This study examines how individual work- and family-related circumstances shape the relationship between working from home and self-perceived changes in career prospects among employed parents, more than a year into the COVID-19 pandemic. We particularly investigate how prior experience with work from home (WFH), occupational status and the intensity of family demands moderate this relationship.
Drawing on several theories, including role conflict and competing demands frameworks, we formulate five hypotheses, which we test using a large multi-country dataset and fixed effects multinomial regression models. The data come from representative samples of working parents with dependent children in Canada, Germany, Italy, Poland, Sweden and the United States, collected in mid-2021.
The findings show that parents who worked from home both before and during the pandemic were more likely to report positive changes to their career prospects, especially men and those in managerial or professional positions. However, these positive associations are contingent on favourable family contexts: they are evident primarily among parents living with a partner and those whose children were not at home for extended periods. In contrast, mothers who began WFH during the pandemic, particularly those living alone or managing prolonged childcare duties, were more likely to report worsened career prospects.
This study advances the literature by highlighting the intersectional and gendered nature of perceived career changes during the pandemic. It shows that the professional benefits of WFH are not universally shared but depend critically on pre-existing work arrangements and family responsibilities. The findings point to the importance of designing context-sensitive career development strategies in increasingly hybrid work environments.
Introduction
Work from home (WFH), which falls into the broader definition of fluid work (Vaiman et al., 2021), represents the deviation from the “standard employment” that takes place within a bounded space at the employers' premises (Minbaeva, 2021). It has surged in prevalence thanks to the development of new technologies and the recent COVID-19 pandemic, which significantly increased the spread, speed and depth of this work transformation (Alfes et al., 2022; Barrero et al., 2021). Importantly, WFH does not appear to be a merely temporary response to the pandemic; its continued prevalence suggests that this type of fluid work is likely to persist (Bick et al., 2023; Buckman et al., 2025). This enduring change is likely to have significant long-term implications for workers, organisations and society at large. For instance, it impacts where people live (Davis et al., 2024; Ramani et al., 2024), work (Akan et al., 2025), whether they have children (Osiewalska et al., 2024), the organisational policies and practices (Forbes et al., 2022), and even crime rates and environmental outcomes (Lee, 2023). These far-reaching consequences underscore the need for continued scholarly attention to the multifaceted effects of WFH.
A recurring theme in studies on the effects of WFH is its potential impact on employees' career prospects (see, e.g. Zhang et al., 2024; Jamali et al., 2023). Pre-pandemic studies highlighted the potential negative impact of WFH on career prospects, often resulting in remote workers being overlooked for in-house training opportunities and promotions (Glass and Noonan, 2016; Martinez and Gomez, 2013). The issue became particularly critical during the pandemic, as WFH became widespread across diverse sectors of the economy. On the one hand, research has shown that during the pandemic, workers who started to WFH because of the pandemic-related lockdowns and WFH mandates expressed concerns about the impact of WFH on their careers (Çoban, 2022; Moens et al., 2022; Fana et al., 2020). Additionally, remote workers frequently experienced higher stress levels and lower productivity during the pandemic (Gibbs et al., 2023; Shirmohammadi et al., 2022), which could have negatively influenced their perceptions of career prospects during this period.
On the other hand, the pandemic period saw a substantial surge in the remote workforce, which could have helped alleviate the flexibility stigma – a belief that employees using flexible working arrangements are less committed and productive (Chung, 2022). This shift may have particularly benefited employees who had already worked from home prior to the pandemic and had likely experienced such stigma in the past. During the pandemic, companies gained increased expertise in managing remote employees and invested in building the necessary infrastructure to support this shift in work dynamics (Bennett and McWhorter, 2021), potentially enhancing career development opportunities for remote workers. Recently, Kasperska (2025) demonstrated that individuals who WFH at least sometimes generally view their career advancement opportunities positively, though the impact of the pandemic on this relationship is still unknown. Furthermore, some studies have documented the positive effects of WFH on productivity and performance (Hackney et al., 2022), with certain groups of employees, particularly managers and professionals, potentially deriving the greatest benefits (Teodorovicz et al., 2022). Consequently, an open question remains as to whether, and in what ways, WFH influenced changes in career prospects among different groups of workers during the pandemic, particularly in relation to their prior (pre-pandemic) experience with remote work and their occupational characteristics.
The ambiguity surrounding the effects of WFH on employees' career prospects is likely also shaped by family-related factors. Pandemic-induced policies not only led to the widespread adoption of WFH but also triggered school and childcare closures, as well as frequent quarantines. On the one hand, for many working parents, WFH became the only viable means of reconciling paid work with caregiving responsibilities at home. On the other hand, it posed substantial challenges, particularly during lockdowns and quarantine periods, when parents were solely responsible for supervising, caring for and educating their children, often during standard working hours (Augustine and Prickett, 2022; Carlson and Petts, 2022; Derndorfer et al., 2021; Farre et al., 2020; Hu et al., 2023; Mandeville et al., 2022). In situations where children were continuously present at home for extended periods, or where a working parent lived alone with children, WFH likely generated acute role conflict (between being a productive employee and an engaged parent) and led to competing demands from both professional and caregiving responsibilities. This, in turn, may have limited the potential positive effects of WFH on career prospects for those who had already worked remotely before the pandemic, and may have had adverse consequences for those who began WFH because of it. For the latter group, the pandemic brought not only an abrupt shift in working arrangements but also a sudden intensification of childcare responsibilities. The convergence of these sudden and simultaneous changes likely undermined their ability to fully engage with professional responsibilities and may also have resulted in negative psychological consequences, thereby negatively affecting their perceptions of career prospects. Nevertheless, to date, no previous study has systematically explored how the effects of WFH on career prospects during the pandemic varied across diverse family-related circumstances.
In this study, we demonstrate that individual heterogeneity in pre-pandemic experience with WFH, occupational characteristics, and family-related circumstances is key to explaining parents’ differential perceptions of changes to their career prospects – an area that has been largely overlooked in previous research. So far, few studies have examined the relationship between WFH and career prospects during or after the COVID-19 pandemic (Kasperska, 2025, is one exception). To the best of our knowledge, no study has yet investigated differences in such an association with a distinction for those who already worked from home before the pandemic and those who started to do so as a result of the pandemic. This distinction is conceptually and analytically important, as perception of changes to career prospects likely differed among these two groups, for several reasons. These reasons include previously experienced flexibility stigma and comparative advantage in telecommuting skills, as well as the know-how in reconciling work and care at home. Differentiating between employees with prior WFH experience and those without during the pandemic is an approach that experts in the field have urged researchers to explore (Caligiuri et al., 2020). However, this differentiation has been largely absent from the WFH literature, except for the recent studies by Kurowska et al. (2025a, b), which highlighted significant differences in (changes to) work–life balance between these two groups. To date, other research has primarily distinguished between voluntary and non-voluntary remote workers (e.g. McAlpine et al., 2024; Huo et al., 2022). Furthermore, these studies have focused on other outcomes, such as feelings of isolation, job autonomy, emotional support or employee innovative behaviour (McAlpine et al., 2024; Huo et al., 2022), rather than on (perceptions of) career prospects.
