Introduction

Although European countries have been actively promoting gender equality in education and the labor market for decades, women still hold a disadvantaged position in households. For example, in the United Kingdom, women’s (especially mothers’) participation in the labor force is still hindered by the fact that they are still responsible for a greater share of housework and face other work-family conflicts. This reduced participation in the labor force, along with the long-hours that women must be spent doing unpaid housework, has a negative effect on women’s mental health and well-being (Alyaemni et al., 2013). In addition, employed women may not always be able to access flexible work arrangements to cope with their work-family conflicts and maintain their careers after childbirth (Chung & van der Horst, 2018). Without flexible working, fixed work schedules and long working hours can lead to role conflicts and mental health problems. However, previous studies indicate that even if flexible work arrangements are available, women may not use them due to fear of losing potential promotion opportunities and facing discrimination from employers (Chung & van der Horst, 2018; Munsch, 2016). This is because the traditional gender norms expect women to work flexibly for family responsibilities, which violates the ideal work norm (Chung & van der Horst, 2018). Similarly, men who use flexible work arrangements are also likely to be stigmatized because the traditional gender norm expects men to be the primary breadwinners and using flexible working violates the expected gender roles for men (Munsch, 2016).

In sum, women in the UK face the dilemma of choosing between paid inflexible work and unpaid housework. However, while the adverse effects of inflexible working and unpaid housework on employees’ well-being are well studied, we know little about which status (doing unpaid housework or inflexible working) has a greater negative impact on women’s mental health. Thus, the first objective of the study is to provide a more comprehensive and integrated theoretical framework in order to compare the effects of unpaid housework and inflexible working on women’s mental well-being. Furthermore, women’s parenting responsibilities are one of the main factors affecting their participation in the labor force, and a major source of housework and work-family conflicts (Grinza et al., 2017; Kalucza et al., 2015; Katz-Wise et al., 2010; J. Kim, 2018; Nomaguchi, 2012). Therefore, the second objective of the study is to examine whether parenthood can moderate the effects of flexible working and unpaid housework on women’s mental well-being.

Unpaid Housework Versus Inflexible Working

Women often face a choice between inflexible working and unpaid housework, both of which have more significant negative impacts on their mental health. However, previous research makes conflicting predictions about which status is worse for women’s mental health, and surprisingly no work has compared the mental health effects between the two. Therefore, we formulate our hypotheses based on conflicting theoretical arguments and research findings.

On the one hand, previous research supports the idea that paid inflexible work is better than unpaid housework. Firstly, work is a central activity in our lives that is the primary source of identity for most adults in contemporary society (Tausig, 2013; Wang et al., 2022b, c). Unemployment or the lack of paid work can lead to the loss not only of income, but also of more intangible psychosocial benefits (Burchell et al., 2020). Specifically, Jahoda’s Latent Deprivation Theory holds that employment is vital for individuals, providing numerous essential latent functions (e.g., time structure, enforced activity, social contact, etc.) and manifest functions (e.g., income) to better individuals’ mental well-being (Jahoda, 1982). In addition, Jahoda (1982) argued that the development of identity, social position and collective integration brought by paid work for individuals cannot be replaced by unpaid work in the capitalist society. The Agency Restriction Theory also holds that the financial hardships brought by unemployment make it difficult for people to manage their lives, and so negatively affect their psychological status (Fryer, 1986). These theoretical frameworks suggest that paid work can benefit women’s mental health by fulfilling psychological needs, even though the work is inflexible. Second, previous research has shown that doing unpaid housework negatively affects women’s mental health and well-being. For example, many studies indicate that the number of hours women spend doing unpaid work is strongly associated with increased symptoms of depression and psychological morbidity (Esteban-Gonzalo et al., 2018). Specifically, studies on women’s unpaid housework in Canada, the United States, Greece and Spain have shown that women experience more stress than men because they have more hours of unpaid work and higher rates of unemployment (Drydakis, 2015; Esteban-Gonzalo et al., 2018; MacDonald et al., 2005; Piovani & Aydiner-Avsar, 2021). In particular, women who undertake unpaid housework need to spend time and effort in addition to their paid work, and yet their efforts are traditionally not recognized as “work” (MacDonald et al., 2005). Thus, Jung and O’Brien conclude in their review that unpaid housework can significantly and negatively impact women’s mental and physical health, as well as their interpersonal, professional, and economic status (Jung & O’Brien, 2019).

