## wpiea2019225-print-pdf

## Source details

**Canonical URL:** [wpiea2019225-print-pdf](https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019225-print-pdf.pdf)

## Other formats

- [Markdown version](/-/media/files/publications/wp/2019/wpiea2019225-print-pdf.pdf.md)
- [Structured JSON version](/-/media/files/publications/wp/2019/wpiea2019225-print-pdf.pdf.json)

---

### Overview
- Unpaid work defined as work not compensated by a wage and comprises: (i) care work for children, the elderly and the sick and (ii) domestic chores—cleaning and household upkeep; construction and repairs; cooking and food production; household management and shopping, volunteering. Both categories include time spent traveling to perform those tasks.
- Time use surveys capture both market (paid) work and non-market (unpaid) activities. Analysis uses average time use statistics by gender for 90 countries and individual-level time use surveys for 18 countries (Sample A), covering around 300,000 individuals older than 15; around 54 percent of the sample is women.

### Key empirical findings
- On average, women do more than two more hours of unpaid work per day than men.
- Bridgman and others (2018) estimate unpaid work accounts for roughly 35 percent of GDP on average in their sample, with a range from 10 percent of GDP in Korea to 60 percent of GDP in Albania.
- UNRISD (2010) finds the value of unpaid work ranges from 10 to 39 percent of GDP.
- Country examples:
  - Women in Hong Kong spend 2.6 hours a day on unpaid work.
  - Women in Mexico spend 7.1 hours a day on unpaid work.
  - In Norway, women do 20 percent more unpaid work than men.
  - In the U.S., women do 60 percent more unpaid work than men.
  - In Japan, women do 4 times as much unpaid work as men.
- Domestic chores account for over 80 percent of total hours spent on unpaid work.
- In high-income countries, women spend 4.1 hours on domestic chores, roughly half an hour less than their peers in emerging and developing economies.
- Time-use surveys for 18 high- and middle-income economies show women spend almost two hours more than men preparing food and cleaning.
- Care work is slightly less than one hour per day on average and is relatively stable across income groups.
- Men in high-income countries increase time spent on care work from 0.2 to 0.4 hours per day and on domestic chores from 1.1 to 1.8 hours per day compared to low and lower-middle income countries.

### Development and distribution patterns
- As countries get richer, hours spent on unpaid work fall, particularly domestic chores; richer economies can afford “engines of liberation.”
- Reduction in unpaid work with development reflects both substitution of household tasks into market provision and redistribution of unpaid work toward men.
- Little or no relationship between hours spent on care work by women and GDP per capita; economic development reduces domestic chores more than care work.
- Despite technological diffusion and family changes, gender imbalances in unpaid work persist: men have increased unpaid work but women still spend from 20 to 1,000 percent more time than men around the world.

### Drivers of unpaid work (cross-country and individual-level)
- Labor market institutions, demographics, and social institutions drive unpaid work.
- Female unpaid work is negatively correlated with female labor force participation.
- Laws, regulations, and social institutional constraints that restrict women’s ability to work, manage wealth, or run businesses are associated with higher female unpaid work and lower male unpaid work.
- Time allocation influenced by family characteristics, education, and employment status.

### Regression specification and data (Samples A and B)
- Individual-level regression (estimated separately for males and females, and for paid and unpaid work):
  - Time_i = β1 HasPartner_i + β2 HasChildren_i + β3 Number_of_Children_i + Γ Education_i + Φ Employment_i + Λ Age_i + κ_c + ε_i
  - Education_i dummies: less than secondary, secondary, more than secondary.
  - Employment_i dummies: full-time, part-time, not in the labor force.
  - Country fixed effects κ_c. Standard errors clustered at the country level.
- Period-trend specification for advanced economies (Sample B):
  - Time_i,t = Φ Period_i,t + κ_c + ε_i,t
  - Period dummies capture average time trends; standard errors clustered at the country-year level.
- Time use sources:
  - Sample A: Time use surveys for 18 advanced and emerging countries.
  - Sample B: Time use surveys since the 1960s for 7 advanced economies.

### Key individual-level regression findings
- Household composition:
  - Women with a partner spend 1.5 hours more on unpaid work than single women.
  - Men with a partner spend 20 minutes more on unpaid work than single men.
- Education:
  - Women with more than secondary education reduce unpaid work by half an hour relative to women with less than secondary education.
  - The half-hour reduction is less than an eighth of the 4 hours that women spend on unpaid work on average.
- Employment and the “second shift”:
  - A woman employed full-time does 5 hours of paid work but reduces unpaid work by only 2.7 hours relative to a woman who is not employed; her work day expands by 2.3 hours (the “second shift”).
  - Women and men show a similar drop in paid work when working part-time; the increase in unpaid work is 50 percent higher for women than for men.
  - The gender imbalance in unpaid work is even larger when moving from full-time employment to not employed.
- Child effects:
  - Women with children spend 37 more minutes on unpaid work than women without children.
  - Time spent on unpaid work increases by 12 minutes, on average, per child for women.
  - Men with children increase unpaid work by 6 minutes, on average, per additional child.

