The effects of employee services on organizational commitment and intentions to quit

The Authors

Soo-Hoon Lee, Old Dominion University, Norfolk, Virginia, USA

Thomas W. Lee, University of Washington, Seattle, Washington, USA

Choon-Fong Lum, National University of Singapore, Singapore

Acknowledgements

The authors acknowledge the generosity of the National University of Singapore in providing a research grant, RP981026, for this study.

Abstract

Purpose – The purpose of this paper is to examine the effects of the provision of employee services on employees' organizational commitment and their intentions to quit as well as their underlying reasons.

Design/methodology/approach – The paper shows that questionnaires were administered at two organizations in Singapore to evaluate employees' attitudes resulting from the provision of employee services. Mediator regression analyses and structural equation modeling were used to test the relationships between the provision of employee services with construed external image of the organization, valence for employee services, organizational identification, organizational commitment, and intentions to quit.

Findings – The paper found that positive employee attitudes arising from the provision of employee services were the result of a positive construed external image of the organization. It also shows that, when employees perceived that outsiders viewed their organization positively, their level of identification with their organization increased.

Research limitations/implications – The paper did not find a significant relationship between employees' valence for employee services and organizational identification, implying that receiving tangible benefits was less important to employees than working for an organization that is viewed positively by outsiders.

Practical implications – The results in this paper suggest that employees' attachment to their organization is enhanced when they perceive that outsiders view their organization positively.

Originality/value – The results in the paper were explained in terms of social identity theory, which suggest that firms must continually invest in socially responsible activities and practices to create a positive corporate image. Employees' perceptions pertaining to how stakeholders view their organization affect their work attitudes.

Article Type:

Research paper

Keyword(s):

Benefits; Identification; Singapore.

Journal:

Personnel Review

Volume:

Number:

Year:

2008

pp:

222-237

Copyright ©

Emerald Group Publishing Limited

ISSN:

0048-3486

Introduction

Traditional employee benefits, such as healthcare plans, retirement plans, or insurance plans, are generally provided to all employees within an organization, irrespective of their work performance, to meet their security and safety needs (Lucero and Allen, 1994). Employee services, however, comprising childcare services, health club memberships, and financial assistance programs, are innovative employee benefits provided to employees to help them integrate their work and family responsibilities (Butler et al., 2004; Fry and Breaugh, 2004). While past studies have examined if the provision of employee services enhanced employees' level of attachment to their organization (Baughman et al., 2003; Kopelman et al., 2006; Thompson et al., 2004), the underlying theoretical reasons for such relationships have not been investigated nor explained.

In view of limitations in past research, this study sought to contribute to existing literature by filling this theoretical gap. We investigated the underlying reasons for the relationship between the provision of employee services and organizational attachment in two steps. First, we examined if organizational identification, a construct related to the extent to which an individual defines himself/herself by the same attributes as the organization, provided a theoretical explanation for the relationship between the provision of employee services with organizational commitment and intentions to quit. Next, we examined two possible reasons for the relationship between the provision of employee services and organizational identification by comparing a tangible outcome that employees would derive from using employee services against an intangible outcome from being associated with an organization that outsiders view positively for providing such services. This study makes a managerial contribution by helping human resource professionals understand how employee benefits programs enhance employees' attachment to the organization.

Conceptual background

Employee services

Anecdotal evidence have shown that employees want to work for companies that recognize the importance of the home, family, and personal time because the difficulties of balancing work and family responsibilities often result in increased work stress (Gutek et al., 1988). Many working parents, for example, have had to leave their jobs for varying periods of time when they had no organizational support (Glass and Estes, 1996). However, when employees were provided with employee services, such as childcare services, eldercare services, wellness programs, or employee assistance programs (EAPs), they managed their work and family responsibilities better and had more positive work attitudes (Butler et al., 2004; Kopelman et al., 2006; Sinclair et al., 2005; Thompson et al., 2004).

