Article body

In an era where skilled-labour shortages place immense pressures on organizations to attract and select the best employees (Lievens et al., 2002; Turban, 2001), a lot of attention has been devoted to understanding how to improve the attractiveness of a firm for job applicants (Chapman et al., 2005). For example, some studies have investigated the role of organizational characteristics, such as reputation or social responsibility, as a factor of attraction (e.g., Collins and Han, 2004; Fombrun et al., 2000; Greening and Turban, 2000; Neville et al., 2005; Turban and Cable, 2003). Others have looked into the role of human resources policies (e.g., Bretz and Judge, 1994; Cable and Judge, 1994; Chapman et al., 2005). However, among those policies, family-friendly policies (FFPs) have received relatively little attention (Breaugh and Frye, 2007). Yet, given the large proportion of women in the labour-market, as well as the growing number of both dual-career families and single-parent families (e.g., Bureau of Labor Statistics, 2005, 2007; Parasuraman and Greenhaus, 1997; Secret, 2005), balancing work and family has become a challenge for many employees (e.g., Duxbury and Higgins, 2003). In North America where, contrary to some European countries, few child-care services are provided by the State (Istace et al., 2004; Silvera, 2002), individuals turn to their employers to request FFPs. Similarly, parental-leaves are more regulated in Europe than in North America (Istace et al., 2004; Silvera, 2002), leaving North-American employers room to base their attractiveness on how generous they are when it comes to personal time off. However, the European Quality of Life Survey 2007 reveals that even in Europe, a good work-life balance is difficult to achieve (European Foundation for the Improvement of Living and Working Conditions, 2007), and many employers offer family-friendly practices to help their employees attain such balance or promote women’s participation in the workforce (Straub, 2007; Zabarauskaite, 2008).

Although some researchers have included FFPs policies in their investigation of organizational attractiveness (Bretz and Judge, 1994; Casper and Buffardi, 2004; Honeycutt and Rosen, 1997; Rau and Hyland, 2002; Rothausen et al., 1998), those studies focus on a limited number of policies, thus making it impossible to compare the effect of various FFPs on the ability for an organization to attract applicants. This study tries to fill that gap by providing an answer to the following research question: “What is the effect of practices designed to reduce work-family conflict on organizational attractiveness in Canada?”

Previous Research and Hypotheses

Work-Family Conflict and Work-Family Balance

Work-family conflict (WFC) is an interrole conflict which appears when the requirements of one role make it difficult for an individual to fulfill the requirements of the other role (Greenhaus and Beutell, 1985). Studies have shown that WFC has negative effects on individuals as well as on organizations (e.g., Frone et al., 1997a; 1997b; Parasuraman et al., 1996). Boyar et al. (2005) have shown that WFC increases absenteeism, which may, in turn, be very costly. As an example, Duxbury and Higgins (2003) estimate that absenteeism due to work WFC costs Canadian employers almost $3 to $5 billion a year.

In order to reduce WFC, some individuals try to reach an acceptable level of work-family balance (Clark, 2000; Greenhaus et al., 2003). However, this represents a challenge for individuals and organizations alike (Lockwood, 2003): while an individual has to learn to reconcile work-related and family-related requirements, organizations have to create a supportive culture in which employees can focus on their jobs once they are at work. In order to reach that goal, several companies have implemented so-called “family-friendly practices” to reduce WFC (Lockwood, 2003).

Family-Friendly Practices

FFPs are human resources practices designed to help employees alleviate the conflict between work and family-roles (Friedman, 1990). They usually fall into four categories: 1) support for child care, dependent care or elder care; 2) time off, including maternity, parental, personal, or family-leave; 3) employee-assistance programs and counselling; and 4) flexible work arrangements (Grover and Crooker, 1995; Lambert, 2000).

Studies have shown that FFPs have many positive effects on individuals, both users and non-users of the practices. They increase attitudes and behaviours such as attachment and loyalty to the organization, organizational citizenship behaviours, and job satisfaction (Breaugh and Frye, 2007; Grover and Crooker, 1995; Lambert, 2000).

