Abstract

This paper outlines a methodological approach to examining individual wellbeing within married couples that incorporates each partner’s allocation of time. Drawing on data from the Sample Survey of Daily Time Use by the Population (orig. Выборочное наблюдение использования суточного фонда времени населением 2024), we construct multiple‑indicator, multiple‑cause (MIMIC) models using ALR‑transformed time‑use regressors to identify the association between wellbeing and the relative redistribution of time. Wellbeing, treated as a latent construct, is operationalised through indicators of human capital (health status and educational attainment) and material resources (income). The results demonstrate a robust and statistically significant relationship between wellbeing and the time‑allocation decisions made by individuals and their partners. The analysis also reveals gendered patterns in this relationship, alongside differences between weekends and working days. The proposed integrated approach enables the estimation of partners’ wellbeing – despite its inherently non-observable nature – while explicitly accounting for each partner’s time allocation.

Keywords

time allocation, wellbeing, time budgets, married couples, gendered division of labour

JEL codes: I31, J16, J18, J22

Introduction

In contemporary research on the wellbeing and quality of life of married couples, comparatively little attention has been paid to the allocation of spouses’ time budgets. Yet partners’ decisions about how to distribute time between paid work, housework, childcare, leisure, and other activities reflect not only individual preferences and priorities, but also social norms, structural constraints, and gendered expectations. Analysing how spouses allocate their time therefore provides important insights into how the organisation of daily life is associated with wellbeing. However, the nature of the relationship between the structure of a family’s time budget and the wellbeing of its members remains insufficiently understood.

A number of studies have demonstrated links between one partner’s time allocation and various aspects of the other’s wellbeing. For example, family functioning and wellbeing are closely associated with work schedules (Fursman 2009). Evidence from US-based IT professionals shows that women whose partners work more than 50 hours per week are more likely to experience stress and time pressure, whereas no comparable association is observed for men whose spouses work overtime (Shafer et al. 2018).

Employment status also plays a significant role: unemployed individuals report lower levels of life satisfaction than those in work (Hoang, Knabe 2021), and job loss affects not only the wellbeing of the unemployed partner but also that of their spouse (Voßemer et al. 2024).

Among dual‑earner couples, higher levels of marital satisfaction and a more equitable division of domestic responsibilities are more commonly observed (Helms et al. 2010). At the same time, the easing of constraints imposed by paid employment influences the distribution of domestic labour in gender‑specific ways: the gender gap widens on days when the wife is not working and the husband is, and narrows in the opposite situation (Álvarez, Miles‑Touya 2019).

Gender inequality in the distribution of time and men’s participation in domestic tasks is also highly relevant to wellbeing. Men’s involvement in housework and childcare has been shown to enhance their wives’ wellbeing (Khawaja, Habib 2007; McCann et al. 2024), while inequalities in these domains constitute a risk factor for women’s health (Hiekel, Kühn 2024). A more equal division of domestic responsibilities is likewise positively associated with couples’ intentions to become parents (Leocádio et al. 2024; Kan et al. 2019; García‑Manglano et al. 2015).

Beyond the structure of time allocation itself, researchers also examine how time is spent – whether jointly or separately. Time spent together is particularly important for happiness and overall life satisfaction (Flood, Genadek 2016). However, regardless of whether couples follow traditional or egalitarian role divisions, the amount of shared time tends to be similar. The arrival of children reduces the time partners spend together, although the magnitude of this effect varies with the child’s age (Flood, Genadek 2016).

Taken together, existing research demonstrates that an individual’s wellbeing is shaped not only by their own lifestyle, as reflected in their time allocation, but also by their partner’s. Differences in the time spouses devote to work, housework and other activities constitute important determinants of individual wellbeing. In this context, examining the temporal determinants of wellbeing using Russian data is particularly timely.

In this study, we test – using Russian data for the first time – the hypothesis that an individual’s wellbeing is associated with both their own time allocation and that of their spouse.

Data

Empirical basis

To test the hypothesis that time allocation within couples is associated with individual wellbeing, we draw on data from the Sample Survey on the Use of Daily Time by the Population, conducted by Rosstat in 2024. The analysis includes observations for individuals aged 20 and over who reported the distribution of their daily time across seven broad activity categories, with a minimum total of 720 minutes (12 hours) per day. Observations in which reported time spent on sleep and paid employment exceeded 1.000 minutes (approximately 16 hours) per day were treated as outliers and excluded from the analysis.

From the full dataset, we selected individuals in registered marriages for whom information on their spouse was available. The household types included in the sample were restricted to ‘married couple’ and ‘married couple with children under 18’. This restriction reflects the aim of the study, which is to identify the association between wellbeing and time allocation among married couples. After applying the selection criteria, the final sample comprised 14.682 couples on weekdays and 14.812 couples at weekends.

The variables used in the analysis are presented in Table 1.

