rather than just one single practice (Bae & Goodman, 2014;Butts et al., 2013;Durst, 1999) because they can complement each other (Lee & Hong, 2011) and, thus, contribute to the perception of an overall organizational philosophy (Grover & Crooker, 1995;Perry-Smith & Blum, 2000).
In the literature, it is also possible to distinguish between studies focusing on the effects of the availability of FFP by the organization and those focusing on the use of FFP by employees. The meta-analytical work of Butts et al. (2013) indicates that FFP availability is more strongly related to employees' attitudes than FFP use. Moreover, it is noteworthy that some studies indicate that even individuals who have no intention of using FFP in the near future are still more attracted to a job in an organization that offers them (Casper & Buffardi, 2004;Rau & Hyland, 2002).
This study analyses the effects of the availability of bundles of FFP (including flexible arrangements and dependent care assistance) on employee attitudes and firm performance.

Work-life balance
The literature on the relationship between the work domain and the family domain identifies three different perspectives -negative, positive, and integrative (Carlson & Grzywacz, 2008). The negative perspective is usually linked with the concept of work-life conflict, which occurs when individuals' work experiences or demands interfere negatively with family life (Greenhaus & Beutell, 1985). The positive perspective is related to the concept of work-life enrichment, which occurs when involvement at work leads to benefits or resources, such as the development of skills or more positive moods, which may have a positive effect on the family (Greenhaus & Powell, 2006). The integrative perspective is related to the concept of WLB, which focuses on an individual's ability to meet the expectations and responsibilities in both work and family domains (Carlson et al., 2009). Work-life balance aims to help all workers to achieve a better fit between their professional and private lives (White et al., 2003), where individuals' circumstances are a major concern (Reiter, 2007). In this paper, we will focus on WLB, given that it integrates and develops the previous concepts (Carlson et al., 2009). However, we will also refer to the studies linking FFP with work-life conflict and work-life enrichment.
Although the topics of work-life enrichment and work-life balance have been less explored in connection with FFP, the literature indicates that FFP contribute to an enhancement of these variables. Lapierre et al. (2018) conducted a meta-analysis on the antecedents of work-life enrichment and found a positive relationship between FFP and work-life enrichment. The work of Chen et al. (2018) and Martinez-Sanchez et al. (2018) also indicate a positive association between FFP and work-family enrichment. Concerning work-life balance, some studies indicate a positive association with FFP (Feeney & Stritch, 2019;Hill et al., 2001Hill et al., , 2003. The available evidence indicates that FFP tend to reduce work-life conflict and enhance work-life enrichment and work-life balance. Although some studies have reported non-significant effects, it stands to reason to expect a positive association between the availability of FFP and WLB. Given that FFP aim to assist and integrate employees' work and family roles and to support the family roles outside the workspace (Frye & Breaugh, 2004;Masterson et al., 2021), it is arguable that the availability of these practices works as a bridge to achieve work-life balance. Moreover, in accordance with signaling theory, the provision of FFP may be interpreted by employees as symbolic of the firm's concern to facilitate the integration of work and family domains (Beauregard & Henry, 2009;Butts et al., 2013;Casper & Harris, 2008;Grover & Crooker, 1995). Therefore, we propose: H1: There is a positive relationship between the availability of FFP and WLB.

