The recession shadowing the COVID-19 pandemic has been frequently characterised as a “shecession,” implying disproportionately negative effects for women. Yet the crisis might more accurately be called a “momcession,” as women’s work losses were driven in large part by the outcomes of mothers specifically. The OECD’s 2020 Risks that Matter survey presents cross-national evidence that when schools and childcare facilities shut down, mothers took on the brunt of additional unpaid care work – and, correspondingly, they experienced labour market penalties and stress. These findings serve as another reminder that governments must consider inequalities in unpaid work and take a gender-sensitive approach when building their policy responses to the COVID-19 crisis.
Caregiving in Crisis: Gender inequality in paid and unpaid work during COVID-19
Abstract
Key findings
The recession shadowing the COVID‑19 pandemic has been frequently and simplistically labelled a “shecession,” implying disproportionately negative effects for women. Across countries, however, one group stands out as faring especially poorly in labour force and unpaid work outcomes: working mothers with school-age or younger children.
International comparisons of the effects of the recession on mothers have been limited thus far due to lags in the cross-national availability of detailed labour force and time use data by parenthood status. The OECD Risks that Matter (RTM) 2020 survey helps to fill a gap in our understanding by combining self-reported employment and caregiving microdata, disaggregated by parenthood status, across 25 OECD countries.
COVID‑19 has laid bare the negative consequences of longstanding gender gaps and norms around caregiving. RTM 2020 reveals that when schools and childcare facilities closed, mothers took on the brunt of the additional unpaid care work – and, correspondingly, they experienced labour market penalties and stress.
Mothers were nearly three times as likely as fathers to report that they took on the majority or all of additional unpaid care work related to school or childcare facility closures: 61.5% of mothers of children under age 12 say they took on the majority or entirety of the extra care work, while 22.4% of fathers report that they did.
Mothers of children under age 12 were the group most likely to move from employed to not employed status between Q4 2019 and Q3 2020, on average across OECD countries.
Gender gaps in a household’s unpaid care were largest, on average, when the father continued to be employed while the mother was not. This relationship was not reciprocated to the same degree in households where the father was out of paid work and the mother was in paid work. Indeed, consistent with existing literature, RTM 2020 data show that mothers’ participation in paid employment did little to mitigate inequality in unpaid work conditions.
Public supports may have helped to lessen gender inequality at home. Additional days of school closures correlated cross-nationally with greater gender inequality in the take‑up of additional unpaid work, while historically higher levels of spending on family supports were associated with smaller gender gaps in the distribution of additional care of children during COVID‑19.
The social policy response to the crisis should include greater public investments in good-quality childcare, education, and out-of-school supports; increasing the length, and incentivising the take‑up, of fathers’ parental leave, which has implications for fathers’ long-term caregiving behaviour; and promoting equal access to telework for women and men, addressing possible penalties in terms of pay and advancement for those using it. Policymakers must embed all of these strategies in a whole‑of-government approach to gender mainstreaming.
A “Shecession,” or a “Momcession”?
The recession shadowing the COVID‑19 pandemic has been frequently labelled a “shecession,” implying disproportionately negative economic effects for women, relative to men. Gender gaps vary across countries, but, on average, cross-national OECD and global estimates indicate that women’s hours worked and women’s employment rates declined at a greater rate than men’s in the early months of the pandemic (Alon et al., 2021[1]; OECD, 2021[2]; Bluedorn et al., 2021[3]). This stands in contrast to what have typically been larger job losses among men in recent recessions (OECD, 2012[4]).
These initial declines were driven by demand- and supply-side shocks on labour. Women disproportionately work in sectors that were hit hardest by the early lockdowns (such as retail and hospitality1). Women are also more likely to work in part-time and irregular jobs, which were more likely to be terminated or furloughed in the early months of COVID‑19 (Bluedorn et al., 2021[3]; Adams-Prassl et al., 2020[5]). Finally, restrictions on labour supply resulting from increased unpaid care responsibilities (e.g. the need to care for children during school closures) were disproportionately borne by women.
After the initial shock, women’s work hours rebounded somewhat in the following quarters and were stabilised throughout the pandemic by women’s continued participation in essential sectors like health care and teaching. Labour force participation and employment rates between the fourth quarters of 2019 and 2020, measured in labour force surveys, show that job losses were spread across men and women, though effects were often larger for women (Figure 1).
Yet many female‑dominated jobs still have not returned in the way that male‑dominated ones have, for example in construction and manufacturing in some countries. And the “pull” factors keeping some women out of the labour market – namely irregular school openings and childcare availability – are still ongoing. Whether (and to what degree) these gender gaps in paid and unpaid work will persist will depend on the nature of the ongoing pandemic and on governments’ responses.
Country-specific studies have helped to illustrate the cause and nature of women’s labour market losses. School and childcare closures were a causal driver of parents’ – and especially women’s – reduced paid work hours. These closures forced parents, and especially women,2 to take on additional unpaid care work in countries as diverse as the United Kingdom (Andrew et al., 2020[6]; Sevilla and Smith, 2020[7]),3 Italy (Del Boca et al., 2020[8]), Spain (Farré et al., 2020[9]), and the United States (Zamarro and Prados, 2021[10]; Alon et al., 2021[1]).
Related to this, women’s work hour reductions in many countries were disproportionately borne by those workers who were unable to stay home and telework (Adams-Prassl et al., 2020[5]; Alon et al., 2021[1]). These were by and large the women who had to choose between continuing to work or caring for their children at home. While the hours worked by women decreased cross-nationally, countries with well-designed care leave systems and/or job retention schemes were often better able to stave off more dramatic drops in women’s work (OECD, 2021[2]; Adams-Prassl et al., 2020[5]).
This research, taken together, suggests that the COVID‑19 “shecession” should more accurately be called a “momcession.” Women’s work losses were driven in large part by the outcomes of mothers, specifically, who often took on additional hours of (unpaid) care of their children during school shutdowns. Yet cross-national comparisons of the effects of the recession on mothers have thus far been limited due to lags in the cross-national availability of detailed labour force microdata by parenthood status.