Table 1.

Download as

CSV

XLSX

Description of the variables used in the analysis

Variable	Description
income	Average monthly monetary income per household member, in roubles
health	Self-assessment of health status, where 1 = very poor, 5 = very good
education	Educational attainment level, where 1 = no basic general education, 8 = highest qualification (postgraduate studies, doctoral studies)
age	Respondent’s age
territory	Respondent’s place of residence, where 1 = city, 0 = village
family	Household composition, number of people
have_ch	Identifier for individuals with children aged 0–14, where 0 = no children of that age, 1 = children of that age are present
work	Number of hours spent in paid employment
household	Number of hours spent on housework
unpaid	Number of hours spent on activities for own final use
child	Number of hours spent on childcare
free	Number of hours spent on leisure activities
basic_needs	Number of hours spent on personal hygiene and self-care
sleep	Number of hours spent sleeping

Measuring wellbeing

In this study, wellbeing is treated as a characteristic that cannot be observed directly but is inferred from a limited, yet theoretically grounded, set of observable indicators. Following the author’s previous work, health status, educational attainment, and income are used as indicators of wellbeing (Yadrishchenskaya 2025).

In the literature, each of these measures is traditionally regarded as a core dimension of wellbeing (Stiglitz et al. 2009; OECD 2013; World Bank, 1980). Income shapes access to material resources (Livingston et al. 2022), while health determines individuals’ capacity to participate fully in work and social life (Livingston et al. 2022; Kahneman, Deaton 2010). Education is associated with the development of innate talents, a clearer understanding of one’s life circumstances, and subjective perceptions of wellbeing and overall life satisfaction (Nikolaev 2018; Ruggeri et al. 2020).

Although these indicators do not capture the full breadth of wellbeing, they represent fundamental and widely applicable dimensions that reflect its multidimensional nature. Accordingly, the following variables are used as indicators of the latent wellbeing variable:

educational attainment (1–8), where 1 denotes no basic general education and 8 denotes postgraduate education level (postgraduate or doctoral studies);
self‑rated health status (1–5), where 1 denotes very poor health and 5 denotes very good health;
average income per household member (expressed in units of the cost of a region-specific fixed basket of consumer goods and services).

Average income per household member is used as the indicator of material wellbeing due to data availability. Expressing this variable in units of the cost of a region-specific fixed basket of goods and services (2024) ensures comparability across the constituent entities of the Russian Federation.

Time‑use indicators

A distinctive feature of time‑use data is their compositional nature: the sum of all time‑use components is constrained by the 24‑hour day (1.440 minutes). An increase in time devoted to one activity necessarily reduces the time available for others, which rules out the independence of potential regressors (Aitchison 1982; Bacon‑Shone 2011; Greenacre 2018; Filzmoser et al. 2018).

This study examines seven broad activity categories:

paid employment,
activities for own final use,
housework,
childcare,
personal hygiene and self‑care,
leisure,
sleep.

A detailed description of the specific activities included in each category is provided in Table A1 in the Appendix. The grouping of activities follows Rosstat’s methodological guidelines and reflects the substantive similarity of the underlying tasks.

Time spent on travel, educational activities and other forms of employment not included in these categories is excluded from the analysis. The standard deviation of total reported time was 72 minutes for women and 77 minutes for men on working days, and 87 and 94 minutes respectively on days off. In this dataset, the sum of the selected activity categories does not equal 24 hours, meaning that the data constitute a ‘partial composition’ (van den Boogaart, Tolosana‑Delgado 2013). The analysis proceeds on the assumption that the selected categories represent the core structure of respondents’ daily time use.

To interpret the results in terms of the redistribution of time across activities, we apply the additive log‑ratio (ALR) transformation. This transformation enables the analysis of the relative structure of time allocation by comparing each component (activity category) with a chosen baseline (Aitchison 1982; van den Boogaart, Tolosana‑Delgado 2013). Sleep is used as the reference component, as it has the lowest proportion of zero values.

Following the transformation, each component is expressed as the logarithm of its ratio to the baseline component:

$alr (x_{i}) = In \frac{x_{i}}{x_{D}}, i = 1, \dots, D - 1$ ,

where x_i is the time (in hours) spent on the i-th activity, and x_D is the time (in hours) spent sleeping (the baseline activity).

The use of these transformed time‑use variables enables the findings to be interpreted in terms of the redistribution of time across different activities.

When applying this approach, particular attention must be paid to the presence of zeros in the data, as these make the transformation impossible (ln0 is undefined) (van den Boogaart, Tolosana‑Delgado 2013). In our case, zeros do not represent missing observations but indicate that the respondent did not engage in a given activity (for example, did not provide childcare). To ensure that the transformation could be applied, zeros were replaced using the multiplicative replacement method (Martín‑Fernández et al. 2003; 2011). This method replaces zeros with small positive values and adjusts the non‑zero components so that the original total is preserved. The procedure was implemented using the zCompositions package in R¹ (Palarea‑Albaladejo, Martín‑Fernández 2015).

