1 Introduction
Working from home is now entrenched in modern working life. In 1996, only 20% of US companies had working from home arrangements, but by 2016 this had grown to 60% (SHRM
2016). In Europe, it was estimated that approximately one out of eight workers work from home at least several times a month on average across the EU 28 countries (Chung
2018). Against the backdrop of a growing number of dual-earner couples, working from home was touted in the 1980s and 1990s as a cost-effective option for having less work–family conflict (Avery and Zabel
2011). However, there is puzzlingly little empirical evidence which suggests that working from home could be an effective way to mitigate work–family conflict (Allen et al.
2015a; Golden et al.
2006; Kossek et al.
2006). Human resource management theory argues that work flexibility, such as the opportunity of working from home benefits employees by giving them more discretion in combining work and family tasks (Appelbaum
2000; Whitener
2001; Ortega
2009). However, the potential downside is that flexible working is coupled to, often implicit, expectations of high effort and commitment, offsetting potential gains for less work–family conflict (Godard
2001; Wright and Raley
2008; Jensen et al.
2013; Kelliher and Anderson
2010; Lott and Chung
2016; see also, Chung and Van der Horst, in this issue; Chung and Van der Lippe, in this issue (introduction)). If these insights are correct, the capacity of working from home to alleviate work–family conflict may depend on how supportive the organizational context is for the work–family needs of workers
beyond providing formal access (cf. Lewis
2001).
There is evidence that a supportive organizational context is important to alleviate work-life conflict (cf. Den Dulk et al.
2016; Chung, in this issue), but a number of scholars indicated that there is void of research studying how cultural-normative organizational contexts are of influence on family-friendly policies and employee outcomes (Allen et al.
2015b; Kossek et al.
2006; Den Dulk et al.
2016). The present paper fills this gap by providing insight in how the organizational context shapes the impact of working from home on work–family conflict. Specifically, we investigate the role of the ideal worker culture (Kelly et al.
2010), managerial work-life support, and co-worker’s engagement in working from home. By studying the role of culture, support and co-worker behaviour, our study answers recent efforts to contextualize work–family research, “abandoning the tight focus on individuals’ experience of the work–family interface” (Williams et al.
2016, p. 521). In this contribution working from home refers to working at or from home during (at least part of) the employees’ contractual working hours (Felstead and Jewson
2000; Peters and Van der Lippe
2007). In most cases information or communication technology is used to interact with others both within and external to the central office (Allen et al.
2015a).
Gender differences in work–family conflict are paramount (Duxbury and Higgins
1991; Hagqvist et al.
2017; See also, the introduction (Chung and Van der Lippe in this issue), as women experience greater tensions between work and family life than men (Crompton
2002). Research focusing on gender differences in utilizing work–family benefits, such as working from home, showed that men and women judge their utility based on different criteria: men considered work–family benefits useful when they believed it benefitted job performance, while women tended to judge their effectiveness based on expected reduction of work–family conflict (Sprung et al.
2015). Gender-biased selection may also account for underutilizing benefits: especially men may refrain from utilizing working from home for the fear of negative career consequences (Greenhaus and Kossek
2014). Findings on gender differences in the impact of work-life benefits on work-life conflict are, however, scattered. Hammer et al. (
2007) showed that using family friendly benefits actually leads to increased perceptions of conflict for women because they use this opportunity to take on more family responsibilities, but our knowledge is very limited whether a supportive organizational context impacts the gains of men and women from working from home differently. Existing research unearthed gender differences in the usefulness of informal organizational support, but it primarily focused on gender similarity between the worker and supervisor (e.g., Foley et al.
2006), and not on gains from formal support, such as working from home. In this contribution we therefore also examine the role of gender in the relation between working from home and work–family conflict. Given the fact that women experience more work–family conflict than men, working from home might help to gain more understanding in why this is the case.
Despite the recent plea of De Menezes and Kelliher (
2011) for multilevel designs to study how flexible working arrangements influence employee outcomes, current research lacks such data. Most studies in the literature on working from home either used worker-level data from a single organization limiting generalizability beyond a given organizational context and the potential to study organizational differences, or data at the organizational level that do not allow to assess how working from home impacts individual outcomes, such as work–family conflict, or study gender differences (Bloom et al.