Figure 1. In many countries women experienced large declines in labour force participation and employment rates between Q4 2019 and Q1 2021
Percentage point changes in labour force participation rates (Panel A) and employment rates (Panel B), disaggregated by gender, 2019 and 2021
Note: Data for Iceland, Switzerland and Turkey refer to the difference in Employment Rate between Q4 2019 and Q4 2020. No data for Chile and Costa Rica. Since January 2021, European countries have introduced a renovated LFS survey questionnaire, with a change in the sampling and structure of the questionnaire. Major changes concern the treatment of full (survey) week work absences where persons on full-week sickness absences irrespective of the duration of absences are employed, those on parental leave receiving job-related income/benefits beyond 3 months with a guarantee to return to the previous job are employed, but lay-offs due to business conditions receiving job-related income or benefits beyond 3 months are no longer employed. This may have increased or decreased the employment level and rates in the countries concerned.
Source: OECD Short-term Labour Market Statistics (https://stats.oecd.org/Index.aspx?DataSetCode=STLABOUR).
The OECD Risks that Matter (RTM)4 2020 data help fill a gap in our understanding of these inequalities by combining self-reported employment5 and caregiving microdata, disaggregated by parenthood status, across 25 OECD countries.
RTM 2020 allows a comparison of men versus women’s self-reported employment status between Q4 2019 and Q3 2020,6 disaggregated by parenthood status (see Figure 2). Mothers of children under 12 years old (hereafter “under‑12s”) were the group most likely to move from employment to out of employment relative to women without children, women with children age 12 and older, men without children, men with children age 12 and older, and men with under‑12s.
When comparing women and men with comparable parenthood status, the gender gap in employment was greatest between mothers and fathers of under‑12s. On average across countries, mothers of under‑12s were over 3 percentage points more likely to have left employment than fathers of under‑12s at some point between the first and the third quarter of 2020 (Figure 2). This represents the early months of the pandemic, and the period with the most intensive public lockdowns. In Germany, for example, mothers of under‑12s were about 6 percentage points more likely to have left employment than fathers of under‑12s (Figure 2, indicated by the blue bar), whereas women who are not mothers or are mothers of children 12 and over had little difference in employment outcomes relative to their male counterparts (indicated by the diamond).
Beyond outright job losses, other job disruptions took their toll on families, too. It is not possible to assess with RTM data to what extent women versus men experienced job disruptions such as layoffs or placement on a job retention scheme in 2020, as RTM 2020 asked about job disruption in the household of the respondent, not the respondent individually7. These job disruption results therefore cannot be disaggregated by gender.
But given widespread school closures during the pandemic, it is unsurprising that about half of respondents with children under age 18 experienced some kind of job disruption in the household due to COVID‑19. Gaps emerge in particular when looking at work hour reductions and taking leave from work. Across the sample, about 15.2% of respondents with children under 18 report a member of their household having their work hours reduced or being put on a part-time job scheme, compared to just 12.3% of other respondents, on average (OECD, 2021[11]).
Figure 2. The gender gap in transitioning from employed to not employed is typically largest among parents of school age and younger children
Gender gaps as percentage point change resulting from gender differences in the share of respondents that were employed and in paid work at some point during Q4 2019, but were not in paid employed during Q3 2020 (% males – percentage females) by parenthood status.
Note: Data refer to adults of prime working age (25 to 54 years old). “Parents with children under 12” are defined as individuals whose youngest child living in the household is under 12 years old. They may or may not have additional children age 12 or older.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Additionally, many parents took leave from work. 15.6% of respondents with children under 18 had at least one member of the household take paid or unpaid leave from work, compared to just 10.8% of respondents without children. The rate of paid or unpaid leave‑taking from work increases further, to 17%, when looking only at parents with younger children under age 12 (OECD, 2021[11]).
It is worth noting, too, that in the households where no one took leave, someone may have already been available to care for children due to a different type of job disruption. Across countries, on average, 41% of households with children that did not take (paid or unpaid) leave reported some other form of job-related disruption, such as job loss, a temporary lay-off, work hours reduction, and/or placement on a job retention scheme (OECD, 2021[11]).
Mothers took on more unpaid work than others
Aside from labour market factors, one of the biggest drivers of women’s reduced employment, labour force participation and reduced working hours during COVID‑19 is the fact that mothers typically took on more additional unpaid childcare obligations than fathers and non-parents. This is the case in most countries that have been studied so far (Andrew et al., 2020[6]; Del Boca et al., 2020[8]; Farré et al., 2020[9]; Sevilla and Smith, 2020[7]; Zamarro and Prados, 2021[10]; Eurofound, 2020[12]).8
RTM 2020 provides new, cross-national data to assess this link. We find that women report taking on more of the additional unpaid childcare work during COVID‑19 than fathers, and we find that the burden of unpaid work is associated with a decreased probability of being in paid employment by Q3 2020.
Mothers were more likely than fathers to take on additional unpaid care work
In the absence of widespread, cross-nationally harmonised time use statistics during COVID‑19,9 RTM 2020 offers evidence across OECD countries that women took on more unpaid care obligations than men when schools and childcare facilities shut down. Respondents whose children were affected by school or childcare facility closures were asked, “In your household, who took on any additional care work as a result of school or childcare facility closures [during COVID‑19]?”
On average, mothers of children under age 12 were nearly three times as likely as fathers to say that they took on all or the majority of additional unpaid care work related to school and/or childcare facility closures (Figure 3). These gender gaps are fairly consistent across ages of minor children (see Annex Table 1.A.1).
61.5% of mothers of under‑12s report that they took on most or all of that additional unpaid care work, compared to just 22.4% of fathers reporting that they took on most or all of the additional unpaid care work – a gap of 39.1 percentage points. The gender gap is smallest in the Netherlands, but even there, the share of mothers reporting taking on the majority or entirety of additional care work is 15.9 percentage points higher than the rate for fathers.
It is important to note that these results present perceptions of who took on more work. Unlike traditional time use surveys, RTM 2020 does not monitor hours spent on specific paid and unpaid activities.
