Why female employment is falling
Low and declining rates of female participation in the workforce imply significant under-utilisation of labour resources in the Indian economy
Despite rapid fertility transition, broad increases in women’s educational attainment, and substantial economic growth over the past two decades, the share of Indian women who work has fallen over time. Currently, only about one-third of India’s half a billion adult females report being part of the labour force. These low and declining rates of female employment are potential causes for concern, because market work for women is often associated with better access to economic opportunities and greater decision-making power within the household. It also implies significant underutilisation of labour resources in the economy.
Why might Indian women be defying global trends in female labour force participation? In a recent International Growth Centre (IGC) study (Afridi, Dinkelman and Mahajan, 2016) we use three rounds of National Sample Surveys (NSS) to examine the demographic and socioeconomic characteristics of women that can account for the fall in their labour force participation rates (or LFPR) in and between 1987-1999 and 1999-2011.
LFPR over time, by gender and urban/rural area, for the sample of adults aged 25-65, show that men work at substantially higher rates than women, and rural women work at higher rates than urban women. Strikingly, rural Indian women experience the largest declines in LFPR over time — by 11 percentage points between 1987 and 2011. Within the sample of rural women, those who are married drive the overall decline in women’s LFPR.
Using decomposition techniques, we divide these changes in female LFPR into the part that can be accounted for by changes in observable characteristics, and the part that can be accounted for by changes in the returns to these characteristics. Our results reveal three broad patterns: Changes in individual attributes (increasing education and changing age distributions) and household factors (increases in household wealth and improvement in men’s education level) fully account for the fall in women’s LFPR between 1987 and 1999. Changes in these variables account for a much smaller share (just over half) of the decline in LFPR between 1999 and 2011. We do not find strong evidence that observable variables correlated with social stigma against women working outside the home (example, caste, religion) can account for a substantial proportion of the fall in women’s LFPR in either period. Increasing education levels among rural married women and the men in their households are the most prominent attributes contributing to the decline in LFPR in both decades.
Why should more women’s education contribute to lower LFPR? There exists a U-shaped relationship between women’s education and their labour force participation in India. As women move from being illiterate to having primary and middle levels of schooling, LFPR falls and only starts to rise as education increases to completed secondary schooling and to the graduate level. Over the last three decades, women in rural India have gained enough education to move younger cohorts from illiteracy to primary and middle schooling. But, in contrast to urban areas, schooling in rural areas has not yet expanded enough to pull most women into high school or further education. Human capital improvements may not yet have reached sufficiently high levels for women to earn high enough returns to market work in rural India.
Even with more education and fewer children, rural women’s time may be relatively more valuable in home production. This could be because women are objectively more productive at home with higher levels of education, or because men’s or women’s preferences for home versus market work change with more education.
We show that the decline in rural married women’s LFPR has been accompanied by an almost equivalent increase in the proportion of women who report domestic work as their primary activity in the past reference year during 1987-2011 (from 55 per cent in 1987 to 69 per cent in 2011). Other results suggest that the decline in the LFPR of women aged 25-40 was larger than the decline for 40-65-year-olds in both decades. These younger cohorts are most likely to have children of school-going age — almost twice as many 614-year-old children as women aged 40-65 — in the NSS, and therefore most likely to experience high returns to childcare at home.
We explore the relationship between education and time spent on childcare and other household activities using data from the 1998 Indian time-use survey. The data suggest distinct increases in women’s time spent in childcare and other chores at higher levels of education, up to higher secondary schooling.
Findings from other research give some reason to think that more education makes women’s time at home more valuable. First, returns to education in home production may have increased during the period of study because returns to children’s human capital were rising. Second, the cross-sectional evidence (from 1998) is in line with broad patterns in some other parts of the world like Brazil, where dramatic improvements in female education and reductions in family size only translated into more market work once women attained over eight years of education.
Our descriptive evidence suggests a compelling explanation for why women’s LFPR in rural India has declined over time. As women have got more education and poverty rates have fallen, the gap between returns to home production versus market work has grown larger.
Market work for women is often associated with better access to economic opportunities and greater decisionmaking power within the household