Furthermore, we extend the existing literature by exploring differences in how WFH relates to perceptions of career prospects among managers, professionals and other workers with children. While previous studies have examined how experiences of WFH varied between managers and other workers (e.g. Oakman et al., 2022; Kirchner et al., 2021; Ipsen et al., 2021), they have not yet addressed differences in their perceptions of career prospects. These differences are likely, as managers and professionals could have leveraged higher job control and autonomy when transferring to WFH during the pandemic (Teodorovicz et al., 2022).
Moreover, prior quantitative research on the topic has predominantly focused on WFH and objective career measures, such as the likelihood of being promoted or experiencing wage growth (Arntz et al., 2022; Golden and Eddleston, 2020; Weeden, 2005), or on the manager's perspective regarding those who WFH, including discriminatory attitudes (Moller et al., 2024a, b) and practices (Matysiak et al., 2025; Kasperska et al., 2024; Wang and Chung, 2023). The employee perspective is, nevertheless, essential because it provides insight into how employees perceive their jobs and careers, which may enhance understanding of their behaviours (Greenhaus and Kossek, 2014). Notably, research indicates that employees' perceptions of their career prospects do not always align with their objective prospects (Greenhaus and Kossek, 2014). Recognising this potential dissonance between objective and subjective career prospects, we contend that understanding employees’ interpretations of their career prospects is crucial.
Finally, previous studies have rarely investigated the role of specific family-related circumstances that, by substantially altering individual experiences of WFH among parents, could have impacted the way they perceived their future career prospects while working from home. Building on role conflict and competing demands theories (Greenhaus and Powell, 2003; Carlson et al., 2000; Greenhaus and Beutell, 1985; Kahn et al., 1964), we argue that acute conflicts between family and work during the pandemic – particularly for parents living alone with children, and parents with children at home for prolonged periods – translated into worse outcomes in terms of subjective career opportunities.
In our study, we leverage a unique, multi-country dataset from representative samples of parents with dependent children in the US, Canada, Germany, Sweden, Italy and Poland (see Kurowska et al., 2023). This enhances the relevance of our findings to diverse contexts. We examine employees’ perceptions in mid-2021, over a year after the pandemic’s outbreak, when the most severe COVID-19 waves had subsided (see Figure 1 in the Appendix), most confinement restrictions in Europe and North America had been lifted – at least temporarily – and vaccination efforts were widespread (Mathieu et al., 2024). This juncture likely provided employees with an opportunity to reflect on how the pandemic had influenced their professional trajectory and career prospects while experiencing a period of relative stability.
By using large samples of employed mothers and fathers, i.e. employees that faced multifaceted challenges during the pandemic (Augustine and Prickett, 2022; Carlson and Petts, 2022; Derndorfer et al., 2021; Farre et al., 2020), we are able to zoom into diverse work- and family-related circumstances, not only exploring their importance for subjective career prospects separately (main effects) but also delving into potential intersectionality of adverse conditions across both work- and family-related situations (interaction effects).
Theoretical framework and hypotheses
The impact of WFH on careers remains ambiguous, particularly in the context of the pandemic-induced rise in remote work (see, for example, Matysiak et al., 2025; Kasperska et al., 2024; Arntz et al., 2022; Golden and Eddleston, 2020; Munsch, 2016; Bloom et al., 2015; Leslie et al., 2012; Weeden, 2005). Below, we discuss how individual heterogeneity in pre-pandemic experience with WFH, occupational characteristics and family-related circumstances is key to explaining parents' differential perceptions of changes to their career prospects.
The role of pre-pandemic experience with working from home
We argue that employees who worked remotely before the pandemic and continued to do so during this period may have perceived changes to their career prospects more positively than the newcomers to WFH. First, it is well-documented that employees in novel and uncertain situations often experience tension and stress (Caligiuri et al., 2020; Stahl and Caligiuri, 2005). For those who began WFH during the pandemic, the sudden transition from in-person to online work brought work intensification, stress and decreased work engagement (Adisa et al., 2023). For workers who were already familiar with the challenges and advantages of remote work, the sudden shift to widespread WFH was unlikely to cause such negative psychological effects. Second, evidence suggests that many initial challenges of remote work were related to technical issues (Wang et al., 2021), and those with prior WFH experience likely held a comparative advantage in addressing such challenges. Third, adapting to simultaneous work and caregiving responsibilities at home, prompted by pandemic-related childcare and school closures as well as quarantines, might have been easier for those who had previous experience balancing remote work with parenting. More importantly, remote work became more widespread and remained at significantly higher levels than before the pandemic, even after the initial lockdown phases (Barrero et al., 2021). This could have fostered greater optimism among remote workers regarding the diminishing of flexibility stigma and, in turn, the potential to enhance their standing within the organization and, consequently, their career prospects. Previous research has documented an increased preference for remote work during the pandemic (Chen et al., 2023), as well as associated productivity gains (Barrero et al., 2021). These findings lend support to the hypothesised mechanism that the pandemic reinforced optimism about career prospects by alleviating the stigma associated with working from home among those who had already been working remotely prior to its onset.
To account for the heterogeneous experiences of employees with and without prior remote work experience, we draw on cognitive appraisal theory (Lazarus and Folkman, 1984). This theory posits that individuals evaluate novel events or developments through, first, assessing whether the situation constitutes a threat or an opportunity (primary appraisal), and second, determining whether they possess the resources to cope effectively with the situation (secondary appraisal).
This perspective is particularly relevant in the context of the abrupt shift to widespread remote work during the COVID-19 pandemic, which represented a sudden and potentially disruptive change for all employees. From this perspective, employees with pre-pandemic remote work experience were more likely to perceive this shift as an opportunity rather than a threat, owing to their greater familiarity with the demands and benefits of remote work, as well as reasonable expectations that the pandemic-induced normalisation of remote work would reduce the flexibility stigma. As a result, they may have appraised their professional situation during the pandemic more favourably – both in comparison to their own pre-pandemic status and relative to colleagues who lacked such experience. In their case, WFH likely contributed to the perception of improved career prospects, also through higher job satisfaction (Felstead and Henseke, 2017; Gajendran and Harrison, 2007). Therefore, we expect a positive relationship between WFH and perceptions of changes to career prospects primarily among parents who had worked from home before the pandemic and continued to do so during this period, rather than among those who started WFH only during the pandemic (H1).
