Declining Female Work Participation Rate

The most striking revelation of the National Sample Survey Office’s (NSSO) 66th round survey is a significant fall in the FWPR between 2004-05 and 2009-10. The mid-quinquennial survey of 2007-08 (64th round) had shown that the rural FWPR had dropped to below the all-time low of 1999-2000. But in 2009-10 it further declined to 20% in usual principal status (UPS) work or employment and 26% in usual principal and subsidiary status (UPSS) work. In urban areas too, the FWPR fell substantially from 13.5% in 2004-05 to below 12% in UPS employment and from close to 17% to below 14% in UPSS work (Figure 1, p 119). With both UPS and UPSS work or employment having lost ground, it appears that relatively more permanent employment as well as short bursts of temporary employment have become less available to women.

It may be recalled that a sharp slump in the FWPR and a decline in the share of women in total employment was a major feature of the first decade of “economic reforms” in India. The findings of the 1999-2000 survey (55th round) had pointed to the displacement of women from employment in the 1990s, undermining the then widely held view that liberalisation and globalisation would lead to the feminisation of labour (Neetha 2009a; Mazumdar 2007). In 2004-05, the 61st round survey seemed to alleviate the gloom by showing a “revival” in the FWPR, albeit primarily driven

october 22, 2011 vol xlvi no 43

Figure 1: Trends in Female Work Participation Rates (UPS and UPSS)

Rural principal + subsidiary status Rural principal status Urban principal + subsidiary status Urban principal status 120 118 121 117 118 119 138 138 166 139 155 152 151 248 245 234 231 242 216 202 261 289 327 299 328 323 340

1983 1987-88 1993-94 1999-2000 2004-05 2007-08 2009-10 Source: Employment and Unemployment Reports, various rounds, NSSO.

by an increase in self-employment (Table 1). Further analysis highlighted a sharp rise in unpaid labour by women as a subcategory among the self-employed (Neetha and Mazumdar 2006). While the share of regular salaried employment increased by a few percentage points to reach a paltry 9% of the female workforce in 2004-05, casual (wage) labour witnessed a precipitate fall of 6 percentage points between 1999-2000 and 2004-05. This was matched and exceeded by a sharp rise in the proportion of self-employed women, which touched 61% in 2004-05, and within this, the share of the sub-category of unpaid women helpers rose to 72.5%.

Table 1: Percentage Distribution of Employment by Status – UPSS

Source: Employment and Unemployment Reports, various rounds, NSSO.

Clearly, even when the FWPR increased, the conditions of women’s work participation in India called for an approach other than the feminisation thesis, which was based on the premise that there would be a higher demand for women in wage work. Further, while the 2004-05 aggregate data on WPRs appeared to give the impression that more women had found jobs or employment, the disaggregated data on employment status suggested that women’s paid work had been substituted on a large scale with unpaid work by them.

With an unusual and unexpected jump in work participation, the analysis of the 61st round highlighted the need to bring the specifics of unpaid labour into discussions on the FWPR. Sanguine

Economic & Political Weekly

EPW

a pproaches to the general increase in WPRs in 2004-05 were already being countered with the argument that difficulties in finding paid work or wage work (jobless growth) were the real reason for the significant increase in self-employment, both among male and female workers (Ghosh 2009). We had argued at the time that the simultaneous expansion in the FWPR and self-employment could be interpreted as desperate attempts by women to prop up crisis-ridden, family-based production by working for no pay and/ or making a scrappy living through intermittent, home-based work for piece rates (defined by the NSSO as self-employment). And this was done with uncertain results in an employment scenario characterised by extraordinary insecurity and volatility (Neetha 2009a; Mazumdar 2008). Moreover, the substitution of women’s paid work with women’s unpaid work between 1999-2000 and 2004-05 could only be explained by family incomes being insufficient to meet petty production and related consumption costs, particularly in agriculture.2 In other words, the increased FWPR in 2004-05, which was largely based on an expansion of unpaid labour and selfemployment, actually signified a mounting crisis in finding paid employment and in adequate family incomes.3

The hypothesis regarding volatility or uncertainty and a crisis inherent in the self-employment-driven increase in the FWPR of 2004-05 appears to have been confirmed with a drop in the proportion of self-employed women from close to 61% in 2004-05 to below 56% in 2007-08 and a further decline to below 54% in 2009-10. Since the deflation of the self-employment balloon has occurred within the broader context of an overall decline in the number of employed women (see Table 1 in conjunction with Figure 1), it is no longer reasonable for anyone to argue that the increases in the FWPR b etween 1999-2000 and 2004-05 reflected growing employment opportunities for women. On the contrary, the abysmal situation of women’s employment in the era of liberalisation has once again been sharply brought to the fore by the surveys of 2007-08 and 2009-10. Significantly, the downturn in women’s employment from 2004-05 to 2009-10 contrasts sharply with the surge in gross d omestic product (GDP) growth rates in the same period.4

Long-term Downward Trend

When looked at in a slightly longer time frame, a long-term downward trend in the FWPR for rural areas is evident, as can be seen in Figure 1. An even more worrisome feature is that the downslide is more pronounced in the case of UPS workers. The proportion of UPS workers among rural females consistently fell between 1983 and 2009-10, apart from in 2004-05. If the UPS figures indicate a consistent process of marginalisation of women’s work in rural employment, the up and down swings in the UPSS work participation figures (Figure 1) reflect the volatility and precariousness of subsidiary, short-term or part-time employment for rural women, something that cannot be explained away by pointing to seasonal variations.