Specifically, in the example of caregiving responsibilities, Rothausen et al. (1998) found that organizations that had on-site childcare facilities had more satisfied employees and higher employee retention rates. However, organizations that were unable to help their employees shoulder eldercare obligations had significantly higher employee absenteeism and intentions to quit, lower work productivity, and more employees with health problems because of caregiving stress (Shoptaugh et al., 2004). In the case of wellness programs, companies actively promote healthy lifestyles because employees face many stress-related illnesses resulting from increased work pressures, workload, job insecurity, and changing work requirements (Wolfe and Parker, 1994). Companies have implemented wellness programs, such as exercise and fitness, smoking cessation, weight control, and stress management programs (Hopkins, 1997), to control escalating health care costs. A positive outcome arising from wellness programs has been higher employee retention because employees have more energy to face stressful work situations (Daley and Parfitt, 1996). Also, anecdotal evidence have shown that EAPs, which are formal intervention programs provided to employees to help them cope with marital or family difficulties and manage problems related to mental health impairment or legal and financial problems (Hopkins, 1997; Sinclair et al., 2005), have been provided by companies because employees who bring their personal problems to work tend to have lower work quality and quantity as well as increased absenteeism (Milne et al., 1994). Organizations that provide employees with EAPs have benefited from improved employee morale, increased productivity, decreased absenteeism, and lower turnover rates (Thompson et al., 2004).

In summary, extant research show that employee services enhance employee retention as employees improve the balance and integration between work and family responsibilities (Fry and Breaugh, 2004; Thompson et al., 2004). Therefore, as shown in Model A of Figure 1, we hypothesize that employees have higher levels of organizational commitment and less intentions to quit when firms provide employee services.

H1a. The provision of employee services is positively related to organizational commitment.

H1b. The provision of employee services is negatively related to intentions to quit.

Nested mediating effects

Organizational identification. The first aim of the study, as discussed previously, sought to examine the relationships between the provision of employee services with organizational commitment and intentions to quit. The second aim of this study sought to examine the underlying reasons for these relationships. We chose social identity theory (Tajfel and Turner, 1985) to explain attitudes pertaining to organizational commitment and intentions to quit. Individuals acquire a social identity when they join a social group and their self-esteem is enhanced when their group compares more favorably with other out-groups (Tajfel and Turner, 1985). Organizational identity is a subtype of social identity in which an individual defines himself/herself by the same attributes that defines his/her organization (Dutton et al., 1994; Mael and Ashforth, 1992). Past research have shown that individuals tend to identify strongly with organizations that have distinctive and positive practices, values, or attributes to enhance their self-esteem, perceived status, and exclusivity (Albert and Whetten, 1985; Ashforth and Mael, 1989).

Many organizations do not provide employee services because of the high costs to do so. Those that do provide employee services enhance their level of attractiveness as an employer because employees who work for a small subset of such firms often feel an accentuated level of distinctiveness (van Knippenberg and van Schie, 2000). When employees perceive that their organization's practices are distinctive and positive, their level of identification with the organization increases (Dutton and Dukerich, 1991; Dutton et al., 1994). Accordingly, the provision of employee services enhances employees' level of organizational identification.

Employees who have high levels of organizational identification have enhanced feelings of belongingness to their organization and are more psychologically attached to it. Because they have a perceived shared destiny with their organization (Dutton et al., 1994; Wan-Huggins et al., 1998) and see the organization to be a part of who they are, they will likely choose to remain with the organization (Dutton et al., 1994). Therefore, employees who have high levels of organizational identification can be expected to more committed to the organization and have lower turnover intentions because they would experience a “psychic loss” if they leave their organization (Mael and Ashforth, 1992; van Knippenberg and van Schie, 2000). Therefore, as shown in Model B of Figure 1, we hypothesize that the provision of employee services enhances organizational commitment and reduces intentions to quit because employees identify more with such organizations.

H2a. Organizational identification mediates the relationship between the provision of employee services and organizational commitment.

H2b. Organizational identification mediates the relationship between the provision of employee services and intentions to quit.