Employee attitudes
According to social exchange theory, when perceiving an organizational concern or favorable treatment, as is the case with FFP availability, employees may feel obliged to reciprocate with positive attitudes that are beneficial to the organization (Beauregard & Henry, 2009;Butts et al., 2013;Chen et al., 2018;Hornung & Glaser, 2010;Muse et al., 2008;Ngo et al., 2009;Talukder, 2019). Organizational commitment and job satisfaction are two of the most researched attitudes in connection with FFP (Butts et al., 2013;Masterson et al., 2021). Organizational commitment is a psychological state regarding the employee's relationship with the organization, which has implications for the decision to maintain membership of the organization (Allen & Meyer, 1990). Job satisfaction is a positive emotional state that results from the evaluation of one's work experiences (Locke, 1969(Locke, , 1976. Several meta-analytical works (Allen et al., 2015;Baltes et al., 1999;Butts et al., 2013;Gajendran & Harrison, 2007) and literature reviews (Beauregard & Henry, 2009;Masterson et al., 2021;Poelmans & Sahibzada, 2004) have documented that FFP provision is positively associated with organizational commitment and job satisfaction. Recent studies continue to indicate a positive effect of FFP on organizational commitment (Martinez-Sanchez et al., 2018;Oyewobi et al., 2022) and job satisfaction (Chen et al., 2018;Kröll & Nüesch, 2019;Martinez-Sanchez et al., 2018).
It is noteworthy that studies on FFP and attitudes have started to include not only attitudes towards the job or the organization, such as job satisfaction and organizational commitment, but also attitudes in the non-work domain, such as life and family satisfaction (Masterson et al., 2021). For the purpose of this study, we will focus on family satisfaction, conceptualized in a similar way to job satisfaction -that is to say, a positive emotional state that results from the evaluation of one's experiences in the family domain of life. Previous studies provide evidence of a positive association between FFP and family satisfaction (Brough et al., 2005;Frye & Breaugh, 2004).
Based on the propositions of social exchange theory and on the empirical evidence concerning the relationship between FFP and employee attitudes, we propose the following hypothesis: H2: There is a positive relationship between the availability of FFP and positive employee attitudes.
H2a: There is a positive relationship between the availability of FFP and organizational commitment.
H2b: There is a positive relationship between the availability of FFP and job satisfaction.
H2c: There is a positive relationship between the availability of FFP and family satisfaction.
In an attempt to understand the processes linking FFP and employee attitudes, some studies have tested mediational models. The majority of these studies investigated work-life conflict and provided evidence of its role as a mediator between FFP and job satisfaction (Fonner & Roloff, 2010;Gajendran & Harrison, 2007;Golden, 2006) and life satisfaction (Fiksenbaum, 2014;Azar et al., 2018) found that work-life conflict mediates between FFP and turnover intentions, a variable that is strongly connected with organizational commitment (Guzeller & Celiker, 2019). Although they did not test a mediation hypothesis, Frye and Breaugh (2004) also found that FFP predicted work-family conflict, which in turn predicted job satisfaction and family satisfaction. As far as work-family enrichment is concerned, Li et al. (2018) and Chen et al. (2018) found evidence that it mediates between FFP and job satisfaction. Concerning work-life balance, recent studies indicate that it mediates between FFP and organizational commitment (Choi et al., 2018;Oyewobi et al., 2022). Additionally, Ferdous et al. (2021) found that work-life balance mediates between FFP and turnover intentions which, as mentioned earlier, is a variable closely related to organizational commitment. This empirical evidence leads us to propose: H3: WLB mediates between the availability of FFP and positive employee attitudes. H3a: WLB mediates between the availability of FFP and organizational commitment.
H3b: WLB mediates between the availability of FFP and job satisfaction. H3c: WLB mediates between the availability of FFP and family satisfaction.