Among urban women, UPS and UPSS work participation rates have more or less followed a similar pattern – apparent stability marked by low levels and stagnancy, with a brief spike in 200405 being the only outlier. We will come back to the significance of these long-term trends in a later section. Before that, attention has to be drawn to the need to make a clearer distinction between employment with a wage or income and employment without any wage or income accruing directly to the worker when analysing trends in women’s employment.

Women’s Unpaid Labour

An aspect of the NSS data that is generally missed in debates on women’s employment trends is how much of women’s work participation is in the form of unpaid labour. Of the three employment or activity status categories of workers recorded by the NSS

– self-employed, regular salaried and casual labour – a higher proportion of the female workforce is always in the self-employed group. Concealed within the aggregate figures for the selfemployed is a large unpaid segment, which contributes to the production economy without receiving any independent payment or income for its labour. Of the three sub-categories within self-employment, own account workers and employers receive wages or incomes. The third sub-category, comprising those who work as helpers in household enterprises, is unpaid. Drawing on unit-level data, Figures 2(a) and (b) describe the characteristics of self-employed women (UPSS) in rural and urban areas respectively across the three large quinquennial rounds from 1993-94 to 2004-05 and the medium-large round of 2007-08.5

Figure 2(a): Characteristics of Rural Self-employed Women Workers (UPSS)

76.01

73.83 74.62

73.28

25.51 1.22 25.44 0.73 23.19 0.8 24.51 0.87 Own account worker Employer Worked as helper in household enterprise (unpaid family worker)

1993-94 1999-2000 2004-05 2007-08

Figure 2(b): Characteristics of Urban Self-employed Women Workers (UPSS)

Worked as helper in household enterprise (unpaid family worker)

Own account worker

51.61 1.97 46.42 52.69 1.34 45.97 49.73 1.56 48.71 55.12 1.39 43.49Employer

1993-94 1999-2000 2004-05 2007-08 Source: Unit-level data, various rounds, NSSO.

Figure 2(a) makes the high number of unpaid workers among rural self-employed women evident. Unpaid workers account for a whopping three-fourths of rural self-employed women in the four rounds, even crossing the three-fourths mark in 2004-05, with only a slight reduction in 2007-08. The sub-category of employers is negligible all through and even own account workers constitute just a quarter of rural self-employed women. Needless to say, when we allude to the unpaid work of women in the NSS data, it does not include domestic work such as cooking, cleaning and childcare, but economic activities. The last can be defined as market activities in the production of goods and services for exchange, as well as non-market activities that result in the production of primary goods for own consumption or relate to own account production (such as construction of houses, roads, wells, manufacture of machinery, tools, and so on, and construction of any private or community facilities free of charge). It would also include working for no payment either as a labourer or a supervisor.

Among urban women, own account workers have a slight edge over helpers or unpaid family workers, while employers are consistently too few to be significant. It should be borne in mind that the NSSO’s definition of own account workers includes those engaged in piece-wage work in homes, on the streets, and so on, and that “self-employed own account workers” is a misnomer for them.6 However, for the purposes of this discussion, the point to note is that the own account workers in the NSS data are paid for their work. Despite the greater relative weight of own account workers among urban self-employed women, unpaid workers constitute a substantial proportion of them, with a high of 49% in 2004-05 and a low of 43% in 2007-08. The peculiar feature of the data on urban self-employed women is that the share of unpaid labour and the FWPR follow a common trajectory. The share of unpaid labour increased when the FWPR rose and declined when the FWPR fell, emphasising the need to look more closely at unpaid work when assessing the implications of any increase in work participation among urban women.

A cursory look at the industrial distribution of unpaid women workers in rural areas confirms the belief that an overwhelming proportion of them are engaged in agriculture (Neetha and Mazumdar 2006). The unpaid workforce in rural areas is largely composed of peasant wives or daughters working as cultivators or supervisors on land owned by either their husbands or in-laws or fathers or parents. In urban areas, unpaid women workers are more evenly distributed across industries. Of the 43% of them in 2007-08, more than two-thirds were in community and personal services (including domestic workers, teachers, launderers, beauticians, and so on), and a little less than one-fourth in manufacturing (primarily home-based, piece-rated work). This makes it difficult to continue accepting the notion that unpaid work by women is a feature of “traditional” production systems or sectors that are outside the contemporary, monetised value exchange economy associated with “modern” capitalist social relations.