2014; Rousseau
2011). Our study makes use of a newly collected, large-scale survey of 259 organizations, 869 work units, and 11,011 employees in multiple economic sectors in the UK, Germany, Hungary, Bulgaria, Spain, Sweden, Finland, the Netherlands, and Portugal (Van der Lippe et al.
2016). This multi-level dataset enables us to study the returns to working from home for workers across different organizational contexts. In line with multilevel theorists, we argue that work–family conflict is more than adding up individual efficiencies from work (Kozlowski and Klein
2000).
To sum up, we try to make the following three contributions: first, in understanding the consequences of working from home for family life, we take into account the workplace context. Second, we pay attention to the gendered effects of working from home and potential differences in how culture, support, and behaviors in the work context impacts the work–family conflict of men and women. Third, we use a large-scale multilevel study in diverse organizational contexts which makes it possible to study how differences among employees in teams in organizations influence work–family conflict.
4 Results
Women participating in the survey experience a slightly higher amount of work–family conflict, but men spend a slightly higher proportion of their regular working time working from home than women do and more often have flexible schedules (see descriptive statistics in Table
1). This latter finding appears to be due to sectoral effects: more than one third of women in the sample are working in health care—compared to only around ten percent of men—where working from home and flexible starting and finishing times are less prevalent. We see furthermore that men work longer hours and that women spend more time on domestic work. Perceptions of managerial support do not differ much between men and women, but men perceive more an ideal worker culture. Moreover, Table
1 shows that women are more found than men in larger organizations.
Table
2 presents the results regarding the relation between working from home—measured as the proportion of working time that employees spend working from home—and work–family conflict, and how this relation varies by perceptions of a supportive organizational culture. Model 1 shows that working from home is associated with higher levels of work–family conflict, supporting our first hypothesis, although the effect is rather weak: 10% (around one standard deviation) increase in the proportion of time devoted to working from home results in 0.05 standard deviation increase in work–family conflict, providing support for H1. Model 2 adds the organizational support factors to the models: as expected, perceived managerial support decreases and ideal worker culture increase work–family conflict. The positive effect of working from home remains. Model 3 to 4 adds the interaction of working from home with organizational context factors (managerial support, ideal worker culture, and co-worker engagement); the one-for-one test of interactions do not indicate that the organizational context moderates the impact of working from home on work–family conflict. Model 5 studies if the relation between working from home and work–family conflict is different for men and women, and results show that working from home increases work–family conflict more for women than for men. In fact, the effect of working from home among men does not reach statistical significance, whilst it significantly increases work–family conflict for women. Model 6 includes all interactions: the interaction between ideal worker culture and working from home becomes positive significant, meaning that working from home is more detrimental for work-life conflict in organizations with an ideal worker culture, providing support for H2b. The hypotheses on the moderating effects of managerial support (H2a) and co-worker engagement in working from home (H2c) are not supported in the completed model either, while the gender and working from home interaction remains significant, providing support for H3. The full model explains 51% of the total variance—after accounting for country, sector, organizational size, and survey mode effects—in work–family conflict across organizations and 40% of the variance across teams. Supportive organizational culture, working from home and their interactions explain 47% of organizational variance unexplained by control variables, and 25% of the team variance, showing that the factors we consider are relevant for understanding the varying level of work–family conflict experienced by workers across organizations.