The role of occupational characteristics
Furthermore, the link between WFH and perceived changes to career prospects can also vary depending on the occupational strata to which employees belong. One reason for this is the differing levels of autonomy and responsibility that managers and professionals experience compared to other occupational groups. Managers and professionals typically enjoy greater work autonomy and engage in tasks distinct from those of other workers (Wheatley, 2017; Gallie et al., 2004). The demand–control model (Karasek, 1979) posits that employees in higher positions tend to have greater control over their work, enabling them to better manage job demands and task completion. Similarly, the job demands-resources model (Demerouti et al., 2001) highlights that high-autonomy roles (typical for managers and professionals) are better buffered against the negative consequences of job demands (such as sudden changes in working arrangements) (Bakker and Demerouti, 2007). When transitioning to WFH, managers and professional workers may have more flexibility and autonomy in managing their time and responsibilities, potentially boosting their productivity and perceived effectiveness in their roles (Kossek et al., 2015). Existing evidence highlights that autonomy is an important mediator in the relationship between WFH and job satisfaction (Gajendran and Harrison, 2007), which, in turn, is linked to productivity levels and, consequently, employment outcomes (Parker et al., 2017; Cordery et al., 2010). Teodorovicz et al. (2022) compared the tasks performed by managers before and during COVID-19 and found that managers exerted more effort (and worked longer hours) after transitioning to WFH during the pandemic. In contrast, non-managers often used the time saved from commuting for personal activities (Teodorovicz et al., 2021). Consequently, the nature of managers’ responsibilities, such as coordination and supervisory tasks (Drucker, 2012), may have led to increased workloads, potentially enhancing their perception of career prospects. Non-managers, however, may have faced challenges adapting to WFH, particularly if their roles required direct supervision or constant coordination with colleagues.
Another reason managers and professionals may have fared better than other occupational groups in the transition to WFH is their often-superior access to resources and support systems within their organisations (Autor and Dorn, 2013; Drucker, 2012). The human capital theory (Becker, 1964) suggests that individuals with greater skills, experience and job-related knowledge, such as managers and professionals, are better positioned to leverage opportunities for career advancement. During the pandemic, they may have been better equipped to manage the transition to WFH, with access to necessary technology, support from IT departments and established communication channels. Other employees may have faced barriers in accessing these resources, potentially affecting their performance and perceived career prospects. Moreover, due to their seniority and experience, managers and professionals generally enjoy higher visibility within their organisations (Drucker, 2012) and may have already established credibility and a strong reputation before the pandemic. Consequently, working remotely may not have significantly diminished their visibility or recognition for their contributions. Whereas, for other groups of workers, especially those early in their careers, the lack of face-to-face interaction in a remote work setting may have hindered their ability to showcase their skills and achievements, potentially affecting their perceived career prospects. Correspondingly, research has shown that the transition to WFH during the pandemic disrupted mentoring relationships, worker interactions and work quality, particularly for junior employees (Emanuel et al., 2023). Therefore, we hypothesise that WFH is positively associated with perceived changes in career prospects among managers and professionals, regardless of their prior experience with WFH before the pandemic (H2).
The role of competing family demands while working from home
More importantly, we acknowledge that the relationship between WFH and employees’ perception of changes in their career prospects during the pandemic might have been moderated by the intensity of role conflicts and competing demands from work and family spheres experienced during this period by working parents.
The role–conflict theory explains how individuals occupy multiple roles within various social structures, such as work and family, each with its own expectations and demands (Michaelides et al., 2023; Carlson and Petts, 2022; Greenhaus and Beutell, 1985). When these roles come into conflict, individuals may experience role conflict, leading to stress, tension and difficulties in managing their responsibilities. For workers with children, diverse childcare and housework responsibilities can result in severe conflicts between their roles as parents and employees (Eddleston and Mulki, 2017), causing psychological strain and negatively impacting their subjective assessment of their (professional) future, including impaired perception of their career prospects.
While role–conflict theory underlines the psychological strain from role conflicts between a productive employee and a present parent, the competing demands framework (Greenhaus and Powell, 2003; Greenhaus and Beutell, 1985; Kahn et al., 1964) highlights more objective constraints, i.e. workers facing overlapping responsibilities that compete for their time and resources, affecting their productivity (Greenhaus and Powell, 2003). Family demands can spill over into work (and vice versa), a phenomenon known as family-to-work and work-to-family spillover (Grzywacz and Marks, 2000; Grzywacz and Demerouti, 2013). For parents, demands from the family sphere frequently spill over into the work sphere, particularly when they WFH (Rosen et al., 2024). We argue that negative family-to-work spillover while working from home may, in turn, negatively impact one’s perception of career prospects, as such circumstances diminish workers' capabilities to fully engage in professional work and objectively threaten their productivity. Previous research has suggested that family-to-work negative spillover can influence both subjective and objective career success (Ballout, 2008). In interviews conducted by Çoban (2022) during the pandemic, some mothers working from home expressed serious concerns about their career plans due to competing family demands, which caused them to drift away from professional responsibilities. The role conflict and competing demands from professional and personal spheres, while WFH can, however, be more acute in some circumstances than in others. When family-to-work spillover is substantial, it can diminish the potential positive effects of WFH on perceived changes in career prospects (Kasperska, 2025). In our study, we consider living alone with children or having children constantly present at home as such circumstances that exacerbate both role conflicts and experiences of competing demands. We also examine potential gender differences, as mothers tend to experience stronger role conflict and competing demands than fathers (Young et al., 2023; Ng et al., 2002; Scott and Alwin, 1989).
Living alone with children
Juggling work and childcare while working from home can be particularly challenging for single parents and adversely affect their careers and mental health, as they often lack support to care for their children during working hours (Almeida et al., 2020; Power, 2020). Evidence suggests that single parents were 13% points more likely than partnered parents to report working less efficiently while working from home compared to their pre-lockdown efficiency levels (Aczel et al., 2021). Therefore, we expect a positive relationship between continued WFH and perceived changes in career prospects during the pandemic among parents living with a partner, but not among those living alone with their children (H3).
Increased childcare demands during the pandemic
Furthermore, dealing with role conflicts and competing demands during the pandemic may have been particularly challenging for working parents whose children were at home for extended periods due to childcare or school closures. Kurowska et al. (2025a) demonstrated that having children at home for prolonged periods due to confinement policies during the pandemic diminished the positive effect of WFH on perceived work–life balance among working parents. The difficulty of managing various roles and responsibilities, coupled with the conflicts arising from them, may have led to doubts about future career prospects, particularly for this group of parents. Therefore, we expect to find a positive relationship between continued WFH and perceived changes in career prospects during the pandemic among parents whose children stayed at home for shorter periods, but not among those whose children stayed at home for longer durations (H4).
Being a mother and a newcomer to WFH
Finally, the way remotely working parents perceived changes to their career prospects during the pandemic may have also differed between men and women, as the experience of competing demands from family and work spheres while working from home was likely not uniform across gender lines. This disparity stems from the differing levels of engagement in housework and childcare responsibilities, which posed significant challenges for working parents during the pandemic. Recent evidence indicates that women, particularly working mothers, bore most of the burden of simultaneously caring for their families and working during the pandemic, potentially with more adverse effects on their future careers (Mandeville et al., 2022; Hu et al., 2023). Although men increased their involvement in childcare and housework, women's contributions grew even more, thereby exacerbating the gender gap in this domain (Landivar et al., 2023; Pabilonia and Vernon, 2023; Hipp and Bünning, 2021; Petts et al., 2021; Ruppanner et al., 2021).