### Recent trends in advanced economies (Sample B)
- Aggregate trends since 1961-1989:
  - Women are doing more paid and less unpaid work; in the 2000s women, on average, spent half an hour less on unpaid work per day than women in 1961-1989.
  - The decline in unpaid work for women was roughly matched by an increase in paid work.
  - Male paid work has declined, while time spent on unpaid work has risen by 40 minutes.
- Subgroup trends:
  - Redistribution within couples: women with a partner have reduced unpaid work and men with a partner now contribute more to unpaid work than in the past.
  - Women employed part-time are doing less unpaid work and working more for pay. No significant change for women employed full-time or not employed.
  - Males, irrespective of employment status, are doing more unpaid work (around 20 minutes per day on average).
  - Educated women (especially more than secondary education) are working more for pay and doing less unpaid work; unpaid work is increasing for all men, particularly for those with less than secondary education. Paid work by men with secondary education or less is declining.

### Cross-country elasticities and findings
- Female labor force participation:
  - Estimated elasticity of unpaid work to female labor force participation: -0.198 for females and 0.689 for males (both statistically significant at the 1 percent level), sample of 57 economies.
- Fertility:
  - Elasticity of female unpaid work to the fertility rate is 0.457 and statistically significant at the 1 percent level (sample of 69 economies).
  - Elasticity for male unpaid care work is 23 percent smaller and not statistically significant.
- Digital access:
  - Elasticity of female unpaid work to internet access is -0.102 and statistically significant at the 10 percent level (sample of 27 economies).
- Social institutions and values:
  - Women do significantly less unpaid work in societies with stronger equality in managing assets and stronger women’s rights in marriage (World Bank women, business, and the law index), conditional on GDP per capita.
  - Societies with more gender inequality experience significantly less unpaid work by men (OECD social institutions and gender index).
  - Male unpaid work is higher in societies where more people consider equal rights key to democracy or disagree with statements such as “men make better business executives than women do,” “men make better political leaders than women do,” or “a woman has to have children to be fulfilled.”

### Flexible work arrangements — selected elasticity coefficients (OECD bivariate regressions, controlling for GDP per capita)
- Female:
  - "Work from home several times a month" = -0.141***.
  - "Work from home less than several times a month" = -0.156***.
  - "Never work from home" = 0.552***.
  - "Employees have some ability to set schedule" = -0.187***.
  - "Working hours set by company" = 0.346***.
- Male:
  - "Work from home several times a month" = 0.230***.
  - "Work from home less than several times a month" = 0.176***.
  - "Never work from home" = -0.783***.
  - "Employees have some ability to set schedule" = 0.310***.
  - "Working hours set by company" = -0.527***.
  - Additional significant male coefficients: "Work from home more than several times a week" = 0.078* and "Employees determine working hours" = 0.165**.

### Model framework and calibration (Ngai and Petrongolo (2017) adaptation)
- Two market sectors: goods (j=g) and services (j=s).
- Female and male labor inputs enter production with sector-specific productivity parameters (A_j_i) and women’s comparative advantage parameters (ξ_j_i).
- Barriers to female participation represented by wedge parameter β_j_i < 1 that lowers female wages relative to their output; raising β_i (toward unity) represents removing barriers.
- Home services produced using female and male labor; household utility depends on composite consumption of goods and services (market and home) and leisure.
- Baseline calibrated to 18 countries: Austria, Chile, Ecuador, Estonia, Finland, France, Hungary, Italy, Japan, Mexico, Netherlands, Norway, New Zealand, Pakistan, Panama, South Africa, United Kingdom, and United States.
- Country-specific parameters derived from: (i) home and market production hours by gender; (ii) working-age population and employment in goods and services by gender; (iii) gender wage gap; (iv) value added by sector.
- Gender elasticity parameters η and η_l set to values calibrated in Ngai and Petrongolo (2017) for the United States.
- Norway chosen as benchmark country; β_i set to unity for Norway and for counterfactual β_i set to unity for all countries (β_i computed as ratio of wage gaps relative to Norway in baseline).

### Counterfactual scenario results — removing wedges (β_i → 1)
- General effects:
  - Reducing barriers to female labor force participation reduces hours women spend in home production and increases men’s participation in home production, lowering the gender gap in total hours of unpaid work.
- Country-specific magnitudes:
  - Pakistan: gap in unpaid work is 45 percent lower in the counterfactual scenario.
  - Mexico, South Africa, United Kingdom: the gap declines by about 20 percent.
- Output gains from removing barriers:
  - Countries with largest gender gaps in unpaid work (e.g., Pakistan and Japan) can gain between three and five percent of additional output by removing these barriers.
  - Countries with relatively low unpaid hours gaps (e.g., Estonia, Finland, Hungary) would benefit little, being already close to Norway in terms of gender hours gaps.
- Sensitivity:
  - Alternative calibration/assumption changes could yield output gains in the order of 7 to 11 percent for the least gender-equal countries in the sample.
- Caveats:
  - Counterfactual isolates labor reallocation following removal of the wedge on women’s relative wages; sectoral productivity levels are unchanged between baseline and counterfactual to isolate wedge effects.
  - Growth in service sector productivity (improvements in child care infrastructure, investment in health and education) would further promote reallocation of female labor away from unpaid work; marketization of home-produced services disproportionately increases women’s paid hours, while reallocation from manufacturing to services can reduce men’s paid hours.