Taking the theoretical arguments and empirical findings mentioned above together, we expect that women who only do unpaid housework might be deprived of social participation and group activities, leading to financial distress and increased depressive symptoms. Therefore, we develop hypothesis 1A:

  • H1A: Women who work in paid inflexible jobs have significantly better mental health than those who only do unpaid housework.

On the other hand, many other studies have found negative impacts of inflexible work arrangements on employees’ mental well-being (Ala-Mursula et al., 2004; Costa et al., 2004; Jeffrey Hill et al., 2008). Firstly, the ‘work-family conflicts’ theory (Frone et al., 1992) holds that inflexible working can increase employees’ work-family conflicts, thereby harming employees’ well-being (Chandola et al., 2019; Chung & van der Lippe, 2018). Both work-to-family and family-to-work conflicts can increase the risk of symptoms of poor mental health, such as ‘burnout’ and depression (Burchell & Fagan, 2004; Kamerāde et al., 2019). Secondly, women are more vulnerable to inflexible working. Chung and van der Lippe (2018) found that women have a higher demand for flexible work arrangements in order to meet caregiving obligations or other unpaid housework than men, especially after transitioning into parenthood (Singley & Hynes, 2005; Sullivan & Lewis, 2001). However, inflexible work hinders mothers’ ability to participate in the labor market because it is hard to square the obligations of motherhood with the fixed time shift and workspace. For example, most jobs have a work-day that finishes at 5:00 p.m., while most primary schools in the UK finish between 3:30 and 4:00 p.m. Thirdly, previous studies found that inflexible work arrangements also lead to greater work stress and burnout (Jeffrey Hill et al., 2008), even for women without children. In addition, nationally representative evidence in the UK shows that women’s lower levels of work schedule control predict higher allostatic load (worse mental health) when controlling the presence of children and their gender role attitudes (Wang et al., 2022, b, c). Fourthly, middle-class women are more likely to work flexibly to devote themselves to more housework due to social restrictions on their gender roles (Clawson, 2014), which was concluded by scholars as the concept ‘do gender’ (West & Zimmerman, 2009). Hence, to alleviate work-family conflicts and be constrained by social norms, women generally tend to have higher demands for flexible working than men. This is because women who are able to stay at home and focus on housework conform to the dominant gender norms, and do not suffer work-family conflicts to the same extent, which may protect their mental health.

Taken together, we argue that women who work in paid inflexible jobs might have a higher risk of work stress and work-family conflicts, leading to a worse mental health status compared with those who only do unpaid housework. Therefore, we develop hypothesis 1B.

  • Hypothesis 1B: Women who only do unpaid housework have significantly better mental health than those who work in paid inflexible jobs.

The Moderating Role of Parenthood

Next, we explore whether the mental health impacts of paid inflexible work and unpaid housework depend on the presence of dependent children, namely the moderating effects of parenthood. Ample research shows that transitioning into parenthood predicts higher demand for using flexible working and doing housework and caring (Baxter et al., 2008; H. Kim et al., 2019), while no evidence shows the impacts of work arrangements or doing housework on the willingness to transition into parenthood. In addition, parenthood has been found to play a moderating role in the relationship between flexible working and work-family conflicts (Abendroth, 2022). We have two competing hypotheses depending on whether inflexible working is better than housework for women’s mental health.

On the one hand, if inflexible working is better for women’s mental health than unpaid housework, we expect that the benefits of inflexible working are more pronounced among women without children for the following reasons. Firstly, the transition into parenthood is an important life event, which will inevitably lead to numerous potential challenges (including changes in social relationships, more housework hours and worse mental well-being) (Evenson & Simon, 2005; Helbig et al., 2006). After transitioning into parenthood, both men’s and women’s family duties and jobs are differentiated (Belsky & Kelly, 1994). Many studies report that females do more housework following the transition to motherhood (Schober, 2013). For example, Baxter et al. (2008) found remarkable consistency in men’s housework hours throughout the transition into fatherhood but a significant increase in women’s housework hours during the transition into motherhood (Baxter et al., 2008). Therefore, women’s mental issues and work-family conflicts brought by inflexible work might be more pronounced among women with children due to the unequal division of parenting responsibilities between men and women.

Secondly, researchers have suggested that the persistent effects of inequalities in family work in husband-wife partnerships result from “doing gender” (Kan, 2008). Doing gender entails consistently establishing and reinforcing the distinction between men and women via acts and accomplishments. Women express and build their feminine identity by doing household chores and caring for their children (Bittman et al., 2003). In addition, for mothers, the happiness of taking care of their children and the satisfaction of their mother’s responsibilities may alleviate their negative emotions. Individuals’ different roles in society have their corresponding responsibilities and obligations. Thus, we formulate the following hypothesis:

  • H2A: If there are mental health benefits of paid inflexible work over unpaid housework, these are more pronounced among women without children.

