Work Participation of Rural Women in Third World: Some Data and Conceptual Biases

Author(s): Bina Agarwal

Source: Economic and Political Weekly, Vol. 20, No. 51/52 (Dec. 21-28, 1985), pp. A155-
A157+A159-A161+A163-A164
Published by: Economic and Political Weekly
Stable URL: https://www.jstor.org/stable/4375146
Accessed: 17-01-2020 12:03 UTC

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide
range of content in a trusted digital archive. We use information technology and tools to increase productivity and
facilitate new forms of scholarship. For more information about JSTOR, please contact [email protected].

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
https://about.jstor.org/terms

Economic and Political Weekly is collaborating with JSTOR to digitize, preserve and extend
access to Economic and Political Weekly

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
Work Participation of Rural Women in Third World
Some Data and Conceptual Biases
Bina Agarwal
There continues to be little appreciation that problems ojfunemployment, poverty and destitution ar
instances gender-specific so that any serious attempt to alleviate these conditions and/or prevent their
aggravation would require a particular focus on the women of poor households. The accuracy of national level
statistics, which usually serve as the principal data input in the framing of development policies, is severely impaired
by biases which lead to an undercounting of women, both as workers and as those available for work. This paper
seeks to spell out the nature and sources of these data biases and attempts to provide pointers on hiow they could
be corrected and some of the data gaps filled.

I intra-household division of labour and in- 1977; Agarwal, 198 l,

come between women and men, or of Secondly, in many cultures, especially in
Introduction
women's independent need for employment a number of African countries, the culti-
PROBLEMS of unemployment, poverty and or other income-earning opportunities. vators are primarily women. Yet this is
destitution have long been the stated con- These pointers relate in particular to the seldom given recognition when programmes
cerns of development policy in most Third aspects spelt out below. to develop agriculture are planned. Usually,
World countries. Yet there continues to be Firstly. the evidence indicates that the in- information on new inputs and training in
little appreciation that, in many instances, terests of women and men belonging to the new practices is provided to the menalone:
these problems are also gender-specific; and same household do not always coincide and, "In the African way ... we speak to the man
that any serious attempt to alleviate these in some situations, are even in conflict. who is the head of the house and assume
conditions and/or prevent their further ag- Hence when schemes are aimed at the men he will pass on the information to other
gravation, would require a particular focus
of a household, the benelits cannot auto- members" [Staudt, 1976: 91; also see
on the women of poor households. The in- matically be assumed to reach the women; Boserup, 1970]. Land titles too are granted
sights provided in this regard by micro- the programmes may even leave the women typically to the men, which limits women's
studies, have received little serious attention. worse off. For example, in the African con- access to credit [Staudt, 1976]. Instances have
At the same time, the accuracy of national- even been noted where agricultural equip-
text, where typically women are responsible
level statistics, which usually serve as the ment has been designed with men in mind
for cultivating the subsistence food crops,
principal data input in the framing of deve- although women are the users [Ahmcd,
the promotion of new rural earning-schemes
lopment policies, is severely impaired by
(cash cropping, poultry-keeping, etc) for the 1978]. As a result, the women continue
biases which lead to an undercounting of men, by national and international agencies, to cultivate with old, low productivity
women, both as workers and as those have in a number of instances adversely techniques.
available for work. There are also lacunae affected the household women. This is Thirdly, in a large number of households,
in data coverage. As a result, in their very because, the women, while continuing to be women are the bread-winners. Although
conceptualisation many of the schemes in- data on this are fragmentary, a fair indica-
solely responsible for producing the family's
stituted to help the poor are misdirected. tion of magnitudes is provided by recent at-
subsistence food needs, also have to con-
This paper seeks to spell out the nature tempts to collate existing information:
tribute obligatory labour to their husbands'
and sources of these data biases. It also at- Buvinic and Youssef [1978] for example,
new activities [Boserup, 1970; Bukh, 1979;
tempts to provide pointers on how they could Hanger and Moris, 1973; Mbilinyi, 1972; estimate that 15-25 per cent of the house-
be corrected and some of the data gaps fil- UNECA, 1972; Wilde, 1967]. However, any holds in many Third World countrieS are
led. The discussion centres on data relating extra cash income so generated accrues likely to be headed by women: 20 per cent
to the rural areas of the non-socialist Third typically to the men alone [Mbilinyi, 1972; on average in the Caribbean and Central
World. The paper is divided into six sections.Palmer, 1977; Wilde, 1967; Bukh, 1979]. And America, 22 per cent in sub-Saharan Africa,
Section I1, which follows, briefly spells out such money tends to be spent by the men 16 per cent in North Africa and the Middle
why thqre needs to be a focus on women largely for their personal needs such as East, 15 per cent in South America and 18.7
separate from men, in planning rural per cent in India. The incidence of poverty
drink, clothes, radios, inputs for their fields,
employment and income-generating sche- etc, with relatively little flowing to the is noted to be much higher among such
mes; and touches on how inadequate and in- women [Bukh, 1979; Hanger and Moris, families than those headed by men. Thert
accurate, gender-related information can 1973; Consortium for International Deve- are indications too that the number of such
lead to schemes being misdirected. In Sec lopment, 1978]. In some instances, women's households is increasing [Birdsall and
tion III the biases, especially in census in- income and consumption are even noted to McGreeley, 1978]. Further, in recent years,
formation, are traced. Section IV deals with have decreased in absolute terms, as they when Food-For-Work (FFW) programmes
some of the inadequacies in data coverage. now have less time and energy to devote to have been initiated in parts of South Asia,
Section V points to some conceptual biases their - independent productive activities many of the women who have sought work
in data use; and Section VI addresses itself are those with dependents, and with little or
[Hanger and Moris, 1973; Palmer, 1977;
to the question what can be done? Bukh, 1979]. In the Asian context also, no financial support from male relatives.
evidence from a number of countries sug- One survey covering 303 female worker
II
gests that the degree to which women from families in a Bangladesh FFW programme,
Need for a Focus on Women in poor rural households gain from a rise in found that 33 per cent were headed by
household income is affected by their extent widows, and 14 per cent by divorced and
Development Schemes and Data
of direct control over such earnings. The deserted women. Almost half the women
Collection were the sole income-earners of their
items on which the household cash is spent,
There are now a fairly large number of for example, is once again noted to vary families,
by and all but seven had dependents
micro-studies that point to the dangers of gender, with the men usually spending a part [Chen and Ghuznavi, 1978, Appendix Tables
State-sponsored schemes which seek to raise on liquor and cigarettes for themselves, and 4 and 10].
the welfare of poor households, based on lit- the women usually spending on family In this context, a lack of information on
tle knowledge or appreciation either of the necessities alone [Arens and van Beurden, the incidence and nature of unemploymdnt