### Policy implications and recommendations
- Investments in infrastructure and public services:
  - Invest in water, sanitation, electricity, transport, security services, health, and education to reduce very low-productivity domestic tasks (e.g., fetching water) and narrow gender gaps in labor force participation.
  - Investments in digital infrastructure to foster internet access can reduce unpaid work and support participation in the gig economy.
- Childcare, elderly care, and family services:
  - Provision of childcare and early education, and higher public spending on families, are associated with lower unpaid work by women and facilitate women’s participation in formal employment.
- Tax and fiscal policy:
  - High tax rates for secondary earners discourage labor force participation and raise women’s unpaid work; tax systems that encourage joint filing can result in much higher marginal tax rates for the secondary earner.
  - Higher secondary-earner tax rates lead to higher unpaid work by women, conditional on GDP per capita and the single-earner tax rate; the effect is stronger the higher the expected earnings of the secondary earner.
- Labor market policies and regulation:
  - Active labor market programs, better labor regulations, and efficient labor markets reduce female unpaid work by facilitating job matching and retaining women in the labor market.
  - Higher wage gaps encourage women to do more unpaid work, implying lack of “equal pay for equal work” creates disincentives for women to engage in paid work.
- Family-friendly policies:
  - Parental leave for both men and women and family leave for caring responsibilities foster women’s paid work.
  - Encouraging parity between paternity and maternity leave is preferable; longer periods of paternity leave can increase male contribution of care work even after the leave period and reduce the within-household gender wage gap.
  - Flexible work arrangements help redistribute unpaid work between men and women.
- Human capital and legal rights:
  - Invest in women’s human capital through education and health care.
  - Enshrine women’s rights in law and implement family-friendly policies such as parental leave and taxation of secondary earners.
- Macroeconomic gains:
  - Removing barriers to women’s paid employment can produce substantial output gains, particularly in countries with large gender gaps in unpaid work; estimated gains in this study range from about three to five percent for the most affected countries under the baseline calibration, with alternative calibrations yielding potential gains of 7 to 11 percent for the least gender-equal countries in the sample.

*Source: wpiea2019225-print-pdf (references, appendix, and sections I–IV as provided).*

### REFERENCES ____________________________________________________________________________________________ 30

### I. INTRODUCTION

### Overview
- Unpaid work is defined as work not compensated by a wage and broadly comprises: (i) care work for children, the elderly and the sick and (ii) domestic chores—cleaning and household upkeep; construction and repairs; cooking and food production; household management and shopping, volunteering. Both categories include time spent traveling to perform those tasks.
- Time use surveys capture both market (paid) work and non-market (unpaid) activities and are used in this paper to analyze unpaid work patterns.
- The analysis uses average time use statistics by gender for 90 countries and individual-level time use surveys for 18 countries (Sample A), covering around 300,000 individuals older than 15; around 54 percent of the sample is women.

### Key empirical findings
- On average, women do more than two more hours of unpaid work per day than men.
- Bridgman and others (2018) estimate unpaid work accounts for roughly 35 percent of GDP on average in their sample, with a range from 10 percent of GDP in Korea to 60 percent of GDP in Albania.
- UNRISD (2010) finds the value of unpaid work ranges from 10 to 39 percent of GDP.
- Country examples:
  - Women in Hong Kong spend 2.6 hours a day on unpaid work.
  - Women in Mexico spend 7.1 hours a day on unpaid work.
  - In Norway, women do 20 percent more unpaid work than men.
  - In the U.S., women do 60 percent more unpaid work than men.
  - In Japan, women do 4 times as much unpaid work as men.
- Domestic chores account for over 80 percent of total hours spent on unpaid work.
- In high-income countries, women spend 4.1 hours on domestic chores, roughly half an hour less than their peers in emerging and developing economies.
- Time-use surveys for 18 high- and middle-income economies show women spend almost two hours more than men preparing food and cleaning.
- Care work is slightly less than one hour per day on average and is relatively stable across income groups.
- Men in high-income countries increase time spent on care work from 0.2 to 0.4 hours per day and on domestic chores from 1.1 to 1.8 hours per day compared to low and lower-middle income countries.

### Development and distribution patterns
- As countries get richer, hours spent on unpaid work fall, particularly domestic chores; richer economies can afford “engines of liberation.”
- Reduction in unpaid work with development reflects both substitution of household tasks into market provision and redistribution of unpaid work toward men.
- There is little or no relationship between hours spent on care work by women and GDP per capita; economic development reduces domestic chores more than care work.
- Despite technological diffusion and family changes, gender imbalances in unpaid work persist in most countries; men have increased unpaid work but women still spend from 20 to 1,000 percent more time than men around the world.

### Drivers of unpaid work
- Unpaid work is driven by labor market institutions, demographics, and social institutions.
- Female unpaid work is negatively correlated with female labor force participation.
- Laws, regulations, and social institutional constraints that restrict women’s ability to work, manage wealth, or run businesses are associated with higher female unpaid work and lower male unpaid work.
- Time allocation is influenced by family characteristics, education, and employment status.