Yet fathers, too, corroborate that their partner took on more of the additional care work than they did – albeit to a lesser degree. While 22.4% of fathers self-report taking on the majority or entirety of additional unpaid care work, 25.9% of fathers report that their partner took on the majority of entirely of additional unpaid care work.
The most common answer for fathers of under‑12s is that the additional unpaid care work was split evenly between them and their partner: 40.8% of fathers claim this, compared to 20.7% of mothers.
This inequality in unpaid work is negatively associated with women’s employment. When analysed in a linear regression, being a mother and carrying out the majority or the entirety of additional unpaid care work is associated with a significant 0.053 percentage point increase in the probability of transitioning from employed status in Q4 2019 to not employed status in Q3 2020.10 In other words, a high unpaid care work burden is highly correlated with moving out of paid work. Of course, causality likely moves in both directions: the additional available time for a respondent who left employment could drive them to take on more unpaid work, or an added unpaid work burden could result in respondent’s departure from paid employment.
Figure 3. Mothers were nearly three times as likely as fathers to say they took on most or all of the additional unpaid care work caused by school and childcare closures
Share of parents with at least one child under 12 who report that they took on all or the majority of the additional care work caused by school or childcare facility closures, by gender, 2020
Note: Respondents were asked who in their household took on any additional care work as a result of school or childcare facility closures. The response options were “Entirely you”, “Mostly you, “Equally shared between you and your spouse/partner”, “Mostly your spouse/partner”, “Entirely your spouse/partner”, “Mostly someone else (another member of your household or someone from outside your household)”, and “A mixture of you (and/or your spouse/partner) and someone else”. Respondents could also choose “Can’t choose / difficult to say” or “Not applicable” as options. This figure presents the aggregation of results from respondents who answered “Entirely you” or “Mostly you”. Parents with children under 12 are defined as respondents whose youngest child living in the household is under 12 years old.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Mothers’ participation in paid work did little to mitigate inequality in unpaid work
Academic research suggests a shift toward more paternal care. Even if mothers took on more of the additional unpaid care work, researchers and policy makers have suggested that COVID‑19 may mark the start of a change towards more egalitarian norms and behaviour. This convergence notion suggests that by getting to know and appreciate the effort that goes into family care work at home, fathers might permanently increase their share of such work (Boll, Müller and Schüller, 2021[13]).
However, this shift towards more paternal caregiving seems to be more pronounced when fathers are unemployed or furloughed rather than working from home, suggesting that the shift emerged out of a supply-side shock (fathers have more time), rather than the demand-side shock (higher need for childcare and housework) (Sevilla and Smith, 2020[7]). Fathers’ childcare provision is much more sensitive to employment status than that of women’s (Sevilla and Smith, 2020[7]; Zamarro and Prados, 2021[10]).
Importantly, fathers’ involvement in childcare seems to be highly sensitive to their partners’ working arrangements. The main driver of the shift toward more paternal care was the inability of working mothers to telework (Boll, Müller and Schüller, 2021[13]), further suggesting that fathers’ involvement came out of necessity (mothers cannot) rather than opportunity (fathers can) ( (Boll, Müller and Schüller, 2021[13]; Margaria, 2021[14]). Women working from home in the United Kingdom, for example, were found during COVID‑19 to have been doing as many additional hours of childcare as men on furlough (Sevilla and Smith, 2020[7]), and a study of time use data in the United States finds that moms working from home spend over 20 minutes more per day on childcare than telecommuting dads (Lyttelton, Zang and Musick, 2021[15]). This gap is even larger when looking at non-care housework (ibid.).
RTM 2020 data corroborate this general consensus that working mothers11 were not immune to the additional unpaid work burden. Working mothers with an employed partner are more than twice as likely as working men with an employed partner to say that they took on the majority or all of the additional unpaid care work (Figure 4). 53.4% of employed mothers, versus 20.9% of employed fathers, report taking on the majority of entirety of additional care work when schools and childcare facilities closed.
The gender gaps in caregiving are largest, by far, when a mother is not employed and the father is employed. 76.9% of not employed mothers (with employed partners) say that they took on most or all of the additional unpaid care work. When the situation is reversed – the father is not employed while the mother is employed – only 24.5% of fathers self-report taking on most or all of the additional unpaid care work.
Figure 4. The gender gap in unpaid caregiving persists even when both parents are employed
Share of employed parents with: 1) an employed partner and 2) at least one child under 18 who report that they took on all or the majority of the additional care work caused by school or childcare facility closures during COVID‑19, by gender, 2020
Note: Respondents were asked who in their household took on any additional care work as a result of school or childcare facility closures. The response options were “Entirely you”, “Mostly you, “Equally shared between you and your spouse/partner”, “Mostly your spouse/partner”, “Entirely your spouse/partner”, “Mostly someone else (another member of your household or someone from outside your household)”, and “A mixture of you (and/or your spouse/partner) and someone else”. Respondents could also choose “Can’t choose / difficult to say” or “Not applicable” as options. This figure presents the aggregated total from respondents who answered “Entirely you” or “Mostly you”. Parents here are defined as respondents with at least one child under the age of 18. Employment status is recorded in September 2020.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
This inequality in caregiving during COVID‑19 is unsurprising in light of longstanding gender norms and gender inequalities in unpaid work (Figure 5). Throughout OECD countries, and indeed throughout the world, women have historically spent far more time on unpaid work than men. Public policies will influence whether COVID‑19 deepens or mitigates these inequalities in the longer run.
Figure 5. Women have historically spent almost twice as much time in unpaid work than men throughout the OECD
Estimates of daily time spent on unpaid work, such as routine housework, shopping, care for household members, childcare, adult care, care for non-household members, volunteering, travel related to household activities, and other unpaid activities, in minutes per day, by gender, most recent year
Note: Figure presents time use estimates from 2019 in USA, from 2016 in Japan and the Netherlands, from 2015 in Canada, from 2014‑15 in Turkey and UK, from 2014 in Korea and Mexico, from 2013‑14 in Italy, from 2013 in Belgium, Greece, Luxembourg and Poland, from 2012‑13 in Germany, from 2010‑11 in Norway, from 2010 in Hungary and Sweden, from 2009‑10 in Estonia, Finland, France, New Zealand and Spain, from 2008‑09 in Austria, from 2006 in Australia, from 2005 in Ireland, from 2003 in Latvia and Lithuania, from 2001 in Denmark, from 2000‑01 in Slovenia, and from 1999 in Portugal.