In addition, the pressure caused by increased responsibilities during the pandemic negatively impacted mothers’ sleep, overall well-being, and mental health (Ruppanner et al., 2021). Access to WFH has been identified as a critical resource in enabling mothers to maintain their work engagement throughout the pandemic (Scarborough et al., 2023). However, it is important to recognise that WFH likely served more as a coping mechanism for managing increased caregiving demands rather than fundamentally challenging traditional norms that place domestic responsibilities on mothers. This is evident in research conducted during the pandemic, which shows that mothers working remotely either maintained or increased their commitment to childcare (Shockley et al., 2021). In contrast, the influence of remote work on fathers’ engagement was less pronounced and often contingent on their partners’ employment circumstances (Martucci, 2023; Hank and Steinbach, 2021; Hipp and Bünning, 2021; Zoch et al., 2021; Del Boca et al., 2020). As a result of these increased demands, working mothers reported a decline in their productivity levels compared to the pre-pandemic period, whereas fathers did not report similar experiences (Feng and Savani, 2020). Consequently, engaging in WFH may have come at a considerable cost to mothers' careers, even though it helped them maintain employment.
Considering all the above evidence, we expect to find a negative relationship between WFH and changes in career prospects during the pandemic for mothers who started working from home only during the COVID-19 pandemic (H5). Furthermore, we expect H3 and H4 to hold particularly true for mothers.
Data and methods
We use data from a unique, cross-country comparative dataset, the Familydemic Harmonised Dataset. The data collection for this survey was conducted online between June and September 2021 on representative samples of working-age mothers and fathers with dependent children, living in six countries: Canada, Italy, Germany, Poland, Sweden and the United States. The survey collected comprehensive, current and retrospective information on the lives of respondents and their families over the period starting just before the outbreak of COVID-19 (February 2020) until the time of the interview (June–September 2021). A detailed description of the choice of countries, the process of survey development, sampling, data collection and data harmonisation can be found in the Scientific Data article, available open access (see: Kurowska et al., 2023).
The total sample of this study amounts to 9,634 parents, with a slight majority of fathers (51%) and individuals with a university degree (51%). Most are under age 40 (57%) and have at least two children (58%). Regarding career prospects, a higher proportion of men (20%) indicated improvement in career prospects than women (18%). Yet, most workers reported no change (66% of men and 67% of women). There are interesting country differences, with the USA (49%) having the highest share of workers who report improved career prospects. Whilst the highest shares of no change to career prospects are reported in Germany, Poland and Sweden, all above 70%. Table 1 in Appendix shows descriptive statistics for the variables used in this study.
The dependent variable, perceived changes to career prospects, has been constructed based on respondents choosing the most relevant ending to the following statement: “Comparing the current situation with the month before COVID-19, my career prospects …”. The possible answers included: “1. Improved a lot, 2. Somewhat improved, 3. Did not change, 4. Somewhat deteriorated, 5. Deteriorated a lot”. We have created a variable with three values: no change in career prospects (answer 3), improvement in career prospects (answers 1 and 2) and worsening of career prospects (answers 4 and 5).
Our main explanatory variable related to the use of WFH has three values:
No use – No use, denoting parents who neither worked from home before the pandemic (February 2020) nor at the moment of the survey (June/September 2021);
No use – Use, denoting parents who did not WFH before the pandemic (February 2020) but did it at the moment of the survey (June/September 2021) and
Use – Use, denoting parents who did WFH before the pandemic (February 2020) as well as at the moment of the survey (June/September 2021).
Such operationalisation of the main explanatory variable enables us to test our Hypotheses (in particular Hypothesis 1) as they differentiate between those who have worked from home already before the pandemic and those who only started to WFH during the pandemic.
Our moderating variables include the number of months the children were at home and the cohabiting status of the parent. The length of time the child(ren) was at home due to childcare/school closures during the pandemic is measured with three values: (1) 0–1 month; (2) 2–5 months; (3) 6 months or more. Cohabiting status is measured as a dummy variable with those living with a partner and those who are single parents or live alone with children. The moderator related to occupational strata is recoded from ISCO 1-digit occupational codes, with 1–3 representing the managerial and professional group, and 4–10 defined as other occupations.
We estimate fixed effects multinomial logistic regression models with standard errors clustered on the country variable. We run separate models for mothers and fathers as the perceived change in their career prospects can be affected differently by the same covariates. To compare the moderation effects, we run the interaction models in which we interact WFH with the presented moderators. In all models, we control for age (in years: 20–30; 31–39; 40–49; 50–59), education (less than tertiary; tertiary), number of children (1; 2; 3+), employment sector (public; private; other), change in working hours during the pandemic (increased; no change; decreased), life satisfaction at the moment of the survey (dissatisfied; satisfied), employment type at the moment of the survey (permanent job; temporary job), whether the respondent lost or left a job during the pandemic (yes; no) and whether his/her hours of work were reduced (yes; no) as well as pregnancy occurrence during the pandemic (for mothers only) (yes; no). In models that do not use an occupational group as a moderator, we also control for occupation (ISCO 1-digit codes).
While interpreting our findings, we use predicted probabilities (estimated marginal means). Demonstrating findings using predicted probabilities, as compared to coefficients or odds ratios, is recommended as the most accurate and straightforward inference in multinomial regressions (Paolino, 2021). We evaluate whether the difference between two predicted probabilities is significant by comparing 83% confidence intervals (CIs). We do it following Austin and Hux (2002), who showed that two means differ from each other with the p-value at the p-level of 0.05 if 83% CIs do not overlap. The coefficients of the estimated models can be found in Tables 2-5 (Panels A, C) in the Appendix.
In the Appendix, we also include a robustness check, using different specifications of our main explanatory variable (Tables 2-5 panels B, D) and sensitivity analysis (Figures 2-9), running models on the subsample of countries. All details of these additional analyses are included in Appendix.
Findings
Our findings indicate that both mothers and fathers who continued to WFH throughout the pandemic had a substantially higher predicted probability of reporting improved career prospects than those who worked solely in the office, thereby providing evidence for H1 (Figure 1). The gaps between those who worked solely in the office and those who continued to WFH are 6% points (pp.) for mothers and 10 pp. for fathers. Notably, fathers who started to WFH also reported improved career prospects as compared to those who worked solely from the office by 4 pp. Among mothers, the difference between these groups is not statistically significant. Moreover, mothers who started to WFH during the pandemic were more likely to report worsened career prospects (by 5 pp.) than those who solely worked on-site and those who continued to WFH. This finding provides evidence for H5.