Structure of Female Workforce

With unpaid workers separated from all other categories of incomeearning or paid workers, Figures 3(a) and (b) (p 121) present a useful description of the broad structure of the female workforce in rural and urban areas from 1993-94 to 2007-08. As is evident from Figure 3(a), unpaid workers account for a very large proportion of the rural female workforce, consistently above 40% and with an all-time high of 48% in 2004-05. The share of paid or income-earning self-employment hovers only around 15%. What is striking is that in rural areas casual labour accounts for the majority of the paid female workforce. If one removes unpaid workers from the count and considers only the distribution within paid or income-earning workers in 2007-08, more than two-thirds (67%) were casual labourers, followed by 26% income-earning self-employed women and 7% regular salary-earning women.7

Across these four surveys between 1993-94 and 2007-08, the essential status distribution of the paid female workforce in rural areas remained unchanged, with casual labour dominating. Casual labour accounted for 64% of the women in the paid workforce in

october 22, 2011 vol xlvi no 43

rural areas even during the exceptional year of 2004-05 when a sharp increase in unpaid self-employment mirrored the fall in the proportion of casual labour. Of course, 2007-08 saw the reverse of this with the share of casual labour rising by 5 percentage points and the share of unpaid self-employed workers falling by the same proportion. There appears to be a consistent inverse relationship between the shares of unpaid workers and casual labour, the former declining when the latter rises and vice versa through the years. This suggests that a significant number of unpaid self-employed women tend to go for casual work when they are either able to find such work or are in special need, and return to being unpaid workers when casual work is unavailable or too difficult to perform, or when the demands of unpaid labour increase to a level incompatible with outside work. The increase in casual labour in the 64th round shows that it is the most significant form of work for income among women in rural areas.8 Unit data for 2009-10 is likely to show a further progress of this trend.

Figure 3(a): Percentage Distribution of Rural Female Workforce by Employment Status (UPSS)

Unpaid workers Paid self-employed workers Regular workers Casual workers 1993-94 1999-2000 2004-05 2007-08 43 42 48 43 16 15 15 15 3 3 4 4 39 40 33 38

Source: Unit-level data, various rounds, NSSO.

Figure 3(b): Percentage Distribution of Urban Female Workforce by Employment

Status (UPSS) Unpaid workers Paid self-employed workers Regular workers Casual workers 1993-94 1999-2000 2004-05 2007-08 21 21 23 18 25 24 24 24 28 33 36 38 26 21 17 20

Source: Unit-level data, various rounds, NSSO.

Interestingly, when the absolute number of paid female selfemployed workers (own account workers and employers) fell in rural India by more than 2,20,000 between 2004-05 and 2007-08, the number of female casual workers increased by more than four million.9 Further disaggregation of the data shows that female casual labour in public works increased by only 1,40,000 between 2004-05 and 2007-08. On the other hand, female casual labour in “other than public works” increased by well over 3.8 million in the two to three years between the two rounds to reach

42.24 million in 2007-08. The nature of this increase needs more detailed analysis.

In urban areas, the separation of unpaid work draws attention to the emergence of regular salaried workers as the most significant segment of paid workers. Figure 3(b) shows that regular workers were 38% of the urban female workforce (UPSS) in 2007-08, but if only the paid workforce is considered, the share of regular

Economic & Political Weekly

EPW

workers would be 46%, followed by self-employment (29%) and casual labour (24%). The sizeable proportion and increase in regular salaried workers is largely accounted for by the rapidly expanding number of women in education or teaching and in the category of paid domestic workers. The phenomenal increase in the number and proportion of women in paid domestic service between 1999-2000 and 2004-05 has been brought out by extracting the sub-categories of housemaids, cooks and governesses and baby-sitters from the broader industrial category of workers10 in private households (Neetha 2009b). Unfortunately, the 64th round (as also the 66th) has merged these sub-categories with gardeners, chowkidars (watchmen), and the like, making comparisons with 2004-05 and before difficult.11

To return to the principal issue that this part of the paper seeks to address, it has to be noted that even in urban areas unpaid work by women is not insignificant. Further, across the survey rounds of 2004-05 and 2007-08, a see-saw contest between the shares of unpaid labour and casual work can be observed, perhaps a muted echo of the same phenomenon in rural areas. It appears that even in desperate times, regular salaried domestic work is either not available or rejected in favour of other forms of casual labour by some, if not all, women making the transition from unpaid to paid work in urban India.

Paid Work Participation Rates

It is our contention that most of the studies on women’s work and employment based on the NSS data (including some of our own) have hitherto failed to incorporate the structural importance of women’s unpaid work in their analyses. They are usually based on UPS or UPSS figures, with the words “workforce” and “employment” used interchangeably, and without distinguishing between unpaid and paid work. As a result, the numbers and proportions used for analysis give us a somewhat skewed picture of women’s employment in the country.

Of course, it is not only women who are involved in economically productive but unpaid work. However, among men it is neither on the same scale as among women, nor does it have the same social significance. Unlike in the case of most men, women’s unpaid labour in productive activities is deeply rooted in the patriarchal nature of their households or families and property rights. It is therefore a material articulation of their lack of freedom and independence in the economy and society as a whole. The continued mediation by households or families of a large part of women’s unpaid but productive or economic labour and the incomes derived from it represent conditions of economic d ependence and patriarchal domination. In the case of men, the dependent relationship represented by unpaid labour is likely to be of a temporary nature in relation to parental or family property or a family activity that will either be inherited or broken free of at some stage in their lives. In the case of women, it is more likely to be of a durable nature, extending from their dependence on natal kin to dependence on their husband and his kin.