Source: OECD Gender Data Portal, https://www.oecd.org/gender/data/.
Gender gaps in worries, distress, and perceptions
RTM 2020 supports existing evidence that gender gaps emerged or were heightened in worry and distress during COVID‑19.
Mothers experienced disproportionately high levels of stress
Country-specific research has found that mothers’ were disproportionately stressed during the pandemic. Zamarro and Prados (2021[7]) find that at the onset of COVID-19, large gaps emerged or persisted in indicators of psychological distress between mothers of school-aged or younger children and all other comparison groups,12 and that these gaps were still present four months into the pandemic in the United States. Xue and McMunn (2021[14]) find similar results for the United Kingdom and highlight the vulnerability of single mothers. Lyttelton et al. (2021) link emotional distress to women’s teleworking and unpaid work obligations in the United States. They find that mothers who teleworked during the pandemic were more likely to report feeling depressed, anxious, and lonely than telecommuting fathers – but they found no gender gaps in anxiety levels among parents who were commuting to workplaces during the crisis (Lyttelton, Zang and Musick, 2021[15]). In addition, qualitative research in sociology has found that mothers who greatly increased the time they spend caring for children disproportionately reported increased stress, anxiety, and frustration, especially when the shock was associated with (paid) work pressure and disrupted childcare arrangements (McCrory Calarco et al., 2020[16]).
Gender gaps in psychological distress are also found in Eurofound’s April and July 2020 opt-in, online surveys in European Union countries (Eurofound, 2020[12]).13 In the first months of the pandemic, women were more likely than men to report work-life conflict, and mothers of young children were more likely than men to report feeling tense, lonely, and/or depressed.
This gender gap in distress is perhaps unsurprising given that 2020 time use research (however limited) shows that women took on both more of the additional care work and a larger share of less gratifying unpaid work, such as laundry, cleaning, and physical childcare. Men, in contrast, spent more time on repairs and grocery shopping (Farré et al., 2020[9]; Del Boca et al., 2020[8]) – perhaps related to the fact that in some countries, leaving the house during confinement was only allowed for grocery shopping (Farré et al., 2020[9]). These mental health findings are situated in a broader context of women being more worried than men, cross-nationally, about household finances during COVID-19 (OECD, 2021[17]).
In RTM 2020, women are more likely than men to respond “yes” to “[Has] your (or at least one member of your household’s) mental health and well-being been affected by the pandemic and crisis?” Mothers are more likely than fathers to report the presence of mental health issues within the household (Figure 6), though the degree to which they are referring to themselves or to family members cannot be established with these data.
Figure 6. Mothers are more likely than fathers to report the pandemic has affected their own or their household’s mental health and well-being
Share of mothers versus fathers of under-12s reporting mental health or well-being challenges (to themselves or a family member) during the pandemic, 2020.
Note: Figure presents the share of mothers and fathers of children under age 12 answering affirmatively to the question “Has your (or at least one member of your household’s) mental health and well-being been affected by the pandemic and crisis?”
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Parents worry about accessing good-quality childcare and education
Looking forward to 2022, a majority of parents – 61.3% – are also very worried about accessing good-quality and affordable children or education for their children (Figure 7). The rate is over 80% of parents in Turkey, Mexico, Estonia, Greece and Chile. On average across countries, mothers are slightly more concerned than fathers: 63.8% of mothers of under‑12s, and 58.3% of fathers, say they that are somewhat or very concerned about being able to access good-quality childcare or education (Annex Table 1.A.2). In Spain, the United States, Mexico and Chile fathers are more likely than mothers to say they worry about accessing good-quality childcare or education.
Figure 7. Mothers and fathers worry about accessing good-quality childcare or education for their children
Share of respondents who are “somewhat concerned” or “very concerned” about not being able to access good-quality childcare or education for their children (or young members of your family) in the next year or two, by gender and parental status
Note: Respondents were asked how concerned they are not being able to access good-quality child care or education for their children (or young members of their family). Possible responses were “Not at all concerned”, “Not so concerned”, “Somewhat concerned”, and “Very concerned”. Respondents could also choose “Can’t choose”. Figure presents the aggregated total of respondents answering “somewhat concerned” or “very concerned”.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
The promise and peril of telework, technology, and work-life balance
COVID‑19 has had obvious effects on the nature of work-life balance for people who remained employed. Telework seems to have helped protect women, and especially mothers, from completely exiting the labour force, and in many countries the gender gap in the change in paid hours worked was largest among people who had to go to work on-site. Simply put, many mothers were forced to choose between working or caring for their children to a degree that fathers were not.
Yet there were negative consequences for teleworking mothers, as well. Among teleworkers, the gender gaps in employment status change were small, but the productivity of mothers seems to have suffered more (Alon et al., 2021[1]). Research suggests that teleworking mothers were more likely to be interrupted during work hours (Andrew et al., 2020[6]), and mothers who teleworked faced a higher childcare burden than mothers who could not telework and instead had to commute to work (Boll, Müller and Schüller, 2021[13]).
Although stigma against teleworking has diminished during the pandemic (Barrero, Bloom and Davis, 2021[18]), pre‑pandemic research found that flexibility stigma is gendered. Men were more likely to discriminate against workers making use of flexible work arrangements, while women, and especially mothers, were more likely to suffer from such discrimination (Chung, 2018[19]). In pre‑COVID‑19 research, employers had also been found to disadvantage teleworking employees in promotions even if they had higher productivity than workers on-site (Bloom et al., 2013[20]).
Perhaps reflecting this benefit-cost trade‑off – the ability to retain a job, but awareness of care obligations at home – RTM 2020 data show that fathers are slightly more likely to feel that technology like that used for telework will make it more likely for their job and work hours to become compatible with private life. Having said that, both mothers and fathers are optimistic, on average, that technology will improve work-life balance.