Several theoretical explanations have been put forth to shed light on these results. The justice literature, together with social exchange and organizational support theories, all suggest that employers who implement FFPs appear fair in the way they distribute resources and supportive of the needs of their employees, prompting employees to reciprocate by adopting positive attitudes and behaviours (Aselage and Eisenberger, 2003; Grover, 1991; Grover and Crooker, 1995; Lambert, 2000). It should then seem natural that this positive perception of the organization spills over to job applicants. Yet, few studies have investigated the effect of FFPs on organizational attractiveness.

Family-Friendly Practices and Organizational Attractiveness

Bretz and Judge (1994) found that salary and promotion opportunities were the most significant predictors of job acceptation among candidates. However, they also discovered that candidates with a high level of WFC were more attracted to organizations with FFPs than candidates with a low level of WFC. This result was later confirmed by Rau and Hyland (2002) who found that WFC acts as a moderator between two FFPs (flexible scheduling and telecommuting) and organizational attraction.

Contrary to these results, Casper and Buffardi (2004) did not find a moderation effect of WFC between FFPs and organizational attraction. However, they found that flexible scheduling and dependent care assistance both had a positive effect on applicants’ intentions to pursue jobs with potential employers, regardless of their level of WFC. Their results also indicated that anticipated organizational support fully mediated the effects of work schedule flexibility and dependent care assistance on job pursuit intentions. Although they did not specifically investigate the effect of WFC, Honeycutt and Rosen (1997) found results comparable to Casper and Buffardi’s. In their study, career and policy flexibility were positively related to organizational attractiveness, regardless of the subjects’ gender and parental status. However, individuals who described themselves as family persons or as both family and career persons were significantly more attracted to organizations that offered flexibility than individuals who described themselves as career persons.

Rothausen et al. (1998) studied the effect of actually using a FFP. Their results indicated that employees who either use the organizational on-site child care centre, have used it, or plan to use it in the future, are more likely than the non-users to believe that this practice has a positive effect on recruitment and retention. This finding suggests that the effect of FFPs on organizational attractiveness depends on the applicants needs.

These studies indicate that FFPs may have an influence on applicant decisions. However, they present several limitations that make it difficult to compare the effect of various FFPs on the ability for an organization to attract applicants. First, several studies have investigated the effect of only one or two practices, thus failing to provide information as to which practices, among the set of dozens of FFPs, are more likely to attract applicants. Second, other studies have investigated the general effect of FFPs without distinguishing among particular practices. For example, Honeycutt and Rosen use the general phrase “flexible career path” without further explaining which practices are included. Although these studies provide valuable information on the effect of FFPs, they fail to identify which practices, if any, contribute more to organizational attractiveness. The study presented here endeavours to answer some of those criticisms by investigating the distinct effect of various FFPs on applicant attraction.

Based on previous research, we believe that organizations with FFPs are more attractive than those who do not offer such arrangements, which leads to our first hypothesis.

H1: Organizations that offer FFPs are more attractive to applicants than those who do not offer such practices.

However, since several FFPs may contribute to this greater attractiveness, we found it worthwhile to investigate the effect of each practice. For example, finding affordable and quality child care is one major concern for working parents (Duxbury and Higgins, 2003; Secret, 2005). Therefore, on-site child care may represent an interesting option for working parents as well as for individuals who plan to have children and continue to work (Rothausen et al., 1998). This leads us to the following hypothesis:

H1a: Organizations that offer on-site child care are more attractive to applicants than those who do not offer such a practice.

Offering personal time off, including maternity, parental or family-leave, constitutes the second category of FFPs. Although this practice is regulated by law, surveys indicate that offering personal leave above what is legally required is highly valued by employees (Lockwood, 2003). Therefore, we believe that a generous personal leave policy is a popular fringe benefit. This is captured in hypothesis H1b.

H1b: Organizations that offer personal leaves above what is legally required are more attractive to applicants than those who do not offer such a practice.

Finally, flexible work arrangements are among the practices that seem to have the greater impact on work-family balance (e.g., Hill et al., 2004; Lockwood, 2003; Rau et Hyland, 2002; Thomas and Ganster, 1995). Flexible work arrangements usually include flexible work scheduling, part-time work, and telecommuting, in order to provide employees with flexibility regarding either when or where the job will be performed. Indeed, Hill et al. (2004) and Thomas and Ganster (1995) found a relationship between flexible work hours and work-family balance.

