Measuring women’s work participation: why it is important to get it right Premium

By valorising women’s diverse activities, we may treat their vulnerabilities as strengths and fail to acknowledge the lack of access to income-generating activities for women

The 2022-23 Economic Survey (ES) should be credited for acknowledging that official statistics may have inadequately captured the true nature of women’s work in India. This inadequacy affects both how we perceive women’s economic contributions and the policies we develop to enhance women’s economic participation. 

The Economic Survey has listed three sources of measurement error: (1) using very broad categories that club productive economic activity with domestic unpaid work (such as in activity code 93 in NSS, or National Sample Survey, employment surveys); (2) asking single-shot questions without probing to categorise women’s work; and (3) not capturing expenditure-saving work that contributes to household wellbeing. The Economic Survey further recommends revising labour force surveys (for example, the Periodic Labour Force Survey) to bring them in line with methodologies recommended by the International Labour Organization (ILO), listing a pre-defined set of activities, to eliminate underreporting.

This attention to improving measurement of women’s labour market activities is welcome. The National Council of Applied Economics Research’s  (NCAER’s) National Data Innovation Center carried out an experiment in 2019 under the Delhi Metropolitan Area Study (DMAS) in the Delhi National Capital Region. The experiment compared a one-shot questionnaire following the National Sample Survey Organisation (NSSO) method of enumerating usual principal and subsidiary activity status. These questions were followed by more detailed questions. In the latter, respondents were asked about engagement in a predefined and exhaustive set of activities, along with the time spent in each of these activities. The results from the DMAS survey indicated an underestimation of women’s work by five percentage points following the activity listing method when including work on the family farm/leased-in land, self-employment, and daily wage work. Underestimation was to the tune of 16 percentage points when including livestock care. This was primarily driven by estimates from the rural area, where underestimation was  34 percentage points, but only 2 percentage points in urban areas. Note that these comparisons are for the same women, allowing us to get precise estimates of measurement error.

While it is difficult to generalise based on urban and rural areas surrounding Delhi, these results highlight deep-rooted challenges in measuring women’s work. Interviewers as well as respondents tend to overlook women’s work, particularly work in family-based enterprises. Our observations from field visits suggest that while male farmers think of themselves as employed in farming, women farmers see farm activities as simply extensions of their household duties. 

From a policy perspective, ignoring measurement errors can lead to a focus on the declining labour force participation rate (LFPR) of women with low education whereas the focus needs to be on moderately educated women as indicated in a study by Esha Chatterjee, Sonalde Desai, and Reeve Vanneman (2019), which shows that occupational sex segregation along with lack of demand for moderately educated women could have contributed to low LFPR. Moderately educated Indian women are less likely to opt for manual labour in family farms or in wage work, but at the same time remain excluded from non-manual semi-skilled jobs such as bus driving or masonry work as well as some of the white collar jobs such as sales. This has implications for targeting skilling programmes for women in non-traditional sectors. Measurement issues also have implications for designing social protection measures, with undercounting economically active women translating into lesser budgetary allocation for informal women workers.

While the Economic Survey should be commended for recommendations to improve measurement of women’s work, its proposal to expand the definition of what counts as employment is somewhat ingenious. The 2019 Time Use Survey shows that women often engage in a variety of activities that can be classified as “expenditure-saving”. This includes the collection of firewood, fetching water, and preparing atta from wheat to reduce purchasing costs. It also includes tutoring children and cooking. The Economic Survey suggests that we should recognise this “work” along with “employment” and collect this information through redesigned surveys.  

Politically this is quite appealing, and consistent with one stream of feminist activism. Expanding the definition of work immediately expands the number of women workers — including almost the total adult female population in India — and softens the stigma of low female work participation rates in India. It also dovetails with the feminist advocacy that has sought recognition of women’s unpaid work. However, this approach has two shortcomings. One, it conflates activities within and outside of the System of National Accounts (SNA) production boundary where own-use production of goods such as growing vegetables or caring for livestock is seen as being within the production boundary while tutoring children is generally considered outside the production boundary. Two, activities that are undertaken due to a shortage of infrastructure are suddenly seen as being valuable rather than classified as “domestic drudgery”. Women’s participation in these activities (for example, fetching firewood) may be necessary for household survival, but few women would choose to rely on cooking with wood if they can easily get access to gas stoves. Our research has shown that women’s engagement in salaried work provides an impetus for households to switch away from such domestic drudgery and invest in time-saving infrastructure such as LPG cylinders.