### Modeling and counterfactual policy experiment
- Methodology: a model featuring market and unpaid home production by women and men is used to quantify policy effects.
- Counterfactual: barriers to women’s paid work across 18 advanced and emerging economies are reduced to the level of Norway.
- Estimated gains from reducing constraints to women’s paid work:
  - Pakistan and Japan: policies that reduce constraints to women’s paid work yield between three to four percent of GDP and over 35 percent reduction in the unpaid hours gender gap.
- Countries with more equal allocation of paid and unpaid work gain less from targeted policies.
- The model predicts that further productivity growth of the services sector would encourage gradual marketization of household production and higher female labor force participation.

### Policy implications and channels
- Reducing and redistributing unpaid work is macro-critical because constraints to female labor force participation can result in resource misallocation and lower economic growth.
- Occupational downgrading and higher prevalence of part-time work among women contribute to gender wage gaps and feedback into gender inequality in unpaid work.
- Policies that reduce barriers to women’s paid work, improve household technologies, expand public services and infrastructure, and address social institutional constraints can reduce female unpaid work and increase female labor force participation.

### Contribution and paper structure
- The paper focuses on unpaid work from the perspective of gender inequality, using cross-country averages and individual-level time use surveys.
- It complements existing literature by providing further disaggregation on individual characteristics driving levels and trends in unpaid work and by quantifying gains from policies that reduce and redistribute unpaid work.
- Organization:
  - Section II: explores unpaid work with cross-country and individual level data.
  - Section III: examines potential drivers of unpaid work.
  - Section IV: identifies policies to reduce and redistribute unpaid work.
  - Section V: quantifies gains of reducing imbalances in unpaid work.
  - Section VI: concludes.

*Source: wpiea2019225-print-pdf (references, appendix, and sections I–II as provided).*

### 2. Female time allocation in hours, with

### wpiea2019225-print-pdf - 2. Female time allocation in hours, with respect to men

### Regression specification and data
- Time use sources:
  - Time use surveys for 18 advanced and emerging countries (Sample A).
  - Time use surveys since the 1960s for 7 advanced economies (Sample B).
- Individual-level regression (estimated separately for males and females, and for paid and unpaid work):
  - Time_i = β1 HasPartner_i + β2 HasChildren_i + β3 Number_of_Children_i + Γ Education_i + Φ Employment_i + Λ Age_i + κ_c + ε_i
  - Variables: HasPartner_i, HasChildren_i, Number_of_Children_i; Education_i dummies (less than secondary, secondary, more than secondary); Employment_i dummies (full-time, part-time, not in the labor force); country fixed effects κ_c. Standard errors clustered at the country level.
- Period-trend specification for advanced economies (Sample B):
  - Time_i,t = Φ Period_i,t + κ_c + ε_i,t
  - Period dummies capture average time trends; standard errors clustered at the country-year level.

### Key regression findings (individual-level)
- Household composition effects:
  - Women with a partner spend 1.5 hours more on unpaid work than single women.
  - Men with a partner spend 20 minutes more on unpaid work than single men.
- Education effects:
  - Women with more than secondary education reduce unpaid work by half an hour relative to women with less than secondary education.
  - The half-hour reduction is less than an eighth of the 4 hours that women spend on unpaid work on average.
- Employment effects and the “second shift”:
  - A woman employed full-time does 5 hours of paid work but reduces unpaid work by only 2.7 hours relative to a woman who is not employed; her work day expands by 2.3 hours (the “second shift”).
  - Women and men show a similar drop in paid work when working part-time; the increase in unpaid work is 50 percent higher for women than for men.
  - The gender imbalance in unpaid work is even larger when moving from full-time employment to not employed.
- Child effects:
  - Women with children spend 37 more minutes on unpaid work than women without children.
  - Time spent on unpaid work increases by 12 minutes, on average, per child for women.
  - Men with children increase unpaid work by 6 minutes, on average, per additional child.

### Recent trends in advanced economies (Sample B)
- Aggregate trends since 1961-1989:
  - Women are doing more paid and less unpaid work; in the 2000s women, on average, spent half an hour less on unpaid work per day than women in 1961-1989.
  - The decline in unpaid work for women was roughly matched by an increase in paid work.
  - Male paid work has declined, while time spent on unpaid work has risen by 40 minutes.
- Subgroup trends:
  - Redistribution within couples: women with a partner have reduced unpaid work and men with a partner now contribute more to unpaid work than in the past.
  - Women employed part-time are doing less unpaid work and working more for pay. No significant change for women employed full-time or not employed.
  - Males, irrespective of employment status, are doing more unpaid work (around 20 minutes per day on average).
  - Educated women (especially more than secondary education) are working more for pay and doing less unpaid work; unpaid work is increasing for all men, particularly for those with less than secondary education. Paid work by men with secondary education or less is declining.