A few studies have investigated the relationship between flexible work hours and organizational attractiveness. Among those, Rau and Hyland (2002) found a main effect of flextime on applicant attraction. This main effect was however moderated by role conflict, subjects with high role conflict being more attracted to organizations that offer flextime than subjects with low role conflicts. Similarly, Casper and Buffardi (2004) found that availability of flexible hours was associated with job pursuit intentions. However, contrary to their expectations, this relationship was not moderated by WFC.

Telecommuting, or teleworking, has also been associated with applicant attraction (Rau and Hyland, 2002). However, just as was the case for flextime, this main effect was moderated by role conflict, subjects with high role conflicts being more attracted to organizations that offer telecommuting than subjects with low role conflicts.

Although flexible work arrangements have seldom been studied as antecedents of applicant attraction, the indications that they are associated with positive work-related attitudes allow us to formulate the following hypotheses:

H1c: Organizations that offer flexible work scheduling are more attractive to applicants than those who do not offer such a practice.

H1d: Organizations that offer teleworking are more attractive to applicants than those who do not offer such a practice.

Moderating Effect of Desire for Segmentation

Despite the anticipated main effect of FFPs on applicant attraction, we cannot disregard the fact that several studies have found moderating effects of either social identity (Honeycutt and Rosen, 1997), family responsibilities (Scandura and Lankau, 1997) or interrole conflict (Bretz and Judge, 1994; Rau and Hyland, 2002). Boundary theory may provide an explanation to this moderating effect. This body of literature examines how individuals create and maintain boundaries to simplify and order time, space, and events, so as to make sense of them (Ashforth et al., 2000; Clark, 2000). Defining boundaries allows individuals to focus their attention on what is currently salient to them (Nippert-Eng, 1996). However, this strategy makes it difficult to transition from one role to another in order to address the specific demands arising from each of them (Ashforth et al., 2000). Therefore, attention has been devoted to the strategies that people develop to manage boundaries between roles.

According to Nippert-Eng (1996) and Ashforth et al. (2000), boundary management strategies can be arrayed on a continuum spanning from high role segmentation to high role integration. Role segmentation refers to the separation of roles (i.e., a low degree of permeability between roles), whereas role integration refers to the blurring of the boundaries between roles (Rothbard et al., 2005). Traditional work arrangements are characterized by strict temporal and spatial boundaries: employees are expected to work from 9 to 5, Monday through Friday, on their employers’ premises (Rau and Hyland, 2002). Therefore, in a traditional work arrangement, roles are rather segmented. FFPs tend to break these strict boundaries and move toward role integration. For example, teleworking allows employees to fulfill their work-related role at home, thus blurring the boundary between work and personal space.

Rau and Hyland (2002) and Rothbard et al. (2005) ordered some FFPs on the segmentation-integration continuum. According to them, flexible working hours lean towards the segmentation end of the continuum, because flextime promotes impermeable temporal and spatial boundaries between work and non-work related activities. Telecommuting, on the other hand, is an integrating strategy because it allows flexibility in both spatial and temporal boundaries (Rau and Hyland, 2002). Rothbard et al. (2005) added on-site child care on the integration end of the continuum because it represents a blurring of the work-family spatial boundary. Other FFPs have yet to be positioned along the integration-segmentation continuum. We posit that, similarly to flexible work scheduling, additional personal leave tends to be a segmentation strategy because it provides employees with extra time to attend to their personal obligations, while keeping the work-role clearly distinct.

Not only do boundary management strategies lie on the integration-segmentation continuum, but individual preferences also fall between integration and segmentation. Nippert-Eng (1996) and Hartmann (1997) found that individuals differ in the degree to which they segment or integrate their professional and personal roles. For example, people who tend to integrate both roles (i.e., integrators) might show pictures of their families to their coworkers or sometimes bring work home. On the other hand, people with a tendency to segment (i.e., segmentors) rarely discuss their family or personal problems with coworkers, and prefer to stay in the workplace longer rather than bring work home (Rothbard et al., 2005).