Variables with more than 80 per cent zero values were excluded from the composition and entered into the model as separate regressors in their original units (hours). For example, for men on both working days and days off, the childcare variable was excluded; for both men and women on days off, the paid‑employment variable was excluded. To assess the robustness of the results, additional specifications using a lower threshold of 60 per cent zero values were estimated.

Descriptive statistics

The distribution of responses for each wellbeing indicator is presented in Figure 1. Both men and women most commonly rate their health as ‘satisfactory’ and report either vocational qualifications or incomplete higher education. In half of all the married couples, the average income per household member exceeds the cost of a fixed basket of goods and services by no more than a factor of two.

Figure 1.

Distribution of responses for the three wellbeing indicators – selfassessed health status, educational attainment, and average income per household member – relative to the cost of a fixed basket of goods and services. Source: author’s calculations based on data from the Sample Survey on the Use of Daily Time by the Population 2024.

Figure 2 shows the average structure of time spent on the main activity categories on weekdays and weekends. On average, men devote more time to paid work and leisure, independent of whether the day is a weekday or a weekend. Women, by contrast, spend more time on housework and childcare: on weekdays, their average time spent on these activities is more than three times that of men, and on weekends more than twice as much.

Figure 2.

Average structure of time allocation across activity categories among couples, weekdays and weekends (hours). Source: author’s calculations based on data from the Sample Survey of the Population’s Daily Time Use 2024.

Descriptive statistics for time spent on the main activities are provided in Table A2 in the Appendix.

Research methodology

To examine the nature of the association between unobserved wellbeing and time allocation, we estimate multiple‑indicator, multiple‑cause (MIMIC) models (Jöreskog, Goldberger 1975). This approach assumes the existence of a latent wellbeing variable that is reflected in a set of observable indicators (the measurement component) and explained by a set of causes (the structural component) (Zwierzchowski, Panek 2020). MIMIC models have previously been used to estimate subjective wellbeing and quality of life (Jewczak, Korczak 2023; Rahman et al. 2004).

The choice of this modelling strategy rests on the assumption that wellbeing is multidimensional. Unlike alternative approaches, the MIMIC framework does not require selecting a single measure of wellbeing or constructing a composite index in advance. Instead, the latent variable is estimated on the basis that each indicator represents an incomplete manifestation of the underlying construct. A key advantage of the model is the simultaneous estimation of parameters for both the measurement and structural components within a single framework, while accounting for measurement error. In sequential analytical approaches, measurement error is typically ignored at the regression stage.

Formally, the measurement component can be expressed as follows:

y = Λ wellbeing + ε,

where y is the vector of indicators of the unobserved wellbeing variable, Λ is the vector of factor loadings, wellbeing is the unobserved (latent) wellbeing variable, and ε is the vector of measurement errors.

$y (\begin{matrix} income \\ health \\ education \end{matrix}), Λ = (\begin{array}{l} λ_{1} \\ λ_{2} \\ λ_{3} \end{array}), ε = (\begin{array}{l} ε_{1} \\ ε_{2} \\ ε_{3} \end{array})$ .

To ensure comparability of scales, all indicators were first standardised using the scale function in R. For model identification, the unstandardised factor loading of the first indicator (income) was set at 1. Standardised factor loadings are used in the interpretation of the results, as they indicate the extent to which each indicator reflects variation in the latent wellbeing construct.

The structural component of the model includes the following causes:

basic sociodemographic characteristics (age, age squared, household size, place of residence and presence of children);
relative measures of time spent on the main activities, represented by ALR‑transformed time‑use variables for both the individual and their spouse;
specific activities outside the composition, namely paid employment on days off for men and women, and childcare for men on both working days and days off.

The structural part of the model is specified as follows:

wellbeing = Γ^TX + ζ,

where Γ is the vector of coefficients for the observed factors, Х is the vector of observed factors, and ζ is the error term of the structural part.

$X = (\begin{matrix} age \\ age \\ family \\ territory \\ have_ch \\ alr_w_work \\ alr_w_household \\ alr_w_unpaid \\ alr_w_child \\ alr_w_free \\ alr_w_basic needs \\ alr_h_work \\ alr_h_household \\ alr_h_unpaid \\ alr_h_free \\ alr_h_basic needs \\ h_child \end{matrix}), Γ = (\begin{array}{l} γ_{1} \\ γ_{2} \\ \dots \\ γ_{n} \end{array})$ .

The variables containing alr_w represent the transformed time‑use regressors for women, while those with alr_h denote the corresponding regressors for men. The variable h_child captures the time men spend on childcare in its original units (hours). It should be noted that the components of vector X may vary depending on the day of the week.