On the other hand, if inflexible working is worse than unpaid housework, we expect the relative benefits of unpaid housework to be more pronounced among women with children. For women with children, inflexible working is more likely to cause work-family conflicts since taking care of children needs to be handled flexibly (e.g. getting up in the middle of the night to breastfeed or pick up the child from school during working hours). In other words, mothers who do only unpaid housework can more effectively deal with the tasks of parenting when compared with mothers in paid work. The conflicts between work and family are sourced from the distinct expectations brought by different social roles (Frone et al., 1992; Grönlund, 2007), which can lead to a series of mental health problems, such as burnout, mental distress and anxiety (Jeffrey Hill et al., 2008). Therefore, women with children will benefit more from unpaid housework, as it will enable them to avoid conflicts between parenting and work.

By contrast, for women without children, work-family conflicts are much less pronounced. Thus, their unpaid housework may no longer provide any benefits for their mental health, but would instead reduce their social engagement and connections, leading to mental health problems (Fryer, 1986; Jahoda, 1982). Moreover, for women without children, doing unpaid housework will not only deprive them and their families of a source of income, but will also decrease their sense of self-realization and self-confidence, leading to a range of mental health problems (Jahoda, 1982). Taken together with the arguments above, we hypothesize that:

  • H2B: If there are mental health benefits of unpaid housework over paid inflexible work, these are more pronounced among women with children.

Method

Data and Sample

This study used the second (2010–2012), fourth (2012–2014), sixth (2014–2016), eighth (2016–2018) and tenth (2018–2020) waves of the United Kingdom Household Longitudinal Study (UKHLS) as these waves contain consistent measures of flexible work arrangements and mental health. The UKHLS is the largest longitudinal study in the UK and its purpose is to collect high-quality longitudinal data on various aspects of social life in order to help gain a deeper understanding of the long-term effects of socio-economic changes and policy interventions on the wellbeing of the UK population. In the first wave, the UKHLS used a stratified and clustered sampling design to collect a sample of around 40,000 households. The households in the first wave were visited and interviewed every year to collect data on changing household and individual lives. The average response rate is around 60% among the five waves. To construct the analytic sample, we excluded male respondents as our focus is on females. We also excluded full-time students, retired and long-term disabled who are not relevant to this study. Moreover, we excluded self-employed people, who are usually a highly self-selected population group. After excluding a small number of missing data (around 5%), our final sample included 23,552 respondents and 60,441 person-wave observations. See Table 1 for more sample details.

Table 1 Sample descriptive statistics
Full size table

Measures

Dependent Variable

Mental health was measured using the 12-item General Health Questionnaire (GHQ-12). GHQ-12 is a widely used and validated scale, which can represent respondents’ depressive and anxiety symptoms, sleeping problems, and overall happiness (Goldberg & Hillier, 1979). In the UKHLS, the sum score of respondents’ answers to the 12-items was converted to a single continuous scale ranging from 0 (the least distressed) to 36 (the most distressed), with a higher score indicating better mental health since the score has been reverse coded.

Independent Variable and Moderator

The key independent variable was employment/job status, which is broken down into four categories. The first category refers to female respondents who are full-time homemakers and without paid jobs. The second category refers to unemployed women who are looking for jobs and who can start a job within two weeks. The third and fourth categories comprise respondents who are in paid work without and with flexible work arrangements (e.g. part-time working, working term-time only, job sharing, flexi-time, compressed week, annualized hours, and working from home). We also conducted further analyses to distinguish between three types of flexible work arrangements: reduced hours, flexible time and flexible place. The moderator in this study is parenthood, which measures whether respondents have any dependent children aged 15 or below in their household. The threshold (15 years old) was used in the dataset because in the UK children when turning 16 years old would lose a range of child benefits and start to be regarded as adults (for more details, see https://www.gov.uk/child-benefit-16-19).