Figure 8. Fathers are slightly more optimistic than mothers that technology will make their job and work hours more compatible with private life
Percent of respondents who think that is “very likely” or “likely” that technology will help their job and work hours become more compatible with their private life by gender and parental status
Note: Respondents were asked how likely they think it is that technology will help their job and working hours become more compatible with their private life over the next five years. The possible responses were “Very unlikely”, “Unlikely”, “Likely”, and “Very likely”. Respondents could also choose “Can’t choose”. Figure presents the aggregated total of respondents who selected “very likely” or “likely”.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Public policies to support parents
The evidence presented in this paper suggests that mothers were disproportionately burdened by the added unpaid care obligations caused by COVID‑19 in 2020. While fathers provided additional unpaid care when schools and childcare facilities closed, mothers did much more. This inequality in unpaid work, in turn, corresponded with mothers’ negative employment outcomes, higher levels of stress, and worries about accessing childcare and school.
These pandemic outcomes reflect longstanding inequalities in unpaid work, women’s weaker labour force attachment, and women’s relatively lower wages, on average, all of which contributed to many women becoming the “default” caregiver when the crisis arose.
Governments must take a gender-sensitive approach and consider inequalities in unpaid work when building their public policy response to the crisis. This policy approach should include greater public investments in good-quality childcare, education, and out-of-school supports; increasing the length of, and mandating take‑up of, paternity leave, which has implications for fathers’ long-term caregiving behaviour; and normalising the use of telework across men and women, so that women are not penalised for using it. Finally, policy makers must embed all of these strategies in a whole‑of-government approach to gender mainstreaming.
Change gender norms by incentivising fathers to take parental leave
Father’s caregiving for children around birth can have important downstream effects on equality in unpaid caregiving as children age. Many countries have been trying to encourage a more equal division of paid and unpaid work across men and women through father-specific parental leave programmes, which incentivise fathers to leave work to care for young children immediately after (or for a longer period) after birth (OECD, 2019[21]). A few countries, such as Portugal and Spain, have begun mandating a period of obligatory paternity leave for fathers after a child’s birth. Indeed, when countries were surveyed on how best to increase men’s unpaid care work at home, the most common answers were changing boys’ and men’s attitudes towards caregiving and ensuring that men do not experience discrimination when they take leave from work to care for dependents (OECD, 2017[22]).
In addition to helping child development and fathers’ well-being, fathers’ participation in parental leave can promote gender equality through different mechanisms. Fathers’ leave‑taking can help change the within-household distribution of unpaid work, as both mothers and fathers are involved in parenting tasks soon after birth – an important period for establishing caregiving habits. Fathers’ leave‑taking may also strengthen women’s labour market participation (both entry and hours) by enabling mothers to return to work earlier and lessening the risk of statistical discrimination against childbearing-age women overall. Quasi‑experimental research in countries as diverse as Canada, Germany, Norway, and Spain finds that the offer of paid paternity or parental leave reserved for fathers does increase take‑up by fathers, and that, after take‑up, fathers-only parental leave produces positive effects on fathers’ downstream hours spent on unpaid work (OECD, 2019[21]). Having said that, more evaluations are needed on “what works;” the offer of father-specific leave, in and of itself, is often not sufficient. Relatively generous leave schemes reserved for fathers in Korea and Japan, for example, have not yet corresponded with widespread increases in fathers’ caregiving.
Strengthen social protection systems to improve access to childcare and out-of-school supports
COVID‑19 obviously presented an unexpected and overwhelming shock to schooling and early childhood education and care (ECEC) systems. In addition to a range of other challenges, countries that faced long-lasting shutdowns of schools also tended to experience larger gender gaps in the distribution of added unpaid care work (Figure 9). Countries are attempting to take lessons from the crisis and modernise educational systems to support children in the new normal,14 but governments should also consider how to encourage gender equality in caregiving for children outside of school.
Figure 9. Longer periods of school closures are associated with greater gender gaps in caregiving
Gender gap in the take‑up of the majority or entirety of additional unpaid care work (mothers minus fathers, 2020) and days of full school closure, RTM OECD countries, 2020‑21
Note: The y-axis (gender gap) is estimated as the share of mothers of U‑12s who took on most or all of the additional care work minus the share of fathers of U‑12s who took on most or all of the additional care work during school and childcare facility closures during COVID‑19 (see note to Annex Table 1.A.1 for description of question and response choices). The x-axis, days of full school closure, are calculated based on the period between 11 March 2020 – 2 February 2021 by UNESCO’s global monitoring of school closures caused by COVID‑19. The definition of school closure status used is in line with UNESCO’s methodology: schools are considered fully closed when the closures affect most or all of the schoolchildren enrolled at pre‑primary, primary, lower and upper secondary levels. Both Mexico and USA are excluded they are as outliers in the UNESCO data on school closures. Mexico is one of few countries in the world in which UNESCO reports schools were closed for around 14 months. The United States is listed as having no days of full closures, but there was in fact a high degree of state-level and local-level variation in closures across the fifty states.
Sources: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM and UNESCO Global Monitoring of School Closures, http://COVID-19.uis.unesco.org/global-monitoring-school-closures-COVID-19/.
To some extent the gender gaps during COVID‑19 reflect longstanding weaknesses in countries’ social protection systems and their historical reliance on women as a crucial pillar of the social safety net. The gender gap in COVID‑19 care work tends to be lower in countries that have historically spent more on family policies like childcare, family allowances, maternity and parental leave, and other cash benefits (Figure 10).
Figure 10. Gender gaps in caregiving are smaller in countries that spend more on family supports
Gender gap in the take‑up of the majority or entirety of additional unpaid care work (mothers minus fathers, 2020) and per capita spending on family benefits and programmes (2017)
Note: The y-axis (gender gap) is estimated as the share of mothers of U‑12s who took on most or all of the additional care work minus the share of fathers of U‑12s who took on most or all of the additional care work during school and childcare facility closures between during COVID‑19 (see note to Annex Table 1.A.1 for description of question and response choices). The x-axis presents OECD SOCX data on public and mandatory private programmes by social policy area and type of support (cash/in kind). The policy area family includes benefits that support families. Benefits are often related to the costs associated with raising children or with the support of other dependants. Types of programmes include family allowances, maternity and parental leave, and other cash benefits, as well as spending on early childhood education and care, home help/accommodation, and other benefits in kind. SOCX includes public spending on early childhood education and care up for children under age 6, but SOCX does not include public spending on education beyond that age. SOCX data is from 2017 and measured in per head, at current prices and current PPPS, in US Dollars.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM and OECD Social Expenditure Database, https://www.oecd.org/social/expenditure.htm.