- Economic and Political Weekly, Vol XX, No 51 and 52

Review of Agriculture, December 21-28, 1985 A-l55

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
 Review of Agriculture December 1985 ECONOMIC AND POLITICAL WEEKLY

among women, especially as it relates to mates of women's work participation that where women's involvement in other than
poverty among women, clearly has signifi- are based on such data. domestic work is considered unprestigious,
cant implications. Employment projects for In India, the decennial censuses constitute this leads to an under-estimation of women's
the poor, for instance, would be misdirected one of the.two main official sources of data work participation in non-domestic work,
where those needing work are women heads on rural women's work participation; the and of her 'availability' for such work,
of households, but the projects in their work other is the National Sample Survey (espe- especially outside the home. Youssef [1974:
Content, location and period of work avail- cially, the 27th and 32nd rounds). Both 12], for instance, notes from survey evidence
ability take no account of this. Misinforma- sources, however, are inadequate on a relating to muslim countries that ...
tion about the availability of women for number of counts. While a comprehensive farmers are generaHy reluctant to go on
employment may also take the form of not review of the statistics will not be attemp- record as having their wives and daughters
recruiting them when a project is started.' ted here, some of the specific biases which reported officially (i e, in census count) in
Undeniably, the filling in of data gaps or tend to creep in, in relation to women, will the labour market ..." Further, women's
the correction of data biases would be only be discussed in detail. economical proddctive work could be seen
a small step in reducing lacunae in scheme- To begin with, it is useful to address by the men as a part of housework and so
formulation. Inadequate information or ourselves to some basic and seemingly reported.
misinformation is not the sole cause of straightforward questions which need Typically, no attempt is made to seek out
misconceived income-generating schemes,answering,or and examine the problem of women respondents. This is partly but not
of the failure to pay attention to questions answering them on the basis of existing data entirely a reflection of the fact that census
of women's poverty and unemployment. The sources. For instance, consider the questions enumerators are usually imen (village revenue
gap between available data and their use in below: officials, school teachers, officials of local
policy can be just as critical as the gaps and (a) What proportion of rural women may bodies, etc). This limits access to women
biases in the data themselves. And the gap be counted as 'workers'? respondents, particularly in communities
between policy and implementation can be (b) What proportion of the women not so where female seclusion is. practised. Also,
even more critical. Nevertheless, it is impor- counted are 'available' for work? even if the women were questioned, their
tant to recognise that insofar as the existing (c) For what period are rural women, both responses could still be affected by the
data base serves as an input into policies in those included among the 'workers' and enumerator's gender.
many Third World countries, data inade- those not so included but 'willing' to It is of course true that in many instances,
quacies that permit unsubstantiated conjec- work, available for work? This is impor- women and men may share the same cultural
turing and the perpetuation of misconcep- tant to know in view of the seasonal perceptions regarding women's roles. The
tions regarding women's work roles and their nature of the work characterising agri- tendency of women to identify themselves
income and employment needs, can have culture. as 'only housewives', even when they are
serious repercussions on the effectiveness of (d) What are the characteristics of the economically active, reflects a common
economic resources allocated to develop- employed and 'involuntarily' unemp- cultural pattern in large parts of the Third
ment schemes. loyed women in terms of their status as World. Buvinic and Youssef [1978: 9]
Data shortcomings as they relate to issues household heads, their economic posi- observe how in a micro-survey of poor rural
concerning rural women and work, stem on tion (including tenure status), their age, households in Honduras, many of the
the one hand.from the method of data col- their education/skill levels and the women who declared themselves as house-
lection and the concepts and definitions number of their dependents? wives, were in fact the bread-winners: irt ad-
used, and on the other hand from the con- A count of those not working but 'available' dition to housework, they performed an
tent of the data collected. An additional pro-for work would be needed to measure the average of two economic activities for which
blem relates to conceptual biases in the use number of women 'involuntarily' unemploy-they received payment. Yet the question: "Do
of the data for analysis, as, for instance, in ed. Information on the period for which you work?" elicited the culturally accepted
women's work being discounted on a priori work is sought would give an idea of female response: "No, I'm a housewife". Similarly,
assumptions of their lower productivity under-employment. This information Deere [1977: 8] notes that Peruvian peasant
relative to men's. These data and conceptual together with that on some general charact- women declare their chief occupation as
problems are discussed in the sections that eristics of women workers and those 'avail- "their home'* because such an answer is con-
follow. For purposes of illustration, the able' would be necessary for formulating sidered "right and proper", irrespective of
discussion is based primarily on Indian data policies to provide income-earning oppor- the actual time they spend on agricultural
sources, although wherever possible, ex- tunities. The discussion that follows is' focus- activities. Here, in addition to the gender of
amples are drawn from other countries as ed on the nature and sources of biases in the the enumerator/respondent, the way ques-
well. Most of the issues highlighted have existing data which seriously affect their ac- tions are framed can be significant. The
general applicability and would be relevant curacy and usefulness. words used in formulating questions also ac-
when considering the use of such statistics quire importance. Part of the misreporting
in large parts of the Third World. (i) Respondent and Enumerator stems from semantic differences which
Biases underlie the interpretation of words such as
III 'work' 'housework' etc, [Abdullah and
These biases may be seen to stem from a
Biases in Data Collection variety of factors such as the gender of the Zeidenstein, 1976]. Unless questions are for-
One of the main sources of macro-data respondent and the enumerator, cultural mulated with sufficient care, the answer will,
on work participation in most countries is perceptions of women's roles, and the type reflect the woman's own assessment of her
the census. It is usually the primary and work, with a strong likelihood of her under-
of questions asked, including the words used
often the only means available for assessingto frame the questions. estimating the economically productive con-
and monitoring, over time, the impact of tent of that work, insofar as she herself con-
Usually, in India and in most parts of the
development policies on the pattern and level Third World, the person from whom infor- siders it a part of 'domestic duties' and
of male-female labour force participation. mation is sought and obtained is the male reports it as such.
It is also one of the most widely-used sources head of household or other male household As one recent survey of UN data on the
for cross-country comparisons of labour member. Hence when questions relating to position of women noted:
force participation rates. It, therefore, women's work status and. her availability for There is one major distorting device
becomes especially important to take ac- work are asked, the answers tend to reflect operating on all data collection concerning
count of the biases which affect the esti- a male perspective. In a cultural setting women, above and beyond interpretation dif-