Control Variables

Based on previous research, this study identified several potential unobserved confounders that can significantly influence women’s mental health (Alyaemni et al., 2013; Esteban-Gonzalo et al., 2018; Seedat & Rondon, 2021). Since the study used fixed-effects regression models, it controlled for all confounders except those that vary over time. It is worth noting that women’s marital status has been found that correlates to their use of work arrangements (Chung & van der Lippe, 2018) and unpaid housework loads (MacDonald et al., 2005), especially for women with the presence of children. For example, evidence from the American Time Use Surveys (ATUS) indicates that married mothers spend more time on unpaid housework than those who are single or divorced (Pepin et al., 2018). Therefore, it is necessary to control women’s marital status when investigating the impacts of their work arrangements and the role of parenthood. The final selected demographic confounders included age, marital status, logged household income, the presence of longstanding illness, and the number of children.

Analytic Strategy

This study uses two-way fixed-effects regression models as the analytic strategy. Compared with cross-sectional analysis, fixed-effects regression can eliminate any unobserved heterogeneity by only comparing variation within individuals (Allison, 2009). This means that we can identify whether changes in employment status for individual women correlate with changes in their mental health. Thus, this study is able to eliminate the confounding effects arising from all time-constant variables, thereby producing a more accurate estimate of the relationship between employment status and women’s mental health (Allison, 2009; Li & Wang, 2022). We first analyzed the impact of employment status on women’s mental health and started our analyses with the following model.

$${Mental\ health}_{it}={\alpha}_t+{\beta}_1 Employment/{job}_{it}+{\beta}_2{Covariates}_{it}+{T}_t+{\mu}_i+{\varepsilon}_{it}$$

Where Mental healthit refers to mental health status of individual i at time point t, αt refers to the intercept that may vary across time, β1 is the coefficient for the key independent variable (employment and job statuses), β2 is the coefficient for covariates, Tt refers to the effect of time, μi refers to the time constant error term which will be excluded during the estimation, and εit refers to the time-varying error term. Then, we examine whether the presence of children can moderate the impacts of employment status on women’s mental health.

Results

The Impacts of Employment Status on women’s Mental Health

Table 2 reports several fixed effects models examining the effects of women’s employment status on their mental health. Model 1 shows that, compared with women who do only unpaid housework, women who use paid inflexible work have significantly better mental health (coefficient = 0.59, SE = 0.12, p < 0.01). Further analysis shows that women who use flexible work arrangements have significantly better mental health than women who do only unpaid housework (coefficient = 0.72, SE = 0.12, p < 0.01). In addition, the results also show that unemployed women have worse mental health than women who do only unpaid housework (coefficient = −0.59, SE = 0.14, p < 0.01). The effects of paid inflexible work on women’s mental health remain significantly positive from model 1 to model 4, even after controlling for a series of time-varying confounders. Further additional analyses show that women who use flexible work arrangements have significantly better mental health than those who do not work flexibly (coefficient = 0.13, SE = 0.06, p = 0.035). In sum, the results of models in Table 2 are consistent with hypothesis 1A, confirming that paid inflexible work results in significantly better mental health than unpaid housework alone.

Table 2 Two-way fixed effects models examining the effects of different employment/job statuses on women’s mental health
Full size table

The Interaction Effects between Different Employment/Job Statuses on women’s Mental Health

Table 3 describes the interaction between employment status and the presence of children. The results of model 1 in Table 3 generally show a significant positive interaction effect between paid inflexible work and the presence of children (coefficient = -0.49, SE = 0.24, p < 0.05). There is also a significant interaction effect between flexible work (of all types) and the presence of children (coefficient = -0.51, SE = 0.23, p < 0.05). Moreover, the results of models 2, 3 and 4 in Table 3 also show that there are significant and positive interaction effects between different types of paid flexible work and the presence of children. By contrast, the interaction effects between unemployment and the presence of children are insignificant, as shown in Table 3. Figure 1 plots the coefficients of the interaction terms in model 1 based on Table 3. The results of model 1 and the information in Fig. 1 confirm hypothesis 2A that the mental health benefits of paid inflexible work over unpaid housework are more pronounced among women without children. Similarly, the mental health benefits of paid flexible work over unpaid housework are also more pronounced among women without children.

Table 3 Two-way fixed effects models examine the interaction effects between different employment/job statuses on women’s mental health
Full size table
Fig. 1
figure 1

The interaction effects between different employment/job statuses on women’s mental health

Full size image

Discussion and Conclusions

Although previous research widely documents the negative impacts of inflexible working and unpaid housework on women’s mental health, no research so far has integrated both areas of research into one framework and compared the effects of both statuses on the mental health of women. To fill in this gap in the literature, this study used a nationally representative sample of 23,552 British workers and fixed-effects models to examine whether paid inflexible work is better for women’s mental health than unpaid housework. Moreover, it investigated whether the presence of children moderates this effect. The research yielded the following important findings.