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of work from home” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. The vertical axis for predicted probabilities differs across the graphs. The details of the graphs are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.14 to 0.24 in increments of 0.02. The graph shows three points with vertical confidence interval bars; the points appear at 0.166 for “No use–No use”, 0.20 for “No use–Use”, and 0.226 for “Use–Use”. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.14 to 0.24 in increments of 0.02. The graph shows three points with vertical confidence interval bars; the points appear at 0.142 for “No use–No use”, 0.187 for “No use–Use”, and 0.148 for “Use–Use”. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The graph shows three points with vertical confidence interval bars; the points appear at 0.17 for “No use–No use”, 0.213 for “No use–Use”, and 0.27 for “Use–Use”. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The graph shows three points with vertical confidence interval bars; the points appear at approximately 0.146 for “No use–No use”, 0.16 for “No use–Use”, and 0.135 for “Use–Use”. Note: All numerical values are approximated.Predicted probabilities of reporting improved/worsened career prospects by WFH, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: Separate multinomial logit models were estimated for mothers and fathers, predicting the likelihood of reporting improved or worsened career prospects compared to reporting no change (the reference category). Confidence Intervals of 83% are presented. Non-overlapping confidence intervals between categories suggest statistically significant differences. The explanatory variable on the use of WFH has three levels: No use – No use (those who never used WFH), No use – Use (those who started to WFH during the pandemic) and Use – Use (those who continued to WFH during the pandemic). Control variables include: age, education, number of children, employment sector, change in working hours during the pandemic, life satisfaction, employment type, whether the respondent lost or left a job during the pandemic and whether his/her hours of work were reduced as well as pregnancy occurrence during the pandemic (for women only). Source: Created by authors
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of work from home” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. The vertical axis for predicted probabilities differs across the graphs. The details of the graphs are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.14 to 0.24 in increments of 0.02. The graph shows three points with vertical confidence interval bars; the points appear at 0.166 for “No use–No use”, 0.20 for “No use–Use”, and 0.226 for “Use–Use”. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.14 to 0.24 in increments of 0.02. The graph shows three points with vertical confidence interval bars; the points appear at 0.142 for “No use–No use”, 0.187 for “No use–Use”, and 0.148 for “Use–Use”. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The graph shows three points with vertical confidence interval bars; the points appear at 0.17 for “No use–No use”, 0.213 for “No use–Use”, and 0.27 for “Use–Use”. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The graph shows three points with vertical confidence interval bars; the points appear at approximately 0.146 for “No use–No use”, 0.16 for “No use–Use”, and 0.135 for “Use–Use”. Note: All numerical values are approximated.Predicted probabilities of reporting improved/worsened career prospects by WFH, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: Separate multinomial logit models were estimated for mothers and fathers, predicting the likelihood of reporting improved or worsened career prospects compared to reporting no change (the reference category). Confidence Intervals of 83% are presented. Non-overlapping confidence intervals between categories suggest statistically significant differences. The explanatory variable on the use of WFH has three levels: No use – No use (those who never used WFH), No use – Use (those who started to WFH during the pandemic) and Use – Use (those who continued to WFH during the pandemic). Control variables include: age, education, number of children, employment sector, change in working hours during the pandemic, life satisfaction, employment type, whether the respondent lost or left a job during the pandemic and whether his/her hours of work were reduced as well as pregnancy occurrence during the pandemic (for women only). Source: Created by authors
We also find that occupational strata are important in moderating the link between WFH and perceived changes in career prospects (Figure 2). As expected (H2), we find that managers and professionals had a higher probability of reporting improvement to career prospects not only if they used to WFH already before the pandemic, but also if they only started to WFH during the pandemic. The same is not true of other occupational groups.
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of work from home” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. Each category shows two points with vertical confindence interval bars. A legend at the bottom indicates that the blue points represent “Professional and managerial occupations” and red points represent “Other occupations”. The vertical axis ranges differ across the graphs. The details of the graph are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.10 to 0.25 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.178; Other occupations: 0.147. No use–Use: Professional and managerial occupations: 0.218; Other occupations: 0.157. Use–Use: Professional and managerial occupations: 0.242; Other occupations: 0.196. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.10 to 0.25 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.136; Other occupations: 0.158. No use–Use: Professional and managerial occupations: 0.192; Other occupations: 0.172. Use–Use: Professional and managerial occupations: 0.142; Other occupations: 0.157. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.178; Other occupations: 0.157. No use–Use: Professional and managerial occupations: 0.226; Other occupations: 0.192. Use–Use: Professional and managerial occupations: 0.282; Other occupations: 0.282. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.158; Other occupations: 0.127. No use–Use: Professional and managerial occupations: 0.158; Other occupations: 0.187. Use–Use: Professional and managerial occupations: 0.137; Other occupations: 0.157. Note: All the numerical values are approximated.Predicted probabilities of reporting improved/worsened career prospects by WFH and occupational category, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: The same as in Figure 1. Source: Created by authors
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of work from home” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. Each category shows two points with vertical confindence interval bars. A legend at the bottom indicates that the blue points represent “Professional and managerial occupations” and red points represent “Other occupations”. The vertical axis ranges differ across the graphs. The details of the graph are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.10 to 0.25 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.178; Other occupations: 0.147. No use–Use: Professional and managerial occupations: 0.218; Other occupations: 0.157. Use–Use: Professional and managerial occupations: 0.242; Other occupations: 0.196. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.10 to 0.25 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.136; Other occupations: 0.158. No use–Use: Professional and managerial occupations: 0.192; Other occupations: 0.172. Use–Use: Professional and managerial occupations: 0.142; Other occupations: 0.157. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.178; Other occupations: 0.157. No use–Use: Professional and managerial occupations: 0.226; Other occupations: 0.192. Use–Use: Professional and managerial occupations: 0.282; Other occupations: 0.282. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Professional and managerial occupations: 0.158; Other occupations: 0.127. No use–Use: Professional and managerial occupations: 0.158; Other occupations: 0.187. Use–Use: Professional and managerial occupations: 0.137; Other occupations: 0.157. Note: All the numerical values are approximated.Predicted probabilities of reporting improved/worsened career prospects by WFH and occupational category, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: The same as in Figure 1. Source: Created by authors
Yet, for mothers in managerial or professional positions who started to WFH, we also evidence an increased likelihood of reporting worsened career prospects compared to managerial and professional mothers working from the office. The same is not true for other occupational groups of mothers. This means that for mothers, being a newcomer to the world of remote work during the pandemic did not have unequivocal positive effects if one was a manager or a professional. Finally, for mothers and fathers who were already used to WFH before the pandemic, both managers/professionals and non-managers/non-professionals experienced improved career prospects.
In line with H3, the positive association between WFH and perceived changes to career prospects is only evidenced for those living with partners (Figure 3). We do not find positive associations for single individuals or those with a partner but who live separately. Notably, mothers who started to WFH during the pandemic were more likely to report worsened career prospects than those who worked solely on-site and those who continued to WFH regardless of their partnership-cohabiting status. We find no such negative association for men, but partnered fathers who continued to WFH during COVID-19 were less likely to report worsened career prospects (by 1 pp).