A-156

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
ECONOMIC AND POLITICAL WEEKLY Review of Agriculture December 1985

ferences or collection facilities. This is a set worker' adopted in the census years, 1951, seasonal and task-specific nature of women's
of cultural assumptions about the secondary 1961, 1971, and discussed here, see Census work in agriculture would also have con-
importance of anything women do; it pro- of India, Centenary Monograph No 2, 1971: tributed to women being so classified under
duces under-registration of women from 637, 638, 657, 658, 684.) this definition. Men usually undertake a
birth to death, an under-enumeration of In the 1951 Indian Census, the population greater diversity of agricultural tasks spread
women in employment, independently of was divided into 'self-supporting persons', out over the year. Although a supplemen-
other forces that also create under-counting'earning- dependents' and 'non-earning tary question was asked as to whether a per-
of populations in general. [Boulding, et al, dependents'. Self-supporting persons were son had participated in any ' secondary
1976: 6]. those earning at least enough for their own -work', the addition of the count of female
Such cultural assumptions reinforce defini-
maintenance. Earning dependents were those 'secondary workers' so obtained to the
tional biases on what constitutes 'working', who secured an income which was regular female 'main workers' is still not found to
as discussed below.
but insufficient to fully support them. These provide an accurate estimate of rural female
(ii) Definitional Biases two categories comprised the 'workers'. participation rates in 1971 [Sinha, 1972,
Household members who jointly cultivated 1975].
In different Third World countries, cen- land and thus jointly earned the household In the latest 1981 Census, 'working' has
sus definitions tend to vary between two ex- income, were classified as workers. However,again been defined as participating in any
tremes, depending on the inclusion or exclu- for the women providing unpaid family economically productive activity. In this
sion of unpaid female family labour. On the *labour on the farm to be so classified, clearly Census, there is- an improvement over the
one hand there are countries such as Thrkey required that they be perceived as joint 1971 Census in that the initial sorting ques-
and.Thailand, where all women in agri- cultivators. For the reasons mentioned tion for separating workers from non-
cultural households are included in the earlier, there was a greater likelihood of their workers is: have you worked at any time at
labour force, under the assumption that both
being reported as housewives and thus be- all last year? The workers have then been
women and men in such households con- ing counted as non-earning dependents. divided into 'main' and 'marginal' depen-
tribute in some way or other to activities ding on whether or not they-have worked for
relating to agricultural production. In these
In censuses subsequent to 1951, the defini-
tion of 'worker' has shifted from the earn- the major part ( i e, over 183 days) of the
countries the agricultural labour force in-
ings criterion to one of 'gainful activity'. In previous year (Census of India, 1981). Once
cludes roughly the same number of women
1961, the persons classified as 'workers' were again, most women are likely to be categoris-
as men. In many North African and Latin
either seasonally employed who had done ed as 'marginal' workers by this criterion,
American countries, on the other hand, all in so far as they might be, or are reported
some regular work of more than one
farmers' wives are counted as housewives, to be, housewives for the major part of the
hour a day throughout the greater part of
and not included in estimates of the labour year. Nevertheless, the 1981 Census does pro-
the working season .' with or without
force unless they participate in the produc- vide greater scope than did the 1971 Census
remuneration, or were in 'regular' employ-
tion of goods and services for sale [Boserup, for counting women workers at least in the
ment and had been employed during any of
1973].
the fifteen days preceding the day of enume- 'marginal' category. The importance of what
As a result of these definitional differen-
ration. An adult woman engaged in house- the initial sorting question is also highlighted
ces, the impact of structural changes in the in these definitional comparisons.
hold duties but doing no other productive
economy on female participation rates, as,
work to augment the family's resources was The negative effect of definitions adopted,
for instance, the impact of urbanisation, will
considered a. non-worker. The 1961 Census has also been observed in the Censuses of
appear to differ greatly between the two sets
thus favoured the inclusion of women as other countries. Deere [1977], for instance,
of countries. In the first set, unless all the
workers, since the criterion for being notes that in the Peruvian Census, cultural
women of rural agricultural households who
seasonally employed was broad enough to norms are such that a man residing in the
migrate to the urban areas can readily find
capture many of the 'marginal' rural peasant household is 'automatically' con-
jobs, at least some will now be counted
workers. However, that the term 'productive sidered the main agriculturist, while peasant
among the non-workers, whereas earlier they
work' as applied to women, might be inter- women are at best regarded as 'helpers' or
were all counted as workers. Hence in such
preted as production for sale, was a strong unpaid family workers. Unpaid family
countries, femnale participation rates will
show a decline with urbanisation. In the
possibility, given a supplementary clarifi- workers are counted as being economically
catory clause relating specifically to women, active only if they satisfy a time criterion.
second set of countries in contrast, female
which defined an adult woman 'worker' as By Deere's estimates, in the 1972 Census, due
activity rates will show an increase over the
one who -.. . in addition to her household to this criterion some 80 per cent of the
initial low rates, as more women get paid
work, engaged herself in work such as rice women who considered agricultural or live-
jobs and can no longer be ignored as
workers. Census data may therefore show pounding for sale or wages, or in domestic stock activities as their main occupation
declining or increasing female participation
service for wages for others, or minding cat-were not counted as economically active.
rates, depending on which method of classi- tle or selling firewood, or making and sell-
A general under-enumeration of women's
fying unpaid family labour is used. Likewise,
ing cowdung cakes or grass, etc" [my em-
involvement in gainful activity can also oc-
inter-country comparisons of female activity phasis, Census of India, Centenary Mono- cur due to the overall fuzziness which sur-
rates can reflect spurious rather than real dif-
graph, 1971: 658].
rounds the demarcation of 'domestic' from
ferences in women's involvement in work. In the 1971 Census, the definition of 'productive' work. One recent study on
Some countries, however, fall in-between 'worker' was particularly, biased Sudan noted that only 10 per cent of the
against
these two extremes. These are countries the inclusion of women. Only those who Sudanese w6men were counted as econo-
which include among its workers those un- reported themselves as participating in mically active in 1967-68: the estimate had
paid family women who by some specified economically productive work as their main overlooked women's work in agriculture,
criteria can be termed as having participated activity (in terms of time spent on the task) animal husbandry and marketing, water and
in 'productive' work. However, the defini- were counted as workers. This meant that fuel carrying, grain grinding, etc [UNECA,
tions adopted can still be such as to have a women who"e main activity was domestic 1975]. Often women's involvement in the
disproportionately negative effect on the work or, due to the cultural biases mention- cleaning and grading of agricultural produce
enumeration of rural women workers. For ed, was reported as being domestic work, contributes to the value added of articles
example, consider the Indian case where the were counted as non-workers even though sold in the market, yet such work tends to
criteria for defining a 'worker' have varied they may well have contributed a significant be classified as 'domestic work'.
between censuses. (For the definitions of amount of labour time in the fields. The However, even if consensus could be