### Cross-country drivers of unpaid work (aggregate regressions and elasticities)
- Female labor force participation:
  - Negative correlation between female labor force participation rate and hours of unpaid work by women, robust to controlling for GDP per capita.
  - Estimated elasticity of unpaid work to female labor force participation: -0.198 for females and 0.689 for males (both statistically significant at the 1 percent level), sample of 57 economies.
- Structural transformation and sectoral composition:
  - Expansion of the services sector is associated with a decline in female unpaid work.
- Fertility:
  - Cross-country: elasticity of female unpaid work to the fertility rate is 0.457 and statistically significant at the 1 percent level (sample of 69 economies).
  - Elasticity for male unpaid care work is 23 percent smaller and not statistically significant.
- Social institutions and gender norms:
  - Women do significantly less unpaid work in societies with stronger equality in managing assets and stronger women’s rights in marriage (World Bank women, business, and the law index), conditional on GDP per capita.
  - Societies with more gender inequality experience significantly less unpaid work by men (OECD social institutions and gender index).
- Values and perceptions:
  - Unpaid work is less unevenly distributed in societies that place lower value on gender inequality in jobs, politics, and education, conditional on GDP per capita.
  - Male unpaid work is higher in societies where more people consider equal rights key to democracy or disagree with statements such as “men make better business executives than women do,” “men make better political leaders than women do,” or “a woman has to have children to be fulfilled.”
- Digital access:
  - Controlling for development, women in countries with higher internet access spend significantly less time on unpaid work.
  - Elasticity of female unpaid work to internet access is -0.102 and statistically significant at the 10 percent level (sample of 27 economies).

### Policy implications and recommendations
- Investments in infrastructure and public services:
  - Invest in water, sanitation, electricity, transport, security services, health, and education to reduce very low-productivity domestic tasks (e.g., fetching water) and narrow gender gaps in labor force participation.
  - Investments in digital infrastructure to foster internet access can reduce unpaid work and support participation in the gig economy.
- Childcare, elderly care, and family services:
  - Provision of childcare and early education, and higher public spending on families, are associated with lower unpaid work by women and facilitate women’s participation in formal employment.
- Tax and fiscal policy:
  - High tax rates for secondary earners discourage labor force participation and raise women’s unpaid work; tax systems that encourage joint filing can result in much higher marginal tax rates for the secondary earner.
  - Higher secondary-earner tax rates lead to higher unpaid work by women, conditional on GDP per capita and the single-earner tax rate; the effect is stronger the higher the expected earnings of the secondary earner.
- Labor market policies and regulation:
  - Active labor market programs, better labor regulations, and efficient labor markets reduce female unpaid work by facilitating job matching and retaining women in the labor market.
  - Higher wage gaps encourage women to do more unpaid work, implying lack of “equal pay for equal work” creates disincentives for women to engage in paid work.
- Family-friendly policies:
  - Parental leave for both men and women and family leave for caring responsibilities foster women’s paid work.
  - The length of maternity leave does not seem to affect time spent on paid work by men, but both maternity and paternity leaves support women’s paid work.

*Source: Time use surveys and cross-country analyses as presented in wpiea2019225-print-pdf (sections 2–IV).*

### 1. Labor market efficiency 2. Active labor market policies

### 1. Labor market efficiency 2. Active labor market policies

### Parental leave, paternity parity, and flexible work — empirical associations
- Extended maternity leave may have a negative effect on female labor outcomes by keeping women outside of the labor force for longer periods when their skills may depreciate, exacerbating the “motherhood penalty” (Ruhm, 1998; Kleven and others, 2019).
- Encouraging greater parity between paternity and maternity leave is preferable (Elborgh-Woytek and others, 2012).
- Longer periods of paternity leave can increase male contribution of care work even after the leave period and reduce the within-household gender wage gap (Fernández-Cornejo and others, 2018; Andersen, 2018).
- Flexible work arrangements help redistribute unpaid work between men and women. The government, businesses, and trade unions can change the organization of paid work to better balance work and family.

### Elasticities of unpaid work to flexible work arrangements (OECD bivariate regression results, controlling for GDP per capita)
- Regression results reported with robust standard errors; statistical significance denoted as: *** p<0.01, ** p<0.05, * p<0.10.
- Selected coefficients (each cell = coefficient of log of time use category by gender on log of each flexible-work variable, controlling for GDP per capita):
  - Female: "Work from home several times a month" = -0.141***.
  - Female: "Work from home less than several times a month" = -0.156***.
  - Female: "Never work from home" = 0.552***.
  - Female: "Employees have some ability to set schedule" = -0.187***.
  - Female: "Working hours set by company" = 0.346***.
  - Male: "Work from home several times a month" = 0.230***.
  - Male: "Work from home less than several times a month" = 0.176***.
  - Male: "Never work from home" = -0.783***.
  - Male: "Employees have some ability to set schedule" = 0.310***.
  - Male: "Working hours set by company" = -0.527***.
  - Additional significant male coefficients include: "Work from home more than several times a week" = 0.078* and "Employees determine working hours" = 0.165**.
- Note: Table reports many more coefficients for both female- and male-employee flexible-work variables; above are illustrative significant values preserved exactly as reported.