Depending on the strength of their desire for segmentation, individuals may react differently to FFPs. Indeed, Rothbard et al. (2005) found that segmentors who were offered child care services displayed lower levels of job satisfaction and organizational commitment than integrators. Conversely, segmentors who were given access to flexible scheduling displayed higher levels of organizational commitment than integrators. Following Rothbard et al. (2005), we believe that desire for segmentation or integration will moderate the relationship between FFPs and applicant attraction:

H2: Desire for segmentation moderates the relationship between FFPs and applicant attraction.

More specifically, we believe that individuals with a strong desire for segmentation will be less attracted to organizations that offer integration practices, such as child care or telecommuting, than individuals with a lower level of desire for segmentation. On the other hand, individuals with a strong desire for segmentation will be attracted to organizations that offer flexible scheduling and additional personal leaves, as these practices favour segmentation.

Corporate Reputation as a Factor of Attraction

In addition to human resources policies, organizational characteristics have been shown to have an effect on applicant attraction. Among those characteristics, reputation received some support from empirical studies. For example, Turban and Cable (2003) found that organizational reputation positively influences both the number and the quality of applicants. Similarly, Greening and Turban (2000) found that corporate social performance, which is related to corporate reputation (Neville et al., 2005), was positively related not only to applicant attraction, but also to job pursuit intention, likelihood of being granted an interview, and job acceptance.

Despite these contradictory results, we posit that corporate reputation has a positive impact on applicant attraction:

H3: Organizations with a good corporate reputation will be more attractive to applicants than those with a poor corporate reputation.

Method

Research Design

In order to test our hypotheses, we chose to use a policy-capturing research design. As described below, we constructed different scenarios in order to experimentally manipulate the within-subject independent variables (FFPs and firm reputation). By presenting subjects with scenarios that contain different information controlled by the researcher, policy capturing is designed to study how people use information to make decisions (Martocchio and Judge, 1994). A policy-capturing design presents several advantages. First, it limits both priming artefacts and social desirability bias due to self-reported measures (Judge and Bretz, 1992; Rynes et al., 1983; Williamson et al., 2002). Second, it increases the realism of an experimental research design by including several factors relevant to decision making in a scenario, as opposed to one factor of interest (Rynes et al., 1983). Finally, due to their high level of experimental control, policy-capturing designs display high levels of internal validity (Deshpande and Schoderbek, 1993; Graham and Cable, 2001) and allow researchers to make causal inferences. In order to develop realistic scenarios related to both independent variables, we conducted two pre-tests, as briefly described below.

Scenario-Constructing Pre-Tests

Pre-test 1. The purpose of pre-test 1 was to make sure that subjects could identify, in our scenarios, whether or not the company offered FFPs, and if so, which ones. Based on real job ads, we developed a general scenario describing a company with job openings. This scenario was tested as the no family-friendly scenario (scenario 1). To construct the FFPs scenarios, we used real practice descriptions, as found on various corporate web sites, to indicate that the company offered on-site child care (scenario 2), personal leaves above what is legally required (scenario 3), flexible work scheduling (scenario 4) and teleworking (scenario 5). For example, the flexible scheduling scenario stated that “the company offers a flexible work schedule that allows employees to take care of both their professional and personal duties.” We then presented each scenario to subjects recruited in a Canadian university continuing education management class. After reading each scenario, subjects were asked to complete a one-page questionnaire about their perceptions of the company.

Thirty subjects voluntarily participated in pre-test 1. Scenarios 1, 2, 3 and 4 were correctly identified by a large majority of subjects. However, the results for scenario 5 were not as clear: only 18 subjects (60 %) correctly identified teleworking. Based on these results, we decided to reinforce the description of the teleworking practice to be used in the main study.

Pre-test 2. In order for the scenarios used in the main study to be more realistic, we decided to use the names of real companies. However, since hypothesis 3 posits that corporate reputation influences organizational attractiveness, a second pre-test was used to identify companies with a good corporate reputation and companies with a poor corporate reputation. Therefore, we drew up a list of 30 companies that regularly recruit on campus and used the corporate web sites to develop a realistic one-paragraph description of each of the companies.

Eighty-two subjects were again recruited in two other continuing education classes. Due to time constraints, each subject evaluated five randomly assigned companies. After reading each company description, subjects were asked to indicate to what extent they agreed with 20 statements. We created the Reputation Quotient measure by averaging answers to the 20 items (a = .95). An ANOVA allowed us to identify companies with a good corporate reputation, and companies with a neutral-to-poor reputation. From these groups, we selected two companies with a good corporate reputation – Cirque du Soleil (M = 3.99, SD = .53) and Xerox (M = 3.69, SD = .38) – and two companies with a neutral-to-poor reputation – Nortel Networks (M = 3.04, SD = .56) and Accenture (M = 3.00, SD = .20) – to be used in the main study.