Trends in the unpaid work of women may also be an indicator of some general social conditions, which makes it possible to gauge levels of household, family or class distress by increases or decreases in unpaid work, as exemplified in the observations on the 61st round data. However, the employment situation or opportunities for women in the economy and society, and related possibilities for economic independence or empowerment (or lack of them) can be better measured by focusing on paid employment. Such measurement becomes all the more important and necessary in the present time, given the depth of commercial penetration in all forms of production and the connections between even subsistence production and the money economy. That incomeearning or paid work is not separately classified in the published NSS reports on employment and unemployment has no doubt discouraged the wider use of its measurement and analyses of trends in the data. So, even among those who regularly deal with employment data, too few realise how small the share is of incomeearning women in the workforce and how large it is of women who are tied down by absolute dependence on their households or families, without the potential or opportunity to free themselves.

Figure 4 presents the WPRs after excluding unpaid workers among women and men in rural and urban areas in the four survey rounds from 1993-94 to 2007-08. For easy comparison, Figure 5 presents the commonly used standard WPRs in the same rounds, that is, without excluding unpaid labour. Sharper differences are seen between rural and urban areas in the case of women as well as men when what we have termed the paid worker participation rate (PWPR) and the standard WPR are compared. The high incidence of unpaid labour of both women and men in the rural economy is largely explained by the family labour aspect of peasant farming. However, the expanding gender differentiation in paid work since 1999-2000 needs more explanation. The proportion of paid workers in the rural female population shows a declining trend from 1993-94 to 2007-08, which is not consistent with the up and down features of the standard FWPR. Among rural males, on the other hand, the proportion of paid workers increased between 1999-2000 and 2007-08, consistent with the trend in the standard male WPR. This would suggest that

Figure 4: Usual (PS+SS) Paid Work Participation Rates

WPRs excluding unpaid workers

Female PWPR (rural) Female PWPR (urban) Male PWPR (rural) Male PWPR (urban) 1993-94 1999-2000 2004-05 2007-08 18.8 17.3 17.1 16.3 12.3 11.0 12.7 11.3 45.7 44.9 45.6 46.7 47.8 47.8 50.2 51.3

Source: Unit-level data, various rounds, NSSO.

Figure 5: Usual (PS+SS) Work Participation Rates

WPRs including unpaid workers 55.4

Rural female Urban female Rural male Urban male 1993-94 1999-2000 2004-05 2007-08 32.8 29.9 32.7 28.9 15.5 13.9 16.6 13.8 55.3 53.1 54.6 54.8 52.1 51.8 54.9

rural women are either more severely constrained from taking up paid work or facing more direct discrimination in the paid job market. Either way, greater attention needs to be paid to reversing the socially regressive trend of declining rates of paid employment for rural women.

We might add that the 2009-10 survey shows a marginal drop in the standard usual male WPR (principal status and subsidiary status) for rural India between 2007-08 and 2009-10, from 54.8% to 54.7%, though no such drop is visible across the unpunctuated quinquennium from 2004-05. For rural women, the drop in the standard WPR is of far greater magnitude, from 28.9% in 2007-08 to 26.1% in 2009-10, and is more sharp across the quinquennium from 2004-05. We expect that when unit data for 2009-10 becomes available it will show that the relatively sharp drop in the standard FWPR is reflected in a further widening of the gender gap in rural paid WPRs.

In urban India, the gap between men and women in both paid work participation as well as usual work participation is substantially more than in rural areas. Among urban males, there were significant increases in both the PWPR as well as the usual WPR between 1999-2000 and 2007-08, although, surprisingly, the proportion of unpaid male workers appears to be relatively higher than among women. The picture for urban females, on the other hand, shows a continuation of very low levels of work participation by both measures. But the proportion of unpaid women workers is substantially lower than in rural areas. The 2007-08 survey shows a worrisome fall in both the PWPR and the usual WPR for urban women, underlining that the most distinctive period of corporate-led high growth in the country has yielded negative results even for urban women’s employment.12 (The link between the nature of such high growth and falling rates of employment for both rural and urban women needs to be explored further than is possible here.) It is striking that if one considers only paid work, the wide gap between the FWPRs in rural and urban areas narrows down. Nevertheless, it remains that while the urbanisation process has afforded increasing employment opportunities for men, the same does not appear to hold true for women.

The slowing down of the economy between 2007-08 and 2009-10 seems to have led to a reduction in the usual urban male WPR (PS+SS) by 1.1% (from 55.4% in 2007-08 to 54.3% in 2009-10), although the reduction across the quinquennial period from 2004-05 is negligible. The employment situation of urban women, however, appears to have been impervious to the fall in GDP growth between 2007-08 and 2009-10 with the usual WPR remaining the same in both the survey rounds. It is more than likely that the decline in the PWPR among urban females seen in the 2007-08 survey will be reflected in 2009-10.