### Model-based quantification — framework and calibration
- Model adapted from Ngai and Petrongolo (2017) to quantify effects of lowering barriers to women’s participation in paid market sector.
- Key model elements:
  - Two market sectors: goods (j=g) and services (j=s).
  - Female and male labor inputs enter production with sector-specific productivity parameters (A_j_i) and women’s comparative advantage parameters (ξ_j_i).
  - Barriers to female participation represented by wedge parameter β_j_i < 1 that lowers female wages relative to their output; raising β_i (toward unity) represents removing barriers.
  - Home services produced using female and male labor; household utility depends on composite consumption of goods and services (market and home) and leisure.
- Calibration:
  - Baseline calibrated to 18 countries: Austria, Chile, Ecuador, Estonia, Finland, France, Hungary, Italy, Japan, Mexico, Netherlands, Norway, New Zealand, Pakistan, Panama, South Africa, United Kingdom, and United States.
  - Country-specific parameters derived from: (i) home and market production hours by gender (Bridgman and others, 2018 and national time use surveys); (ii) working-age population and employment in goods and services by gender (World Bank WDI and ILOSTAT); (iii) gender wage gap (ILOSTAT); (iv) value added by sector (World Bank WDI and Bridgman et al., 2018).
  - Gender elasticity parameters η and η_l set to values calibrated in Ngai and Petrongolo (2017) for the United States.
  - Norway chosen as benchmark country; β_i set to unity for Norway and for counterfactual β_i set to unity for all countries (β_i computed as ratio of wage gaps relative to Norway in baseline).

### Counterfactual scenario results — removing wedges (β_i → 1)
- Main general effects:
  - Reducing barriers to female labor force participation reduces number of hours women spend in home production and increases men’s participation in home production, lowering the gender gap in total hours of unpaid work.
- Country-specific magnitude examples (exact reported figures):
  - Pakistan: gap in unpaid work is 45 percent lower in the counterfactual scenario.
  - Mexico, South Africa, United Kingdom: the gap declines by about 20 percent.
- Output gains from removing barriers:
  - Countries with largest gender gaps in unpaid work (e.g., Pakistan and Japan) can gain between three and five percent of additional output by removing these barriers.
  - Countries with relatively low unpaid hours gaps (e.g., Estonia, Finland, Hungary) would benefit little, being already close to Norway in terms of gender hours gaps.
- Alternative calibration/assumption sensitivity:
  - Changing assumptions on barrier parameter β_i and adjusting calibration could yield output gains in the order of 7 to 11 percent for the least gender-equal countries in the sample.
- Caveats:
  - The counterfactual isolates labor reallocation following removal of the wedge on women’s relative wages; sectoral productivity levels are unchanged between baseline and counterfactual to isolate wedge effects.
  - Growth in service sector productivity (improvements in child care infrastructure, investment in health and education) would further promote reallocation of female labor away from unpaid work; marketization of home-produced services disproportionately increases women’s paid hours, while reallocation from manufacturing to services can reduce men’s paid hours.

### Conclusions and policy recommendations
- Empirical summary:
  - Unpaid work is a substantial part of economic activity and is shouldered disproportionately by women.
  - While gender imbalances in unpaid work have declined, they remain significant: even in the most egalitarian countries, women do at least 20 percent more unpaid work than men, with the vast majority of unpaid work comprised of domestic chores rather than care work.
  - The burden of female unpaid work declines with development via marketization and structural transformation, and social institutions and values matter for redistributing unpaid work.
- Policy recommendations to reduce and redistribute unpaid work:
  - Invest in infrastructure and public services (water, electricity, security in developing countries; digital connectivity everywhere) to reduce unpaid work.
  - Foster provision of childcare and elderly care to replace unpaid work with paid work.
  - Invest in women’s human capital through education and health care.
  - Enshrine women’s rights in law and implement family-friendly policies such as parental leave and taxation of secondary earners.
  - Enhance the efficiency of labor markets and promote flexible work arrangements.
- Potential macroeconomic gains:
  - Removing barriers to women’s paid employment can produce substantial output gains, particularly in countries with large gender gaps in unpaid work; estimated gains in this study range from about three to five percent for the most affected countries under the baseline calibration, with alternative calibrations yielding potential gains of 7 to 11 percent for the least gender-equal countries in the sample.

*Source: Excerpt from wpiea2019225-print-pdf — 1. Labor market efficiency 2. Active labor market policies.*