In addition, the Reputation Quotient measure used in pre-test 2 (Fombrun et al., 2000) proved to be too long to be used in the main study. Based on Neville et al. (2005), we believe that each stakeholder may use a different dimension to form an impression of corporate performance. For example, job candidates may attach more importance to emotional appeal and workplace environment than to financial performance. Therefore, faced with the constraint of reducing the length of the questionnaire, we decided to include in the main study only two of the six dimensions of the original Reputation Quotient: emotional appeal and workplace environment. In order to reflect the fact that we wanted to focus on the work environment, we added one item: “This company offers a good work environment.”

Final Study Procedure and Measures

Based on the results of the pre-tests, we developed five family-friendly scenarios (including no FFP) for four companies (Cirque du Soleil, Xerox, Accenture and Nortel Networks). This amounted to 20 different scenarios. Given the large number of scenarios, a complete block design study, in which each subject would have to evaluate each scenario, seemed too time-constraining to conduct. However, Graham and Cable (2001) found that an incomplete block design generated results similar to a complete block design. In addition, subjects in the incomplete design expressed more positive attitudes toward the study than subjects in the complete design. Therefore, we opted for an incomplete block design in which each subject evaluated four randomly selected scenarios.

Part-time students attending a continuing education management class in a large Canadian university were asked to participate in the main study. Choosing students allowed us to compare our results with those of other studies that have used college students to investigate job choice decisions. However, contrary to some of these studies, our sample is composed of individuals who work full-time, while studying part-time, and expect to look for a better job in a near future. Since those students are generally older than full-time undergraduate students, they are more likely to have a family, and therefore more likely to be particularly sensitive to FFPs.

Upon agreement to participate, subjects were told that the study investigated job choice decisions and were given the questionnaire. In addition to demographic information, the first page of questions measured the subject’s desire for segmentation (Rothbard et al., 2005). Next, each subject read the description of a company corresponding to one of the 20 scenarios. After reading each scenario, subjects were asked to answer questions about the company. These questions included manipulation checks, as well as the measure of the dependant variable.

Results

Sample Characteristics and Manipulation Checks

A total of 110 subjects (66 women, 44 men) volunteered to participate in the study. Among them, 21 (19%) indicated that they were looking for a job, while 82 (74.5%) indicated that they held a job. A large proportion of subjects were either married (37.3%) or cohabitating with a spouse (38.2%). Twenty-nine of them (26.4%) had children and 2 (2 %) had dependents.

Based on the results of pre-test 1, we had constructed five FFP scenarios. In all scenarios, the FFP was correctly identified by a large majority of subjects (no family-friendly practice: 80%; on-site child care: 89.6%; generous personal leaves: 94.6%; flexible scheduling: 93.5%; teleworking: 76.8%).

Following pre-test 2, we had selected two companies with a good reputation and two companies with a neutral-to-poor reputation. In order to check that manipulation, we conducted an ANOVA. The results indicated that the companies had significantly different reputations: F(3, 438) = 21.71, p ≤ .05. However, contrary to expectations, the post-hoc test revealed three groups rather than two. As expected, Cirque du Soleil was perceived as having a good reputation (M = 3.81, SD = .57) and Nortel was perceived as having the poorest reputation (M = 3.14, SD = .82). However, both Accenture and Xerox obtained similar scores, forming a neutral reputation group (Accenture: M = 3.41, SD = .59; Xerox: M = 3.56, SD = .51). This may be explained by the fact that all reputation scores varied slightly between the pre-test and the final study. In the case of Accenture, the difference was significant, t(38.86) = -3.64, p ≤ .001. Further analyses will thus include three levels for corporate reputation.

Test of hypotheses

The means, standard deviations, and correlations among the variables are presented in Table 1.