In general form, the multiple‑indicator, multiple‑cause model can be represented as the following system:

${\begin{cases} y = Λ wellbeing + ε \\ wellbeing = Γ^{T} X + ζ \end{cases}$ .

The model was estimated using maximum likelihood, implemented via the R lavaan package (Rosseel 2012; Chang et al. 2020). To test the hypothesis that each spouse’s wellbeing is associated with both their own time allocation and that of their partner, separate models were estimated for men and women. To identify differences in the nature of these associations depending on whether it is a weekday or a weekend, models were also estimated separately for weekdays and weekends.

It should be emphasised that, within this methodological framework, the results are interpreted solely in terms of association (presence and direction) whilst causal effects are not specified, owing to potential endogeneity.

Findings

The modelling results are presented in Table 2. Across all four specifications (1–4), the highest factor loading is observed for the self‑rated health indicator, indicating that variation in the latent wellbeing construct is most strongly reflected in this measure. The factor loadings for income and education are similar in magnitude but less prominent.

Table 2.

Download as

CSV

XLSX

Estimation results for Models (1–4)

	Women, working day	Women, day off	Men, working day	Men, day off
Indicators:	Factor loadings
Income	0.22	0.165	0.2	0.171
Health estimate	0.648	0.702	0.725	0.751
Education	0.332	0.321	0.322	0.316
Regressors:	Coefficient estimates (Standard errors)
Age of women (men)	-0.005***	-0.001	-0.007***	-0.003***
Age of women (men)	(0.001)	(0.001)	(0.001)	(0.001)
Woman’s (man’s) age^2	0.000***	0.000***	0.000	0.000***
Woman’s (man’s) age^2	(0.000)	(0.000)	(0.000)	(0.000)
Household composition	-0.003	0.002	0.004	0.005**
Household composition	(0.003)	(0.002)	(0.003)	(0.002)
Place of residence	0.053***	0.036***	0.034***	0.027***
Place of residence	(0.005)	(0.004)	(0.004)	(0.004)
Presence of children	0.005	0.006	0.006	0.003
Presence of children	(0.008)	(0.006)	(0.007)	(0.006)
Proportion of time spent in paid employment, women (at weekends – excluded from the analysis)	0.014***	-0.003**	-0.005***	-0.007***
	(0.002)	(0.001)	(0.001)	(0.001)
Relative time spent on housework, women	-0.011***	-0.008***	-0.01***	-0.017***
Relative time spent on housework, women	(0.003)	(0.002)	(0.003)	(0.002)
Time spent on activities for own final use, women	-0.008***	-0.003**	-0.014***	-0.009***
Time spent on activities for own final use, women	(0.002)	(0.001)	(0.002)	(0.001)
Relative time spent on childcare, women	-0.005*	-0.005**	-0.01***	-0.007***
Relative time spent on childcare, women	(0.003)	(0.002)	(0.002)	(0.002)
Relative time spent on leisure, women	0.002	-0.002	0.004*	0.002
Relative time spent on leisure, women	(0.003)	(0.002)	(0.002)	(0.002)
Relative time spent on personal hygiene and care, women	0.01	0.013**	0.011**	0.011**
Relative time spent on personal hygiene and care, women	(0.006)	(0.005)	(0.005)	(0.005)
Relative time spent on paid employment, men (at weekends – excluded from the analysis)	0.01**	-0.004***	0.025***	0.001
	(0.002)	(0.001)	(0.002)	(0.001)
Relative time spent on housework, men	0.000	-0.001	0.005***	0.005***
Relative time spent on housework, men	(0.002)	(0.001)	(0.002)	(0.001)
Relative time spent on activities for own final use, men	-0.007***	-0.004***	0.004**	0.003**
Relative time spent on activities for own final use, men	(0.002)	(0.001)	(0.002)	(0.001)
Relative time spent on leisure, men	0.003	0	0.005**	0.002
Relative time spent on leisure, men	(0.003)	(0.003)	(0.002)	(0.002)
Relative time spent on personal hygiene and care, men	-0.002	-0.012**	-0.013**	-0.011**
Relative time spent on personal hygiene and care, men	(0.006)	(0.005)	(0.005)	(0.005)
Childcare, men (not included in the model)	-0.002	-0.003	0.002	0.001
Childcare, men (not included in the model)	(0.005)	(0.002)	(0.004)	(0.002)
Number of observations	14.682	14.812	14.682	14.812
CFI	0.975	0.977	0.976	0.979
TLI	0.96	0.964	0.962	0.966
RMSEA	0.1	0.092	0.1	0.092
SRMR	0.048	0.043	0.047	0.043

All estimated models exhibit relatively high values of the root mean square error of approximation (RMSEA), exceeding the commonly recommended threshold of 0.08. However, high Comparative Fit Index (CFI) and Tucker–Lewis Index (TLI) values, together with low Standardised Root Mean Square Residual (SRMR) values, indicate that the models fit the data well overall.