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of work from home” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. Each category shows two points with vertical confidence interval bars. A legend at the bottom indicates that the blue points represent “Live with partners” and red points represent “Live separately or single”. The vertical axis ranges differ across the graphs. The details of the graph are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.05 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Live with partners: 0.162; Live separately or single: 0.207. No use–Use: Live with partners: 0.201; Live separately or single: 0.208. Use–Use: Live with partners: 0.224; Live separately or single: 0.235. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.05 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Live with partners: 0.146; Live separately or single: 0.132. No use–Use: Live with partners: 0.185; Live separately or single: 0.218. Use–Use: Live with partners: 0.154; Live separately or single: 0.113. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.1. The data is as follows: No use–No use: Live with partners: 0.171; Live separately or single: 0.198. No use–Use: Live with partners: 0.223; Live separately or single: 0.159. Use–Use: Live with partners: 0.276; Live separately or single: 0.245. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.1 to 0.30 in increments of 0.1. The data is as follows: No use–No use: Live with partners: 0.147; Live separately or single: 0.173. No use–Use: Live with partners: 0.164; Live separately or single: 0.147. Use–Use: Live with partners: 0.13; Live separately or single: 0.18. Note: All the numerical values are approximated.Predicted probabilities of improved/worsened career prospects by WFH and the cohabitation relationship status, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: The same as in Figure 1. Source: Created by authors
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of work from home” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. Each category shows two points with vertical confidence interval bars. A legend at the bottom indicates that the blue points represent “Live with partners” and red points represent “Live separately or single”. The vertical axis ranges differ across the graphs. The details of the graph are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.05 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Live with partners: 0.162; Live separately or single: 0.207. No use–Use: Live with partners: 0.201; Live separately or single: 0.208. Use–Use: Live with partners: 0.224; Live separately or single: 0.235. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.05 to 0.30 in increments of 0.05. The data is as follows: No use–No use: Live with partners: 0.146; Live separately or single: 0.132. No use–Use: Live with partners: 0.185; Live separately or single: 0.218. Use–Use: Live with partners: 0.154; Live separately or single: 0.113. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.1. The data is as follows: No use–No use: Live with partners: 0.171; Live separately or single: 0.198. No use–Use: Live with partners: 0.223; Live separately or single: 0.159. Use–Use: Live with partners: 0.276; Live separately or single: 0.245. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.1 to 0.30 in increments of 0.1. The data is as follows: No use–No use: Live with partners: 0.147; Live separately or single: 0.173. No use–Use: Live with partners: 0.164; Live separately or single: 0.147. Use–Use: Live with partners: 0.13; Live separately or single: 0.18. Note: All the numerical values are approximated.Predicted probabilities of improved/worsened career prospects by WFH and the cohabitation relationship status, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: The same as in Figure 1. Source: Created by authors
The presence of children at home also proved to be an important moderator of the studied relationships (Figure 4). The positive association between WFH and improved perception of one’s career prospects are particularly visible for those whose children spent no longer than 1 month at home. This provides support for H4. Findings indicate that women who started WFH during the pandemic and continued to do so reported a positive change in their career prospects if the child(ren) stayed at home for a short period (0–1 month). This positive relationship gradually fades away with longer periods, and for mothers whose children have stayed at home for 6 months or longer, the positive association is no longer observable. In fact, such mothers are more likely to report worsened career prospects if they started to WFH during the pandemic (by 6 pp). Fathers who continued to WFH appear to experience improved career prospects regardless of the length of the period the child stayed home. We find no statistically significant difference in reported changes to career prospects between fathers who started to WFH and those working only in the office for any length of time that children stayed at home.
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of W F H” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. Each category contains three points with vertical confidence interval bars. A legend below the plots indicates that the points represent three groups according to the number of months children spent at home: blue – “0 to 1”, red – “2 to 5”, and green – “6 plus”. The vertical axis ranges differ across the graphs. The details of the graph are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.184; 2 to 5: 0.164; 6 plus: 0.158. No use–Use: 0 to 1: 0.253; 2 to 5: 0.176; 6 plus: 0.171. Use–Use: 0 to 1: 0.256; 2 to 5: 0.233; 6 plus: 0.188. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.148; 2 to 5: 0.147; 6 plus: 0.155. No use–Use: 0 to 1: 0.185; 2 to 5: 0.182; 6 plus: 0.213. Use–Use: 0 to 1: 0.15; 2 to 5: 0.133; 6 plus: 0.178. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.35 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.197; 2 to 5: 0.150; 6 plus: 0.152. No use–Use: 0 to 1: 0.257; 2 to 5: 0.182; 6 plus: 0.184. Use–Use: 0 to 1: 0.265; 2 to 5: 0.306; 6 plus: 0.254. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.10 to 0.35 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.151; 2 to 5: 0.149; 6 plus: 0.149. No use–Use: 0 to 1: 0.128; 2 to 5: 0.188; 6 plus: 0.184. Use–Use: 0 to 1: 0.131; 2 to 5: 0.14; 6 plus: 0.149. Note: All the numerical values are approximated.Predicted probabilities of reporting improved/worsened career prospects by WFH and the number of months a child stayed at home without formal care, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: The same as in Figure 1. Source: Created by authors
The figure shows four scatter plots arranged in a two-by-two grid. The horizontal axis of each graph is labeled “Change in the use of W F H” and shows three categories: “No use–No use”, “No use–Use”, and “Use–Use”. Each category contains three points with vertical confidence interval bars. A legend below the plots indicates that the points represent three groups according to the number of months children spent at home: blue – “0 to 1”, red – “2 to 5”, and green – “6 plus”. The vertical axis ranges differ across the graphs. The details of the graph are as follows: The first graph in the first row is labeled “Improved” under “Mothers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.184; 2 to 5: 0.164; 6 plus: 0.158. No use–Use: 0 to 1: 0.253; 2 to 5: 0.176; 6 plus: 0.171. Use–Use: 0 to 1: 0.256; 2 to 5: 0.233; 6 plus: 0.188. The second graph in the first row is labeled “Worsened” under “Mothers”. The vertical axis ranges from 0.10 to 0.30 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.148; 2 to 5: 0.147; 6 plus: 0.155. No use–Use: 0 to 1: 0.185; 2 to 5: 0.182; 6 plus: 0.213. Use–Use: 0 to 1: 0.15; 2 to 5: 0.133; 6 plus: 0.178. The first graph in the second row is labeled “Improved” under “Fathers”. The vertical axis ranges from 0.10 to 0.35 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.197; 2 to 5: 0.150; 6 plus: 0.152. No use–Use: 0 to 1: 0.257; 2 to 5: 0.182; 6 plus: 0.184. Use–Use: 0 to 1: 0.265; 2 to 5: 0.306; 6 plus: 0.254. The second graph in the second row is labeled “Worsened” under “Fathers”. The vertical axis ranges from 0.10 to 0.35 in increments of 0.05. The data is as follows: No use–No use: 0 to 1: 0.151; 2 to 5: 0.149; 6 plus: 0.149. No use–Use: 0 to 1: 0.128; 2 to 5: 0.188; 6 plus: 0.184. Use–Use: 0 to 1: 0.131; 2 to 5: 0.14; 6 plus: 0.149. Note: All the numerical values are approximated.Predicted probabilities of reporting improved/worsened career prospects by WFH and the number of months a child stayed at home without formal care, estimated separately for mothers and fathers: multinomial logit models, CIs 83%. Notes: The same as in Figure 1. Source: Created by authors
Discussion
In this article, we have shown how the individual work- and family-related circumstances and their interrelations impact the way the experience of WFH during the pandemic translated to parents’ perceived changes to their career prospects. These heterogeneities have been largely overlooked in previous research on the role of WFH for the career prospects of employees.