Industrial Distribution of Workforce

The numbers of usual workers in India across broad industrial categories, estimated on the basis of census projections of male and female populations in rural and urban areas, for the NSS survey years from 1993-94 to 2009-10 (along with the percentage shares of the various industrial categories) are given in Table 2 (p 123). Apart from the four quinquennial surveys, the 2007-08 mid-point survey has also been included to get a more nuanced picture of recent developments, as well as for inferring trends in paid and

Source: Employment and Unemployment Reports, various rounds, NSSO.

october 22, 2011 vol xlvi no 43

paid employment to 2007-08.

It may be emphasised that a widening

female workers is evident in the paid workforce as well as in the standard estimate of the UPSS workforce. Yet, the fall in the number of paid women workers between 2004-05 and 2007-08 is not as

dramatic as in the standard estimate.

As the tables show, a relatively high

concentration of the female workforce

Table 3: All-India Industrial Distribution of Workforce (Rural + Urban) Usual (PS + SS) Excluding Unpaid Helpers (1993-94 to 2007-08, in 000s)

in agriculture has remained a feature of both the standard UPSS workforce as well as that of only paid workers. However, while the number of male workers in agriculture has been increasing across each of the quinquennial surveys, the female workforce as a whole has seen a reduction of some 21 million between 2004-05 and 2009-10, more than 20 million of it from agriculture.13 Moreover, the share of agriculture in the standard male workforce, which appeared to be rapidly declining between 1993-94 and 2004-05, was arrested to a significant degree in the last quinquen-

(100.0)

nium. But this was precisely the period when the pace of decline of the share of the female workforce in agriculture a lmost doubled.14 Of course, a major part of the reduction (as per the standard estimate) represents a withdrawal of unpaid family helpers from the workforce, as indicated by a comparison with paid women workers. We have argued in this paper that paid work is a better

Figures in parentheses are the sectoral shares in total employment.

measure of employment opportunities

Source: Unit-level data, various rounds, NSSO.

for women and the withdrawal of such

Table 4: Female Share of Paid and UPSS Workforce across Industrial Categories (Rural + Urban)

Industry Female Share of Paid Workforce Female Share of Standard UPSS Workforce 1993-94 1999-2000 2004-05 2007-08 1993-94 1999-2000 2004-05 2007-08 2009-10 the agricultural workforce without a

Agriculture 31.5 30.8 30.7 28.7 39.4 38.8 41.6 38.5 36.2 concomitant increase in their overall

Mining and quarrying 17.7 14.5 14.7 15.5 17.7 16.2 15.5 16.2 12.7 number in paid work cannot be seen as Manufacturing 23.4 21.0 25.2 22.0 28.9 28.2 30.9 26.3 27.5

Electricity, water and the like 5.6 3.8 4.7 3.3 6.1 4.1 4.8 3.4 11.5

Construction 12.8 11.1 10.5 10.6 13.3 11.3 10.5 10.7 14.9

Trade, hotels and restaurant 9.7 9.1 7.6 7.2 13.7 12.7 12.3 10.9 11.1

Transport, storage and communications 2.5 2.7 2.5 2.6 2.6 3.0 2.9 2.9 2.6 share of output as their own contribu-

Other services 22.7 25.1 29.5 27.5 24.1 26.7 30.3 28.3 28.4 tion to production.

All workers 25.6 24.0 24.0 22.4 32.5 31.0 32.5 29.4 28.0 While very little movement took

place in the number of paid women unpaid labour for which, as mentioned earlier, we have data only workers in agriculture between 1993-94 and 2004-05 (less than up to 2007-08. Table 3 presents the estimated numbers of workers 1,00,000 in each quinquennium), a sudden jump of 5,00,000 is after excluding unpaid helpers across the three quinquennial noticeable between 2004-05 and 2007-08. Since the share of rounds to 2004-05 and also for 2007-08. Table 4 gives the share casual workers in the paid workforce in rural areas also increased in of women workers in standard UPSS estimates for each of the this period, it would appear that while a substantial section of

Economic & Political Weekly

EPW

october 22, 2011 vol xlvi no 43

unpaid female workers was being withdrawn from the workforce, a significant number was also turning to casual labour. It is also possible that the Mahatma Gandhi National Rural Employment Guarantee Scheme (MGNREGS) contributed to the surge in the number of paid women workers in agriculture.

The industry that has probably absorbed the largest numbers of workers moving out of agriculture is construction, whose share in the male workforce equalled that of manufacturing in 2009-10. Among female workers, manufacturing seems to have held its ground as the second largest employer after agriculture despite having ejected some 3.5 million women (according to the standard estimate) between 2004-05 and 2009-10. Nevertheless, a jump in the number of women in construction undoubtedly occurred after 2004-05 (by more than 3 million according to standard workforce estimates), taking the share of women in the sector to a two-decade high of 15% in 2009-10. This was in contrast to the period from 1993-94 to 2004-05 when there was an observable process of masculinisation in the construction sector, which was confirmed by a decline in the share of women in the industry (Table 4).

