“How India Fails Its Women” proclaimed The Economist in a cover story in 2018. The short answer was: “patriarchal social mores supersede economic opportunity in a way more associated with Middle Eastern countries”. The article suggested that women in India were either dropping out of paid work voluntarily or due to social conservatism seen in the “ … enduring stigma of women being seen as ‘having to toil’".

Paraphrasing (in reverse) a common adage, this explanation is an excellent example of two rights making a wrong, almost. Or two pieces of a larger jigsaw that appear to fit, but in the absence of other pieces produce a partial and hence misleading picture.

There is no doubt that women’s participation in paid work has been declining over the last 25 years. The report of the Periodic Labour Force Survey (PLFS) for 2017-18, conducted by the National Sample Survey (NSS), shows the trend clearly. While men’s labour force participation rates (LFPR) have always been higher than those for women, they have remained constant over 1993-94 to 2017-18. Women’s LFPRs have declined sharply and the entire decline is driven by rural women, whose LFPRs have declined from roughly 32 percent to 18 percent over the period.

It is also true especially in the last five years that right-wing, conservative and reactionary forces have been on the rise. Tendencies that might have once been regarded as extremist or fringe are now firmly occupying the mainstream terrain. Among the several facets of this shift to the right is the not-so-covert attack on women’s economic independence, their sexual freedoms, their desire to choose their partners, and their expressing a mind of their own. A deadly combination of misogyny and hatred of minorities underlies a large number of these attacks. For instance, the attacks on the so-called “love jihads”, on couples in public spaces by self-appointed vigilantes of the “anti-Romeo squads”, the targeting of inter-caste marriages between Dalit men and upper caste women, and the horrific crimes against women in public spaces, a recent instance being the mind-numbing, unimaginably brutal rape of 8-year old Asifa.

Then, of course, there are crimes stemming from “pure” unadulterated misogyny and an unabashed patriarchal mindset which extols the virtues of the ideal woman – the ever-sacrificing mother, daughter, wife and sister – the archetypal sati-savitri, adarsh Hindu nari (fiercely loyal to her husband above anything else, whose life is exclusively devoted to the service of her husband). While these tendencies are not confined to the ruling Hindutva coalition, it is a fact that in 2017, of all parties, BJP had the largest number of MPs and MLAs with cases of crimes against women, including abduction, rape and forcing minor girls into prostitution (Hindustan Times 2017). It would be interesting to see how this figure has changed after the 2019 elections.

Why is Two Plus Two Not Four?

Why then don’t these two very powerful and real trends add up to a simple explanation of women’s withdrawal from the labour force, precisely as The Economist analyses? Several reasons.

The Grey Zone: Unpaid, Invisible, Fractured Work

One, the focus on a dichotomous indicator (in the labour force, or out of it) misses a crucial dimension about women’s work in specific regional contexts such as South Asia. Indeed, at the two ends of the spectrum are at the one end women who clearly work outside the home for pay and at the other those who clearly don’t out of choice, being exclusively involved in care activities such as cooking, cleaning, routine household chores, and looking after children and the elderly. One can clearly say if these two groups of women are in the labour force or out of it.

However, the majority of women in South Asia are in between these two extremes. These are women whose involvement in economic work (activities that are within the standard boundaries of the System of National Accounts, i.e. counted as economic activities when national income or GDP is measured) lies in a grey zone. These are women who might work in the house or outside, and whose work might be paid or unpaid, and whose work might be continuous throughout the year, or seasonal; and it might be full time or part-time.

A woman might be involved in the family business, or the main activity that provides the livelihood for the family. For example, she could be involved in livestock rearing or farming or helping with the kirana shop, or involved in artisanal activity, such as making baskets, or weaving or pots. If these are family activities, then her contribution to economic work (over and above her “care” work), would not be paid. In such a case, it is highly likely that she would not be seen as a worker, neither by her family nor by herself. These would be women involved in unpaid economic work, either at home or outside but in family business.

Similarly, there would be workers paid by piece rate, working either at home (for example, rolling bidis, or stitching garments, or doing small labour-intensive jobs such as sticking bindis) or outside, either throughout the year or occasionally, and either for the full day or part of the day. Most such work would be poorly paid and irregular. When such women are asked in labour force surveys “over and above your domestic responsibilities, do you work”, some answer in the affirmative, and some do not. The latter don’t get counted as “workers”.