Because the time‑use variables were transformed using the ALR method, the interpretation of the coefficients for these regressors is as follows: a positive coefficient indicates that an increase in the ratio of time spent on the activity in question to time spent sleeping is associated with a higher level of latent wellbeing; a negative coefficient indicates an association with lower wellbeing.

For women on working days, an increase in the ratio of time spent in paid employment relative to sleep is associated with higher wellbeing, whereas increases in the ratios for housework and activities for own final use are associated with lower wellbeing. Viewed in this light, the ‘double burden’ borne by women may have a negative impact on their wellbeing.

A husband’s redistribution of time towards paid employment is positively associated with his wife’s wellbeing on working days, while a redistribution towards activities for his own final use is negatively associated with it.

On days off, a shift in time allocation towards any form of activity – whether paid work or housework – is associated with lower wellbeing for women. A positive association is observed only for ‘personal hygiene and self‑care,’ which includes activities related to restoring energy and health. A husband’s engagement in activities on his day off (whether paid or unpaid for his own final use) is likewise negatively associated with his wife’s wellbeing.

For men, the pattern of associations differs markedly. Time allocated to paid employment on working days is positively associated with the latent wellbeing variable, which may reflect aspirations for professional fulfilment and the material wellbeing of the household. A positive association – albeit weaker than paid employment – is also observed for activities such as housework and work for own final use. This may be interpreted as a beneficial effect of participating in the shared management of household affairs.

Conversely, a greater time burden on wives is negatively associated with men’s wellbeing: the more time a wife allocates to paid work and domestic duties (relative to sleep), the lower, on average, her husband’s wellbeing. In other words, a wife’s overwork – both in the labour market and within the household – is associated with reduced wellbeing for her spouse. By contrast, a relative increase in the time a wife devotes to self‑care is positively associated with her husband’s wellbeing.

On days off, the time men spend in paid employment becomes less salient, while their involvement in household chores and activities for their own personal use is positively associated with wellbeing. An increase in the time wives spend on paid work and housework on days off continues to be associated, on average, with lower wellbeing among husbands.

Taken together, across all four specifications, the results indicate that on working days a wife’s wellbeing is associated with both her own labour‑market participation and her husband’s employment. At the same time, a husband’s wellbeing is, on average, lower when his wife carries a heavy workload – whether in paid employment or in housework. At weekends, any redistribution of women’s time towards labour activities (relative to sleep) is negatively associated with both their own wellbeing and that of their husbands. By contrast, women’s engagement in self‑care activities is positively associated with the wellbeing of both partners.

Regarding basic sociodemographic characteristics, household size is not significant for latent wellbeing in most specifications. Living in an urban area is positively associated with the latent variable, which may reflect the benefits of better developed infrastructure. The presence of children aged 0–14 is not significant in any of the estimated models.

Conclusions

This study applied a methodological framework to estimate the association between latent wellbeing and time allocation within married couples. Drawing on ALR‑transformed time‑use data from the 2024 Sample Survey of Daily Time Use by the Population enabled assessing how the relative allocation of spouses’ time is associated with individual wellbeing.

The MIMIC models revealed consistent associations between the composite wellbeing measure and spouses’ time‑allocation decisions, with particular attention to gender differences in how time devoted to specific activities relates to wellbeing.

For women, wellbeing on working days is positively associated with a relative increase in time spent in paid employment, suggesting the importance of self‑fulfilment through work. Greater involvement in housework and activities for own final use, however, is associated with lower wellbeing, indicating that the ‘double burden’ may have a detrimental effect. On weekends, higher wellbeing among women is associated with refraining from both paid and unpaid work in favour of self‑care and recuperation.

For men, wellbeing on working days is more strongly associated with paid employment, consistent with the traditional breadwinner role. However, a wife’s heavy workload – both in the labour market and in the household, relative to her recovery time – is negatively associated with her husband’s wellbeing. Notably, on days off, men’s participation in household duties is positively associated with his individual wellbeing, whereas high levels of wives’ involvement are negatively associated with it. Overall, women’s wellbeing appears to depend on achieving a balance between work and recovery, while men’s wellbeing is linked to the extent of their wives’ workload.

The proposed conceptualisation of latent wellbeing brings together fundamental, widely applicable and accessible indicators, avoiding the reduction of wellbeing to a single measure. Nonetheless, it does not capture the full diversity of wellbeing outcomes, such as subjective happiness, life satisfaction, or satisfaction with specific life domains.