### REFERENCES

### wpiea2019225-print-pdf - REFERENCES

### References (selected list from source)
- Aguiar, M., M. Bils, K.K. Charles, E. Hurst, 2018. “Leisure Luxuries and the Labor Supply of Young Men,” University of Chicago Booth Unpublished Manuscript.
- Alvarez, J., 2019. “Structural Transformation and the Rise of Female Labor Force Participation Across the World,” IMF Working Paper.
- Andersen, S. H., 2018. “Paternity Leave and the Motherhood Penalty: New Causal Evidence,” Journal of Marriage and Family 80, 1125-1143.
- Bick, A., N. Fuchs-Schundeln, and D. Lagakos, 2018. “How Do Hours Worked Vary with Income? Cross-Country Evidence and Implications,” American Economic Review 108(1), 170-199.
- Blau, F. and L. Kahn, 2017. “The Gender Wage Gap: Extent, Trends, and Sources,” Journal of Economic Literature 55(3), 789-865.
- Bridgman, B., 2013. "Home Productivity," BEA Working Papers 0091, Bureau of Economic Analysis.
- Bridgman, B., G. Duernecker, and B. Herrendorf, 2018. "Structural Transformation, Marketization, and Household Production Around the World," Journal of Development Economics 133, 102-126.
- Brussevich, M., E. Dabla-Norris, C. Kamunge, P. S. Khalid, and K. Kochhar, 2018. “Gender, Technology, and the Future of Work,” IMF Staff Discussion Note 18/07.
- Connolly, S. and M. Gregory, 2008. “Moving Down: Women's Part‐Time Work and Occupational Change in Britain 1991–2001,” The Economic Journal 118(526), F52-F76.
- Cubas, G., 2016. “Distortions, Infrastructure, and Female Labor Supply in Developing Countries,” European Economic Review 87, 194-215.
- Duernecker, G. and B. Herrendorf, 2018. “On the Allocation of Time. A Quantitative Analysis of the Role of Taxes and Productivities,” European Economic Review 102, 169–187.
- Duval-Hernández, R, L. Fang and R. L. Ngai, 2017. “Taxes and Market Hours. The Role of Gender and Skill,” CEPR Discussion Paper 12292.
- Elborgh-Woytek, K., M. Newiak, K. Kochhar, S. Fabrizio, K. Kpodar, P. Wingender, B. Clements, and G. Schwartz, 2013. “Women, Work, and the Economy: Macroeconomic Gains from Gender Equity,” IMF Staff Discussion Note 13/10.
- Elson, D., 2017. “Recognize, Reduce, and Redistribute Unpaid Care Work: How to Close the Gender Gap,” New Labor Forum 26(2), 52-61.
- Fang, L. and C. McDaniel, 2017. “Home hours in the United States and Europe,” B.E. Journal of Macroeconomics, 17(1), 1-27.
- Feldstein, M. and D. Feenberg, 1996. “The Taxation of Two-Earner Families,” in Empirical Foundations of Household Taxation, edited by M. Feldstein and J. Poterba, Chicago, University of Chicago Press.
- Fernández-Cornejo, J. A., E. Del-Pozo, L. Escot, and C. Castellanos-Serrano, 2018. “Can an Egalitarian Reform in the Parental Leave System Reduce the Motherhood Labor Penalty? Some Evidence from Spain,” Revista Española de Sociología 27(3), 45-64.
- Fisher, K. and J. Gershuny, 2016. “Multinational Time Use Study: User’s Guide and Documentation”, Version 9.
- Floro, M. and E. M. King, 2016, “The Present and Future of Time-Use Analysis in Developing Countries,” Asia-Pacific Population Journal 31(1), 5-42.
- Gallen, Y., 2018. “Motherhood and the Gender Productivity Gap,” Becker Friedman Institute for Research in Economics Working Paper No. 2018-41, University of Chicago.
- Garnero, A., S. Kampelmann, and F. Rycx, 2014. “Part-time Work, Wages and Productivity: Evidence from Belgian Matched Panel Data,” Industrial and Labor Relations Review, 67(3), 926-954.
- Gonzales, C., S. Jain-Chandra, K. Kochhar, and M. Newiak, 2015. “Fair Play: More Equal Laws Boost Female Labor Force Participation,” IMF Staff Discussion Note 15/02.
- Hochschild, A. and A. Machung, 1989. “The Second Shift: Working Parents and the Revolution at Home,” New York, Viking.
- Ichino, A., M. Olsson, B. Petrongolo, P. Skogman Thoursie, 2019. “Economic Incentives, Home Production and Gender Identity Norms,” IZA Discussion Paper No. 12391.
- Ilahi, N. and F. Grimard, 2000. “Public Infrastructure and Private Costs: Water Supply and Time Allocation of Women in Rural Pakistan,” 49(1), 45-75.
- Jain-Chandra, S., K. Kochhar, M. Newiak, Y. Yang, and E. Zoli, 2018. “Gender Equality: Which Policies Have the Biggest Bang for the Buck?” IMF Working Paper 18/105.
- Kleven, H., C. Landais, J. Posch, A. Steinhauer, J. Zweimüller, 2019. “The Impact of Family Policies on the Dynamics of Gender Inequality,” Slides.
- LaLumia, S., 2008. “The Effects of Joint Taxation of Married Couples on Labor Supply and Non-wage Income,” Journal of Public Economics 92, 1698-1719.
- Ngai, L. R. and B. Petrongolo, 2017. “Gender Gaps and the Rise of the Service Economy,” Working Paper.
- Olivetti, C. and B. Petrongolo, 2017. “The Economic Consequences of Family Policies: Lessons from a Century of Legislation in High-Income Countries,” Journal of Economic Perspectives, 31, 205-230.
- Ostry, J.D., J. Alvarez, R.A. Espinoza, C. Papageorgiou, 2018. “Economic Gains from Gender Inclusion: New Mechanisms, New Evidence,” IMF Staff Discussion Note 18/06.
- Peri, G., 2014. "Do Immigrant Workers Depress the Wages of Native Workers?," IZA World of Labor, Institute for the Study of Labor (IZA), 1–10, May.
- Ruhm, C. J., 1998. “The Economic Consequences of Parental Leave Mandates: Lessons from Europe,” The Quarterly Journal of Economics 113(1), 285-317.
- UNRISD, 2010. “Why Care Matters for Social Development,” Research and Policy Brief 9.
- Welzel, C., 2013. “Freedom Rising: Human Empowerment and the Quest for Emancipation,” New York, Cambridge University Press.
- World Economic Forum, 2017. “The Global Gender Gap Report.”