TABLE 1

Correlations among Variables

Correlations among Variables

** p < .01; * p < .05

-> See the list of tables

Test of hypothesis 1. Hypothesis 1 stated that organizations that offer FFPs are more attractive to applicants than those who do not offer such practices. As expected, the results indicated that FFPs do have a significant main effect on attractiveness: F(4, 109) = 7.35, p < .001. The organizations that did not offer FFPs received significantly lower scores on attractiveness (M = 3.05, SD = .84) than any other organization (on-site child care: M = 3.45, SD = .90; teleworking: M = 3.55, SD = .84; flexible scheduling: M = 3.70, SD = .76; generous personal leaves: M = 3.77, SD = .76). Two scenarios, personal leaves and flexible scheduling, received significantly higher scores on attractiveness. An additional analysis revealed that childless subjects were significantly more attracted to the organizations that provided on-site child care (M = 3.68, SD = .84) than were parents (M = 2.95, SD = .87, t(62) = 3.21, p < .05).

Table 2 presents the paired-comparisons using the least squares means method. It shows that compared to scenario 1, all scenarios received significantly higher scores on organizational attractiveness. Therefore, hypothesis 1, as well as hypotheses 1a, 1b, 1c and 1d, are supported.

TABLE 2

Differences of Least Squares Means for Attractiveness among Scenarios

Differences of Least Squares Means for Attractiveness among Scenarios

-> See the list of tables

Test of hypothesis 2. According to hypothesis 2, desire for segmentation moderates the relationship between FFPs and applicant attraction. In order to test the interaction, we divided subjects into three groups: the low desire for segmentation group included 49 subjects whose desire for segmentation was below 3; the medium desire for segmentation group included 132 subjects whose desire for segmentation was comprised between 3 and 4; and the high desire for segmentation group included 123 subjects whose desire for segmentation was higher than 4.

The results are presented in Table 3. They indicate that despite a significant main effect of all variables, the interaction between desire for segmentation and FFPs was not significant: F(8, 105) = 1.51, p = .16. Therefore, hypothesis 2 is not supported.

TABLE 3

Test of Fixed Effects of all Independent Variables on Organizational Attractiveness

Test of Fixed Effects of all Independent Variables on Organizational Attractiveness

** p < .01; * p < .05

-> See the list of tables

However, the results revealed an unexpected three-way interaction: F(16, 105) = 2.21, p ≤ .05. This suggests that the anticipated interaction between desire for segmentation and FFPs may be significant for some values of the reputation variable. Additional analyses revealed that in the good reputation group, subjects with a high desire for segmentation were significantly more attracted to organizations that offered teleworking (M = 3.20, SD = 1.95) than were subjects with a low desire for segmentation (M = 2.70, SD = 1.84, t(105) = -2.36, p ≤ .05). This result is contrary to our expectations because teleworking is an integration strategy and therefore should be more appealing to integrators than to segmentors. All other paired-comparisons revealed to be not significant. In the neutral reputation group as well, only one paired-comparison was significant: individuals with a high desire for segmentation were significantly more attracted to organizations that offered flexible scheduling (M = 4.11, SD = .59) than were individuals with a low desire for segmentation (M = 2.95, SD = .57, t(105) = -2.91, p ≤ .05). This result is consistent with our belief that segmentors are attracted to organizations that offer practices that favour segmentation. In the poor reputation group, none of the paired-comparisons was significant.

In addition, Table 3 shows an unexpected main effect for desire for segmentation: F(2, 105) = 6.22, p ≤ .05. Individuals with a low level of desire for segmentation were the least attracted to the organizations (M = 3.29, SD = .97), followed by those with a medium level of desire for segmentation (M = 3.49, SD = .84) and subjects with a high level of desire for segmentation (M = 3.54, SD = .87). However, only the difference between low level and high level groups was significant, t(105) = -2.43, p < .05.

Test of hypothesis 3. Hypothesis 3 stated that organizations with a good corporate reputation are more attractive to applicants than those with a poor corporate reputation. As expected, the results indicated that corporate reputation does have a significant main effect of attractiveness: F(2, 109) = 3.69, p ≤ .05. Therefore, hypothesis 3 is supported. More specifically, the company with a good reputation received the highest score on attractiveness (M = 3.59, SD = .83), followed by the companies with a neutral reputation (M = 3.55, SD = .80) and the company with a poor reputation (M = 3.27, SD = 1.01). However, the difference in attractiveness between good and neutral reputation is not significant: companies with a neutral reputation are as attractive to applicants as companies with a good reputation. Both, however, are significantly more attractive than companies with a poor reputation.