Although the use of transformed time variables addressed the compositional nature of the data, further comparison with models based on original (absolute) indicators, as well as analyses of associations between wellbeing and participation in specific activities, would be valuable for testing the robustness of the findings. Future research should also examine more closely the role of children in the household. In this study, the presence of children was not significant in any specification, which may reflect variation in associations depending on the child’s age. In the original dataset, more than 80 per cent of men reported zero time spent on childcare. Identifying subgroups of families with non‑zero childcare time would allow for a more nuanced interpretation.

At this stage, the results suggest that the balance of time allocation is an important factor in the wellbeing of each spouse. The proposed approach offers a basis for further investigation of the multidimensional construct of wellbeing, taking into account patterns of time allocation across different household types.

References

Aitchison J (1982) The statistical analysis of compositional data. Journal of the Royal Statistical Society: Series B (Methodological): 44(2): 139–60. https://doi.org/10.1111/j.2517-6161.1982.tb01195.x

Álvarez B, Miles-Touya D (2019) Gender imbalance in housework allocation: a question of time? Review of Economics of the Household: 17: 1257–87. https://doi.org/10.1007/s11150-019-09467-w

Bacon-Shone J (2011) A short history of compositional data analysis / In: Pawlowsky-Glahn V, Buccianti A (eds) Compositional Data Analysis. https://doi.org/10.1002/9781119976462.ch1

Chang C, Gardiner J, Houang R, Yan-Liang Y (2020) Comparing multiple statistical software for multiple-indicator, multiple-cause modeling: an application of gender disparity in adult cognitive functioning using MIDUS II dataset. BMC Medical Research Methodology: 20: 275. https://doi.org/10.1186/s12874-020-01150-4

Filzmoser P, Hron K, Templ M (2018) Applied Compositional Data Analysis. Springer, Cham. https://doi.org/10.1007/978-3-319-96422-5

Flood SM, Genadek KR (2016) Time for each other: Work and family constraints among couples. Journal of Marriage and Family: 78(1): 142–64. https://doi.org/10.1111/jomf.12255

Fursman L (2009) Parent’s long work hours and the impact on family life. Social Policy Journal of New Zealand: (35): 55-67. URL: https://www.msd.govt.nz/about-msd-and-our-work/publications-resources/journals-and-magazines/social-policy-journal/spj35/35-parents-long-work-hours.html

García-Manglano J, Nollenberger N, Sevilla A (2015) Gender, time-use, and fertility recovery in industrialized countries. In: International Encyclopedia of the Social & Behavioral Sciences (Second Edition). Elsevier, 775-80. https://doi.org/10.1016/B978-0-08-097086-8.31104-7

Greenacre M (2018) Compositional Data Analysis in Practice (1st ed.). Chapman and Hall/CRC. https://doi.org/10.1201/9780429455537

Helms HM, Walls JK, Crouter AC, McHale SM (2010) Provider role attitudes, marital satisfaction, role overload, and housework: A dyadic approach. Journal of Family Psychology: 24(5): 568–77. https://doi.org/10.1037/a0020637

Hiekel N, Kühn M (2024) Lessons from the pandemic: Gender inequality in childcare and the emergence of a gender mental health gap among parents in Germany. Demographic Research: 51: 49-80. https://doi.org/10.4054/DemRes.2024.51.3

Hoang T, Knabe A (2021) Time use, unemployment, and well-being: An empirical analysis using British Time-Use Data. Journal of Happiness Studies: 22: 2525–48. https://doi.org/10.1007/s10902-020-00320-x

Jewczak M, Korczak K (2023) Measuring welfare with the use of the MIMIC approach: evidence for Poland. GeoJournal: 88: 5109–27. https://doi.org/10.1007/s10708-023-10910-w

Jöreskog KG, Goldberger AS (1975) Estimation of a Model with Multiple Indicators and Multiple Causes of a Single Latent Variable. Journal of the American Statistical Association: 70(351a): 631–639. https://doi.org/10.1080/01621459.1975.10482485

Kahneman D, Deaton A (2010) High income improves evaluation of life but not emotional well-being. Proceedings of the National Academy of Sciences: 107(38): 16489–93. https://doi.org/10.1073/pnas.1011492107

Kan M-Y, Hertog E, Kolpashnikova K (2019) Housework share and fertility preference in four East Asian countries in 2006 and 2012. Demographic Research: 41: 1021–46. https://doi.org/10.4054/DemRes.2019.41.35

Khawaja M, Habib RR (2007) Husbands’ involvement in housework and women’s psychosocial health: findings from a population-based study in Lebanon. American Journal of Public Health: 97: 860–6. https://doi.org/10.2105/AJPH.2005.080374

Leocádio V, Verona AP, Wajnman S (2024) Exploring the association between gender equality in the family and fertility intentions: an explanation of the findings in low-fertility countries. Genus: 80: 26. https://doi.org/10.1186/s41118-024-00234-z

Livingston V, Jackson-Nevels B, Reddy VV (2022) Social, cultural, and economic determinants of well-being. Encyclopedia 2(3): 1183–99. https://doi.org/10.3390/encyclopedia2030079