Construed external image. Next, we examined two possible reasons for the relationship between the provision of employee services and organizational identification. As shown in Model C of Figure 1, one reason may be explained by employees' positive construed external image of their organization for having innovative human resource management practices. The construed external image of the firm represents how the employee thinks outsiders view his/her organization (Carmeli and Tishler, 2004; Dutton et al., 1994). Employees interpret and assess their organization's external prestige by comparing its practices that are considered central, enduring, and distinctive with other firms. When employees believe that their organization is perceived positively by outsiders for having socially-valued characteristics, they feel proud to belong to and would identify with their organization to enhance their self-esteem (Dutton et al., 1994; Smidts et al., 2001; Wan-Huggins et al., 1998). Thus, Mael and Ashforth (1992) found that a sample of alumni had higher levels of identification with their Alma Mater when they perceived that outsiders assessed their institution positively.

No past studies have explicitly examined if the provision of employee services enhances employees' construed external image of their firm, representing a gap in existing literature. A firm that provides employee services not only goes beyond providing the minimum legally mandated requirements and traditional benefits, it is adopting an innovative organizational practice that is not offered by every organization. In improving employees' welfare by helping them balance work and family responsibilities, the organization boosts its reputation as a family-friendly, socially responsible employer (McWilliams and Siegel, 2001). Thus, we posit that the provision of employee services will have a positive effect on employees' level of organizational identification because it enhances the construed external image of the firm as having a distinctive human resource management practice.

H3a. Construed external image of the firm mediates the relationship between the provision of employee services and organizational identification.

Valence. An alternative reason for the relationship between the provision of employee services and employees' level of organizational identification is explained by employees' valence for such services (shown in Model C of Figure 1). Valence is a measure of an individual's utility, desirability, salience, attractiveness, or value for a specific outcome or practice (Vroom, 1964). An employee who places greater importance for employee services has higher valence for such services. Currently, no studies have tested the relationship between valence for an organizational practice with organizational identification, providing us with an opportunity to add to existing literature. Given that employees tend to use those benefits that they value more frequently (Wilson et al., 1985), they would be more attached to their organization if they cannot obtain the same benefits in another firm. As few organizations provide employee services, we hypothesize that those employees who highly value employee services would have higher levels of organizational identification with organizations that provide employee services.

H3b. Valence for the employee services mediates the relationship between the provision of employee services and organizational identification.

In summary, this study served two purposes. First, it examined the relationships between the provision of employee services with employees' level of organizational commitment and their intentions to quit. Second, it investigated the theoretical reasons underlying these relationships.

Methodology

Sample and procedure

More than one organization is needed in organizational identification research because this construct involves comparison among organizations over a group-level attitude (Kopelman et al., 2006). Thus, we attempted to obtain at least two organizations to participate in our study by sending letters to 32 human resource managers in Singapore in 1999 whose firms were highlighted in the media that year to be among the first in the country to provide employee services. We chose Singapore for two reasons. First, the provision of employee services was unheard of in Singapore firms until that year. Second, the high female labor force participation of 56.3 per cent in 1998 (ILO, 2003) made employee services a potentially important means to help working women balance their work and family responsibilities. Hence, the setting provided an opportunity to tap into employees' initial reactions toward employee services.

Of the two organizations that agreed to participate in the survey, the first was an electronics manufacturing firm that employed over 8,000 employees in 11 plants. The second was a housing and construction firm that employed just under 9,200 employees in seven business units. In both firms, over 50 per cent the work force were foreign workers who had come to Singapore without their families to work on short-term contracts in low-skilled assembly-line or construction work respectively. These individuals were not given employee services. Among the employees in the manufacturing firm who were offered employee services, they included the skilled technicians, professional staff such as mechanical engineers, and management. Among the employees in the construction firm who were offered employee services, they included the building and project supervisors, professional staff such as civil engineers, and management. Since past studies limited generalizability by considering only one subgroup of workers, e.g. married employees or those with children (Boyer et al., 2003), we overcame this limitation by sampling a diverse mix of permanent, professional staff (van Dick et al., 2004). Because professional employees have greater market value and more job opportunities, focusing on professional staff as the sampling frame enabled us to test if the provision of employee services enhanced their level of organizational attachment (Jandeska and Kraimer, 2005). In both firms, the human resource managers specified the number of questionnaire surveys they would distribute on our behalf to their employees in one of their plants or business units.