Firstly, the predictions made by previous research are inconsistent with the existing evidence about which employment status is better for women’s mental health. This study contributes an important insight into the literature about the health consequences of women’s choice between paid inflexible work and unpaid housework. On the one hand, the study finds that women who work in paid inflexible jobs have significantly better mental health than those who do only unpaid housework. This supports Jahoda’s Latent Deprivation Theory and Fryer’s Agency Restriction Argument. Both unemployment and unpaid housework worsen women’s mental health by depriving them of the latent and manifest functions that work plays in people’s lives. Conversely, paid inflexible work can perform these important functions. On the other hand, the findings provide important insights into the debates about the work-family conflicts brought by inflexible working and the stigma towards flexible working. Although previous theoretical arguments and empirical findings raised concerns about the potential negative impacts of inflexible and flexible working (Chung, 2018), our findings suggest that both are better than unpaid housework, at least when it comes to women’s mental health. The findings are overall consistent with two nationally representative studies that confirmed that paid work (whether flexible or inflexible) is better than unemployment (including unpaid housework) for women’s mental health (Kamerāde et al., 2019; Wang et al., 2021).

Secondly, the study extends previous studies on paid inflexible work and unpaid housework by considering the moderating effects of the presence of children. Our findings indicate that the mental health benefits of paid inflexible work over unpaid housework are indeed more pronounced among women without children. The moderating role of children was identified using a longitudinal analysis on a nationally representative cohort. This is a significant contribution to the research on women’s work and mental health. Conversely, while many studies have found that paid inflexible work increases women’s work-family conflicts and burnout (Beutell, 2010; Jeffrey Hill et al., 2008; Singley & Hynes, 2005), we did not know whether mothers would face the same negative mental effects of unpaid housework. The presence of children potentially leads to more housework, but it was unclear whether this would bring more work-family conflicts and so stress, or whether women with children would be able to adapt (Kalucza et al., 2015; Singley & Hynes, 2005; Yavorsky et al., 2015). Consistent with the former, our findings suggest that individuals’ mental health and subjective well-being are adaptive to life events or life goal changes (Diener, 1984; Zhao et al., 2017). Once they have children, most women (especially middle-class women) shift the focus of their lives from work to family (Baxter et al., 2008; Belsky & Kelly, 1994; Singley & Hynes, 2005). Therefore, our findings partially support the idea that women are more likely to devote themselves to more childcare and housework after transitioning into parenthood, which is significant for how we understand gender inequality in the UK. Future research needs to consider the presence of children as a social determinant of women’s mental health, and to further investigate the moderating role of parenthood in the health impact of different employment statuses.

Thirdly, there are some limitations to this study, which point to potential areas for future research. One shortcoming of this study is that it only focuses on mental health. Future research can compare the effects of employment and economic status on a broader range of wellbeing and quality of life indicators such as life satisfaction, social participation, and whether participants are satisfied with their marriages. Moreover, while the study proposes a number of possible explanations for why inflexible working is better for women’s mental health than doing only unpaid housework, it does not test these possible mechanisms due to limitations in the data. As a result, future research should test the potential mechanisms underlying the mental health advantages of inflexible working over unpaid housework, such as having a more structured work schedule and greater financial security. Finally, although this study used fixed-effects regression models to eliminate many unobserved variables, it cannot establish any causal relationships because it did not control for factors that vary over time. Further research can also explore whether the impacts of women’s employment statuses differ across countries or occupational classes. Due to the differences in cultural norms and macroeconomic environments between different countries, researchers should carry out cross-national or cross-cultural comparative studies to see whether the results of this study are robust across different social groups.

In conclusion, this study shows that women who work in paid inflexible jobs have significantly better mental health than those who do only unpaid housework, especially for women without children. The trade-off between paid inflexible work and unpaid housework is important for addressing women’s mental health problems. By comparing the mental health impacts of paid inflexible working and unpaid housework, this study provides a more comprehensive and integrated theoretical framework to study women’s mental health at both domestic and public spheres. Although flexible working is an important component of job quality, future research could profitably compare the mental health effects of other job characteristics such as social environment, job prospects with doing unpaid housework. Regarding policy implications, the results of this study suggest that the government’s priority should guarantee women’s employment because even employment in jobs without flexible work arrangements is of great importance for women’s mental health, highlighting the manifest and latent functions of employment. Finally, as flexible working can provide additional benefits to employed women’s mental health, government and managers should consider providing more flexible working opportunities for women with and without the presence of children.