It is noticeable that there is very little difference in the numbers of male and female construction workers whether one considers paid workers alone or the standard estimate of workers in all the survey rounds till 2007-08, a feature that is likely to hold good for 2009-10 as well. It would therefore appear that the earlier decline in the female share of construction jobs is now being reversed. Again, it is possible that the MGNREGS has played a role here. There are, however, limits to how much this upward swing in women’s employment in construction can mitigate the broader picture of reduced employment for women in India. Further, with a contractor-based circular migration regime now dominating the construction labour market, the movement from agriculture to construction for survival is often of a temporary and partial nature, particularly for women.15

The only other industry where women’s share has shown an increase is electricity, water, and the like. Here an inexplicable jump in the number and proportion of women seems to have taken place between 2007-08 and 2009-10, although even after the increase the industry accounts for a negligible 0.1% of the female workforce. In every other major industry, according to the standard measure, women’s share of employment fell in 2009-10 in comparison to 2004-05. The most significant fall was in manufacturing, where the number of female workers declined by some 20% between 2004-05 and 2009-10. In the same period, the number of women working in trade, hotels and restaurants fell by 10%. Even transport, storage and communications and, surprisingly, other services saw a reduction in the number of women, although to a lesser degree.

A comparison of the figures for only paid workers and standard estimates of the workforce in manufacturing gives us some insight into changes in the sector and its gendered impact. The recent drop in the number of women in manufacturing has been largely due to a decline in unpaid work. The fallout of the 2007-08 global financial crisis and the closing down of many traditional household industries where women were a part of the family labour seem to have resulted in their being pushed out of the workforce. Further, the decline in the number of unpaid women in manufacturing may indicate a contraction in subcontracting part processes to home-based workers, which appeared to offer some employment to women in an earlier phase of the expansion of modern but unorganised small and medium industries.

At an overall level, a comparison of both standard and paid measures of employment by broad industrial categories for 2007-08 indicates that the decline in the number of women in the standard UPSS estimate of employment in 2009-10 is likely to be reflected in women’s paid employment as well. While the dramatic fall in the estimated number of female workers according to the standard measure has been primarily driven by a withdrawal of large numbers from agriculture and manufacturing, estimates of the number of paid workers show a slightly different picture. The decrease in the number of paid women workers may not be as large, but it remains true that apart from agriculture and construction, in every other major industry there appears to have been a reduction in the absolute numbers of paid women workers. Thus, whichever way one looks at the employment data, the first decade of the 21th century has ended in a grim situation for women’s employment, and further marginalisation of women in the economy.

Advancing the Debate on Counting Women’s Work

Concern about the marginalisation of women in employment has engaged those involved in women’s studies since its inception in the 1970s. In the 1960s, before the advent of specialised women’s studies, declines in female work participation had been noticed, but generally thought to be transitory phenomena as the economy moved from subsistence agriculture and household industry to modern industry.16 It was also assumed that the decline would be automatically corrected by a “countervailing force”, identified with educational progress and substantial increases in income (Nath 1968). Analysis of historical trends in developed countries, particularly in Europe, had given rise to a developmental view which held that although income increases initially aggravated the downward trend in female work participation, further gains in income reversed the decline (Sinha 1967). Such assumptions and perspectives were, of course, proven to be erroneous in the case of India on several counts.

Following the report of the Committee on the Status of Women in India (GOI 1974), the long-term overall decline, both in percentage of workers to total female population and in their percentage to the total labour force, began to be viewed as evidence that women’s economic participation had been adversely affected by the nature of the development process itself. The CSWI postulated that ruination of domestic industry (deindustrialisation) from the colonial period onwards (with relatively greater ruin in the female labour-intensive sectors) had eroded the non-agricultural occupations of women, while the externally induced process of limited modernisation had excluded them from the limited opportunities in the modern sector. The result was a permanent shift of women to the periphery of the economy (GOI 1974; Banerjee 1998).

From the latter half of the 1970s and through the 1980s, studies on the WPR attempted to further explain the processes underpinning

october 22, 2011 vol xlvi no 43

the decline in female employment in India, while simultaneously focusing on the question of actual under-enumeration or undercounting of women’s work in national data sets. There was sufficient observable evidence that the vast majority of women were not idle but actively contributing to their household economy. The question therefore arose as to why and how their work was not being counted. Considerable attention was paid to problems in measurement of the extent of decline in the FWPR, especially in the context of changes in the definitions of workers across censuses and the ensuing problems of comparability.17

At a more conceptual level, the need to evolve alternative methods of capturing women’s work and the inadequacy of the existing modes of data collection on women’s work were linked to the invisibility of a large part of women’s productive labour and its contribution to family subsistence in both the non-monetised and monetised segments of the economy.18 The two sectors where such invisibility prevailed to a great extent were agriculture and household industry, and two decades of discussion and debate contributed to generating a greater awareness of the need to make women’s work more visible in the macro data, particularly in agriculture. Questions were occasionally raised about whether domestic work within families and productive work (economic activities that were also often within families) could or should be separated (Jain 1985; Sen 1985). On this, the separation between economic and non-economic activities has held on quite firmly.

The census as the primary source of historical trends was the database at the heart of these initial discussions and it is only from the 1990s that the quinquennial NSSO surveys became more central. In the 1990s, there was an understandable eagerness to assess the impact of the structural shift in the policy framework towards liberalisation after 1991 without having to wait for the 2001 Census reports. Given publication delays and the limitations of the published economic data of the 2001 Census, the NSS, which is thought to be more amenable to capturing women’s economic activities, has continued as the principal data source for most of the studies and analyses of women’s work and employment trends even after 2001.