Firm performance
Research has shown that firms that implement FFP indeed benefit from it in a variety of ways. Firms that have implemented FFP are more likely to obtain financial gains when comparing the benefits versus the costs of introducing FFP (Bae & Goodman, 2014;Dex & Scheibl, 1999;Perry-Smith & Blum, 2000). To that end, there is evidence of positive repercussions, such as an increase in firm and labor productivity (Bae & Goodman, 2014;Clifton & Shepard, 2004;Glass & Finley, 2002;Lee & Kim, 2010;Ngo et al., 2009;Wood & de Menezes, 2010), firm and employee performance (de Menezes & Kelliher, 2017;Gray, 2002;Lee & Hong, 2011;Martínez-Sánchez et al., 2008;Perry-Smith & Blum, 2000;Stavrou, 2005), firm profitability (Lau, 2000;Lau & May, 1998;Lee & DeVoe, 2012), and even higher shareholder returns (Arthur, 2003). On the other hand, FFP are likely to be a source of sustained competitive advantage and attraction (Perry-Smith & Blum, 2000) and can achieve higher levels of market performance (Ngo et al., 2009). Nevertheless, some studies conclude that firms with FFP do not outperform firms with non-family-friendly policies (Bloom & Van Reenen, 2006;Preece & Filbeck, 1999). Other studies find a neutral relationship between FFP and performance. For example, Medina-Garrido et al. (2021) note that neither the existence nor the accessibility of FFP has a direct, positive impact on performance. Nevertheless, they find that the existence of FFP does have an indirect effect through the well-being generated by such policies, which show the beneficial effect of such practices. In a similar line, Shahzad et al. (2022) find that, despite no direct relationship on performance (or satisfaction), reducing the work-life conflict can be dependent on the level of supervisor support, and the absence of it increases burnout and turnover intention.
In light of the evidence, we will position our hypothesis in line with the studies that find a positive relation between FFP and performance because we assume that, when firms offer FFP that help employees to juggle family and work commitments, firms' performance will benefit. Therefore, we propose: H4: There is a positive relationship between the availability of FFP and firm performance.
Few studies have been conducted on the processes linking FFP and firm performance. We could find only two studies testing the mediating role of employee attitudes. Muse et al. (2008) found that organizational commitment mediates between FFP and performance. De Menezes and Kelliher (2017) found that job satisfaction and organizational commitment mediate between FFP and performance. Based on this evidence, we propose: H5: Employee positive attitudes mediate between the availability of FFP and firm performance.
H5a: Organizational commitment mediates between the availability of FFP and firm performance.
H5b: Job satisfaction mediates between the availability of FFP and firm performance.
H5c: Family satisfaction mediates between the availability of FFP and firm performance.
We were unable to find studies testing the mediating role of work-life balance on the relationship between FFP and firm performance. However, some inferences can be derived from previous studies that suggest that FFP reduce work-life conflict and help in improving firm performance (Garg & Aggarwal, 2018). Therefore, we propose: H6: Work life balance mediates between the availability of FFP and firm performance.
The research model in Fig. 1 depicts the relationships under study.
Given that the relationships under study depend on the individual circumstances of employees (Masterson et al., 2021), we will also conduct multigroup analysis to test for significant differences in these relationships. Thus, we will investigate the differences in gender, marital status, number of children, level of education, annual income, managerial position, and firm size.
Gender has received a significant scholarly interest but, despite the extensive research, the results are still mixed. For example, while some studies show that women are more dissatisfied or face greater challenges in reconciling family and work (Georgellis et al., 2012;Lyonette et al., 2007;Tan et al., 2022), other studies challenge this common perception and show that men and women report similar challenges when it comes to reconciling family and work spheres (e.g., Shockley et al., 2017). Others show that men struggle more with such reconciliation when fulfilling the role of fathers (e.g., McLaughlin & Muldoon 2014). In fact, recent evidence sug- gests that work-family conflicts are increasing, and the role of a supportive spouse or partner is pivotal in reducing the adverse effects of the strained part (Cerrato & Cifre, 2018;Liao et al., 2019;Pluut et al., 2018). Not surprisingly, marital status seems to matter -sharing responsibilities can alleviate the burden coming from the competing demands of work and family, especially when there are children involved (e.g., Bächmann et al., 2020;Eagle et al., 1997Eagle et al., , 1998Michel et al., 2011).
Men and especially women with higher levels of education and in higher positions (and arguably higher income) have more resources to achieve greater balance in all spheres of their lives (e.g., Dilmaghani & Tabvuma 2019). Not surprisingly, conflicts are more prevalent among employees with less resources (Allen et al., 2000;Carlson & Kacmar, 2000). Finally, large firms have more resources at their disposal and may provide a variety of different FFP -for example, flexible work times or flexible leave arrangements that may cushion the conflict between family and work (e.g., Nayak & Pandey 2022) and have benefits for both employers and employees (Shabir & Gani, 2020).

Method
A total of 724 complete responses were collected in 2018 through an online questionnaire, sent to 12,301 general company emails provided by Informa D&B, a database that comprises all public limited firms in Portugal. Thus, the response rate was approximately 5.89%. It was made clear that any member of the company could answer the questionnaire. This method of gathering information is efficient, easy, and low cost (Lefever et al., 2007), while preserving the anonymity of the respondents. Data was analyzed using Smart PLS 3.0 (Ringle et al., 2015).

Sample
The majority of respondents are female (64.5%) with a degree (80.2%), married or living together (70.7%) with children (60.5%) with an annual level of income lower than 30,000€ (72.8%) and holding a managerial position (64.5%). The average age of the respondents is 40.2 years. The great majority of the participants work in private (95.9%), micro, or small size (67.4%) non-family business firms (61.5%) that sell abroad (67.3%) ( Table 1).