Discussion

This study suggests that FFPs influence applicants’ decision processes. All the organizations described as offering a practice designed to reduce WFC were perceived as more attractive than organizations that did not offer any such practice. The two practices that seem most likely to increase attractiveness are personal leaves and flexible scheduling.

Interestingly, the positive effect of FFPs on organizational attractiveness does not seem to be related to current use of a practice. In our study, childless subjects were significantly more attracted to the organizations that provided on-site child care than were parents. This may be due to the fact that the subjects with children already had an alternative child care system, which made this practice less of a decision-making factor in their case. On the other hand, our childless subjects may plan to have children. Since they have not had to look for a child care arrangement yet, on-site child care may appear to be the most convenient option. Also of interest are the results of additional analyses that revealed no difference in the effect of FFPs between men and women in our sample. This may be explained by the fact that, as shown in many data, child care and elder care are now equally shared between men and women in many Canadian families (Duxbury and Higgins, 2003).

Not surprisingly, our results also show that corporate reputation influences organizational attractiveness. This is consistent with Turban and Cable’s study (2003) that found reputation to be positively related to both quantity and quality of candidates. However, we found no difference in attractiveness between organizations with a good reputation and those with an average reputation. Not having a wonderful reputation does not seem to hinder an organization’s effort to interest applicants. On the other hand, having a poor reputation significantly diminishes their chances of attracting candidates. This finding suggests that organizations must be careful not to make any decisions that could harm their reputation. Although this statement holds true in general, it is of particular importance for recruitment purposes, as qualified applicants receive more and more job offers to choose from. However, corporate reputation is a volatile concept, often based on emotions rather than reason (Groenland, 2002), and an organization’s reputation may plummet faster than it may rise.

In addition, the interaction between reputation and FFPs revealed to be not significant. This suggests that an organization can not compensate for the effect of having a poor reputation with the implementation of FFPs. Even though offering FFPs may send the message that an organization cares for its employees’ well-being, other company actions may overshadow and tarnish the organization’s general reputation.

Finally, contrary to expectations, the interaction between FFPs and desire for segmentation did not prove to significantly affect attractiveness. Not only was the overall “Family-friendly practices x Desire for segmentation” interaction not significant, but a closer look at the results for the significant three-way interaction suggested that they might even contradict expectations. For example, we found that in the good reputation group, subjects with a high desire for segmentation were significantly more attracted to organizations that offered teleworking than were subjects with a low desire for segmentation. Yet, we believed that telecommuting favours integration, and therefore should be less attractive to individuals with a high desire for segmentation.

There may be several explanations to these unexpected results. First, contrary to expectations, telecommuting may not be perceived by potential applicants as an integration strategy. Rau and Hyland (2002) posited that teleworking allows flexibility in both spatial and temporal boundaries. However, since telecommuting is seldom offered on a full-time basis (Golden and Veiga, 2005; Schweitzer and Duxbury, 2006), occasional teleworkers may not change their entire child care arrangement just because they sometimes work outside the organization premises. In addition, employers often require that full-time teleworkers make sure that personal duties do not interfere with work (Madsen, 2006). Therefore, the classification of telecommuting as an integration strategy should be nuanced.

A second potential explanation is suggested by the main effect of desire for segmentation on organizational attractiveness. Regardless of other variables, subjects with a high desire for segmentation were simply more likely to apply for a job. Individuals with a high desire for segmentation may perceive that holding a salaried position is the best way to make a clear distinction between their professional and their personal lives. Indeed, people who fear that they will not intrinsically be able to create and maintain boundaries between their work and family roles may choose work arrangements that will deliberately accentuate these boundaries. On the other hand, individuals with a low desire for segmentation may look for self-employment opportunities or free-lancing positions.