A continued debate on the invisibility of women’s work, including issues of definitional limitations in data sources, was in a sense circumscribed during the 1990s by a wave of discussions on the informal sector that came into its own in the 1980s. Arguably, it could be said that the new preoccupation with the informal sector shifted the focus from the classical developmental dualistic approach comprising agriculture (as the backward subsistence sector) and industry (as the modern sector)19 to a more diffused set of occupations and employments that could not always be catalogued and categorised within the dualistic framework.20 While there were clear failings in the formulaic assumptions of traditional developmental approaches, which were brought out by empirical evidence on informal and unorganised petty production in urban areas and then on shrinking organised sector employment, the discussions on the informal sector contributed little to the great question of agrarian social relations. In relation to women, home-based work became a primary point of discussion that united the earlier discourses on the invisibility of women’s work and the later preoccupation with the informal sector (NCEUS 2007).

Economic & Political Weekly

EPW

Further evidence of the uncounted labour of women in economic activities came from a 1998 time use survey conducted by the Central Statistics Office (CSO) in some states, which showed that the WPRs for women as per the time use data was 58.2% and 30.9% for rural and urban areas respectively against 25.3% and 12.8% according to current weekly status in the closest NSSO employment survey.21 The time use survey was used to press a case for computing a value to women’s unpaid labour on the grounds of giving it visibility in official statistics and providing a basis for unpaid workers to claim their due share from the state exchequer to improve their conditions. Implicit to the valuation argument was a critique of the GDP in its present form as “a limited variable to project the status of well-being in an economy” (Hirway 2005).

As becomes clear from this quick survey, invisibility issues have dominated discussions on women’s work in India’s macrodata sets, with the brief digression towards the feminisation of labour thesis in the 1990s faltering in the face of empirical evidence. While there can be little doubt about the evidence that shows a significant part of women’s economic activities, labour or work does not get recorded, in this paper, we argue that even less attention has been paid to specifying and counting paid employment among women and that there is a need to do so. We are well aware of the limitations of such a measure. Women’s entry to paid employment is not necessarily an indicator of well-being and may even indicate pauperisation, impoverishment and greater levels of vulnerability for many of those doing so. In addition, without supportive social institutions that loosen women’s bonds to care and domestic responsibilities, their entry to paid employment exacts a heavy price through the double burden of work inside and outside the home. Nevertheless, the present commonly used method of lumping paid and unpaid workers together without differentiation is hardly the best method for analysis of employment trends, particularly among women.

Changing Role of Work and Employment

The separation of paid and unpaid work at the macro level, while providing a better understanding of trends in women’s employment, also directs attention to the changing role of work and employment in shaping some of the qualitative changes taking place in gender relations. There is no disputing that money incomes have become more significant and necessary in meeting even subsistence consumption, making for a differentiation between the typically income-earning male members and the unpaid female workers in households. Such differentiation naturally alters the nature of cooperation or interdependence and conflict and power in families or households, particularly of the labouring poor. With relatively less-diversified women being more concentrated in petty production functions in the form of unpaid labour (as distinct from a situation where both sexes were working together in a common production process), qualitative changes and shifts in gender relations within families as well as in society at large are inevitable.

On a more general level, it is worth noting that even with all the hype about expanding opportunities for women’s employment, so much so that it has become a kind of common sense, particularly among the educated classes, the shocking reality is that if one removes unpaid labour from the work participation count, in 2007-08, only 15% of the female population in the country fully dependent on others. Given the further decline in the FWPR received wages or incomes for their labour. The corresponding in 2009-10, this extreme situation of financial dependence is figure for males was 51%. In other words, when GDP growth rates likely to have worsened. The scale of such economic and financial reached an all-time high, 85% of the female population had no dependence is perhaps the most significant factor in the continuemployment or income and was economically and financially ing subordinate status of women in Indian society.

Notes

1 Workforce figures in this paper have been calculated using census segment-wise population projections and NSS segment-wise worker population ratios.

2 It may be remembered that it was from the same 2004-05 NSSO survey round that the NCEUS calculated that 77% of India’s population was living on less than Rs 20 a day (NCEUS 2007).

3 The compulsion for substituting paid work with the unpaid labour of women between 1999-2000 and 2004-05 was influenced by the pervasive agrarian crisis, which was marked by rising input costs and declining returns in peasant agriculture that appeared to be prompting a section of male workers and peasants to move out of agriculture. The supplementary incomes of poor peasants were also affected by the non-availability or collapse of casual work in rural areas and falling real wages for casual work in urban areas.

4 As is well known, India’s growth story picked up from 2003-04, pushing GDP growth to an average of 8.8% per annum till 2007-08 compared to an average of 5.4% in the preceding six years. Even following a slowdown after 2007-08, GDP growth in 2009-10 stood at a high 7.44%.

5 At the time of writing, unit-level data (without which the proportion of unpaid helpers cannot be calculated) was not available for 2009-10.