### Appendix: Data Sources — A. Time Use Averages
- Two primary sources for time use averages by country are the SDGs Gender Indicators (referred to as SDG) and the OECD time-use surveys (referred to as OECD).
- Coverage:
  - SDG: covers both developing and advanced economies; data available for 90 countries; some countries have multiple observations in the period 2000 to 2017, many have only one datapoint.
  - OECD: covers 28 OECD countries and includes breakdowns for unpaid work, paid work, leisure, personal care, and other.
- SDG unpaid work disaggregation:
  - Unpaid domestic chores include: food preparation, dishwashing, cleaning and upkeep of the dwelling, laundry, ironing, gardening, caring for pets, shopping, installation, servicing and repair of personal and household goods.
  - Unpaid care work includes: childcare, and care of the sick, elderly or disabled household and family members.
  - Both categories include traveling associated to the tasks.
- SDG linkage: data compiled to monitor attainment of Target 5.4 as part of Goal 5.

### Appendix: Data Sources — B. Time Use Surveys
- MTUS (Multinational Time Use Study) provides harmonized time use surveys for 23 countries with consistent variables and individual/household characteristics.
- Additional microdata included:
  - “Time Use Survey 2007” for Pakistan
  - “Encuesta Nacional de Uso del Tiempo 2014” for Mexico
  - “Encuesta Nacional de Uso del Tiempo 2015” for Chile
- Time use categories aggregated into: paid work or study, unpaid work, personal care, leisure, and other.
- Two microdata samples defined:
  - Sample A: covers the 18 countries using only the most recent time use survey for each country and excludes countries whose most recent survey is from before 2000.
  - Sample B: used to study time trends; includes countries with multiple surveys and is restricted to Canada (7 surveys), Finland (4 surveys), France (5 surveys), Netherlands (7 surveys), Norway (4 surveys), United Kingdom (7 surveys), and United States (16 surveys).

### Appendix: Data Sources — C. Other Variables (variable — source)
- A woman has to have children to be fulfilled — The World Values Survey
- Ability of women to be paid without prejudice — The World Bank's Women, Business and the Law Index
- Ability of women to run a business — The World Bank's Women, Business and the Law Index
- Ability of women to travel — The World Bank's Women, Business and the Law Index
- Ability of women to work without prejudice — The World Bank's Women, Business and the Law Index
- Active labor market policies — The World Economic Forum's Global Competitiveness Index 4.0
- Discrimination in the family — The OECD's Social Institutions and Gender Index
- Employees choose a between set schedules — The OECD's Family Database
- Employees determine working hours — The OECD's Family Database
- Employees have some ability to set schedule — The OECD's Family Database
- Employees set schedules within limits — The OECD's Family Database
- Equal rights are key to democracy — The World Values Survey
- Equality of managing assets — The World Bank's Women, Business and the Law Index
- Female labor force participation — The World Economic Forum's Global Competitiveness Report
- Fertility rate, total (births per woman) — The World Bank's World Development Indicators
- GDP per capita, PPP (constant 2011 international $) — The World Bank's World Development Indicators
- Gender equality in education — The World Values Survey based on Welzel (2013)
- Gender equality in jobs — The World Values Survey based on Welzel (2013)
- Gender equality in politics — The World Values Survey based on Welzel (2013)
- Husband and wife should both contribute to income — The World Values Survey
- Individuals using the Internet (percentage of population) — The World Bank's World Development Indicators
- Labor market efficiency — The World Economic Forum's Global Competitiveness Report
- Length of paid maternity and parental leave for mothers in weeks — The OECD's Family Database
- Length of paid paternity and parental leave for fathers in weeks — The OECD's Family Database
- Men make better business executives than women do — The World Values Survey
- Men make better political leaders than women do — The World Values Survey
- Never work from home — The OECD's Family Database
- Percentage of children aged 0-2 in formal childcare and pre-school — The OECD's Family Database
- Public spending on early education/care as a percentage of GDP — The OECD's Family Database
- Reproductive choices — The World Values Survey
- Restricted access to productive and financial resources — The OECD's Social Institutions and Gender Index
- Restricted physical integrity — The OECD's Social Institutions and Gender Index
- Share of employment in services — The International Labour Organization
- Son preference — The OECD's Social Institutions and Gender Index
- Tax rate entering work at 100% of average earnings, with 2 children, 1 partner earning 67% of the average — The OECD's Family Database
- Tax rate entering work at 50% of average earnings, with 2 children, 1 partner earning 67% of the average — The OECD's Family Database
- Tax rate entering work at 67% of average earnings, with 2 children, 1 partner earning 67% of the average — The OECD's Family Database
- Total public social expenditure on families as a percentage of GDP — The OECD's Family Database
- Women's rights in marriage — The World Bank's Women, Business and the Law Index
- Work from home less than several times a month — The OECD's Family Database
- Work from home more than several times a week — The OECD's Family Database
- Work from home several times a month — The OECD's Family Database
- Working hours set by company — The OECD's Family Database
- Working rights when having children — The World Bank's Women, Business and the Law Index

*Source: wpiea2019225-print-pdf - REFERENCES (content provided).*

---


_Source: https://www.imf.org/-/media/files/publications/wp/2019/wpiea2019225-print-pdf.pdf_