Contextual factors in formal policy and retention relationships
As noted above with work-life outcomes, one reason for the lack of consistency in studies examining family-friendly policies and retention-related outcomes is likely to be the presence of moderating variables. In this section, we discuss common variables that have been mentioned and examined as moderators of these relationships.
Gender has been most commonly studied as a moderating variable. One study reported no significant moderating effect of gender on the relationship between work flexibility and organizational commitment (Ngo & Tsang, 1998), but other researchers found that women whose organizations offered flexible schedules reported higher commitment than women whose organizations did not offer such policies. However, for men, commitment was not affected by the availability of flexibility (Scandura & Lankau, 1997). Moreover, adding another level of complexity by examining availability and use of schedule flexibility in tandem with gender, Casper and Harris (2008) reported a significant three-way interaction. For women, schedule flexibility was positively related to commitment regardless of use, but for men, schedule flexibility was positively associated with commitment only when use was high. A negative relationship was observed when use was low. The same relationships were tested for dependant care support, but no significant interactions were observed. Overall, it seems that for women availability of flexible policies alone communicates a positive message about organizational support, whereas men benefit in terms of commitment only when using these policies.
Other streams of research highlight indirect evidence for gender differences in the formal benefit-retention relationship. First, there is some evidence of differential access to flexible work arrangements based on the gender composition of the occupation in the US (Glauber, 2011). Specifically, a curvilinear relationship exists, such that the probability of a person’s organization offering flexible scheduling is highest when occupations have a more even gender ratio (i.e., around 55% of workers in that occupation are female). When the proportion of females is low, the odds of flexibility being offered are also lower; the same trend (albeit with a weaker effect) is observed when the proportion of women is very high. Given that, by definition, men are more likely to be in occupations that are male- dominated, this finding suggests that men may have less access to flexibility. Clearly, without access to policies one cannot possibly reap their retention-related benefits. Of note is that these findings may not generalize outside the US, as there is considerable variation in the gap in gender differences in access to flexibility in Europe. Notably, in a comparison of 30 European countries, in all cases (except Greece and Malta) the gender differences were such that men had greater access to flexibility (Plantega & Remery, 2009).
In a qualitative study focused on employees’ views about the implementation of a Results Only Work Environment (ROWE; an initiative aimed at reorienting employees and managers towards measurable results while playing down the need to be physically present at work for a set number of hours each day) at Best Buy corporation, Kelly, Ammons, Chermack and Moen (2010) reported that men and women differed in their reactions. Women, especially mothers, were very enthusiastic about working in a ROWE arrangement. Men, particularly those early in their careers, were more ambivalent, focusing on the potential negative career repercussions (e.g., being passed over for promotion) of using flexibility. Quantitative data from a study in Greece support this finding, as men reported greater concerns about career costs associated with using flexible work arrangements than did women (Giannikis & Mihail, 2011). It is unclear whether flexibility users do suffer penalties, as three studies (Cohen & Single, 2001; Glass, 2004; Judiesch & Lyness, 1999) found that flexible benefit use results in fewer promotions and pay increases for both men and women, whereas others found that users of flexible work practices earn higher wages (Leslie, Manchester, Park & Mehng, 2012; Weeden, 2005). Regardless, it appears that women seem more attracted to flexible work arrangements, perhaps because they focus more on the potential work-life benefits (Butler, Gasser & Smart, 2004; Sharpe, Hermsen & Billings, 2002; Sullivan & Smithson, 2007; Vandello, Hettinger, Bosson & Siddiqi, 2013).
Moving beyond gender, preference for role segmentation versus integration is mentioned frequently in discussions of use and efficacy of formal policies. According to boundary theory (Ashforth et al., 2000; Nippert-Eng, 1996), people create boundaries around different life roles to help manage them and simplify their complex environments, but there are individual differences in the extent to which people prefer and are able to keep boundaries permeable and flexible, which in turn influences the relative segmentation or integration of roles (Ashforth et al., 2000; Kossek, Lautsch & Eaton, 2006; Trefalt, 2013). Preference for segmentation versus integration lies along a continuum, with the most extreme preference for each serving as poles at each end. Thus, most individuals are not qualified as pure segmenters or integrators, but rather fall between the two extremes.
Researchers argue that many types of formal family-friendly policies alter the nature of the workplace in terms of the integration or segmentation that it allows (Rau & Hyland, 2002; Rothbard, Phillips & Dumas, 2005; Shockley & Allen, 2010). For example, because telecommuting typically involves working from home, the place where the family domain resides, it is likely to lead to role blurring, an undesirable state for those who prefer to segment their roles. A similar argument can be made for on-site daycare. On the other hand, whether or not flextime is an integrating or segmenting policy has been debated. Some argue that it allows for greater integration (Rau & Hyland, 2002; Shockley & Allen, 2010) by making boundaries less permeable, but others argue that it allows employees to alter their schedules so that role blurring does not have to occur, making it more of a segmenting policy (Rothbard et al., 2005). Empirical results support both positions, as segmentation preferences negatively relate to use of flextime (Shockley & Allen, 2010), but moderate the relationship between flextime use and organizational commitment, such that the association is positive for segmenters but negative for integrators. In other words, those who prefer to segment benefit from flextime in terms of greater organizational commitment, whereas the opposite trend is true for integrators, suggesting it may be a policy that is more conducive to segmentation. Whatever the case, this individual difference variable has important implications for work-life management and family- friendly policies, although the nature of these associations remains unclear (Allen, Cho & Meier, 2014).
With regard to the specific policy of telecommuting, a sense of professional and social isolation that may accompany being physically removed from the workplace is a matter of concern (Cooper & Kurland, 2002; Golden, Veiga & Dino, 2008; Greer & Payne, 2014). Although telecommuting may grant additional flexibility and ability to manage work and family demands, it can also easily lead to seclusion and loneliness, and may ultimately result in turnover. Arrangements that are not fully remote may offset this effect; in fact, research suggests that job satisfaction is maximized when an employee telecommutes around 15 hours a week (Golden, 2006; Golden & Veiga, 2005). This hypothesis has not been investigated with retention-specific outcomes, but given the association between job satisfaction and turnover (Podsakoff, LePine & LePine, 2007), a credible link exists. Additional research suggests that the association between professional isolation and turnover intentions among teleworkers is weaker when they have greater access to communicationenabling technologies, as this may work as a partial substitute for face-to-face interactions (Golden et al., 2008).
Finally, several researchers have discussed the role of the informal work environment. Specifically, when available, policies are not perceived to be usable when the culture is highly focused on face-time or the supervisor does not support policy use, so that even when policies are available, employees are less likely to use them (Allen, 2001; Breaugh & Frye, 2008; Eaton, 2003; Thompson, Beauvais & Lyness, 1999: Shockley & Allen, 2010). Additionally, when policies are used despite a non-supportive environment, any gains (i.e., better work-life management) may be offset by the stress that comes with taking advantage of the policy (Shockley & Allen, 2007). In general, if the organizational norms and values are not consistent with use of available policies, there is cause to suspect that their value will be greatly undermined (Kirby & Krone, 2002; Lobel & Kossek, 1996; Ryan & Kossek, 2008; Wallace & Young, 2008). We discuss the issue of informal support and colleagues next.