A-157

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC AND POLITICAL WEEKLY Review of Agriculture December 1985

reached regarding what constitutes 'domestic state and 46 pet cent to 27.37 per cent in the of women's availability for work depends
work', the fundamental bias would remain, latter (figures taken from Census of India, greatly on the attitude of the respondent to
namely the exclusion of this work from 1981) women's employment. Secondly, whether or
'gainful work'- and hence from national in- not a women is available for work can
Since :he effects of cultural norms on the
come and employment statistics. The incon- Census Uigures relating to women will tend depend crucially on where the work is pro-
sistency that this entails is seldom recognis- to vary between states, a. cross-sectional vided. The questiorn asked-are you avail-
ed. In the 1971 Indian Census instructions, anialysis for measuring the impact of inter- able for work?-would usually be interpreted
for instance, it is stated that ". a servant statc di ffe 'Xncc tr todernisation/urbanisa- as being available for work outside the
who work as a cook in his or her employer' tion. etc ont woicn, also presents a problem. home. Hence many women involved essen-
home for wages will be considered economi- tially in domestic work for most of the year
(iv) Measurement of Female
cally active, but a housewife, even if she may and who may be willing to undertake paid
work much more than a paid servant, in hav-
Unemployment work in the home but not outside, would not
Even more problcmatic, however, is the be counted among the unemployed and
ing to cook for -the family or looking after
measurement of fernzle unemployment. In would be outside the labour force even by
the household will not be treated as econo-
mically active.." [Census of India, Centenary the 1961, 1971 and i981 Indian Censuses, as the 'usual' status classification. Thirdly,
Monograph, 1971: 693]. The question about well as in the Indian Natioinal Sample Survey there would be women who are usually
the value of a housewife's work continues (NSS) data discussed further on, unemploy- employed (say during the busy season) but
to be part of an unresolved debate. ment is measured essentially in terms of a who 'withdraw' into dotnestic work when
person's javailability' for work. The question they cannot find work during certain periods
(iii) Changes in Definitions and of 'availability for work' becomes difficult of the year (say in the slack season) and who
Problems of Comparability to tackle since (a) it touches not merely on do not report themselves even as being
Perhaps the most obvious advantage of the number of people seeking work or available for work during such periods-
Census data is that changes over time can available for work but also on the period for "the discouraged drop-outs" [Bardhan,
be studied for the entire population. Defini- which they are available; and (b) the concent
1978]. While such women would get counted
tional changes in censuses reduce com- of availability is subject to specific biases 6? as workers under the 'usual' status classifica-
parability in general and tend to affect interpretation when applied,to women. In
tion, they are likely to be termed as engaged
statistics on female labour in particular. This the Indian Censuses unemployment among
in domestic work and outside the labour
is especially because the principal casualty rural women would have been under-
force under the 'current' status classification,
of definitional changes tends to be the estimated insofar as relatively few would
when the survey reference week falls in such
category of 'marginal workers' and rural have been reported as actively seeking work,
periods.3 This would not merely lead to a
women more often than rural men, as noted, partly because of the cultural bias (in repor- lower estimate of the female labour force by
tend to be so categorised. Time-series, inter- ting) mentioned earlier, and partly because
'current' status relatixe to 'usual' status, but
census extrapolation, therefore, becomes they would (in actuality) tend to fall back more crucially, it would lead to an overall
difficult and often suspect. on housework when they do not readily find underestimation of the effective female time-
For Indian data, we have already noted work, and be classified as housewives.
rate of unemployment which is based solely
that the 1961 and 1971 Censuses have been The Indian Censuses do not provide data
on 'current' status measurements.
rendered virtually non-comparable in terms on the time intensity of work effect-an
The latest 32nd NS$ round has gone some
of female labour force participation.2 aspect which in fact can prove to be of way forward in providing information which
Given that the actual changes in rural female primary importance for a meaningful
could correct the second bias mentioned, by
work participation between 1961 and 1971 measurement of rural unemployment in
including in the survey a set of probing ques-
cannot readily be separated from spurious Third World countries, where unemploy-
tions, to be put to women categorised as
changes due to definitional modifications, ment time rather than unemployed persons
engaged in domestic work, such as questions
the apparent enormous decline in this par- would be the significant factor. Some light
on the willingness to work if work were
ticipation between the two censuses (from is thrown on this aspect by the Indian NSS,
made available at the home, the nature of
31.42 per cent in 1961 to 13.44 per cent in which provides information both on employ-
work desired, etc. It is noteworthy that 23
1971, on an all-India basis) cannot be related ed/unemployed persons and on the intensity per cent of rural women categorised as
meaningfully to other economic/structural of work effect during the survey's reference
engaged in domestic work and hence
changes in the country. What is noteworthy period.
counted outside the labour force even by the
is that where matrilineal tribal communities In the 27th NSS round survey, for in-
'usual' status classification, reported a will-
dominate, as in Meghalaya and -Nagaland, stance, the classification of the labour force
ingness to accept work if such work were
even by the 1971 Census definitions, a high (people employed plus those unemployed available at their residence. [For details on
proportion of the rural female population viz, seeking or available for work) was
both the 27th and 32nd NSS rounds see Na-
is reported to be doing productive work as undertaken both according to 'usual' status
tional Sample Survey Organisation, 1981.]
its main activity, aind falls within the (i e, by main activity pursued during the past However, both the 27th and 32nd rounds
'worker' category-37.97 per cent in Megha- year or more and likely to continue in the suffer from the third bias mentioned. In this
laya and 47.90 per cent in Nagaland, rela- future) and 'current' status ( i e, by principal, -ontext, it has been suggested by one author
tive to below 20 per cent in most other states. time-wise activity pursued d'uring the week (who examined the 27th round) that for a
A part of the higher female participation in preceding the survey). The 'usual' status woman whose 'usual' status is that of a
1971 in these two states, vis-a-vis other states, classification gives an idea of the number 'worker', the days spent in domestic work
probably reflects the effect of cultural fac- of persons employed or unemployed, while during the survey reference week could be
tors on actual work participatibn, but a part the 'current' status classificationi also helpscounted as unemployed days even when she
would also be due to the more accurate measure the time-rate of employment and
has not reported as being 'available' for work
reporting of such work participation. unemployment,For in- [Bardhan, 1978]. Of course to the extent that
stance, states without a dominant matrilineal However, the asseszmaient of femrale un-some of the time spent on domestic work
comnmunity, such as Madhya Pradesh and
employment obtained from the survey, in may represent voluntary. withdrawal, this
Andhra Pradesh, which also reported higk terms of both the number of unemployed measure would somewhat over-estimate in-
rural female work participation in 1961, and their time-intensity of unemployment, voluntary unemployment. One might, how-
showed a considerable decline in 1971, from is likely to be biased for several reasons. ever, suggest that there is a base level of
48.60 per cent to 20.75 per cent in the formerFirstly, as in the census, the measurement domestic work beyond which women would