Limitations

There are some shortcomings with this study that should be acknowledged. Firstly, even though the vast majority of subjects in our sample were currently employed, they were still part-time students, raising questions of the validity of our results. However, we believe that the use of student subjects in this context is appropriate because these students are likely to be looking for a job in the near future, and therefore were acting in role during the study. In addition, student subjects are likely to experience a three-way work-study-family conflict, making them even more sensitive to the efforts made by a potential employer to provide FFPs. Since our research used older students, we believe that our subjects did not systematically differ from non-students (Peterson, 2001). Yet, different results might have been obtained with a more diverse or a less educated sample. Having a larger proportion of subjects with children might also have changed the results, although, surprisingly, non-parents turned out to be more attracted to the organizations that provided on-site child care than were parents.

Secondly, due to constraints in the length of the questionnaire, we decided not to use the entire measure of Reputation Quotient developed by Fombrun et al. (2000). By using only the subset of indicators that we believed was more important to job applicants, we may have over-evaluated the effect of reputation on attraction. Our results, however, are consistent with studies that found corporate reputation to be positively related to applicant attraction (Turban and Cable, 2003). Nevertheless, it is possible that the use of the entire measure instrument would have generated other results.

Our research design is also subject to limitations. We chose to limit our study to four FFPs: on-site child care; generous personal leaves; flexible scheduling; and teleworking. However, other practices could have been included in the initial design, such as on-site convenience, tuition reimbursement or health care centre. Although we believe that adding these practices would have rendered the research design tedious and therefore would have discouraged participation, we think that their effect is worth exploring. As it was, the research design called for 20 scenarios, which were impossible for each subject to rate. Therefore, we had to resort to using an incomplete block design, which may have increased between-subject error. Finally, our design did not allow us to test the potential cumulative effect of FFPs. There may be a limit to how much FFPs may increase organizational attractiveness. If this is the case, implementing additional practices may only have a marginal effect.

Implications for Practice and Future Research

Contrary to many studies that either investigated one or two practices, or used a general measure of family-friendliness, this research investigated the effect of several FFPs on organizational attractiveness. The results suggest that although all the practices studied here have a positive effect on applicants’ perceptions, offering generous personal leaves and flexible work scheduling seems to have the most impact. Therefore, organizations that cannot afford to implement a large variety of FFPs would be well-advised to first consider these two. In addition, since implementing even one practice increases the likelihood of attracting candidates, organizations facing labour shortage should consider offering opportunities to reduce WFC. And those that do offer such practices should make sure that they are displayed prominently in their job ads, so as to be noticed by potential candidates.

Although this study was conducted in Canada, our findings may benefit companies located in other countries where, traditionally, employers were not involved in the provision of FFPs. For example, in Europe, corporate implantation of FFPs may vary depending on the welfare state system. However, studies indicate that more and more European companies are implementing FFPs in order to encourage women’s participation in the workforce, especially in management positions (Straub, 2007).

Future research should now focus on three avenues. First, we chose to include in our study the practices that are most commonly used in organizations (Guérin et al., 1997). However, other FFPs, such as programs promoting well-being, have received a lot of attention lately and should be further studied. Second, our research design did not allow us to test the combined effect of several practices. It may be the case that implementing a vast array of FFPs dramatically increases an organization’s attractiveness, even in the face of less positive characteristics, such as an average pay level. Therefore, further research should look into the combined effect of FFPs. Third, in order to limit the number of scenarios, we decided to hold other decision factors constant, such as pay or promotion opportunities. Yet, previous studies have found that these are major predictors of organizational attractiveness. Thus, studies should now focus on the role of FFPs in combination with other decision factors.

Corporate reputation and its effect on organizational attractiveness is another area that requires further investigation. Our study suggests that having a good reputation is not necessary to attract applicants, but avoiding a bad reputation is a must. However, our research design did not allow us to investigate which dimensions of corporate reputation, if any, are more likely than others to influence applicants’ perception. In addition, corporate reputation evolves rapidly, and researchers might consider exploring how corporate reputation is formed, how it influences organizational attractiveness, and how organizations can maintain a good reputation among job seekers.

Finally, our study explored the role of desire for segmentation in the recruitment process. To our surprise, desire for segmentation was positively related to organizational attraction; studies should be conducted to better understand the concept of segmentation, how it differs from actual WFC, and how it relates to an individual’s interest in an organization. In addition, our data did not support our argument that some FFPs encourage role segmentation while others favour role integration. An in-depth analysis of the effect of FFPs on work-family boundaries may help shed light on this surprising result.