The respondents mailed the completed survey anonymously to one of the authors in a self-addressed stamped envelope. A total of 212 individuals out of 290 responded to the questionnaire. In total, 37 out of 55 employees from the manufacturing firm participated in the survey, yielding a response rate of 67.3 per cent, while 175 out of 235 employees from the housing and construction firm participated in the study, yielding a response rate of 74.5 per cent. Despite unequal sample sizes, t-tests between the respondents in the two companies, assuming non-equal variances, found no statistical differences in gender, education level, and number of dependents at p < 0.05. Hence, the two samples were pooled for further analysis.

Measures

Employee services. The manufacturing company provided employees with only wellness program comprising talks and campaigns related to smoking cessation, health fitness, and stress reduction, while the housing and construction firm provided employees with dependent care, where employees chose between subsidized childcare or eldercare, wellness program comprising smoking cessation workshops, exercise programs, and stress reduction programs, which included taichi, meditation, and deep breathing exercises, and EAPs comprising subsidized housing loans. The provision of employee services (Employee Services) was categorized as a dummy variable. Employees from the manufacturing company who were provided with only one type of employee services, namely, wellness program, were coded as “0”, while employees from the housing and construction company who were provided with three types of employee services, namely, dependent care, wellness program, and EAPs, were coded as “1”. By comparing differences in employee attitudes, we assessed if employees accentuated their firm's distinctiveness from the different levels of employee services that their organizations provided.

Construed external image. The respondents evaluated four Likert-scaled items on a scale of 1 (strongly disagree) to 5 (strongly agree) on their perceptions regarding the extent to which their organization is known to outsiders' for its concern for employees and its innovativeness in reengineering its human resource management processes to satisfy its internal customers. Specifically, the items were derived from Dutton and Dukerich's (1991) inductive field study and adapted from Riordan et al.'s (1997) corporate image scale in which the respondents assessed outsiders' perceptions that “employees in the organization are constantly encouraged to experiment to find new and better ways of doing things”, “the company is known for its strong employee relations in showing concern for employee welfare”, “many people are aware of the company's fair treatment towards women and the disabled in terms of employment and career opportunities”, and “management is known for innovations in reengineering and business processes to improve customer service or employee benefits”. Exploratory factor analysis using principal components analysis with varimax rotation yielded only one factor with an eigenvalue that is greater than 1.0 (eigenvalue=2.47). Discriminant validity (reported in the Statistical Analysis section) was high, showing that this construct is distinct from other organizational attachment constructs, namely organizational commitment and intentions to quit, as well as organizational identification. With a reliable Cronbach's alpha of 0.79, the mean value for these four items was calculated and labeled as Image.

Valence. Valence refers to the value that individuals place on employee services in meeting their needs (Sinclair et al., 2005). Given that each person has different needs, the respondents evaluated the level of importance to them in receiving childcare program, eldercare program, wellness programs, and EAPs on a scale of 1 (not at all important) to 5 (very important). Exploratory factor analysis using principal components analysis with varimax rotation yielded only one factor with an eigenvalue that is greater than 1.0 (eigenvalue=2.52). With a reliable Cronbach's alpha of 0.80, the mean value of these four items was calculated and labeled as Valence.