A-159

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
 Review of Agriculture December 1985 ECONOMIC AND POLITICAL WEFICLY

be available for work if there were ap- that 10 per cent of these households were Venezuelan Census, do the Census and
propriate work. The time devoted to domes-headed by women, 16 per cent of whom were micro-survey estimates tally. The identifica-
tic work during the busiest agricultural married, 79 per cent were divorced and 4 pertion of female household heads and their
season of the year could be one possible, cent were single. characteristics, particularly in terms of their
even if crude, measure of this. Buvinic and Youssef, as noted, have economic responsibilities and their economic
Some of the biases noted are those of sought to provide some estimates of likely means is clearly important for planning ap-
most research and researchers in this field female-headed households, by applying propriate
the income-generating schemes.
and are found even in micro-level, region-typology of such households gleaned from Accurate information is also necessary on
specific data. Sometimes cross-checking can micro-data to aggregate data from 74 non- the seasonal nature of agricultural work, and
help to pinpoint a bias. For instance, in socialist Third World countries. Using Cen-the seasonal fluctuations in the need for ad-
analysing farm survey data relating to a sam-sus information they identify the women ditional work. This is again a generally
ple of 99 farms in the Andhra Pradesh State who, because of their current marital/family neglected feature in macro-data collection.
of India, I found that 34 of the farms ac- status, are likely to become household heads. Even the Indian NSS data, which consider
tively used female family labour for crop While this does not give the exact number the time intensity of work effort, do not
production activities: the average annual of such women in these countries, it provides adequately capture this dimension. Any at-
labour hours per adult family woman total- an approximation of likely magnitudes and tempt at correction would also need to take
led 105 for all agricultural operations. Yet trends, a;nd reinforces the need for a more account of the possible divergence in the
only 6 of these 34 farms reported family reliable quantification of the incidence of seasonal pattern of men's and women's
women as being involved in agricultural such households. work, in that their periods of high and low
work- [Agarwal, 1981]. Deere [1977] similarly They also provide several examples, largely labour input and availability, may not
points to the difference between the results from Africa and Central America, of how match. Much would depehd on the task-
obtained when estimates of women's work Census data tend to under-enumerate specificity of female labour use on the
participation in Peruvian agriculture were female-headed households. They note that farms. A pre-dominance of female labour
based on an actual observation of activities in countries where male supremacy is the in the cultivation of certain crops such as
(through survey data), and those obtained norm, both the enumerators and the respon- paddy relative to say wheat, and in the per-
when census definitions were used to dents are pre-disposed to identify a man forming of certain operations such as sow-
separate the economically active from the rather than a woman as the household head. ing, weeding and harvesting relative to other
inactive. Even a boy of 12 could be cited or could cite operations such as ploughing (which is
himself as a household head in the-absence usually done almost exclusively by men),
IV
of a male adult! would tend to introduce differences between
Coverage Inadequacies In these circumstances, even when defini- men and women in their 'availability' at a
In addition to the noted biases, existing tions are somewhat more specific, the given point in time. As a result, attempts at
macro-data sources are severely limnited in chances are that men rather than women will
alleviating male unemployment in certain
their ability to provide details about the be identified. Consider the following defini-seasons (even insofar as such flexibility could
characteristics of employed and unemployedtions: In the Honduran Census, a household be introduced in work schemes) may not, at
women, such as their status as household head is "that person who provides the chief the same time, help reduce female un-
heads, the number of dependents they have, source of income for thea household or who employment.
their economic status and agrarian class, is regarded as such by its members" [my em-In this context, an important additional
their levels of skills/education and the phasis, Buvinic and Youssef, 1978: 8,9]. dimension is the possible difference between
periods for which they need work. Again in the Philippine Census the head is different socio-economic classes in the sexual
While the collection of data on family "a person-who provides the chief source of division of labour. The tendency of women
headship is an existing feature of many cen-income for the household unit; it is the eldest to withdraw from work in the fields, with
suses in Third World countries (as in India, person, male or female, who is responsible an improvement in tht economic positions
the Philippines, and a number of African for the organisation of the household or who of their households (provided there are other
countries), the information is often not is regarded as such by the members" [my em- earning members), as noted in parts of Asia
available irr published Census reports or, phasis,
if Buvinic and Youssef, 1978]. [Boserup, 1970; Epstein, 1973; Agarwal,
available, is distorted by cultural and defini-Sometimes the Census instructions them- 1984] as well as in Africa [Brausch, 1964],
tional biases. Also, as noted earlier, infor- selves are biased: The instructions forwould the be another aspect requiring considera-
mation gleaned from micro-studies, frag- 1951 Indian Census were as follows: "The tion in employment planning, and hence in
mentary though it might be, points to an in-, head of the household is the person on data gathering on rural women's work.
crease in female-headed households in the whom falls the chief responsibility for- the
V
Third World, as a result of male migration, maintenance of the household. You need not,
or of the dissolution of marriages, or of the however, make any enquiry about this and Some Conceptual Biases in
desertion of women by their husbands, such you should treat as the head of, the house- Data Une
that the women are left economically respon- hold any person who is actually acknowledg- In addition to the lacunae and shortcom-
sible for their own and their dependents' ed as such"(!) [my emphasis; Census of In- ings of available data, their actual use in
survival. dia, '1971: 636].
economic analysis reflects a male,bias. This
However, no macro-level estimates exist in The effect of these biases is clearly is particularly apparent when we consider
most cases. In the Indian Census, for exam- established through a comparison of Cen-the question of the weights to be given to
ple, data on family headship is collected but sus data with micro-studies. For example,female labour time vis-a-vis male labour
prior to the 1961 Census, the published Buvinic and Youssef noted that in the 1973time.
reports do not provide comprehensive esti- Costa Rican Census, only 16.4 per cent of In empirical studies dealing with aspects
mates. Specific sample studies give some all households were classified as headed by of total farm employment, an aggregation
limited idea of this. Dandekar and Unde's women. A sample study of 82 women drawn often carried out is the conversion of female
[1964] study for West Bengal, India, for in- from poor households from different partslabour time to 3/4 or V2 of male labour time
stance, noted on the basis of the 1951 Cen- of the country found that 62 per cent of the[for instance, Bishnoi, 1966; Gollas, 1970;
sus survey covering 248,511 households sample women were -de facto household Kahlon, et al, 1973; Norman, 1973; Sanghvi,
(rural plus urban), which formed 5 per cent heads with economic responsibility for the 1969; Sidhu, 1974], with labour time usual-
of the households in 12 districts of the state,family. Only in rare cases, such as in the ly being measured in hours or in standahl

A-160

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
ECONOMIC AND POLITICAL WEEKLY Review of Agriculture December 1985