Our findings reveal that, on average, those parents who had already worked from home before the pandemic, rather than the newcomers to the world of remote work, reported an improvement in their career prospects. This subjective improvement may have stemmed from their familiarity with WFH, potentially decreasing anxiety related to the sudden change of working mode and elevating their status within the company due to their adeptness in remote work (Adisa et al., 2023; Wang et al., 2021). These employees may have drawn on their prior experience in managing work and family responsibilities while working from home, making them better prepared to handle the increased family-to-work spillovers caused by greater childcare demands during the pandemic (Augustine and Prickett, 2022; Carlson and Petts, 2022; Derndorfer et al., 2021; Farre et al., 2020). More importantly, as remote work became increasingly widespread during the pandemic, it may have contributed to a decline in the perceived flexibility stigma, particularly among those who had been working from home prior to the pandemic and had likely encountered such stigma. Furthermore, the persistently high rates of remote work long after the initial wave of the pandemic and initial signs of the weakening of flexibility stigma (Chung, 2022) likely fostered optimism that telework would become the new norm in the post-pandemic world (Barrero et al., 2021).
Our study also evidenced the privileged situation of (male) managers and professionals working from home, who were more likely to report improved career prospects than those working from the office, regardless of whether they had previous experience with WFH or not. This is likely attributable to managers and professionals being able to take advantage of higher job autonomy, flexibility and control over resources (Autor and Dorn, 2013; Drucker, 2012), as well as access to support systems (Teodorovicz et al., 2021), compared to other occupational groups. These advantages likely enabled them to adapt more flexibly to changing work arrangements and emerging demands during the pandemic, intensify their professional engagement and capitalise on new opportunities created by the rapidly evolving work environment.
At the same time, we have shown that circumstances of increased role conflicts (between being a productive employee and a present parent) and competing demands from family and work responsibilities (particularly childcare interfering with work at home) substantially diminished the positive effects of WFH on subjective changes to career prospects among the above-mentioned groups. First, the positive relationship between continued WFH and perception of one’s career prospects was not observed for those mothers and fathers who lived alone with their children. Single parents faced unique challenges in managing both paid work and caregiving responsibilities (Almeida et al., 2020; Power, 2020), which could negatively impact their ability to remain productive and focused on work. This finding is consistent with recent evidence showing that single parents were substantially more likely than partnered parents to report working less efficiently when engaging in WFH than before the lockdown (Aczel et al., 2021).
Second, we found that for mothers (also partnered ones), the positive effect of continuing WFH was weaker if children were at home for extended periods due to pandemic-related confinement measures. Juggling work, childcare and schooling placed significant strain on parents, particularly mothers, who experienced a disproportionate increase in caregiving and household duties during the pandemic (Augustine and Prickett, 2022; Carlson and Petts, 2022; Derndorfer et al., 2021; Farre et al., 2020). It has been widely documented that mothers faced increased demands and spent more time on childcare and housework than fathers during the pandemic (Landivar et al., 2023; Hu et al., 2023; Feng and Savani, 2020; Mandeville et al., 2022; Pabilonia and Vernon, 2023; Hipp and Bünning, 2021; Petts et al., 2021; Shockley et al., 2021; Ruppanner et al., 2021). Stefanova et al. (2023) found that during the COVID-19 pandemic, women allocated more time to caregiving and significantly less time to work than men. Correspondingly, women reported feeling more dissatisfied with the division of housework labour than men (Craig and Churchill, 2021). These strains must have been even more difficult for mothers, who were the newcomers to the world of work. Our findings show that for mothers, being a newcomer to WFH was related to reporting worsened career prospects, mainly when the mother lived without a partner and her children stayed home for extended periods.
Third, the greater increase in childcare responsibilities faced by mothers during the pandemic may also help explain another finding from our study, that only fathers in managerial or professional roles (unlike mothers in these roles) reported improved career prospects, even if they were new to remote work. Mothers working from home, who were burdened with increased childcare responsibilities during the pandemic to a greater extent than fathers, were less able to leverage their privileged occupational positions in terms of job autonomy, control over resources and access to support systems, to improve their productivity and standing within the company (Guy and Arthur, 2020; Zanhour and Sumpter, 2024). Our findings correspond well with previous evidence that confirms that the more caregiving women performed during the pandemic relative to other tasks, the more negative their self-reported career outcomes were, specifically in terms of career self-efficacy and aspirations (Stefanova et al., 2023).
Overall, our findings underscore the differential professional realities experienced by varying groups of working parents, with pronounced differences across gender lines. Fathers tend to benefit more straightforwardly from WFH regarding career advancement (regardless of occupational gradient), while mothers, due to more acute experiences of role conflict and childcare demands that successfully compete with work demands, are less likely to take advantage of the opportunities offered by WFH. Thus, our findings are relevant for supporting working parents in the post-pandemic context, as increased family demands extend beyond the pandemic itself. Various groups of employees, such as those caring for children with disabilities or ageing parents, could benefit from improved alignment between professional responsibilities and caregiving duties while working from home. In this regard, our results have continued relevance for contemporary work–life integration challenges.
Importantly, recent literature suggests that managerial stigma toward remote workers persists in the post-pandemic period (Matysiak et al., 2025; Chung and Wang, 2024; Kasperska et al., 2024; Moller et al., 2024a). This is despite remote employees being as productive as their office-based counterparts (Bloom et al., 2024). Emerging post-pandemic evidence from the United States further reveals that remote-working employees with children perceive themselves as likely to be penalised through reduced pay and diminished job satisfaction for engaging in this mode of fluid working (Moller et al., 2024b). This is particularly noteworthy given that the initial phase of the pandemic and widespread use of WFH were accompanied by a diminishing of flexibility stigma and improvement in managerial attitudes towards those working remotely (Chung, 2022). Future (ideally cross-national) research should investigate the psychological mechanisms underlying these changes, including the potential role of negative emotions such as resentment or disappointment with the continuation of flexibility stigma and differential treatment of those working remotely. In light of these developments, our study offers a timely and valuable foundation for exploring these shifts in employee attitudes in the future.
Limitations
Although our study encompasses multiple countries and a large sample of over 9,000 employees, a limitation of this research is the use of cross-sectional data. Consequently, we cannot establish causality or definitively determine the direction of the effects (Antonakis et al., 2010). Endogeneity remains a potential concern, including reverse causality, as the perception of changes in career prospects could have, at least partially, influenced the choice to WFH. The issue of reverse causality is only partially mitigated by the fact that the dependent variable is measured at the time of the interview, while the main explanatory variable includes retrospective data from before and during the pandemic.