Any gender-sensitive attempts to rebuild social protection systems and labour markets after the crisis must seriously invest in strengthening good-quality, widely accessible, and affordable ECEC and out-of-school supports. During the crisis, many countries addressed the challenges caused by childcare closures with paid or unpaid leave for parents and emergency care for essential workers. But in countries where childcare is more often privately provided, such as the United States, many facility shutdowns during COVID‑19 have become permanent, and there is now an even greater shortage of childcare spaces than pre‑COVID‑19 (Queisser, 2021[23]).
High childcare costs and an inadequate supply of ECEC are long-standing structural factors disincentivising parents from entering into paid employment – particularly those with low income potential, and particularly mothers. These women often then commit to a larger share of unpaid work at home. Affordable, accessible and good-quality ECEC is crucial for ensuring equality of opportunity in children’s’ learning outcomes and in parents’ employment outcomes (OECD, 2020[24]). Policymakers must design such measures in a way that boosts parents’ work incentives and ensure that support reaches the parents who need it most (Queisser, 2021[23]).
Parents recognise ECEC and other family supports as a key government intervention (Figure 11). On average across countries, when considering the taxes they might have to pay and the benefits they might receive, 63.9% of parents report that they would like the government to spend more or much more on family supports like parental leave, childcare benefits and services, and child benefits. Mothers are about five percentage points more likely than fathers to identify this as a policy priority.
Figure 11. Parents call on governments for more support
Share of respondents who say they would pay “more” or “much more” when asked, “Thinking about the taxes you might have to pay and the benefits you and your family might receive, would you like to see the government spend less, spend the same, or spend more in the area of family supports (e.g. parental leave, childcare benefits and services, child benefits, etc.)?”, by parenthood status, 2020
Note: Figure presents the aggregated share of RTM 2020 respondents, sorted by parent status, who say they would pay “more” or “much more” when asked, “Thinking about the taxes you might have to pay and the benefits you and your family might receive, would you like to see the government spend less, spend the same, or spend more in the area of Family supports (e.g. parental leave, childcare benefits and services, child benefits, etc.)?”
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Make telework gender-sensitive
As discussed earlier, the benefits and costs of teleworking are gendered. While teleworking allowed many women to avoid choosing between earning income or taking care of their children, for this very reason it also disproportionately hampered women’s productivity and increased their stress. Thus, while teleworking is a promising tool to help reconcile paid and unpaid responsibilities, it will not necessarily further co-responsibility for parenting nor increase gender equality in workplaces unless these are conscious goals (Chung et al., 2021[25]). With the expected increase in teleworking post-pandemic, it is important to lay down foundations so as to not simultaneously exacerbate the unequal distribution of unpaid work.
Recent policy recommendations at the EU level call for investing in digital infrastructure, training, and connections for all, with special attention to women, to enable them to fully participate in labour markets and address any form of digital divide; targeted actions and campaigns to reduce and break down stereotypical thinking that may harm women who telework; and embedding any discussion of telework within a broader campaign to improve access to high-quality and affordable care supports throughout the life cycle, to lessen women’s unpaid work obligations at home (European Economic and Social Committee, 2021[26]).
At the firm level, some researchers have argued for mandatory telework days to prevent a “race to the bottom” in terms of office presence post-pandemic (Barrero, Bloom and Davis, 2021[18]). This may act in the same vein as mandatory paternity leave to encourage men to take a more active role in unpaid care work. It could also help limit women’s disadvantage in the ability to spend excess hours at a workplace.
This research topic is gaining momentum now. Researchers must continue to evaluate the gendered effects of telework – and how to combat them – under post-pandemic conditions when children are ostensibly back in school and formal care, and with an eye towards trends in the future of work.
Embed a gender lens in all aspects of the recovery from COVID‑19
Finally, it is crucial to embed a gender-sensitive approach when designing policies for the recovery from COVID‑19. Beyond targeted policy actions, such as increasing ECEC spending, governments must engage in gender mainstreaming – in other words, embedding a gender lens into all aspects of governance (OECD, 2021[27]). This is crucial in ensuring that policies do not unintentionally hurt women by reinforcing existing gender stereotypes and inequalities. Short-term recommendations include the collection and analysis of gender-disaggregated data to monitor the differential effects of the pandemic on men and women as well as incorporating advice from experts in gender issues. In the long term, governments should aim to build policy that alleviates structural gender inequalities and to build gender-inclusive frameworks and plans – taking into consideration inequalities at home, in the economy, and in society at large.
Acknowledgements
This report was written by Valerie Frey and Lyydia Alajääskö. It was prepared in the OECD Directorate for Employment, Labour and Social Affairs (ELS), under the senior leadership of Stefano Scarpetta (Director of ELS), Mark Pearson (Deputy Director of ELS) and Monika Queisser (Senior Counsellor and Head of Social Policy).
The authors thank Willem Adema, Carlotta Balestra, Chris Clarke, Jonas Fluchtmann, Raphaela Hyee, Maxime Ladaique, Valentina Patrini, Stefano Scarpetta and Monika Queisser for insightful feedback. Liv Gudmundson and Lucy Hulett prepared the brief for publication, and Jayne Maddock and Alastair Wood provided logistical, publication and communications support.
References
[5] Adams-Prassl, A. et al. (2020), Inequality in the Impact of the Coronavirus Shock: Evidence from Real Time Surveys, Institute of Labor Economics, https://www.iza.org/publications/dp/13183/inequality-in-the-impact-of-the-coronavirus-shock-evidence-from-real-time-surveys (accessed on 20 September 2021).