Martín-Fernández JA, Barceló-Vidal C, Pawlowsky-Glahn V (2003) Dealing with zeros and missing values in compositional data sets using nonparametric imputation. Mathematical Geology: 35: 253–78. https://doi.org/10.1023/A:1023866030544

Martín-Fernández JA, Palarea-Albaladejo J, Olea RA (2011) Dealing with zeros. In: Pawlowsky-Glahn V, Buccianti A (eds) Compositional Data Analysis. https://doi.org/10.1002/9781119976462.ch4

McCann JK, Freire S, de Oliveira CVR et al. (2024) Father involvement is a protective factor for maternal mental health in Western Kenya. SSM – Mental Health: 5: 100318. https://doi.org/10.1016/j.ssmmh.2024.100318

Nikolaev B (2018) Does higher education increase hedonic and eudaimonic happiness? Journal of Happiness Studies: 19: 483–504. https://doi.org/10.1007/s10902-016-9833-y

Palarea-Albaladejo J, Martín-Fernández JA (2015) zCompositions – R package for multivariate imputation of left-censored data under a compositional approach. Chemometrics and Intelligent Laboratory Systems: 143: 85–96. https://doi.org/10.1016/j.chemolab.2015.02.019

Rahman T, Mittelhammer RC, Wandschneider P (2004) A latent variable MIMIC approach to inferring the quality of life / Proceedings of the 2004 Annual Meeting of the American Agricultural Economics Association. Denver, CO, August 1-4. URL: https://ageconsearch.umn.edu/record/20351/files/sp04ra01.pdf

Rosseel Y (2012) lavaan: An R package for structural equation modeling. Journal of Statistical Software: 48(2): 1–36. https://doi.org/10.18637/jss.v048.i02

Ruggeri K, Garcia-Garzon E, Maguire А et al. (2020) Well-being is more than happiness and life satisfaction: A multidimensional analysis of 21 countries. Health and Quality of Life Outcomes: 18: 192. https://doi.org/10.1186/s12955-020-01423-y

Shafer EF, Kelly EL, Buxton OM, Berkman LF (2018) Partners’ overwork and individuals’ wellbeing and experienced relationship quality. Community, Work & Family: 21(4): 410–28. https://doi.org/10.1080/13668803.2017.1311839

van den Boogaart KG, Tolosana-Delgado R (2013) Analyzing Compositional Data with R. Springer Berlin, Heidelberg. https://doi.org/10.1007/978-3-642-36809-7

Voßemer J, Baranowska-Rataj A, Heyne S, Loter K (2024) Partner’s unemployment and subjective well-being: The mediating role of relationship functioning. Advances in Life Course Research: 60: 100606. https://doi.org/10.1016/j.alcr.2024.100606

Yadrishchenskaya KYu (2025) Well-being of Russia’s population based on daily time allocation data. Lomonosov Economics Journal: 60(6): 329–66. (In Russ.) https://doi.org/10.55959/MSU0130-0105-6-60-6-14

Zwierzchowski J, Panek T (2020) Measurement of subjective well-being under capability approach in Poland. Polish Sociological Review: 210(2): 157–78. https://doi.org/10.26412/psr210.02

Other sources of information

OECD (2013) OECD Guidelines on Measuring Subjective Well-being. OECD Publishing, Paris. https://doi.org/10.1787/9789264191655-en

Stiglitz JE, Sen A, Fitoussi J-P (2009) Report by the Commission on the Measurement of Economic Performance and Social Progress. Paris. URL: https://ec.europa.eu/eurostat/documents/8131721/8131772/Stiglitz-Sen-Fitoussi-Commission-report.pdf

World Bank (1980) World Development Report, 1980. The World Bank, Washington, D.C. URL: https://documents1.worldbank.org/curated/en/430051469672162445/pdf/108800REPLACEMENT0WDR01980.pdf

Information about the author

Ksenia Yuryevna Yadrishchenskaya – postgraduate student at the Faculty of Economics of Lomonosov Moscow State University, Moscow, 119991, Russia. Email: bashkirova_ks@mail.ru.

Appendix

Table A1.