Family-friendly policies.
Two items measure the availability of FFP. The first relates to dependent care support provided by the organization and is an additive index that includes issues such as the availability of kindergartens at work, financial aid for kindergartens, information services on kindergartens, and support for the elderly. The second item is also an additive index and includes issues related to flexible working arrangements available in the organization, such as flexible schedule, compressed work week, work from home, and part-time work. All issues were taken from the Family-Supportive Orga-nization Perceptions scale, originally developed by Allen (2001) and lately adapted and translated to Portuguese by Chambel and Santos (2009)). Similar to Chambel and Santos (2009), we excluded the two issues related to maternity and paternity license since both of them are mandatory under Portuguese law.
Work-life balance. The present study includes 6 items from the scale of Carlson et al. (2009), measured on a 5-point Likert scale, ranging from 1 "Strongly disagree" to 5 "Strongly agree". A sample item of this scale is: "I am able to accomplish the expectations that my supervisors and my family have for me". Job satisfaction.
To assess the respondents' perception of their job satisfaction, we used the Portuguese version of the Brayfield and Rothe (1951) survey, adapted and translated to Portuguese by Chambel and Pinto (2008). The scale has 5 items, measured on a 5-point Likert scale, ranging from 1 "Strongly disagree" to 5 "Strongly agree". A sample item of this scale is: "I feel very satisfied with my professional life".
Family satisfaction. Following Frye and Breaugh (2004), in order to assess the respondents' perceptions concerning their satisfaction with family, we used a 5-item scale, similar to the job satisfaction scale, where the word "job" was substituted by the word "family". Respondents were asked to choose the option that best described their situation using a 5-point Likert scale, ranging from 1 "Strongly disagree" to 5 "Strongly agree". A sample item of this scale is: "I feel very satisfied with my family".
Firm performance. Firm performance was measured using a subjective self-report scale of 9 items, developed by Wiklund and Shepherd (2003). Respondents were asked to compare their current company and its two main competitors over the last three years in nine different areas of performance (sales growth, revenue growth, growth in the number of employees, net profit margin, product/service innovation, adoption of new technology, product/service quality, product/service variety, and customer satisfaction), using a 5-point Likert scale, ranging from 1 "Much lower" to 5 "Much higher".
Organizational commitment was measured using the 6-item scale developed by Berg et al. (2003). Respondents were asked to select the option that best described their situation with a 4-point Likert scale, ranging from 1 "Strongly disagree" to 4 "Strongly disagree". A sample item from this scale is "I am willing to work harder than I have to in order to help this company succeed".

Results
Structural Equations Modelling (SEM) was used to test the hypothesis under study. We chose PLS (Partial Least Squares) to carry out data analysis given its value in exploratory research. It estimates a less restricted model -the composite factor model -and it is less prone to the consequences of misspecifications in subparts of the model because it adopts a limited-information approach (Henseler et al., 2014). Additionally, preliminary normality tests using the Kolmogorov-Smirnov test with Lillefors significance correction in the SPSS (Statistical Package for Social Sciences) show that a large majority of the variables included in this study do not follow a normal distribution, suggesting the need to use PLS (Ringle et al., 2015).

Descriptive analysis
Initial analyses with PLS showed cases of poor item reliability, leading to the exclusion of two items from the analyses. We decided to keep some items with loadings between 0.5 and 0.7, after analyzing the impact of deleting them in the composite reliability and in the AVE, as suggested by Hair et al. (2017). Table 2 shows the final items used for each construct, as well as their means and standard deviations.

Measurement validity and reliability
As shown in Table 2, all composite reliabilities for latent variables are above the acceptable internal consistency level of 0.7 (Hair et al., 2011). The standardized load-ings of indicators are all larger than 0.4 (Table 2), which also confirms indicator reliability (Hair et al. 2017).
Concerning convergent validity, we can observe ( Table 2) that the average variance extracted (AVE) by each latent variable exceeds the threshold of 0.5, indicating that the constructs are unidimensional (Fornell & Larcker, 1981). We also calculated bootstrap t-statistics of the indicators' standardized loadings to complement the analysis of convergent validity (Anderson & Gerbing, 1988). As Table 1 shows, they were significant at the 1% significance level. We can, therefore, conclude that the measurement model has a high convergent validity.
To analyze if there is discriminant validity, we compared the square root of average variances extracted (AVE) with the correlations for each pair of latent variables (Fornell & Larcker, 1981). As Table 3 shows, the square roots of the AVE are higher than the correlations for all pairs. Therefore, each latent variable shares greater variance with its own measurement than with other constructs, providing evidence of discriminant validity.
Considering that both independent and dependent variables were captured by the same response, there is the possibility of common method bias, which may affect the study's validity. Therefore, we used the full collinearity assessment approach of Kock (2015) to test for common method bias. All the variance inflation factor values (VIF) were lower than the 3.3 threshold, which indicates that the model is free from common method bias.