Organizational identification. Organizational identification was measured on a scale of 1 (strongly disagree) to 5 (strongly agree) using the six items from Mael and Ashforth (1992): “when someone criticizes the company, I would feel bothered by the criticisms”, “I am very interested in what others think about the company”, “when I talk about this company, I usually say ‘we’ rather than ‘they’”, “when the company is successful, I feel that it is my personal success”, “when someone praises this company, it feels like a personal compliment”, and “if a story in the media criticized the company, I would be affected emotionally”. Factor analysis using principal components analysis with varimax rotation yielded only one factor with an eigenvalue that is greater than 1.0 (eigenvalue=3.39). With a high Cronbach's alpha of 0.86, the mean value of these six items was calculated and labeled as Identification.

Organizational commitment. Organizational commitment was measured on a scale of 1 (strongly disagree) to 5 (strongly agree) using a shortened Organizational Commitment Questionnaire (Mowday and Steers, 1979). The seven items, representing the three components of organizational commitment pertaining to belief in the organization's values, willingness to exert effort, and desire to maintain membership, were: “I am willing to put in a great deal of effort beyond that normally expected in order to help this organization be successful”, “I talk about this organization to my friends as a great organization to work for”, “I would accept almost any type of job assignment in order to keep working for this organization”, “I find that my values and my organization's values are very similar”, “the organization really inspires the very best in me in the way of job performance”, “often, I find it difficult to agree with this organization's policies on important matters relating to its employees” (reverse coded), and “I really care about the fate of this organization”. Factor analysis using principal components analysis with varimax rotation yielded only one factor with an eigenvalue that is greater than 1.0 (eigenvalue=3.67). With a reliable Cronbach's alpha of 0.83, the mean value of these seven items was calculated and labeled as Commitment.

Intentions to quit. We chose intentions to quit as the outcome variable instead of actual turnover because leaving was not an option for most employees at the height of the Asian financial crisis in 1999 due to a lack of job alternatives during the severe economic recession (Carsten and Spector, 1987). However, individuals could still engage in other forms of organizational withdrawal, such as having thoughts of quitting when the economy improved (Boyer et al., 2003). We also chose intentions to quit because it is a measure of organizational withdrawal that is a strong predictor of actual turnover (Shoptaugh et al., 2004). Intentions to quit was measured on a scale of 1 (strongly disagree) to 5 (strongly agree) using the three items from the Michigan Organizational Assessment Questionnaire (Cammann et al., 1983) for this construct. Factor analysis using principal components analysis with varimax rotation yielded only one factor with an eigenvalue that is greater than 1.0 (eigenvalue=2.51). With a high Cronbach's alpha of 0.90, the mean of these three items was calculated and labeled as Quit.

Control variables. The respondents provided information on their gender (Gender), age (Age), job tenure (Tenure), highest educational level attained (Education), annual household income (Income) as well as the number of dependants in their family (Dependent). We chose these demographic variables as control variables because they are related to employees' need for employee services and attitudes pertaining to organizational commitment and intentions to quit. For example, employee services tend to be more important to women (Ingram and Simons, 1995), younger employees (Vanderkolk and Young, 1991), and individuals with many dependents (Bardoel et al., 1999) because these individuals typically have more dependent-care demands or responsibilities at home. Also, individuals from lower-income households prefer to work for organizations that help them pay for dependent-care so that they could remain on the workforce (Edin and Lein, 1997). In the relationships between these control variables and organizational attachment, men were found to have higher levels of organizational commitment in a Chinese sample (Chen and Francesco, 2000), while employees with more education had lower organizational commitment and greater turnover intentions because they had more job opportunities (Mathieu and Zajac, 1990). On the other hand, older employees and those with long tenure within an organization had less desires to change jobs and fewer job opportunities (Arnold and Feldman, 1982).

Statistical analyses

A confirmatory factor analysis was conducted for all the 24 questionnaire items (Anderson and Gerbing, 1988) to assess the unidimensionality of the five constructs, namely Image, Valence, Identification, Commitment, and Quit. The goodness of fit index (GFI), comparative fit index (CFI), and nonnormed fit index (NNFI) were 0.86, 0.97, and 0.97 respectively, indicating that the five constructs were distinct from each other. As the items were adopted from existing literature and the goodness of fit statistics were high, these measures were used for hypotheses testing.