6 For these piece-rated workers, the nomenclature of own account worker can be somewhat misleading. Piece wages are paid for work that is farmed out to workers by entrepreneurs, traders and other employers, and the workers themselves actually work for wages, albeit mostly on piece rates. For example, the hundreds of thousands of women rolling beedis at home for beedi magnates or their contractors are clearly not really working in own account enterprises, but they are counted as own account workers in the NSS.

7 Among rural males, casual labour accounts for 39% of the paid male workforce and self-employment for a little over 50%.

8 Even in 1993-94 and 1999-2000, the share of casual labour in the paid female workforce in rural areas stood at 68.4% and 68.9% respectively.

9 The number of female own account workers plus employers in rural India dropped from 1,89,64,700 in 2004-05 to 1,67,34,800 in 2007-08, while female casual labour increased from 38,562,000 to 42,581,400.

10 Gleaning the number and proportion of domestic workers from industrial categories (National Industrial Classification; NIC) had emerged as the conceptually most appropriate method a principal researcher on employment trends for domestic workers in India could adopt.

11 Under the revised NIC of 2004, followed by the survey for 2007-08 (as also 2009-10), we can no longer make any clear distinction, for example, between security guards in private households (a rapidly expanding segment of male employment in urban areas) and domestic workers (an increasingly feminised occupation). It is obvious then that the data is failing us. Thus, perhaps an equally important problem is the official statisticians’ disconnect from the discussions on labour issues (in this case in relation to paid domestic workers) and their apparent indifference to the need to develop tools for a better understanding of this lowend service-sector employment.

12 Since 2003-04 and till the global crisis erupted, the pace of aggregate growth of India’s economy moved to levels unprecedented in its history, driven by a sharp rise in investment levels, led by the private corporate sector (see Mazumdar 2010).

126

october 22, 2011

References

Banerjee, Nirmala (1998): “Whatever Happened to the Dreams of Modernity? The Nehruvian Era and Woman’s Position”, Economic & Political Weekly, Vol 33, No 17.

Duvvury, Nata (1989): “Women in Agriculture: A Review of the Indian Literature”, Economic & Political Weekly, Vol 24, No 43.

Ghosh, Jayati (2009): Never Done and Poorly Paid: Women’s Work in Globalising India (New Delhi: Women Unlimited).

Government of India (1974): Towards Equality: Report of the Committee on the Status of Women in India (New Delhi: Ministry of Education and Social Welfare).

Hirway, Indira (2005): “Measurements Based on Time Use Statistics: Some Issues”, paper presented at a seminar on “Unpaid Work and Economy: Gender, Poverty and Millennium Development Goal”, 1-3 October, Levy Economics Institute, New York.

Jain, Devaki (1985): “Domestic Work: Its Implication for Enumeration on Women’s Work and Employment” in K Saradamoni (ed.), Women, Work and Society, Indian Statistical Institute, Calcutta.

Krishnaraj, Maithreyi (1985): “Review of Indian Studies on Women and Work” in K Saradamoni (ed.), Women, Work and Society, Indian Statistical Institute, Calcutta.

Mazumdar, Indrani (2007): Women Workers and Globalisation: Emergent Contradictions in India (Kolkata: Stree).

– (2008): “Women’s Unpaid Labour in Neo-Liberal India”, Indian Historical Review, Vol 35, No 2.

Mazumdar, Surajit (2010): “On the Sustainability of India’s Non-inclusive High Growth”, ISID Working Paper, 12.

Nath, Kamala (1968): “Women in the Working Force in India”, Economic & Political Weekly, Vol 3, No 31.

National Commission for Enterprises in the Unorganised Sector (NCEUS) (2007): Report on the Conditions of Work and Promotion of Livelihoods in the Unorganised Sector (New Delhi: NCEUS).

National Sample Survey Office (NSSO) (2011): “Key Indicators of Employment and Unemployment in India 2009-10”, Ministry of Statistics and Programme Implementation, Kolkata, New Delhi.

– (various rounds): Employment and Unemployment Survey Reports, 1993-94, 1999-2000, 2004-05, 2007-08 and 2009-10, Ministry of Statistics and Programme Implementation, Kolkata, New Delhi.

Neetha, N (2009a): “Feminisation or Segregation? Insights from a Disaggregate Analysis of NSSO Data” in Bhaswati Das and Vimal Khawas (ed.), Gender Issues in Development: Concerns for the 21st Century (New Delhi: Rawat Publications).

– (2009b): “Contours of Domestic Service: Characteristics, Work Relations and Regulation”, Indian Journal of Labour Economics, Vol 52, No 3.

Neetha, N and Indrani Mazumdar (2006): “Globalisation and Women’s Work: Trends in Women’s Unpaid Labour in India”, paper presented at a GADNET/CWDS seminar on “A World in Transition: New Challenges for Gender Justice”, 13-15 December, Centre for Women’s Development Studies, New Delhi.

Sen, Gita and Chiranjib Sen (1985): “Women’s Domestic Work and Economic Activity”, Results from the National Sample Survey, Economic & Political Weekly, Vol 20, No 17.

Sinha, J N (1967): “How to Use 1961 Census Working Force Data: A Comment”, Economic & Political Weekly, Vol 2, No 15.

vol xlvi no 43