Model estimation results
Then, we proceeded with the analysis of the structural model in order to test our research hypotheses (Henseler et al., 2009). Using bootstrapping and pseudo t-tests, we analyzed the significance of the path coefficients. Since some of the path coefficients presented a t value above 1.96 (p < 0.05), they were deleted. Figure 2 depicts the final structural model. Table 4 shows all significant direct effects in the model and the effect sizes. FFP have a positive and significant impact on WLB (β = 0.150; p < 0.01) validating H1. FFP also have a positive and significant impact on Organizational Commitment (β = 0.260; p < 0.01) and Job Satisfaction (β = 0.230; p < 0.01), but not on Family Satisfaction and, therefore, H2 was only partially validated. Besides, FFP have a positive and significant impact on Firm Performance (β = 0.105; p < 0.01), validating H4.
To evaluate the explanatory power of the model, we analyzed the coefficient of determination (R 2 ) of the endogenous constructs   (Table 6). The blindfolding procedure was used to calculate Stone-Geiser's Q 2 to evaluate the predictive relevance of the model. As Q 2 > 0 for all the endogenous constructs   (Table 6), we conclude, therefore, that the model has predictive relevance (Hair et al., 2011).
To analyze the effects of demographic variables on the relationships found, we conducted multi-group analyses. Only two groups were considered for each variable: male and female for Gender; married and not married for Marital Status; child and no child for Parental Status; graduate degree and no graduate degree for Education; managerial position and no managerial position for Hierarchy Level; <30.000€ and > = 30.000€ for Annual Income Level, and < 50 employees and > = 50 employees for Company Size.
There is a significant difference for Gender in the relationship between FFP and Job Satisfaction (β female = 0.297, β male = 0.107; p = 0.009) and in the relationship between Work-Life Balance and Family Satisfaction (β female = 0.279, β male = 0.421; p = 0.031). Females perceive a greater impact of FFP on Job Satisfaction than males and the latter perceive a stronger impact of Work-Life Balance on Family Satisfaction when compared to their counterparts. There is also a significant difference for Education in the relationship between Work-life Balance and Family Satisfaction (β no−degree = 0.495, β degree = 0.302; p = 0.000). Participants without a graduate degree perceive a greater impact of Work-Life Balance on Family Satisfaction than those with a graduate degree. We also found a significant difference for Annual Income Level in the relationship between FFP and Organizational Commitment (β income<30.000 €=0.309, β income>30.000€ = 0.118; p = 0.023) and in the relationship between FFP and Work-Life Balance (β income<30.000 €=0.233, β income>=30.000€ =-0.048; p = 0.003). Participants with a lower annual income (< 30.000€) perceive a stronger impact of FFP on Organizational Commitment and Work-Life Balance than participants with a higher income. There are no differences for Marital Status, Parental Status, Hierarchy Level, and Company Size.