8 hour days. The conversion is typically done and cotton there was found to be no use of in the agrarian economy is needed. For this,
on an a priori assumption that female male labour on most plots, so that no a detailed analysis of their work by activity
is bour is less efficient than male labour. satisfactory comparison could in fact have would be essential, to capture the seasonal,
Sometimes, the fact that women are general- been made. operation-wise, crop-specific variations in
ly paid 3/4 or 1/2 of what men are paid is used-Indeed, little systematic research has gone the work, as well as to make a more ap-
into actually comparing the relative efficien-
to justify the use of these values as conversion propriate and explicit distinction between
rates. Sanghvi, for instance, even while cy of men and women in given tasks. In this domestic and non-domestic work. Such arp
noting that in certain operations such as cot- overall research void, studies such as the one analysis would also be necessary for identit
ton picking, women are more efficient than conducted by the Government Potato Seed fying and measuring the likely impact (time
men, still converts women's labour time Farm in Mattewara (Punjab, India) during saving or using, income generating or reduc-
spent in this task to ?/4 of its value, since 1973-74
he to test the efficiency of different ing) of technological change and agri-
finds that women are paid ?/4 of what men; potato-digging equipment, stand out. In this cultural m'odernisation programmes on rural
(taken as the standard unit) are paid per study it was found that for the equipment women's work.
hour. The discounting of women's labour tested, women were over three times as effi- In this context, the possible differential ef-
time vis-a-vis male labour time on the basis cient as the men. (see Table). fect of technological innovations on women
of differential wage rates is also common This study while by no means conclusive, belonging to different socio-economic
in the frequently-used reports of 'Studies in is nevertheless indicative, and underlines the classes needs to be kept in mind, and infor-
the Economics of Farm Management' in need for a requestioning of hitherto a priori mation obtained by class divisions: certain
India [see examples given in Malya, 1979]. assumptions and for more detailed research. innovations may reduce the work burden of
This is clearly a questionable rationale, women in higher income households, and
VI at the same time also reduce the income-
since it ignores the fact that differences in
wages paid are not necessarily indicative of What Can Be Done? earning capacity of the poorest women. An
differences in productivity. The employer's example of this is the introduction of rice
The above review clearly points to the
own bias and his ability to command female mills in parts of Asia, which have saved
need for taking corrective measures in the
labour at cheaper rates, for instance, would many women of small cultivator households
data gathering process and for re-examining
be reflected in women receiving a lower wage the burden of hand-pounding the rice, but
analytical concepts as they relate to women,
than men for identical agricultural tasks. have deprived the women of landless house-
so that the shortcomings highlighted can be
The question of the relative efficiency of holds, who performed the work for the large
overcome and a more adequate date base
different types of labour is in fact a com- cultivators, of a crucial source of livelihood
provided.
plex one, related, among other things, to the [see Collier, et al, 1974, for Java; Abdullah
While it is not always easy to suggest or and Zeidenstein, 1975, for Bangladesh; and
task being performed, the technique/imple-
to implement corrective measures, especially Harriss, 1977, for India].
ment being used, and the age and nutritional
for overcoming respondent and enumerator Seeking data on the time-allocation pat-
intake of the worker. If a systematic conver-
biases introduced by cultural perceptions, terns of rural women, on a national basis,
sion has to be done then it needs to be done
some move towards obtaining more accurate is perhaps not a practical proposition. How-
for both men and women. Some attempts
data can nevertheless be made. For instance, ever, detailed, region-specific, micro-level
at weighting, by one or more of these
a specific attempt can be made to enlist research focusing on these aspects could help
criteria, have been made [e g, Njoku 1971,
female enumerators and to seek out female to provide more appropriate definitions for
cited by Spencer, 1972, p 14, by agricultural
yesponden:s. While this cannot entirely over-wider-based data gathering. It would be
operations, gender and age; Gollas, 1970;
come the bias, in that women respondents'
Norman, 1973; and Waddell and Krinks, possible, for example, to specify more clearly
self-perceptions could still colour the which activities should be taken as economi-
1968, by gender and age]. Again, however,
answers given, it can help to reduce the ex- cally productive so that the women involved
the weighting is on a priori grounds. It is
tent of the bias. In some cases the accuracy in' such activities could be counted as
nevertheless of interest to note that Njoku,
of the data could be considerably improved 'workers'; this would also help to reduce
w,ho alone among those mentioned also
by this as, for instance, where women's and respondent biases. Insofar as the pattern of
looks at tasks performed, assigns and equal
men's interests within the household clearly activities itself differs between regions, more
or higher weight to female labour, relative
diverge.4 region-appropriate definitions could be
to male labour, in all operations other than
land clearing (where a conversion rate of 3/4 Further, insofar as the definitions adopted adopted. Also, data obtained through such
is used for women), this being an operation to identify the women who are employed or micro-studies could help provide correction
requiring greater physical strength. Similarly, unemployed tend to aggravate cultural factors for macro-data.
as noted in some other studies [for instance, biases, appropriate modifications in these In this context a greater use could be made
Agarwal, 1983; and Spencer, 1972, assume definitions and the way questions are framed of existing farm survey information. In
a one-to-one equivalence for male and would help in reducing the extent of the bias. India, for example, data gathered annually
female labour], in the particular operations This wouldhold true too where information by the agricultural universities of different
(e g, weeding, harvesting) in which women is collected about the characteristics of the states, under the 'Cost ' of Cultivation
women, such as whether they are heads of
are observed to be employed most common- Surveys',5 constitute a rich and relatively lit-
ly, there is no obvious a priori-reason to their households. tle expldred source of information on female
assume a lesser efficiency. Also, a sounder empirical base for asses- labour participation in crop production
Observations on comparative male-female sing the time contribution of rural women activities.6
efficiency are, in any case, rendered difficult, TABLE
when there is a sex-typing of tasks, so that
there is a predominance of women in cer- Type of Equipment. Picking Rate Per Hours Taken for Potato Yield
Labourer in Metres the Same Job kg/20 Metres
tain tasks and of men in others. In the earlier
Per Minute
mentioned study by Sanghvi, for instance, it
was not indicated how the conclusion about Men Women Men Women Men Women

the higher efficiency of women in cottoil Spring time cultivator 1.6 5.2 192 53 17.1 23.6
picking and their lower efficiency in the
Potato digger elevator 1.4 4.0 185 69 18.2 23.9
harvesting of other crops such as ground-
nut was arrived at, when in both groundnut Source: Kaul,. Kalhat and Shyam [1974].

A-161

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
 ECONOMIC AND POLITICAL WEEKLY Review of Agriculture December 1985