Next, to test the relationships between Employee Services with Commitment and Quit as stated in H1a and H1b respectively, we used hierarchical regression analysis to partial out the effects from the demographic variables. Then, we used hierarchical mediator regression analysis to assess the mediating effects of Identification on the relationships between Employee Services with Commitment and Quit as stated in H2a and H2b respectively. A variable acts as a mediator when three conditions are met: the independent and outcome variables covary significantly, variations in the mediator significantly account for variations in the outcome variable, and when the mediator is added into the model, the relationship between the independent and outcome variables becomes non-significant (Baron and Kenny, 1986). Thereafter, we assessed the mediating effects of Image and Valence on the relationship between Employee Services and Identification as stated in H3a and H3b respectively.

Finally, we followed a nesting approach in structural equation modeling using LISREL 8.54 (Jöreskog and Sörbom, 1984) to compare the three models sequentially with one another (Anderson and Gerbing, 1988) as depicted in Figure 1. LISREL provides statistical information that can be used to assess the overall fit of the data to each model as well as indicate where the statistically significant relationships are within each model. In LISREL, we used three measures to assess the overall fit of the data to each of the models:

GFI.
CFI.
Root mean square error of approximation (RMSEA).

The GFI and CFI values should be close to one, while the RMSEA should be under 0.08 to indicate good overall fit (Hair et al., 1998).

Results

The correlations among the variables are shown in Table I. Age was highly correlated with Tenure (r=0.91, p < 0.01) and so Tenure was excluded as a control variable in the regression analyses. The variance inflation factors were less than 2.0, indicating no serious multicollinearity problems among the predictor variables.

Table II shows the hierarchical regression results pertaining to H1, H2 and H3. H1a and H1b predicted that the provision of employee services would be related to higher organizational commitment and lower intentions to quit respectively. Models 1 and 3 in Table II indicate that after controlling for Gender, Age, Education, Income, and Dependent, Employee Services was positively related to Commitment at p < 0.01 but negatively related to Quit at p < 0.05 respectively. Hence, H1a and H1b were supported by the data.

H2a and H2b predicted that organizational identification would mediate the relationships between the provision of employee services with organizational commitment and intentions to quit respectively. Models 2 and 4 in Table II show that after controlling for the demographic variables, Identification fully mediated both the relationships between Employee Services with Commitment and Quit at p < 0.01. When Identification was considered, the relationships between Employee Services with Commitment and Quit became statistically non-significant at p < 0.05. Hence, H2a and H2b were supported by the data.

H3a and H3b predicted that the relationship between the provision of employee services and employees' level of organizational identification were mediated by an enhanced construed external image of the firm and employees' valence for the employee services respectively. Model 5 in Table II shows that the relationship between Employees Services and Identification was statistically significant at p < 0.05. Model 6 of Table II shows that when Valence and Image were added, the relationship between Employee Services and Identification became statistically non-significant at p < 0.05. In addition, Image fully mediated the relationship between Employee Services and Identification at p < 0.01 but Valence was not a statistically significant mediator between Employee Services and Identification at p < 0.05. Thus, H3a was supported but H3b was not supported by the data.

The LISREL results for the relationships depicted in Models A, B, and C in Figure 1 are reported in Table III. Overall, the results in Table III indicate that the data had relatively good fit with the models as the GFI, CFI, and RMSEA were within acceptable ranges. Model A in Table III shows that Employee Services is positively related to Commitment but negatively related to Quit at p < 0.01, providing support for H1a and H1b respectively. Model B in Table III shows that Employee Services enhanced Identification, which enhanced Commitment and lowered Quit at p < 0.01, providing support for H2a and H2b respectively. Model C in Table III shows that Employee Services enhanced Image and Valence at p < 0.01 and p < 0.05 respectively. While an enhanced Image increased Identification at p < 0.01, Valence was not significantly related to Identification at p < 0.05, providing support for H3a but not H3b. In summary, Model C in Table III indicates that after controlling for the demographic variables, Employee Services enhanced Image, which increased Identification, leading to higher levels of Commitment and lower levels of Quit. The statistically significant relationships in Model C are shown in Figure 2, providing support for all the hypotheses except H3b.