Table 2
Hierarchical linear regression analyses of work–family conflict on working from home
Working from home (proportion of working month × 10) | 0.05*** (0.01) | 0.03*** (0.01) | 0.03*** (0.01) | 0.04*** (0.01) | 0.01† (0.01) | 0.02 (0.01) | 0.08*** (0.02) | 0.01 (0.01) |
Managerial support | | − 0.17*** (0.01) | − 0.17*** (0.01) | − 0.17*** (0.01) | − 0.17*** (0.01) | − 0.17*** (0.01) | − 0.17*** (0.01) | − 0.18*** (0.02) |
Ideal worker culture | | 0.30*** (0.01) | 0.30*** (0.01) | 0.30*** (0.01) | 0.30*** (0.01) | 0.29*** (0.01) | 0.29*** (0.01) | 0.30*** (0.02) |
Proportion of co-workers working from home | | 0.01 (0.01) | 0.01 (0.01) | 0.01 (0.01) | 0.01 (0.01) | 0.01 (0.01) | 0.03*** (0.01) | − 0.01 (0.01) |
Working from home × Managerial support | | | 0.00 (0.01) | | | 0.01 (0.01) | − 0.00 (0.01) | 0.02† (0.01) |
Working from home × Ideal worker culture | | | 0.01† (0.01) | | | 0.01* (0.01) | 0.01 (0.01) | 0.02† (0.01) |
Working from home × Pr. co-workers WFH | | | | − 0.00 (0.00) | | − 0.00 (0.00) | − 0.01* (0.00) | 0.00 (0.00) |
Working from home × Female | | | | | 0.04*** (0.01) | 0.04*** (0.01) | | |
Female gender | 0.04† (0.02) | 0.04* (0.02) | 0.04* (0.02) | 0.04* (0.02) | 0.02 (0.02) | 0.02 (0.02) | | |
Supervisory position | 0.20*** (0.02) | 0.17*** (0.02) | 0.17*** (0.02) | 0.17*** (0.02) | 0.17*** (0.02) | 0.17*** (0.02) | 0.19*** (0.03) | 0.16*** (0.03) |
Job autonomy | 0.15*** (0.01) | 0.07*** (0.01) | 0.07*** (0.01) | 0.07*** (0.01) | 0.07*** (0.01) | 0.07*** (0.01) | 0.06*** (0.01) | 0.10*** (0.02) |
Organizational tenure (years) | 0.00*** (0.00) | 0.00 (0.00) | 0.00 (0.00) | 0.00 (0.00) | 0.00 (0.00) | 0.00 (0.00) | − 0.00 (0.00) | 0.00*** (0.00) |
Flexible schedule (0/1) | − 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | − 0.02 (0.03) | 0.01 (0.03) |
Contracted hours (sqwrt) | 0.14*** (0.01) | 0.12*** (0.01) | 0.12*** (0.01) | 0.12*** (0.01) | 0.12*** (0.01) | 0.12*** (0.01) | 0.11*** (0.02) | 0.12*** (0.02) |
Commuting time (sqrt) | 0.04*** (0.01) | 0.04*** (0.00) | 0.04*** (0.00) | 0.04*** (0.00) | 0.04*** (0.00) | 0.04*** (0.00) | 0.04*** (0.01) | 0.04*** (0.01) |
Higher educated | 0.07*** (0.02) | 0.03 (0.02) | 0.03 (0.02) | 0.03 (0.02) | 0.03 (0.02) | 0.03 (0.02) | 0.03 (0.03) | 0.04 (0.03) |
Age | 0.02*** (0.01) | 0.02*** (0.01) | 0.02*** (0.01) | 0.02*** (0.01) | 0.02*** (0.01) | 0.02* (0.01) | 0.02* (0.01) | 0.01 (0.01) |
Age2 | − 0.00*** (0.00) | − 0.00*** (0.00) | − 0.00*** (0.00) | − 0.00*** (0.00) | − 0.00*** (0.00) | − 0.00*** (0.00) | − 0.00* (0.00) | − 0.00† (0.00) |
Having a partner | 0.06*** (0.02) | 0.06*** (0.02) | 0.06*** (0.02) | 0.06*** (0.02) | 0.06*** (0.02) | 0.06*** (0.02) | 0.05† (0.03) | 0.08* (0.03) |
Having a young child | − 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | 0.00 (0.02) | − 0.03 (0.03) | 0.04 (0.03) |
Hours of domestic work (sqrt) | 0.04*** (0.01) | 0.03*** (0.01) | 0.03*** (0.01) | 0.03*** (0.01) | 0.03*** (0.01) | 0.03*** (0.01) | 0.05*** (0.01) | 0.01 (0.01) |
Organization size (log) | − 0.01 (0.01) | − 0.01 (0.01) | − 0.01 (0.01) | − 0.01 (0.01) | − 0.01 (0.01) | − 0.01 (0.01) | 0.01 (0.01) | − 0.024 (0.02) |
Intercept | − 1.77*** (0.18) | − 1.56*** (0.16) | − 1.56*** (0.16) | − 1.57*** (0.16) | − 1.54*** (0.16) | − 1.54*** (0.16) | − 1.75*** (0.21) | − 1.11*** (0.24) |
σ(organization) | 0.19*** (0.02) | 0.14*** (0.02) | 0.14*** (0.02) | 0.14*** (0.02) | 0.14*** (0.02) | 0.14*** (0.02) | 0.11*** (0.02) | 0.18*** (0.02) |
σ(work unit) | 0.17*** (0.02) | 0.15*** (0.01) | 0.15*** (0.01) | 0.15*** (0.01) | 0.15*** (0.01) | 0.15*** (0.01) | 0.17*** (0.02) | 0.11*** (0.04) |
σ(Residual) | 0.85*** (0.01) | 0.80*** (0.01) | 0.80*** (0.01) | 0.80*** (0.01) | 0.80*** (0.01) | 0.80*** (0.01) | 0.80*** (0.01) | 0.80*** (0.01) |
Number of employees | 11,011 | 11,011 | 11,011 | 11,011 | 11,011 | 11,011 | 6069 | 4751 |
Number of work units | 868 | 868 | 868 | 868 | 868 | 868 | 781 | 759 |
Number of organizations | 257 | 257 | 257 | 257 | 257 | 257 | 256 | 252 |