Then there would be women who work full time (for pay), but occasionally. This means that they don’t work throughout the year. These women are most likely to fall through the cracks of the statistical system. How? When surveyors come to them, they ask if they had worked the previous day (the daily rate measure of labour force participation), or in the last seven days (the weekly measure), or for the majority of the time in the last year (the usual or principal status measure) or for any stretch of 30 days in the last year. The first two questions are easily answered, and they would be classified as workers according to the daily or weekly status, depending on their response.

On the third question, their answer would most likely be “no” to the majority time question. The likelihood of their being able to work for 30 days at a stretch depends critically on the availability of work. Suppose, a woman worked on a construction site for 25 days, but not in the last seven days (and therefore not the previous day). Such a woman would not be a worker according to her weekly or daily status. And, both because she did not work for a majority of the time in the last year and because she fell short of the 30-day cut-off, she would not be counted as a worker based on her usual status. Thus, despite actually being in the labour force, she would be counted as out of the labour force.

A short summary of these points is: women are “working”, but are not being counted as such. Their participation in economic work is invisible.

The Demand-Side Story

Two, the whole focus on labour force participation reduces the issue of women’s involvement to a labour supply issue. If women’s involvement is seen only as a supply-side story, then the attention, quite naturally, would be on factors that inhibit women’s ability or inclination to go out of the house and work. Thus, the spotlight turns to constraints such as the stigma attached to working outside the home, which may or may not be internalised by women, or a rise in religious fundamentalism (of both the Hindu and Muslim variety), or a resurgence of a patriarchal mindset, which asserts the supremacy of the male breadwinner model in which the man earns and the woman cooks, cleans and cares for the household.

As we noted in the previous example of the construction worker, whether a woman is working or not critically depends on whether there is work available. The fall in rural women’s LFPR should make us turn the spotlight on the nature of work availability, especially the non-farm opportunities. Yet, the literature on this topic tends to favour the supply-side explanations. Commentaries on the demand-side story are not absent, but less common.

Putting these two factors together, we see that (a) more women work than indicated in the official statistics, and (b) if they don’t, the lack of suitable work opportunities has something to do with it, over and above any other constraints arising from within households, or communities.

Understanding Women’s Economic Work

A great deal of the focus in this discussion is on the decline in the participation of women in the labour force. However, an equally (if not more) important issue is the persistently low level of women’s LFPR in India, which is lower than in neighbouring Bangladesh and Sri Lanka. In joint work with Naila Kabeer (Deshpande and Kabeer 2019), we have explored factors that shape the low level of women’s participation in work. Our results are based on a large primary household survey in seven districts in West Bengal. We collected data on all the indicators included in the official surveys and on additional variables that are usually not included in surveys.

Since we wanted to focus on the specific internal constraints that inhibit women from working, we asked specific questions. Were they were primarily responsible for child care, for elderly care, and for standard domestic chores (cooking, washing clothes etc.)? Did they cover their heads/faces always, sometimes, or never? The latter is taken as a proxy for cultural conservatism; indeed, internationally, women covering their faces in public spaces is often attacked as an oppressive practice. Of course, the context in the West is different in that covering heads/faces is associated with being Muslim. In India, the practice is followed by both Hindus and Muslims. In recognition of that we label it more broadly as “veiling”, and not as wearing a burqa or hijab.

We implemented simple changes to the official survey questionnaires in order to get better estimates of women’s work that lies in the grey zone. Accordingly, our estimates are higher than official estimates. But even with improved measurement, just a little over half (52 percent) get counted as “working”. This means that even after work in the grey zone is included, women’s participation in work in India is low.

Critical Role of Domestic Chores

We then investigated the main constraints on women's ability to "work" by identifying factors that make women more likely to be in regular work,as well as in the grey zone (economic, expenditure saving work) relative to "not working" (neither involved in regular work, nor in expenditure saving economic activities).

Academic research on women’s work has identified a set of factors that are commonly used as explanatory variables. For instance, the age of women is said to influence a woman’s ability to work. The likelihood of women working increases with age up to a point and then starts declining. Marriage is another factor. Married women are less likely to work, relative to those who were never married, widowed, separated or divorced. Education too is a factor: women with the lowest and highest education levels are more likely to be working compared to women with middle-levels of education. Location matters: all-India estimates reveal that rural women are more likely to be working than urban women.