Discussion
The results show that FFP availability has a positive impact on firm performance. The results also provide evidence that FFP availability enhances positive employee attitudes. Thus, FFP provision is positively associated with organizational commitment and job satisfaction as suggested by the literature (e.g., Beauregard & Henry 2009;Masterson et al., 2021;Poelmans & Sahibzada, 2004). Although previous studies indicate a positive association between FFP and family satisfaction (Brough et al., 2005;Frye & Breaugh, 2004), we could not find a signification relationship between those variables. Thus, FFP availability explains job-related outcomes (organizational commitment and job satisfaction) but not family-related outcomes (family satisfac-  Table 6 Explanatory Power and Predictive Relevance tion). A possible explanation is that there are other relevant variables explaining family satisfaction in addition to FFP availability that were not considered in our model. We also found that employees' positive attitudes (organizational commitment and job satisfaction) mediate the relationship between FFP and firm performance, as already proposed by de Menezes and Kelliher (2017). On the other hand, the study reveals that WLB mediates between the availability of FFP and positive employee attitudes -namely, organizational commitment, job satisfaction, and family satisfaction. These results evidence the benefits for organizations in providing FFP, which help employees to cope with their family responsibilities. Employee attitudes and WLB may, thus, be considered as relevant links between FFP availability and firm performance.
FFP have a greater impact on female's job satisfaction than on their male counterparts. Although the existent literature has revealed mixed effects of FFP on gender (Masterson et al., 2021), given the fact that women tend to have more family responsibilities than men, it is understandable that working in an organization that provides FFP makes them feel greater satisfaction at work. FFP seem to have a stronger impact on organizational commitment and WLB for participants with a lower annual income. A possible explanation lies in the fact that people with lower incomes experience greater difficulty in paying for extra help and, therefore, they value the fact that their employer organization provides FFP, enhancing their WLB. Employees in this situation tend to reciprocate with organizational commitment, as proposed by social exchange theory (Blau, 1964). Male participants and participants without a university degree perceive a greater impact of Work-Life Balance on family satisfaction than women or participants with a university degree.

Theoretical contributions
Our study contributes to the management and business literature by exploring the missing links between the availability of FFP and firm performance. While previous studies have tended to focus either on employee attitudes or firm performance, we tested a comprehensive model that includes both types of effect.
Since, to date, it was still unclear whether firms benefit from FFP (Beauregard & Henry, 2009;Masterson et al., 2021), we show that WLB plays a mediating role between the availability of FFP and employee attitudes: organizational commitment, job satisfaction, and family satisfaction. To the best of our knowledge, this mediation role has only been previously tested for organizational commitment (Choi et al., 2018) but not for the other employee attitudes. Otherwise, WLB mediates the relationship between FFP and firm performance. Considering that this has been an under-researched topic (Allen, 2001;Beauregard & Henry, 2009;Chen et al., 2018;Kelly et al., 2008;Masterson et al., 2021), we hope that this study contributes by shedding some light on this subject.

Managerial contributions
The results of this study show that, indeed, it pays off to offer FFP. However, the impact of FFP on firm performance is not perceived in the same way by everyone, and not all links are conducive to performance. Women and men have different perceptions of the impact of FFP on performance, with women having higher job satisfaction because of it. Men perceive a stronger impact of WLB on Family Satisfaction. Managers interested in increasing employees' satisfaction should consider these gender differences when designing flexible benefits packages.
Likewise, employees with lower levels of education and lower levels of annual income have a stronger perception of the impact of FFP on Organizational Commitment and WLB. Managers should also consider offering FFP to employees with lower levels of education and income because they seem to value them and reciprocate with Organizational Commitment, which is closely linked to the issue of employee retention. Managers need to be aware that the policies may be adapted and tailored according to employees' characteristics in order to increase the benefits that they can extract from FFP.

Limitations and future research opportunities
The current study has some limitations that offer scope for future research. First, the model only explains 10.7% of family satisfaction. Therefore, future studies should consider including additional variables (e.g., self-core evaluations) that have proved to be relevant to family satisfaction (Boyar & Mosley, 2007). Second, in this study, we offer some links that affect, both directly and indirectly, the relationship between FFP and firm performance. Future studies could dig deeper into other unexplored links in this relationship -for example, to see if the relationship is different for different industries or business cycle phases. Moreover, the effects of individual characteristics, such as workload and the role that psychological and physiological conditions play in moderating the relationship between FFP and performance, could also be explored. For example, employees experiencing burnout or anxiety can see the merits of FFP differently due to the difficulties in reconciling their personal and private lives. Third, we suggest that future research on the topic should consider using objective measures of performance, given that one limitation of this study was the use of a self-report measure of firm performance. For example, the use of accounting measures, such as return on assets (ROA), or market-based measures, such as earnings-per-share or dividend payout, could add to our understanding of the impact on performance. Finally, we suggest the use of longitudinal studies, so that the effects of FFP availability may be studied in different time periods.
Despite these limitations, the study contributes to an unveiling of the direct and indirect links between FFP and firm performance, and also to the identification of individual characteristics that HRM managers need to consider when setting up FFP policies. Overall, the results suggest that FFP can affect performance, but one measure does not fit all. Thus, firms need to take into account the profile of their employees when setting up FFP policies that strive for balance between work and non-work spheres of employees' lives.