Download as

CSV

XLSX

Description of broad categories of activities

Category of activities	Code	Description
Sleep	K_101	Sleep
Personal hygiene and care	K_201	Food and drink consumption
	K_202	Personal hygiene and self-care
	K_203	Self-management of one’s own health/self-care
	K_204	Receiving personal care and grooming services from others
	K_205	Receiving health/medical care services from others
	K_206	Other/unspecified types of personal care, health care, and medical care
Paid employment	K_301	Time spent at work (including short breaks and work-related travel)
	K_303	Other/unspecified employment-related activities
	K_304	Job-related training and learning
	K_305	Occasional or one-off paid work
Housework	K_501	Food preparation and cooking
	K_503	Cleaning up after meals
	K_502	Food supply storage and organisation
	K_504	Home cleaning
	K_604	Laundry
	K_605	Ironing
	K_601	Home water and heating management
	K_602	Tidying household items and materials
	K_505	Caring for houseplants and tidying the garden areas
	K_603	Sorting and taking out rubbish
	K_704	Walking and caring for pets
Housework	K_606	Other or unspecified care for textiles and footwear
	K_901	Shopping (including online)
	K_902	Seeking and obtaining commercial or administrative services
	K_903	Running and managing the household
Activities for own final use	K_701	Gardening and horticulture
	K_702	Caring for livestock, poultry, and other domestic animals
	K_703	Hunting, fishing, gathering wild foods, forestry activities
	K_705	Other or unspecified types of gardening, forestry, and animal-care activities
	K_801	House construction and major renovation
	K_802	Routine home maintenance
	K_803	Installation, maintenance and repair of household goods and office equipment
	K_804	Vehicle maintenance and minor repairs
	K_805	Producing and processing goods for own final use
Childcare	K_1001	Childcare (feeding, washing, physical and medical care)
	K_1002	The child’s education
	K_1003	Reading, playing, and talking with the child
	K_1004	Accompanying the child
	K_1005	Other childcare (passive care)
	K_1204	Caring for one’s own children living in another household
Leisure	K_1305	Socialising with other household members
	K_1306	Face-to-face socialising with relatives, friends, and acquaintances
	K_1307	Communication via audio and video systems
	K_1308	Text messaging and online communication
	K_1309	Celebrating special occasions together
	K_1303	Attending funerals and memorial events
	K_1310	Taking part in public cultural or historical celebrations
	K_1311	Other social interaction
	K_1202	Construction and repair work to assist other households
	K_1203	Assisting other households with paid work and farming
	K_1205	Assisting another household with childcare
	K_1206	Assisting an adult member of another household
	K_1207	Other friendly help for another household
	K_1101	Providing physical care for an adult household member
	K_1102	Providing passive care for adult household members, other support for an adult household member
Leisure	K_1201	Volunteering and organisational work for the community
	K_1301	Taking part in public meetings
	K_1304	Taking part in civic and related duties
	K_1302	Taking part in religious rites
	K_1602	Programming and other productive nonemployment computer use
	K_1401	Attending or visiting cultural, entertainment and sporting venues
	K_1503	Taking part in sport
	K_1504	Doing physical exercises
	K_1505	Other sports-related activities
	K_1501	Leisure, reflection, and relaxation
	K_1502	Casual walking without intent
	K_1601	Hobby activities
	K_1701	Playing non-digital games
	K_1603	Searching for interesting information online
	K_1702	Playing computer games
	K_1703	Playing mobile games
	K_1704	Playing home console games
	K_1801	Reading periodicals. Reading for leisure.
	K_1802	Reading books
	K_1803	Watching television and videos
	K_1804	Listening to radio and audio content

Table A2.

Download as

CSV

XLSX

Descriptive statistics for indicators of time spent on the main activity categories

		Sleep	Personal hygiene and care	Paid employment	Housework	Activities for own final use	Childcare	Leisure
Men, working day	Average	8.37	2.97	5.37	0.88	0.99	0.17	3.39
	Median	8.33	2.83	7.50	0.50	0.00	0.00	2.83
	Standard deviation	1.56	1.10	4.16	1.16	1.84	0.52	2.46
	Proportion of zeros	0.00%	0.00%	33.96%	30.98%	62.28%	83.59%	2.97%
Women, working day	Average	8.32	2.91	4.04	3.34	0.55	0.62	2.58
	Median	8.17	2.83	5.00	2.83	0.00	0.00	2.17
	Standard deviation	1.41	1.01	3.95	2.16	1.23	1.47	1.96
	Proportion of zeros	0.00%	0.00%	45.76%	1.55%	74.32%	71.93%	5.55%
Men, day off	Average	9.41	3.25	0.54	1.50	1.98	0.33	5.12
	Median	9.33	3.17	0.00	1.17	1.00	0.00	5.00
	Standard deviation	1.69	1.13	2.09	1.43	2.36	0.93	2.56
	Proportion of zeros	0.00%	0.00%	91.69%	18.03%	40.58%	81.55%	1.06%
Women, day off	Average	9.26	3.18	0.25	4.42	0.98	0.56	3.75
	Median	9.17	3.00	0.00	4.33	0.00	0.00	3.50
	Standard deviation	1.51	1.10	1.40	2.08	1.60	1.32	2.13
	Proportion of zeros	0.00%	0.00%	95.54%	0.96%	60.81%	73.79%	2.15%

1 Multiplicative replacement of count zeros: https://search.r-project.org/CRAN/refmans/zCompositions/html/cmultRepl.html.