[1] Alon, T. et al. (2021), “From Mancession to Shecession: Women’s Employment in Regular and Pandemic Recessions”, NBER Working Paper.
[6] Andrew, A. et al. (2020), “The Gendered Division of Paid and Domestic Work under Lockdown”, IZA Discussion Paper 13500.
[18] Barrero, J., N. Bloom and S. Davis (2021), “Why Working From Home Will Stick”, Working paper.
[20] Bloom, N. et al. (2013), “Does Working from Home Work? Evidence from a Chinese Experiment | Stanford Graduate School of Business”, No. 3109, https://www.gsb.stanford.edu/faculty-research/working-papers/does-working-home-work-evidence-chinese-experiment (accessed on 11 October 2021).
[3] Bluedorn, J. et al. (2021), Gender and Employment in the COVID-19 Recession: Evidence on “She-cessions”, IMF, https://www.imf.org/en/Publications/WP/Issues/2021/03/31/Gender-and-Employment-in-the-COVID-19-Recession-Evidence-on-She-cessions-50316 (accessed on 20 September 2021).
[13] Boll, C., D. Müller and S. Schüller (2021), “Neither Backlash nor Convergence: Dynamics of Intracouple Childcare Division after the First COVID-19 Lockdown and Subsequent Reopening in Germany”, IZA Discussion Paper, https://covid-19.iza.org/publications/dp14375/.
[19] Chung, H. (2018), “Gender, Flexibility Stigma and the Perceived Negative Consequences of Flexible Working in the UK”, Social Indicators Research 2018 151:2, Vol. 151/2, pp. 521-545, http://dx.doi.org/10.1007/S11205-018-2036-7.
[25] Chung, H. et al. (2021), “Covid-19, Flexible Working, and Implications for Gender Equality in the United Kingdom”, Gender and Society, Vol. 35/2, http://dx.doi.org/10.1177/08912432211001304.
[8] Del Boca, D. et al. (2020), “Women’s and men’s work, housework and childcare, before and during COVID-19”, Review of Economics of the Household, Vol. 18/4, http://dx.doi.org/10.1007/s11150-020-09502-1.
[12] Eurofound (2020), Living, Working and COVID-19, http://dx.doi.org/doi:10.2806/467608 (accessed on 28 October 2021).
[26] European Economic and Social Committee (2021), Teleworking and gender equality: Conditions so that teleworking does not exacerbate the unequal distribution of unpaid care and domestic work between women and men and for it to be an engine for promoting gender equality | European Economic and Social Committee, https://www.eesc.europa.eu/en/our-work/opinions-information-reports/opinions/teleworking-and-gender-equality-conditions-so-teleworking-does-not-exacerbate-unequal-distribution-unpaid-care-and (accessed on 15 October 2021).
[9] Farré, L. et al. (2020), “How the COVID-19 Lockdown Affected Gender Inequality in Paid and Unpaid Work in Spain”, IZA Discussion Paper 13434.
[28] Hupkau, C. and B. Petrongolo (2020), “Work, Care and Gender during the COVID-19 Crisis*”, Fiscal Studies, Vol. 41/3, pp. 623-651, http://dx.doi.org/10.1111/1475-5890.12245.
[15] Lyttelton, T., E. Zang and K. Musick (2021), “Telecommuting and Gender Inequalities in Parents’ Paid and Unpaid Work Before and During the COVID-19 Pandemic”, Journal of Marriage and Family, https://doi.org/10.1111/jomf.12810.
[14] Margaria, A. (2021), Fathers, Childcare and COVID-19, http://dx.doi.org/10.1007/s10691-021-09454-6.
[16] McCrory Calarco, J. et al. (2020), ““Let’s not pretend it’s fun”: How COVID-19-related school and childcare closures are camaging mothers’ well-being”, OSF Preprints.
[29] OECD (2021), Bringing Household Services Out of the Shadows: Formalising Non-Care Work in and Around the House, OECD Publishing, Paris, https://dx.doi.org/10.1787/fbea8f6e-en.
[17] OECD (2021), Main Findings from the 2020 Risks that Matter Survey, OECD Publishing, Paris, https://dx.doi.org/10.1787/b9e85cf5-en.
[2] OECD (2021), OECD Employment Outlook 2021: Navigating the COVID-19 Crisis and Recovery, OECD Publishing, Paris, https://doi.org/10.1787/5a700c4b-en.
[11] OECD (2021), “Risks that matter 2020: The long reach of COVID-19”, OECD Policy Responses to Coronavirus (COVID-19), OECD Publishing, Paris, https://dx.doi.org/10.1787/44932654-en.
[27] OECD (2021), Towards gender-inclusive recovery, OECD Publishing, Paris, https://doi.org/10.1787/ab597807-en.
[24] OECD (2020), “Is Childcare affordable?”, OECD, Paris, https://www.oecd.org/els/family/OECD-Is-Childcare-Affordable.pdf.
[21] OECD (2019), Part-time and Partly Equal: Gender and Work in the Netherlands, OECD Publishing, Paris, https://dx.doi.org/10.1787/204235cf-en.
[22] OECD (2017), The Pursuit of Gender Equality: An Uphill Battle, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264281318-en.
[4] OECD (2012), Closing the Gender Gap: Act Now, OECD Publishing, Paris, https://dx.doi.org/10.1787/9789264179370-en.
[23] Queisser, M. (2021), “COVID-19 and OECD labour markets: What impact on gender gaps?”, Intereconomics, Vol. 56/5.
[7] Sevilla, A. and S. Smith (2020), “Baby steps: The gender division of childcare during the COVID-19 pandemic”, Oxford Review of Economic Policy, Vol. 36, http://dx.doi.org/10.1093/oxrep/graa027.
[10] Zamarro, G. and M. Prados (2021), “Gender differences in couples’ division of childcare, work and mental health during COVID-19”, Review of Economics of the Household, Vol. 19/1, http://dx.doi.org/10.1007/s11150-020-09534-7.