Another limitation is the possibility of sample selection bias, as individuals with better career prospects may self-select into WFH arrangements. Such concerns have been previously noted in the literature on WFH based on survey data (Allen et al., 2015). In our study, the category of individuals who began WFH during the pandemic may be less affected by selection issues, as for many, WFH was imposed by the unique circumstances of the pandemic. Future research should address these limitations by employing (quasi)experimental designs, which would enable the identification of causal effects and mitigate sample selection bias.
While our sample included respondents from only six countries, these countries represented diverse socio-economic contexts and the main types of welfare regimes. This diversity enhances the generalizability of our findings, as it reflects a range of policy and cultural environments. Nevertheless, conducting cross-country comparisons posed challenges due to the limited number of countries. For instance, hierarchical modelling could have provided a more comprehensive analysis; however, the number of countries would need to be at least 25–30 to yield unbiased results (Bryan and Jenkins, 2016). Additionally, due to the limited number of respondents per country, we were unable to estimate triple interactions (gender × WFH × moderator) by country to investigate whether our results hold within each country included in our pooled sample. Nonetheless, we conducted additional sensitivity analyses on two subsamples of countries, excluding those with particularly favourable remote-work policies, a high prevalence of working from home, or family-friendly policies that could have influenced our results. The main findings remained consistent across all models (see the Appendix). Future studies should investigate the role of contextual differences and national particularities in the relationship between WFH and perceived career prospects using datasets that enable such analyses. Furthermore, future (quasi)experimental studies, as well as in-depth qualitative analyses, should try to explore in detail the hypothesised mechanisms behind the evidenced relationships in our study.
Further analysis could also benefit from mitigating potential bias arising from variations in remote work policies across different industries or organisational cultures. Although this study controls for sector, remote work policies could have differed across industries within a particular sector. Likely, unequal access to remote work infrastructure, changes in policies or varying attitudes toward WFH across organisations (e.g. Kasperska et al., 2025) may have influenced the results, but these factors could not be fully captured in our analysis. Future research could also consider the long-term implications of WFH and provide evidence on the post-pandemic period, and not only the change in career prospects that occurred during the pandemic, as explored in this study.
Finally, varying personal-level characteristics, such as personality traits, work motivation and actual job autonomy, could significantly influence how employees perceive changes to their work situations and career prospects. Future research should expand this analysis to include such factors to gain a more comprehensive understanding of how different employee profiles interact with WFH scenarios.
Theoretical and practical contributions
Despite these limitations, our study contributes to a deeper understanding of the complex effects of the expansion of WFH during the COVID-19 pandemic on how employees with children perceive their career prospects across various work- and family-related contexts.
First, we highlight the critical role of prior experience with WFH before the pandemic and how this experience significantly influenced employees’ positive perceptions of career prospects while working from home during the pandemic. In doing so, we address a research gap that has received limited attention, with some exceptions (e.g. Kasperska, 2025), despite calls from experts for further investigation (Caligiuri et al., 2020). We also propose theoretical explanations, embedded in cognitive appraisal theory, for the empirically observed advantage of those who had already been working from home before the pandemic. These explanations go beyond the notion of an initial comparative advantage in remote work know-how and the ability to balance work and caregiving responsibilities more effectively than those new to remote work. We emphasise the potential role of increased optimism among pre-pandemic remote workers, driven by the rapid and widespread normalisation of remote work during the pandemic. This optimism likely stemmed from the belief that the flexibility stigma, something many of them had likely experienced before the pandemic, would diminish or even disappear. This novel line of reasoning invites future research to further empirically test our theoretical proposal.
Second, we extend theoretical considerations on the adverse psychological outcomes of role conflicts (Carlson and Petts, 2022; Greenhaus and Beutell, 1985) between being a present parent and a productive employee, as well as objective outcomes resulting from competing demands (Greenhaus and Powell, 2003; Kahn et al., 1964) between family and work spheres. Previous literature has predominantly focused on the negative effects of these dynamics on (mental) well-being (e.g. Hecht, 2001), job satisfaction (e.g. Kossek and Ozeki, 1998), productivity (e.g. Deole et al., 2023) or work commitment (Rosen et al., 2024). We expand these discussions by demonstrating that situations of heightened family-related demands may limit the positive effects of WFH on employees' perceptions of their career prospects and, in some cases, even lead to negative outcomes, particularly for mothers.
Third, we contribute to existing theoretical debates on the privileged position of managerial and professional workers in the context of WFH during the pandemic. We further highlight the gendered differences in how this privileged position translated into perceptions of changes in career prospects. In particular, we show that interferences from the family sphere, especially among mothers, who are bearing care responsibilities to a much larger extent than fathers, likely hindered their ability to fully capitalise on job autonomy, flexibility and access to organisational resources while working from home.
Our study has practical implications that go beyond the pandemic context. The findings clearly demonstrate that those with increased demands in their personal spheres are not able to benefit from fluid working as much as other employees. Organisations could use personalised work plans to reflect individual employees’ needs, preferences and personal responsibilities. Empirical evidence supports the notion that flexibility and individualisation in work structures are highly valued by employees (Kossek et al., 2015). Additionally, organisations should actively monitor the experiences of remote workers by gathering feedback and committing to continuous improvement, which is essential in navigating rapidly changing work environments. Particular attention should be given to the perspectives (and sentiments) of fluid employees, especially given the previously discussed evidence of unmet employee expectations regarding career development for remote workers and the persistence of flexibility stigma. Recent media discourse has highlighted growing tensions between employers and employees, with some organisations mandating a return to the office, often prompting strong resistance from the employees (Partridge, 2025; The Economist, 2023). In this context, it is vital for organisations to establish open channels of communication with employees to discuss expectations, explore the feasibility of fluid working policies and ensure transparency in their implementation.
Finally, in light of the ongoing technological revolution, increasing automation, and the rapid advancement of AI, employees may find themselves adapting to new work solutions while simultaneously facing competing demands from their personal lives (Alfes et al., 2022). Our findings underscore that while some employees adapt effectively to such new challenges (especially those with more resources and prior exposure), others may experience heightened strain or reduced capacity to adjust. This calls for the development of appropriate support mechanisms to facilitate seamless technological transitions while accommodating familial obligations. Organisations should invest in training programs to help employees start using new technologies or enrolling in fluid working arrangements, equipping them with digital skills and knowledge of available tools. By implementing these suggestions, organisations can better address the diverse needs of their workforce, foster inclusive and effective remote work practices and support employees in managing their careers efficiently.
The authors gratefully acknowledge financial support from the National Science Centre in Poland (NCN, grant no. 2020/37/B/HS5/02703). We thank participants of internal seminars at the Interdisciplinary Centre of Labour Market and Family Dynamics (LabFam) and attendees of the Work and Family Researchers Network conference for their helpful feedback on earlier drafts. We are also grateful to the anonymous reviewers of Personnel Review for their constructive comments and suggestions.
The supplementary material for this article can be found online.