[B]eing primarily responsible for routine domestic tasks such as cooking, cleaning and household maintenance as well as for care of the elderly, lowers the probability of women working.

In our survey, we found that urban women are marginally more likely to be in conventional economic activity than rural. We attribute this to the fact that a large part of our sample is from Kolkata and Howrah -- the two most economically vibrant centres in the state of West Bengal. There is inter-district variation in LFPRs as well, which mirrors larger surveys that find both inter-state and inter-district variation. 

The unique finding from our survey is that over and above the standard explanations for low women’s participation that is to be found in the literature, being primarily responsible for routine domestic tasks such as cooking, cleaning and household maintenance as well as for care of the elderly, lowers the probability of women working.

If, after controlling for the standard explanatory factors, domestic chores emerge as an important determinant of women’s labour force participation, the question that arises is: to what extent do the low LFPRs found in India, in particular, but more broadly in South Asia and MENA (Middle-East and North Africa), reflect international differences in women’s involvement in housework? There is some indicative evidence that in these regions, women indeed spend more time on what can be broadly defined as unpaid care work (i.e., care of persons, housework or other voluntary care work), relative to a range of other developing and developed countries in the world. According to the Organisation for Economic Co-operation and Development(OECD) data, in 2014 the female-to-male ratio of time devoted to unpaid care work was 10.25 and 9.83 in Pakistan and India, respectively ,compared to 1.85 in UK and 1.61 in the US. (India and Pakistan are the two countries with the lowest female LFPRs in South Asia)

Factors traditionally viewed as cultural norms that constrain women’s participation in paid work, such as the practice of veiling or adherence to Islam, are insignificant in our analysis after the conventional variables have been accounted for. Given that the primary responsibility of domestic chores falls on the woman, we suggest that the conventional definition of cultural norms needs to be revised, and shifted to focus on the real culprit, viz, the cultural norm that places the burden of domestic chores almost exclusively on women.

Do women really want to participate in paid work, or have they either internalised the male breadwinner model that relegates them to take care of the home and the family?

Is There an Unmet Demand for Work?

Do women really want to participate in paid work, or have they either internalised the male breadwinner model that relegates them to take care of the home and the family? What about the “income effect”, according to which women work only if necessary for economic reasons and withdraws from work as soon as they don’t need to? What about the marriage penalty, i.e. women dropping out of the labour force once they are married? Thus, women’s work might be a sign of economic compulsions of trying to make two ends meet, rather than an expression of their desire for economic independence.

[B]eing out of the labour force is less a matter of choice for large numbers of women and more a reflection of the demands of unpaid domestic responsibilities.

We explored the evidence for this in our survey. Married women are less likely to be working than unmarried women. But marriage in India is near universal (making marriage the most common career choice for women), and asking women to choose either marriage or paid work is not a fair or a realistic choice. We asked women who were not working if they would accept paid work if it was made available at or near their homes. The percentage that said “yes” was 73.5%. When questioned further, 18.7% expressed a preference for regular full-time, 7.8% for regular part-time, 67.8% for occasional full time and 5.78% for occasional part time. It would appear that there was indeed a major unmet demand for paid work, whether regular or occasional, full time or part time, as long as the work in question was compatible with their domestic responsibilities. Based on this, we suggest that being out of the labour force is less a matter of choice for large numbers of women and more a reflection of the demands of unpaid domestic responsibilities.

Rising Open Unemployment

Labour force participation rates consist of women working and women seeking work or available for work (but not currently working), i.e. women who are in the labour force whether employed or not. Developing countries typically have underemployment or disguised unemployment, where individuals are engaged in very low productivity, subsistence activities and do not declare themselves openly unemployed. When jobs are few and far between, women typically withdraw from the labour force rather than declare themselves as seeking work, i.e. being openly unemployed. One feature of the 2017-18 NSS data is the staggering rise in open unemployment, which again is driven by rural women, a clear indication of the unmet demand for work.

What is the Role of Stigma or Fear of Sexual Violence?