Annex 1.A. Supplementary tables
Annex Table 1.A.1. Percentage of respondents who self-report that they took up the majority or the entirety of any additional care work resulting from school or childcare facility closures, by gender and children’s age
Note: Respondents were asked who in their household took on any additional care work as a result of school or childcare facility closures. The response options were “Entirely you”, “Mostly you, “Equally shared between you and your spouse/partner”, “Mostly your spouse/partner”, “Entirely your spouse/partner”, “Mostly someone else (another member of your household or someone from outside your household)”, and “A mixture of you (and/or your spouse/partner) and someone else”. Respondents could also choose “Can’t choose / difficult to say” or “Not applicable” as options. The results here present the aggregated results for individuals who answered “Entirely you” or “Mostly you”. Parents with children under 18/12 are individuals whose youngest child living in the household is under 18/12 years old, respectively. The gender gaps are calculated by subtracting the share of males from the share of females in the respective groups.
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Annex Table 1.A.2. Percentage of respondents who worry about accessing good-quality childcare or education for their children (or young members of their family) in the next year or two, by gender and parental status
Aggregated share of respondents who are “somewhat concerned” or “very concerned” about not being able to access good-quality childcare or education for their children (or young members of your family) in the next year or two, by gender and parental status
Note: Respondents were asked how concerned they are about not being able to access good-quality child care or education for their children (or young members of their family). Possible responses were “Not at all concerned”, “Not so concerned”, “Somewhat concerned”, and “Very concerned”. Respondents could also choose “Can’t choose”. Table 2 presents the aggregated results of “Somewhat concerned” and “very concerned.”
Source: OECD Secretariat estimates based on the OECD Risks That Matter 2020 survey, http://oe.cd/RTM.
Contact
Stefano SCARPETTA (✉ stefano.scarpetta@oecd.org)
Mark PEARSON ((✉ mark.pearson@oecd.org)
Monika QUEISSER (✉ monika.queisser@oecd.org)
Valerie FREY (✉ valerie.frey@oecd.org)
Notes
← 1. On average across OECD countries, women make up about 53% of workers in food and beverage services (e.g. cafés, restaurants and catering); 60% in accommodation services (e.g. hotels); and 62% in the retail sector.
← 2. Fathers generally increased their unpaid care work following the increase in caregiving responsibility related to school closures during the pandemic, but generally women took on more ( (Andrew et al., 2020[6]; Sevilla and Smith, 2020[7]; Del Boca et al., 2020[8]; Farré et al., 2020[9]; OECD, 2021[29]; Zamarro and Prados, 2021[10]; Alon et al., 2021[1]).
← 3. Hupkau and Petrongolo (2020), using longitudinal data in the U.K., find little evidence of a gender gap in working hours or job loss across genders in the United Kingdom at the start of the pandemic. However within households they find that women provided on average a larger share of increased childcare needs (Hupkau and Petrongolo, 2020[28]).
← 4. Details on the survey, including sampling and questionnaires, can be found on the OECD Risks that Matter website and in recent reports : https://www.oecd.org/social/risks-that-matter.htm
← 5. RTM is an attitudinal survey that includes background questions on job status, work history and job disruption to help illustrate the employment status of respondents and households. Accurate employment figures are best derived from traditional, regular labour force surveys (LFS) and administrative data like tax records. RTM cannot be directly compared with LFS results for a few reasons. RTM asks a simple, retrospective question on 2019 employment status over a period of several months, from October to December 2019, which does not track with any traditional LFS figures (e.g. a monthly or quarterly unemployment rate). RTM also has a smaller sample (n=1000) per country than traditional LFSs.
← 6. Respondents were asked “Were you employed and in paid work at any point during the last three months of 2019?” [yes/no], and “What is your current employment status now [Q3 2020)? Are you now currently employed?”
← 7. The full list of job disruptions and indicators in RTM 2020 are detailed further in (OECD, 2021[11]).
← 8. While some evidence from Germany suggests that the increase in additional care work was shared equally between men and women, the overall division of work remained largely unchanged and women – in light of longstanding inequalities – were still more likely than men to consider the division of work unfair. In the United Kingdom there is also some mixed evidence on whether additional care work was equally shared (OECD, 2021[29]).
← 9. Official and high-quality time use data are still being collected across countries, and it is difficult, cross-nationally, to provide exact breakdowns of hours spent on care. Only a few OECD countries’ national statistical offices, like those in the United Kingdom and the United States, have published public time use data from 2020. See for example https://www.ons.gov.uk/economy/nationalaccounts/satelliteaccounts/bulletins/coronavirusandhowpeoplespenttheirtimeunderrestrictions/28marchto26april2020 and https://www.bls.gov/tus/. Cross-nationally, one provider of time use estimates is Eurofound’s “Living, Working and COVID‑19” survey, which has run several waves of survey questions across EU countries since spring 2020. Eurofound asked respondents “Last month, on average, how many hours per week were you involved in [caring for/educating your children or grandchildren?” However there are important caveats in terms of sampling. The Living, Working and COVID‑19 survey should not be compared to the national survey office estimates, as it recruited participants via social media and used “snowball” sampling methods, with varying response rates across European countries (https://www.eurofound.europa.eu/sites/default/files/wpef20005.pdf). The survey also of course does not cover the entirety of the OECD.
← 10. This regression includes controls for the provision of additional care, gender, age group, education, type of employment status, income, and country fixed effects. The independent variable of interest is an interaction term between gender (women) and providing the majority of entire or additional unpaid care work.
← 11. Parents in this section include all parents with minor children, i.e. under age 18, as restricting by younger age group and employment status leads to very small subgroups for comparison.
← 12. Mothers fared worse in psychological distress relative to the comparison groups of women without school-age or younger children; men with school-age or younger children; and men without school-age or younger children (Zamarro and Prados, 2021[10]).
← 13. There may be self-selection issues among respondents as the Living, Working and COVID‑19 survey recruited participants via social media and used “snowball” sampling methods, with varying response rates across European countries (https://www.eurofound.europa.eu/sites/default/files/wpef20005.pdf).
← 14. See, for example, a stocktaking of education policy responses at https://www.oecd.org/education/state-of-school-education-one‑year-into-COVID-19.htm.