Discussion and conclusions

The results of this study showed that the provision of employee services, a tangible benefit aimed at helping employees balance their work and family obligations, was positively related to employees' level of organizational commitment but negatively related to their intentions to quit. Hence, the results supported extant research, which showed that employee services enhanced employee attachment to and retention in their organization.

However, the key contribution of this study was to answer the question “why?” as no studies have tried to explore a theoretical reason for these relationships. We found support that one underlying reason for a significant relationship between the provision of employee services and attitudes related to organizational attachment is explained by concepts from social identity theory. The results indicated that the provision of employee services enhanced the image of the organization to outsiders that it cares about its employees and is innovative in its human resource management practices and processes. Given that employee services are not provided by every organization, firms that do so are perceived to be distinctive. This perception enhances employees' level of identification with their organization, resulting in positive attitudes. The results corroborate with the principle of reciprocity (Gouldner, 1960), which suggests that people tend to return in kind what they receive from another. When a firm endows its employees with benefits that enhance their work and family life, employees make a conscious effort to reciprocate the goodwill they receive by being more committed to the organization. The practical implication of this study suggests that firms must continually invest in activities that not only benefit their stockholders, customers, and other external business partners, but must have practices and processes that benefit their internal customers to enhance employees' work attitudes.

The study did not find significant relationships between employees' valence for employee services and their level of organizational identification. This represents an opportunity for future research because the non-significant results are also similar to Sinclair et al.'s (2005) findings in which they found no direct relationships between benefit importance and organizational commitment. The non-significant relationship may suggest that the tangible and immediate enjoyment of employee services are not as critical to employees as the strategic signal that the provision of employee services carries. In other words, employees may identify more with their organization for demonstrating that it cares about their long-term welfare and views them as an important stakeholder group as opposed to merely meeting their immediate needs. Our suggestion corroborates with Thompson et al.'s (2004) study in which they found that employees' perceptions of emotional support were significantly related to affective commitment while tangible family support practices had no significant relationship with commitment.

A limitation of this study is the assessment of intentions to quit rather than actual quits. Although economic uncertainties in Singapore during the time of the study limited the number of actual quits, following up on actual quit rates may provide a more robust model on the impact of employee services. Hence, future research should assess the impact of employee services on actual turnover data as well as its financial impact on the firm's bottom line. More research is also needed to identify other innovative, family-friendly or diversity-friendly human resource management policies that would enhance employees' work attitudes. A second limitation of this study arose from a two-firm comparison as opposed to having multi-firm comparisons, thereby constraining our ability to control for firm-level differences using hierarchical multi-level analysis.

In conclusion, firms that provide employee services help their employees balance their work and family responsibilities, thereby enhancing employee retention and preventing a premature loss of human capital in the workforce. Given that the industrial structure of most developed nations have transformed towards a knowledge-based economy, retaining knowledge workers and high-performing employees are vitally important for organizational success and global competitiveness (Jackson and Schuler, 1990; Offerman and Gowing, 1990). As higher levels of construed external image increase employees' level of organizational identification with their company, which in turn enhances employee retention and other desirable employee attitudes and behaviors, managers have even more reasons to engage in activities that augment the profile of the firm as a socially responsible corporate citizen.

Figure 1Models depicting the relationships between the provision of employee services with organizational commitment and intentions to quit

Figure 2Final model depicting the relationships between the provision of employee services with organizational commitment and intentions to quit

Table IMeans, standard deviations, and correlations

Table IIMediator regression analyses

Table IIISequential LISREL analyses and path estimates

References

Albert, S., Whetten, D.A. (1985), "Organizational identity", in Cummings, L.L., Staw, B.M. (Eds),Research in Organizational Behavior, JAI Press, Greenwich, CT, Vol. 7 pp.263-95.