What explains the decline in female labour force participation in India?

It is often argued that female labour force participation is declining in India due to rising incomes that allow more women to stay at home, which is the preferred household choice in a predominantly patriarchal society. However, this column contends that the trend is mainly explained by a scarcity of suitable jobs opportunities outside of farming and close to place of residence.

Trends in India's female labour force participation ought to be a matter for concern for anyone who cares about gender equality. As argued in the recent Gender and Jobs World Development Reports (World Bank, 2011 and 2012), gainful work by women, and especially paid employment, are correlated with their agency in decision-making at the household level and in society more broadly. Women's work is also associated with better development outcomes, including greater investments in children's education and health. India has one of the lowest female participation rates in the world, ranking 120 among the 131 countries for which data are available (International Labour Organization (ILO), 2013). Even within South Asia, it is in sixth position among eight countries, just above Pakistan and Afghanistan. Moreover, female labour force participation rates (LFPR) in India have been declining for several years now. The trend has been particularly strong in rural areas, where female LFPR fell from 49% of working-age population in 2004-05 to 36% in 2011-12.

In a recent paper (Chatterjee, Murgai and Rama 2015), we argue that the explanation for this worrying decline is a jobs deficit. What most working women do in India does not match the image of a nine-to-five job, with a pay check at the end of the month. A majority are farmers. Many others have marginal jobs, not involving an employer, or not ensuring steady work throughout the year. This makes it challenging to assess who actually participates in the labour force. In our paper, we argue that a fourth of the observed decline in female LFPR could be a measurement artifact, and should actually be relabelled as an increase in unemployment. Perhaps more importantly, the number of farming jobs has been shrinking without a commensurate increase in other employment opportunities, especially in the rapidly-expanding areas that are neither truly rural nor fully urban. Our results suggest that more than half the decline in female LFPR is due to the scarcity of jobs relative to the working-age population, and to the change in the composition of the available jobs away from farming.

Voluntary withdrawal from labour force?

A different explanation, the so-called 'income effect', has been the focus of most academic research in recent years¹. Higher earnings have gradually allowed more rural women to stay at home, it is argued, and this is a preferred household choice in a predominantly patriarchal society. Other considerations in the recent literature include growing school attendance of young girls and the lack of institutional child support in the context of shrinking family sizes. These explanations are consistent with the fact that female LFPR is much lower in urban areas, as the latter enjoy higher incomes, offer more school choices, and host fewer extended families. While reference is increasingly made in the literature to the availability of jobs, or lack thereof, labour demand is not rigorously introduced in the analyses. Our results confirm the importance of the supply-side factors emphasised by the literature, but show that these factors can explain less than a quarter of the recent decline in India's female LFPR.

Our paper presents evidence pointing to a less 'voluntary' withdrawal from the labour force than the income effect explanation implies. At an aggregate level, districts which experienced faster wage growth also saw larger declines in rural female LFPR. But the estimated correlation is such that a doubling of wages in real terms – which is what happened in India between 2004-05 and 2011-12 – would only result in a decline in rural female LFPR by around 3 percentage points. Looking more closely, the criterion to assess labour force participation turns out to be excessively stringent. As per the National Sample Survey (NSS) definitions of 'Usual Status' and 'Principal Status', a person is not considered part of the labour force if he or she has not been looking for a job for at least six months during the survey year. As a result, women who were registered with a placement agency, or who worked or sought work under the Mahatma Gandhi National Rural Employment Act (MNREGA), are counted as being out of the labour force. Elsewhere they would most probably be considered unemployed, hence as part of the labour force. If their number is taken into account, the rural female LFPR would be about 3 percentage points higher.

Figure 1. Female unemployment along the rural-urban gradation

Source: Estimates based on the 2004-05 and 2011-12 NSS Employment and Unemployment Surveys.

Scarcity of suitable jobs for women

Beyond the income effect and measurement issues, the main driver of the decline in female LFPR is the transformation of job opportunities at the local level. As elsewhere, urbanisation in India involves rural migration to cities. But it is also characterised by a densification of rural areas, or cities moving to people. As a result, the rural-urban divide is giving way to a rural-urban gradation. One of the innovations of our paper concerns the measurement of job opportunities at a highly disaggregated level along this gradation. When we treat locations as a rural-urban continuum, rather than as a dichotomous choice, it becomes evident that the decline in farming jobs has happened mainly in small villages with less than 5,000 inhabitants. And this decline has not been offset by job creation out of agriculture (Figure 1).

In a traditional society, women's work outside the home is acceptable if it takes place in environments perceived as safe and allows for flexibility for multi-tasking. From this perspective, female LFPR can be expected to depend on the availability of suitable jobs such as farming, which are flexible and mainly 'at home'. But our analysis shows that there is a 'valley' of suitable job opportunities along the rural-urban gradation in India. In recent years, suitable job opportunities for women have declined dramatically in villages and small towns. And it is in this intermediate space, neither truly rural nor fully urban, that a large share of India's urbanisation is taking place.

The other innovation in our paper is the explicit incorporation of these job opportunities in the estimation of female labour force participation decisions. We do so in a way that avoids circularity, as causality could also go the other way – from participation decisions to employment levels. Once the local employment structure is incorporated in the analysis, the explanatory power of the supply-side effects considered in the literature is weakened considerably. Interestingly, urbanisation becomes a much less relevant factor in the participation decision than the gap in female LFPR between rural and urban areas would suggest. For a similar local employment structure, there is virtually no difference in participation rates along the rural-urban gradation. Thus, it is the employment opportunities around where a household lives which matter for female LFPR, and not so much where the household actually lives.

The nature of jobs matters

The types of jobs available also matters. For the same employment level, relative to the working-age population, locations with more farming or casual wage jobs see higher female LFPR. The literature on female labour force participation in advanced countries often makes the argument that women prefer 'service sector' or 'white collar' jobs (Goldin 1994, Mammen and Paxson 2000). But in India where women bear the bulk of the child care and domestic work burden, the flexibility offered by working in the farm or in casual jobs is highly valued. Indeed, three-quarters of women who were willing to accept work, if made available, favoured regular part-time jobs. Unfortunately, there are not so many of them available in India.

Notes:

The literature includes studies by Olsen et al. (2006), Himanshu (2011), Rangarajan et al. (2011), Kannan et al. (2012), Neff et al. (2012), Bhalla et al. (2012), Abraham (2013), Klassen et al. (2013) and Chand et al. (2014).