Model 6 is estimated for men and women separately
6 to investigate gender differences. We also estimated a pooled model including three-way interactions between working from home, organizational context variables, and gender to formally test H4 (full results available in online Appendix Table A1, Model 7) as well as a less parsimonious full gender interaction model (results available in online Appendix Table A1, Model 8). The results of model 6 show clearly different patterns for men and women: ideal worker culture and managerial support increase the positive relation between working from home and work–family conflict but only for men and the findings are only significant on the .1 level (2-sided). When formally tested, the three-way interaction between gender, managerial support and managerial support is significant on the .1 level (2-sided) (b = − .02, SE = .01,
p = .08). In the full interaction model, however, the effect is of similar size, but not significant on the .1 level (2-sided) (b = − .02, SE = .01,
p = .12). These results provide, although modest, support for H4 that working from home is more beneficial in combination with managerial work-life support for women than for men with regards to less work-life conflict. In addition, if more co-workers work from home own working from home is less negatively related to work–family conflict among women, but this is not the case for men. Again, when formally tested, the negative moderating effect of co-worker engagement in working from home is stronger among women, although only significant on the .1 level (2-sided) (b = − .01, SE = .00,
p = 0.06), but both in the pooled and full interaction models. This result is also in line with H4.
5 Conclusion
Empirical research on the influence of working from home on the work–family interface does not produce convincing outcomes regarding a decrease of work–family conflict. In this contribution we argued that working from home will lead to more facilitation of work and family life, if a supportive work context is available. Specifically, managerial support, ideal worker culture, as well as the number of colleagues working from home are expected to moderate the relation between working from home and work–family conflict. Gender differences are explored therein. Using unique 2016 data from 11,011 employees in 869 teams in 259 work establishments in 9 European countries, we tested several hypotheses. Our multi-level approach resulted in four substantive findings.
The first conclusion is that elements of the work context help to explain how working from home can alleviate or increase work–family conflict for both men and women. We studied under what conditions working from home leads to work–family conflict, and conclude that if an ideal worker culture exists, where the norm is to work hard, to work overtime, and to take work home at night or in the weekend to get ahead in the organization, working from home will lead to more work–family conflict. This implies that in organizations where no such culture exists, results in less work–family conflict. This is in line with research from Kelly et al. (
2010) who also showed the existence of the ideal worker culture. Working part-time which alleviates the time pressure, is no option in these organizations (Rose et al.
2013). However, we didn’t find that the perception of a supportive manager is important as well for work–family conflict. It could be that the backbone of work–family conflict is an ideal worker culture, which creates demands that interfere with family responsibilities. The original research by Thompson et al. (
1999) already showed more influence of the ideal worker culture than managerial support on work–family conflict. However, Anderson et al. (
2002) reported managerial support to be more important than the ideal worker culture. Differences could be due to the samples used. Anderson et al. made use of a large scale sample of employed adults, not necessarily nested in organizations.