To take account of the se.asonal aspects would not work due to social and religious communication]. Clearly, in such cases, the
of unemployment, the periods of data barriers. When women beggars in the area gender of the respondent becomes a signifl-
gathering need to be more closely aligned were approached, they said: "Yes, we would cant factor.
with the lean and heavy periods of women's like t? do earthwork, but they do not recruit 5 These data are collected under the 'Com-
and men's agricultural activities. Some ex- women" [Chen and Ghuznavi, 1978: 13]. prehensive Scheme for Studying the Cost of
isting data sources such as the Indian NSS 2 A sample re-survey was carried out to pro- Cultivation of Principal Crops' usually by
could easily be adapted in this way, with
vide an adjustment factor to make the two the agricultural universities of the different
censuses comparable. However, strong reser- states, for the Directorate of Economics and
surveys being conducted in appropriate
vations have been expressed by some scholars Statistics, Ministry of Agriculture, New
periods of the year. To the extent that the
regarding the accuracy of the adjustment fac- Delhi, India.
lean and heavy periods of work diverge bet-
tor so obtained [e g, Sinha, 1975]. 6 For a recent, analysis based on this data
ween women and men, this might require
3 Bardhan [1978] for example, notes from thesource of the impact of high yielding variety
conducting additional surveys; or a separate
27th round data for the State of W Bengal, rice on the use of different types of female
NSS round devoted to assessing women's that of the women who by 'usual' status were labour in crop production, in three states of
unemployment, using female respondents classified as casual agricultural labourers, India, see Agarwal [1984].
and female enumerators as far as possible, only 2.5 per cent reported domestic work as
could be conducted. Initially perhaps this their principal current activity when the References
could be attempted on a pilot basis. The reference week fell in the relatively busy
question of the correct weighting of,women's
agricultural period, while as many as 14.0 per Abdullah, T A and S A Zeidenstein [1975]:
work-time and productivity, likewise needs cent reported this when the survey week fell 'Socio-Economic Implications of Introduc-
to be tackled from a sounder empirical base. in the slack period. This fluctuation is seen ing HYV Rice Production on Rural Women
To sum up, if we are genuinely concerned as reflective of periodic entries and involun- in Bangladesh', paper No, 34, presented at
that the data gathered provide an honest pic- tary withdrawals from the labour force. the International Seminar on Socio-Econo-
ture of the dimensions of unemployment 4 Consider the case of a Food Patrol Survey mic Implications of Introducing HYVs in
and poverty *that characterise the under- which was carried out in New Guinea in Bahgladesh, Bangladesh Academy of Rural
developed world, and if such data are to 1972, for assertaining how soon the potato Development, Ford Foundation, Dacca,
crop would be ready so that famine rations April.
serve as an appropriate base for policy for-
mulation and for initiating developmental
could be withdrawn. The crop was cultivated - [19761: 'Finding Ways to Learn about Rural
by the women; this gave them authority over Women: Experiences from a Pilot Project
schemes, ensuring their accuracy becomes
the family's food supply. Rations, however, in Bangladesh', The Ford Foundation,
crucial. In this context, the inadequacies of
were usually supplied to the men, which gave Dacca, Report No 44, November.
available data on various aspects of rural
them powers of patronage in the household Agarwal, Bina [1981]: 'Agricultural Modernisa-
women's work participation, require sober
and political influence in the village. The tion and Third World Women: Pointers
and immediate attention.
Patrol found that the information it obtained from the Literature ana an Empirical
The practical problem of successfully in- depended critically on whether it first met Analysis', Rural Employment Policy
corporating these data into policy would of the men or the women. Men claimed that the Research Programme Working Paper
course still loom large. The issues this raises crop was a failure; women took the Patrol No WEP 1O/WP 21, International Labour
are complex and a discussion on them is out- to the fields and proved that the crop was Office, Geneva.
side the purview of this paper. Nevertheless prospering and would soon be ready for [1983]: "Mechanisation in Indian
it must be pointed out that instituting work harvesting [Ingrid Palmer, personal Agriculture: An Analytical Study Based on
schemes which are flexible enough to fit in
with the time pattern and location of work
for which the women indicate their 'avail-
ability', and which are in keeping with an ex-
isting sexual and social division of labour,
may not always be practicable. At the same
time, the incorporation of these women in
state instituted income-generating schemes
which do not have this flexibility, may re-
quire a basic re-structuring of work tasks Akos Ostor
both within households and between house-
holds. Insofar as it may not be possible to CULTURE AND POWER
institute such changes without altering the
social and political context within which Legend, Ritual, Trade and Revolution in a Bengali Society
work is conducted and work schemes are
conceived and implemented, the question of In what is a unique book, Akos Ostor brings together a whole set of new data, nor-
rural women and work in fact raises more mally considered se-parately and in isolation, and provides a hitherto untried.style
fundamental issues than are apparent at first and method of analyzing them. The result is an interpretive study of cultural form
glance. and meaning in the social relations of everyday life in a Bengali town, which also
contributes a fresh and imaginative methodology to anthropological theory.
Notes Akos Ostor does not proceed from the geographical setting to religion, economics,
[I would like to express my gratitude to the and politics through the caste system. Rather than taking these analytic constructs
following people for their helpful comments on and domains for granted he designates them as the problematic of the study.
an earlier draft: Maureen Mackintosh, Ingrid
224pp/225x140mm/Rs 130 (hb)/1984
Palmer, Kate Young, Raghav Gaiha and Akmal
Hussain. ]

1 Researchers looking at one FEW programme

in Bangladesh, on noting the absence of ~SA~iEPUBLICATIONS INDIA PRIVATE LMTED
women workers, made enquiries. The local
project staff assured them that the women

A-163

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms
Review of Agriculture December 1985 ECONOMIC AND POLITICAL WEEKLY

