Unpaid care work—the housework and care of persons that occurs in homes and communities of all societies on an unpaid basis—is an area that has generally been neglected by economists, as well as by many development actors. Yet the amount of unpaid care work carried out, the way that the burden of this work is distributed among different actors, and the proportion and kinds of care work that are unpaid or paid, have important implications for the well-being of individuals and households, as well as for the economic growth and well-being of nations.
This paper summarizes and compares findings from analysis of time use data from Argentina, Nicaragua, India, the Republic of Korea, South Africa and Tanzania for a project of the United Nations Research Institute for Social Development (UNRISD) on Political and Social Economy of Care
. The project as a whole aims to explore the way in which care—and care of persons in particular—is provided by the institutions of family/household, state, market and community, and by the people within these institutions. The analysis presented in this paper focuses on the quantitative aspects of unpaid care provided by individuals in households.
The paper consists of nine sections, as follows.
• Key concepts
introduces time use–related concepts which are utilized in later discussion in the paper.
• Background to the surveys in the six countries
describes the source of the data used for analysis in each of these countries. This is important to the extent that some of the variation across countries reported in the paper might reflect methodological, rather than “real”, differences between the countries.
• Basic gender patterns
presents a set of graphs derived from standardized sex-disaggregated tables compiled for each country. These graphs give a sense of the variation in male and female levels of engagement in, and the time spent on, employment-related work, unpaid care work and care of persons more narrowly defined.
• Distribution of time spent on care
explores the distributions that lie behind the averages that usually form the basis of time use analysis. The various country graphs confirm that while the amount of time spent by men on unpaid care work and person care tends to cluster at the lower end of distribution, there are substantial numbers of women who spend long hours on care work.
• The Tobit estimations
reports on the econometric analysis conducted in each of the countries to determine the main factors influencing the time spent on unpaid care work and person care across the six countries.
• Gender combined with other factors
discusses differences and similarities across countries in the way gender interacts with other factors explored in the Tobit estimations in determining how much care is undertaken by different individuals. In particular, it looks at how time spent differs between women and men in each of the countries in relation to the presence of young children in the household, employment status and age.
• The care dependency ratio
presents country results for a care dependency ratio proposed by the project as an indicator of care demand, in contrast to other sections that focus primarily on the supply of care.
• The monetary value of unpaid care work
discusses various approaches to assigning value to unpaid care work, and compares the results with a range of macroeconomic indicators for the six countries. These indicators include gross domestic product (GDP), paid work, government revenue and government expenditure on social services.
• The conclusion
offers some final remarks on the relevance of the findings.
The paper confirms some constant basic gender patterns in engagement in System of National Accounts (SNA) work, and unpaid care work, across the six countries. For all countries, the mean time spent on unpaid care work by women is more than twice that for men. The gender gap is most marked in India, where women spend nearly 10 times as much time on unpaid care work than men. Conversely, men tend to spend more time than women on SNA work across all countries. Again, India has the largest gender difference, with men spending nearly two and a half times as much time on SNA work as women.
When SNA and unpaid care work are combined, women are found to do noticeably more work than men in all countries. The volume of the total work done by men ranges from 74 per cent of the total amount done by women in South Africa to 94 per cent of the amount done by women in India. When the distribution of men and women in terms of time spent on unpaid care work is examined, there are far more men than women who do not do this type of work at all. Among those who do, there is strong clustering at points representing short times spent on this work. In contrast, there is high variability among women in the amount of unpaid care work done and, as a consequence, a notable level of inequality, with some women spending considerable time on it.
Tobit estimations confirm that, as expected, being male tends to result in doing less unpaid care work across all countries. This factor has the greatest influence (largest coefficient in absolute terms) of all tested factors in every country except Argentina. For all countries, having a (young) child in the household tends to increase the amount of unpaid care work done. The coefficient for age is always positive, while that for age squared is negative. This suggests an initial increase in the amount of unpaid care work done with increasing age, followed by a decrease. The amount of unpaid care work tends to decrease with increases in income, while being employed tends to decrease the amount of unpaid care work done in all countries except Tanzania. For most countries, being married tends to increase the amount of unpaid care work done.
Overall, there are at least as many differences as similarities across countries. In particular, there are significant variations in the “size” of care work done in the sense of the level of participation rates, average times spent by women and men on different activities, and absolute and relative differences between women and men. Some of these reflect methodological differences in terms of instruments, number of days covered, classification schemes, age group covered and so on. However, the methodological differences cannot explain away more than a small proportion of the variations.
The differences between countries in this paper thus confirm that gender is not “god-given” and immutable. Instead, gender is something that varies across countries and cultures. For policy purposes, however, what happens within a particular country is as important, if not more so, than cross-country comparisons. This paper, as well as the individual country research papers, present cross-sectional comparisons of different groups within a particular country at a particular point in time. Longitudinal comparisons of patterns of time use within a particular country are also needed. Countries therefore need to conduct time use surveys at regular intervals, using a standard methodology that allows reliable comparisons over time. This would be similar to the current practice of ongoing labour force surveys, although time use surveys would not need to be conducted as regularly as some labour force surveys because time use patterns are unlikely to shift as quickly.