The second conclusion is that the relation between working from home and work–family conflict is clearly gendered. Working from home leads to more work–family conflict for women. The fact that they are working from home makes the boundaries between work–and family life permeable which increases the chance on work–family interference especially for women (see also, Kim in this issue; Kurowska in this issue). This is in line with many research findings on related topics showing that the work and family domain are more linked for women than men (Van der Lippe
2007; Van der Horst et al.
2014). When studying ambitions of men and women, women more often take the linkage between work and family life into account, whereas men are more fully concentrated on their work career (Van der Horst et al.
2014); looking at work–family conflict, women also more often report to be influenced by their family obligations and men by work obligations (Van der Lippe et al.
2006). Our finding adds to the extant literature on gender differences between experiencing work and family life.
The third conclusion is that the work context appears to work differently for men than for women. Having many co-workers working home alleviates some of the influence of working from home on work–family conflict for women, but men are not influenced by this. This type of support may also especially help women, as they experience more stress in general than men (Matud
2004). One could imagine that they are more sensitive to this behavior by co-workers and are more likely to benefit from it. Co-workers working from home are more likely to be understanding of the potential conflicting situations which helps women more than men. Note that this conclusion is only true for workplaces where and jobs for which it is possible at all to work from home (so for example not for nurses). Unexpectedly, although only marginally significant, managerial support increases the positive relation between working from home and work–family conflict for men but not for women. The influence of ideal worker culture seems to be mainly driven mainly by male respondents, although the differences between men and women are very slight. Norms in organizations might still dictate a full commitment of men to work (Van der Lippe
1994; Treas and Drobnic
2010), which heightens the experiences of work–family conflict. This has perhaps also to do with the feminity stigma (Williams et al.
2013; Chung in this issue), where men may face a double stigma when using flexible working arrangements due to the fact that such arrangements deviate away from the ideal worker norm but also from masculinity. For women, work ethic is also positively associated with their job, but only if women’s gender role values are taken into account, which negatively relate to women’s work (Stam et al.
2014).
The fourth conclusion is that it is important to take a multi-level approach to this issue. Our contribution showed the importance of incorporating the team level and organizational context to understand why working from home is related to work–family conflict, and why this works differently for men and women. Not having information on the teams and organizations where these men and women work would make it possible to describe that women experience more work–family conflict when working from home, but not what the role of colleagues in the team and the organization thereby is.
We recognize that there are also limitations to the research presented here. First and foremost, it is likely that there is selection of men and women who experience work–family conflict, and maybe they are selected into organizations that provide working from home policies. There might even be a possibility of reverse causation: organizations in which men and women experience comparatively more work–family conflict may like to work from home. We have tried to take this partly into account by controlling for a number of family responsibilities and organizational factors. This gives us confidence with respect to our research strategy, but nevertheless, we encourage new research using a long time panel set up, in which organizations and their employees are followed over some time to better unravel the causes and consequences of work–family conflict. Second, we have focused in this study on the experienced work–family culture in the organization by the employee. Although we still argue this to be important, we can imagine that focusing on the actual behavior (and the gender) of the manager of the employee, will contribute to understanding the influence of working from home and work–family conflict better (Hammer et al.
2009; Matthews et al.
2013). Third, work–family conflict is a subjective feeling, and we encourage other researchers to also study more behavioral consequences such as time spent on parenting, time spent on housework, the amount of leisure activities, and having contacts with family and friends, which may be all behavioral outcomes of work–family conflict. However, we moved beyond existing work–family conflict scales by also including leisure time as a potential source of conflict.
Our results leave several other questions unanswered, such as what happens with men when they work from home. Managerial support and organizational demands both pointed in the direction of having more work–family conflict, but it is unclear what is actually happening. We therefore propose qualitative research where the researcher studies processes of support and the culture in one or two organizations for a longer period of time. Also reactions of the team when working from home as a male employee are important to take into account. Another way to go would be to study more closely which tasks are performed at home when working from home. Tasks which apparently lead to more work–family conflict, than when performed at the office. Moreover, it would be informative to study why people work from home, is it to reduce work–family conflict as is the assumption in this paper, or also to save commuting time (Bailey and Kurland
2002). All in all, our study showed working from home leads to more work–family conflict, especially when workers perceive an ideal worker culture at their workplace. Women experience more work–family conflict from working home than men, but the more colleagues are working from home the less conflict these women experience.