the Punjab", Allied Publishers, India. ment", Agricultural Development Council,tation, University of Illinois.
- [1984]: 'Rural Women and the High Yielding Research and Training Network Reprint, Palmer, Ingrid [1977]: 'Rural Women and the
Variety Rice Technology', Economic and New York, September. Basic-Needs Approach to Development', In-
Political Weekly, Review of Agriculture, Consortium for International Development ternational Labour Review, Vol 15, No 1,
March 31, Vol 19, No 13. [1978]: Proceedings and Papers of the In- January-February.
Ahmed, Iftikhar [1978]- 'Technological Change' ternational Conference on Women and Sanghvi, Prafulla [1969]: "Surplus Manpower
and the Condition of- Rural. Women: A Food, Vol III, University of ArizonaTuscon, in Agriculture and Economic Development,
Preliminary Assessment', World Employ- Arizona, January 8-11. with Special Reference to India", Asia
ment Programme Research Working Paper, Dandekar, Kumudini and D B Unde [1964]: Publishing House, India.
WEP 2-22/WP 39, International Labour 'Households in West Bengal and Their
Sidhu, Surjit S [1974]: 'Economics of Technical
Office, October. Headship'. Artha Vijnana, Vol 6, No 1, Change in Wheat Production in the Indian
Arens, Jenneke and Jos van Beurden [1977]: March.
Punjab', American Journal of Agricultural
"Ihagrapur: Poor Peasants and Women in Deere, Carmen Diana [1977]: 'The Agricultural
Economics, Vol 56, No 2, May.
a Village in Bangladesh", Amsterdam. Division of Labour by Sex: Myths, Facts
Bardhan, Pranab [19781: 'Some Employment and Contradictions in the Northern Peru- Sinha. J N [19721: 'A Rational View of Census
and Unemployment Characteristics of vian Sierra, paper presented to the panel on Economic Data', Indian Journal of In-
Rural Women: An Analysis of NSS Data 'Women: The New Marginals in the Deve- dustrial Relations, Vol 8, No 2, October.
for West Bengal 1972-73', Economic and lopment Process', Joint National Meeting -[1975]: 'Viewpoints on Employment
Political Weekly, Review of Agriculture, of the Latin American Studies Association Statistics in India', paper presented at the
'Vol 13, No 2, March 25. and the African Studies Association, National Seminar on Social Statistics, New
Birdsall, Nancy and William P McGreevey Houston, Texas, November 2-5. Delhi, March.
Epstein, T Scarlett [1973]: "South India: Yester- Spencer, Dunstan, S C [1972]: "Micro-Level
f1978]: 'The Second Sex in the Third World:
Is Female Poverty a Developrient Issue', day, Today and Tomorrow"' Macmillan, Farm Management and Production Econo-
paper prepared for the International Center London. mics Research among Thaditional African
for Research, Washington DC, July 21. Gollas, M [1970]: "Surplus Labour and
Farmers: Lessons from Sierra "Leone",
Bishnoi, R N [1966]: 'Pattern of Employment Economic Development: The Guatemalan
Michigan State University, African Rural
and the Nature and Causes of Unemploy- Case". University of Wisconsin, Land: Emvloyment Paper No 3, September.
ment in Agriculture', Indian Journal of Tenure Center, Research Paper No 39,
January. Staudt, Kathleen A [1976]: 'Women Farmers
Agricultural Economics, Vol >1, No 1,
Hanger, Jane and Jon Moris [1973]: 'Women and Inequities in African Services', Rural
January-March.
and the Household Economy' in Robert Africana, No 29, Winter, African Studies
Boserup, Ester [1970]: "Women's Role in
Chambers and Jon Moris (Eds), "An Center, Michigan State University,
Economic Development", St Martin's Press,
Michigan.
Irrigated Rice Settlement in' Kenya"' Africa
New York.
- [1973]: 'Eng,loyment of Women in Develop- Studien No 83, Ifo-Institut Fur Wirtschaft-
UNECA [1972]: 'Women The Neglected
ing Countries', in "Population Growth and sforchung, Munchen, Afrika-Studien Stelle,
Human Resources for African Develop-
Economic Development in the Third Munchen, ment', Canadian Journal of African
World", edited by Leon Tabah, International Harriss, Barbara [1977]: 'Paddy-Milling: Pro- Studies, Vol 6, No 2.
Union for the Scientific Study of Popu- blems in Policy and Choice of Technology',
lation. in "Green Revolution? Technology and -[1975]: 'Employment of Women in the
Boulding, Elise, Shirley A Nuss, Dorothy Lee Change in Rice-Growing Areas of Tamil Sudan', paper contributed by the United
Carson, Michael A, Greenstein [1976]: Nadu and Sri Lanka" (ed) by B H Farmer, Nations Economic Commission for Africa,
"Handbook of International Data on Macmillan Press, London and Basingstoke. for the International Labour Organisation
Women", Sage Publications, Halsted Press. Kahlon, A S, S S Miglani and S K Mehta [1973]: Comprehensive Employment Mission,
Brausch, Georges [1964]: '-Change and Con- "Studies. in the Economics of Farm Sudan, January.
tinuity in the Gezira Region of the Sudan', Mariagement in Ferozepur District (Punjab):
Waddell, E W and P A Krinks [1968]: "The
International Social Science Journal, Report for the Year 1967-68", Directorate
Organisation of Production and Distribu-
Vol XVI, -No 3. of Economics and Statistics, Ministry of tion among the Orokaiva", New Guinea
Bukh, Jette [1979]: "The Village Women in Agriculture, New Delhi.
Research Bulletin No 24, New Guinea
Ghana", Center for Development Research, Kaul, R N, H S Kalhat and Shyam [1974]: Research Unit, Australian National Univer-
Scandinavian Institute of African Studies, 'Potato Digger Elecators: Performance
sity, September.
Uppsala. Studies'. mimeo, Department of Farm,
Buvinic, Mayra and Nadia H Youssef [1978]: Pover.and Machinery, Punjab Agricultural Wilde, John C de [1967]: "Experiences with
'Women-Headed Households: The Ignored University, Punjab. Agricultural Development in Tropial Africa:
Factor in Development Planning', Report Mbilinyi, Marjorie J [1972]: 'The State of Vol 1, The Synthesis", The John Hopkins
submitted to AID/WID, International Women in Tanzania', Canadian Journal of Press, Baltimore, Maryland.
Center for Research on Women, Washing- African Studies, Vol 6, No 2. Youssef, Nadia H [1974]: 'Muslim Women and
con DC, March. Malya, Meenakshi M [1979]: 'A Note on the Agricultural Production: Are They Under-
Chen, Marty and Ruby Ghuznavi [1978]: Methodological Problems of Using the
counted or Actually Dispensible?', mimeo,
"Women in Food-For-Work: The Bangla- Employment Data from the Farm Manage- Population Research Laboratory, Depart-
desh Experience", Oxfam. ment Studies', Indian Journal of ment of Sociology. University of Southern
Census of India [1971]: "Indian Census Agricultural Economics, Vol 34, No 2, California, Los Angeles.
Through a Hundred Years", Part I, Census April-June.
Centenary Monograph No 2 by D Natara- National Sample Survey Organisation [1981]:
jan, Office of the Registrar General, Sarvekshana, Vol V, Nos I and 2, July- AAN NEW SERIES
ASIAN ECONOMIC & SOCIAL REVIEW will be
Ministry of Home Affairs, New Delhi. October, Government of India. published from January 1986 as a monthly serial with an
Norman, D W [1973]: 'Methodology and Pro-
Census of India J1981]: "Provisional Popula- up-to-date appraisal of industry, banking, rural - urban
development trends through its cohumns in articles, notes
tion Totals-Workers and Non-Workers" blems of Farm Management Investigations: with comments from the editor, if so required. Publication
Paper 3 of 1981, Series 1, Office of the Experiences from Northern Nigeria', authority-resolution of the First International Scholars'
Conference on Asian cooperation held in 1975 and
Registrar General and Census Commis- Michigan State University, African Rurai published quarterly fro i 1976.

sioner, for India, Government of India. Employment Paper No 8, April. YEARLY SUBSCRIPTION RATE
India... Rs. 150.00 Foreign... Rs. 750.00
Collier, William L, Jusuf Colter, Sinarhadi, Njoku, Athanausius, 0 [1971]: 'Labour Utilisa- order from
INDIAN INSTITUTE OF ASIAN STUDIES
Robert d'A Shaw [1974]: "Choice of Tech- tion in Traditional Agriculture: The Case 23/354, Azadnagar, Jaiprakash Road, Andheri,--
nique in Rice Milling on Java: A Com- of Sierra Leone Rice Farms', Ph D disser- BOMBAY-400058. TELEPHONE: 573264

A-164

This content downloaded from 14.139.235.3 on Fri, 17 Jan 2020 12:03:31 UTC
All use subject to https://about.jstor.org/terms