We have now seen that there are other components of the puzzle that need to be joined, or other dots that need to be connected before the full picture about women’s low participation in work becomes clear.

What exactly is the role of stigma in explaining low participation by women? It is hard to get a clear answer to this because we would need hard evidence of rising intolerance towards women working outside the home which we do not have. Also consider this. Urban LFPRs have always been lower than rural. If stigma is the main reason underlying this gap, then it follows that urban women have faced greater stigma than rural. But the entire decline in LFPRs is due to rural women. Does this mean that stigma, which might be greater in urban areas has remained roughly constant but has increased in rural areas? This does not make sense. Finally, the stigma of working outside the home as a mark of low status is typically characteristic of upper caste women. Dalit and Adivasi women have always worked outside the home in far greater proportions. But the recent decline in LFPR is larger for them than their upper caste sisters.

The only set of explanations that fits all these facts is a combination of the following: (non)availability of work that is compatible with domestic responsibility, i.e. either at or near home or easy to get to.

What about a fear of sexual violence? Recent studies (for example, Chakraborty et al 2018) find that perceptions of violence deter women from working outside the home in the sense that either women are less likely to work in regions with greater violence against women, or that increased reports of sexual violence reduce the probability of urban women working outside the home. Both these stories are entirely plausible. However, they explain the variation in LFPRs across regions and time, i.e. why some regions see lower LFPRs than some others, or why LFPRs for urban women were lower during certain time periods relative to others. They do not shed light on the persistence of the low average labour force participation of Indian women. 

Internal Migration

According to the Census of 2011, nearly 70% of internal migrants were women. Exploitative, unsafe, informal working conditions with poor pay continue to characterise the lives of a large number of women migrants who are vulnerable to sexual violence. Yet, women are taking huge risks and are migrating in growing proportions.

Marriage-induced migration continues to be the single-largest cause of women’s internal migration, but its importance has declined over the last three decades. Between 2001 and 2011, the proportion of women migrating for work increased by 101 percent, which was more than double the rate for men (48.7 percent). Women who cited “business” as a reason for migration increased by 153 percent during 2001-11, more than four times the rate for men (Vivek 2017). Even women who migrated for marriage ended up looking for work and/or working. Thus, migration for marriage does not preclude women’s participation in work – again, it all boils down to the availability of suitable work.

Summing Up: The Full Picture

Persistently low rates of female labour force participation remain a challenge in India. A part of the issue, easier to fix, is the mis-measurement. But even correcting for that, women’s involvement in paid work remains low. The larger issue, harder to fix, is how to get more women in paid work. As has been suggested, women not joining or dropping out of the workforce is nothing short of a national tragedy (Desai 2019). This piece started with citing an editorial from The Economist. The same analysis calculated that if India were to “rebalance its workforce” (i.e. correct the gender imbalance), India would be 27% richer. Desai argues that India’s demographic dividend is much celebrated, but it is the squandering of the gender dividend that we need to be concerned about.

Education levels of Indian women are rapidly increasing (faster than those for men), and while the share of agricultural work has declined for both men and women, men have been able to find employment in other sectors. But this is not the case for women: “a man with Class 10 education can be a postal carrier, a truck driver or a mechanic; these opportunities are not open to women” (Desai, 2019).

Increased participation of women in paid work has several larger ramifications. In South Korea and Bangladesh, a rise in women’s labour force participation contributed substantially to a lowering of son preference. In India too, stereotypes about girls and boys are changing. Earlier boys were seen as pillars of old-age support, and girls as "paraya dhan" (literally, someone else's treasure) and thus not available for elderly care of parents. Qualitative studies more recently reveal that parents now see girls as more reliable and more likely to provide old-age support despite their marital commitments, because of their intrinsically caring nature. Boys are often seen as selfish and uncaring, not to be counted upon to fulfil their traditional duties towards parents. 

There has also been an improvement in the sex ratio at birth (which continues to be masculine, but less so). The total fertility rate is now at replacement level; family sizes have fallen.

Women are getting educated rapidly and they want to work. But, one, there aren’t sufficient suitable opportunities (the demand side). Two, the notion of suitability rests on compatibility of work with their “primary” responsibility of domestic chores. This, not religion or veiling, is the real cultural norm that constrains women’s labour supply.